Tesi: "Robotic system identification"

1

Dang, Kim Son Mechanical &amp Manufacturing Engineering Faculty of Engineering UNSW. "Design and control of autonomous crop tracking robotic weeder : GreenWeeder". Publisher:University of New South Wales. Mechanical & Manufacturing Engineering, 2009. http://handle.unsw.edu.au/1959.4/44418.

Testo completo

Abstract (sommario):

This thesis reports the design and control of the ??GreenWeeder??, a non-herbicidal autonomous weeding robot, in order to autonomously track crop rows for weeding through electrocution in the inter-row space. The four wheel mobile robot platform was designed and built with a motorised Ackerman steering system allowing the robot to steer up to 30 degree left and right. It was also equipped with an electronically geared rear wheel drive, a pair of stereo cameras, a SICK LMS-291 laser range finder to localize itself with respect to the crop rows, a GPS system for obtaining the robot position in the field and a long-range communication system for tele-supervision by operators. The first prototype of the robot electrocution system was also designed and constructed to ignite 22kV electrical arcs to destroy weeds. Its operation was tested in the research field of the University of Sydney and the results of this experiment were analysed to improve the efficiency of this first prototype. An improved prototype of the electrocution system was then constructed and attached to a cradle extending out at the back of the mobile robot platform. The testing of this improved prototype was conducted at Lansdowne farm, a research field of the University of Sydney. After the construction of the robot platform, the robot control was considered with the demands for robot localization with respect to crop rows, an autonomously tracking control system and a manual control mode in order to take the robot to transportation vehicles. Firstly, the robot localization was accomplished by utilizing SICK LMS-291 laser range finder sensor for the sensing and perception of the robot. Secondly, the robot control system was developed with a PID controller, a second order model of the robot system and a first order filter. The PID controller is in the standard form with the filtered derivative and the PI part being in automatic reset configuration. The second order model was identified using Matlab System Identification toolbox based on the robot kinematic analysis. The first order filter is utilized for filtering out the lateral deviations of the robot with respect to the crop rows received from the SICK laser sensor. A Simulink simulation model of the robot control system was also built in order to fine-tune PID and filter parameters. Thirdly, the manual control mode of the robot was produced. In this mode, a joystick can be attached to a notebook to wirelessly drive the robot around or it can be plugged into a USB port at the back of the robot to drive it without the notebook. After the robot control was implemented and simulated, some experiments were conducted with the robot autonomously tracking a strip of reflective tape mimicking a crop row stuck into the ground of a laboratory. Depending on distances from the row assigned to the controller, the robot tried to keep those distances away from the row. The results showed the lateral errors of the robot with respect to the row were approximately 4.5 cm which were sufficient for our current agricultural application.

Gli stili APA, Harvard, Vancouver, ISO e altri

2

Giantomassi, Andrea. "Modeling, estimation and identification of complex system dynamics: issues and solutions". Doctoral thesis, Università Politecnica delle Marche, 2012. http://hdl.handle.net/11566/242023.

Testo completo

Abstract (sommario):

La modellazione dei sistemi è di fondamentale importanza in tutte le discipline, sono utili per l’analisi, la previsione o la simulazione dei sistemi. Esistono due pratiche per definire modelli: modellazione e di identificazione. La modellazione è basata su leggi note. L’identificazione consiste nella selezione di un modello sulla base delle osservazioni effettuate sul sistema. In questo lavoro si è dato un contributo all’identificazione e stima di dinamiche complesse di sistemi. Con attenzione ai sistemi reali, sono proposte tre soluzioni. Il primo argomento riguarda un inceneritore per rifiuti solidi urbani, dove i modelli matematici sono troppo complessi per essere utilizzati. La soluzione data è in grado di stimare e predire, la produzione di vapore di un inceneritore RSU. L’algoritmo di apprendimento si basa su reti di funzioni a base radiale e combina la tecnica Minimal Resource Allocating Network con un filtro di Kalman esteso adattativo per aggiornare i parametri della rete. Il secondo problema riguarda la compensazione degli errori di controllo per un manipolatore industriale. Se un contro è ben progettato l’errore di controllo non può essere compensato. Tuttavia nel controllo Sliding Mode discreto, l’errore di controllo presenta dinamiche residue. Si propongono sue approcci per compensare l’incertezza, l’obiettivo è sviluppare un SMC discreto più robusto con due soluzioni, una basata sullo stimatore di incertezza del modello, e un predittore autosintonizzante. La diagnosi guasti ha ricevuto un crescente interesse degli ultimi anni. L’ultimo argomento riguarda una procedura di rilevamento guasti e isolamento per la rilevazione e l’analisi di difetti di motori elettrici a fine linea di un impianto di produzione di cappe. L’obiettivo consiste nel rilevare e identificare i motori difettosi per l’analisi di qualità. Un approccio diagnostico basato sull’analisi dei segnali è preferibile per le caratteristiche dei segnali acquisiti e per la soluzione di implementazione.
Models of real systems are of fundamental importance in all disciplines, and they are useful for system analysis, prediction or simulation of a real system. Two practices exist to define models: modeling by physical laws and by identification. Physical modeling is based on known laws. Identification consists in the selection of a model in a specified class on the basis of observations performed on the system to be described. A contribution to complex system dynamics identification and estimation is given. With particular attention to real systems, three solutions are discussed. The first issue deals with a Municipal Solid Waste incinerator, where first principles mathematical models are too complex to be implemented. The procedure proposed is able to estimate and predict, the steam production of a MSW incinerator. The learning algorithm is based on radial basis function networks and combines the Minimal Resource Allocating Network technique with an adaptive extended Kalman filter to update the network parameters. The second issue regard the control error compensation for an industrial manipulator. If a controller is well designed the control error cannot be compensated. However in the discrete Sliding Mode Controller, control errors carry information about residual dynamics. Two approaches are proposed for uncertainties compensations, the objective is to develop a more robust and accurate discrete SMC using two solutions, a model based uncertainty estimator, and an auto-tuning predictor. Fault Detection and Diagnosis has received an increasing interest in years. The last issue regard a Fault Detection and Isolation procedure that is applied for the defects detection and analysis of electrical motors at the end of the production line in a hoods production plant. The objective consists of detect and identify defective motors for the quality analysis. A signal based FDI approach is preferred for the characteristics of acquired signals and for the implementation solution.

Gli stili APA, Harvard, Vancouver, ISO e altri

3

Mahé, Antoine. "Identification de systèmes par réseaux de neurones pour la commande prédictive". Electronic Thesis or Diss., CentraleSupélec, 2020. http://www.theses.fr/2020CSUP0010.

Testo completo

Abstract (sommario):

Le développement de la robotique mobile permet la réalisation de tâches de plus enplus variées de façon automatisée. Le projet GRoNe dans lequel s’inscrit cette thèse, a pour objectif de développer les connaissances et les expérimentations sur ce sujet. Dans ce cadre, l’automatisation est un élément clé.Le développement d’algorithmes de contrôle efficaces est une étape importante en ce sens. Les contrôleurs à commandes prédictives ont démontré de nombreux avantages en robotique mobile.La réalisation de ces algorithmes requiert la conception de modèles précis des systèmes afin de prédire leur évolution. Le problème de la modélisation en robotique est traité par l’identification de système. L’apprentissage automatique est fréquemment utilisé dans ce contexte. Afin de modéliser des systèmes robotiques des corpus de données des robots en fonctionement ontété collectés en simulation et sur le terrain. Différentes architectures de réseaux de neurones ont été comparées. Cependant les échantillons collectés ne correspondent pas toujours à la situation dans laquelle les modèles doivent être utilisés. Ce problème est traité en priorisant certains éléments. Pour cela, deux méthodes sont comparées. La réalisation de modèles n’est pas une fin en soit, il s’agit d’une composante de l’algorithme de contrôle. Il importe d’étudier les différents modèles au sein du contrôleur. L’application de ce contrôleur à un drone et un bateau, tant en simulation qu’en cas réel, permet d’étudier les avantages de son utilisation. Finalement un modèle priorisé est utilisé au sein d’un algorithme à commande prédictive sur le bateau réel afin de réaliser le suivi de la berge d'un lac artificiel
Developing mobile robotic allow to address ever more complex task autonomously.This thesis is part of the GRoNe project which aim at improving knowledge and experimentation on this topic. In this context automation is a key element. Developing efficient control algorithm is a step in that direction. Model predictive control has shown good result and interesting advantages in mobile robotic. Implementing this algorithm require precise system modelling in order to predict their evolution. In robotic modelling is usually solved by system identification. In this context machine learning is often a powerful tool. In order to model robotic system, data collection of their behaviour both in simulation and on the real platform have been collected. Several neural network architecture have been compared. Collected sample may not correspond to the condition of target task making part of the training irrelevant. A solution to that problem is to use prioritization during the training. Two prioritization scheme are compared. Modelling is only a step toward control. Thus it is important to test the obtained model as part of the whole control algorithm. The application of this controller to a drone and a boat, in simulation as well as on the real platform, allow to study its advantages. In the end a model train with prioritization is used in a model predictive controller on the real boat to perform shore following in an artificial lac

Gli stili APA, Harvard, Vancouver, ISO e altri

4

Tuna, Eser Erdem. "PERCEPTION AND CONTROL OF AN MRI-GUIDED ROBOTIC CATHETER IN DEFORMABLE ENVIRONMENTS". Case Western Reserve University School of Graduate Studies / OhioLINK, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=case1619795928790909.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

5

Åkesson, Ulrik. "Design of a multi-camera system for object identification, localisation, and visual servoing". Thesis, Mälardalens högskola, Akademin för innovation, design och teknik, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:mdh:diva-44082.

Testo completo

Abstract (sommario):

In this thesis, the development of a stereo camera system for an intelligent tool is presented. The task of the system is to identify and localise objects so that the tool can guide a robot. Different approaches to object detection have been implemented and evaluated and the systems ability to localise objects has been tested. The results show that the system can achieve a localisation accuracy below 5 mm.

Gli stili APA, Harvard, Vancouver, ISO e altri

6

Tout, Bilal. "Identification of human-robot systems in physical interaction : application to muscle activity detection". Electronic Thesis or Diss., Valenciennes, Université Polytechnique Hauts-de-France, 2024. https://ged.uphf.fr/nuxeo/site/esupversions/36d9eab3-c170-4e40-abb6-e6b4e27aeee2.

Testo completo

Abstract (sommario):

Ces dernières années, l’interaction physique humain-robot est devenue un sujet de recherche important, par exemple pour des applications de rééducation. Cette thèse vise à améliorer ces interactions, dans le cadre du développement de contrôleurs basés modèles, par des approches d’identification paramétrique des modèles des systèmes en interaction. Le but est de développer des méthodes d’identification en tenant compte de la variabilité et de la complexité du corps humain, et en n'utilisant que les capteurs du système robotique pour éviter l'ajout de capteurs externes. Les différentes approches présentées dans cette thèse sont testées expérimentalement avec un système robotique à un degré de liberté (1-DDL) permettant d'interagir avec la main d'une personne.Après un 1er chapitre présentant l'état de l'art, le 2e chapitre aborde les méthodes d'identification développées en robotique ainsi que la problématique du filtrage, analysée en simulation et expérimentalement. La question du réglage du filtre passe-bas est adressée, et en particulier le choix de la fréquence de coupure qui reste délicate pour un système non-linéaire. Pour surmonter ces difficultés, une technique de filtrage utilisant un filtre de Kalman étendu (EKF) est développée à partir du modèle dynamique du robot. La formulation EKF proposée permet un réglage en fonction des propriétés connues du capteur et de la confiance dans l’estimation initiale des paramètres. Cette méthode est comparée en simulation puis expérimentalement avec différentes méthodes existantes en analysant la sensibilité à l’initialisation et au réglage du filtre. Les résultats montrent que la méthode proposée est prometteuse si l’EKF est correctement réglé.Le 3e chapitre porte sur l'identification en continu des paramètres du modèle dynamique d'un système passif en interaction avec un système robotique, en combinant des méthodes d’identification de la charge utile avec des algorithmes en ligne, sans capteurs externes. Ces méthodes sont validées en simulation et expérimentalement en utilisant le système à 1-DDL dont la poignée est attachée à des bandes élastiques pour imiter une articulation humaine passive. L’analyse de l’effet de l’ajustement des méthodes en ligne met en évidence qu’un compromis est nécessaire entre la vitesse de convergence et la précision des estimations des paramètres. Enfin, la comparaison des méthodes d’identification de la charge utile montre que les méthodes identifiant séparément les paramètres du système robotique et de l’humain passif donnent une meilleure précision et une plus faible complexité de calcul.Le 4e chapitre porte sur l'identification durant l'interaction humain-système robotique. Un modèle à raideur quadratique est proposé afin de mieux représenter le comportement de l’articulation humaine passive qu’un modèle linéaire. Par la suite, ce modèle est utilisé avec une méthode d’identification itérative basée sur le rejet d’outliers, pour détecter l’activité musculaire de l’humain sans capteurs externes. Cette méthode est comparée expérimentalement à une méthode non-itérative utilisant des signaux d’électromyographie (EMG), en adaptant le système à 1-DDL pour interagir avec le poignet et permettre d’évaluer l’activité des muscles fléchisseurs et extenseurs de deux sujets. La méthode itérative proposée sans signaux EMG donne des résultats proches de ceux obtenus avec la méthode utilisant les signaux EMG lorsqu’un modèle représentant bien le comportement de l’articulation humaine passive est choisi. Les résultats de détection de l’activité musculaire obtenus avec ces deux méthodes montrent un niveau de similarité satisfaisant avec ceux obtenus directement à partir des signaux EMG
Over the last years, physical human-robot interaction has become an important research subject, for example for rehabilitation applications. This PhD aims at improving these interactions, as part of model-based controllers development, using parametric identification approaches to identify models of the systems in interaction. The goal is to develop identification methods taking into account the variability and complexity of the human body, and only using the sensor of the robotic system to avoid adding external sensors. The different approaches presented in this thesis are tested experimentally on a one degree of freedom (1-DOF) system allowing the interaction with a person’s hand.After a 1st chapter presenting the state-of-the-art, the 2nd chapter tackles the identification methods developed in robotics as well as the issue of data filtering, analyzed both in simulation and experimentally. The question of the low-pass filter tuning is addressed, and in particular the choice of the cut-off frequency which remains delicate for a nonlinear system. To overcome these difficulties, a filtering technique using an extended Kalman filter (EKF) is developed from the robot dynamic model. The proposed EKF formulation allows a filter tuning depending on the known properties of the sensor and on the confidence on the initial parameters estimations. This method is compared in simulation and experimentally to different existing methods by analyzing its sensitivity to initialization and filter tuning. Results show that the proposed method is promising if the EKF is correctly tuned.The 3rd chapter concerns the continuous identification of the parameters of the model of a passive system interacting with a robotic system, by combining payload identification methods with online identification algorithms, without external sensors. These methods are validated in simulation and experimentally with the 1-DOF system whose handle is attached to elastic rubber bands to emulate a passive human joint. The analysis of the effects of the online methods tuning highlights a necessary trade-off between the convergence speed and the accuracy of the parameters estimates. Finally, the comparison of the payload identification methods shows that methods identifying separately the robotic system and the passive human parameters give better accuracy and a lower computation complexity.The 4th chapter deals with the identification during the human-robot interaction. A quadratic stiffness model is proposed to better fit the passive human joint behavior than a linear stiffness model. Then, this model is used with an iterative identification method based on outlier rejection technique, to detect the human user muscle activity without external sensors. This method is compared experimentally to a non-iterative method that uses electromyography (EMG), by adapting the 1-DOF system to interact with the wrist and to allow the detection of the flexor and extensor muscle activity of two human users. The proposed iterative identification method not using EMG signals achieves results close to those obtained with the non-iterative method using EMG signals when a model that correctly represents the passive human joint behavior is selected. The muscle activity detection results obtained with both methods show a satisfactory level of similarity compared to those obtained directly from EMG signals

Gli stili APA, Harvard, Vancouver, ISO e altri

7

Wang, Zeya. "Robotisation de la fabrication additive par procédé arc-fil : Identification et amélioration de la commande". Electronic Thesis or Diss., Université de Lorraine, 2022. http://www.theses.fr/2022LORR0068.

Testo completo

Abstract (sommario):

La fabrication additive de pièces métalliques a fait l'objet d'un vif intérêt ces dernières années comme une solution technologique importante pour la réalisation de pièces complexes. Parmi les différents procédés de la fabrication additive métallique, la fabrication additive arc-fil (FAAF) utilisant le soudage CMT (Cold metal transfer) est retenue pour notre étude grâce à son taux de dépôt important, faible coût des équipements et peu de perte de matière par projections lors de la fabrication. Dans la littérature, il est constaté que l'un des problèmes les plus importants qui empêchent l'application industrielle du procédé FAAF est la mauvaise précision géométrique des pièces fabriquées à cause de l'instabilité du procédé et du manque de contrôle-commande fiable pour traiter les irrégularités pendant le dépôt. L'objectif de ce travail est d'améliorer la stabilité et la performance géométrique du procédé. Dans ce travail, un système expérimental est mis en œuvre pour robotiser le procédé et contrôler la géométrie des pièces déposées. Le procédé est modélisé par les réseaux de neurones artificiels et un système contrôle-commande est développé permettant de commander la géométrie du dépôt et de réduire les erreurs de fabrication. De plus, une stratégie d'amélioration est appliquée afin de réduire les instabilités géométriques aux deux extrémités du cordon ; une méthode de contrôle in situ est également développée pour détecter les défauts internes des pièces déposées
Additive manufacturing of metallic parts has gained significant popularity in recent years as an important technological solution for the production of complex parts. Among the different processes of metal additive manufacturing, the wire-arc additive manufacturing (WAAM) using CMT (Cold metal transfer) welding is taken for our study because of its high deposition rate, low cost of equipment and little loss of material (low spatter) during manufacturing. In the literature review, it can be noted that one of the most important problems that prevent the industrial application of the WAAM is the poor geometric accuracy of the manufactured parts due to the instability of the process and the lack of reliable control system to deal with irregularities during deposition. The focus of this work is to improve the stability and geometric performance of the process. In this work, an experimental system is implemented to robotize the process and to monitor the geometry of the deposited parts. The process is modeled by artificial neural networks and a control system is developed to regulate the geometry of the deposit and to reduce manufacturing errors. Furthermore, an improvement strategy is applied in order to reduce the geometric instabilities at the ends of the bead; an in-situ monitoring method is also developed to detect the internal defects of deposited parts

Gli stili APA, Harvard, Vancouver, ISO e altri

8

Cetin, Murat. "Performance identification and multi-criteria redundancy resolution for robotic systems /". Digital version accessible at:, 1999. http://wwwlib.umi.com/cr/utexas/main.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

9

Leijonhufvud, Peder, e Emil Bråkenhielm. "Image Processing for Improved Bacteria Classification". Thesis, Linköpings universitet, Programvara och system, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-167416.

Testo completo

Abstract (sommario):

Mastitis is a common disease among cows in dairy farms. Diagnosis of the infection is today done manually, by analyzing bacteria growth on agar plates. However, classifiers are being developed for automated diagnostics using images of agar plates. Input images need to be of reasonable quality and consistent in terms of scale, positioning, perspective, and rotation for accurate classification. Therefore, this thesis investigates if a combination of image processing techniques can be used to match each input image to a pre-defined reference model. A method was proposed to identify important key points needed to register the input image to the reference model. The key points were defined by identifying the agar plate, its compartments, and its rotation within the image. The results showed that image registration with the correct key points was sufficient enough to match images of agar plates to a reference model despite any varieties in scale, position, perspective, or rotation. However, the accuracy depended on the identification of the salient features of the agar plate. Ultimately, the work proposes an approach using image registration to transform images of agar plates based on a pre-defined reference model, rather than a reference image.

Gli stili APA, Harvard, Vancouver, ISO e altri

10

Larsson, Joel, e Rasmus Hedberg. "Development of machine learning models for object identification of parasite eggs using microscopy". Thesis, Uppsala universitet, Signaler och system, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-414386.

Testo completo

Abstract (sommario):

Over one billion people in developing countries are afflicted by parasitic infections caused by soil-transmitted helminths. These infections are treatable with cheap and safe medicine that is widely available. However, diagnosis of these infections has proven to be a bottleneck by the fact that it is time-consuming, requires expensive equipment and trained personnel to be consistent and accurate. This study aimed to investigate the viability and performance of five machine learning models and a 'modular neural network' approach to localize and classify the following parasite eggs in microscopic images: Ascaris lumbricoides, Trichuris trichuria, Hookworm and Schistosoma mansoni. These models were implemented and evaluated on the Nvidia Jetson AGX Xavier to establish that they fulfilled the specifications of 95\% specificity and sensitivity, but also a speed requirement of 40000 images per 24 hours. The results show that R-FCN ResNet101 was the best model produced in this study, which performed the best on average. However, it did not fulfill the specifications entirely but is still considered a success due to being an improvement to the current implementation at Etteplan. Evaluation of the modular neural network approach would require further investigation to verify the performance of the system, but the results indicate it could be a possible improvement to the off-the-shelf machine learning models. To conclude, the study showed that the data and data infrastructure provided by Etteplan has proven to be a very powerful tool in training machine learning models to classify and localize parasite eggs in stool samples. However, expansion of the data to reduce the imbalance between the representations of the classes but also include more patient information could improve the training and evaluation process of the models.

Gli stili APA, Harvard, Vancouver, ISO e altri

11

Johansson, Victor. "3D Position Estimation of a Person of Interest in Multiple Video Sequences : Person of Interest Recognition". Thesis, Linköpings universitet, Datorseende, 2013. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-97970.

Testo completo

Abstract (sommario):

Because of the increase in the number of security cameras, there is more video footage available than a human could efficiently process. In combination with the fact that computers are getting more efficient, it is getting more and more interesting to solve the problem of detecting and recognizing people automatically. Therefore a method is proposed for estimating a 3D-path of a person of interest in multiple, non overlapping, monocular cameras. This project is a collaboration between two master theses. This thesis will focus on recognizing a person of interest from several possible candidates, as well as estimating the 3D-position of a person and providing a graphical user interface for the system. The recognition of the person of interest includes keeping track of said person frame by frame, and identifying said person in video sequences where the person of interest has not been seen before. The final product is able to both detect and recognize people in video, as well as estimating their 3D-position relative to the camera. The product is modular and any part can be improved or changed completely, without changing the rest of the product. This results in a highly versatile product which can be tailored for any given situation.

Gli stili APA, Harvard, Vancouver, ISO e altri

12

Elhami, Mohammad Reza. "Modelling, identification and compensation of friction in robot control systems". Thesis, University of Liverpool, 1996. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.243266.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

13

Moberg, Stig. "On Modeling and Control of Flexible Manipulators". Licentiate thesis, Linköping University, Linköping University, Automatic Control, 2007. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-10463.

Testo completo

Abstract (sommario):

Industrial robot manipulators are general-purpose machines used for industrial automation in order to increase productivity, flexibility, and quality. Other reasons for using industrial robots are cost saving, and elimination of heavy and health-hazardous work. Robot motion control is a key competence for robot manufacturers, and the current development is focused on increasing the robot performance, reducing the robot cost, improving safety, and introducing new functionalities. Therefore, there is a need to continuously improve the models and control methods in order to fulfil all conflicting requirements, such as increased performance for a robot with lower weight, and thus lower mechanical stiffness and more complicated vibration modes. One reason for this development of the robot mechanical structure is of course cost-reduction, but other benefits are lower power consumption, improved dexterity, safety issues, and low environmental impact.

This thesis deals with three different aspects of modeling and control of flexible, i.e., elastic, manipulators. For an accurate description of a modern industrial manipulator, the traditional flexible joint model, described in literature, is not sufficient. An improved model where the elasticity is described by a number of localized multidimensional spring-damper pairs is therefore proposed. This model is called the extended flexible joint model. This work describes identification, feedforward control, and feedback control, using this model.

The proposed identification method is a frequency-domain non-linear gray-box method, which is evaluated by the identification of a modern six-axes robot manipulator. The identified model gives a good description of the global behavior of this robot.

The inverse dynamics control problem is discussed, and a solution methodology is proposed. This methodology is based on a differential algebraic equation (DAE) formulation of the problem. Feedforward control of a two-axes manipulator is then studied using this DAE approach.

Finally, a benchmark problem for robust feedback control of a single-axis extended flexible joint model is presented and some proposed solutions are analyzed.

Gli stili APA, Harvard, Vancouver, ISO e altri

14

Deyle, Travis. "Ultra high frequency (UHF) radio-frequency identification (RFID) for robot perception and mobile manipulation". Diss., Georgia Institute of Technology, 2011. http://hdl.handle.net/1853/42903.

Testo completo

Abstract (sommario):

Personal robots with autonomy, mobility, and manipulation capabilities have the potential to dramatically improve quality of life for various user populations, such as older adults and individuals with motor impairments. Unfortunately, unstructured environments present many challenges that hinder robot deployment in ordinary homes. This thesis seeks to address some of these challenges through a new robotic sensing modality that leverages a small amount of environmental augmentation in the form of Ultra High Frequency (UHF) Radio-Frequency Identification (RFID) tags. Previous research has demonstrated the utility of infrastructure tags (affixed to walls) for robot localization; in this thesis, we specifically focus on tagging objects. Owing to their low-cost and passive (battery-free) operation, users can apply UHF RFID tags to hundreds of objects throughout their homes. The tags provide two valuable properties for robots: a unique identifier and receive signal strength indicator (RSSI, the strength of a tag's response). This thesis explores robot behaviors and radio frequency perception techniques using robot-mounted UHF RFID readers that enable a robot to efficiently discover, locate, and interact with UHF RFID tags applied to objects and people of interest. The behaviors and algorithms explicitly rely on the robot's mobility and manipulation capabilities to provide multiple opportunistic views of the complex electromagnetic landscape inside a home environment. The electromagnetic properties of RFID tags change when applied to common household objects. Objects can have varied material properties, can be placed in diverse orientations, and be relocated to completely new environments. We present a new class of optimization-based techniques for RFID sensing that are robust to the variation in tag performance caused by these complexities. We discuss a hybrid global-local search algorithm where a robot employing long-range directional antennas searches for tagged objects by maximizing expected RSSI measurements; that is, the robot attempts to position itself (1) near a desired tagged object and (2) oriented towards it. The robot first performs a sparse, global RFID search to locate a pose in the neighborhood of the tagged object, followed by a series of local search behaviors (bearing estimation and RFID servoing) to refine the robot's state within the local basin of attraction. We report on RFID search experiments performed in Georgia Tech's Aware Home (a real home). Our optimization-based approach yields superior performance compared to state of the art tag localization algorithms, does not require RF sensor models, is easy to implement, and generalizes to other short-range RFID sensor systems embedded in a robot's end effector. We demonstrate proof of concept applications, such as medication delivery and multi-sensor fusion, using these techniques. Through our experimental results, we show that UHF RFID is a complementary sensing modality that can assist robots in unstructured human environments.

Gli stili APA, Harvard, Vancouver, ISO e altri

15

Turesson, Eric. "Multi-camera Computer Vision for Object Tracking: A comparative study". Thesis, Blekinge Tekniska Högskola, Institutionen för datavetenskap, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-21810.

Testo completo

Abstract (sommario):

Background: Video surveillance is a growing area where it can help with deterring crime, support investigation or to help gather statistics. These are just some areas where video surveillance can aid society. However, there is an improvement that could increase the efficiency of video surveillance by introducing tracking. More specifically, tracking between cameras in a network. Automating this process could reduce the need for humans to monitor and review since the tracking can track and inform the relevant people on its own. This has a wide array of usability areas, such as forensic investigation, crime alerting, or tracking down people who have disappeared. Objectives: What we want to investigate is the common setup of real-time multi-target multi-camera tracking (MTMCT) systems. Next up, we want to investigate how the components in an MTMCT system affect each other and the complete system. Lastly, we want to see how image enhancement can affect the MTMCT. Methods: To achieve our objectives, we have conducted a systematic literature review to gather information. Using the information, we implemented an MTMCT system where we evaluated the components to see how they interact in the complete system. Lastly, we implemented two image enhancement techniques to see how they affect the MTMCT. Results: As we have discovered, most often, MTMCT is constructed using a detection for discovering object, tracking to keep track of the objects in a single camera and a re-identification method to ensure that objects across cameras have the same ID. The different components have quite a considerable effect on each other where they can sabotage and improve each other. An example could be that the quality of the bounding boxes affect the data which re-identification can extract. We discovered that the image enhancement we used did not introduce any significant improvement. Conclusions: The most common structure for MTMCT are detection, tracking and re-identification. From our finding, we can see that all the component affect each other, but re-identification is the one that is mostly affected by the other components and the image enhancement. The two tested image enhancement techniques could not introduce enough improvement, but other image enhancement could be used to make the MTMCT perform better. The MTMCT system we constructed did not manage to reach real-time.

Gli stili APA, Harvard, Vancouver, ISO e altri

16

Gunnar, Johan. "Dynamical Analysis and System Identification of the Gantry-Tau Parallel Manipulator". Thesis, Linköping University, Department of Electrical Engineering, 2005. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-5322.

Testo completo

Abstract (sommario):

This report presents work done in the field of linear and nonlinear system identification on robots. The subject of study has been a new parallel manipulator called Gantry-Tau. The work shall be seen as one of the first steps in the dynamical analysis of the robot. All practical work presented in the report was conducted on a prototype situated at University of Queensland.

The actuators have been analysed and modelled with the aim to gain knowledge of weaknesses and dynamical behaviour. The analysis resulted in a study of nonlinear grey-box identification of hysteresis in the drive train of the actuators. A very compact nonlinear hysteresis model was used together with a three-step identification procedure. The results show that a model of the nonlinear system can be successfully identified from measurement data.

Finally a method for estimation of parameters in the model for the inverse dynamics of the leg structure has been investigated. It turns out that the investigated method is not able to give accurate estimates. This is thought to be a result of unmodelled behaviour in the system and noisy data.

Gli stili APA, Harvard, Vancouver, ISO e altri

17

Hemakumara, Madu Prasad. "UAV Parameter Estimation with Gaussian Process Approximations". Thesis, The University of Sydney, 2013. http://hdl.handle.net/2123/9414.

Testo completo

Abstract (sommario):

Unmanned Aerial Vehicles (UAVs) provide an alternative to manned aircraft for risk associated missions and applications where sizing constraints require miniaturized flying platforms. UAVs are currently utilised in an array of applications ranging from civilian research to military battlegrounds. A part of the development process for UAVs includes constructing a flight model. This model can be used for modern flight controller design and to develop high fidelity flight simulators. Furthermore, it also has a role in analysing stability, control and handling qualities of the platform. Developing such a model involves estimating stability and control parameters from flight data. These map the platform's control inputs to its dynamic response. The modeling process is labor intensive and requires coarse approximations. Similarly, models constructed through flight tests are only applicable to a narrow flight envelope and classical system identification approaches require prior knowledge of the model structure, which, in some instances may only be partially known. This thesis attempts to find a solution to these problems by introducing a new system identification method based on dependent Gaussian processes. The new method would allow for high fidelity non-linear flight dynamic models to be constructed through experimental data. The work is divided into two main components. The first part entails the development of an algorithm that captures cross coupling between input parameters, and learns the system stability and control derivatives. The algorithm also captures any dependencies embodied in the outputs. The second part focuses on reducing the heavy computational cost, which is a deterrent to learning the model from large test flight data sets. In addition, it explores the capabilities of the model to capture any non-stationary behavior in the aerodynamic coefficients. A modeling technique was developed that uses an additive sparse model to combine global and local Gaussian processes to learn a multi-output system. Having a combined approximation makes the model suitable for all regions of the flight envelope. In an attempt to capture the global properties, a new sampling method is introduced to gather information about the output correlations. Local properties were captured using a non-stationary covariance function with KD-trees for neighbourhood selection. This makes the model scalable to learn from high dimensional large-scale data sets. The thesis provides both theoretical underpinnings and practical applications of this approach. The theory was tested in simulation on a highly coupled oblique wing aircraft and was demonstrated on a delta-wing UAV platform using real flight data. The results were compared against an alternative parametric model and demonstrated robustness, improved identification of coupling between flight modes, sound ability to provide uncertainty estimates, and potential to be applied to a broader flight envelope.

Gli stili APA, Harvard, Vancouver, ISO e altri

18

Brunot, Mathieu. "Identification of rigid industrial robots - A system identification perspective". Phd thesis, Toulouse, INPT, 2017. http://oatao.univ-toulouse.fr/20776/1/BRUNOT_Mathieu_20776.pdf.

Testo completo

Abstract (sommario):

In modern manufacturing, industrial robots are essential components that allow saving cost, increase quality and productivity for instance. To achieve such goals, high accuracy and speed are simultaneously required. The design of control laws compliant with such requirements demands high-fidelity mathematical models of those robots. For this purpose, dynamic models are built from experimental data. The main objective of this thesis is thus to provide robotic engineers with automatic tools for identifying dynamic models of industrial robot arms. To achieve this aim, a comparative analysis of the existing methods dealing with robot identification is made. That allows discerning the advantages and the limitations of each method. From those observations, contributions are presented on three axes. First, the study focuses on the estimation of the joint velocities and accelerations from the measured position, which is required for the model construction. The usual method is based on a home-made prefiltering process that needs a reliable knowledge of the system’s bandwidths, whereas the system is still unknown. To overcome this dilemma, we propose a method able to estimate the joint derivatives automatically, without any setting from the user. The second axis is dedicated to the identification of the controller. For the vast majority of the method its knowledge is indeed required. Unfortunately, for copyright reasons, that is not always available to the user. To deal with this issue, two methods are suggested. Their basic philosophy is to identify the control law in a first step before identifying the dynamic model of the robot in a second one. The first method consists in identifying the control law in a parametric way, whereas the second one relies on a non-parametric identification. Finally, the third axis deals with the home-made setting of the decimate filter. The identification of the noise filter is introduced similarly to methods developed in the system identification community. This allows estimating automatically the dynamic parameters with low covariance and it brings some information about the noise circulation through the closed-loop system. All the proposed methodologies are validated on an industrial robot with 6 degrees of freedom. Perspectives are outlined for future developments on robotic systems identification and other complex problems.

Gli stili APA, Harvard, Vancouver, ISO e altri

19

Craig, David. "Modeling and Control of a Magnetically Levitated Microrobotic System". Thesis, University of Waterloo, 2006. http://hdl.handle.net/10012/2844.

Testo completo

Abstract (sommario):

Magnetically levitated microrobotic systems have shown a great deal of promise for micromanipulation tasks. A new large-gap magnetic suspension system has recently been developed at the University of Waterloo in order to develop microrobotic systems for various applications. In order to achieve motion with the system, a model is needed in order to facilitate the design of various aspects of the system, such as the microrobot and the controller. In order to derive equations of motion for the system attempts were made to characterize the force produced by the magnetic drive unit in terms of a simple analytical equation. The force produced by the magnetic drive unit was estimated with the aid of a finite element model. The derived equations were able to predict the general trend of the force curves, and with sufficient parameter tweaking the error between the force estimated by the finite element model and the force estimated by the analytical equation could be minimized. System models describing the motion of the system in the horizontal and vertical directions are identified and compared to the actual system response. The vertical position response is identified through a least squares parameter estimate of the closed loop response combined with a partial reconstruction of the root locus diagram, with the model structure based on the known dynamics of a simplified form of magnetic levitation. This model was able to provide a reasonable prediction of the system response for a variety of PID controllers under a variety of input conditions. The horizontal models are identified using a least-squares parameter estimate of the open loop characteristics of the system. The horizontal models are able to provide a reasonable prediction of the system response under PD and PID control. Full spatial motion of a microrobot prototype is demonstrated over a working range of 20x22x30 mm³, with PID controller parameters and reference trajectories adjusted to minimize disturbances. The RMS error at steady state is on the order of 0. 020 mm for vertical positioning and 0. 008 mm for horizontal positioning. A linear quadratic regulator implemented for vertical position control was able to reduce the vertical position RMS error to 0. 014 mm.

Gli stili APA, Harvard, Vancouver, ISO e altri

20

Meira, Anrafel Silva. "Identificação não linear de um manipulador eletromecânico de três graus de liberdade". Universidade Federal da Paraíba, 2014. http://tede.biblioteca.ufpb.br:8080/handle/tede/7571.

Testo completo

Abstract (sommario):

Submitted by Maria Suzana Diniz (msuzanad@hotmail.com) on 2015-11-10T13:06:52Z No. of bitstreams: 1 arquivototal.pdf: 2081019 bytes, checksum: 810ebb6d466319b898b20b865caa5d4f (MD5)
Made available in DSpace on 2015-11-10T13:06:52Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2081019 bytes, checksum: 810ebb6d466319b898b20b865caa5d4f (MD5) Previous issue date: 2014-12-12
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES
This paper presents multivariable nonlinear mathematical models with estimable parameters in online identification for an electromechanical manipulator, thus enabling practical applications of adaptive control techniques. The manipulator comprises three rotary joints and three links appointed link 1, 2 and 3. The total displacement of the link 1 is 180° and of the link 2 is 110 °, with each of these links being driven by a direct current motor, while the link 3 has its motion controlled by a mechanical system that always keeps the horizontal position. The identification process started of the nonlinear autoregressive model with exogenous variables (NARX) for multiple inputs and multiple outputs, using the error reduction rate (ERR) method, coupled nonlinear, decoupled nonlinear and decoupled linear models were determined to the link 1 and link 2 of the manipulator. The Recursive Least Squares (RLS) estimator is used to estimate the parameters of the representative model of the robot manipulator links 1 and 2, thus verifying the efficiency of the models obtained in the online identification.
Este trabalho apresenta modelos matemáticos não lineares, multivariáveis com parâmetros estimáveis em identificação online para um manipulador eletromecânico, permitindo assim aplicações práticas de técnicas de controle adaptativo. O manipulador é composto por três juntas rotativas e três elos nomeados de elo 1, 2 e 3. O deslocamento total do elo 1 é de 180° e do elo 2 é de 110°, sendo cada um desses elos comandado por um motor de corrente contínua, enquanto o elo 3 tem o seu movimento comandado por um sistema mecânico que sempre o mantém na posição horizontal. O processo de identificação partiu do modelo auto-regressivo com entradas exógenas NARX (Nonlinear Autoregressive Model with Exogenous Variables) para múltiplas entradas e múltiplas saídas MIMO (Multiples Inputs and Multiplex Outputs). Utilizando o método da taxa de redução de erro ERR (Error Reduction Rate), determinando os modelos não lineares acoplados, não lineares desacoplados e lineares desacoplados para o elo 1 e elo 2 do manipulador. O estimador dos Mínimos Quadrados Recursivos (MQR) é utilizado para a estimação dos parâmetros do modelo representativo dos elos 1 e 2 do robô manipulador, verificando assim as eficiências dos modelos obtidos na identificação online.

Gli stili APA, Harvard, Vancouver, ISO e altri

21

Gautam, Deepthi Riggs Lloyd Stephen. "A robot mounted electromagnetic induction system for identification of a UXO free corridor". Auburn, Ala., 2009. http://hdl.handle.net/10415/1926.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

22

Balaniuk, Remis. "Identification structurelle". Phd thesis, Grenoble INPG, 1996. https://theses.hal.science/tel-00004974.

Testo completo

Abstract (sommario):

Dans ce mémoire nous proposons une méthode originale d'acquisition de modèles : l'identification structurelle. Nous nous plaçons dans un cadre intermédiaire entre les méthodes de modélisation classiques et les méthodes basées sur l'apprentissage. Nous montrons que pour le cas d'une classe particulière mais assez générale de fonctions il est possible d'inférer automatiquement la forme d'équation qui représente au mieux un certain processus physique, évitant ainsi l'effort de caractérisation du modelé par le concepteur. L'acquisition des modèles est faite suivant un protocole expérimental dans lequel l'identification de paramètres est à des problèmes n'ayant qu'une seule dimension d'entrée, réduisant ainsi la quantité de données requise. Les modèles générés par la méthode sont facilement différentiables, améliorables et réutilisables. La méthode peut être particulièrement utile en robotique ou l'on rencontre souvent le type fonctionnel considéré

Gli stili APA, Harvard, Vancouver, ISO e altri

23

Pinto, Carlos Roberto Alves. "Controle adaptativo aplicado em dois elos de um robô manipulador eletromecânico de cinco graus de liberdade". Universidade Federal da Paraíba, 2011. http://tede.biblioteca.ufpb.br:8080/handle/tede/5318.

Testo completo

Abstract (sommario):

Made available in DSpace on 2015-05-08T14:59:35Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2036431 bytes, checksum: 88e5aa7eb3c25813fc9bdaac159e943a (MD5) Previous issue date: 2011-11-25
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
This research aims at developing adaptive decentralized and centralized controllers for two links of an electromechanical manipulator robot of five degrees of freedom (5 DOF). The manipulator robot is consisted of five rotational joints, four links and a claw. Five DC motors are used to drive the robot and the motion transmission of the motors for the joints is achieved by gear trains. The measurements of the angular positions of the joints are made by potentiometers. Models of the manipulator robots, which are coupled and nonlinear, are obtained by using Newton- Euler and Lagrange equations. In this research, the models of the links of the manipulator robot are obtained in real time for each sampling period. The parameters of the links, to be controlled, are identified by recursive least squares (RLS) method resulted from imposed excitation to the motors which activate the links and the responses obtained by means of the joints. These parameters are used in the designs of adaptive controllers for the positions control of the link joints in question. Experimental results are presented, as well as evaluation of the achieved performance by the controlled links of the robot.
O objetivo do presente trabalho é apresentar o projeto de controladores adaptativos descentralizados e centralizados para dois elos de um robô manipulador eletromecânico de cinco graus de liberdade (5 GDL). O robô manipulador é composto por cinco juntas rotacionais, por quatro elos e uma garra. Cinco motores de corrente contínua são utilizados para o acionamento do robô. A transmissão do movimento dos motores para as juntas é realizada através de trens de engrenagens. As medidas das posições angulares das juntas são realizadas por potenciômetros. Modelos de robôs manipuladores são obtidos usando equações de Newton Euler ou de Lagrange; e são acoplados e não lineares. Neste trabalho, os modelos dos elos do robô manipulador são obtidos em tempo real, para cada período de amostragem. Os parâmetros dos elos, a serem controlados, são identificados pelo método dos mínimos quadrados recursivo (MQR), em função de excitações impostas aos motores que acionam os elos e das respostas obtidas nas juntas e são usados nos projetos dos controladores adaptativos, para o controle das posições das juntas dos elos em análise. Finalizando, são apresentados resultados experimentais bem como a avaliação do desempenho obtido pelos elos controlados do robô.

Gli stili APA, Harvard, Vancouver, ISO e altri

24

Parisi, Aaron Thomas. "An Application of Sliding Mode Control to Model-Based Reinforcement Learning". DigitalCommons@CalPoly, 2019. https://digitalcommons.calpoly.edu/theses/2054.

Testo completo

Abstract (sommario):

The state-of-art model-free reinforcement learning algorithms can generate admissible controls for complicated systems with no prior knowledge of the system dynamics, so long as sufficient (oftentimes millions) of samples are available from the environ- ment. On the other hand, model-based reinforcement learning approaches seek to leverage known optimal or robust control to reinforcement learning tasks by mod- elling the system dynamics and applying well established control algorithms to the system model. Sliding-mode controllers are robust to system disturbance and modelling errors, and have been widely used for high-order nonlinear system control. This thesis studies the application of sliding mode control to model-based reinforcement learning. Computer simulation results demonstrate that sliding-mode control is viable in the setting of reinforcement learning. While the system performance may suffer from problems such as deviations in state estimation, limitations in the capacity of the system model to express the system dynamics, and the need for many samples to converge, this approach still performs comparably to conventional model-free reinforcement learning methods.

Gli stili APA, Harvard, Vancouver, ISO e altri

25

Wernholt, Erik. "Multivariable Frequency-Domain Identification of Industrial Robots". Doctoral thesis, Linköping : Department of Electrical Engineering, Linköping University, 2007. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-10149.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

26

Janeke, Hanna. "Modellering, identifiering och reglering av skannern i ett laserbatymetrisystem". Thesis, Linköping University, Department of Electrical Engineering, 2005. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-2823.

Testo completo

Abstract (sommario):

The purpose with this masters thesis was to model the scanner in a system for laser bathymetry. The model was then used to develop a controller for the scanner so a good search pattern was achieved.

Two different types of models have been tested, a physical model and a Black Box model of Box Jenkins type. The physical model has been derived from Lagranges equations. Identification experiments have been used to compute the Black Box model and to find the unknown parameters in the physical model.

Three different controllers have been tested, a PID controller, a model predictive controller and a controller with feedforward. The controller with feedforward gave the best result. By softening the reference signal and using feedforward a good search pattern was achieved.

Gli stili APA, Harvard, Vancouver, ISO e altri

27

Faria, Bruno Guedes. "Identificação dinamica longitudinal de um dirigivel robotico autonomo". [s.n.], 2005. http://repositorio.unicamp.br/jspui/handle/REPOSIP/260166.

Testo completo

Abstract (sommario):

Orientadores: Paulo Augusto Valente Ferreira, Ely Carneiro de Paiva
Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação
Made available in DSpace on 2018-08-04T03:59:44Z (GMT). No. of bitstreams: 1 Faria_BrunoGuedes_M.pdf: 2340545 bytes, checksum: 440ff7a9aa46f5a39514e81423363750 (MD5) Previous issue date: 2005
Resumo: Nos últimos anos tem-se observado um crescente interesse de empresas e instituições de pesquisa pelo desenvolvimento de veículos robóticos, dotados de diferentes níveis de capacidade de operação autônoma, objetivando a execução de diversas tarefas. Dentro deste contexto o CenPRA, Centro de Pesquisas Renato Archer, propôs o Projeto AURORA. O Projeto AURORA (Autonomous Unmanned Remote mOnitoring Robotic Airship) tem como seu principal objetivo o desenvolvimento de protótipos de veículos aéreos tele-operados, e a obtenção de veículos telemonitorados, através do desenvolvimento de sistemas com graus de autonomia crescentes. Para que se possam agregar níveis crescentes de autonomia ao veículo, é essencial incrementar seu sistema de controle e navegação de maneira gradativa. Por esse motivo o aprimoramento das estratégias de controle do sistema é essencial. Assim, é primordial possuir um modelo fidedigno do sistema físico em questão, pois somente dessa forma é possível elaborar leis de controle e testá-las imediatamente em simulação antes de partir para os ensaios práticos no veículo real. Além disso, um modelo adequado é essencial para a simulação do vôo do dirigível de forma a permitir a análise preliminar de seu comportamento diante de uma nova missão. O principal objetivo deste trabalho é a implementação e validação de metodologias para a identificação do modelo dinâmico longitudinal do dirigível. Foram abordadas três metodologias para a identificação do modelo dinâmico do dirigível: a identificação estacionária, que identifica os coeficientes aerodinâmicos do dirigível a partir de um vôo estacionário, a identificação dinâmica, que identifica esses coeficientes e a dinâmica linearizada do veículo a partir de um vôo com entradas de perturbação conhecidas e, finalmente, a identificação por meio de estratégias evolutivas, que procura otimizar alguns parâmetros do modelo dinâmico. As três metodologias foram testadas, validadas e comparadas através de ensaios de simulação, utilizando-se o simulador do dirigível AS800 do Projeto AURORA
Abstract: In recent years many research institutions and companies have been demonstrating a growing interest in the development of unmanned aerial vehicles with different autonomous operation levels in order to allow for the performance of many types of tasks. Within this context, CenPRA (Renato Archer Research Center) proposed the Project AURORA. Project AURORA (Autonomous Unmanned Remote Monitoring Robotic Airship) aims at the development of unmanned airships remotely operated with a view to the creation of an autonomous flight airship by the incorporation of increasing levels of autonomy. In order to increase the vehicle autonomy level, the development of a proportionally enhanced control and navigation systems is essential. It is extremely important to have a very accurate model of the physical airship system, given that this is the only way to design control laws for the vehicle and test them in simulation before performing actual flight tests. Moreover, an accurate model is essential to predict the vehicle behavior in simulation before any real flight demanding a new type of mission. The definition of identification methodologies for the AS800 airship system identification is the main scope of this work. Three methodologies were considered to allow the airship dynamic model identification: stationary identification, which identifies aerodynamic coefficients from stationary stabilized flight conditions; dynamic identification, which identifies these coefficients and the vehicle linear dynamics from the application of known inputs into the system; and, finally, through evolution strategies, which uses an evolutionary approach for the optimization of the aerodynamic coefficients of the dynamic model. All the methodologies were tested, validated and compared through simulation experiments by using the AS800 airship simulator of the Project AURORA
Mestrado
Automação
Mestre em Engenharia Elétrica

Gli stili APA, Harvard, Vancouver, ISO e altri

28

Parde, Natalie. "Reading with Robots: A Platform to Promote Cognitive Exercise through Identification and Discussion of Creative Metaphor in Books". Thesis, University of North Texas, 2018. https://digital.library.unt.edu/ark:/67531/metadc1248384/.

Testo completo

Abstract (sommario):

Maintaining cognitive health is often a pressing concern for aging adults, and given the world's shifting age demographics, it is impractical to assume that older adults will be able to rely on individualized human support for doing so. Recently, interest has turned toward technology as an alternative. Companion robots offer an attractive vehicle for facilitating cognitive exercise, but the language technologies guiding their interactions are still nascent; in elder-focused human-robot systems proposed to date, interactions have been limited to motion or buttons and canned speech. The incapacity of these systems to autonomously participate in conversational discourse limits their ability to engage users at a cognitively meaningful level. I addressed this limitation by developing a platform for human-robot book discussions, designed to promote cognitive exercise by encouraging users to consider the authors' underlying intentions in employing creative metaphors. The choice of book discussions as the backdrop for these conversations has an empirical basis in neuro- and social science research that has found that reading often, even in late adulthood, has been correlated with a decreased likelihood to exhibit symptoms of cognitive decline. The more targeted focus on novel metaphors within those conversations stems from prior work showing that processing novel metaphors is a cognitively challenging task, for young adults and even more so in older adults with and without dementia. A central contribution arising from the work was the creation of the first computational method for modelling metaphor novelty in word pairs. I show that the method outperforms baseline strategies as well as a standard metaphor detection approach, and additionally discover that incorporating a sentence-based classifier as a preliminary filtering step when applying the model to new books results in a better final set of scored word pairs. I trained and evaluated my methods using new, large corpora from two sources, and release those corpora to the research community. In developing the corpora, an additional contribution was the discovery that training a supervised regression model to automatically aggregate the crowdsourced annotations outperformed existing label aggregation strategies. Finally, I show that automatically-generated questions adhering to the Questioning the Author strategy are comparable to human-generated questions in terms of naturalness, sensibility, and question depth; the automatically-generated questions score slightly higher than human-generated questions in terms of clarity. I close by presenting findings from a usability evaluation in which users engaged in thirty-minute book discussions with a robot using the platform, showing that users find the platform to be likeable and engaging.

Gli stili APA, Harvard, Vancouver, ISO e altri

29

Bayram, Alican. "Identification Of Kinematic Parameters Using Pose Measurements And Building A Flexible Interface". Master's thesis, METU, 2012. http://etd.lib.metu.edu.tr/upload/12614819/index.pdf.

Testo completo

Abstract (sommario):

Robot manipulators are considered as the key element in flexible manufacturing systems. Nonetheless, for a successful accomplishment of robot integration, the robots need to be accurate. The leading source of inaccuracy is the mismatch between the prediction made by the robot controller and the actual system. This work presents techniques for identification of actual kinematic parameters and pose accuracy compensation using a laser-based 3-D measurement system. In identification stage, both direct search and gradient methods are utilized. A computer simulation of the identification is performed using virtual position measurements. Moreover, experimentation is performed on industrial robot FANUC Robot R-2000iB/210F to test full pose and relative position accuracy improvements. In addition, accuracy obtained by classical parametric methodology is improved by the implementation of artificial neural networks. Neuro-parametric method proves an enhanced improvement in simulation results. The whole proposed theory is reflected by developed simulation software throughout this work while achieving accuracy nine times better when comparing before and after implementation.

Gli stili APA, Harvard, Vancouver, ISO e altri

30

Boots, Byron. "Spectral Approaches to Learning Predictive Representations". Research Showcase @ CMU, 2012. http://repository.cmu.edu/dissertations/131.

Testo completo

Abstract (sommario):

A central problem in artificial intelligence is to choose actions to maximize reward in a partially observable, uncertain environment. To do so, we must obtain an accurate environment model, and then plan to maximize reward. However, for complex domains, specifying a model by hand can be a time consuming process. This motivates an alternative approach: learning a model directly from observations. Unfortunately, learning algorithms often recover a model that is too inaccurate to support planning or too large and complex for planning to succeed; or, they require excessive prior domain knowledge or fail to provide guarantees such as statistical consistency. To address this gap, we propose spectral subspace identification algorithms which provably learn compact, accurate, predictive models of partially observable dynamical systems directly from sequences of action-observation pairs. Our research agenda includes several variations of this general approach: spectral methods for classical models like Kalman filters and hidden Markov models, batch algorithms and online algorithms, and kernel-based algorithms for learning models in high- and infinite-dimensional feature spaces. All of these approaches share a common framework: the model’s belief space is represented as predictions of observable quantities and spectral algorithms are applied to learn the model parameters. Unlike the popular EM algorithm, spectral learning algorithms are statistically consistent, computationally efficient, and easy to implement using established matrixalgebra techniques. We evaluate our learning algorithms on a series of prediction and planning tasks involving simulated data and real robotic systems.

Gli stili APA, Harvard, Vancouver, ISO e altri

31

Jiang, Bing. "Ubiquitous monitoring of distributed infrastructures /". Thesis, Connect to this title online; UW restricted, 2006. http://hdl.handle.net/1773/6118.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

32

Barreto, Guerra Jean Paul. "Design of a mobile robot’s control system for obstacle identification and avoidance using sensor fusion and model predictive control". Master's thesis, Pontificia Universidad Católica del Perú, 2017. http://tesis.pucp.edu.pe/repositorio/handle/123456789/9507.

Testo completo

Abstract (sommario):

The aim of this master thesis is to design a control system based on model predictive control (MPC) with sensor data fusion for obstacle avoidance. Since the amount of obtained data is larger due to multiple sensors, the required sampling time has to be larger enough in comparison with the calculation time of the optimal problem. Then it is proposed a simplification of the mobile robot model in order to reduce this optimization time. The sensor data fusion technique uses the range information of a laser scanner and the data of a mono-camera acquired from image processing techniques. In image processing different detection algorithms are proposed such as shape and color detection. Therefore an estimation of the obstacles dimension and distance is explained obtaining accurate results. Finally a data fusion for obstacle determination is developed in order to use this information in the optimization control problem as a path constraint. The obtained results show the mobile robot behavior in trajectories tracking and obstacle avoidance problems by comparing two different sampling times. It is concluded that the mobile robot reaches the final desired position while avoiding the detected obstacles along the trajectory.
Ziel dieser Masterarbeit ist, einen Steuerungsentwurf auf Basis der modellprädiktiven Regelung (MPC) mit Sensordatenfusion und zur Hindernisvermeidung. Da die Menge der erhaltenen Daten aufgrund mehrerer Sensoren größer ist, muss die erforderliche Abtastzeit im Vergleich zur Rechenzeit des optimalen Problems größer sein. In der Arbeit wird eine Vereinfachung des mobilen Robotermodells vorgeschlagen, um diese Optimierungszeit zu reduzieren. Die Sensordaten-Fusionstechnik verwendet die Bereichsinformation eines Laserscanners und die Daten einer Monokamera, die durch Bildverarbeitungstechniken gewonnen werden. Bei der Bildverarbeitung werden verschiedene Erfassungsalgorithmen vorgeschlagen, wie z. B. Muster- und Farbdetektion. Eine Schätzung der Hindernisdimension und -distanz wird erklärt, um genaue Ergebnisse zu erzielen. Schließlich wird eine Datenfusion zur Hindernisbestimmung entwickelt, um diese Information im Optimalsteuerungsproblem als Pfadbeschränkung zu nutzen. Die erzielten Ergebnisse zeigen das Verhalten des mobilen Roboters bei Trajektorienverfolgungsund Hindernisvermeidungsproblemen, indem zwei verschiedene Abtastzeiten verglichen werden. Es wird gefolgert, dass der mobile Roboter die endgültige gewünschte Position erreicht, während die erkannten Hindernisse entlang der Trajektorie vermieden werden.
Tesis

Gli stili APA, Harvard, Vancouver, ISO e altri

33

Carvalho, James Sidney Freitas de. "Controle adaptativo aplicado em um robô manipulador de dois graus de liberdade planar". Universidade Federal da Paraíba, 2009. http://tede.biblioteca.ufpb.br:8080/handle/tede/5366.

Testo completo

Abstract (sommario):

Made available in DSpace on 2015-05-08T14:59:51Z (GMT). No. of bitstreams: 1 parte1.pdf: 1731249 bytes, checksum: b526d4df9a40f0fe59df8d314c9314af (MD5) Previous issue date: 2009-10-28
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
The objective of this thesis was to model and control in real time a planar manipulator robot with two degrees of freedom (2-DOF), comprising a rotational and a prismatic link. The rotational link is an aluminum U-channel activated by a motor-reducer DC. The prismatic link comprises a double-acting pneumatic cylinder and a pass-through rod fixed inside the U-channel and activated by a 5-way electropneumatic proportional valve with three positions. A 10 turn potentiometer senses the angular position of the rotational link, and a potentiometric ruler senses its linear position of the prismatic link. The mathematical model that represents the manipulator robot, whose parameters are estimated in real time by the recursive least squares (RLS) method, is obtained as a function of the inputs fed into the manipulator and its measured outputs, considering the coupling between the links, based on a structure predefined for this purpose. After modeling the system, self-adjustable adaptive controls of generalized minimum variance (GMV) are designed and implemented, which control the position of the manipulator robot according to trajectories specified for the two links. Lastly, the estimated and experimental responses of the system are presented and compared, considering its links operating in uncoupled and coupled mode under the action of the designed controls.
A presente tese tem como objetivo o modelamento e controle em tempo real de um robô manipulador de dois graus de liberdade planar, constituído de um elo rotacional e outro prismático. O elo rotacional é um perfil U em alumínio e tem como atuador um motor-redutor de corrente contínua. O elo prismático é composto por um cilindro pneumático de dupla ação e haste passante, fixado no interior do perfil U e tem como atuador uma válvula eletropneumática proporcional de cinco vias e três posições. O sensoriamento de posição angular do elo rotacional é realizado por meio de um potenciômetro de dez voltas, e o de posição linear do elo prismático é realizado através de uma régua potenciométrica. É obtido o modelo matemático representativo do robô manipulador cujos parâmetros são estimados em tempo real pelo método dos mínimos quadrados recursivo (MQR), em função de entradas impostas ao manipulador e das saídas obtidas considerando o acoplamento entre os elos, a partir de uma estrutura pré-definida para este fim. De posse do modelo do sistema, controladores adaptativos de variância mínima generalizado (GMV) auto-ajustáveis são projetados e implementados visando o controle de posição do robô manipulador, conforme trajetórias especificadas para ambos os elos. Resultados de simulações e experimentais dos modelos estimados e respostas do sistema, considerando seus elos operando de forma desacoplada e acoplada sob a ação dos controladores, são apresentados e comparados.

Gli stili APA, Harvard, Vancouver, ISO e altri

34

Meira, Anrafel Silva. "Controle adaptativo aplicado em um robô manipulador eletropneumático de três graus de liberdade cartesiano". Universidade Federal da Paraíba, 2010. http://tede.biblioteca.ufpb.br:8080/handle/tede/5410.

Testo completo

Abstract (sommario):

Made available in DSpace on 2015-05-08T15:00:05Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 1708281 bytes, checksum: 415dd2aa16def4aa6286d61c1482fef0 (MD5) Previous issue date: 2010-12-19
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
This work presents a design of adaptive controllers for cartesian electropneumatic manipulator robot with three degrees of freedom (3DOF) cartesian. The manipulator is basically composed of three valves and three electro-pneumatic cylinders. It use the estimator of the recursive least squares (RLS) to obtain the representative model of the links of the manipulator robot. The model is used in the design of adaptive controllers Dahlin Minimum Variance (DMV) with the criterion of Favier and Hassani for position control of links robot. The experimental results are presented, as well as evaluation of the achieved performance by the manipulator robot.
Este trabalho apresenta um projeto de controladores adaptativos para um robô manipulador eletropneumático de três graus de liberdade (3GDL) cartesiano. O robô manipulador é composto basicamente por três válvulas eletropneumáticas e três cilindros pneumáticos. É utilizado o estimador dos mínimos quadrados recursivos (MQR) para a obtenção do modelo representativo dos elos do robô manipulador. O modelo obtido é utilizado no projeto dos controladores adaptativos Dahlin Variância Mínima (DMV) com o critério de Favier e Hassani para o controle de posição dos elos do robô. Os resultados experimentais são apresentados, assim como a avaliação do desempenho obtido pelo robô manipulador.

Gli stili APA, Harvard, Vancouver, ISO e altri

35

Raptis, Ioannis A. "Linear and Nonlinear Control of Unmanned Rotorcraft". Scholar Commons, 2009. http://scholarcommons.usf.edu/etd/3482.

Testo completo

Abstract (sommario):

The main characteristic attribute of the rotorcraft is the use of rotary wings to produce the thrust force necessary for motion. Therefore, rotorcraft have an advantage relative to fixed wing aircraft because they do not require any relative velocity to produce aerodynamic forces. Rotorcraft have been used in a wide range of missions of civilian and military applications. Particular interest has been concentrated in applications related to search and rescue in environments that impose restrictions to human presence and interference. The main representative of the rotorcraft family is the helicopter. Small scale helicopters retain all the flight characteristics and physical principles of their full scale counterpart. In addition, they are naturally more agile and dexterous compared to full scale helicopters. Their flight capabilities, reduced size and cost have monopolized the attention of the Unmanned Aerial Vehicles research community for the development of low cost and efficient autonomous flight platforms. Helicopters are highly nonlinear systems with significant dynamic coupling. In general, they are considered to be much more unstable than fixed wing aircraft and constant control must be sustained at all times. The goal of this dissertation is to investigate the challenging design problem of autonomous flight controllers for small scale helicopters. A typical flight control system is composed of a mathematical algorithm that produces the appropriate command signals required to perform autonomous flight. Modern control techniques are model based, since the controller architecture depends on the dynamic description of the system to be controlled. This principle applies to the helicopter as well, therefore, the flight control problem is tightly connected with the helicopter modeling. The helicopter dynamics can be represented by both linear and nonlinear models of ordinary differential equations. Theoretically, the validity of the linear models is restricted in a certain region around a specific operating point. Contrary, nonlinear models provide a global description of the helicopter dynamics. This work proposes several detailed control designs based on both dynamic representations of small scale helicopters. The controller objective is for the helicopter to autonomously track predefined position (or velocity) and heading reference trajectories. The controllers performance is evaluated using X-Plane, a realistic and commercially available flight simulator.

Gli stili APA, Harvard, Vancouver, ISO e altri

36

North, Ben. "Learning dynamical models for visual tracking". Thesis, University of Oxford, 1998. http://ora.ox.ac.uk/objects/uuid:6ed12552-4c30-4d80-88ef-7245be2d8fb8.

Testo completo

Abstract (sommario):

Using some form of dynamical model in a visual tracking system is a well-known method for increasing robustness and indeed performance in general. Often, quite simple models are used and can be effective, but prior knowledge of the likely motion of the tracking target can often be exploited by using a specially-tailored model. Specifying such a model by hand, while possible, is a time-consuming and error-prone process. Much more desirable is for an automated system to learn a model from training data. A dynamical model learnt in this manner can also be a source of useful information in its own right, and a set of dynamical models can provide discriminatory power for use in classification problems. Methods exist to perform such learning, but are limited in that they assume the availability of 'ground truth' data. In a visual tracking system, this is rarely the case. A learning system must work from visual data alone, and this thesis develops methods for learning dynamical models while explicitly taking account of the nature of the training data --- they are noisy measurements. The algorithms are developed within two tracking frameworks. The Kalman filter is a simple and fast approach, applicable where the visual clutter is limited. The recently-developed Condensation algorithm is capable of tracking in more demanding situations, and can also employ a wider range of dynamical models than the Kalman filter, for instance multi-mode models. The success of the learning algorithms is demonstrated experimentally. When using a Kalman filter, the dynamical models learnt using the algorithms presented here produce better tracking when compared with those learnt using current methods. Learning directly from training data gathered using Condensation is an entirely new technique, and experiments show that many aspects of a multi-mode system can be successfully identified using very little prior information. Significant computational effort is required by the implementation of the methods, and there is scope for improvement in this regard. Other possibilities for future work include investigation of the strong links this work has with learning problems in other areas. Most notable is the study of the 'graphical models' commonly used in expert systems, where the ideas presented here promise to give insight and perhaps lead to new techniques.

Gli stili APA, Harvard, Vancouver, ISO e altri

37

Kommaraju, Mallik. "Predictor development for controlling real-time applications over the Internet". Texas A&M University, 2005. http://hdl.handle.net/1969.1/4813.

Testo completo

Abstract (sommario):

Over the past decade there has been a growing demand for interactive multimedia applications deployed over public IP networks. To achieve acceptable Quality of Ser- vice (QoS) without significantly modifying the existing infrastructure, the end-to-end applications need to optimize their behavior and adapt according to network char- acteristics. Most existing application optimization techniques are based on reactive strategies, i.e. reacting to occurrences of congestion. We propose the use of predic- tive control to address the problem in an anticipatory manner. This research deals with developing models to predict end-to-end single flow characteristics of Wide Area Networks (WANs). A novel signal, in the form of single flow packet accumulation, is proposed for feedback purposes. This thesis presents a variety of effective predictors for the above signal using Auto-Regressive (AR) models, Radial Basis Functions (RBF) and Sparse Basis Functions (SBF). The study consists of three sections. We first develop time- series models to predict the accumulation signal. Since encoder bit-rate is the most logical and generic control input, a statistical analysis is conducted to analyze the effect of input bit-rate on end-to-end delay and the accumulation signal. Finally, models are developed using this bit-rate as an input to predict the resulting accu- mulation signal. The predictors are evaluated based on Noise-to-Signal Ratio (NSR) along with their accuracy with increasing accumulation levels. In time-series models, RBF gave the best NSR closely followed by AR models. Analysis based on accu- racy with increasing accumulation levels showed AR to be better in some cases. The study on effect of bit-rate revealed that bit-rate may not be a good control input on all paths. Models such as Auto-Regressive with Exogenous input (ARX) and RBF were used to develop models to predict the accumulation signal using bit-rate as a modeling input. ARX and RBF models were found to give comparable accuracy, with RBF being slightly better.

Gli stili APA, Harvard, Vancouver, ISO e altri

38

Calliess, Jan-Peter. "Conservative decision-making and inference in uncertain dynamical systems". Thesis, University of Oxford, 2014. http://ora.ox.ac.uk/objects/uuid:b7206c3a-8d76-4454-a258-ea1e5bd1c63e.

Testo completo

Abstract (sommario):

The demand for automated decision making, learning and inference in uncertain, risk sensitive and dynamically changing situations presents a challenge: to design computational approaches that promise to be widely deployable and flexible to adapt on the one hand, while offering reliable guarantees on safety on the other. The tension between these desiderata has created a gap that, in spite of intensive research and contributions made from a wide range of communities, remains to be filled. This represents an intriguing challenge that provided motivation for much of the work presented in this thesis. With these desiderata in mind, this thesis makes a number of contributions towards the development of algorithms for automated decision-making and inference under uncertainty. To facilitate inference over unobserved effects of actions, we develop machine learning approaches that are suitable for the construction of models over dynamical laws that provide uncertainty bounds around their predictions. As an example application for conservative decision-making, we apply our learning and inference methods to control in uncertain dynamical systems. Owing to the uncertainty bounds, we can derive performance guarantees of the resulting learning-based controllers. Furthermore, our simulations demonstrate that the resulting decision-making algorithms are effective in learning and controlling under uncertain dynamics and can outperform alternative methods. Another set of contributions is made in multi-agent decision-making which we cast in the general framework of optimisation with interaction constraints. The constraints necessitate coordination, for which we develop several methods. As a particularly challenging application domain, our exposition focusses on collision avoidance. Here we consider coordination both in discrete-time and continuous-time dynamical systems. In the continuous-time case, inference is required to ensure that decisions are made that avoid collisions with adjustably high certainty even when computation is inevitably finite. In both discrete-time and finite-time settings, we introduce conservative decision-making. That is, even with finite computation, a coordination outcome is guaranteed to satisfy collision-avoidance constraints with adjustably high confidence relative to the current uncertain model. Our methods are illustrated in simulations in the context of collision avoidance in graphs, multi-commodity flow problems, distributed stochastic model-predictive control, as well as in collision-prediction and avoidance in stochastic differential systems. Finally, we provide an example of how to combine some of our different methods into a multi-agent predictive controller that coordinates learning agents with uncertain beliefs over their dynamics. Utilising the guarantees established for our learning algorithms, the resulting mechanism can provide collision avoidance guarantees relative to the a posteriori epistemic beliefs over the agents' dynamics.

Gli stili APA, Harvard, Vancouver, ISO e altri

39

Joukhadar, Ammar. "Simulation dynamique et applications robotiques". Phd thesis, Grenoble INPG, 1997. http://www.theses.fr/1997INPG0069.

Testo completo

Abstract (sommario):

Nous décrivons, dans cette thèse, des modèles et des algorithmes conçus pour produire des simulations dynamiques efficaces et consistantes, dans le contexte de la Robotique d'intervention (c'est-à-dire, pour les tâches robotiques qui impliquent des contraintes fortes sur la nature de l'interaction entre des objets qui ne sont pas forcément rigides). Ces modèles et ces algorithmes ont été intégrés et implantés dans le système Robot_Phy qui peut être potentiellement reconfiguré pour traiter une grande variété de tâches d'intervention, comme la manipulation dextre d'un objet par une main robotique, la manipulation d'un objet non rigide, la téléprogrammation du mouvement d'un véhicule tout-terrain, ou encore des tâches chirurgicales assistées par robot (par exemple, le positionnement d'un ligament artificiel dans la chirurgie du genou). L'approche utilise une nouvelle technique de modélisation physique pour produire des simulations dynamiques qui sont à la fois efficaces et consistantes avec les lois de la physique. Les avantages par rapport aux travaux antérieurs dans le domaine de la robotique et de la synthèse d'image sont: le développement d'une structure unique pour traiter simultanément le mouvement, les déformations, et les interactions; et l'incorporation d'algorithmes et de modèles appropriés pour obtenir un temps d'exécution efficace en assurant un comportement consistant avec les lois de la physique. Les contributions principales de ce système sont: l'intégration de la notion du système masse/ressort avec la dynamique d'objets rigides, la discrétisation adaptative basée sur la notion de matrice d'inertie et de centre d'inertie, le pas de temps adaptatif basé sur la notion de l'énergie mécanique pour optimiser le temps de calcul et éviter la divergence numérique, la détection rapide du contact entre polyèdres déformables, et l'identification des paramètres physiques en utilisant les algorithmes génétiques
We describe models and algorithms designed to produce efficient and physically consistent dynamic simulations. These models and algorithms have been implemented within the RobotΦ system\cite(RAP95) which can potentially be configured for a large variety of interven\-tion-style tasks such as dextrous manipulations with a robot hand; manipulation of non-rigid objects; tele-programming of the motions of an all-terrain vehicle; and some robot assisted surgery tasks (e. G. Positioning of an artificial ligament in knee surgery). The approach uses a novel physically based modeling technique to produce dynamic simulations which are both efficient and consistent according to the laws of the Physics. The main advances over previous works in Robotics and Computer Graphics fields are twofold: the development of a unique framework for simultaneously processing motions, deformations, and physical interactions; and the incorporation of appropriate models and algorithms for obtaining efficient processing times while insuring consistent physical behaviors

Gli stili APA, Harvard, Vancouver, ISO e altri

40

Huard, Benoît. "Contribution à la modélisation non-linéaire et à la commande d'un actionneur robotique intégré pour la manipulation". Thesis, Poitiers, 2013. http://www.theses.fr/2013POIT2262/document.

Testo completo

Abstract (sommario):

La réalisation de tâches de manipulation dextres requiert une complexité aussi bien dans la conception de préhenseur robotique que dans la synthèse de leurs lois de commande. Une optimisation de la mécatronique de ces systèmes permet de répondre aux contraintes d'intégration fonctionnelle en se passant de capteurs de force terminaux. L'utilisation de mécanismes réversibles rend alors possible la détermination du positionnement du système dans l'espace libre et la détection de son interaction avec les objets manipulés, grâce aux mesures proprioceptives inhérentes aux actionneurs électriques. L'objectif de cette thèse est de parvenir synthétiser, dans le contexte articulaire (un degré-de-liberté), une commande adaptée à la manipulation en tenant compte de ces particularités mécaniques. La méthode proposée est basée sur une commande robuste par rapport aux non-linéarités structurelles dues aux effets gravitationnels et aux frottements secs d'une part et par rapport aux rigidités variables des objets manipulés. L'approche choisie nécessite la connaissance précise de la configuration du système étudié à chaque instant. Une représentation dynamique de son comportement permet de synthétiser un capteur logiciel pour l'estimation des grandeurs indispensables à la commande. Ces différentes étapes sont validées par des essais expérimentaux pour justifier la démarche choisie menant à une commande adaptée à la manipulation d'objets
The realization of dexterous manipulation tasks requires a complexity in robotic hands design as well as in their control laws synthesis. A mecatronical optimization of these systems helps to answer for functional integration constraints by avoiding external force sensors. Back-drivable mechanics allows the free-space positioning determination of such system as far as the detection of its interaction with a manipulated object thanks to proprioceptives measures at electric actuator level. The objective of this thesis is to synthesize a control law adapted to object manipulation by taking into account these mechanical properties in a one degree-of-freedom case. The proposed method is based on a robust control according to structural non-linearities due to gravitational effects and dry frictions on the one hand and with regard to a variable rigidity of manipulated objects on the other hand. The chosen approach requires a precise knowledge of the system configuration at all time. A dynamic representation of its behavior enables a software sensor synthesis for the exteroceptives variables estimation in a control law application purpose. The different steps are experimentally validated in order to justify the chosen approach leading to object manipulation

Gli stili APA, Harvard, Vancouver, ISO e altri

41

Oueslati, Marouene. "Contribution à la modélisation dynamique, l'identification et la synthèse de lois de commande adaptées aux axes flexibles d'un robot industriel". Phd thesis, Ecole nationale supérieure d'arts et métiers - ENSAM, 2013. http://pastel.archives-ouvertes.fr/pastel-01016462.

Testo completo

Abstract (sommario):

Les robots industriels représentent un moyen de production sophistiqués pour l'industrie manufacturière d'aujourd'hui. Ces manipulateurs sont plus agiles, plus flexibles et moins coûteux que les machines-outils spécialisées. L'exploitation de ces avantages fait l'objet d'une demande croissante de l'industrie. La dynamique de ces manipulateurs est soumise à des nombreuses sources d'imprécision. En effet les défauts de la chaîne de transmission, ou encore les éléments de liaisons peuvent être le siège de déformations et de vibrations dégradant sensiblement leur précision. Ces phénomènes physiques sont d'autant plus difficiles à compenser que seul un sous ensemble des états du système est mesuré par les codeurs moteurs. La structure de commande industrielle actuelle d'un robot n'agit donc pas directement sur ces phénomènes. Il est nécessaire alors de progresser sur le front de l'amélioration de la précision par l'adaptation de la commande à ces nouvelles exigences. Un état de l'art met en évidence un manque de travaux qui traitent de l'élaboration d'anticipations adaptées aux axes d'un robot et intégrant les phénomènes de déformation. En outre, la planification de trajectoire n'est classiquement pas remise en cause et peu évoquée. Elle représente pourtant un moyen d'action éprouvé afin d'améliorer les performances dynamiques en suivi de profil. L'approche proposée dans ce mémoire se veut une alternative à ces méthodes. Elle est basée sur une exploitation d'un modèle dynamique représentatif et détaillé. Il intègre les principaux phénomènes physiques mis en évidence tels que les effets de la gravité, les systèmes mécaniques de compensation, les forces de frottement et la flexibilité articulaire. Cette modélisation associée à des méthodes d'identification expérimentale est exploitée afin de déduire une structure de commande. Elle permet la réduction des déformations élastiques et des vibrations par une action sur la précommande et sur la loi de mouvement adaptée. Ainsi, nous introduisons une méthode d'estimation non asymptotique appliquée en robotique, afin d'estimer rapidement les paramètres vibratoires de ce dernier et contribue à une réactualisation des modèles exploités. Des résultats expérimentaux montrent que cette méthodologie mène à une amélioration des performances de positionnement par rapport à la commande industrielle.

Gli stili APA, Harvard, Vancouver, ISO e altri

42

Mendes, Ellon Paiva. "Identifica??o em tempo real de modelo din?mico de rob? m?vel com acionamento diferencial e zona morta". Universidade Federal do Rio Grande do Norte, 2012. http://repositorio.ufrn.br:8080/jspui/handle/123456789/15387.

Testo completo

Abstract (sommario):

Made available in DSpace on 2014-12-17T14:55:56Z (GMT). No. of bitstreams: 1 EllonPM_DISSERT.pdf: 1231242 bytes, checksum: 49456bef5c0d0bfdc5bf49d689568b60 (MD5) Previous issue date: 2012-01-27
Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico
Several mobile robots show non-linear behavior, mainly due friction phenomena between the mechanical parts of the robot or between the robot and the ground. Linear models are efficient in some cases, but it is necessary take the robot non-linearity in consideration when precise displacement and positioning are desired. In this work a parametric model identification procedure for a mobile robot with differential drive that considers the dead-zone in the robot actuators is proposed. The method consists in dividing the system into Hammerstein systems and then uses the key-term separation principle to present the input-output relations which shows the parameters from both linear and non-linear blocks. The parameters are then simultaneously estimated through a recursive least squares algorithm. The results shows that is possible to identify the dead-zone thresholds together with the linear parameters
V?rios rob?s m?veis apresentam comportamentos n?o-lineares, principalmente ocasionados por fen?menos de atrito entre as partes mec?nicas do rob? ou entre o rob? e o solo. Modelagens puramente lineares apresentam-se eficientes em alguns casos, mas ? preciso levar em considera??o as n?o-linearidades do rob? quando se deseja movimentos ou posicionamentos precisos. Este trabalho prop?e um procedimento de identifica??o param?trica do modelo de um rob? m?vel com acionamento diferencial, no qual s?o consideradas as n?o-linearidades do tipo zona-morta presentes nos atuadores do rob?. A proposta baseia-se no modelo de Hammerstein para dividir o sistema em blocos lineares e n?o-lineares. O princ?pio da separa??o do termo chave ? utilizado para demonstrar a rela??o entre as entradas e sa?das do sistema com os par?metros tanto da parcela linear quanto da n?o-linear. Os par?metros de ambas as parcelas s?o identificados simultaneamente, atrav?s de um algoritmo de m?nimos quadrados recursivo. Os resultados mostram que ? poss?vel identificar o valor os limites da zona-morta assim como os par?metros da parcela linear do modelo do sistema

Gli stili APA, Harvard, Vancouver, ISO e altri

43

Barreto, Guilherme de Alencar. "Redes neurais não-supervisionadas temporais para identificação e controle de sistemas dinâmicos". Universidade de São Paulo, 2003. http://www.teses.usp.br/teses/disponiveis/18/18133/tde-25112015-115752/.

Testo completo

Abstract (sommario):

A pesquisa em redes neurais artificiais (RNAs) está atualmente experimentando um crescente interesse por modelos que utilizem a variável tempo como um grau de liberdade extra a ser explorado nas representações neurais. Esta ênfase na codificação temporal (temporal coding) tem provocado debates inflamados nas neurociências e áreas correlatas, mas nos últimos anos o surgimento de um grande volume de dados comportamentais e fisiológicos vêm dando suporte ao papel-chave desempenhado por este tipo de representação no cérebro [BALLARD et al. (1998)]. Contribuições ao estudo da representação temporal em redes neurais vêm sendo observadas nos mais diversos tópicos de pesquisa, tais como sistemas dinâmicos não-lineares, redes oscilatórias, redes caóticas, redes com neurônios pulsantes e redes acopladas por pulsos. Como conseqüência, várias tarefas de processamento da informação têm sido investigada via codificação temporal, a saber: classificação de padrões, aprendizagem, memória associativa, controle sensório-motor, identificação de sistemas dinâmicos e robótica. Freqüentemente, porém, não fica muito claro até que ponto a modelagem dos aspectos temporais de uma tarefa contribui para aumentar a capacidade de processamento da informação de modelos neurais. Esta tese busca apresentar, de uma maneira clara e abrangente, os principais conceitos e resultados referentes à proposição de dois modelos de redes neurais não-supervisionadas (RNATs), e como estas lançam mão da codificação temporal para desempenhar melhor a tarefa que lhes é confiada. O primeiro modelo, chamado rede competitiva Hebbiana temporal (competitive temporal Hebbian - CTH), é aplicado especificamente em tarefas de aprendizagem e reprodução de trajetórias do robô manipulador PUMA 560. A rede CTH é uma rede neural competitiva cuja a principal característica é o aprendizado rápido, em apenas uma época de treinamento, de várias trajetórias complexas contendo ) elementos repetidos. As relações temporais da tarefa, representadas pela ordem temporal da trajetória, são capturadas por pesos laterais treinados via aprendizagem hebbiana. As propriedades computacionais da rede CTH são avaliadas através de simulações, bem como através da implementação de um sistema de controle distribuído para o robô PUMA 560 real. O desempenho da rede CTH é superior ao de métodos tabulares (look-up table) tradicionais de aprendizagem de trajetórias robóticas e ao de outras técnicas baseadas em redes neurais, tais como redes recorrentes supervisionadas e modelos de memória associativa bidirecional (BAM). O segundo modelo, chamado rede Auto-Organizável NARX (Self-Organizing NARX-SONARX), é baseado no conhecido algoritmo SOM, proposto por KOHONEN (1997). Do ponto de vista computacional, as propriedades de rede SONARX são avaliadas em diferentes domínios de aplicação, tais como predição de séries temporais caóticas, identificação de um atuador hidráulico e no controle preditivo de uma planta não-linear. Do ponto de vista teórico, demonstra-se que a rede SONARX pode ser utilizada como aproximador assintótico de mapeamentos dinâmicos não-lineares, graças a uma nova técnica de modelagem neural, chamada Memória Associativa Temporal via Quantização Vetorial (MATQV). A MATQV, assim como a aprendizagem hebbiana da rede CTH, é uma técnica de aprendizado associativo temporal. A rede SONARX é comparada com modelos NARX supervisionados, implementados a partir das redes MLP e RBF. Em todos os testes realizados para cada uma das tarefas citadas no parágrafo anterior, a rede SONARX tem desempenho similar ou melhor do que o apresentado por modelos supervisionados tradicionais, com um custo computacional consideravelmente menor. A rede SONARX é também comparada com a rede CTH na parendizagem de trajetórias robóticas complexas, com o intuito de destacar as principais diferenças entre os dois ) tipos de aprendizado associativo. Esta tese também propõe uma taxonomia matemática, baseada na representação por espaço de estados da teoria de sistemas, que visa classificar redes neurais não-supervisionadas temporais com ênfase em suas propriedades computacionais. Esta taxonomia tem como principal objetivo unificar a descrição de modelos neurais dinâmicos, facilitando a análise e a comparação entre diferentes arquiteturas, contrastando suas características representacionais e operacionais. Como exemplo, as redes CTH e a SONARX serão descritas usando a taxonomia proposta.
Neural network research is currently witnessing a significant shift of emphasis towards temporal coding, which uses time as an extra degree of freedom in neural representations. Temporal coding is passionately debated in neuroscience and related fields, but in the last few years a large volume of physiological and behavioral data has emerged that supports a key role for temporal coding in the brain [BALLARD et al. (1998)]. In neural networks, a great deal of research is undertaken under the topics of nonlinear dynamics, oscillatory and chaotic networks, spiking neurons, and pulse-coupled networks. Various information processing tasks are investigated using temporal coding, including pattern classification, learning, associative memory, inference, motor control, dynamical systems identification and control, and robotics. Progress has been made that substantially advances the state-of-the-art of neural computing. In many instances, however, it is unclear whether, and to what extent, the temporal aspects of the models contribute to information processing capabilities. This thesis seeks to present, in a clear and collective way, the main issues and results regarding the proposal of two unsupervised neural models, emphasizing how these networks make use of temporal coding to perform better in the task they are engaged in. The first model, called Competitive Temporal Hebbian (CTH) network, is applied specifically to learning and reproduction of trajectories of a PUMA 560 robot. The CTH model is a competitive neural network whose main characteristic is the fast learning, in just one training epoch, of multiple trajectories containing repeated elements. The temporal relationships within the task, represented by the temporal order of the elements of a given trajectory, are coded in lateral synaptic trained with hebbian learning. The computational properties of the CTH network are assessed through simulations, as well ) as through the practical implementation of a distributed control system for the real PUMA 560 robot. The CTH performs better than conventional look-up table methods for robot trajectory learning, and better than other neural-based techniques, such as supervised recurrent networks and bidirectional associative memory models. The second model, called Self-Organizing NARX (SONARX) network, is based on the well-known SOM algorithm by KOHONEN (1997). From the computational view-point, the properties of the SONARX model are evaluated in different application domains, such as prediction of chaotic time series, identification of an hydraulic actuator and predictive control of a non-linear plant. From the theoretic viewpoint, it is shown that the SONARX model can be seen as an asymptotic approximator for nonlinear dynamical mappings, thanks to a new neural modelling technique, called Vector-Quantized Temporal Associative Memory (VQTAM). This VQTAM, just like the hebbian learning rule of the CTH network, is a temporal associative memory techniques. The SONARX network is compared with supervised NARX models which based on the MLP and RBF networks. For all simulations, in each one of the forementioned application domains, the SONARX network had a similar and sometimes better performance than those observed for standard supervised models, with the additional advantage of a lower computational cost. The SONARX model is also compared with the CTH network in trajectory reproduction tasks, in order to contrast the main differences between these two types of temporal associative learning models. In this thesis, it is also proposed a mathematical taxonomy, based on the state-space representation of dynamical systems, for classification of unsupervised temporal neural networks with emphasis in their computational properties. The main goal of this taxonomy is to unify the description of dynamic neural models, ) facilitating the analysis and comparison of different architectures by constrasting their representational and operational characteristics. Is is shown how the CTH and SONARX models can be described using the proposed taxonomy.

Gli stili APA, Harvard, Vancouver, ISO e altri

44

LI, HUNG-MING, e 李鴻明. "System Identification and Control Parameter Tuning Techniques for Robotic Arm". Thesis, 2016. http://ndltd.ncl.edu.tw/handle/53290376594957561863.

Testo completo

Abstract (sommario):

碩士
國立中正大學
機械工程系研究所
104
Articulated robotic arms in the medical, military, industrial and other places have a considerable number of applications. They are gradually replacing many repetitive, high risk and assembly work and other matters. Recently, the requirements on robotic arms in speed, accuracy and precision are gradually increased. In order to achieve the goal of high speed and precision, the first and most important issue is to understand the internal characteristics of the motor. Through servo tuning, one can improve the overall stability and precision of the robot arm and further enhance the effectiveness of robotic arm. This thesis proposes a set of servo tuning procedures and inertia estimation method for articulated mechanical arm. Due to the articulated arm is a variable structure system, dynamic behavior will be changed under different postures. Single servo parameter set may not meet the dynamic characteristics of robot arms under different position. To perform parameter adjustment must understand the relationship between the drive parameters which can be done by performing each axis system identification and through experimental spectral response. The thesis paper presents a set of servo tuning method which will assist the user to find out the optimal parameters, enhance the performance of the robot arm, and achieve high speed and high precision objectives. Usually, the end effector of the robotic arm has different gripper performing different actions. Under this situation, load inertia will be changed. Inertia estimation method proposed in this thesis can effectively estimate robotic arm load inertia estimation and modify servo parameters to achieve optimum performances of mechanical arm.

Gli stili APA, Harvard, Vancouver, ISO e altri

45

Assadi, Hamed. "Posture dependent dynamics in robotic machining". Thesis, 2019. http://hdl.handle.net/1828/10879.

Testo completo

Abstract (sommario):

Compared to conventional machine tools, industrial robots offer great advantages such as multitasking, larger workspace, and lower price. However, these advantages of robots are undermined by their high structural flexibility leading to excessive deflections, severe vibrations, and ultimately violating dimensional tolerances and poor surface finish. Modeling the dynamics of robots under machining (e.g. milling and drilling) forces is essential for reducing deflections and vibrations during the process. Although modeling the dynamics of traditional machining systems is a well-studied subject, the existing modeling approaches are not applicable to robotic manipulators because of the posture-dependent dynamics of industrial robots. Within this context, the presented thesis aims to predict the stability of vibrations during robotic machining operations through prediction of posture dependent dynamic behavior of robots. A rigid-body modeling approach is used to identify the dynamic parameters of the robotic manipulator based on least squares estimation method. Next, by adopting a rigid link flexible joint model and employing experimental modal analysis to identify the joint stiffness and damping parameters, posture dependent dynamic response prediction of the robot is achieved. Finally, the posture-dependent milling stability is presented as a function of the predicted tool center point transfer function, spindle speed, and axial depth of cut. A Staubli TX200 robot and a Kuka KR90 robot are used as experimental case studies.
Graduate

Gli stili APA, Harvard, Vancouver, ISO e altri

46

Amadio, Fabio. "Gaussian Processes for Data-Driven Modeling and Control in Robotic Applications". Doctoral thesis, 2021. http://hdl.handle.net/11577/3408057.

Testo completo

Abstract (sommario):

Negli ultimi anni, il campo d’azione della robotica non si limita più alla sola industria, ma si sta espandendo in sempre più aspetti della vita umana. Per operare con successo in questa grande varietà di contesti, la prossima generazione di robot dovrà diventare sempre più autonoma e pronta ad adattarsi a diversi scenari. In tale contesto, le tecniche di Machine Learning e Reinforcement Learning possono fornire degli strumenti importanti per affrontare le sfide che coinvolgono una diffusione su larga scala dei sistemi robotici. Questi metodi data-driven potranno, potenzialmente, dotare i robot dei mezzi per gestire l’incertezza che caratterizza gli ambienti non strutturati nei quali verranno impiegati. A tale riguardo, i Gaussian Process si sono affermati come una tecnica di Machine Learning molto potente e flessibile. Essi possono essere utilizzati per risolvere complessi problemi di regressione, fornendo direttamente una stima dell’incertezza associata alle previsioni. In questa tesi, presentiamo soluzioni basate su Gaussian Process per diversi problemi di modellazione e controllo. In particolare, abbiamo sviluppato un nuovo algoritmo Model-Based Reinforcement Learning, chiamato Monte Carlo Probabilistic Inference for Learning COntrol (MC-PILCO), che può imparare autonomamente come controllare un sistema. L’algoritmo è stato poi modificato in modo tale da essere capace di gestire la presenza di misure parziali dello stato, condizione comune quando si lavora su sistemi meccanici. La tesi prosegue presentando una strategia data-driven per il controllo di robot, che utilizza dei Gaussian Process per stimare la dinamica inversa, utilizzandola poi all’interno di uno schema di controllo basato su feedback linearization. Infine, concludiamo esplorando le possibilità offerte dai Gaussian Process per la modellizzazione di dinamiche ad alta dimensionalità. Dei Gaussian Process sono stati impiegati sia per imparare una mappa che proietta le osservazioni ad alta dimensione in uno spazio latente di dimensione ridotta, che per stimare una funzione di transizione appropriata all’interno di questo spazio latente. Abbiamo chiamato il modello complessivo Controlled Gaussian Process Latent Variable Model (CGPDM), e lo abbiamo utilizzato per modellare la dinamica di un pezzo di tessuto manipolato da un robot. In ciascuno dei problemi studiati, le nostre soluzioni sono state valutate empiricamente usando dati reali e simulati.

Gli stili APA, Harvard, Vancouver, ISO e altri

47

Dahal, Pranesh. "Design of Object Identification System Based on Machine Vision". Thesis, 2015. http://ethesis.nitrkl.ac.in/7459/1/2015_BT_Design_P_Dahal.pdf.

Testo completo

Abstract (sommario):

Object sorting is an important aspect in almost all the industries. Production industries like food, chemical, petroleum and textile industries have to sort objects on numerous parameters. Various automated object sorting systems are required to avoid human flaws, with increase in productivity and reduce the overall time. Objective of the present work is to develop a part identification system using machine vision. Due to the advantage of LabVIEW in controlling hardware effectively it is employed in the present work. The Vision camera once identifies an object based on its attributes like color shape and size, immediately a signal should be communicated with the controller for separating that object. In this work the signal is shown as a glowing LED. Also the number of objects of particular category passing on the conveyor is counted and displayed to illustrate moving objects identification. A low speed conveyor belt is fabricated with different test objects passing over it. For identifying colors, wavelength data is used, for identifying the shape geometric pattern matching is used and for identifying the size edge detection is applied. The developed G-programming environment generates a graphic user interface in front panel. Ability to count the objects of specific attribute is tested for different trail runs. Thesis is organized as follows: Chapter 1 contains introduction to machine vision system, its components, the objectives of present study and literature review of similar works. Chapter 2 deals with various methods used in vision and their implementation in LabVIEW as done in this work was presented in chapter 3. Chapter 4 gives brief conclusions and future scope of present work.

Gli stili APA, Harvard, Vancouver, ISO e altri

48

Romano, Pedro Sousa. "A cooperative active perception approach for swarm robotics". Master's thesis, 2018. http://hdl.handle.net/10071/17143.

Testo completo

Abstract (sommario):

More than half a century after modern robotics first emerged, we still face a landscape in which most of the work done by robots is predetermined, rather than autonomous. A strong understanding of the environment is one of the key factors for autonomy, enabling the robots to make correct decisions based on the environment surrounding them. Classic methods for obtaining robotic controllers are based on manual specification, but become less trivial as the complexity scales. Artificial intelligence methods like evolutionary algorithms were introduced to synthesize robotic controllers by optimizing an artificial neural network to a given fitness function that measures the robots’ performance to solve a predetermined task. In this work, a novel approach to swarm robotics environment perception is studied, with a behavior model based on the cooperative identification of objects that fly around an environment, followed by an action based on the result of the identification process. Controllers are obtained via evolutionary methods. Results show a controller with a high identification and correct decision rates. The work is followed by a study on scaling up that approach to multiple environments. Experiments are done on terrain, marine and aerial environments, as well as on ideal, noisy and hybrid scenarios. In the hybrid scenario, different evolution samples are done in different environments. Results show the way these controllers are able to adapt to each scenario and conclude a hybrid evolution is the best fit to generate a more robust and environment independent controller to solve our task.
Mais de um século após a robótica moderna ter surgido, ainda nos deparamos com um cenário onde a maioria do trabalho executado por robôs é pré-determinado, ao invés de autónomo. Uma forte compreensão do ambiente é um dos pontos chave para a autonomia, permitindo aos robôs tomarem decisões corretas baseadas no ambiente que os rodeia. Abordagens mais clássicas para obter controladores de robótica são baseadas na especificação manual, mas tornam-se menos apropriadas à medida que a complexidade aumenta. Métodos de inteligência artificial como algoritmos evolucionários foram introduzidos para obter controladores de robótica através da otimização de uma rede neuronal artificial para uma função de fitness que mede a aptidão dos robôs para resolver uma determinada tarefa. Neste trabalho, é apresentada uma nova abordagem para perceção do ambiente por um enxame de robôs, com um modelo de comportamento baseado na identificação cooperativa de objetos que circulam no ambiente, seguida de uma atuação baseada no resultado da identificação. Os controladores são obtidos através de métodos evolucionários. Os resultados apesentam um controlador com uma alta taxa de identificação e de decisão. Segue-se um estudo sobre o escalonamento da abordagem a múltiplos ambientes. São feitas experiencias num ambiente terrestre, marinho e aéreo, bem como num contexto ideal, ruidoso e híbrido. No contexto híbrido, diferentes samples da evolução ocorrem em diferentes ambientes. Os resultados demonstram a forma como cada controlador se adapta aos restantes ambientes e concluem que a evolução híbrida foi a mais capaz de gerar um controlador robusto e transversal aos diferentes ambientes. Palavras-chave: Robótica evolucionária, Sistemas multi-robô, Cooperação, Perceção, Identificação de objetos, Inteligência artificial, Aprendizagem automática, Redes neuronais, Múltiplos ambientes.

Gli stili APA, Harvard, Vancouver, ISO e altri

49

Biglarbegian, Mohammad. "Systematic Design of Type-2 Fuzzy Logic Systems for Modeling and Control with Applications to Modular and Reconfigurable Robots". Thesis, 2010. http://hdl.handle.net/10012/5301.

Testo completo

Abstract (sommario):

Fuzzy logic systems (FLSs) are well known in the literature for their ability to model linguistics and system uncertainties. Due to this ability, FLSs have been successfully used in modeling and control applications such as medicine, finance, communications, and operations research. Moreover, the ability of higher order fuzzy systems to handle system uncertainty has become an interesting topic of research in the field. In particular, type-2 FLSs (T2 FLSs), systems consisting of fuzzy sets with fuzzy grades of membership, a feature that type-1 (T1) does not offer, are most well-known for this capability. The structure of T2 FLSs allows for the incorporation of uncertainty in the input membership grades, a common situation in reasoning with physical systems. General T2 FLSs have a complex structure, thus making them difficult to adopt on a large scale. As a result, interval T2 FLSs (IT2 FLSs), a special class of T2 FLSs, have recently shown great potential in various applications with input-output (I/O) system uncertainties. Due to the sophisticated mathematical structure of IT2 FLSs, little to no systematic analysis has been reported in the literature to use such systems in control design. Moreover, to date, designers have distanced themselves from adopting such systems on a wide scale because of their design complexity. Furthermore, the very few existing control methods utilizing IT2 fuzzy logic control systems (IT2 FLCSs) do not guarantee the stability of their system. Therefore, this thesis presents a systematic method for designing stable IT2 Takagi-Sugeno-Kang (IT2 TSK) fuzzy systems when antecedents are T2 fuzzy sets and consequents are crisp numbers (A2-C0). Five new inference mechanisms are proposed that have closed-form I/O mappings, making them more feasible for FLCS stability analysis. The thesis focuses on control applications for when (a) both plant and controller use A2-C0 TSK models, and (b) the plant uses T1 Takagi-Sugeno (T1 TS) and the controller uses IT2 TS models. In both cases, sufficient stability conditions for the stability of the closed-loop system are derived. Furthermore, novel linear matrix inequality-based algorithms are developed for satisfying the stability conditions. Numerical analyses are included to validate the effectiveness of the new inference methods. Case studies reveal that a well-tuned IT2 TS FLCS using the proposed inference engine can potentially outperform its T1 TSK counterpart, a result of IT2 having greater structural flexibility than T1. Moreover, due to the simple nature of the proposed inference engine, it is easy to implement in real-time control systems. In addition, a novel design methodology is proposed for IT2 TSK FLC for modular and reconfigurable robot (MRR) manipulators with uncertain dynamic parameters. A mathematical framework for the design of IT2 TSK FLCs is developed for tracking purposes that can be effectively used in real-time applications. To verify the effectiveness of the proposed controller, experiments are performed on an MRR with two degrees of freedom which exhibits dynamic coupling behavior. Results show that the developed controller can outperform some well-known linear and nonlinear controllers for different configurations. Therefore, the proposed structure can be adopted for the position control of MRRs with unknown dynamic parameters in trajectory-tracking applications. Finally, a rigorous mathematical analysis of the robustness of FLSs (both T1 and IT2) is presented in the thesis and entails a formulation of the robustness of FLSs as a constraint multi-objective optimization problem. Consequently, a procedure is proposed for the design of robust IT2 FLSs. Several examples are presented to demonstrate the effectiveness of the proposed methodologies. It was concluded that both T1 and IT2 FLSs can be designed to achieve robust behavior in various applications. IT2 FLSs, having a more flexible structure than T1 FLSs, exhibited relatively small approximation errors in the several examples investigated. The rigorous methodologies presented in this thesis lay the mathematical foundations for analyzing the stability and facilitating the design of stabilizing IT2 FLCSs. In addition, the proposed control technique for tracking purposes of MRRs will provide control engineers with tools to control dynamic systems with uncertainty and changing parameters. Finally, the systematic approach developed for the analysis and design of robust T1 and IT2 FLSs is of great practical value in various modeling and control applications.

Gli stili APA, Harvard, Vancouver, ISO e altri

Tesi sul tema "Robotic system identification"

Cita una fonte nei formati APA, MLA, Chicago, Harvard e in molti altri stili