Dissertations / Theses on the topic 'Robots – Dynamique'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Robots – Dynamique.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Jubien, Anthony. "Identification dynamique des robots à flexibilités articulaires." Nantes, 2014. http://archive.bu.univ-nantes.fr/pollux/show.action?id=f6f60215-c628-4adc-b6f8-bda76fb14f34.
Full textThis work is the result of collaboration between IRCCyN and ONERA on dynamic identification of robots with joint flexibilities, used for example in new applications for collaborative robotics. The usual identification technique in robotics requires the actual data of motor positions and the actual data of elastic deformations, which are usually not available in industrial robots. Recently, a new technique called DIDIM (Direct and Inverse Dynamic Identification Models), which uses only the data of motor torques, has been proposed and validated on rigid robots. This thesis proposes an extension of DIDIM, which uses no actual position data at all, to the case of robots with joint flexibilities. First, a comparative study on a rigid 6-axis robot with 61 parameters, shows the superiority of DIDIM over a conventional method CLOE (Closed- Loop Output Error) in position: DIDIM converges 100 times faster and is strongly more robust with respect to errors in the initial conditions. Second, DIDIM is extended to robots with joint flexibilities in a three steps procedure: a rigid model identification at low frequencies, an approximated identification of the flexible mode and of the inertia of each side of the flexibility, and finally the overall accurate identification of the full flexible dynamic model. A first experimental validation is performed on a test bench robot with one joint and one flexibility. A second validation in simulation on the 7 axes Kuka Light Weight Robot shows the effectiveness of DIDIM applied to industrial robots with joint flexibilities, in the case where the actual control law is known
Gallot, Guillaume. "Modélisation Dynamique et Commande d'un robot Anguille." Phd thesis, Ecole centrale de nantes - ECN, 2007. http://tel.archives-ouvertes.fr/tel-00306695.
Full textA travers cet objectif, la thèse se porte dans un premier temps sur la modélisation dynamique du robot sous la forme d'un mécanisme hybride (structures de robots parallèles montées en série) permettant ainsi d'être le plus proche possible du prototype construit. Pour cela nous avons utilisé les algorithmes récursifs de Newton-Euler pour les modèles dynamiques inverse et direct en les généralisant au cas des robots à base mobile. Nous avons également proposé un modèle de contact fluide-structure pour simuler le comportement du robot dans l'eau. Pour tester ces algorithmes, nous avons simulé le comportement du robot lors de différents types de nage et en avons tiré des conclusions qui nous ont guidées dans la conception du prototype.
Dans un deuxième temps, à partir d'un générateur de mouvements à base de CPGs (ou réseau de neurones), nous avons étudié des lois de commande pour réaliser des simulations de nage en boucle fermée. Ainsi, nous avons abordé les problèmes de la nage vers des points cible et l'évitement d'obstacles pour la nage en milieu confiné.
Marot, Jérôme. "Contribution à la synthèse dynamique optimale de la marche." Poitiers, 2007. http://www.theses.fr/2007POIT2340.
Full textNumerical synthesis of gait is carried out by solving a dynamic optimisation problem. The human-like locomotion system considered is made up of 12 internal degrees of freedom. Kinematic and dynamic models are stated according to fundamental determinant of gait which is the way the biped links up its ground-foot supports when walking. The support sequence adopted consists of a single support phase and two double support sub-phases. Special attention is devoted to the localisation of the centre of pressure under each foot, which are key indicators of the equilibrium state of the biped. Movement synthesis carried out using dynamic optimisation consists of extracting a feasible solution from dynamics equation by minimizing a performance criterion which is the time-integral of quadratic actuating torques. The problem stated is recast into a parametric optimisation problem using C3-spline approximations for the joint coordinates, and Ci-spline (0 ≤ i ≤ 2) for closure forces at front foot. The new optimisation variables are discrete values of these splines functions at times equally distributed along the motion time. Further global optimisation variables are accounted for: the step length and the relative length of the three step phases. In its final form, the problem is solved using SQP algorithms. Input data characterising the step to be generated can be limited to the only walking speed
Prades, Julien. "Dynamique linéarisée totale : Application aux robots parallèles." Thesis, Montpellier, 2018. http://www.theses.fr/2018MONTS106/document.
Full textThe research work of this thesis manuscript focus on the analysis of the frequency of robots’ vibrations. Our applications mainly revolve around architectures with parallel kinematics. First we examined parallel robots which are redundant in actuation and for which we are considering an increase of their oscillations’ frequency using the internal forces inherent to this type of structure. The aim is to use their actuation is the tensioning of their structure, and consequently, by analogy with a vibrating-wire, to enhance theiroscillation frequency. We have studied several redundancy planar robots and we demonstrate that in the case of robots which are typically designed to be stiff, the impact of added internal forces is of low relevance. The continuation of our research supports the following proposal: “High dynamics trajectories have an impact on the oscillation frequency of the mobile platform.” Indeed parallel robots, when designed to be light, can reach greater accelerations. We chose to concentrate on the study of the impact that dynamic effects canhave on the oscillation frequency of those robots’ mobile platform. The robots examined for our developments are planar parallel robots whether they have redundant actuation or not. We offer to study this impact based on a prime order development of the dynamic model. This linearisation of the dynamic model is intended to be more complete than those suggested by literature. We explain and verify the validity of our approach with a study on the link between speed and oscillation frequency on PR robots (pendulum on a vertical sliding guide) and RR robots ( double pendulum rotating horizontally). Then we will generalize our first order model and apply it to the four robots ( PRR-2 PRR-3, PRR-4, and Dual-V) to see if we are able to identify a pattern regarding the evolution ofoscillation frequencies. We observe that, depending on the trajectories, the dynamics have a low but noticeable, and often positive, impact on the increase of oscillation frequency of the mobile platform. However, since the trajectories and speed input laws are imposed, we have no choice but to be subjected to this impact
Ibrahim, Ouarda. "Contribution à la modélisation dynamique des robots parallèles et des robots hybrides." Nantes, 2006. http://www.theses.fr/2006NANT2096.
Full textIn this thesis, we propose novel methods for the dynamic modeling of parallel robots and hybrid robots. These methods are essentially based on the recursive Newton-Euler algorithm, which is easy to program and efficient during the real-time execution. For the dynamic modeling of the parallel robots, the proposed method takes into account all the dynamics of these robots what gets complete models of these structures that can be exploitable in advanced control algorithms and in simulation. The dynamics of the platform is calculated as a function of the Cartesian variables of the platform, whereas the dynamic models of legs are calculated as a function of the joint variables, This choice allows to obtain relation less complicated than the calculation of both dynamics in the joint space. To illustrate the various steps of the method we applied it to six parallel robots with different structures and mobilities, each of them presents particularities that guided us in the clarification of the method. Then, we studied the modeling of the hybrid robots composed of series of parallel modules. The proposed method is a recursive method based on calculation of the efforts applied to the various modules of the structure the ones with respect to the others and with respect to the environment. The inverse dynamic model generalizes, for the hybrid robots, the formulation of Newton-Euler of Luh developed for serial robots, whereas the direct dynamic model generalizes the algorithm of Featherstone. Then finally, the modeling of hybrid robots was extended to robots with a mobile base, for which we exploited the principle of a generalized module composed of the succeeding modules, the use of this principle allowed us to calculate the acceleration of the mobile base of the robot. To illustrate the feasibility of this method we applied it to two modules of the robot Anguille
Fried, Georges. "Contribution à la modélisation cinématique et dynamique des robots à architecture parallèle : calibrage en modes statique et dynamique." Paris 12, 1997. http://www.theses.fr/1997PA120060.
Full textÖzgür, Erol. "Des droites à la dynamique des robots parallèles." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2012. http://tel.archives-ouvertes.fr/tel-00777361.
Full textRennuit, Antoine. "Contribution au contrôle des humains virtuels interactifs." Nantes, 2006. http://www.theses.fr/2006NANT2080.
Full textVirtual humans contribute to the evolution of techniques related to concurrent engineering. They allow to test the behaviour of humans towards a digital mock-up of the product being designed. We develop an animation architecture that enables the interaction of virtual humans with the digital mock-up. After describing a state of the art of techniques linked to virtual humans' animation, we deal with passivity loss problems that can arise during interaction. We propose solutions to the problem. In a second step we tackle with constraints linked to the evolution of the virtual human in its environment: non penetration of the environment, joint limits, and above all balance. We develop a simple balance controller that only regard coplanar contacts (this is the case of walk movements), then we extend this controller to a more generic case allowing non coplanar contacts. We developed several types of control modes (free, constrained, automatic…). In the last section we try to schedule them, being careful to continuity between each mode. We then describe the results obtained thanks to our animation platform
David, Anthony. "Etablissement de critères analytiques pour la propulsion de systèmes marcheurs : application à l'initiation de la marche du robot bipède Rabbit." Orléans, 2006. http://www.theses.fr/2006ORLE2021.
Full textDHAOUI, AMOR. "Commande dynamique en temps-reel des robots manipulateurs." Paris 7, 1994. http://www.theses.fr/1994PA077232.
Full textAit, Mohamed Abderrahim. "Commande dynamique des robots redondants dans l'espace opérationnel." Nantes, 1995. http://www.theses.fr/1995NANT2026.
Full textYU, NIAN FEI. "Programmation hors-ligne des robots : dynamique et calibrage." Paris, ENSAM, 1993. http://www.theses.fr/1993ENAM0030.
Full textSans, Mariano. "Étude dynamique et aide à la conception d'un robot mobile à pattes." Toulouse, ENSAE, 1986. http://www.theses.fr/1986ESAE0002.
Full textBelkoura, Lotfi. "Modèle dynamique de robots et méthode d'identification des paramètres intertiels." Lille 1, 1993. http://www.theses.fr/1993LIL10140.
Full textThieffry, Maxime. "Commande dynamique de robots déformables basée sur un modèle numérique." Thesis, Valenciennes, Université Polytechnique Hauts-de-France, 2019. http://www.theses.fr/2019VALE0040/document.
Full textThis thesis focuses on the design of closed-loop control laws for the specific needs of dynamic control of soft robots, without being too restrictive regarding the robots geometry. It covers the entire development of the controller, from the modeling step to the practical experimental validation. In addition to the theoretical studies, different experimental setups are used to illustrate the results. A cable-driven soft robot and a pressurized soft arm are used to test the control algorithms. Through these different setups, we show that the method can handle different types of actuation, different geometries and mechanical properties. This emphasizes one of the interests of the method, its genericity. From a theoretical point a view, large-scale dynamical systems along with model reduction algorithms are studied. Indeed, modeling soft structures implies solving equations coming from continuum mechanics using the Finite Element Method (FEM). This provides an accurate model of the robots but it requires to discretize the structure into a mesh composed of thousands of elements, yielding to large-scale dynamical systems. This leads to work with models of large dimensions, that are not suitable to design control algorithms. A first part is dedicated to the study of the large-scale dynamic model and its control, without using model reduction. We present a way to control the large-scale system using the knowledge of an open-loop Lyapunov function. Then, this work investigates model reduction algorithms to design low order controllers and observers to drive soft robots. The validated control laws are based on linear models. This is a known limitation of this work as it constrains the guaranteed domain of the controller. This manuscript ends with a discussion that offers a way to extend the results towards nonlinear models. The idea is to linearize the large-scale nonlinear model around several operating points and interpolate between these points to cover a wider workspace
Harton, David, and David Harton. "Modélisation, conception mécanique, étude cinématique et dynamique d'un robot hybride redondant à (6+3) degrés de liberté." Master's thesis, Université Laval, 2020. http://hdl.handle.net/20.500.11794/38156.
Full textTableau d'honneur de la Faculté des études supérieures et postdoctorales, 2019-2020
Les robots collaboratifs prennent de plus en plus de place sur les lignes de production au sein des entreprises manufacturières. Leur facilité d'installation et d'utilisation ainsi que leur caractère sécuritaire constituent des avantages liés à leur utilisation. Les robots collaboratifs sériels sont les plus populaires dans l'industrie. Le principal avantage de ceux-ci est leur grand espace de travail. Cependant, l'inertie des architectures sérielles est généralement élevée, limitant ainsi les performances dynamiques du robot. Les robots parallèles sont plus avantageux sur ce point. Un principal avantage des robots parallèles collaboratifs est que les actionneurs sont situés près de la base, diminuant ainsi l'inertie, comparativement aux robot sériels. Cependant il existe peu de robots parallèles collaboratifs sur le marché. Dans ce mémoire est présenté un concept de robot hybride cinématiquement redondant utilisé pour des applications de coopération humain-robot à faible impédance. Ce robot d'architecture 3-[R(RR-RRR)SR] possède (6+3) degrés de liberté (ddl). La redondance du robot permet d'augmenter l'espace du travail notamment en rotation (comparativement à celui d'un robot non redondant d'architecture semblable) en diminuant le nombre de configurations singulières de type II dans l'espace de travail. Le robot est composé de trois jambes d'architecture hybride ayant chacune trois ddl et trois actionneurs ainsi qu'une plateforme composée d'un mécanisme parallèle plan à trois ddl. Les trois degrés de liberté redondants sont utilisés à la plateforme, afin d'y opérer une pince à partir des actionneurs aux jambes. Ce robot possède de grandes capacités en rotation, soient +-90° en inclinaison et en torsion. Ce robot est conçu de manière à ce qu'il soit rétrocommandable et qu'il ait une faible impédance et une faible inertie. Il ne possède aucun réducteur aux actionneurs. Le concept du robot présenté dans ce document est modulaire. En effet, l'architecture des jambes et de la plateforme peuvent différer légèrement afin d'adapter le robot à une application spécifique. Dans le cas présent, des jambes hybrides et une plateforme plane sont choisies pour des fins de simplicité et de maximisation de l'espace de travail. Dans ce document, les modèles cinématiques et dynamiques du robot, de la plateforme et des jambes sont présentés. Les étapes de conception mécanique ainsi qu'une étude de la sensibilité cinématique du robot sont également détaillés.
Les robots collaboratifs prennent de plus en plus de place sur les lignes de production au sein des entreprises manufacturières. Leur facilité d'installation et d'utilisation ainsi que leur caractère sécuritaire constituent des avantages liés à leur utilisation. Les robots collaboratifs sériels sont les plus populaires dans l'industrie. Le principal avantage de ceux-ci est leur grand espace de travail. Cependant, l'inertie des architectures sérielles est généralement élevée, limitant ainsi les performances dynamiques du robot. Les robots parallèles sont plus avantageux sur ce point. Un principal avantage des robots parallèles collaboratifs est que les actionneurs sont situés près de la base, diminuant ainsi l'inertie, comparativement aux robot sériels. Cependant il existe peu de robots parallèles collaboratifs sur le marché. Dans ce mémoire est présenté un concept de robot hybride cinématiquement redondant utilisé pour des applications de coopération humain-robot à faible impédance. Ce robot d'architecture 3-[R(RR-RRR)SR] possède (6+3) degrés de liberté (ddl). La redondance du robot permet d'augmenter l'espace du travail notamment en rotation (comparativement à celui d'un robot non redondant d'architecture semblable) en diminuant le nombre de configurations singulières de type II dans l'espace de travail. Le robot est composé de trois jambes d'architecture hybride ayant chacune trois ddl et trois actionneurs ainsi qu'une plateforme composée d'un mécanisme parallèle plan à trois ddl. Les trois degrés de liberté redondants sont utilisés à la plateforme, afin d'y opérer une pince à partir des actionneurs aux jambes. Ce robot possède de grandes capacités en rotation, soient +-90° en inclinaison et en torsion. Ce robot est conçu de manière à ce qu'il soit rétrocommandable et qu'il ait une faible impédance et une faible inertie. Il ne possède aucun réducteur aux actionneurs. Le concept du robot présenté dans ce document est modulaire. En effet, l'architecture des jambes et de la plateforme peuvent différer légèrement afin d'adapter le robot à une application spécifique. Dans le cas présent, des jambes hybrides et une plateforme plane sont choisies pour des fins de simplicité et de maximisation de l'espace de travail. Dans ce document, les modèles cinématiques et dynamiques du robot, de la plateforme et des jambes sont présentés. Les étapes de conception mécanique ainsi qu'une étude de la sensibilité cinématique du robot sont également détaillés.
Collaborative robots become present on production lines in factories. Their easiness of installation and use and their safety features make them more attractive. Serial collaborative robots are the most popular in the industry. Their main advantage is their large workspace. However, the inertia of the members of serial robots is the main limitation of the dynamic performances. Parallel robots are more attractive on this aspect. The main advantage of parallel robots is that their actuators are located near the base, decreasing the inertia compared to serial robots. However, there are few parallel collaborative robots on the market. In this Master's thesis, a novel concept of a redundant hybrid robot used for low impedance physical human-robot interaction (pHRI) applications is presented. This robot has a 3-[R(RRRRR) SR] architecture and (6+3) degrees of freedom (dof). Redundancy allows to get a larger workspace especially in rotation (compared to a non-redundant robot with the same architecture) by avoiding some type II singularity configurations in the workspace. The robot has three 3-dof hybrid legs having three actuators, and the platform, which is a 3-dof parallel planar mechanism. The three redundant degrees of freedom are used at the platform to actuate a gripper from the leg actuators. The robot has a large rotational workspace, namely > +-90° in tilt and torsion. This robot is designed to be backdrivable, with a low impedance and a low inertia. The actuators have no gearbox. The robot presented in this document is modular. Indeed, the leg architecture and the platform may differ depending on the application. In the present case, hybrid legs and planar platform are chosen for simplicity and workspace maximisation purposes. In this document, kinematic and dynamic models of the robot are presented. The main mechanical design steps and a study of the kinetic sensitivity are also detailed.
Collaborative robots become present on production lines in factories. Their easiness of installation and use and their safety features make them more attractive. Serial collaborative robots are the most popular in the industry. Their main advantage is their large workspace. However, the inertia of the members of serial robots is the main limitation of the dynamic performances. Parallel robots are more attractive on this aspect. The main advantage of parallel robots is that their actuators are located near the base, decreasing the inertia compared to serial robots. However, there are few parallel collaborative robots on the market. In this Master's thesis, a novel concept of a redundant hybrid robot used for low impedance physical human-robot interaction (pHRI) applications is presented. This robot has a 3-[R(RRRRR) SR] architecture and (6+3) degrees of freedom (dof). Redundancy allows to get a larger workspace especially in rotation (compared to a non-redundant robot with the same architecture) by avoiding some type II singularity configurations in the workspace. The robot has three 3-dof hybrid legs having three actuators, and the platform, which is a 3-dof parallel planar mechanism. The three redundant degrees of freedom are used at the platform to actuate a gripper from the leg actuators. The robot has a large rotational workspace, namely > +-90° in tilt and torsion. This robot is designed to be backdrivable, with a low impedance and a low inertia. The actuators have no gearbox. The robot presented in this document is modular. Indeed, the leg architecture and the platform may differ depending on the application. In the present case, hybrid legs and planar platform are chosen for simplicity and workspace maximisation purposes. In this document, kinematic and dynamic models of the robot are presented. The main mechanical design steps and a study of the kinetic sensitivity are also detailed.
Guegan, Sylvain. "Contribution à la modélisation et l'identification dynamique des robots parallèles." Nantes, 2003. http://www.theses.fr/2003NANT2112.
Full textRognant, Mathieu. "Modélisation et identification élastodynamiques des robots manipulateurs : application à l'étude d'une nouvelle structure de transstockeur." Rennes, INSA, 2010. http://www.theses.fr/2010ISAR0004.
Full textThe stacker crane tgd50 is a robot manipulator developed by industrial companies sydel and sedep and used in automated warehouses. The main objective of this phd is the study of the elastodynamic behavior of this manipulator. It aims to define how to modify the structure to improve its performance. Initially, an experimental diagnosis of this stacker crane has been achieved, in order to bring out limitations of this manipulator, and to validate modeling assumptions. A modeling methodology that can describe elastodynamic behavior of the tgd50 stacker crane is then developed, integrating the various assumptions made during the diagnosis phase. This methodology is experimentally validated, and implemented in a design process, leading to the development of a new structure whose workspace is augmented for a footprint equal to the former one. Finally, the experimental approach of diagnosis and modeling methodology are successfully implemented for identification on a 3d delta-like parallel structure. Results show both relevance and genericity of procedures suggested in this thesis. These tools are adapted to mechanical as well as identification for control
Hourieh, Abdullah. "Approche géométrique basée sur des fonctions de coordination pour la marche dynamique d'un bipède." Versailles-St Quentin en Yvelines, 2009. http://www.theses.fr/2009VERS0007.
Full textIn this thesis, a 3d dynamic walking gait of a mockup is proposed. The walking strategy is based on elementary coordination functions in the sagittal and frontal planes, in the simulation environment "ADAMS". The study starts by establishing coordination functions to generate the oscillation movement of the mockup in the frontal plane. The determination of some essential factors of the walk movement is obtained from human walking analysis. A leg lifting and a pelvis rotation are added to the oscillation movement to realize a first walking type. Increasing the step length has been done by a new configuration with legs being outside the pelvis plane and via the flexions of those legs. Finally, an energetic approach is proposed to enhance the walking stability of the biped system
Large, Fréderic. "Navigation autonome d'un robot mobile en environnement dynamique et incertain." Chambéry, 2003. https://tel.archives-ouvertes.fr/tel-00147376.
Full textThis thesis proposes two methods of navigation that allows a robot to move at high speeds in a partially unknown environment amidst moving obstacles. Both methods are based on a novel formalism inspired by the v-obstacle concept, allowing a fast estimation of the risk of collision associated with any feasible movement of the robot. The first method makes the robot react faster by computing only its next move. It is more suitable for fast changing environments, but may cause the robot to be blocked. The second method aims to build a complete trajectory to the goal. It can be suspended at any time in order to return an intermediate result and takes into account changes in the environment. The computed trajectories remain coherent between two successive decisions of the robot, they stay away from unpredictable obstacles and limit the cases that prevent the robot from reaching its goal. Our approach is compared to other existing methods. Its use in practical applications is shown in simulation. Several adaptations of our methods for implementation on real robots are presented. One of these adaptations leads to improved motion execution with artificial neural networks. Experimental results are given and analysed
Large, Fréderic Laugier Christian. "Navigation autonome d'un robot mobile en environnement dynamique et incertain." [S.l.] : [s.n.], 2003. http://emotion.inrialpes.fr/bibemotion/2003/Lar03/large-these-fr-03.pdf.
Full textConti, Patricia. "Contribution à la commande dynamique-adaptative des robots manipulateurs." Toulouse 3, 1987. http://www.theses.fr/1987TOU30220.
Full textMejri, Seifeddine. "Identification et modélisation du comportement dynamique des robots d'usinage." Thesis, Clermont-Ferrand 2, 2016. http://www.theses.fr/2016CLF22688/document.
Full textMachining robots have major advantages over cartesian machine tools because of their flexibility, their ability to reach inaccessible areas on a complex part, and their important workspace. However, their lack of rigidity and precision is still a limit for precision tasks. The stresses generated by the cutting forces and inertia are important and cause static and dynamic deformations of the structure which result in problems of workpiece surface. The aim of the thesis work is to characterize the dynamic behavior of robots during machining operation. This work followed a three-step approach : Modeling a first model considered as a reference where the robot is at rest. Then the identification of the dynamic behavior in service. Finally, the prediction of the cutting stability using the robot dynamic model. The originality of this work is the development of new operational modal identification methods. They integrate the machining conditions and result into a more accurate model than the first model of reference without being biased by harmonics. Finally, guidlines of robot’s configurations and excitation forces’ direction are proposed to ensure the robotic machining stability
Conti, Patricia. "Contribution à la commande dynamique-adaptative des robots manipulateurs." Grenoble 2 : ANRT, 1987. http://catalogue.bnf.fr/ark:/12148/cb37604046n.
Full textCurtelin, Jean-Jacques. "Identification des paramètres du modèle dynamique d'un robot en chaîne complexe : Aspects Méthodologiques et Micro-Informatiques." Lyon, INSA, 1989. http://www.theses.fr/1989ISAL0009.
Full textLuo, Qiuyue. "Marche Bipède 3D Auto-Stabilisante." Thesis, Ecole centrale de Nantes, 2020. http://www.theses.fr/2020ECDN0010.
Full textHumanoid robot, which can walk by two legs and perform skillful tasks using both arms with hands, could be considered as one of the ultimate robots. However, bipedal walking remains a complex phenomenon that has not been fully understood. The thesis is dedicated to find some physical insights that can explain the stability of periodic walking on horizontal floor. In human walking, the gait is usually expressed as a function of a phasing variable based on the internal state instead of time. The controlled variables (swing foot trajectories, vertical oscillation of center of mass, upper body motion,etc.)oftherobotsarebasedonaphasingvariable via the use of virtual constraints and the step timing is not explicitly imposed but implicitly adapted under disturbances. Firstly, simplified models of the robot: the linear inverted pendulum (LIP) model and variable length inverted pendulum (VLIP) model are used to study control strategies. The proposed control strategy for the LIP and VLIP models is extended through the proposed essential model to control a complete humanoid model. The walking algorithm proposed above is applied on the humanoid robots Romeo and TALOS
Marin, Hernandez Antonio Devy Michel. "Vision dynamique pour la navigation d'un robot mobile." Toulouse : INP Toulouse, 2004. http://ethesis.inp-toulouse.fr/archive/00000009.
Full textSeguin, Pascal. "Développement d'une technique d'optimisation paramétrique pour la synthèse de mouvements à dynamique régulière : application à la marche." Poitiers, 2003. http://www.theses.fr/2003POIT2326.
Full textThis work is aimed at optimizing motions of constrained dynamics systems. A parametric optimisation method is developed. It consists in approximating joint motion coordinates using spline functions of class C3, made up of 4-order polynomials linked at uniformly distributed knots, in order to avoid jerks at connecting points. Joint actuating torques as well as interaction forces associated with closure constraints of closed kinematic chains, are expressed, through dynamics equations, as functions depending on both the time and the optimisation parameters. The objective function to be minimized is obtained by integrating quadratic torques and interaction forces along the motion time. The initial dynamic optimisation problem is then recast as a parametric optimisation problem, which is solved using existing computing codes. This technique is used to carry out optimal synthesis of sagittal gait. The walking velocity is the only data required for generating an optimal step
Ait-Ahmed, Mourad. "Contribution à la modélisation géométrique et dynamique des robots parallèles." Toulouse 3, 1993. http://www.theses.fr/1993TOU30080.
Full textChevallereau, Christine. "Contribution à la commande dynamique de robots dans l'espace opérationnel." Nantes, 1988. http://www.theses.fr/1988NANT2005.
Full textBessonnet, Guy. "Optimisation dynamique des mouvements point à point de robots manipulateurs." Poitiers, 1992. http://www.theses.fr/1992POIT2001.
Full textBENALLEGUE, ABDELAZIZ. "Contribution a la commande dynamique adptative des robots manipulateurs rapides." Paris 6, 1991. http://www.theses.fr/1991PA066408.
Full textChevallereau, Christine. "Contribution à la commande dynamique de robots dans l'espace opérationnel." Grenoble 2 : ANRT, 1988. http://catalogue.bnf.fr/ark:/12148/cb376127182.
Full textPaccot, Flavien. "Contribution à la commande dynamique référencée capteurs de robots parallèles." Phd thesis, Clermont-Ferrand 2, 2009. http://www.theses.fr/2009CLF21929.
Full textPaccot, Flavien. "Contribution à la commande dynamique référencée capteurs de robots parallèles." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2009. http://tel.archives-ouvertes.fr/tel-00725568.
Full textZamzami, Ziad. "Analyse et exploitation de la dynamique naturelle des robots sous-actionnés." Thesis, Sorbonne université, 2018. http://www.theses.fr/2018SORUS408.
Full textHumans and animals are capable of performing graceful and agile movements. One of the key ingredients for such complex behaviors is motor coordination to exploit their natural dynamics which results in a synergistic motion that surpasses their physical limits. Despite the existence of powerful tools such as nonlinear trajectory optimization, they are usually treated as a black box. This thesis introduces the Dynamical Coupling Map (DCM), a novel graphical analysis technique, to help gain insight into the resulting trajectory of the optimization and analyze the capability of underactuated robots while taking into consideration underactuation constraints and torque limits. The DCM analysis technique along with the three novel performance measures aptly named Natural Dynamics Indexes (NDI) demonstrate the pivotal role of exploiting the natural dynamics to exceed the robot's physical capability which is dictated by the input torques limits. Furthermore, the merits of the DCM analysis are demonstrated on several dynamic maneuvers such as the swing-up motion of a simplified model of a gymnast on high bar and standing high jump for a high-dimensional humanoid robot with arm swing. In addition, the DCM is extended from a posteriori analysis tool to a fundamental heuristic for generating dynamic motions by leveraging the probabilistic motion planning approach with a new natural dynamic based heuristic. Thus, introducing the Natural Dynamics based Tree (NDT), a novel sampling-based kinodynamic planning algorithm which plans a dynamic motion based on the natural dynamics of the system resulting in a more computational efficient motion planning and with high-quality motion solution
Sabourin, Christophe. "Approche bio-inspirée pour le contrôle de la marche dynamique d'un bipède sous-actionné : validation expérimentale sur le robot RABBIT." Orléans, 2004. http://www.theses.fr/2004ORLE2065.
Full textCotton, Sébastien. "Modélisation, dynamique et estimation du centre de masse de robots humanoïdes." Thesis, Montpellier 2, 2010. http://www.theses.fr/2010MON20018/document.
Full textBefore they can interact with men, humanoid robots must be strongly enhanced in their modeling, their control and their design. Contrary to manipulator robots, the notion of center of mass is predominant in humanoid robots and will be central to the management of their balance. In this context, this thesis aims to provide accurate modeling of the center of mass of humanoid robots, whose complexity is increasing. Indeed, the models used today to determine the trajectory of center of mass are simplified models of humanoid robots. The works of this thesis revolve around three major contributions : kinematics and dynamics modeling as well as the estimation of the center of mass of humanoid robots. The first part proposes a transformation of the tree structure of the humanoid in a virtual serial chain locating its center of mass and allowing an adapted control of the latter. The dynamics of the robot is then expressed in the center of mass space allowing an accurate description of its acceleration. As such, the concept of dynamic manipulability of the center of mass is introduced. Finally, through the modeling in a virtual chain, a methodology that is today a reference in the field of center of mass estimation in humans is proposed. Many experiments show throughout this thesis the application and usefulness of this work
Villard, Claude. "Etude du comportement dynamique d'un robot quadrupède." Paris 6, 1993. http://www.theses.fr/1993PA066660.
Full textHabert, Olivier. "Modélisation dynamique d'un environnement intérieur pour robot mobile." Nancy 1, 1995. http://www.theses.fr/1995NAN10051.
Full textBorojéni, Diane. "Modélisation cinématique et dynamique des systèmes poly-articulés à chaînes ouvertes ou fermées : cas des robots parallèles." Paris 12, 2006. https://athena.u-pec.fr/primo-explore/search?query=any,exact,990002390270204611&vid=upec.
Full textArchitecture of parallel robots is so complex that it often makes the established models inapplicable to real time dynamic control. The real time constraint implies an optimization of the computational complexity of the kinematic and dynamic models. Therefore the interest is to parallelise calculations in order to implement them on a multi-processor system. This parallelization is made possible by introducing a new approach. The proposed approach is based on a global formalism. The extension of this formalism is performed for parallel robots. Considering the parallel robot as an entity composed of several serial robots moving a common load, the factorized expressions of the kinematic Jacobian and inertia matrices expressed in joint and operational space, are determined. This approach allows moreover a physical interpretation of the kinematics of the manipulator, by expressing the Jacobian matrix of the parallel robot as the sum of the Jacobian of a segment considered as opened, with a corrective term representing the closing of the kinematic chain. These factorizations present the base for developing parallel algorithms with the aim of future implementation on multi-processors architecture
Doubliez, Paul-François. "Marche de robots bipèdes anthropomorphes : simulation et validation expérimentale." Versailles-St Quentin en Yvelines, 2011. http://www.theses.fr/2011VERS0016.
Full textDespite the large number of existing humanoid robots, three main scientific issues remain: the decision-making autonomy, the autonomy in energy and the agile dynamic walking. This thesis focuses on this last issue. Firstly, a three-dimensional static walking generated by analytical geometrical inverse kinematics using active toes has been generated. Secondly, in order to generate a three-dimensional dynamic walking, the motions were studied separately in the frontal plane and in the sagittal one, with numerical models of robots with point-contacts feet. In the frontal plane, stable oscillations were generated by adapting energetically the foot placement and the push force. In the sagittal plane, a high level energy control of the propulsion force with the rear leg was successively tested and validated thanks to different numerical bipeds: with a robot with two rotational hip joints and two translational knee joints, and then with a robot with two rotational knee joints. Thirdly, in order to generalize the studied principles, a robot with ten bodies and nine rotational joints including active toes and discretized-contact feet allowed the dynamic crossing of an obstacle up to twenty centimetres high. Then, our geometrical techniques of walk pattern generation and our energy control of the propulsion have been implemented in the case of a three-dimensional dynamic walking. Finally, walking experimentations were realized with the ROBIAN biped robot prototype, and allowed to identify the transition between static and dynamic walking
Poirier, Mathieu. "Planification de tâches de manipulation par pivotement pour un robot humanoïde." Toulouse, INSA, 2009. http://eprint.insa-toulouse.fr/archive/00000306/.
Full textThis manuscript highlights the ability of a humanoid robot to perform a task hard to achieve by other types of robots. Here we are interested in manipulation of bulky objects. Such manipulation tasks are performed with great difficulty and involve several skills, such as taking into account motion of the body as one whole system (or whole-body motion) and perfect synchronization between the different limbs, arms and legs. We introduce here a whole-body motion planner that gives a humanoid robot the ability to automatically find a strategy for moving bulky objects by pivoting, taking into account a number of constraints : collision avoidance, limbs coordination, arms and legs, and stability control throughout the motion. Based on formal controllability results, defined upstream, the planner also inherits completeness of the probabilistic random sampling methods on which it is built. The geometric and kinematic capabilities of the planner proposed are also demonstrated through simulations and real experiments. Then we focus on solving more complex problems, giving the robot the ability to release and regrasp the object to manipulate if it is stuck in a narrow passage
Rostami, Mostafa. "Contribution à l'étude dynamique de la phase unipodale de la marche sagittale, et étude expérimentale du comportement dynamique d'un membre locomoteur anthropomorphe de robot bipède." Poitiers, 1999. http://www.theses.fr/1999POIT2281.
Full textBertrand, Sylvain. "A control strategy for high-speed running within unknown environments for a 3D bipedal robot." Versailles-St Quentin en Yvelines, 2013. http://www.theses.fr/2013VERS0003.
Full textThis thesis aims to improve insight into bipedal locomotion based on the robotics and biomechanics advances. The research conducted in this thesis is mainly centered on the running motion to gain insight in a gait that remains not fully understood in the control of bipedal robots. Based on an intuitive approach, such as the one used for walking with the robot M2 by J. E. Pratt and G. Pratt, a control strategy for high-speed running with a 3D complex bipedal robot is suggested and evaluated in simulation. Taking into account biomechanics observations, such as the spring-mass model, the simulated bipedal robot is able to run from 2. 7m/s up to 7m/s, to deal with external perturbations such as strongly uneven terrains and substantial external pushes, and to exhibit a very human-like gait
Li, Ying. "Analyse dynamique des systèmes multicorps flexibles et contrôle des robots déformables." Châtenay-Malabry, Ecole centrale de Paris, 1995. http://www.theses.fr/1995ECAP0464.
Full textGagnol, Vincent. "Modélisation du comportement dynamique des électrobroches UGV." Phd thesis, Clermont-Ferrand 2, 2006. https://theses.hal.science/docs/00/69/52/43/PDF/2006CLF21670.pdf.
Full textGagnol, Vincent. "Modélisation du comportement dynamique des électrobroches UGV." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2006. http://tel.archives-ouvertes.fr/tel-00695243.
Full textJiang, Xiaoling. "Dynamic trajectory planning and synthesis for fully-actuated cable-suspended parallel robots." Doctoral thesis, Université Laval, 2017. http://hdl.handle.net/20.500.11794/28389.
Full textIn the trend that robots are required to operate at increasingly high speeds and in large workspaces, dynamic trajectories of fully actuated cable-suspended parallel robots (CSPRs) that can extend beyond the robots’ static workspace are designed. Due to the unilateral property of cables, strategies to explore trajectories during which cable tensions can be guaranteed to remain positive are proposed. The planning and synthesis of dynamic periodic trajectories are first investigated for three-DOF point-mass, three-DOF planar, and six-DOF CSPRs. Based on an analytical approach, pure translation trajectories and more complex motion that includes changes in position and orientation are produced. A passive mechanical system that is equivalent to a CSPR is introduced to provide insight and facilitate the design of such trajectories. The dynamic differential equations that govern the translational component of the trajectories are shown to become linear under some conditions. Natural frequencies of the equivalent linear system are obtained and a generalization of periodic trajectories is accomplished by the integration of the linear system of differential equations. Natural trajectories associated with equivalent springs of constant stiffness and without any restriction of the amplitude are obtained. Using this formulation, the rotational component of three-DOF planar CSPRs becomes a nonlinear spring whose trajectories can be found in literature, which largely reduces the complexity of its trajectory planning. For six-DOF CSPRs, tilt and torsion angles are used to define the rotational component of the trajectories and the mathematical conditions corresponding to the linear trajectories are obtained. The above periodic trajectories provide insight into the fundamental properties of the mechanism and can be used in some specific applications. However, most practical situations require that the robot moves from one target point to another. Thus, a point-to-point dynamic trajectory planning technique for reaching a series of points for a point-mass three-DOF CSPR is proposed. Each trajectory segment is designed to have zero velocity at its endpoints. This formulation allows for trajectories that extend beyond the static workspace of the robot. A basis motion is introduced, which is a mathematical function that can be adapted for each coordinate direction along each trajectory segment. Kinematic constraints are satisfied through the selection of the coefficients for this function. Dynamic constraints are imposed by defining feasible regions within the workspace for each segment endpoint, based on the previous endpoint. This scheme is expanded to the trajectory planning of a six-DOF CSPR. Each trajectory segment is designed to have zero translational and rotational velocity at its endpoints; transitions between segments have translational and rotational acceleration. Additionally, smooth interpolation and singularity avoidance is achieved by using a unit quaternion to represent the rotational component of the trajectories. Then, a dynamic transition trajectory planning technique for three-DOF point-mass CSPRs is proposed to satisfy a real application where the robot is required to move from one trajectory to the next. This trajectory is designed to automatically chain multiple pre-generated trajectories beyond the static workspace in sequence with different starting points, as well as have the ability of starting from/ending with a resting position, while ensuring continuity up to the acceleration level. Two consecutive target trajectories are involved in the transition trajectory by using proper time functions, such that a goal trajectory is gradually reached by approaching the amplitude parameters and frequencies from those of a source trajectory. Additionally, each transition is based on the optimization of the departure point from its source trajectory and a minimum time for the transition to its goal trajectory. An example is provided to demonstrate the novel trajectory-planning technique. The robot is requested to start from the state of rest, merge into two consecutive ellipses, a straight line and a circle in sequence and then go back to the state of rest. Finally, experimental validation of the periodic and point-to-point trajectories is implemented on the prototype of a three-DOF point-mass CSPR and a six-DOF CSPR. Supplementary video files are included to demonstrate the results.
Ricard, Rémi. "Compensation dynamique de mécanismes parallèles." Thesis, Université Laval, 2006. http://www.theses.ulaval.ca/2006/23770/23770.pdf.
Full text