Dissertations / Theses on the topic 'Automated fuzzy controllers'

Dissertation (PhD)--University of Stellenbosch, 2000.
ENGLISH ABSTRACT: A standard fuzzy logic controller is not robust enough to guarantee consistent closed-loop performance for highly non-linear plants. A finely tuned closed-loop response loses relevance as the system dynamics change with operating conditions. The self-adaptive fuzzy logic controller can track changes in the system parameters and modify the controller parameters accordingly. In most cases, self-adaptive fuzzy logic controllers are complex and rely on some form of mathematical plant model. The multi-mode fuzzy logic controller extends the working range of a standard fuzzy logic controller by incorporating knowledge of the non-linear system dynamics into the control rule-base. The complexity of the controller and difficulty in finding control rules have limited the application of multi-mode fuzzy logic controllers. An automated design algorithm is proposed for the design of a multi-mode control rule-base using qualitative plant knowledge. The design algorithm is cost function-based. The closed-loop response, local to a domain of the non-linear state space, can be tuned by manipulation of the cost function weights. Global closed-loop response tuning can be done by manipulation of the controller input gains. Alternatively, a self-learning or self-adaptive algorithm can be used in a model reference adaptive control architecture to optimise the control rule-base. Control rules responsible for unacceptable closed-loop performance are identified and their consequences modified. The validity of the proposed design method is evaluated in five case studies. The case studies illustrate the advantages of the multi-mode fuzzy logic controller. The results indicate that the proposed self-adaptive algorithm can be used to optimise a rule-base given a required closed-loop specification. If the system does not conform to the model reference adaptive architecture then the intuitive nature of the cost function based design algorithm proves to be an effective method for rule-base tuning.
AFRIKAANSE OPSOMMING: Standaard wasige logika beheerders is nie noodwendig robuust genoeg om goeie geslote lus werkverrigting vir hoogs nie-liniere aanlegte te waarborg nie. In Perfek ge-optimeerde beheerder se geslote lus werkverrigting mag verswak indien die aanleg-parameters weens bedryfstoestande verander. Self-aanpassende beheerders kan die verandering in die aanleg-parameters volg en die beheerder dienooreenkomstig optimeer. As In reël is In self-aanpassende beheerder kompleks en afhanklik van In wiskundige model van die aanleg. Die multi-modus wasige logika beheerder vergroot die werksbereik van die standaard wasige logika beheerder deur kennis aangaande die stelsel se bedryfstoestand en stelselparameters in die reël-basis in te bou. Die aanwending van die multi-modus beheerder word tans beperk deur die struktuur kompleksiteit en moeilike optimering van die reël-basis. In Ge-outomatiseerde multi-modus reël-basis ontwerps-algoritme wat gebruik maak van kwalitatiewe kennis van die aanleg en In kostefunksie word in hierdie proefskrif voorgestel. Die geslote lus gedrag beperk tot In gebied in die toestands-ruimte kan ge-optimeer word deur die kostefunksie gewigte te manipuleer. Die globale werkverrigting kan ge-optimeer word met die beheerder intree aanwinste. In Self-aanpassende algoritme in In model-verwysings aanpassende argitektuur word as altematieftot reël-basis optimering voorgestel. Reëls verantwoordelik vir swak werkverrigting word ge-identifiseer en verbeter deur modifikasie van die reëls se gevolgtrekkings. Die voorgestelde ontwerps-metode word deur middel van vyf gevallestudies ondersoek. Die studies dui die voordele van die multi-modus struktuur aan. Die self-aanpassende argitektuur is In kragtige hulpbron om In reël-basis te optimeer vir In gegewe geslote lus spesifikasie. Hierdie proefskrif toon aan dat indien die stelsel nie aan die vereistes van In model verwysingstelsel voldoen nie, is die kostefunksie benadering tot reël-basis ontwerp In aantreklike en intuïtief verstaanbare opsie om die reël-basis te optimeer.

The purpose of this research work is to propose an improved method for real-time sensitivity analysis (SA) applicable to large-scale complex systems. Borrowed from the EventTracker principle of the interrelation of causal events, it deploys the Rank Order Clustering (ROC) method to automatically group every relevant system input to parameters that represent the system state (i.e. output). The fundamental principle of event modelling is that the state of a given system is a function of every acquirable piece of knowledge or data (input) of events that occur within the system and its wider operational environment unless proven otherwise. It therefore strives to build the theoretical and practical foundation for the engineering of input data. The event modelling platform proposed attempts to filter unwanted data, and more importantly, include information that was thought to be irrelevant at the outset of the design process. The underpinning logic of the proposed Event Clustering technique (EventiC) is to build causal relationship between the events that trigger the inputs and outputs of the system. EventiC groups inputs with relevant corresponding outputs and measures the impact of each input variable on the output variables in short spans of time (relative real-time). It is believed that this grouping of relevant input-output event data by order of its importance in real-time is the key contribution to knowledge in this subject area. Our motivation is that components of current complex and organised systems are capable of generating and sharing information within their network of interrelated devices and systems. In addition to being an intelligent recorder of events, EventiC could also be a platform for preliminary data and knowledge construction. This improvement in the quality, and at times the quantity of input data, may lead to improved higher level mathematical formalism. It is hoped that better models will translate into superior controls and decision making. It is therefore believed that the projected outcome of this research work can be used to predict, stabilize (control), and optimize (operational research) the work of complex systems in the shortest possible time. For proof of concept, EventiC was designed using the MATLAB package and implemented using real-time data from the monitoring and control system of a typical cement manufacturing plant. The purpose for this deployment was to test and validate the concept, and to demonstrate whether the clusters of input data and their levels of importance against system performance indicators could be approved by industry experts. EventiC was used as an input variable selection tool for improving the existing fuzzy controller of the plant. Finally, EventiC was compared with its predecessor EventTracker using the same case study. The results revealed improvements in both computational efficiency and the quality of input variable selection.

We start by discussing fuzzy sets and the algebra of fuzzy sets. We consider some properties of fuzzy modeling tools. This is followed by considering the Mamdani and Sugeno models for designing fuzzy controllers. Various methods for using sets of data for desining controllers are discussed. This is followed by a chapter illustrating the use of genetic algorithms in designing and optimizing fuzzy controllers.Finally we look at some previous applications of fuzzy control in telecommunication networks, and illustrate a simple application that was developed as part of the present work.

Over the last few years there has been an increasing interest in the area of type-2 fuzzy logic sets and systems in academic and industrial circles. Within robotic research the majority of type-2 fuzzy logic investigations has been centred on large autonomous mobile robots, where resource availability (memory and computing power) is not an issue. These large robots usually have a variation of a Unix operating system on board. This allows the implementation of complex fuzzy logic systems to control the motors. Specifically the implementation of interval and geometric type-2 fuzzy logic controllers is of interest as they are shown to outperform type-1 fuzzy logic controllers in uncertain environments. However when it comes to using micro robots it is not practical to use type-1 and type-2 fuzzy logic controllers, due to the lack of memory and the processor time needed to calculate a control output value. The choice of motor controller is usually either fixed pre-set values, a variable scaled value or a PID controller to generate wheel velocities. In this research novel ways of implementing type-1 and interval type-2 fuzzy logic controllers on micro robots with limited resources are investigated. The solution thatis being proposed is the use of pre-calculated 3D surfaces generated by an off-line Fuzzy Logic System covering the expected ranges of the input and output variables. The surfaces are then loaded into the memory of the micro robots and can be accessed by the motor controller. The aim of the research is to test if there is an advantage of using type-2 fuzzy logic controllers implemented as surfaces over type-1 and PID controllers on a micro robot with limited resources. Control surfaces were generated for both type-1 and average interval type-2 fuzzy logic controllers. Each control surface was then accessed using bilinear interpolation to provide the crisp output value that was used to control the motor. Previously when this method has been used a single surface was employed to hold the information. This thesis presents the novel approach of the dual surface type-2 fuzzy logic controller on micro robots. The lower and upper values that are averaged for the classic interval type-2 controller are generated as surfaces and installed on the micro robots. The advantage is that nuances and features of both the lower and upper surfaces are available to be exploited, rather than being lost due to the averaging process. Having conducted the experiments it is concluded that the best approach to controlling micro robots is to use fuzzy logic controllers over the classical PID controllers where ever possible. When fuzzy controllers are used then type-2 fuzzy controllers (dual or single surface) should be used over type-1 fuzzy controllers when applied as surfaces on micro robots. When a type-2 fuzzy controller is used then the novel dual surface type-2 fuzzy logic controller should be used over the classic average surface. The novel dual surface controller offers a dynamic, weighted, adaptive and superior response over all the other fuzzy controllers examined.

Este trabalho teve por objetivo o desenvolvimento de um sistema de controle de navegação para um veículo autônomo utilizando lógica difusa. Atribui-se o nome de SNAT (Sistema de navegação-autônoma para um triciclo). O sistema é responsável pela automatização do processo de navegação de um veículo de testes monitorado e controlado de uma base remota mediante enlace de telemetria de dados. O usuário do sistema, pela tela principal da aplicação indica os pontos a seguir, cada um destes é referenciado pela sua latitude e longitude no mapa. Uma vez iniciada a navegação, o sistema de controle recebe periodicamente pacotes de dados da posição e direção do veículo; estes dados são processados pelo controlador difuso que gera os comandos para atuação sobre o veículo. O controlador difuso traduz as expressões qualitativas, comuns na comunicação humana em valores numéricos que representam a direção e a velocidade para manter ou corrigir a navegação do veículo de testes pelo rumo desejado. Nesta dissertação são apresentados diversos aspectos do desenvolvimento do controlador de navegação difusa, os mecanismos de comunicação e acionamento do veículo de uma base remota e os testes realizados que mostram os resultados do SNAT controlando automaticamente a navegação do veículo de testes.
This work had as a goal the development of a fuzzy logic based navigation control system for an autonomous vehicle. The developed control system was named SNAT (from the equivalent in portuguese of autonomous navigation system for a tricycle). The system controls and monitors the navigation of the vehicle from a remote base station using a telemetry data link. The user indicates, using a developed application, the navigation waypoints, referencing them by their latitude and longitude coordinates in a map. When the navigation starts, the system receives periodically data packets indicating the position and direction of the vehicle, these data are processed by the fuzzy control which returns commands to actuate over the vehicle. The fuzzy control translates qualitative expressions, common in human communication, into numerical values that represent the speed and the direction to keep the vehicle as near as possible to the desired navigation route. Many development aspects of the fuzzy controller and of the communication and actuation over the vehicle from the remote base are presented. Also are presented some results of the control system over navigation of the vehicle, which demonstrate that the system operates in a quite satisfactory manner.

Esta dissertação apresenta uma aplicação de inteligência artificial, a lógica fuzzy, estruturada como um controlador fuzzy, para a sintonia dos ganhos Kp e Ki de um controlador PI. Este último, configurado num conversor de frequência para a operação de um conjunto motor-bomba do sistema de abastecimento de água potável na região denominada de Recalque Mercês (RMER), na cidade de Curitiba, compreendendo 4 bairros com 140 km de tubulação e mais de 13 mil ligações a clientes. Hoje, o sistema funciona com controle PI tradicional de ganhos fixos, variando a velocidade do conjunto motorbomba, controlados por um sensor de pressão à jusante do bombeamento. Contudo, a variação de consumo de água potável fornecida é muito grande durante o período de um dia, sendo máxima no entardecer e mínima durante a madrugada, e com características sazonais, o consumo aumenta com o calor e diminui no frio, o que faz esse tipo de controle fixo ser ineficiente em determinados momentos. A proposta é adotar o controle fuzzy para alterar os coeficientes do controlador PI ao longo do dia, de forma autônoma, de acordo com o consumo instantâneo e assim, alcançar uma melhor eficiência energética do sistema, prezando pela qualidade do serviço prestado à população. Isto é, garantindo a vazão e a pressão de água para o consumidor. Por meio do MatLab®, mais precisamente os toolbox’s de simulação, Simulink® e SimScape®, foram criados dois modelos, um com o controle atual e o outro com a proposta inteligente, de uma abreviação da rede e do sistema de bombeamento da elevatória existente, a diminuição de 4 conjuntos motor-bomba para apenas 1 (um) e num trajeto único, perfazendo 5,4 km de distância, até o local denominado de ponto crítico, onde é mais longe, alto e difícil de se manter a pressão no patamar de 10 m.c.a. (1 bar), por norma. O controlador fuzzy desenvolvido possui 2 entradas, o nível do reservatório e a pressão à jusante do conjunto motor-bomba, e 2 saídas, os ganhos proporcional e integral para o PI. Com a estrutura baseado em regras, foram criadas 123 regras de inferência para que o sistema obtenha um desempenho favorável. As simulações foram realizadas com a referência de dados – 3 meses de histórico – de medições do reservatório e do sensor de pressão que é a realimentação do controle PI existente. Assim, a eficiência energética de 6,15% ao mês, em termos de energia elétrica consumida, foi conseguida, comparando com o modelo de controle tradicional.
This dissertation presents an application of artificial intelligence, the fuzzy logic, structured as a fuzzy controller, for the tuning of the Kp and Ki gains of a PI controller. The latter, configured in a frequency converter for the operation of a motor-pump assembly of the drinking water supply system in the region called Recalque Mercês (RMER), in the city of Curitiba, comprising 4 districts with 140 km of pipeline and more than 13 thousand connections to customers. Today, the system operates with PI traditional control with fixed gains, varying the speed of the motor-pump assembly, controlled by a pressure sensor downstream to the pumping. However, the variation in the consumption of potable water supplied is very large during the period of one day, being maximum at dusk and minimum during the dawn, and with seasonal characteristics, consumption increases with heat and decreases in the cold, which makes static control inefficient at certain times. The proposal is to adopt the fuzzy control to change the coefficients of the PI controller throughout the day, autonomously, according to the instantaneous consumption and thus achieve a better energy efficiency of the system, appreciating the quality of service provided to the population. That is, ensuring the flow and pressure of water to the consumer. Through MatLab®, simulation tools, Simulink® and SimScape®, two models were created, one with the current control and the other with the intelligent proposal. An abbreviation of the network and the existing pumping system of the lift, the reduction of 4 engine-pump assemblies to only 1 (one) and a single path, making 5.4 km distance, to the socalled critical point location, where it is further, higher and more difficult to keep the pressure at the 10 mwg (1 bar), as a rule. The developed fuzzy controller has 2 inputs, the reservoir level and the downstream pressure of the motor-pump assembly, and 2 outputs, proportional and integral gains for the PI. With the rulesbased framework, 123 inference rules were created for the system to perform favorably. The simulations were carried out with the reference of data – 3 months of historical – of measurements of the reservoir and pressure sensor that is the feedback of the existing PI control. Thus, the energy efficiency of 6.15% per month, in terms of electric power consumed, was achieved, compared to the traditional control model.

碩士
國立臺北科技大學
機電整合研究所
92
For the sake of reducing the seeking time of DVD-ROM, this thesis proposed a hybrid track seeking fuzzy controller (HTSFC) for DVD ROM's track seeking control. It is composed of the switching time controller, the speed controller and the start-up voltage regulator in long track seeking, and the speed controller was designed based on modulated span membership functions. An on-line design scheme based on genetic algorithm is also proposed for learning these 3 fuzzy controllers in long track seeking. Furthermore, we use the fuzzy gain regulator to do short track seeking of DVD-ROM in this thesis. An on-line design scheme based on modulated span membership functions and two kinds of genetic algorithm are also proposed for this regulator. By reducing the searching time through the HTSFC of auto learning on-line for DVD-ROM's track seeking to get the best performance of DVD-ROM is the prime object of this paper.

碩士
義守大學
電機工程學系
89
In many fuzzy control applications, most of researchers have found that the steady state error generally exists in fuzzy control problems. Usually, the system’s error and error change are used as two input parameters of the fuzzy control mechanism. According to the mathematical deduction, the output of the fuzzy controller can be looked as the PD controller with adjustable parameters, if the singleton fuzzifier and center average defuzzifier are adopted in the fuzzy control mechanism. It is the main reason of the steady state error existing in many fuzzy control problems. In this thesis, an automatic fuzzy controller is developed. In this controller, the corresponding values of fuzzy membership function used in fuzzfier and defuzzifier can be automatically adapted based on the error signal of the system output. Even for a very small steady state error, the fuzzy control mechanism can also produce a proper response to generate an adequate output force to overcome and eliminate the steady state error. In this research, the effects of transient time and overshoot of the system will be investigated and reported. Some simulations will be experimented to demonstrate the feasibility and the superiority of the fuzzy controller we developed.

碩士
中華大學
電機工程學系碩士班
88
To play music with friends is a wonderful experience for most instrument players. In addition to numerous of practices, playing music with other members is also a very important exercise in order to have wonderful performances. The purpose of playing music with others is to learn to listen to others'' playing, follow a suitable tempo and play in harmony. However, it is very unlikely that one can gather all his/her members every time one wants to practice. Therefore, there are the so-called accompaniment records available, such as "Piano-Concerto-Minus-One" which takes away the solo part from a piano concerto recording. The problem with these kinds of products is that a player can only follow a fixed tempo and there isn''t the necessary interaction between the user and the accompaniment. The objective of this thesis is to develop an automatic music accompaniment system such that the system can closely follow the player''s tempo, but not strictly. Because players usually have to listen to the accompaniment part, they virtually interact with the machine. What we need to do is to analyze the player’s tempo and design a controller to adjust the tempo of the accompaniment part in order to match the player’s tempo. As a matter of fact, different players have their own styles of following other members’ playing. The parameters of the controller have to be different for different players. In this thesis, a real time MIDI controller and a neuro-fuzzy based approach are used. The proposed Neuro-Fuzzy Based Tempo Controller produces the tempo adjustment parameters to the MIDI controller to achieve the automatic accompaniment work. The structure of the neuro-fuzzy controller is designed based on performance behavior suggested by professional players. Because the architecture of the controller is different from common feed-forward types, a new Forward-Backward training algorithm is proposed. Real-time verification experiments are performed on subjects with different performance capability levels. Though the performance of the system is very acceptable, it is still not up to the standard set by a professional pianist whose performance analysis is also included in this thesis.

by Lee Wai Man.
Thesis (M.Phil.)--Chinese University of Hong Kong, 1995.
Includes bibliographical references (leaves 118-[124]).
Chapter 1 --- Introduction --- p.1
Chapter 1.1 --- Introduction --- p.1
Chapter 1.2 --- Review of Previous Work --- p.3
Chapter 1.3 --- Scope of the Thesis --- p.4
Chapter 2 --- Background Knowledge of Fuzzy Control System --- p.7
Chapter 2.1 --- Introduction --- p.7
Chapter 2.2 --- Fuzzy Sets --- p.7
Chapter 2.2.1 --- Properties of Fuzzy Sets --- p.10
Chapter 2.2.2 --- Operations on Fuzzy Sets --- p.13
Chapter 2.3 --- Fuzzy Models --- p.14
Chapter 2.3.1 --- Linguistic Model --- p.15
Chapter 2.3.2 --- Takagi-Sugeno-Kang (TSK) Fuzzy Model --- p.16
Chapter 2.4 --- Fuzzy Inference System --- p.17
Chapter 2.4.1 --- Fuzzifier --- p.18
Chapter 2.4.2 --- Knowledge Base --- p.19
Chapter 2.4.3 --- Inference Engine --- p.19
Chapter 2.4.4 --- Defuzzifier --- p.20
Chapter 2.4.5 --- Product-Sum-Gravity Inference --- p.21
Chapter 3 --- Decomposition of Fuzzy Rules --- p.25
Chapter 3.1 --- Introduction --- p.25
Chapter 3.2 --- Decomposability of Fuzzy Inference System --- p.26
Chapter 3.3 --- The Decomposability condition --- p.29
Chapter 3.4 --- Determining Decomposed Parameters --- p.32
Chapter 3.5 --- Decomposable Approximation --- p.35
Chapter 3.5.1 --- Linear Approximation --- p.38
Chapter 3.5.2 --- Case Study --- p.40
Chapter 3.6 --- Limitation of Decomposable Approximation --- p.42
Chapter 3.7 --- Approximation Index --- p.44
Chapter 3.7.1 --- Case Study --- p.48
Chapter 3.8 --- Decomposable TSK Model --- p.52
Chapter 3.8.1 --- Case Study --- p.54
Chapter 3.9 --- Conclusion --- p.56
Chapter 4 --- Fuzzy Identification --- p.58
Chapter 4.1 --- Introduction --- p.58
Chapter 4.2 --- Least-squares Estimation --- p.59
Chapter 4.3 --- LSE Formulation of Various Fuzzy Models --- p.63
Chapter 4.3.1 --- Linguistic Model --- p.63
Chapter 4.3.2 --- TSK Model --- p.69
Chapter 4.3.3 --- Decomposable System --- p.75
Chapter 4.3.4 --- Comparative Case Study --- p.79
Chapter 4.4 --- Fuzzy Regional System Identification --- p.81
Chapter 4.4.1 --- Case Study --- p.86
Chapter 4.5 --- Recursive Estimation --- p.86
Chapter 4.5.1 --- Case Study --- p.90
Chapter 4.6 --- Conclusion --- p.90
Chapter 5 --- Performance-Based Fuzzy Gain Controller --- p.92
Chapter 5.1 --- Introduction --- p.92
Chapter 5.2 --- Conventional Fuzzy Control --- p.93
Chapter 5.3 --- Fuzzy Gain Control --- p.95
Chapter 5.4 --- Design Algorithm --- p.97
Chapter 5.5 --- Stability Design Approach --- p.98
Chapter 5.6 --- Simulation Case Study --- p.102
Chapter 5.7 --- Conclusion --- p.106
Chapter 6 --- Identification/Control Design Example --- p.107
Chapter 7 --- Conclusion --- p.115
Bibliography --- p.118

Thesis (M.Sc.)--University of Alberta, 2010.
Title from PDF file main screen (viewed on July 7, 2010). A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfillment of the requirements for the degree of Master of Science in Software Engineering and Intelligent Systems, [Department of] Electrical and Computer Engineering, University of Alberta. Includes bibliographical references.

A project report submitted to the Faculty of Engineering, University of Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Science in Engineering. Johannesburg 1997.
This project report describes the design of a fuzzy logic controller for automatic generation control (AGC) in Eskom in 1995 and the process of re-optimisation of the fuzzy logic controller in 1997. The main purpose of the AGC controller is to determine the shortfall or surplus generation of electricity for South Africa. The difficulties associated with optimising the original AGC controller, the design,implementation and optimisation of the fuzzy controller are described in detail. [Abbreviated Abstract. Open document to view full version]
AC2017

碩士
國立虎尾科技大學
電機工程系碩士班
106
This paper proposes a hydroponic system using fuzzy logic control. Regulating electrical conductivity and pH of nutrient solutions. The system can detect the EC value and pH value of the nutrient solution through the sensor at any time and pass the fuzzy control method. Control the working time of the adjusting solution pump. The purpose is to increase and decrease the EC value and pH of the nutrient solution. Maintain the EC value and pH of the nutrient solution. Maintain normal plant growth. The control system uses the Raspberry Pi3 development board to achieve and solve the problem of EC value and pH adjustment of the nutrient solution. The output is the ON/OFF operation time of the pump calculated by the fuzzy control. Meet the set value with the appropriate adjustment.

碩士
國立交通大學
電機與控制工程系所
96
In this thesis, a novel fuzzy/lead-lag switching control design is developed for the permanent magnetic (PM) DC brushless motor to improve its velocity response performance in both the bandwidth and stability. In the current feedback loop, the noise due to the Hall sensor, PWM switching, and environments has been practically reduced by applying the proposed medium filter and its velocity performance is significantly improved. Furthermore, to improve performance and stability of the velocity loop, both operations in linear and nonlinear ranges are concerned. In the linear operation region, the lead-lag compensator has been designed to effectively increase the bandwidth. On the other hand, the fuzzy controller has been designed to be more efficient in the nonlinear range due to the saturated current, the larger operation range, or significant load variation. Finally, the auto-switching fuzzy/lead-lag controller is designed by integrating two controllers to properly operate in various regions with a proper switching rule. The present controller is implemented on the TI TMS320 F2812 DSP and experimental results indicate that the proposed controller significantly improves both system bandwidth and response performance in the overshoot and steady-state oscillation under various conditions.

碩士
國立中興大學
電機工程學系
91
This thesis aims at the automatic gain controller (AGC) design for a IEEE 802.11b baseband processor receiver. We propose Fuzzy PI control strategy to regulate the amplitude of the input signal to the baseband A/D converter. The operational delay and the system convergence are analyzed through a software package Synopsys and FPGA emulators Altera FLEX10K50E. Robustness and fault tolerance are verified through experimental results.

碩士
大同大學
電機工程學系(所)
95
The automatic generation fuzzy neural network (AGFNN) controller with supervisory control for permanent magnet linear synchronous motor (PMLSM) is proposed in this thesis. It comprises an AGFNN controller, which has ability of neuron automatic generation with on-line learning and a supervisory controller, which is designed to stabilize the system states around a bounded region. The Mahalanobis distance (M-distance) formula is employed that the neural network has the ability of identification of the neurons will be generated or not. To improve the learning speed of back-propagation algorithm in AGFNN controller, a switching law and a momentum term are used in this study. The design of supervisory controller is derived in the Lyapunov sense; thus, the stability of the control system can be guaranteed. Simulation results show that the proposed controller is robust with regard to plant parameter variations and external load disturbance.

碩士
大同大學
電機工程學系(所)
94
This thesis presents the design and implementation of an Automatic Generation Fuzzy Neural Network (ADFNN) controller suitable for real-time control of the speed control of the permanent-magnet synchronous motor (PMSM) to track periodic step and sinusoidal reference inputs. The structure and parameter learning are done automatic and online. The structure learning is based on the partition of input space, and the parameter learning is based on the supervised gradient decent method using a delta law. Several simulation results are provided to demonstrate fast learning rate and accurate tracking performance of the proposed ADFNN control stratagem under the occurrence of parameter variations and external disturbance.

碩士
國立臺灣海洋大學
導航與通訊系
93
According to flight records, most aircraft accidents occurred during final approach or landing. Among these accidents, some causes are attributed to weather factors. Nowadays, most aircraft have installed the Automatic Landing System (ALS) which helps aircraft landing stably and reduces pilot’s work loading greatly. But control schemes of the conventional ALS usually use gain-scheduling and adaptive control techniques. If the flight conditions are beyond the preset envelope, the ALS is disabled and the pilot takes over. In order to facilitate the performance of the ALS, this study presents three intelligent controllers (Linguistic Fuzzy Neural Controller, Functional Fuzzy Neural Controller, and RAN Controller) with Particle Swarm Optimization (PSO) to guide the aircraft to a safe landing. The designed intelligent controller can enable the aircraft to adapt to wide range of disturbances. PSO algorithm is used to adjust control parameters of the pitch autopilot and to increase the ALS adaptability to different environments. In addition, the feedforward Neural Network controller is realized on a fixed-point DSP chip, and simulated with VisSim simulation software. Meanwhile, the procedure also uses PSO algorithm to enhance the DSP controller’s capability to adapt to more extensive turbulences.

碩士
大同大學
電機工程學系(所)
95
In this thesis, a radial basis function network (RBFN) based automatic generation fuzzy neural network (AGFNN) is proposed to control the rotor position of the permanent magnet linear synchronous motor (PMLSM) to track the period reference trajectories. The proposed RBFN based AGFNN not only has the advantages of the back-propagation algorithm, in which the parameter of the connected weights are adjusted but also has the advantages of the switching law, momentum term and RBFN, in which the tracking error and steady state responses will be betterment. The structure learning is based on the Mahalanobis distance and the parameter learning is based on the back-propagation algorithm. The simulation results of the proposed RBFN-based AGFNN with the periodic reference trajectories show that the tracking error and steady state responses have better performance, own the robustness performance under the parameter variation and external load disturbance.

Frequency Control of large interconnected power systems is governed by means of Automatic Generation Control (AGC), which regulates the system frequency and tie line power interchange at its nominal parameter set points. Conventional approaches to AGC controller design is centered around the Proportional, Integral and Derivative (PID) controller structures, which have found widespread application within industry. However, the dynamic changes experienced throughout the life cycle of power systems have many contributing factors, in part attributed to unknown knowledge of system behavior, neglected process dynamics and a limited knowledge of system interactions, which makes modeling for AGC systems particularly trying for conventional AGC controller design approaches. Therefore, in this study, Genetic - Fuzzy controllers (GA - Fuzzy) are applied as plausible candidates for Automatic Generation Controller design and application. In GA - Fuzzy controllers, genetic algorithms which are based on the foundation of evolutionary heuristics are used as a global search method for FLC design. This is particularly motivated by the fact that Fuzzy controllers, especially where there are large data sets, unknown process knowledge and insu cient expert data available, FLC controller design proves to be a daunting task. Therefore, this thesis explores the automatic design of FLC controllers through evolutionary heuristics and applies the designed controller to the AGC problem of large interconnected power systems. The design methodology followed is to understand power system interactions through power plant modeling and the simulation power plant models for the basis for AGC controller design. It is shown in this study that the performance of the GA - Fuzzy controller have favourable characteristics in terms of robust performance, robustness properties and compares favorably with conventional AGC controller techniques. The analysis of the GA - Fuzzy controller shows that problem formulation and chromosome encoding of the problem search space forms an important prerequisite for controller design by evolutionary methods. Therefore the study concludes by stating that GA - Fuzzy controllers are plausible for application within the power industry because of its desirable attributes and that future work would include extending this research into areas of renewable energy for study and application.