林巍聳臺灣大學：電機工程學研究所張良豪Chang, Liang-HaoLiang-HaoChang2007-11-262018-07-062007-11-262018-07-062005http://ntur.lib.ntu.edu.tw//handle/246246/53067目前對於輪式行動機器人動態模式的研究大都假設在輪子純轉動的條件下，也就是必須考慮沒有滑動的情形，而我們的研究主要是針對輪式行動機器人在泥濘的路面上行走所產生的效應，則不得不考慮滑動的影響。主要的工作為推導在考慮路面條件下機器人的數學模型，以及發展一個具有評斷機制與學習的控制器，整個控制器的架構包含一個模糊姿態控制器和一個自評自調速度控制器。而姿態控制器根據姿態誤差經由經驗法則而產生一組速度命令以減少姿態的誤差。而速度控制器可在未知路面情況下根據自我學習與最佳化的能力而使得受控系統的速度反應能追隨期望的速度命令。利用電腦模擬，藉由輪式行動機器人行走在多個泥濘的路上驗證我們所提出的控制架構。而結果顯示自評自調控制器可成功的在未知的路面情況的條件下運用。Previous researches on dynamic behavior about the wheeled mobile robot are mostly assuming a purely rolling case, i.e. no slipping. Instead, this research focuses on the wheeled mobile robot moving on muddy surface, on which the wheels may slip. The main contributions are deriving the mathematical model of the robot taking the road condition into consideration and developing the controller based on adaptive critic design. The controller is composed of a fuzzy posture controller and an adaptive critic velocity controller. The posture controller adopts posture error as inputs and produces desired velocities by fuzzy logic inference. The adaptive critic velocity controller conducts the linear and angular velocities with the capabilities of self-learning and optimization in response to unknown road condition and unmodeled nonlinearity. By computer simulation, the proposed design has been verified carefully by driving the wheeled mobile robot on variant muddy surface. The results show that the proposed adaptive critic velocity controller is a successful design to comply with unknown road conditions.List of contents Chapter 1 Introduction 1 1.1 Motivation and Objective 1 1.2 Literature Survey 3 1.2.1 Constraint System 3 1.2.2 Adaptive Critic Design 5 1.3 Contribution 8 1.4 Organization of Thesis 9 Chapter 2 Model of A Wheeled Mobile Robot Moving on Muddy Surface 11 2.1 The architecture of the target WMR 11 2.2 Wheel Rolling Motion 13 2.3 Model of the target WMR moving on muddy surface 14 Chapter 3 Adaptive Critic Design of the WMR Control System 24 3.1 Architecture of the WMR control system 24 3.2 Adaptive critic design of the velocity controller 25 3.2.1 Adaptive critic design 25 3.2.2 Updating process 26 3.2.3 The Architecture of adaptive critic velocity controller 27 3.2.4 The utility function 28 3.2.5 The action neural network 28 3.2.6 The critic neural network 29 3.2.7 The Approximation Model 31 3.2.8 Training algorithm of the adaptive critic velocity controller 32 3.3 The fuzzy based posture controller 34 3.3.1 Basics of fuzzy control 34 3.3.2 Fuzzy based posture controller 35 Chapter 4 Simulation Results 38 4.1 The Objective and environment of the Simulation 38 4.2 Preliminary training of the adaptive critic velocity controller 38 4.2.1 Test the self-learning from scratch 39 4.2.2 Test the performance of the adaptive critic velocity controller by tracking an arbitrary velocity command 43 4.2.3 Test the performance in acceleration, deceleration and constant velocity 46 4.3 Comparison of the adaptive critic and fuzzy based velocity controllers in trajectory tracking. 47 4.3.1 Autonomous navigation performing a left turn in case 1&2 48 4.3.2 Adaptation in Various Road Conditions. 54 Chapter 5 Conclusions 58 Reference: 591355241 bytesapplication/pdfen-US自主控制適應性評價autonomous navigationadaptive criticthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/53067/1/ntu-94-R92921061-1.pdf