微管道机器人的智能模糊神经网络控制
|
摘要
随着微电子机械系统(Micro Electric Mechanical System)科学与技术的发展，微机器人的研究有着广泛的应用前景和社会需求，微机器人技术可应用于生物医学、航空航天、国防、工业、农业及家庭等领域。微机器人技术是MEMS科学与技术的一个重要内容，它是多学科技术的综合，由驱动器、传动装置、传感器、控制器、电源等的高度集成。目前，智能机器人的控制是自动控制的一个重要研究方向，由于机器人自身的特点，对微机器人实现自主控制有着特别的意义。
本文基于国内外已取得科研成果，主要针对15～20mm工业管道的微机器人控制技术的若干理论和实际问题，进行研究。主要工作有以下几方面。
(1)本文在查阅和掌握大量有关文献资料的基础上，阐述了微管道机器人的发展现状及展望，并介绍了有关微管道机器人的关键技术。
(2)阐述了智能控制—模糊神经网络控制系统的特点，并阐明了神经网络—模糊推理融合控制的必要性。
(3)分析了电磁式微管道机器人的移动原理，并根据微管道机器人的运动特点及其运行环境，分析了其运动的动力学稳定性。提出了管道内受限微机器人运动的动力学模型，并根据这一模型利用奇异摄动理论对微管道机器人管内运动稳定性进行了研究。从理论上确认了微管道机器人管内运动的可能性和稳定性。
(4)将模糊神经网络技术用于微管道机器人的控制，以实现微管道机器人的自主控制。根据提出的机器人的运行环境、神经网络结构等参数高度集成的模糊神经网络自学习算法(该算法包括基于环境识别的自监督学习算法和增强模糊神经网络结构等参数自学习算法)。对微管道机器人的自主控制进行了计算机仿真研究，并得出了初步的方针结果。
With the development of micro electric mechanical science and technology , the studies of micro robot have board application fields and social demands. Micro robots can be used in biomedicine, aeronautics and astronautics, national defense, Industry, agriculture and family fields. Micro robot technique is an important content of the micro electric mechanical science and technology. It is a synthesis of many different disciplines, and a high integration of actuators, drive equipments, sensors, controller and power supplies. The control of intelligent robots is a key study area of modern automatic controls. Owing to the features of robot, it is very important to achieve self-determination control of micro robots.
Based on the results have been made in this area so far, the author made a study on some theoretical and practical problems for the control techniques of micro pipe robot. The major works are shown as follows:
(1) Based on the amount of recently literature in this thesis, the development status quo and prospect are described, and introduced the key technology of micro-pipe robot.
(2) The characteristic of intelligent control-Fuzzy neural network control system is expatiated, and illustrated the necessity
(3) Based on the moving environment and the moving features of the robot, the dynamics stability of the micro robot moving is analyzed. The author built the dynamics model of its constrained motion with frictional contact, and its moving stability is analyzed by singular perturbation theory. The stability and possibility of the micro robot moving in pipe are affirmed in theory.
(4) The micro pipe robot is controlled by fuzzy neural networks, to achieve the self-determination control of the robot. A new self-learning algorithm that is highly integrated with moving environment of the robot and neural networks
structure is presented in this paper. According to the new self-learning algorithm, a self-learning control of the micro pipe robots is realized by computer simulation, and the simulating results are given.
本文基于国内外已取得科研成果，主要针对15～20mm工业管道的微机器人控制技术的若干理论和实际问题，进行研究。主要工作有以下几方面。
(1)本文在查阅和掌握大量有关文献资料的基础上，阐述了微管道机器人的发展现状及展望，并介绍了有关微管道机器人的关键技术。
(2)阐述了智能控制—模糊神经网络控制系统的特点，并阐明了神经网络—模糊推理融合控制的必要性。
(3)分析了电磁式微管道机器人的移动原理，并根据微管道机器人的运动特点及其运行环境，分析了其运动的动力学稳定性。提出了管道内受限微机器人运动的动力学模型，并根据这一模型利用奇异摄动理论对微管道机器人管内运动稳定性进行了研究。从理论上确认了微管道机器人管内运动的可能性和稳定性。
(4)将模糊神经网络技术用于微管道机器人的控制，以实现微管道机器人的自主控制。根据提出的机器人的运行环境、神经网络结构等参数高度集成的模糊神经网络自学习算法(该算法包括基于环境识别的自监督学习算法和增强模糊神经网络结构等参数自学习算法)。对微管道机器人的自主控制进行了计算机仿真研究，并得出了初步的方针结果。
With the development of micro electric mechanical science and technology , the studies of micro robot have board application fields and social demands. Micro robots can be used in biomedicine, aeronautics and astronautics, national defense, Industry, agriculture and family fields. Micro robot technique is an important content of the micro electric mechanical science and technology. It is a synthesis of many different disciplines, and a high integration of actuators, drive equipments, sensors, controller and power supplies. The control of intelligent robots is a key study area of modern automatic controls. Owing to the features of robot, it is very important to achieve self-determination control of micro robots.
Based on the results have been made in this area so far, the author made a study on some theoretical and practical problems for the control techniques of micro pipe robot. The major works are shown as follows:
(1) Based on the amount of recently literature in this thesis, the development status quo and prospect are described, and introduced the key technology of micro-pipe robot.
(2) The characteristic of intelligent control-Fuzzy neural network control system is expatiated, and illustrated the necessity
(3) Based on the moving environment and the moving features of the robot, the dynamics stability of the micro robot moving is analyzed. The author built the dynamics model of its constrained motion with frictional contact, and its moving stability is analyzed by singular perturbation theory. The stability and possibility of the micro robot moving in pipe are affirmed in theory.
(4) The micro pipe robot is controlled by fuzzy neural networks, to achieve the self-determination control of the robot. A new self-learning algorithm that is highly integrated with moving environment of the robot and neural networks
structure is presented in this paper. According to the new self-learning algorithm, a self-learning control of the micro pipe robots is realized by computer simulation, and the simulating results are given.
引文
1．刘文耀，杜君文．MEMS的概念及国内外研究状况．机械工程师，2001 (2)：8～10
2．张寿柏，蔡柄初．微电子机械系统(MEMS)研究现状及展望，薄膜科学与技术，2001．8 (3)
3．李志坚．微电子机械系统(MEMS)发展展望．电子科技导报，1999．1
4. Hiroyuki Fujita, A decade of MEMS and its future, IEEE 1997,0-7803-3744-1\97, P1-8
5．裴晓亮．微型机器人的发展动向．机器人技术及应用．1997 (4)：5～7
6．朱涛，谈大龙．微操作机器人系统及其关键技术研究．组合机床及自动化加工技术，2001 (11)：33～36
7．蔡昌骏，李彦明．承压管道检测机器人的多通讯系统．机械电子．2003 (4)：63～67
8. Toshio FUKUTA, Microrobot and application, IEEE 1993,0-7803-0985-5/93.P1-8
9．李人厚．智能控制理论与方法．西安：西安电子科技大学出版社，1999．10
10．权太范．信息融合 神经网络—模糊推理理论与应用．北京：国防工业出版社，2002．8
11．汤兵勇．模糊控制理论与应用技术．北京：清华大学出版社，2002．9
12．傅京孙，蔡自兴，徐光佑．人工智能及其应用(第二版)．北京：清华大学出版社，2000．5
13. Bj.Haug, S.C.Wu, Dynamics of mechanical systems with coulomb friction striction, impact and constraint addition-deletion-Ⅰ. Theory. Mech. mach. Theory,21.1996.P401-406
14. Cheng Lianglun,Yang Yimin. Study on In-pipe Micro Robot Dynamics and Intelligent Control, Pioneering international Symposium on Motion and Vibration Control in Mechatronics, April 6-7,1999, Tokyo, Japan.
15. J.Trinkle, J.Pang, S.Sudarky, and G.Lo. On dynamic multi-rigid-body contract problem with coulomb friction, ZAMM, Vol.77, 1997. P267～269
16. A.P. Ambler, R.J.Popplestone. Inferring the Position of Bodies from Specified Spatial Relationships, Artificial intelligence, 6(2), 1975.P157～174
17. R. Fealhustone, The dynamics of rigid body systems with multiple concurrent contacts. Robotics Reasearchm, The Third Int.Sym., Eds., O.D.Faugeras and G.Giralt 1986
18. I.Han and B.J.Gilmore. Multi-body impact motion with friction-analysis, simulation and experiment validation. Proc. ASME Design Technical conf. 1990
19．张东明，竺长安等．电磁驱动式微型机器人的移动机理和设计．机器人．1995 (17)：P223～228
20．孙麟治，孙萍等．细小管道内爬行的微机器人．光学精密工程．1995．05
21．程良伦，杨宜民．微管道机器人的管内运动稳定性分析．机器人，1999．07
22．R．E 奥马利著，江福汝，尚汉翼等译．奇异摄动引论．北京：科学出版社，1988．3
23．毕学涛编．高等动力学，天津大学出版社
24．蔡自兴．机器人学．北京：清华大学出版社，2001．9
25．李士勇等．模糊控制与智能控制理论与应用．哈尔滨：哈尔滨工业大学出版社，1997．7
26．罗志增，蒋静坪．机器人感觉与多信息融合．北京：机械工业出版社，2002．7
27．孙增圻等．智能控制理论与技术，北京：清华大学出版社，1997．4
28．王士同编．神经模糊系统及其应用，北京：北京航空航天大学出版社，1998．5
29. Lin C T and Lin C S C. Reinforcement structure/ parameter learning for Neural-network-based fuzzy logic control system. IEEE Trans. Fuzzy Systems, 1994,2 (1): P46-63
30. I.F. Chung. A Reinforcement Neuro-Fuzzy Combiner for Multiobjective Control. Accepted for publication to IEEE Transactions on Systems, Man, and Cybernetics, Part B, 1998.
31. C. T. Lin and C.P.Jou. Controlling Chaos by GA-based Reinforcement Learning Neural Network. Accepted for publication to IEEE Transactions on Neural Network. 1998.
32. J.H.Lai. Application of neural Fuzzy Network to Pyrometer Correction and Temperature Control in Rapid Thermal Processing. Accepted for publication to IEEE Transactions on Fuzzy Systems, 1998
33. C.F. Juang. Genetic Reinforcement Learning Through Symbiotic Evolution for Fuzzy Controller Design. Accepted for publication to IEEE Transactions on Systems, Man, and Cybernetics, Part B, 1998.
34. Qiao L et al. Self-supervised learning algorithm of environment recognition in driving vehicle. IEEE Trans. Syst., Man and cybern., 1996.26. (6): P843-850
35. Chen S B, Wu L and Wang Q L. Self-learning fuzzy neural network for control of uncertain systems with time delays. IEEE Trans. Sys., Man and cybern., 1997.27. (1): P143-148
36．张培强．MATLAB 语言．北京：中国科技大学出版社，1995．1
37．施阳，李俊．MATLAB 语言工具箱．西安：西安电子科技大学出版社，1999．9
38．范影乐，杨胜天等．MATLAB仿真应用详解．北京：人民邮电出版社，2001．7
39．楼顺天，胡昌华，张伟．基于MATLAB的系统分析与设计产—模糊系统．西安：西安电子科技大学出版社，2001．5
40．楼顺天，胡昌华，张伟．基于MATLAB的系统分析与设计—神经网络．西安：西安电
子科技大学出版社，2002．9
41．欧阳黎明．MATLAB控制系统设计—工程师工具软件应用系列．北京：国防工业大学出版社．2001．1
2．张寿柏，蔡柄初．微电子机械系统(MEMS)研究现状及展望，薄膜科学与技术，2001．8 (3)
3．李志坚．微电子机械系统(MEMS)发展展望．电子科技导报，1999．1
4. Hiroyuki Fujita, A decade of MEMS and its future, IEEE 1997,0-7803-3744-1\97, P1-8
5．裴晓亮．微型机器人的发展动向．机器人技术及应用．1997 (4)：5～7
6．朱涛，谈大龙．微操作机器人系统及其关键技术研究．组合机床及自动化加工技术，2001 (11)：33～36
7．蔡昌骏，李彦明．承压管道检测机器人的多通讯系统．机械电子．2003 (4)：63～67
8. Toshio FUKUTA, Microrobot and application, IEEE 1993,0-7803-0985-5/93.P1-8
9．李人厚．智能控制理论与方法．西安：西安电子科技大学出版社，1999．10
10．权太范．信息融合 神经网络—模糊推理理论与应用．北京：国防工业出版社，2002．8
11．汤兵勇．模糊控制理论与应用技术．北京：清华大学出版社，2002．9
12．傅京孙，蔡自兴，徐光佑．人工智能及其应用(第二版)．北京：清华大学出版社，2000．5
13. Bj.Haug, S.C.Wu, Dynamics of mechanical systems with coulomb friction striction, impact and constraint addition-deletion-Ⅰ. Theory. Mech. mach. Theory,21.1996.P401-406
14. Cheng Lianglun,Yang Yimin. Study on In-pipe Micro Robot Dynamics and Intelligent Control, Pioneering international Symposium on Motion and Vibration Control in Mechatronics, April 6-7,1999, Tokyo, Japan.
15. J.Trinkle, J.Pang, S.Sudarky, and G.Lo. On dynamic multi-rigid-body contract problem with coulomb friction, ZAMM, Vol.77, 1997. P267～269
16. A.P. Ambler, R.J.Popplestone. Inferring the Position of Bodies from Specified Spatial Relationships, Artificial intelligence, 6(2), 1975.P157～174
17. R. Fealhustone, The dynamics of rigid body systems with multiple concurrent contacts. Robotics Reasearchm, The Third Int.Sym., Eds., O.D.Faugeras and G.Giralt 1986
18. I.Han and B.J.Gilmore. Multi-body impact motion with friction-analysis, simulation and experiment validation. Proc. ASME Design Technical conf. 1990
19．张东明，竺长安等．电磁驱动式微型机器人的移动机理和设计．机器人．1995 (17)：P223～228
20．孙麟治，孙萍等．细小管道内爬行的微机器人．光学精密工程．1995．05
21．程良伦，杨宜民．微管道机器人的管内运动稳定性分析．机器人，1999．07
22．R．E 奥马利著，江福汝，尚汉翼等译．奇异摄动引论．北京：科学出版社，1988．3
23．毕学涛编．高等动力学，天津大学出版社
24．蔡自兴．机器人学．北京：清华大学出版社，2001．9
25．李士勇等．模糊控制与智能控制理论与应用．哈尔滨：哈尔滨工业大学出版社，1997．7
26．罗志增，蒋静坪．机器人感觉与多信息融合．北京：机械工业出版社，2002．7
27．孙增圻等．智能控制理论与技术，北京：清华大学出版社，1997．4
28．王士同编．神经模糊系统及其应用，北京：北京航空航天大学出版社，1998．5
29. Lin C T and Lin C S C. Reinforcement structure/ parameter learning for Neural-network-based fuzzy logic control system. IEEE Trans. Fuzzy Systems, 1994,2 (1): P46-63
30. I.F. Chung. A Reinforcement Neuro-Fuzzy Combiner for Multiobjective Control. Accepted for publication to IEEE Transactions on Systems, Man, and Cybernetics, Part B, 1998.
31. C. T. Lin and C.P.Jou. Controlling Chaos by GA-based Reinforcement Learning Neural Network. Accepted for publication to IEEE Transactions on Neural Network. 1998.
32. J.H.Lai. Application of neural Fuzzy Network to Pyrometer Correction and Temperature Control in Rapid Thermal Processing. Accepted for publication to IEEE Transactions on Fuzzy Systems, 1998
33. C.F. Juang. Genetic Reinforcement Learning Through Symbiotic Evolution for Fuzzy Controller Design. Accepted for publication to IEEE Transactions on Systems, Man, and Cybernetics, Part B, 1998.
34. Qiao L et al. Self-supervised learning algorithm of environment recognition in driving vehicle. IEEE Trans. Syst., Man and cybern., 1996.26. (6): P843-850
35. Chen S B, Wu L and Wang Q L. Self-learning fuzzy neural network for control of uncertain systems with time delays. IEEE Trans. Sys., Man and cybern., 1997.27. (1): P143-148
36．张培强．MATLAB 语言．北京：中国科技大学出版社，1995．1
37．施阳，李俊．MATLAB 语言工具箱．西安：西安电子科技大学出版社，1999．9
38．范影乐，杨胜天等．MATLAB仿真应用详解．北京：人民邮电出版社，2001．7
39．楼顺天，胡昌华，张伟．基于MATLAB的系统分析与设计产—模糊系统．西安：西安电子科技大学出版社，2001．5
40．楼顺天，胡昌华，张伟．基于MATLAB的系统分析与设计—神经网络．西安：西安电
子科技大学出版社，2002．9
41．欧阳黎明．MATLAB控制系统设计—工程师工具软件应用系列．北京：国防工业大学出版社．2001．1