用户名: 密码: 验证码:
移动机器人轨迹跟踪模糊变结构控制的研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
一.问题的提出
    移动机器人是机器人学科的一个重要分支,移动机器人研究是20世纪80年代以后兴起的一个比较新的研究领域。随着计算机、传感器、控制等领域的技术进步,移动机器人的发展也进入了一个更高的阶段。移动机器人通常指的是地面可移动机器人,主要应用在军事领域和民用服务领域。另外从更广的意义上来说,空间探测机器人和水下机器人也属于移动机器人的范畴。随着时代的发展,移动机器人必将会更加贴近我们的生活,同时在军事中也将发挥更大的作用,因此对这一课题的研究有着深刻的实际意义。但移动机器人在工作过程中必须要面对外界大量的不确定性信息的存在,如何提高移动机器人控制系统的鲁棒性,使得移动机器人在复杂的外界环境中能够快速准确地进行位姿的调整,精确地实现轨迹跟踪成为移动机器人发展中的一项关键技术。本文的目的是寻求一种易于实时硬件实现的强鲁棒性控制算法,使移动机器人在规划好的路径上能实现快速、精确的轨迹跟踪。
    二.移动机器人的整体结构
    移动机器人整体结构由三大部分组成:感知机构,控制机构,移动机构。
    感知机构 本移动机器人采用视觉导航,感知机构主要由图像采集单元,A/D转换单元和图像处理单元组成。
    控制机构 采用单片机为控制算法的核心处理单元,直流伺服电机为动力驱动装置实时调整移动机器人的位姿。
    移动机构 采用两驱动轮加前后万向轮的轮式移动机构,速度传感器安装在车轮轴上实时检测移动机器人的运行速度。
    三.研究难点
    1.移动机器人是一个十分复杂的多输入多输出非线性系统,它具有时变、强耦合和非线性的动力学特征,其控制是十分复杂的。由于测量和建模的不精确,再加上负载的变化以及外部扰动的影响,实际上无法得到机器人精确、完整的运动模型,我们必须面对机器人大量不确定性因素的存在。
    
    2.轮式移动机器人需要考虑与外界环境的滚动接触等非完整约束因素,实际上属于一个非完整约束系统。这类系统由于以下的一些新特性,所以其控制问题也变得非常复杂。
    1).非完整约束本质上是动态约束,由于不能通过积分等运算将其转化为简单的代数约束方程,使得其控制和运动规划等问题变得相当困难。
    2).非完整约束系统不能采用光滑或连续的纯状态反馈实现整体的精确线性化,在光滑连续的纯状态反馈下不能实现平衡点的渐进稳定。因此目前大多数工作主要基于构造时变连续状态反馈及非连续状态反馈控制器。这类方法不仅要求准确的运动学或动力学模型,而且衰减或收敛速度也较慢 ,因此用于实时控制时往往效果较差。
    四.本文所作的研究
    本文针对以上分析的难点并结合本移动机器人实验装置的特点做了如下一些研究:
    1.由于非完整移动机器人系统不能采用光滑或时不变反馈控制来实现其渐进运动规划和跟踪,我们采用非连续的滑模变结构控制。滑模变结构控制的最大优点就在于它能够克服系统的不确定性,对干扰和未建模动态具有很强的鲁棒性。缺点是用于补偿干扰和未建模动态的高控制增益和在滑动面附近控制行为的高频转换会产生抖振现象。为此我们设计了模糊变结构控制器,引入模糊逻辑规则来削弱抖振的不利影响。模糊变结构控制器是一种混合控制器,兼有模糊控制和变结构控制的优点;既解决了模糊控制系统的稳定性和鲁棒性问题,又削弱了变结构控制中的抖振。其基本方法是在变结构控制系统的趋近阶段通过模糊逻辑调整控制作用来补偿未建模动力学的影响。由于不确定性直接影响状态轨迹的趋近速度,这就使得趋近速度也带有一定程度的不确定性,所以我们在设计模糊逻辑规则时选择系统状态和趋近速度两个变量作为模糊控制器的输入,使得系统轨迹既能快速趋近滑动面又能降低抖振,从而提高了变结构控制系统的品质。
    2.建立了包含伺服电机在内的系统动力学模型,尽可能减小系统的未建模动力学参数引起的误差。由于非完整约束系统不能采用光滑或连续的纯状态反馈实现整体的精确线性化,我们结合本移动机器人实验装置特点,通过合适地选取系统的输出,实现了动力学模型的输入/输出精确线性化,在此基础上设计了模糊变结构控制器。
    3.针对移动机器人实验装置建立了运动学模型,确定了移动机器人的位姿和控制参数之间的关系。传统的方法采用分别建立伺服电机驱动电压和两
    
    
    驱动轮转速之间的关系来调整移动机器人的位姿,这样将会产生耦合。本文直接建立起移动机器人的方向角和两伺服电机的输入差动电压之间的关系,去除了耦合,提高了控制的精度,降低了程序设计的复杂性。
    4.离线状态下建立了伺服驱动电机额定负载下的电压-转速特性曲线,经过换算进而得到脉宽调制系数-电压-转速之间的函数关系,大大简化了程序的设计,提高了控制的实时性。
    5.硬件电路设计上采用带有两路脉宽调制端口和看门狗电路的微处理器P80C552作为控制算法的核心单元,提高了处理的实时性和可靠性。同时采用一片大容量、可编程的芯片PSD834F2取代了传统电路设计中必须用到的数据存储器、程序存储器、锁存器等器件,简化了硬件电路的设计量,减小了电路板的体积。
1. Put forward of the question
    The mobile robot is an important branch of robotics, and the study of the mobile robot is a new research field which is raised after the 80’s of last century. With the advancements of the fields of computer, sensor and control,the developments of the mobile robot also enter into a higher stage. In general, the mobile robot refers to the robot moving on the ground, and it mainly includes the mobile robot which is used in military affairs and in our daily life. In addition, from wider meaning space exploration robot and under-water robot also belong to the category of the mobile robots. With the development of times, it must press close to our life even more, and it also will play a bigger role too in the military affairs at the same time. So there is a deep actual meaning on this subject’s research. But the mobile robot must face the existence of a large amount of uncertain information of external world during the course of its job, so it becomes a key technique of how to enhance the robustness of mobile robots’ control system, which make it can carry on the adjustment of the location accurately and fast and the tracking trajectory accurately in the complicated external environment. The purpose of the paper is to seek a kind of control algorithms with great robustness which is easily realized through real-time hardware, and it can make the mobile robot to track the layout trajectory accurately and fast.
    2. The framework of the mobile robot
    The whole structure of the mobile robot is made up of three major parts: sense organization, control organization, moving organization.
    (1) Sense organization: The mobile robot uses vision navigation, and the sense organization is mainly made up of the image gathered unit, A/D conversion unit and image processing unit.
    (2) Control organization: The mobile robot uses single-chip microcomputer as the core processing unit of the control algorithms, direct current servomotor as the power drive setting to adjust the state of the mobile robot.
    (3) Moving organization: The mobile robot uses the wheeled structure of two
    
    
    driving wheel and universal wheel. Speed sensor is fixed on the wheel axes to detect the real-time speed of the mobile robots.
    3. Difficulties of the study
    (1) The mobile robot is a very complexly multi-input and multi-output non-linear system, and it has the dynamics characteristic of time variant, coupling and non-linear. So it is complicated to control. Because of the imprecision of measuring and modeling and the influence of changed-load and outside flexibility, in fact we are unable to get the model of the mobile robot accurately and integrally, so we must face the existence of a large number of uncertain factors of the mobile robot.
    (2) The nonholonomic constraints factor such as the roll-contact with outside environment needs to be considered, so in fact the mobile robot belongs to a kind of nonholonomic constraints system. It is very complicated to control this kind of systems because of the following some new characteristic.
    1) Essentially, nonholonomic constraints are dynamic constraints, so we are unable to change it into simple algebra constraints equation by integral operation, which make it very difficult to control and locomotion layout.
    2) Nonholonomic constraints system can’t realize linearization exactly and asymptotical stability of balance point by sliding or continuous pure state feedback. So most of the works are mainly based on constructing of sliding, continuous pure state feedback and discontinuous state feedback controller. This method need the exact kinematics model and dynamics model, and the speed of attenuate is slow, so it has undesirable effect to be used on real-time control.
    4. Research of the paper
    To the question that we have analyzed above, combining with the characteristic of our experimental apparatus we make some research as follows:
    (1) Because the nonholonomic mobile robot is unable be controlled through sliding and continuous pure state feedback, we use discontinuous sli
引文
[1] 王丰尧,滑模变结构控制,机械工业出版社,1995
    [2] 姚琼荟,黄继起,变结构控制系统,重庆大学出版社,1997
    [3] 胡跃明,变结构控制理论与应用,科学出版社,2003
    [4] 高为炳,变结构控制的理论及设计方法,科学出版社,1996
    [5] 王顺晃,智能控制系统及其应用,机械工业出版社,1999
    [6] 王灏,毛宗源,机器人的智能控制方法,国防工业出版社,2002
    [7] 郑堤,唐可洪,机电一体化设计基础,机械工业出版社 1997
    [8] 诸静,模糊控制原理与应用,机械工业出版社,1998
    [9] 胡寿松,自动控制原理(第3版),国防工业大学出版社 1998
    [10] 邓星钟,机电传动控制,华中理工大学出版社,1998
    [11] 范影乐,MATLAB仿真应用详解,人民邮电出版社,2001
    [12] 张国良,曾静,模糊控制及其MATLAB应用,西安交通大学出版社,2002
    [13] 闻新,周露,MATLAB模糊逻辑工具箱的分析与应用,科学出版社,2001
    [14] 张友德,菲利浦80C51系列单片机原理与应用技术手册,北京航空航天大学出版社,1992
    [15] 李华, MSC-51系列单片机实接口技术用,北京航空航天大学出版社2000
    [16] 张毅刚,彭喜源,MCS-51单片机应用设计,哈尔滨工业大学出版社,1999
    [17] 吴金戌,8051单片机实践与应用,清华大学出版社,2002
    [18] 单片机C程序设计及应用实例,人民邮电出版社,2003
    [19] 潘新民,单片微型计算机实用系统设计,人民邮电出版社,1992
    [20] 陈汝全,实用微机与单片机控制技术,电子科技大学出版社,2002
    [21] 王晓明,电动机的单片机控制,北京航空航天大学出版社,2002
    [22] 范逸之,江文贤,C++ Builder与RS-232串行通信控制,清华大学出版社,2002
    [23] 可编程单片机外围器件PSD8XXF系列数据手册及应用笔记,武汉力源电子股份有限公司,1998
    [24] 孙涵芳,单片机现场可编程外围芯片PSD的原理及应用,北京航空航天大学出版社,1998
    
    [25] 阎石,数字电子技术基础,高等教育出版社,1998
    [26] 夏宇闻,Verilog数字系统设计教程,北京航空航天大学出版社,2003
    [27] Donald E.Thomas ,硬件描述语言Verilog,清华大学出版社,2001
    [28] 赵雅兴,FPGA原理、设计与应用,天津大学出版社,2002
    [29] 张大鹏,模式识别与图象处理并行计算机系统设计,哈尔滨工业大学出版社,1998
    [30] 章晋,图象处理和分析 清华大学出版社 2000
    [31] 徐守时,信号与系统理论、方法和应用,中国科学技术大学出版社,1999
    [32] Rafael C.Gonzalez ,Richard E.Woods Digital Image Processing(Second Edition),电子工业出版社,2003
    [33] Young K-KD. Controller Design for A Manipulator Using Theory of Variable Structure Systems. IEEE Trans,SystMan and Cyber. 1 978,8(2 ):2 1 0 - 1 2 8.
    [34] Yury Stepanenko,Yong Cao,Chun-Yi Su. Variable Structure Control of Robotic Manipulator with PID Sliding Surface.Int JOf Robust And Nonlinear Control,1 998,(8):79- 90
    [35] Yeung K S,Chen Y P. A New Controller Design for Manipulators Using the Theory of Variable Structure System.IEEE Trans,1988,35(2 ):995- 1003.
    [36] Chen Y- F,Mita T,Wakui S. A New and Simple Algorithm for Sliding Mode Trajectory Control of the Robot Arm.IEEE Trans,1990 ,35(7):828- 829.
    [37] Man et al. A Robust MIMO Terminal Sliding Mode Control Scheme for Rigid Robotic Manipulators. IEEE Trans,on Autom Contr,1 994,39(8),2462-2469 .
    [38] Slotine J- JE. The Robust Control of Robot Manipulators. Int JRob Res,1 985,(4):49- 54.
    [39] Dong WJ and Huo W. Quasi-exponential stabilization of wheedled mobile robots with uncertainty[A]. Preprint of 14th Triennial World Congress[C]
    
    
    1999 .
    [40] Luo Z H, Machida K and Funaki M. Design and contol of a Cartesian nonholononic robot [J].J .Robotic Syst.1998.
    [41] Laumoikl J P .Robot Motion Planning and Control [M] .London:Springer Verlag,1998
    [42] Li Q X, Hu Y M and Zhou Q J. Robust output tacking of mobile robots [J]. Control Theory and Applications,1998, 15(4) :515-524.
    [43] Chung J H ,Velinsky S A. Motion control of a mobile manipulator by input output linearization [A] .In ASME International Mechanical Engineering Congress and Exposition Dynamic Systems and Control Division[C] ,Dallas,
    TX,USA ,1997 ,481-486.
    [44] Camrlas de Wit C,Siciliano B and Bastin G. Throry of Robot Control[M].
    London: Springer Verlag,1998.
    [45] Hu Y M, Zhou Q J and Pei H L. Theory and application of nonholonomic control systems [J]. Control Theory and Applications,1996,13(1):1-10.
    [46] Kolmanovesky and McClamroch N H. Development in nonholonomic control problems [J]. IEEE Control Systems,1995,15,(12):20-36.
    [47] Krishnan H and McCiamroch N H. Tracking in nonlinear sifferential algebraic control systems with applications to constrained robot systems[J]. Automatica,1994,30(12):1885-1897.
    [48] Yu W L, Sobel K M. Robust Elgenstructure Assignment Structure State Space Uncertainty. Control and Dynamics,1991,14(3):621~628.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700