网络遥操作机器人系统稳定性分析和规划方法研究
文献类型：学位论文
| 作者 | 景兴建 |
| 学位类别 | 博士 |
| 答辩日期 | 2005-01-23 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 中国科学院沈阳自动化研究所 |
| 导师 | 王越超 ; 谈大龙 |
| 关键词 | 机器人 时滞系统 稳定性 人工协调场 行为动力学 |
| 其他题名 | Stability and Planning of Internet-Based Tele-Robotic Systems |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 伴随着互联网技术在人们日常生产和生活中的深入发展，近年来，在控制界和机器人领域也悄然兴起一场新的研究热潮。基于网络的机器人遥操作系统成为了控制专家和机器人学学者共同的研究热点。但和其诱人的应用前景和潜在价值相比，由于网络引入而带来的系统实现问题则是一道必须克服的难关。除了大规模的分布式系统的实现本身所存在的各种问题外，还要克服网络任意时变时延、丢包以及乱序等现象所带来的稳定性、透明度以及控制设计等问题。现有的控制理论和方法面对这些问题是相形见绌的，新的理论或方法有待进一步开发和研究。从现有的一些结果看，为了提高系统的安全性以及操作性能，提高从端机器人的自主性是目前的主要技术手段之一。在对现有的解决方法和理论的优缺点进行深入分析的基础上，本文提出了主端操作与从端自主性相结合的控制系统设计思想。从网络对系统的影响看，丢包和乱序等问题都可转化为时延问题，所以研究随机时延对系统的影响是系统分析和设计的关键。而机器人自主性则是解决网络遥操作系统稳定性和透明度的有效手段。基于这一思想，本文深入地分析了两方面基本问题，即网络时变时延对系统稳定性的影响和机器人自主运动规划问题。在时滞系统的稳定性分析方面，针对任意时变时延的情况，本文分别对系统在存在不确定性和扰动情况下的指数稳定性、渐近稳定性以及一致稳定性等进行了研究，得到了一系列保守性小、计算简单的稳定性条件。对一些结果比如Razumikhin条件等进行了有效地扩展和推广，使得其保守性更小而且更方便应用。本文还提出了一组新的Krasovskii范函，利用这些新的范函，所得结果保守性大大降低，并为基于Lyapunov-Krasovskii定理的方法处理任意时变时延的问题提供了一种新的分析方法，而传统该类方法一般只能处理时延变化率小于1的情况。在动态不确定环境下机器人的自主运动规划方面，本文针对传统人工势场法所存在的不足，从矢量场的角度提出了一个新的更具一般意义的人工协调场。在二维局部坐标系下，人工协调场具有二维正交力分量，改变了传统的人工势场简单的“排斥－吸引”作用模式，丰富了人工场在任务空间内的信息表达能力，提高了行为决策的鲁棒性和准确性。人工协调场可更好地设计与机器人的运动状态和环境风险度相关，从而有效地克服传统人工势场的不足。从实际分析看，传统人工势场只是人工协调场的一个特例，而协调场是人工势场方法的一个有意义的发展。通过对机器人与环境交互行为的建模和分析，本文还提出了一种基于行为动力学的规划控制方法。基于行为动力学的最优控制，规划问题转化为在机器人加速度空间内的一个优化问题。最后，对于网络机器人控制系统的设计，给出了几个镇定控制器的设计结果。同时，本文着重讨论了遥操作系统主端操作命令与从端自主规划相结合的问题，给出了两种设计方法。分析结果验证了，自主性的提高对机器人遥操作系统的稳定性和操作性能的提高都是极其有利的。通过本文对基于网络的机器人系统一些基本问题的深入探讨，为今后进一步的研究工作打下了坚实的基础。 |
| 索取号 | TP242/J74/2005 |
| 英文摘要 | The internet is introduced into control systems, and thus being formed into larger and more efficient systems as so called Networked Control Systems or Internet-Based Tele-Robotic Systems. Such systems have great challenges in both application and theory. However, we have to deal with the stability and control issues induced by the time-varying communication delays, packet dropout, and asynchrony, etc. The conventional control theory and methods may not be directly used. In order to guarantee the stability and performance of internet-based tele-robotic systems, improving the autonomy of the slave robot is one of the most effective techniques used in many existing methods. With respect to the advantages and disadvantages of the existing results, we adopted the idea that teleoperation is combined with the slave autonomy in the construction of the control system of the tele-robotic system. Hence, in order to realize a tele-robotic system via internet, great efforts are made in two fundamental problems in this paper, that is, stability of the time-delay systems and motion-planning problems of tele-robots. For the stability of time-delay systems, the time-delay is assumed to be arbitrary time-varying. The robust exponential stability, asymptotical stability and uniform bounded stability of a class of uncertain time-delay systems are studied respectively. And some new stability conditions are derived. Some existing results such as Razumikhn condition are extended to less conservative forms, and some novel Lyapunov-Krasovskii functional are also proposed, which directly lead to some new less conservative stability conditions formulated into LMIs and also provide new approaches to the stability and control problems of autonomous systems with arbitrary time-varying delays. Note that the conventional Lyapunov-Krasovskii theorem based methods can only deal with the time delay with its derivative less than 1. As for the motion-planning problems of mobile robots in uncertain dynamic environments, in order to overcome the drawbacks of the conventional artificial potential field based methods, a new improved artificial potential field is proposed---artificial coordinating field. The artificial coordinating field can be easily used in more complex environments without local minima under some less conservative assumptions on the environments. The obstacles can be moving, unknown and U-shape. Cooperative collision avoidance and coordination between multiple mobile robots can also be easily realized. The conventional artificial potential field is just a special case of the artificial coordinating field. Moreover, another new approach is also proposed, which is based on the behavior dynamics of the mobile robots. The behavior dynamics is obtained by modeling the dynamic process of the interaction between the robot and its local environment. It provides a unique insight to the motion planning problem from the control of the behavior dynamics point of view. Finally, some stabilization controllers are proposed based on the networked control system model for the internet-tele-robotic systems using LMIs. The theoretic analysis indicates that the slave autonomy does help to improve the stability and transparency of the whole tele-robotic system. The results presented in this paper make a solid theoretic basis for our further research. |
| 语种 | 中文 |
| 公开日期 | 2012-08-29 |
| 产权排序 | 1 |
| 分类号 | TP242 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/9533]  |
| 专题 | 沈阳自动化研究所_机器人学研究室 |
| 推荐引用方式 GB/T 7714 | 景兴建. 网络遥操作机器人系统稳定性分析和规划方法研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2005. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。