智能轮椅控制方法与控制系统研究
文献类型：学位论文
| 作者 | 朱智平 |
| 学位类别 | 工学博士 |
| 答辩日期 | 2010-05-30 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 中国科学院自动化研究所 |
| 导师 | 原魁 |
| 关键词 | 智能轮椅 轨迹跟踪 滑模控制 视觉伺服 路径规划 Intelligent Wheelchair Trajectory Tracking Sliding Mode Control Visual Servo Path Planning |
| 其他题名 | Research on Control Methods and Control System for Intelligent Wheelchair |
| 学位专业 | 控制理论与控制工程 |
| 中文摘要 | 随着社会经济的发展，人口寿命和伤残期的延长，如何提高老年人和残疾人健康护理，增强他们的生活自理能力已经成为全社会关注的热点问题。智能轮椅作为一种服务机器人，已经突破了传统轮椅和电动轮椅的概念，具有人机交互、自主避障、取物开门等更智能、更安全的功能。在国家863重点项目所属子课题“智能轮椅概念样机及目标识别技术的研究”的支持下，本文针对安装有机械臂智能轮椅的控制系统及其理论与算法进行了深入研究，论文的主要工作和创新之处如下： 第一，设计实现了安装机械臂智能轮椅的软、硬件系统，分析建立了其运动学和动力学模型；根据五自由度机械臂的结构特点，采用位置分解的思想，利用几何投影的位姿反解算法对机械臂的逆运动学进行求解。 第二，在智能轮椅的控制方面：针对智能轮椅的轨迹跟踪控制问题，综合考虑轮式轮椅移动平台的运动学约束和控制系统的驱动输入限制，提出了一种输入受限的Lyapunov直接法的轨迹跟踪算法；在串联机械臂的动力学模型中引入电机模型，利用Lyapunov直接法对机械臂系统PID控制的稳定性进行了分析。 第三，在底层伺服电机控制方面：1、提出一种自适应时变滑模面变结构伺服电机控制算法，通过实时调整滑模面减少趋近运动模态的时间从而加快误差的收敛速度；2、设计实现了基于动态滑模的驱动电机控制方法。通过仿真与实验，验证了算法的有效性和可行性。 最后，结合智能轮椅的双目视觉系统和“eye-in-hand”视觉系统结构，设计实现了智能轮椅完成目标抓取与开门任务的视觉控制策略和算法。给出了一种开门路径规划的优化准则函数，将开门路径规划问题转化为最优化问题；结合智能轮椅的运动学模型和优化准则函数，提出了一种基于几何方法的开门路径规划算法。 |
| 英文摘要 | With the social development, the problem of aging society is becoming more serious. How to enhance the healthcare and self-care ability of the elderly and the disabled had been paid more attention to. Intelligent wheelchairs, as a kind of service robots, have not only been a means of transportation like traditional wheelchairs or power wheelchairs, but provided many functions such as Human-Computer Interaction(HCI), intelligent obstacle avoidance, object grasping and door opening. Supported by the National High Technology Research and Development Programme of China “Research on Design of Intelligent Wheelchair Functional Prototype and Target Recognition Technique”, this thesis has studied on the control methods and control system of intelligent wheelchair mounted manipulator. The novel work and contributions of this thesis can be summarized as follows: Firstly, the hardware and software system of intelligent wheelchair mounted manipulator is designed. The kinematics and dynamics of intelligent wheelchair are built. Considering the manipulator architecture, we analyzed inverse kinematics of manipulator using a geometric approach. Secondly, by analyzing the trajectory tracking model of wheelchair, a new trajectory tracking controller of intelligent wheelchair with input constraints is proposed, which is based on Lyapunov’s direct method. A dynamic model including motor dynamics is developed for manipulator and the stability of PID controller is analyzed by using Lyaponuv’s direct method. Thirdly, two DC motor controllers are developed: (i) an adaptive time-varying sliding surface controller; and (ii) a dynamical sliding mode controller. The former controller can short the reaching phase by regulating the slope of the sliding surface. The feasibility of the methods is verified by simulation and experimental results. Finally, combining the binocular vision system and the “eye-in-hand” visual framework, the visual control strategy for intelligent wheelchair grasping object is proposed. By analyzing the path planning problem for opening door, we presented an optimization criterion function, by which the path planning can be transformed into optimization. Then, the opening door path planning algorithm based on geometric method is proposed. The experimental results show that the algorithms are effective. |
| 语种 | 中文 |
| 其他标识符 | 200718014628029 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/6265]  |
| 专题 | 毕业生_博士学位论文 |
| 推荐引用方式 GB/T 7714 | 朱智平. 智能轮椅控制方法与控制系统研究[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2010. |
入库方式： OAI收割
来源：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。