混合驱动式多关节双机械臂运动控制研究

发布时间：2018-12-28 15:28

【摘要】：气动人工肌肉是一种新型的拉伸型气动执行元件,具有重量轻、输出力-直径比大、柔顺性好、类似生物肌肉的特点,在仿生、康复等技术领域中展现出了良好的应用前景。本论文基于机器人学和现代仿生学,仿照人体上肢结构,设计了一种基于气动人工肌肉和步进电机混合驱动的多关节双机械臂,为气动人工肌肉在仿生机器人中应用探路。本论文主要完成了多关节双机械臂的机构设计、运动学与动力学建模、控制系统软硬件设计、系统建模仿真以及运动控制实验研究与分析。首先仿照人体手臂,完成了仿人多关节双机械臂的机构设计。使用气动人工肌肉驱动机械臂的肩、肘关节,步进电机驱动辅助运动关节,所设计的机械臂结构上兼具柔顺性和紧凑性;同时,根据机器人学,完成了多关节双机械臂结构的运动学和动力学建模,包括完备的位姿方程正、逆解,速度雅可比矩阵和动力学方程。然后,在前人研究的基础上,结合多关节双机械臂动力学方程,借助Matlab/Simulink工具,完成系统模型的搭建,并在此基础上研究比较了常规PID和模糊自适应整定PID对肩关节的控制效果,完成了两种控制策略的仿真分析。仿真结果表明模糊自适应整定PID能够有效抑制关节角度偏差,提高响应速度和系统稳定性。最后,完成多关节双机械臂整套样机的软硬件开发,并在此样机的基础上进行了运动控制实验研究。通过实验,分析比较了常规PID和模糊自适应整定PID的控制效果。实验证明了模糊自适应整定PID控制下,多关节双机械臂串行、并行运动以及协调运动时,每个关节均具有较好定位精度和响应速度。本论文实验研究为仿生机器人的进一步研究提供了实践参考。
[Abstract]:Pneumatic artificial muscle is a new type of stretching pneumatic actuator with light weight, large ratio of output force to diameter, good flexibility and similar to biological muscle. It has shown a good application prospect in bionics, rehabilitation and other technical fields. In this paper, based on robotics and modern bionics, and imitating the structure of human upper limb, a kind of multi-joint double manipulator driven by pneumatic artificial muscle and stepping motor is designed, which can be used to explore the application of pneumatic artificial muscle in bionic robot. In this paper, the mechanism design, kinematics and dynamics modeling, hardware and software design of control system, system modeling and simulation, and the research and analysis of motion control experiment are completed. At first, the mechanism design of humanoid multi-joint double manipulator is completed, which mimics the human arm. The pneumatic artificial muscle is used to drive the shoulder and elbow joints of the manipulator and the stepper motor to drive the auxiliary motion joints. The designed arm has both flexibility and compactness in structure. At the same time, the kinematics and dynamics modeling of the multi-joint dual-manipulator structure is completed according to robotics, including the complete forward and inverse solutions of the pose equation, the velocity Jacobian matrix and the dynamic equation. Then, on the basis of previous studies, combined with the dynamics equation of multi-joint and dual-manipulator, the system model was built by Matlab/Simulink tool. On this basis, the control effect of conventional PID and fuzzy adaptive tuning PID on shoulder joint is studied and compared, and the simulation analysis of two control strategies is completed. The simulation results show that the fuzzy adaptive tuning PID can effectively suppress the angle deviation of joints and improve the response speed and system stability. Finally, the software and hardware development of the whole prototype is completed, and the motion control experiment is carried out on the basis of the prototype. Through experiments, the control effects of conventional PID and fuzzy adaptive tuning PID are analyzed and compared. It is proved by experiments that each joint has better positioning accuracy and response speed under fuzzy adaptive tuning PID control when multi-joint dual manipulator moves in serial parallel and coordinated motion. The experimental research in this paper provides a practical reference for the further research of bionic robot.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH138.5

【参考文献】