气动肌腱拮抗式仿生关节建模与运动控制  

Modeling and Motion Control of Antagonistic Bionic Joint Driven by Pneumatic Muscles Actuators

作　　者：艾宁义 杨冬[1] 李铁军[1] 马涛 AI Ning-yi;YANG Dong;LI Tie-jun;MA Tao(School of Mechanical Engineering, Hebei University of Technology, Tianjin 300132)

机构地区：[1]河北工业大学机械工程学院,天津300132

出　　处：《液压与气动》2020年第8期12-21,共10页Chinese Hydraulics & Pneumatics

基　　金：国家重点研发计划(2018YFB1306902);国家自然科学基金重点项目(U1813222);河北省自然科学基金(E2019202338)。

摘　　要：从仿生角度出发,基于人体手指关节的骨骼结构和肌肉发力特点,设计出一种气动肌腱驱动、腱传动的拮抗式仿生关节。针对气动肌腱的高度非线性特性,通过搭建实验平台,测试肌腱性能,采用最小二乘法建立单根气动肌腱的数学模型。基于该关节采用腱传动特点,建立关节力矩输出模型。建立关节位置/刚度解算模型,根据关节位置、关节刚度及关节输出力矩解算气动肌腱理论输入压力。采用模糊PID算法和传统PID算法实现对该关节位置闭环控制、关节刚度开环控制。实验结果表明,2种方法均能较好的跟随关节刚度,同时模糊PID算法的位置控制精度明显优于传统PID算法。From the perspective of bionics,based on the bone structure and the muscle strength characteristics of human finger joints,a bionic joint is designed.This joint is based on the pneumatic muscles actuators actuation and the tendon transmission.Due to the highly nonlinear characteristics of the pneumatic muscles actuators,the performance test is carried out on the experiment platform.the mathematical model of a single pneumatic muscles actuator is established by using the least square method.Based on the characteristics of the tendon transmission,the torque output model of the joint is established.According to joint position,joint stiffness and joint driving torque,joint position/stiffness's calculating model is built to compute theoretical pressure of pneumatic muscles actuators.The fuzzy PID algorithm and the traditional PID algorithm are adopted.Both of algorithms realize the closed-loop control of joint position and the open-loop control of joint stiffness.The experimental results show that both of algorithms are able to follow the joint stiffness well.the position control precision of the fuzzy PID algorithm is obviously superior to that of the traditional PID algorithm.

关 键 词：气动肌腱 腱传动 位置/刚度控制 模糊PID 试验验证 

分 类 号：TH138[机械工程—机械制造及自动化] TP241[自动化与计算机技术—检测技术与自动化装置]