上海交通大学学报

上海交通大学学报（自然版） ›› 2019, Vol. 53 ›› Issue (3): 341-347.doi: 10.16183/j.cnki.jsjtu.2019.03.012

• 学报（中文） • 上一篇    下一篇

未知环境下水下机械手智能抓取的自适应阻抗控制

张建军，刘卫东，李乐，程瑞锋，郑海峰   

  1. 西北工业大学 航海学院， 西安 710072
  • 出版日期:2019-03-28 发布日期:2019-03-28
  • 通讯作者: 刘卫东，男，教授，博士生导师，电话(Tel.):029-88492193; E-mail: liuwd@nwpu.edu.cn.
  • 作者简介:张建军(1983-)，男，河南省周口市人，博士生，主要从事水下机械手控制及遥操作研究.
  • 基金资助:
    国家自然科学基金(61473224)，国家重点研发计划(2016YFC0301700)，高等学校学科创新引智计划(B18041)资助项目

Adaptive Impedance Control for Underwater Manipulator Intelligent Grasping in Unknown Environment

ZHANG Jianjun,LIU Weidong,LI Le,CHENG Ruifeng,ZHENG Haifeng   

  1. School of Marine Science and Technology, Northwestern Polytechnical University, Xi’an 710072, China
  • Online:2019-03-28 Published:2019-03-28

PDF (PC)

374

摘要： 为了满足水下机械手在未知环境下对目标抓取的多样性，保证抓住、抓牢并最大限度地避免目标损伤，提出了自适应阻抗控制方法.构建了基于位置的阻抗模型的力跟踪控制系统，采用递推最小二乘法辨识目标的阻抗参数，根据阻抗参数与机械手的运动特征，利用模糊推理方法在线调整抓取力的期望值，并根据期望抓取力与实际抓取力的误差设计自适应比例-积分-微分(PID)控制器来调整期望位置，以实现在跟踪目标位置的同时对期望抓取力信号的跟踪，并利用MATLAB/Simulink软件平台进行仿真实验.结果表明：自适应阻抗控制方法在自由空间和约束空间均具有良好的力、位移的跟踪性能；对期望抓取力的实时调整满足抓取目标的多样性，期望位置的自适应调整能够实现对期望抓取力的跟踪.

关键词: 水下机械手, 阻抗模型, 自适应, 模糊推理, 递推最小二乘法

Abstract: In order to realize the diversity of the unknown environment target capture of underwater manipulator, the adaptive impedance control method is proposed to ensure grasping, holding and minimizing the damage of the target. A force tracking control system based on position impedance model is established, and the target impedance parameter is identified by recursive least square method, and the expected value of grasping force is adjusted online according to the impedance parameter and the manipulator motion characteristic with fuzzy inference. According to the expectation force and grasping force error, the adaptive PID controller adjusts the desired position, realizes the tracking of different expected force signals while tracking the target position, and realizes the simulation verification on the MATLAB/Simulink platform. The results show that the adaptive impedance control has good force and displacement tracking performance in free space and constrained space. The realtime expectation force adjustment satisfies the diversity of grasping target, and the adaptive adjustment of the desired position realizes the tracking of the desired force.

Key words: underwater manipulator, impedance model, adaptive, fuzzy inference, recursive least square method

中图分类号: