应用草原放牧算法的机器人动态参数迭代辨识  

Iterative identification of robot dynamic parameters using grassland grazing algorithm

作　　者：吴鸿宇 库祥臣[1] WU Hongyu;KU Xiangchen(School of Mechatronics Engineering,Henan University of Science and Technology,Luoyang 471003,China)

机构地区：[1]河南科技大学机电工程学院,河南洛阳471003

出　　处：《河南科技学院学报（自然科学版）》2024年第4期56-65,共10页Journal of Henan Institute of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金(51675161);河南省重大科技专项(171100210300)。

摘　　要：准确的动态模型参数是机器人导纳控制、轨迹跟踪和力控制的基础.常用的最小二乘法难以满足动态参数辨识的高精度要求.应用草原放牧算法,通过迭代更新各采样点的数据权重,能够避免观测数据中的异常值干扰辨识计算.同时,为充分激发机器人各连杆的动态特性,采用fmincon函数优化参数辨识实验所用的激励轨迹,并设计对照实验以探究激励轨迹优化对参数辨识的影响.辨识方法在六自由度机器人身上得到实验验证:相比加权最小二乘法,运用草原放牧算法辨识的动态参数来预测机器人各关节的驱动力矩,残差平均降低18.12%.这为机器人的力矩跟踪控制提供了一种新思路.The accurate dynamic model parameters are essential for robot admittance control,trajectory tracking,and force control.However,the commonly used least square method fails to meet the high precision requirements of dynamic parameter identification.In this paper,we propose using the grassland grazing algorithm to iteratively update the weights of each sampling point in order to avoid outliers in the observation data.Additionally,we utilize the fmincon function to optimize the excitation trajectory used in the parameter identification experiment,aiming to fully stimulate the dynamic characteristics of each connecting rod of the robot.A control experiment is designed to explore how optimizing the excitation trajectory influences parameter identification.Experimental results on a 6-DOF robot validate our identification method:compared with weighted least square method,our proposed method reduces average residual error by 18.12%when predicting driving torque for each joint of the robot.This also provides a new approach for torque tracking control.

关 键 词：机器人 动态参数辨识 草原放牧算法 加权迭代 关节力矩预测 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]