基于接触力-姿态模型的机器人非刚体轴孔装配偏角预测方法  

Deflection angle prediction method based on contact force attitude model for robot non-rigid body shaft hole assembly

作　　者：胥芳[1] 沈一丰 陈教料[1] 王佳才 XU Fang;SHEN Yifeng;CHEN Jiaoliao;WANG Jiacai(College of Mechanical Engineering,Zhejiang University of Technology,Hangzhou 310014)

机构地区：[1]浙江工业大学机械工程学院,杭州310014

出　　处：《高技术通讯》2022年第11期1188-1195,共8页Chinese High Technology Letters

基　　金：国家重点研发计划(2018YFB1309404);浙江省公益技术研究计划(LGG18E050023)资助项目。

摘　　要：针对机器人非刚体轴孔装配过程中的偏差角识别问题,提出了一种基于接触力-姿态模型的装配偏角预测方法。对非刚体轴孔装配过程中不同的接触状态进行静力学理论推导和弹性形变计算,进而建立接触力-姿态模型。在此基础上,结合动态罚函数和灰狼算法对模型进行可行域内的寻优求解,得到装配偏角。通过机器人的非刚体轴孔装配实验,采集不同接触状态和装配偏角下的力觉数据,并将其代入接触力-姿态模型得到预测偏角。结果表明,装配偏角在0~10°的情况下,本文所提方法的非刚体轴孔装配偏角预测误差在0.5°以内,可为后续的成功装配提供有效的工件装配姿态信息。Aiming at the identification of deviation angle in the robotic assembly of non-rigid shaft and hole,an assembly deviation prediction method based on contact force attitude model is proposed.According to derivation of statics theory and calculation of elastic deformation,a contact force attitude model is developed in case of the different contact states during the assembly of non-rigid shaft and hole.A grey wolf algorithm with dynamic penalty function is proposed to solve the model in the feasibility domain to obtain the deviation angle using the force-attitude model.Robotic assembly experiments of non-rigid shaft and hole with various deviation angle and different contact state are carried out to collect the force sensor data,and the deviation angles of the non-rigid shaft-hole robotic assemblies can be obtained via contact force attitude model and the grey wolf algorithm with dynamic penalty function.The results show that the error of the predicted deviation angle is less than 0.5°in case of the actual deviation angle in the range of 0°to 10°in the assembly,which proves the proposed method can effectively predict assembly posture of the workpiece for subsequent successful assembly.

关 键 词：机器人装配 非刚体轴 弹性形变 接触力-姿态模型 动态罚函数 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TG95[自动化与计算机技术—控制科学与工程]