受限空间内机器人在位测量位姿优化方法  

Optimization method for in-situ measurement postures of robots in restricted space

作　　者：宋慧鹏 王英鹏 孙玉文[1,2] SONG Huipeng;WANG Yingpeng;SUN Yuwen(School of Mechanical Engineering,Dalian University of Technology,Dalian 116024,Liaoning,China;National Key Laboratory of High Performance Precision Manufacturing,Dalian 116024,Liaoning,China)

机构地区：[1]大连理工大学机械工程学院,辽宁大连116024 [2]高性能精密制造全国重点实验室,辽宁大连116024

出　　处：《沈阳工业大学学报》2025年第2期137-144,共8页Journal of Shenyang University of Technology

基　　金：国家自然科学基金项目(U22A20202)。

摘　　要：【目的】随着现代制造业对加工精度和效率要求的不断提高,在位测量技术作为一种实时、精准的测量手段,正逐渐成为现代加工过程中不可或缺的一部分。工业机器人因其具有工作空间大、运动灵活性好和易于编程等特点,被广泛应用于复杂零件在位测量与加工过程中。然而受限于测量空间约束的复杂性及测量运动在机器人本体关节空间下的非线性传递表达,机器人末端沿给定路径测量时存在位姿不光顺,甚至发生与障碍物碰撞的情形。为解决该问题,提出了一种基于机器人测量系统冗余运动参数的位姿优化新方法。【方法】首先,建立机器人的正逆运动学模型,结合该模型计算得到测量运动与机器人关节空间的运动传递关系,基于机器人测量系统中绕着测量工具轴线转动的冗余运动参数,构造各路径点处满足关节角限位约束和位姿奇异性约束的机器人测量冗余姿态集合;其次,为避免机器人与障碍物发生碰撞,采用有向包围盒技术对大型复杂曲面构件测量场景进行建模,并基于GJK方法对机器人与障碍物进行碰撞快速检测,筛选出满足无碰撞约束的机器人在位测量姿态集合;最后,基于Dijkstra最短路径方法计算得到关节角变化最小的机器人测量姿态序列,提高了受限空间内机器人在位测量过程中的稳定性。【结果】对本文方法进行实验验证,采集测量过程中机器人关节角数据。结果表明优化后的机器人关节运动更加平稳,各个关节角的累计变化量得到降低,关节角波动幅值和标准差较优化前均显著减小,实现了在受限空间内对工件的平稳测量。【结论】基于机器人冗余运动参数的在位测量位姿优化方法可以在有效避免机器人与障碍物发生全局干涉的同时,显著提高测量过程中机器人位姿的光顺性。[Objective]In modern manufacturing,as the demand for machining accuracy and efficiency continues to rise,in-situ measurement technology,serving as a real-time and precise measurement method,is gradually becoming an indispensable part of modern machining processes.Industrial robots are widely applied in in-situ measurement and processing of complex parts due to their large work space,high motion flexibility,and ease of programming.However,limited by complicate measurement space constraints and the nonlinear transmission and expression of measurement motion in the joint space of the robots,they exhibit posturenon-smoothness at their ends and even suffer collisions during measurement along a predefined path.To tackle this issue,a posture optimization method based on the redundant motion parameter of the robot measurement system was proposed.[Methods]Firstly,the forward and inverse kinematics model of the robot was established,and the motion transmission relationship between the measurement motion and the joint space of the robot was calculated with this model.Considering that there existed a redundant motion parameter in the robot measurement system that rotates around the axis of the measuring tool,a set of redundant measurement postures satisfying the joint angle limitconstraint and posture singularity constraint at each path point were constructed.Subsequently,to avoid collision between the robot and obstacles,an oriented bounding box was employed to model themeasurement of large-size complex curved components,and then the Gilbert-Johnson-Keerthi(GJK)method was implemented for rapid collision detection between the robot and the obstacles to screen a set of robot in-situ measurement postures that meeting collision-free constraints.Finally,the Dijkstra′s shortest path technique was applied to determine the robot measurement posture sequence with the minimum joint angle variation,enhancing the stability of robots during in-site measurement in restricted space.[Results]The proposed method was verified by experiment

关 键 词：机器人 运动学 奇异性 在位测量 位姿优化 碰撞检测 冗余度 光顺 

分 类 号：TP241.2[自动化与计算机技术—检测技术与自动化装置]