基于模拟退火量子遗传算法的焊接机器人轨迹规划  

Simulated annealing quantum genetic algorithm for welding robot trajectory planning

作　　者：金宇杰 龚堰珏 赵罘[1] JIN Yujie;GONG Yanjue;ZHAO Fu(School of Artificial Intelligence,Beijing Technology and Business University,Beijing 100048,China)

机构地区：[1]北京工商大学人工智能学院,北京100048

出　　处：《现代制造工程》2024年第1期33-38,共6页Modern Manufacturing Engineering

基　　金：国家自然科学基金项目(51975006)。

摘　　要：针对焊接机器人在焊接过程中经常出现的轨迹规划问题,以六自由度机械臂PUMA 560为研究对象,采用笛卡尔空间规划轨迹,并利用混合算法,以关节惯性力矩变化量最小为优化目标,对不同末端位置的各个关节惯性力矩进行优化,从而消除了机械臂焊接过程中的运行不稳定、关节运动不平稳等问题。克服了模拟退火算法收敛速度慢及量子遗传算法局部寻优能力差等问题,成功规划出机械臂关节惯性力矩最优轨迹。MATLAB仿真结果表明,模拟退火量子遗传算法收敛时间相比传统遗传算法缩短30.56%,并且优化了关节惯性力矩,验证了该算法的可行性,可为后续研究奠定基础。For solving the trajectory planning problem that often occurs in the welding process of welding robots,a six-degree-of-freedom robotic arm PUMA 560 was taken as the research object,Cartesian space was used to plan the trajectory,and a hybrid algorithm was used to optimize the inertia moment of each joint at different end positions with the minimum amount of joint inertia moment variation as the optimization objective,thus the problems of unstable operation and unstable motion of the robot arm during welding were eliminated.The problems of slow convergence of the simulated annealing algorithm and poor local optimization capability of the quantum genetic algorithm were overcome,and the optimal trajectory of the joint inertia moments of the robot arm was successfully planed.MATLAB simulation results show that,the convergence time of the simulated annealing quantum genetic algorithm is reduced by 30.56%compared with the traditional genetic algorithm,and the inertia moment of the joint is optimized,which verifies the feasibility of the algorithm and lays the foundation for the subsequent research.

关 键 词：机械臂轨迹规划 模拟退火算法 量子遗传算法 惯性力矩 

分 类 号：TH164[机械工程—机械制造及自动化]