改进差分进化算法优化的机器人时间最优轨迹规划算法

摘要

协作型工业机器人在执行焊接、装配任务时,因关节状态无法达到最大约束而影响效率.在确保"人-机-环"安全的前提下,为最大限度地提高机器人的操作速度与响应时间,提出了一种改进差分进化(DE)算法优化的协作型机器人轨迹规划算法.采用蒙特卡洛算法获得机器人的操作空间,并采用广义动量方法进行碰撞检测.同时,以机器人在执行任务过程中运行时间最小为目标,兼顾满足运动学约束、动力学约束和负载约束,以保证机器人在运行过程中的平稳性.利用改进DE的全局寻优能力调整运动参数,进行关节空间的轨迹规划.在UR5机器人的平台进行仿真验证,仿真结果验证了算法的有效性、可行性.%In the process of executing welding and assembling tasks for collaborative industrial robot,the joint state cannot achieve the maximum constraint,thus the efficiency is affected.In order to ensure the"human - robot - environment"safety case,the maximum operating speed and response time of robot,an improved differential evolution(DE) algorithm based trajectory planning algorithm for collaborative robot is proposed.By using Monte Carlo algorithm,the operation space of robot is obtained,and the collision detection is conducted by using eneralized momentum method; at the same time,the minimum operation time of robot during the mission is set as the target,to meet the kinematic constraints,dynamics constraints and load constraints,to ensure stability of the robot in the operation process.The global optimization ability of improved DE is used for adjusting suitable motion parameters and implementing joint space trajectory planning.The simulation verification is carried out on the UR5 robot platform, and the simulation results verify the effectiveness and feasibility of the algorithm.