基于ADPSO算法的机械臂轨迹规划  

Manipulator Trajectory Planning based on ADPSO Algorithm

作　　者：汤小红[1,2] 龚永健 王念娇 张宏 任垒垒 Tang Xiaohong;Gong Yongjian;Wang Nianjiao;Zhang Hong;Ren Leilei(College of Mechanical and Electrical Engineering,Central South University of Forestry and Technology,Changsha 410004,China;College of Mechanical Engineering,Hunan Institute of Science and Technology,Yueyang 414000,China)

机构地区：[1]中南林业科技大学机电工程学院,湖南长沙410004 [2]湖南理工学院机械工程学院,湖南岳阳414000

出　　处：《机械传动》2022年第5期123-128,166,共7页Journal of Mechanical Transmission

摘　　要：焊接机械臂工作路径复杂,对规划轨迹平滑性要求较高,并且规划轨迹需满足各关节运动学约束。提出了带扰动的自适应粒子群(Adaptive particle swarm optimization,ADPSO)算法,可以在满足关节约束条件下规划出时间、能力、跃度最优轨迹。采用5次NURBS曲线插值关节工作路径点,使各关节位置、速度、加速度、跃度曲线均连续光滑。利用ADPSO算法进行多目标最优轨迹规划,首先,将粒子外推思想与粒子群优化(Particle swarm optimization,PSO)算法结合,以增强粒子搜索能力;然后,对搜索所得个体极值与群体极值引入扰动,加快粒子收敛速度。在Matlab环境下进行仿真分析,对比其他智能算法,ADPSO算法的优化效果更好、优化时效性更快。The working path of welding manipulator is complex,which requires high smoothness of the planning trajectory,and the planning trajectory needs to meet the kinematics constraints of each joint.An adaptive particle swarm optimization(ADPSO)algorithm with disturbance is proposed,which can plan the optimal trajectory of time,ability and jump under joint constraints.The quintic NURBS curve is used to interpolate the joint working path points,so that the joint position,velocity,acceleration and jump curves are continuous and smooth.The ADPSO algorithm is used for multi-objective optimal trajectory planning.Firstly,the idea of particle extrapolation is combined with particle swarm optimization(PSO)algorithm to enhance the ability of particle search,and then disturbance is introduced to the individual extremum and group extremum to accelerate the convergence speed of particles.Simulation analysis is carried out in Matlab environment,compared with other intelligent algorithms,ADPSO algorithm has better optimization effect and faster optimization timeliness.

关 键 词：机械臂 轨迹规划 5次 NURBS曲线 自适应粒子群优化算法 多目标优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP241[自动化与计算机技术—控制科学与工程]