基于非支配排序遗传算法的多目标轨迹优化方法  

A Multi-objective Trajectory Optimization Method Based on the NSGA-ⅢAlgorithm

作　　者：杨丽荣[1,2] 刘洋[1] 周俊 YANG Lirong;LIU Yang;ZHOU Jun(School of Mechanical and Electrical Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,Jiangxi,China;Jiangxi Mining and Metallurgy Electromechanical Engineering Technology Research Center,Jiangxi University of Science and Technology,Ganzhou 341000,Jiangxi,China)

机构地区：[1]江西理工大学机电工程学院,江西赣州341000 [2]江西理工大学江西省矿冶机电工程技术研究中心,江西赣州341000

出　　处：《实验室研究与探索》2025年第2期31-36,共6页Research and Exploration In Laboratory

基　　金：赣州市重点研发计划项目(202101124911);江西理工大学精品在线开放课程项目(XZG-22-02-14)。

摘　　要：机器人轨迹规划中,采用五次非均匀有理B样条(NURBS)插值算法可以在一定程度上提高机械臂轨迹规划的平滑性、连续性和稳定性,但无法解决运行时间、能耗及冲击的多目标最优值求解问题。为此,提出了一种面向实验室液压破碎机械臂轨迹优化的非支配排序遗传算法(NSGA-Ⅲ)。根据收敛性指标和间距指标选取Pareto最优解后进行轨迹优化,并与NSGA-Ⅱ对比。结果表明,在通过五次NURBS插值算法得到一条关节轨迹曲线后采用NSGA-Ⅲ进行轨迹优化,可以获得运行时间-能耗-冲击的Pareto最优解。相较于优化前,机械臂运行时间降低16%,最大关节角速度、角加速度及角加加速度分别降低8%、17%、19%。In the trajectory planning of mechanical arms,the employment of a fifth-order non-uniform rational B-spline(NURBS)interpolation algorithm,in a certain extent,can enhance the smoothness,continuity,and stability of the mechanical arm trajectory.However,it fails to address the multi-objective optimal value solution problem concerning time,energy consumption,and impact.Hence,a non-dominated sorting genetic algorithm(NSGA-Ⅲ)oriented towards the trajectory optimization of the laboratory hydraulic crushing mechanical arm is proposed.The Pareto optimal solution is selected based on convergence and distance indicators for trajectory optimization and is compared with the NSGA-Ⅱalgorithm.The results demonstrate that after obtaining a joint trajectory curve through the fifth-order NURBS interpolation algorithm,the NSGA-Ⅲalgorithm is adopted for trajectory optimization,thereby attains the Pareto optimal solution in the running time-energy-impact domain.Compared with the pre-optimization situation,the running time of the mechanical arm is decreased by 16%,and the maximum joint angular velocity,angular acceleration and angular jerk are reduced by 8%,17%and 19%respectively.

关 键 词：多目标轨迹优化 非支配排序遗传算法 参考点机制 收敛性指标 间距指标 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]