Abstract: Aiming at the deficiency that the spraying trajectory and joint trajectory do not satisfy the mixed constraints caused by the mapping nonlinearity between Cartesian space and joint space,a collaborative optimization of robotic spraying trajectory based on Dual-population Chaotic Search Particle Swarm Optimization (DCSPSO) algorithm was proposed.The joint angle sequence was constructed according to the pre-selected trajectory feature points.Aiming at the robotic spraying efficiency and motion stability,the multi-objective optimization model of joint trajectory was established.Using the DCSPSO algorithm,the optimization model was solved to obtain the Pareto optimal solution,which made the joint trajectory satisfy the robotic kinematics constraints.Based on the chord error and film thickness error of the theoretical and feedback trajectory,the spraying trajectory error model was established and the quality of optimal solution was verified to make the spraying trajectory satisfy the machining accuracy constraints.The case indicated that DCSPSO algorithm had stronger global and local search ability than classical multi-objective algorithms such as multi-objective genetic algorithm.Trajectory error model could add feature points reasonably,which made the maximum chord height error of the theoretical and feedback trajectories was reduced from 12.619 mm to 1.587 mm,and the maximum film thickness error was reduced from 11.47 μm to 1.18 μm.

Key words: Cartesian space, joint trajectory, spraying, multi-objective optimization, trajectory error optimization

摘要： 针对笛卡尔空间与关节空间的映射非线性,喷涂轨迹和关节轨迹不满足混合约束的问题,提出基于双种群混沌搜索粒子群优化(DCSPSO）算法的机器人喷涂轨迹协同优化。根据预选取的轨迹特征点构建关节角度序列,以机器人的喷涂效率和运动稳定性为目标建立关节轨迹多目标优化模型,利用DCSPSO算法求解优化模型得到Pareto最优解,使关节轨迹满足机器人运动学约束,最后根据理论轨迹与反馈轨迹的弦高误差和漆膜厚度误差建立喷涂轨迹误差模型,并验证最优解的质量,使喷涂轨迹满足加工精度约束。通过实例表明,DCSPSO算法较多目标遗传算法等经典多目标优化算法具有更强的全局和局部搜索能力,利用轨迹误差模型可合理增加特征点,使理论轨迹与反馈轨迹的最大弦高误差从12.619 mm降至1.587 mm,最大漆膜厚度误差从11.47 mm降至1.18 mm。

关键词: 笛卡尔空间, 关节轨迹, 喷涂, 多目标优化, 轨迹误差优化