论文部分内容阅读
本文介绍了采用进化计算思想研制的机器人轨迹规划系统ＲＴＰ－１，提出多层次动态结构化编码方案，利用低中高三层进化算法分别优化距离、路径和关节角度偏差，基于多层综合优化策略解决多目标多约束工程优化问题，建立了次序相关问题求解的通用框架。在进化算法中，利用拉马克效应加快轨迹规划速度，在ＲＭ－５０１机械手上实现的任意空间直线和空间曲线轨迹规划具有良好的鲁棒性，规划轨迹的相邻臂构型间具有良好的柔顺性，规划轨迹的臂构型序列具有良好的平滑性。本文所采用的机器人轨迹规划方法具有通用性，可推广应用于各种动力学系统的研制。
This paper introduces a robot trajectory planning system RTP-1 based on evolutionary computation, proposes a multi-level dynamic structured coding scheme, and optimizes the distance, path and joint angle deviation by using the low, middle and high three-layer evolutionary algorithms to solve the problem based on multi-layer comprehensive optimization strategy Multi-objective and multi-constrained engineering optimization problem, a general framework for solving order-dependent problems is established. In the evolutionary algorithm, the use of Lamarck effect to speed up the trajectory planning speed, the RM-501 robot to achieve any space linear and spatial trajectory planning with good robustness, planning trajectory adjacent arm configuration has good compliance The trajectories of the arm and arm of the planned orbit have good smoothness. The robot trajectory planning method adopted in this paper has versatility and can be widely applied in the development of various dynamic systems.