冗余度机器人多性能准则优化研究

OPTIMIZING MULTIPLE PERFORMANCE CRITERIA IN THE OPERATION OF A REDUNDANT MANIPULATOR

为了解决在一个冗余度情况下同时优化多个性能准则的困难，基于“优化期望”和“优化分配”的概念，提出了实时在线地决定子任务优先权的新方法．任务被划分为一个高优先权的主任务和数个低优先权的子任务，子任务通过利用主任务的冗余性得到完成．文中利用作者所开发的“改进的加权梯度法（ＭＷＧＰＭ）”求解逆运动学，成功地解决了冗余度机器人的多性能准则同时优化问题．

To solve the difficulty of optimizing multiple performance criteria simultaneously using one redundant only,based on the concepts of 'the optimization expectation(OE)'and 'the optimization assigment(OA)',this paper presentes a new scheme that subtask priority is decided real time and on line.A requied task is devided into a main task with higher priority and several subtasks with lower priority.Subtasks can be performed utilizing redundancy on the main task.A modified weighted gradient projection method (MWGPM)which has remarkable merits is introduced,it is successfully solved that multiple performance criteria for a redundant manipulator are simultaneously optimised.This has been verified by simulation.