摘要
为了研究机器人关节空间轨迹时间最短的优化计算问题,依据混沌优化理论,采用改进的基于混沌变量优化算法,以时间最短为性能指标对轨迹进行优化求解。研究了均匀非周期四阶B样条曲线的优化,每一段B样条曲线的运行时间作为优化参数,优化问题模型包括关节角速度、角加速度、角加加速度及力矩4种约束。给出了PUMA560前三铰B样条轨迹优化算例,优化结果明显优于采用复合形法或有约束随机搜索方法的优化结果;该算法简单,易于实现,求解速度快,任给一组初值得到优化结果的可靠性达90%以上,逼近约束条件的误差几乎为零,进一步减少了运行时间,从而有效地实现了机器人在关节空间的轨迹优化。
In order to study the optimization problem of time-minimal path planning for robot in joint space, modified optimization algorithm based on optimization theory of chaos is proposed, and the time- minimal is considered as performance index. The optimization of fourth-order quasi-uniform B-splines is discussed. Considering the time of each B-spline as optimization parameter, the physical constraints include joint velocity, acceleration, jerk and torques. An example on B-splines path optimization for PU- MA560 robot pre-3 joints is presented for illustration, which has shown that the optimization results of the algorithm are much better than using compound form method or stochastic search technique with constraints. The algorithm which is simple is realized easily and its speed is fast; its reliability of optimization results is above 90% for any initial values; and there is almost no error of approaching constraints; the operation time is decreased. So it is effective for the realization of trajectories optimization through u- sing the algorithm for robot in joint space.
出处
《电机与控制学报》
EI
CSCD
北大核心
2007年第2期174-177,共4页
Electric Machines and Control
基金
湖南科技大学科学研究基金资助(E54100)
关键词
机器人
混沌优化
轨迹规划
robot
chaos optimization
path planning
