空间机器人动力学奇异回避的笛卡尔轨迹规划  

Cartesian Trajectory Planning of Free-Floating Space Robot with Dynamic Singularities Avoidance

作　　者：金荣玉 耿云海[1] JIN Rong-yu;GENG Yun-hai(Research Center of Satellite Technology,Harbin Institute of Technology,Harbin 150001,China)

机构地区：[1]哈尔滨工业大学卫星技术研究所,哈尔滨150001

出　　处：《宇航学报》2020年第8期989-999,共11页Journal of Astronautics

摘　　要：针对空间机器人动力学奇异的回避问题,提出一种基于组合函数的笛卡尔轨迹规划方法。将梯形规划与正弦函数相结合,对机械臂末端的位姿进行参数化。机械臂沿着某一轨迹运动时,根据阻尼最小方差法(DLS)的特点,提出一种判断是否发生动力学奇异的方法,并据此进行轨迹规划。该方法可以保证空间机械臂运动过程中不会遇到动力学奇异。此外,将基座姿态扰动和机械臂运动时间作为目标函数的一部分,最终,轨迹规划问题转化为多目标优化问题,并利用混合整数规划的混沌粒子群优化算法(CPSO)进行求解。该优化算法能够改善标准粒子群算法(PSO)的"早熟"现象。仿真结果表明,新方法能够有效处理动力学奇异问题,减小基座姿态扰动及运动时间。To avoid dynamic singularities, a Cartesian trajectory planning method for a free-floating space robot is developed in this paper. Combining the trapezoidal programming and sine functions, the position and attitude of the end-effector are parameterized. Based on the damped least-squares(DLS), this paper proposes a method to determine whether the singularities occur during the movement of the manipulators. It can guarantee that the space robot cannot meet the singularities with the method. Moreover, considering the base attitude disturbance and the moving time, a cost function is obtained. Therefore, the trajectory planning issue is transferred to an optimization issue. The chaos particle swarm optimization(CPSO) algorithm for mixed integer programming is implemented to find the optimal solution. Compared with particle swarm optimization(PSO), it can improve the premature. The numerical simulations have demonstrated that the designed method can avoid dynamic singularities, and minimize the base attitude disturbance and the moving time effectively.

关 键 词：空间机器人 轨迹规划 动力学奇异 基座姿态扰动 运动时间 混沌粒子群优化算法(CPSO) 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]