基于递推差分进化算法的空间机器人参数辨识

针对空间机器人抓取未知目标时的特性参数以及关节摩擦系数的辨识和在线修正问题，提出一种基于递推差分进化算法的实时参数辨识方法。首先采用静态连续摩擦模型描述机器人的关节摩擦特性，并建立两关节空间机器人的非线性动力学模型，然后基于差分进化算法和递推最大似然估计法推导出递推差分进化算法，并用于空间机器人的参数辨识，最后仿真校验了该辨识方法的有效性。仿真结果表明，该算法的辨识精度优于遗传算法和最小二乘法，辨识速度较快，能满足遥操作的要求，且对于动态信息有较好的跟随性。

Parameter Identification of Space Robot Based on Recursive Different Evolution Algorithm

Aiming at the identification and on-line revision problem of the characteristic parameter of the unknown object captured by the space robot and the friction parameter of the robot joint, a method of real-time patameter identification based on the recursive different evolution algorithm is proposed in this paper. Firstly, a static continuous friction model is introduced to model joint friction, and the nonlinear dynamics model of a two-joint space robot is established. Then the recursive different evolution algorithm ,which is used to solved the parameter identification problem of the space robot, is deduced based on the different evolution and the maximum similar estimate algorithm. At last, the validity of the identification method is proved by simulation results. The simulation results show that the identification accuracy is better than genetic algorithms and least square method，and the  identification rate can meet the teleoperation requirement .This identification method can follow the dynamic information exactly.