六自由度机械臂逆运动控制方法

摘要

Firstly,the paper established the link bar coordinate system based on axis direction of manipulator joint.Secondly,according to Denavit-Hartenberg(D-H)notation,the coordinate system transformation matrix was derived.Thirdly,the forward kinematics model was established based on the coordinate system transformation matrix.Finally,we set up the inverse kinematics model based on reversely solving the forward kinematics model.The inverse kinematics model is a kind of multiple object optimization problems,and the optimal solution of the inverse kinematics model has characteristics of high accuracy with minimized rotation extent,and the solution is continuous.For the sake of solving the inverse kinematics model,we proposed an improved multiple population differential evolution (IMPDE) algorithm,which can destroy the aggregation level of the population by choosing different mutation operation based on mutation probability so as to enhance the ability of global search.The numerical results show that the new algorithm can solve the inverse kinematics model efficiently.%首先根据机械臂关节轴线方向建立了连杆坐标系,并利用Denavit-Hartenberg(D-H)法得到连杆坐标系变换矩阵；其次通过连杆坐标系变换矩阵得到机械臂正运动模型；最后通过反向求解正运动模型得到逆运动控制模型.逆运动控制模型是多目标优化问题,该模型的最优解既可以保证控制精度又可以保证各个关节角旋转幅度达到最小,同时也可以保证解的连续性.为了求解机械臂逆运动模型,提出了一种改进的多种群差分演化算法(IMPDE),新算法以一定概率选择不同的变异模式,可以起到破坏种群的聚集程度的作用,从而达到增强算法全局搜索能力的目的.最后,数值实验表明,新算法可以有效求解机械臂逆运动学模型.