一种基于优化算法的机械手运动学逆解

本文给出一种基于优化算法的机械手运动学逆解的方法 ,这种优化方法基于信赖域方法 ,,具有超线性的收敛速率 .这种方法不仅具有牛顿方法的快速收敛性 ,又具有理想的总体收敛性 .这种方法较 CCD& BFS有明显的优点 ,可以在一般的 PC机上实现实时求解 .在 P II40 0上仅需不到 10 ms就可以求得最优解 .

This paper proposes an approach to the inverse solution for the manipulator kinematics based on the optimal algorithm. This optimal approach, based on confidence threshold algorithm, possesses the superlinear convergence rate. This approach has not only the rapid convergence for Newton's method, but also the desirable overall convergence. Compared with CCD & BFS, this approach has the obvious advantages and implements a real time solution on PC Computer generally. The optimal solution can be solved on PⅡ400, using less than 10ms.