基于神经网络的冗余度四面体变几何桁架机器人位置控制

摘要

A position control scheme for redundant tetrahedron based variable geometry truss manipulators is proposed based on two BP neural networks. The constraint conditions that is the minimum of change of joint variables are set and the objective function of optimization is defined. Based on it, the position control model of redundant tetrahedron based variable geometry truss manipulators is established. The simulation study of position control system for a four-celled tetrahedron variable geometry truss manipulator is carried out.%提出了一个基于双重BP神经网络的冗余度四 面体变几何桁架机器人位置控制方案，通过增加关节角变化最小的约束条件，定义了优化的 目标函数，在此基础上建立了冗余度四面体变几何桁架机器人位置控制模型，并以四重四面 体变几何桁架机器人位置控制为例进行了系统仿真。