模型分块逼近的RBF网络机械臂自适应控制

Adaptive RBF Neural Network Model Block Approach Control of Robotic Manipulators

针对隧道掘进后期维护中喷涂机械臂的控制轨迹精确度不高的问题,研究了一类基于模型分块逼近的机械臂RBF网络自适应控制方法。通过RBF网络分别逼近机械臂动力学中的三个方程系数矩阵,设计自适应律,实现机械臂控制,并对系统的稳定性进行证明。仿真显示模型分块逼近的RBF网络自适应方法具有很好的轨迹跟踪特性。

To the problem of spray lacquer manipulator trajectory tracking accuracy in latest stage of tunnel excavation,a solution is put forward by model block of the approaching RBF neural network adaptive control method. By approaching of RBF network to three of the mechanical arm dynamics equation coefficient matrix,an adaptive lawis designed to control the mechanical arm,which proves system stability. Simulation results showthat model block of the approaching RBF network adaptive trajectory tracking control method has very good properties.