基于小脑模型神经网络的气动六自由度并联平台的复合控制方法研究

针对气动驱动系统的强非线性特性,本文采用基于小脑模型(CMAC)神经网络和PD的复合控制结构,用于气动六自由度并联平台各驱动关节的轨迹跟踪控制,以提高系统的跟踪精度和抗干扰能力。CMAC网络采用不均匀的输入量化函数,在零值误差附近增加量化等级,以提高系统的控制精度;而在误差较大处减小输入量化级数,从而在不增加存储空间的情况下,加大输入信号的范围。系统的仿真和实验结果都表明,与PD控制相比,该方法能有效地提高系统的轨迹跟踪精度。

A Compound Control Method for a Six-DOF Parallel Pneumatic Manipulator Based on CMAC Networks

We present a compound control method based on CMAC(cerebellar model articulation controller) networks and PD controllers,and its applications to the control of sub-systems of a 6-degree of freedom(DOF) parallel pneumatic manipulator to modify the non-linearity of the system in order to improve the precision of trajectory and the anti-disturbance performances of the system.An asymmetric input division function is adopted in the CMAC network. A thin input division is used in the area whose error or differential error is nearly zero in order to increase the precision of the trajectory;and a coarse input division is used to divide the input space roughly to extend the scope of input without adding memory to the system.Simulations and experiments show that the trajectory precision is considerably improved by the compound control method compared with the PD controller.