基于神经网络的超磁致伸缩智能构件滑模控制

提出了一种利用超磁致伸缩材料（giant magnetostrictive material GMM）智能构件精密加工活塞异形孔方法。 为了消除GMM智能构件迟滞非线性影响，提出一种神经网络前馈复合离散滑模变结构控制策略，实现GMM智能构件的精密位移控制。将智能构件的输出位移及其变化率作为小脑模型神经网络（CMAC）输入，构件的输入电流作为网络输出，利用CMAC在线自学习能力建立GMM智能构件的迟滞逆模型，神经网络的建模近似误差以及外界干扰通过离散滑模变结构控制器来消除。仿真结果表明此控制策略能在线建立智能构件的迟滞逆模型，消除迟滞非线性的影响，可实现智能构件的精密位移控制。

Sliding mode control based on neural network for giant magnetostrictive smart component

A new method for precise machining non-cylinder pin hole of piston by using giant magnetostrictive smart component is presented. To eliminate the impact of GMM smart component hysteresis and nonlinearity, a real-time hysteretic compensation control strategy combining a CMAC neural network feedforward controller and a sliding mode controller is proposed to implement the precision position tracking control of the smart component. The input data of CMAC neural network are the current smart component output and the output rate, the output of neural network is smart component input. The inverse hysteresis model of GMM smart component is achieved by CMAC network on-line learning. The model approximate error of CMAC neural network and the external disturbance is eliminated by using discrete sliding controller. Simulation shows that this control strategy can on-line obtain inverse hysteresis model of the smart component, eliminate the hysteretic nonlinear impact and achieve the precision control of the smart component.