基于神经网络PID的轧机AGC力控制

Hydraulic AGC Press System Based on NNPID Controller

以PID控制器为基础,辅助以神经网络的自学习、自适应能力,将神经网络和PID控制融合为一体,设计出一种神经网络PID控制器,应用于具有高度非线性、时变性的板带轧机AGC力控制系统,仿真结果表明,能有效改善系统的动态品质,并具有良好的适应能力。

A new neural network PID （NNPID） controller,which was based on PID by means of NN＇s ability of self-learning and adaptive, was presented. The NNPID controller was designed by combining neural network with PID control strategy. An improved BP algorithm was used in the NN. This paper presented the control for high nonlinear, time-varying hydraulic AGC of rolling mills based on the NNPID controller. The simulations show that dynamic quality of the system is imnroved, and NNPID has good adantability.