基于Hopfield神经网络的板形预测控制模型

THE FLATNESS PREDICTIVE CONTROL MODEL BASED ON HOPFIELD NEURAL NETWORKS

将Hopfield神经网络建模方法与预测控制思想相结合用于带钢板形自动控制,建立了基于Hopfield神经网络的板形的模型算法预测控制系统数学模型。模型经仿真验证表明能较好地跟踪目标值,并具有较强的抗干扰能力。

The modeling of Hopfield neural networks was combined with the predictive control algorithm to the AFC （Auto Flatness Control） system, a model of algorithmic control based on these neural networks were offered for the flatness predictive control. The simulation by computer with the production data showed that this algorithm could trail the target value preferably, and has excellent antidisturbance ability.