根据具有非线性、强耦合、不确定性过程的控制需要,提出了一种基于神经网络的内模控制方法.该方法充分利用神经网络的自学习及非线性逼近能力,建立非线性、强耦合、不确定性过程的动态模型及逆模型.采用这种方法对冷轧过程中带材全局板形进行仿真实验控制,取得了理想的控制效果. According to the requirement of control with nonlinear, strong coupling and uncertain process, an internal model control method based on neural network is proposed. This method makes full use of self-learning and nonlinear approximation ability of neural network to establish nonlinear and strong coupling , The dynamic model of the uncertainty process and the inverse model, the simulation of global shape of the strip in the cold rolling process is controlled by this method, and the ideal control effect is achieved.