小波神经网络在板形板厚控制系统中的应用

Strip Flatness and Gauge Control Based on Wavelet Neural Networks

带材轧制是一个复杂的非线性过程,针对板形和板厚控制相互耦合等特点,本文提出了一种基于小波神经网络的自适应控制新算法。文中系统由两个小波神经网络组成,分别实现综合系统的模型辨识和控制。由于小波变换的紧支性及神经网络的非线性映射能力,模型辨识能准确地辨识板形板厚系统的动态特性,控制器能产生较为复杂的控制规律。仿真结果证明,该板形和板厚控制系统具有良好的自适应跟随和抗扰性能,其控制效果优于传统的解耦PID控制。

Strip rolling is a very complicated nonlinear process. Automatic flatness control （AFC） and automatic gauge control （AGC） interact and couple each other. A novel adaptive control system of AFC-AGC is developed by using two wavelet neural networks （WNNs） as the identifier and the controller. Because WNN has the ability of mapping nonlinear functions and time-frequency localization properties, dynamic characteristics of AFC-AGC can be more precisely identified, and more complex control strategies can be mapped. Simulation results show that the control system has good performance of adaptively tracking targets and resisting disturbances. It is superior to the conventional decoupled PID control by improving the strip flatness and the gauge accuracy.