基于神经元网络的快速监控功能

应用神经元网络理论和计算机仿真实验相结合的方法,对用快速监控理论解决带钢头部厚度超差问题进行了研究。针对板带材轧制过程是一个复杂的非线性过程,难以建立精确数学模型的问题,提出了用神经元网络实现快速监控功能的方法,仿真证明,该方法对于及时地消除厚度偏差的后遗性有着比其他控制方法更好的效果。

Monitor and Control Quickly Function Based on Neural Network

With the header of hot strip's gauge precision has become an important factor affecting the product quality, Monitor and Control Quickly Function problem is studied with the combination of theoretical research and computer simulation. Making use of the nerve cell network method to finish Monitor and Control Quickly Function, cope with the head of the hot strip's gauge deviation that is drawn from the setting up rolling force not accurately. Through the simulation results can see this method that based on the nerve cell network is more effective dealing with the gauge deviation than other gauge control method.