摘要
当重轨矫直参数控制模型的计算参数模型给出的压下量不能满足工艺要求时,需要建立一个自学习模型重新分配压下量。自学习模型中训练好的人工神经网络(ANN)可用于预测下次计算模型的分配系数,从而得出一个较优的初始压下量输入值,采用RBFNN算法从数据库中存储的历史数据中智能学习参数相互之间的关系,给出优化的调整值。
Parameter control model for heavy rail straightening gives a rolling reduction by calculation, and when the rolling reduction can not meet the technical requirements, it need to establish a self-learning model to redistribute of rolling reduction. The trained ANN （Artificial Neural Network） in self-learning model can be used to predict the next distribution coefficient of the calculation model, and get a better input values of roiling reduction, it uses RBFNN algorithm to study the correlation between the parameters from the historical data stored in thedatabase intelligently, and then give a optimal adjusted value.
出处
《微型机与应用》
2010年第18期83-85,共3页
Microcomputer & Its Applications
基金
武汉科技大学冶金装备及其控制教育部重点实验室开放基金资助项目(2009A07)
关键词
重轨矫直
自学习
ANN
智能学习
heavy rail straightening
self-learning
ANN
intelligent learning
