基于人工神经网络的CSP热轧板带力学性能预报  

Prediction of mechanical property of hot-rolled strips by artificial neural network

作　　者：胡于华[1] 程晓茹[1] 戴宇杰[1] 

机构地区：[1]武汉科技大学材料与与冶金学院,湖北武汉430081

出　　处：《南方金属》2007年第1期17-20,共4页Southern Metals

摘　　要：建立了BP神经网络力学性能预报模型.神经网络模型经过训练,得到的预报结果与实际测量的数据相比较表明:BP神经网络预报产品的力学性能精度较高,具有良好的推广价值.A model for predicting the mechanical properties of hot-rolled strips produced by CSP technique was proposed based on BP neural network. A comparison of the properties predicted using a trained artificial neural network（ ANN）, with the measured ones showed that the prediction by the BP neural network was precise and reliable, and hence is suitable for an extended application.

关 键 词：人工神经网络 力学性能 预报 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TG335.5[自动化与计算机技术—控制科学与工程]