Using neural network models for predicting mechanical properties after hot plate rolling processes

被引：22

作者：

Korczak, P ^{[1
]}

Dyja, H ^{[1
]}

Labuda, E ^{[1
]}

机构：

[1] Czestochowa Tech Univ, Fac Met & Mat Engn, PL-42200 Czestochowa, Poland

来源：

JOURNAL OF MATERIALS PROCESSING TECHNOLOGY | 1998年 / 80-1卷

关键词：

artificial neural network; recrystallization; mechanical properties;

D O I：

10.1016/S0924-0136(98)00151-4

中图分类号：

T [工业技术];

学科分类号：

08 ;

摘要：

The neural network model for prediction of mechanical properties of hot rolled plates was constructed in the present work. The use of a neural network gives good results in predicting a non-linear relationship between a steel chemical composition, microstructure parameters, cooling rate and final mechanical properties. The model can be used to describe the spread of the ferritic grain size and the distribution of mechanical properties in the thickness of final plates. (C) 1998 Elsevier Science S.A. All rights reserved.

引用

页码：481 / 486

页数：6