基于人工神经网络和灰色系统的异种钢焊接接头碳迁移过程的预测  

of Carbon Migration in Dissimilar Steel Welded Joints Based on ANN and Grey Modeling

作　　者：邹杨[1] 潘春旭[1] 

机构地区：[1]武汉大学物理科学与技术学院,湖北武汉430072

出　　处：《中国材料科技与设备》2006年第1期57-59,67,共4页Chinese Materials Science Technology & Equipment

基　　金：石油科技中青年创新基金项目（2002CX05）;教育部全国优秀博士学位论文作者专项资金项目（200233）

摘　　要：本文主要通过灰色系统辅助人工神经网络（ANN）的模型时异种钢焊接接头在高温运行过程中的碳迁移现象进行预测。指出了单纯使用神经网络方法的局限性及存在的问题。将计算预测曲线与实际工件高温服役真实曲线进行比较，得到了较好的符合性。This paper proposes to method for prediction of the carbon migration process in dissimilar steel welded joints during heat treatment or high temperature service based on grey modeling assisted artificial neural network （ANN）. The limitation and problems when using a simplex ANN method in the prediction are discussed. It has been shown that the calculated prediction curve using the present method exhibits a good match with the actual measured curve from the workpiece which have been served in high temperature for a long term.

关 键 词：人工神经网络 灰色系统 异种钢焊接 碳迁移 

分 类 号：TG457.11[金属学及工艺—焊接] O621.2[理学—有机化学]