人工神经网络预测动态再结晶的研究进展与关键技术  

Research progress and key technology of artificial neural networks in predicting dynamic recrystallization

作　　者：黎小辉 黎子聪 LI Xiaohui;LI Zicong(School of Mechanical and Electrical Engineering and Automation,Foshan University,Foshan 528225,China)

机构地区：[1]佛山大学机电工程与自动化学院,佛山528225

出　　处：《中国有色金属学报》2024年第8期2547-2558,共12页The Chinese Journal of Nonferrous Metals

基　　金：广东省科技厅科技计划资助项目(2022A0505050081)。

摘　　要：本文综述了近年来人工神经网络在预测金属材料动态再结晶方面的研究进展,通过分析加工参数(温度、应变速率等)对动态再结晶的影响程度,并对比多种输出变量(应力−应变曲线、再结晶体积分数等)在判定动态再结晶机制中的作用,构建了一套基于网络输入输出特征的筛选准则。本文对人工神经网络模型的应用进行了详细介绍,包括原始数据的预处理、超参数的整定以及评估指标的选择。最后,在总结当前人工神经网络在动态再结晶机制预测领域研究现状的基础上,提出了一系列值得深入探索的新方向。This paper reviewed the research progress in recent years on artificial neural networks(ANNs)for predicting the dynamic recrystallization(DRX)of metallic materials.By analyzing the influence degree of the processing parameters(temperature and strain rate)on DRX and comparing the role of multiple output variables(stress-strain curves and volume fractions of recrystallized grains)in determining the DRX mechanisms,a set of screening guidelines based on the input and output characteristics of the network was constructed.This paper provided a detailed description of the application of artificial neural network models,including the preprocessing of raw data,rectification of hyperparameters and selection of evaluation metrics.Finally,based on a summary of the current research status of ANNs in the field of DRX mechanism prediction,this study proposed a series of new directions worth exploring in depth.

关 键 词：数值模拟 动态再结晶 热变形 BP人工神经网络 金属材料 

分 类 号：TG146.23[一般工业技术—材料科学与工程]