金属材料腐蚀预测模型研究进展

Research Progress on Corrosion Prediction Model of Metallic Materials for Electrical Equipment

对现有腐蚀领域常用腐蚀预测方法,包括函数模型、灰色理论模型、神经网络预测模型,剂量响应函数模型和随机森林模型等进行总结分析,并将现有的腐蚀预测模型分为腐蚀-时间和腐蚀-环境预测模型,介绍了不同腐蚀预测模型的特点和适用范围等。最后,根据电力行业的特点对金属材料的腐蚀预测研究提出了一些展望。

Metallic materials for electrical equipment are affected by many factors related with environment during service,and their corrosion behavior is very complex,therefore,which is difficult to be accurately predicted.The development of computer technology and data analysis technology enriches the prediction methods for corrosion behavior of metallic materials with better accuracy.This paper summarizes and analyzes the existing common corrosion prediction methods in the field of corrosion,including function model,grey theory model,neural network prediction model,dose response function model and random forest model etc.,and which then are classified into two types,namely corrosion-time models and corrosion-environment prediction models.Furthermore,the characteristics and application scope of different corrosion prediction models are introduced.Finally,prospects for the corrosion prediction of metallic materials are put forward especially in terms of the demands of power industry.