基于机器学习的铁合金疲劳强度预测

打开文本图片集

中图分类号：TG241 文献标志码：A

文章编号：1001-4934（2025）05-0084-08

Abstract： In this paper， a fatigue strength prediction method based on Convolutional Neural Network-Gated Recurrent Unit （CNN- GRU） network model is proposed based on MakeItFrom to collect data on the composition of material elements，physical properties，and mechanical property parameters that affect fatigue strength，establish a fatigue strength dataset，and propose a fatigue strength prediction method based on CNN-GRU network model. The 2O feature parameters with more practical applications in engineering are used as inputs，and fatigue strength is the output，and various algorithms are trained based on the dataset. Then，the trained machine learning regression prediction model was used to predict the fatigue intensity. Finally，a series of performance evaluations among the machine learning regression prediction models were conducted by evaluation metrics such as Mean Absolute Error （MAE），Root Mean Squared Error（RMSE），Coefficient of Determination R2 and Mean Absolute Percentage Error （MAPE），which proved the superiority of the proposed models and methods used.

Key words： machine learning； fatigue strength; ferrous alloys；regression prediction

0 引言

随着工业技术的不断进步，现代金属材料、机械设备和部件正逐步升级，变得更加智能化和多功能化。（剩余8007字）