机电工程技术2024，Vol.53Issue(7)：241-245,5.DOI:10.3969/j.issn.1009-9492.2024.07.051
汽车底盘零部件疲劳耐久性能测试新方法的研究
Research on the New Method for Fatigue Durability Test of Automobile Chassis Parts
摘要
Abstract
The traditional fatigue durability test of automobile chassis parts requires long time repeated cyclic loading,which is not only highly dependent on equipment and manpower,but also inefficient.A new method for fatigue durability testing of automobile chassis parts is proposed.The main technical feature of this method is the introduction of digital image correlation and convolutional neural network algorithm in the result analysis of fatigue durability test.The binocular high-precision camera and the digital image correlation method are used to obtain the equivalent characteristic strain cloud map of the specimen in the whole process of the test.Subsequently,based on the ResNet-152 convolutional neural network model,the fatigue durability status of the specimens is identified.After two stages of training and testing,the prediction accuracy of the fatigue failure identification model reached over 91%,and an intelligent fatigue failure identification model for the components is formed,thereby achieving the function of determining whether automotive components have fatigue failure cracks.The new test method realizes the online monitoring of the fatigue durability performance of automobile chassis components,which greatly improves the testing efficiency and provides a new idea for the fatigue reliability prediction of automobile structural components.关键词
汽车零部件/疲劳耐久/数字图像相关法/卷积神经网络Key words
automobile parts/fatigue durability/digital image correlations/convolutional neural network分类
交通工程引用本文复制引用
吴奕东,李妮妮,曹伟,刘祎晗..汽车底盘零部件疲劳耐久性能测试新方法的研究[J].机电工程技术,2024,53(7):241-245,5.基金项目
广东省工业摩擦学企业重点实验室项目(2023B1212070025) （2023B1212070025）
