基于工况参数信息的摩擦界面形貌反演  

Friction interface morphology inversion based on working condition parameter information

作　　者：王煜鑫 吴健鹏[1] 杨成冰 王立勇[1] 黄晓赞 WANG Yuxin;WU Jianpeng;YANG Chengbing;WANG Liyong;HUANG Xiaozan(Key Laboratory of Modern Measurement and Control Technology,Ministry of Education,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学现代测控技术教育部重点实验室,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2024年第4期33-40,共8页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金项目(52105084)。

摘　　要：为优化湿式离合器设计,增强传动系统可靠性,并为工程研究与应用提供基础数据,以湿式离合器为研究对象,设计通用机械性能测试仪(universal mechanical testing machine,UMT)摩擦磨损试验机盘-盘实验,在给定工况下应用循环神经网络(recurrent neural network,RNN)算法,构建了湿式离合器摩擦元件三维微界面形貌反演模型。通过对比11组工况下的真实值和反演值,验证了RNN反演模型的准确性,其中测试工况的平均绝对百分比误差为4.04%,决定系数为0.9806。最后,借助反演模型分析了转速和压力2个工况参数对界面形貌特征的影响,湿式摩擦副界面形貌受压力的影响较为显著。To enhance the design of wet clutches,improve the reliability of the transmission system,and provide basic data for engineering research and application,taking wet clutches as the research object,a three-dimensional micro-interface morphology inversion model of wet clutch friction component was constructed by designing a disk-disk experiment of a universal mechanical testing machine(UMT)friction and wear tester,and applying recurrent neural network(RNN)algorithm under given working conditions.The accuracy of the RNN inversion model was verified by comparing the real values and inversion values under 11 sets of working conditions,with a mean absolute percentage error(MAPE)of 4.04%and a coefficient of determination of 0.9806 under the test condition.Finally,the effects of two working condition parameters,rotational speed and pressure,on the interface morphology were analyzed using the inversion model.The interface morphology of the wet friction pair was significantly affected by pressure.

关 键 词：微界面形貌 数据反演 循环神经网络 湿式摩擦副 

分 类 号：TB1[一般工业技术]