机械制造与自动化2024,Vol.53Issue(6) :32-36.DOI:10.19344/j.cnki.issn1671-5276.2024.06.006

基于机器学习的螺栓连接接触状态实时预测

Real-time Prediction of Bolted Joint Contact State Based on Machine Learning

赵伟 马奔奔 唐林 杜宗亮
机械制造与自动化2024,Vol.53Issue(6) :32-36.DOI:10.19344/j.cnki.issn1671-5276.2024.06.006

基于机器学习的螺栓连接接触状态实时预测

Real-time Prediction of Bolted Joint Contact State Based on Machine Learning

赵伟 1马奔奔 1唐林 1杜宗亮2
扫码查看

作者信息

  • 1. 东方电气风电股份有限公司风电研究设计院,四川德阳 618099
  • 2. 大连理工大学工程力学系,辽宁大连 116023
  • 折叠

摘要

螺栓连接系统界面间的接触状态是衡量其工作状态和密封性能的重要指标.由于受到较强接触非线性和过多耦合变量的影响,接触应力分布的实时预测依然是一个难题.借助机器学习的方法,将复杂的螺栓连接问题封装在后台运算里,最终呈现出一个简单的能够实时预测接触应力分布的前台操作.

Abstract

The contact status of bolted joint between system interfaces is an important indicator measuring its working condition and sealing performance,the real-time prediction of contact stress distribution,however,remains a challenge due to the effects from strong contact nonlinearity and exceeding coupling variables.This article,with the help of machine learning technique,presents a simple front operation window capable of predicting the contact stress distribution in real-time by packaging the complicated bolted joints into background computational process.

关键词

螺栓连接/接触应力/机器学习/实时预测

Key words

bolted joints/contact stress/machine learning/real-time prediction

引用本文复制引用

出版年

2024
机械制造与自动化
南京机械工程学会 南京机电产业(集团)有限公司

机械制造与自动化

CSTPCD
影响因子:0.29
ISSN:1671-5276
段落导航相关论文