摘要
针对火工品装配关键工装的剩余使用寿命难以预测以及现场数据难以提供参考的问题,构建集成装配特征提取与最小化不纯度加权和的火工品装配关键工装寿命预测模型,开发了火工品装配关键工装寿命预测系统原型。以主成分分析提取关键工装的装配特征,以随机森林模型对关键工装的剩余使用寿命进行预测。通过实例对比分析证明,相较于传统寿命预测方法,本文所提方法预测准确率提升了0.89%,均方误差、均方根误差和预测用时分别同比降低22.3%、22.3%和24.8%,具有预测精度高、稳定性好和速度快等优点。
To address the difficulties in predicting the remaining life span of key fixtures in assembly of initiating explosive product and providing reference for field data,a life span prediction model of key fixtures for initiating explosive product assembly is established integrating assembly feature extraction and minimization of impurity weighted sum.In addition,a prototype of a life span prediction system for key fixtures in initiating explosive product assembly is developed.In this paper,assembly features of key fixtures are extracted by principal component analysis,while the remaining life span of key fixtures is predicted by random forest model.Comparative analysis shows that,compared with traditional methods,the prediction accuracy of the proposed method is increased by 0.89%,and the mean square error,root mean square error and prediction time are decreased by 22.3%,22.3%and 24.8%,respectively.It is proved that the method has the advantages of high prediction accuracy,good stability and fast prediction speed.
作者
朱晓峰
徐曼菲
刘治红
冷杰武
ZHU Xiaofeng;XU Manfei;LIU Zhihong;LENG Jiewu(State Key Laboratory of Precision Electronics Manufacturing Technology and Equipment,Guangdong University of Technology,Guangzhou 510006,China;Department of Intelligent Manufacturing,Automation Research Institute Co.,Ltd.of China South Industries Group Corporation,Mianyang 621000,China)
出处
《工业工程》
北大核心
2023年第5期149-158,共10页
Industrial Engineering Journal
基金
国防基础科研资助项目(JCKY2020209B005)。
关键词
火工品
关键工装
主成分分析
随机森林
剩余使用寿命
寿命预测
initiating explosive product
key fixture
principal component analysis
random forest
remaining life span
life span prediction
作者简介
朱晓峰(2000-),男,广东省人,硕士研究生,主要研究方向为智能制造;通讯作者:冷杰武(1989-),男,湖南省人,教授,博士,主要研究方向为智能制造。Email:jwleng@gdut.edu.cn。
