摘要
针对特种车辆齿轮箱工作环境恶劣、状态识别困难的现实问题,这里提出了一种基于自适应优化变分模态分解(Variational Mode Decomposition,简称VMD)和奇异值分解(SVD)的特征值提取方法,并结合支持向量机(Support Vector Machine,SVM)构建诊断模型,应用到齿轮箱的状态识别中。首先,针对VMD分解层数K值难确定问题,结合相关系数和阈值提取有效分量,确定最优分解层数K,完成对VMD分解的自适应优化。然后用改进后的VMD算法对振动信号进行分解,用相关系数筛选出蕴含故障信息最丰富的分量进行频谱分析和SVD特征值提取,将特征值输入到构建好的支持向量机诊断模型中,根据输出结果识别齿轮箱状态。研究结果表明,该方法能有效应用于特种车辆齿轮箱状态识别,诊断正确率达到95.36%,为恶劣工况下齿轮箱状态识别提供了一种有效的应用方案。
In order to solve the problem of vehicle gearbox state recognition under strong interference environment,it proposes an eigenvalue extraction method based on improved VMD-SVD,and combines SVM to build a diagnostic model,which is applied to gearbox state recognition.First,combine the correlation coefficient and threshold to extract the effective components,and de⁃termine the optimal decomposition level K.Then,the vibration signal is decomposed by improved VMD,and the fault informa⁃tion is filtered out by the correlation coefficient.Spectrum analysis and SVD eigenvalue extraction are performed on the signal,Input the eigenvalues into the constructed SVM diagnostic model,and identify the gearbox status according to the output.The re⁃search results show that the method can be effectively applied to the state identification of the gearbox of special vehicles,and the accuracy rate reaches 95.36%,which provides an effective application scheme for the state identification of the gearbox under complex working conditions.
作者
何雷
刘溯奇
张皓惟
HE Lei;LIU Suqi;ZHANG Haowei(Liuzhou Railway Vocational and Technical College,Guangxi Liuzhou 545616,China;College of Mechanical and Electronic Engineering,Central South University,Hu’nan Changsha 410083,China)
出处
《机械设计与制造》
北大核心
2025年第7期86-90,96,共6页
Machinery Design & Manufacture
基金
广西高校中青年教师科研基础项目(2022KY1401)
教育部支撑技术项目(625010339)。
关键词
齿轮箱
变分模态分解
奇异值分解
支持向量机
Gearbox
Variational Mode Decomposition
Singular Value Decomposition
Support Vector Machines
