摘要
针对当前铁路货车制动系统故障诊断精度不高、效率低下的问题,以货车空气制动机风压数据为研究对象,提出了一种基于核主成分分析(KPCA)和优化隐马尔可夫模型(HMM)的货车制动系统故障诊断方法。首先,对制动机多通道风压信号进行特征提取,获得特征向量;然后采用KPCA对特征向量进行特征约减,获取主要的信息成分;采用K均值(K-means)算法对隐马尔可夫模型的初始参数进行优化,最后利用优化隐马尔可夫模型对空气制动机进行故障诊断。实验结果表明,与其它诊断模型相比,所提方法具有更高的诊断率和优越性。
Aiming at the problems of low accuracy and low efficiency of fault diagnosis of railway freight car braking system,taking the wind pressure data of freight car air brake as the research object,a fault diagnosis method of freight car braking system based on kernel principal component analysis(KPCA) and optimized hidden Markov model(HMM) is proposed.First,the features were extracted from multi-channel wind pressure signal of air brake;Next,the KPCA was utilized to reduce the feature dimensions and obtain the main information components;Then,the K-means algorithm was used to optimize the initial hidden Markov model parameters;Finally,the optimized model was used to diagnose the air brake.The experimental results show that the proposed method has higher recognition rate and superiority compared with the other diagnosis models.
作者
张鹏飞
岳建海
裴迪
焦静
ZHANG Peng-fei;YUE Jian-hai;PEI Di;JIAO Jing(School of Mechanical,Electronic and Control Engineering,Beijing Jiaotong University,Beijing 100044,China)
出处
《计算机仿真》
北大核心
2022年第5期167-171,244,共6页
Computer Simulation
基金
中国铁路总公司科技研究开发计划重大/重点课题(2017J004-H)。
关键词
核主成分分析
优化隐马尔可夫模型
均值算法
故障诊断
空气制动
Kernel principal component analysis(KPCA)
Optimized hidden Markov model
Means algorithm
Fault diagnosis
Air brake