基于模糊灰关联分析的高速列车运行状态识别  

Running state recognition of high-speed train based on fuzzy grey correlation analysis

作　　者：李家会[1,2] 金炜东[1] 熊莉英[2] 

机构地区：[1]西南交通大学电气工程学院,成都610031 [2]西南科技大学信息工程学院,四川绵阳621010

出　　处：《振动与冲击》2014年第16期188-193,共6页Journal of Vibration and Shock

基　　金：国家自然科学基金(61134002);绵阳市科技计划项目(12G032-4)

摘　　要：针对高速列车运行状态监测问题提出小波包能量熵与模糊灰关联度相结合的运行状态识别方法。对高速运行状态下列车10个关键部位传感器振动信号进行均匀分段及多层小波包分解,将小波包能量熵作为特征值;随机选取四种运行状态下各10段数据求其平均能量熵作为参考序列,其余数据能量熵作为待检测序列,采用灰色理论对参考、待检测序列进行模糊灰关联分析,获得待检测序列对各运行状态隶属度;实现对高速列车运行状态识别。实验结果表明,该方法能有效诊断高速列车运行状态,尤其小样本、故障特征不明显时明显优于支持向量机及概率神经网络方法。Aiming at the high-speed train running state monitoring,a running state recognition method that couples wavelet packet energy entropy with fuzzy grey correlation degree technique was proposed. The vibration signals,which were acquired by ten sensors at the key positions of high-speed running train,were uniformly segmented and then decomposed by using multi-layer wavelet packets. The wavelet packet energy entropies,extracted from the vibration signals,were used as fault features. The average energy entropies of 10 pieces of random data of every running state were used as the reference sequences and the energy entropies of other data were used as the detected sequences. By analyzing the fuzzy grey correlation between the reference sequences and the detected sequences,the membership degree of the detected sequences belonging to four running states of the train was obtained and so the high-speed train running state recognition was realized. The experimental results show that the proposed method can effectively diagnose four running states of high-speed train,especially in the case of small samples and inconspicuous fault features. The proposed method is superior to the mothod of support vector machine and probabilistic neural network.

关 键 词：高速列车 状态识别 模糊灰关联分析 小波包能量熵 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]