Research on Wheel-flat recognition Algorithm for High-speed Train Based on High-order Spectrum Feature Extraction

doi:10.3901/JME.2017.06.102

Abstract

Abstract:

With the increasing of train speed and axle load, the wheel flats occur more frequently. A detection method for wheel flats is proposed based on PSO-SVM, in which high-order spectrum is used to extract the features of the rail vibration responses when the train runs on the rail. A vehicle-track vertical coupling system and wheel flat are modeled and the rail dynamic responses are calculated with the condition of normal wheel and wheel flat. The two-dimensional contour and three-dimensional double spectrum diagrams are obtained by analyzing two kinds of rail vibration responses with high-order spectrum method. Six texture features of the 2D contour map, which are extracted from gray-gradient co-occurrence matrix, are combined with the train speed to detect whether there is a wheel flat by PSO-SVM. Then the peak value of the diagonal slice of the three-dimensional double spectrum map and the frequency of the inner spread of the two-dimensional contour map are added into PSO-SVM to recognize different wheel-flat levels. The results show that the detection accuracy of normal wheel and wheel flats can be up to 100% and the recognition accuracy of wheel flat different levels is 94.6%. Finally, the recognition accuracy are compared and analyzed between high-order spectrum and EMD methods which are used to extract the feature signals. For EMD method, the detection accuracy for wheel flats is 98.3%, and for different wheel flat levels is only 56.4%. The investigation shows the detection method based on high-order spectrum and PSO-SVM can be used to identify wheel flat and its levels effectively.

Key words: gray-gradient co-occurrence matrix, high-order spectrum, PSO-SVM, rail vibration responses, wheel flats

ZHAO Rong, SHI Hongmei. Research on Wheel-flat recognition Algorithm for High-speed Train Based on High-order Spectrum Feature Extraction[J]. Journal of Mechanical Engineering, 2017, 53(6): 102-109.

References

[1]    VITTORIO B，FRANCESCO C，RINALDO C，et al. Wheel-flat diagnostic tool via wavelet transform[J]. Mechanical Systems and Signal Processing，2006，20(8)：1953-1966.
[2] LIANG B，IWNICKI S，BALL A，et al. Adaptive noise cancelling and time-frequency techniques for rail surface defect detection[J]. Mechanical Systems and Signal Processing，2015，54-55：41-54.
[3] 李忠继，魏来，戴焕云，等. 基于Hilbert-Huang变换的车轮扁疤识别方法[J]. 交通运输工程学报，2012，12(4)：33-41.
        LI Zhongji，WEI Lai，DAI Huanyun，et al. Identification method of wheel flat based on Hilbert-Huang transform [J]. Journal of Traffic and Transportation Engineering，2012，12(4)：33-41.
[4] 姜爱国，王雪. 车轮踏面擦伤的集成粗糙神经网络预示诊断[J]. 清华大学学报，2005，45(2)：170-173.
        JIANG Aiguo，WANG Xue. Prognostics for surface scrapes on wheels based on integrated rough neural network[J]. Tsinghua Science and Technology，2005，45(2)：170-173.
[5] 江航，尚春阳，高瑞鹏. 基于EMD和神经网络的轮轨故障噪声诊断识别方法研究[J]. 振动与冲击，2014，33(17)：34-38.
        JIANG Hang，SHANG Chaoyang，GAO Ruipeng. Wheel/rail fault noise diagnosis method based on EMD and neural network[J]. Journal of Vibration and Shock，2014，33(17)：34-38.
[6] 徐姗，黄宜坚. 基于高阶谱和支持向量机的溢流阀的故障诊断[J]. 机床与液压，2013，41(3)：148-151.
        XU Shan，HUANG Yijian. Fault diagnosis of relief valve based on higher order spectrum and support vector machine[J]. Machine Tool & Hydraulics，2013，41(3)：148-151.
[7] 姜万录，郑直，胡浩松. 基于EEMD形态谱和支持向量机复合的滚动轴承故障诊断方法[J]. 工程科学学报，2015，37(增刊1)：72-77.
        JIANG Wanlu，ZHENG Zhi，HU Haosong. Fault diagnosis of ball bearing based on EEMD morphological spectrum and support vector machine[J]. Chinese of Journal of Engineering，2015，37(Suppl.1)：72-77.
[8] 翟婉明. 车辆—轨道耦合动力学[M]. 北京：科学出版社，2007.
        ZHAI Wanming. Vehicle-track coupling dynamics [M]. Beijing：Science Press，2007.
[9] DING Jianming，LIN Jianhui，WANG Guangming，et al. Time-frequency analysis of wheel-rail shock in the presence of wheel flat [J]. Journal of Traffic and Transportation Engineering，2014，1(6)：457-466.
[10] 孙洁娣，靳世久. 基于小波包能量及高阶谱的特征提取方法[J]. 天津大学学报，2010，43(6)：562-566.
        SUN Jiedi，JIN Shijiu. Feature extraction method based on wavelet packet energy and high-order spectrum [J]. Journal of Tianjin University，2010，43(6)：562-566.
[11] 李玉韩，蔡坤宝，张继红. 双谱估计在中医脉象信号分析中的应用[J]. 重庆大学学报，2006，29(5)：44-47.
        LI Yuhan，CAI Kunbao，ZHANG Jihong. Application of bispectrum estimation to analyzing pulse signals [J]. Journal of Chongqing University，2006，29(5)：44-47.
[12] 从蕊，高光甫，樊瑞筱，等. 基于灰度-梯度共生矩阵和模糊核聚类的振动图形识别方法[J]. 振动与冲击，2012，31(21)：73-76.
        CONG Rui，GAO Guangfu，FAN Ruixiao，et al. Vibration image recognition method based on gray-gradient co-occurrence matrix and kernel-based fuzzy clustering[J]. Journal of Vibration and Shock，2012，31(21)：73-76.
[13] 熊安斌，丁其川，赵新刚，等. 基于单通道sEMG分解的手部动作识别方法[J]. 机械工程学报，2016，52(7)：6-13.
        XIONG Anbin，DING Qichuan，ZHAO Xingang，et al. Classification of hand gestures based on single-channel sEMG decomposition[J]. Journal of Mechanical Engineering，2016，52(7)：6-13.
[14] TIMAC G，BRAUT S，ZIGULI R. Comparative analysis of PSO algorithms for PID controller tuning[J]. Chinese Journal of Mechanical Engineering，2014，27(5)：928-936.