基于VMD和BP神经网络的轨道病害诊断方法

doi:10.16180/j.cnki.issn1007-7820.2022.04.007

摘要/Abstract

摘要：

针对从非线性、非稳态的轨枕振动信号中提取病害特征困难的问题,文中提出一种基于变分模态分解和多尺度排列熵的轨道病害特征提取方法,并采用BP神经网络病害诊断模型进行病害识别。利用变分模态分解方法将采集到的振动加速度信号进行分解,得到若干个本征模态分量。计算这些本征模态分量的多尺度排列熵值,将其作为轨道病害的高维特征向量,以实现对轨枕振动信号的降噪和病害特征的提取。通过建立BP神经网络病害诊断模型,将高维特征向量输入到BP网络中进行训练、拟合、验证,并与经验模态分解和BP神经网络结合的方法对比。分析结果表明,文中所提方法识别准确率更高,能够有效地进行病害诊断。

关键词: 变分模态分解, 多尺度排列熵, BP神经网络, 本征模态分量, 降噪, 高维特征向量, 经验模态分解, 病害诊断

Abstract:

In view of the problem of difficulty in extracting disease features from non-linear and unsteady sleeper vibration signals, this study proposes a track disease feature extraction method based on variational modal decomposition and multi-scale permutation entropy, and adopts the BP neural network disease diagnosis model to perform disease identification. The variational modal decomposition method is used to decompose the collected vibration acceleration signals to obtain several eigenmode components. The multi-scale permutation entropy value of these eigenmode components is calculated and used as the high-dimensional feature vector of the track disease to realize the noise reduction of the sleeper vibration signal and the extraction of the disease feature. Through the establishment of a BP neural network disease diagnosis model, high-dimensional feature vectors are input into the BP network for training, fitting, and verification, and compared with the method of combining empirical mode decomposition and BP neural network. The analysis results show that the proposed method has a higher recognition accuracy rate and can effectively diagnose disease.

Key words: variational modal decomposition, multi-scale permutation entropy, BP neural network, eigenmode component, noise reduction, high-dimensional feature vector, empirical mode decomposition, disease diagnosis

中图分类号:

TN911.7

华莉,杨俭,袁天辰,宋瑞刚. 基于VMD和BP神经网络的轨道病害诊断方法[J]. 电子科技, 2022, 35(4): 40-46.

Li HUA,Jian YANG,Tianchen YUAN,Ruigang SONG. Track Disease Diagnosis Method Based on VMD and BP Neural Network[J]. Electronic Science and Technology, 2022, 35(4): 40-46.

图/表 10

图1

图2

图3

图4

图5

图6

表1

图7

图8

表2

参考文献 21

[1]	丁宇. 高速铁路无砟轨道疲劳损伤及疲劳可靠性研究[D]. 北京:北京交通大学, 2020.
	Ding Yu. Research on fatigue damage and fatigue reliability of high-speed railway ballastless track[D]. Beijing:Beijing Jiaotong University, 2020.
[2]	丁政开. 基于机器视觉的钢轨表面缺陷检测技术研究[D]. 北京:北京交通大学, 2017.
	Ding Zhengkai. Study of inspection technology for rail surface defects based on machine vision[D]. Beijing:Beijing Jiaotong University, 2017.
[3]	Yang D, Li D, Kuang K S C. Fatigue crack monitoring in train track steel structures using plastic optical fiber sensor[J]. Measurement Science and Technology, 2017, 28(10):1-10.
[4]	Li J W, Li G N. Research on fault diagnosis of ZPW-2000K track circuit based on RS-BN algorithm[J]. Technical Vjesnik, 2019, 26(4):1091-1097.
[5]	乔新勇, 顾程, 韩立军, 等. 基于VMD多尺度散布熵的柴油机故障诊断方法[J]. 汽车工程, 2020, 42(8):1139-1144.
	Qiao Xinyong, Gu Cheng, Han Lijun, et al. Diesel engine fault diagnosis method based on VMD and multi-scale dispersion entropy[J]. Automotive Engineering, 2020, 45(8):1139-1144.
[6]	牟伟杰, 石林锁, 蔡艳平, 等. 基于EMD-WVD与LNMF的内燃机故障诊断[J]. 振动与冲击, 2016, 35(23):191-196.
	Mou Weijie, Shi Linsuo, Cai Yanping, et al. IC engine fault diagnosis method based on EMD-WVD and LNMF[J]. Journal of Vibration and Shock, 2016, 35(23):191-196.
[7]	方必武, 刘涤尘, 王波, 等. 基于小波变换和改进萤火虫算法优化LSSVM的短期风速预测[J]. 电力系统保护与控制, 2016, 44(8):37-43.
	Fang Biwu, Liu Dichen, Wang Bo, et al. Short-term wind speed forecasting based on WD-CFA-LSSVM model[J]. Power System Protection and Control, 2016, 44(8):37-43.
[8]	陈鹏飞, 陈卫, 高星伟, 等. LMD和支持向量机相结合的齿轮毂故障诊断方法[J]. 机械科学与技术, 2015, 34(10):1599-1603.
	Chen Pengfei, Chen Wei, Gao Xingwei, et al. Fault diagnosis of gear hub based on LMD and support vector machine[J]. Mechanical Science and Technology for Aerospace Engineering, 2015, 34(10):1599-1603.
[9]	Li X, Ma Z Q, Kang D, et al. Fault diagnosis for rolling bearing based on VMD-FRFT[J]. Measurement, 2020, 155(4):1-17.
[10]	吕明珠, 苏晓明, 刘世勋, 等. 基于VMD-SVM的滚动轴承退化状态识别[J]. 机械设计与制造, 2020(1):96-100.
	Lü Mingzhu, Su Xiaoming, Liu Shixun, et al. Degradation state recognition of rolling bearing based on VMD-SVM[J]. Machine Design and Manufacturing, 2020(1):96-100.
[11]	Bandt C, Pompe B. Permutation entropy: a natural complexity measure for time series[J]. Physical Review Letters, 2002, 88(17):1-4.
[12]	张建财, 高军伟. 基于变分模态分解和多尺度排列熵的滚动轴承故障诊断[J]. 噪声与振动控制, 2019, 39(6):181-186.
	Zhang Jiancai, Gao Junwei. Fault diagnosis of train rolling bearing based on variational modal decomposition and multi-scale permutation entropy[J]. Noise and Vibration Control, 2019, 39(6):181-186.
[13]	Dragomiretskiy K, Zosso D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3):531-544.
[14]	刘长良, 武英杰, 甄成钢. 基于变分模态分解和模糊C均值聚类的滚动轴承故障诊断[J]. 中国电机工程学报, 2015, 35(13):3358-3365.
	Liu Changliang, Wu Yingjie, Zhen Chenggang. Rolling bearing fault diagnosis based on variational modal decomposition and fuzzy C means clustering[J]. Proceedings of the CSEE, 2015, 35(13):3358-3365.
[15]	王育炜, 韩秋实, 王红军, 等. 滚动轴承VMD能量熵与PNN故障模式识别研究[J]. 组合机床与自动化加工技术, 2020, 4(4):47-50.
	Wang Yuwei, Han Qiushi, Wang Hongjun, et al. Research on VMD energy entropy and PNN fault pattern recognition of rolling bearings[J]. Modular Machine Tool and Automatic Manufacturing Technique, 2020, 4(4):47-50.
[16]	张长青, 杨楠. 基于混合神经网络的车牌字符识别技术[J]. 电子科技, 2019, 32(9):51-54.
	Zhang Changqing, Yang Nan. License plate character recognition technology based on hybrid neural network[J]. Electronic Science and Technology, 2019, 32(9):51-54.
[17]	孟卫东, 刘畅, 张伟, 等. 基于BP神经网络的变压器故障诊断[J]. 通信电源技术, 2020, 37(2):84-86.
	Meng Weidong, Liu Chang, Zhang Wei, et al. Transformer fault diagnosis based on BP neural network[J]. Telecom Power Technology, 2020, 37(2):84-86.
[18]	邓文杰. 基于聚粒子群算法的神经网络权值优化方法[J]. 计算机技术与发展, 2017, 27(10):16-18.
	Deng Wenjie. A neural network weights optimization method based on clustering particle swarm optimization[J]. Computer Technology and Development, 2017, 27(10):16-18.
[19]	翟婉明. 车辆-轨道耦合动力学[M].4版. 北京: 科学出版社, 2007.
	Zhai Wanming. Vehicle-Track coupling dynamics[M].4th ed. Beijing: Science Press, 2007.
[20]	许敬成, 陈长征. BP神经网络在齿轮箱故障诊断中的应用[J]. 噪声与振动控制, 2018, 38(S2):673-677.
	Xu Jingcheng, Chen Changzheng. Application of BP neural network in fault diagnosis of gear boxs[J]. Noise and Vibration Control, 2018, 38(S2):673-677.
[21]	郭珊山, 吴朝晖, 汪庆, 等. 基于小波分析和神经网络的便携式哮喘病监测系统的校准研究[J]. 电子设计工程, 2017, 25(21):90-95.
	Guo Shanshan, Wu Zhaohui, Wang Qing, et al. Calibration study of portable asthma monitoring system based on the wavelet analysis and neural network[J]. Electronic Design Engineering, 2017, 25(21):90-95.

类型	模态	1	2	3	4	5	6	7	8	9	10
正常	IMF₁	0.77	0.83	0.88	0.93	0.97	1.02	1.06	1.09	1.13	1.16
	IMF₂	0.95	1.12	1.27	1.38	1.48	1.56	1.64	1.69	1.74	1.77
	IMF₃	0.86	0.98	1.08	1.17	1.25	1.33	1.39	1.45	1.51	1.56
	IMF₄	0.75	0.80	0.83	0.85	0.88	0.91	0.92	0.91	0.94	0.94
	IMF₅	0.95	1.13	1.27	1.39	1.49	1.57	1.63	1.69	1.72	0.76
板结	IMF₁	0.81	0.89	0.96	1.02	1.08	1.13	1.16	1.20	1.24	1.26
	IMF₂	0.96	1.14	1.28	1.40	1.50	1.59	1.66	1.71	1.75	1.78
	IMF₃	1.01	1.22	1.36	1.47	1.55	1.63	1.68	1.72	1.74	1.77
	IMF₄	1.12	1.37	1.54	1.65	1.69	1.73	1.75	1.76	1.77	1.79
	IMF₅	1.15	1.42	1.60	1.71	1.76	1.79	1.79	1.79	1.79	1.77
翻浆	IMF₁	0.77	0.83	0.88	0.92	0.97	1.01	1.05	1.08	1.12	1.15
	IMF₂	0.95	1.13	1.27	1.39	1.49	1.57	1.64	1.70	1.74	1.77
	IMF₃	0.74	0.77	0.79	0.82	0.84	0.86	0.87	0.89	0.90	0.92
	IMF₄	0.85	0.97	1.07	1.16	1.23	1.31	1.37	1.43	1.48	1.53
	IMF₅	0.96	1.14	1.28	1.41	1.50	1.59	1.66	1.70	1.75	1.76
空吊	IMF₁	0.75	0.79	0.83	0.87	0.90	0.93	0.96	0.99	1.02	1.04
	IMF₂	0.71	0.74	0.75	0.77	0.78	0.80	0.81	0.82	0.83	0.84
	IMF₃	0.96	1.13	1.27	1.40	1.50	1.58	1.65	1.70	1.74	1.77
	IMF₄	0.78	0.84	0.90	0.95	0.99	1.03	1.07	1.11	1.14	1.17
	IMF₅	0.86	0.98	1.08	1.17	1.25	1.32	1.38	1.44	1.49	1.54

病害诊断方法	准确率/%
EMD+MPE-BP	90.03
VMD+MPE-BP	97.27