作者：王晓东, 宁静, 陈春俊

作者单位：1. 西南交通大学机械工程学院，四川 成都 610031;
2. 西南交通大学 轨道交通运维技术与装备四川省重点实验室，四川 成都 610031

关键词：高速列车;横向稳定性;小幅蛇行;状态识别;1D CNN;LSTM

摘要：

高速列车横向稳定性对列车的行车安全有重大影响，针对列车高速运行时出现的小幅蛇行和蛇行失稳问题，提出基于1D CNN和LSTM的识别方法。以高速列车构架横向加速度信号为研究对象，通过1D CNN自适应地对信号进行特征提取，避免手动提取特征的局限性，经1D CNN提取的特征信号作为LSTM的输入，充分利用LSTM学习加速度信号时间维度上的信息，最后通过全连接层输出识别结果。实验结果表明：基于1D CNN和LSTM的方法能准确识别小幅蛇行、蛇行失稳和正常状态，3种状态的识别率均为100%。

Study on the recognition of lateral stability state of high-speed trains based on 1D CNN and LSTM
WANG Xiaodong, NING Jing, CHEN Chunjun
1. School of Mechanical Engineering, Southwest Jiaotong University, Chengdu 610031, China;
2. Technology and Equipment of Rail Transit Operation and Maintenance Key Laboratory of Sichuan Province,Southwest Jiaotong University, Chengdu 610031, China
Abstract: The lateral stability has a significant impact on the running safety of high-speed trains. Aiming at the problems of small amplitude hunting and hunting instability in high-speed trains, a recognition method based on 1D CNN and LSTM is proposed. 1D CNN adaptively extracts the features from the lateral acceleration signal of the high-speed strain frame, avoiding the limitation of manual feature extraction. The feature information extracted by 1D CNN is used as the input of LSTM, and the information in time dimension of the acceleration signal is learned by LSTM. Finally, the recognition result is obtained through the fully connected layer. The experimental results show that the method based on 1D CNN and LSTM can accurately identify small amplitude hunting, hunting instability and normal state. The recognition rate of the three states is 100%.
Keywords: high-speed train;lateral stability;small amplitude hunting;state recognition;1D CNN;LSTM
2020, 46(11):25-30  收稿日期: 2020-04-20;收到修改稿日期: 2020-05-19
基金项目: 国家自然科学基金项目（51975486，51975487）
作者简介: 王晓东(1992-),男,江西赣州市人,硕士研究生,专业方向为智能化状态监测及故障诊断
参考文献
[1] 张卫华, 李艳, 宋冬利, 等. 高速列车运动稳定性设计方法研究[J]. 西南交通大学学报, 2013, 48(1): 1-9
[2] 池茂儒, 张卫华, 曾京, 等. 蛇行运动对铁道车辆平稳性的影响[J]. 振动工程学报, 2008, 21(6): 639-643
[3] OKAMOTO H, KOTANI M, LIN S P, et al. Stabilization control for hunting motion of railway vehicle by gyroscopic damper (experimental analysis using 1/10 scale vehicle model)[J]. Transactions of the Japan Society of Mechanical Engineers, 2008, 74(738): 235-241
[4] 高速铁路工程动态验收技术规范: TB10761—2013[S]. 北京: 中国铁道出版社, 2013.
[5] 铁道客车行车安全监测诊断系统技术条件: TB/T 3188—2007[S]. 北京: 中国铁道出版社, 2007.
[6] 孙丽霞. 高速列车横向运动稳定性和动态脱轨理论分析及评价方法[D]. 北京: 中国铁道科学研究院, 2014.
[7] 孙永奎, 余志斌, 金炜东. 基于多重分形与SVM的高速列车运行状态识别方法[J]. 西南交通大学学报, 2015, 50(1): 7-12
[8] 秦娜, 金炜东, 黄进, 等. 基于EEMD样本熵的高速列车转向架故障特征提取[J]. 西南交通大学学报, 2014, 49(1): 27-32
[9] 刘峰, 延九磊, 董孝卿, 等. 转向架横向稳定性的实时在线监测[J]. 铁道机车车辆, 2010, 30(2): 5-8
[10] 叶运广, 宁静, 种传杰, 等. 基于MEEMD香农熵-LSSVM的高速列车蛇行失稳诊断方法[J]. 计算机应用研究, 2017, 34(4): 1097-1100
[11] 宋欣武. 转向架蛇行失稳监测的横向位移识别方法研究[D]. 成都: 西南交通大学, 2017.
[12] 刘棋, 宁静, 叶运广, 等. 基于D-S证据理论高速列车横向失稳故障判别研究[J]. 中国测试, 2017, 43(7): 103-107
[13] BOUDIAF A, MOUSSAOUI A, DAHANE A, et al. A comparative study of various methods of bearing faults diagnosis using the case Western Reserve University data[J]. Journal of Failure Analysis and Prevention, 2016, 16(2): 271-284
[14] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444
[15] KIRANYAZ S, INCE T, HAMILA R, et al. Convolutional neural networks for patient-specific ECG classification[C]//2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 2015: 2608-2611.
[16] KIRANYAZ S, AVCI O, ABDELJABER O, et al. 1D convolutional neural networks and applications: A survey[J/OL]. Electrical Engineering and Systems Science, 2019[2020-03-21]. https://arxiv.org/abs/1905.03554.
[17] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural computation, 1997, 9(8): 1735-1780
[18] LIPTON Z C, BERKOWITZ J, ELKAN C. A critical review of recurrent neural networks for sequence learning[J/OL]. Computer Science, 2015[2020-03-21].https://arxiv.org/abs/1506.00019.
[19] NIELSEN M A. Neural networks and deep learning[M]. San Francisco, CA, USA: Determination press, 2015.
[20] 叶运广, 宁静, 种传杰, 等. 高速列车转向架蛇行失稳的MEEMD-LSSVM预测模型[J]. 铁道学报, 2018(1): 38-43
[21] 蔡里军. 基于FPGA的高速列车转向架蛇行失稳检测装置[D]. 成都: 西南交通大学, 2012.