高速列车轴承是列车转向架上的重要部件，其承受着复杂的激励和多变的工况。轴承温度，作为衡量轴承状况的状态量之一，现有方法通过设定阈值判别能够避免轴承故障威胁到火车安全性。因此，对于运行中轴承的温度进行预测和监测，建立一个实时的轴承温度模型，并辅设预警决策，可以有效地提升轴承运行安全并提早诊断发现故障。但针对铁路所采集的轴承状态监测大数据呈现多维特性，并缺少典型的轴承故障类型数据，难以提前对轴承等关键部件进行故障预警，对其进行的研究也少有探究各个轴承之间、关键部件之间、各轴之间状态监测数据存在的线性、非线性耦合关系。为
此，本论文开展相关研究，主要研究工作如下：
（1）提出了一种基于复杂相关性的高速列车轴承温度预测方法。首先分析列车履历数据，使用 Pearson 相关系数分析从而引入高线性相关的关联轴承，将低线性相关特征的数据和关联轴承数据输入 LightGBM 模型，对影响轴承温度的特征进行再次筛选后降维，使模型可以学习到更有价值的特征；其次，基于深度神经网络模型双向门控循环单元建立轴承温度预测模型，通过大量正常轴承的履历数据来训练预测模型；最后进行方法验证。在多个预测的关键参数上进行对比验证得到结论，采用短时预测策略、关联测点进行预测的方式，能够最有效地提高模型的精度和稳定性。
（2） 提出了一种基于多任务学习的高速列车轴承温度异常检测方法。首先把正常工况下的关联轴承温度作为模型输入构建轴承温度预测模型，当实际温度异常时预测值与实际值关联性呈现异常变化。其次，引入多头注意力机制，所构建的模型能够同时对一轴上的轴箱、齿轮箱、电机三类共九个轴承温度进行同时预测。最后建立孤立森林模型，结果证明孤立森林模型具有同时对多个轴承同时进行异常检测的能力。
（3）提出了一种基于置信区间的高速列车轴承温度异常检测方法。结合不确定性的思想，将随机不确定性、认知不确定性、以及混合不确定性等三类不确定性思想添加到模型中，将点预测模型转换为概率预测模型，求得温度的分布概率，通过概率计算的方式直观辅助模型进行轴承温度异常检测。进行实验验证表明，概率预测模型能够实现端对端的直接检测，消除了模型构建后再收集预测值与实际值残差进行二次建模的精度耗损，达到更高置信度。
（4）开发了一套高速列车轴承温度异常检测与预警系统。该系统集成了上述研究成果，可基于高速列车状态检测参数进行轴承异常检测与预警。系统应用实例表明，该系统达到了预期的设计功能。

Bearing of high-speed train is an important part of train bogie, which bears complex excitation and changeable working conditions. Bearing temperature, as one of the state variables to measure the bearing condition, can avoid bearing failure threatening the train safety by setting threshold. Therefore, the prediction and monitoring of the bearing temperature in operation, the establishment of a real-time bearing temperature model, with the aid of early warning decision-making, can effectively improve the safety of the bearing operation and early diagnosis of fault.
However, the bearing condition monitoring big data collected by railways presents multi dimensional characteristics, and lacks typical bearing fault type data. It is difficult to early warning of faults of key components such as bearings, and there are few studies on them that explore the key points of each bearing the linear and non-linear coupling relationship between the components and the condition monitoring data of each axis. To this end, this thesis carries out related research, the main research work is as follows:
(1) A bearing temperature prediction method for high-speed train based on complex correlation is proposed. Firstly, train history data is analyzed, and Pearson correlation coefficient is used to analyze the correlation bearing with high linear correlation. The data with low linear correlation and correlation bearing data are input into LightGBM model, and the features that affect bearing temperature are filtered again to reduce dimension, so that the model can learn more valuable features. Secondly, based on the deep neural network model, the bidirectional Gated Recurrent Unit is realized the bearing temperature prediction model is established by the ring unit, and the prediction model is trained by a large number of normal bearing history data; finally, the method is verified. It is concluded that the short-term prediction strategy and the method of correlated measurement points can effectively improve the accuracy and stability of the model.
(2) A method for detecting abnormal bearing temperature of high-speed train based on Multi-Task Learning is proposed. First, the associated bearing temperature under normal operating conditions is used as the model input to construct a bearing temperature prediction model. When the actual temperature is abnormal, the correlation between the predicted value and the actual value presents an abnormal change. Secondly, with the introduction of a Multi-Head Attention mechanism, the constructed model can simultaneously predict the temperature of nine bearings in three categories: axle box, gear box, and motor on one axle. Finally, an  isolated forest model is established, and the results prove that the isolated forest model has the ability to detect anomalies on multiple bearings at the same time.
(3) A method of bearing temperature anomaly detection for high-speed train based on confidence interval is proposed. Combined with the idea of uncertainty, the Aleatoric Uncertainty, Epistemic Uncertainty and Mixture Uncertainty are added to the model. The point prediction model is transformed into the probability prediction model, and the distribution probability of temperature is obtained. Through probability calculation, the bearing temperature anomaly detection is directly assisted by the model. The experimental results show that the probabilistic prediction model can realize the end-to-end direct detection, eliminate the accuracy loss of secondary modeling by collecting the residual between the predicted value and the actual value after the model is constructed, and achieve higher confidence.
(4) A set of high-speed train bearing temperature anomaly detection and early warning system is developed. The system integrates the above research results, and can detect and warn the abnormal bearing based on the state detection parameters of high-speed train. The application of the system shows that the system achieves the expected design function.

[1].国家铁路局介绍铁路领域“十三五”发展成就[J]. 铁道技术监督,2020,48(11):53.

[2].翟婉明,王开云,陈建政. 铁路货车横向非线性动态行为的理论与试验研究[J]. 机械工程学报,2008,44(11):138-144.

[3].王华胜. 机车车辆区间故障数据置信限近似估计方法[J]. 铁道学报,2012,34(05):15-19.

[4].刘剑锋,刘友梅,桂卫华,刘豫湘,黄志武. 基于模糊预测控制的机车制动控制方法[J]. 中南大学学报(自然科学版),2009,40(05):1329-1335.

[5].赵阳,徐田华,周玉平,赵文天. 基于贝叶斯网络的高铁信号系统车载设备故障诊断方法的研究[J]. 铁道学报,2014,36(11):48-53.

[6].汤武初,王敏杰,陈光东,孙玉超,许立. 高速列车故障轴箱轴承的温度分布研究[J]. 铁道学报,2016,38(07):50-56.

[7].尹诗,侯国莲,于晓东,李宁,王其乐,弓林娟.基于Bi-RNN的风电机组主轴承温度异常检测方法研究[J]. 郑州大学学报(工学版),2019,40(05):45-51.

[8].Xu G, Hou D, Qi H, et al. High-speed train wheel set bearing fault diagnosis and prognostics: A new prognostic model based on extendable useful life[J]. Mechanical Systems and Signal Processing, 2021, 146: 107050.

[9].Chandola V, Banerjee A, Kumar V. Outlier detection: A survey[J]. ACM Computing Surveys, 2007, 14: 15.

[10].Hawkins D M. Identification of outliers[M]. London: Chapman and Hall, 1980.

[11].Xie X, Wang C, Chen S, et al. Real-time illegal parking detection system based on deep learning[C]//Proceedings of the 2017 International Conference on Deep Learning Technologies. 2017: 23-27.

[12].Schlegl T, Seeböck P, Waldstein S M, et al. Unsupervised anomaly detection with generative adversarial networks to guide marker discovery[C]//International conference on information processing in medical imaging. Springer, Cham, 2017: 146-157.

[13].Javaid A, Niyaz Q, Sun W, et al. A deep learning approach for network intrusion detection system[C]//Proceedings of the 9th EAI International Conference on Bio-inspired Information and Communications Technologies (formerly BIONETICS). 2016: 21-26.

[14].Mohammadi M, Al-Fuqaha A, Sorour S, et al. Deep learning for IoT big data and streaming analytics: A survey[J]. IEEE Communications Surveys & Tutorials, 2018, 20(4): 2923-2960.

[15].Peng H K, Marculescu R. Multi-scale compositionality: identifying the compositional structures of social dynamics using deep learning[J]. PloS one, 2015, 10(4): e0118309.

[16].张杰豪. 基于深度学习的行为检测[D].杭州电子科技大学,2019.

[17].胡正平,张乐,尹艳华. 时空深度特征AP聚类的稀疏表示视频异常检测算法[J]. 信号处理,2019,35(03):386-395.

[18].Andrews J T A, Morton E J, Griffin L D. Detecting anomalous data using auto-encoders[J]. International Journal of Machine Learning and Computing, 2016, 6(1): 21.

[19].Chalapathy R, Toth E, Chawla S. Group anomaly detection using deep generative models[C]//Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, Cham, 2018: 173-189.

[20].Horobets O O. Big Data of Source of Statistical Information: On the Example of the Book Publishing Industry[J]. Науковий вісник Національної академії статистики, обліку та аудиту, 2019 (1-2): 7-13.

[21].Zhai Y, Ong Y S, Tsang I W. The emerging" big dimensionality"[J]. IEEE Computational Intelligence Magazine, 2014, 9(3): 14-26.

[22].Thudumu S, Branch P, Jin J, et al. Adaptive clustering for outlier identification in high-dimensional data[C]//International Conference on Algorithms and Architectures for Parallel Processing. Springer, Cham, 2019: 215-228.

[23].Erfani S M, Rajasegarar S, Karunasekera S, et al. High-dimensional and large-scale anomaly detection using a linear one-class SVM with deep learning[J]. Pattern Recognition, 2016, 58: 121-134.

[24].Niu Z, Shi S, Sun J, et al. A survey of outlier detection methodologies and their applications[C]//International Conference on Artificial Intelligence and Computational Intelligence. Springer, Berlin, Heidelberg, 2011: 380-387.

[25].Liu H, Li X, Li J, et al. Efficient outlier detection for high-dimensional data[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2017, 48(12): 2451-2461.

[26].黄星寿,赵培信. 高维线性均值漂移模型的异常值检测[J]. 统计与决策,2019,35(14):68-70.

[27].丁望祥. 多变量时间序列数据聚类和异常检测算法研究[D].南京大学,2020.

[28].徐晓丹. 高维数据的异常检测算法研究[D].浙江工业大学,2020.

[29].Zenati H, Foo C S, Lecouat B, et al. Efficient gan-based anomaly detection[J]. arXiv preprint arXiv:1802.06222, 2018.

[30].Zenati H, Romain M, Foo C S, et al. Adversarially learned anomaly detection[C]//2018 IEEE International conference on data mining (ICDM). IEEE, 2018: 727-736.

[31].Boutros T, Liang M. Detection and diagnosis of bearing and cutting tool faults using hidden Markov models[J]. Mechanical Systems and Signal Processing, 2011, 25(6): 2102-2124.

[32].Jack L B, Nandi A K. Fault detection using support vector machines and artificial neural networks, augmented by genetic algorithms[J]. Mechanical systems and signal processing, 2002, 16(2-3): 373-390.

[33].Liu T I, Singonahalli J H, Iyer N R. Detection of roller bearing defects using expert system and fuzzy logic[J]. Mechanical systems and signal processing, 1996, 10(5): 595-614.

[34].Chen J, Pan J, Li Z, et al. Generator bearing fault diagnosis for wind turbine via empirical wavelet transform using measured vibration signals[J]. Renewable Energy, 2016, 89: 80-92.

[35].Zhang W, Li C, Peng G, et al. A deep convolutional neural network with new training methods for bearing fault diagnosis under noisy environment and different working load[J]. Mechanical Systems and Signal Processing, 2018, 100: 439-453.

[36].Nishimoto S, Saegusa T, Fujimoto Y. Method and apparatus for predicting destruction of a rolling bearing: U.S. Patent 5,001,931[P]. 1991-3-26.

[37].Yang B S, Han T, Hwang W W. Fault diagnosis of rotating machinery based on multi-class support vector machines[J]. Journal of Mechanical Science and Technology, 2005, 19(3): 846-859.

[38].Pandya D H, Upadhyay S H, Harsha S P. Fault diagnosis of rolling element bearing with intrinsic mode function of acoustic emission data using APF-KNN[J]. Expert Systems with Applications, 2013, 40(10): 4137-4145.

[39].Tra V, Kim J, Khan S A, et al. Bearing fault diagnosis under variable speed using convolutional neural networks and the stochastic diagonal levenberg-marquardt algorithm[J]. Sensors, 2017, 17(12): 2834.

[40].Wang W J, Cui L L, Chen D Y. Multi-scale morphology analysis of acoustic emission signal and quantitative diagnosis for bearing fault[J]. Acta Mechanica Sinica, 2016, 32(2): 265-272.

[41].Jardine A K S, Lin D, Banjevic D. A review on machinery diagnostics and prognostics implementing condition-based maintenance[J]. Mechanical systems and signal processing, 2006, 20(7): 1483-1510.

[42].Rai A, Upadhyay S H. A review on signal processing techniques utilized in the fault diagnosis of rolling element bearings[J]. Tribology International, 2016, 96: 289-306.

[43].Lin T R, Yu K, Tan J. Condition monitoring and fault diagnosis of roller element bearing[J]. INTECH, 2017: 39-75.

[44].Center B D. Case western reserve university bearing data[J]. 2011-12-10]. http://csegroups, case. ed./bearingdatacenter/pages/down-loaddata-file, 2011.

[45].Bearing Datacenter, Paderborn University. Accessed: Dec. 2018, [Online]. Available: https://mb.unipaderborn.de/kat/forschung/datacenter/bearing-datacenter/.

[46].Yuan Y, Chen C. Fault detection of rolling bearing based on principal component analysis and empirical mode decomposition[J]. AIMS Mathematics, 2020, 5(6): 5916-5938.

[47].Hao Q, Shen Y, Wang Y, et al. An adaptive extraction method for rail crack acoustic emission signal under strong wheel-rail rolling noise of high-speed railway[J]. Mechanical Systems and Signal Processing, 2021, 154: 107546.

[48].Hoang D T, Tran X T, Van M, et al. A Deep Neural Network-Based Feature Fusion for Bearing Fault Diagnosis[J]. Sensors, 2021, 21(1): 244.

[49].Su H, Song M, Xiong Z. Bearing fault diagnosis method based on EEMD and adaptive redundant lifting scheme packet[J]. Vibroengineering PROCEDIA, 2020, 34: 14-19.

[50].Jiang Y, Wu P, Zeng J, et al. Detection and alleviation of the abnormal vibration of the monorail based on experiment and simulation[J]. Journal of Low Frequency Noise, Vibration and Active Control, 2019, 38(2): 282-295.

[51].Wei X, Guo Y, Jia L, et al. Fault detection of rail vehicle suspension system based on CPCA[C]//2013 Conference on Control and Fault-Tolerant Systems (SysTol). IEEE, 2013: 700-705.

[52].Xia Z, Zhou J, Liang J, et al. Online detection and control of car body low-frequency swaying in railway vehicles[J]. Vehicle System Dynamics, 2021, 59(1): 70-100.

[53].Kou L, Qin Y, Zhao X, et al. Integrating synthetic minority oversampling and gradient boosting decision tree for bogie fault diagnosis in rail vehicles[J]. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, 2019, 233(3): 312-325.

[54].Chong L, Jiahong W, Zhixin Z, et al. Design and evaluation of a remote measurement system for the online monitoring of rail vibration signals[J]. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, 2016, 230(3): 724-733.

[55].刘强,方彤,董一凝,秦泗钊. 基于动态建模与重构的列车轴承故障检测和定位[J]. 自动化学报,2019,45(12):2233-2241.

[56].崔秀国. CRH3型动车组电气系统可靠性研究[D].北京交通大学,2013.

[57].刘德东. CRH2及CRH380A(L)系列动车组轴温实时检测系统故障分析及对策[J]. 中国高新区,2018(10):175.

[58].WANG, Zhiwei, et al. Effect of vehicle vibration environment of high-speed train on dynamic performance of axle box bearing. Vehicle System Dynamics, 2019, 57.4: 543-563.

[59].梁颖,方瑞明. 基于SCADA和支持向量回归的风电机组状态在线评估方法[J]. 电力系统自动化,2013,37(14):7-12+31.

[60].陈德生. 列车轴温规律及红外线轴温探测方式的研究[D].哈尔滨工程大学,2003.

[61].CHENG, Yao; WANG, Zhiwei; ZHANG, Weihua. A novel condition-monitoring method for axle-box bearings of high-speed trains using temperature sensor signals. IEEE Sensors Journal, 2018, 19.1: 205-213.

[62].罗怡澜. 高速列车轴承异常温升预警方法研究[M] 成都:西南交通大学，2018.

[63].谢国, 王竹欣, 黑新宏, 高橋聖, 望月宽. 面向热轴故障的高速列车轴温阈值预测模型[J] 交通运输工程学报, Vol.03,129-137, 2018.

[64].Ma W, Tan S, Hei X, et al. A prediction method based on stepwise regression analysis for train axle temperature[C]//2016 12th International Conference on Computational Intelligence and Security (CIS). IEEE, 2016: 386-390.

[65].Sun L, Hei X, Xie G, et al. Data Based Fault Diagnosis of Hot Axle for High-Speed Train[C]//2018 Chinese Automation Congress (CAC). IEEE, 220-225.

[66].祁明明. 基于NSET的高速列车轴箱轴承温度异常检测研究[D].兰州交通大学,2020.

[67].宋佳音. 高速列车轴承温升趋势预测方法研究[D].兰州交通大学,2020.

[68].Luo C, Yang D, Huang J, et al. LSTM-based temperature prediction for hot-axles of locomotives[C]//ITM web of conferences. EDP Sciences, 2017, 12: 01013.

[69].Wang X, Liu F, Chen Y. Application of Spatio-Temporal Attention Mechanism in Temperature Prediction of High-Speed Train Bogie[C]//2019 Prognostics and System Health Management Conference (PHM-Paris). IEEE, 2019: 327-331.

[70].曹源,王玉珏,马连川,陈磊. 基于DTW的车辆轴温监测方法[J]. 交通运输工程学报,2015,15(03):78-84+100.

[71].杨云,薛元贺. 基于动态时间规整和证据理论的机车轴温监测预警研究[J]. 铁道科学与工程学报,2020,17(03):714-721.

[72].刘强,詹志强,王硕,刘英翔,方彤. 数据驱动的高速列车轴承多模态运行监控与故障诊断[J]. 中国科学:信息科学,2020,50(04):527-539.

[73].王竹欣. 基于数据的高速列车轴温动态阈值估计及故障诊断[D].西安理工大学,2018.

[74].谭思雨. 高速列车车轴故障预警与诊断方法研究[D].西安理工大学,2019.

[75].WANG, Chenchong, et al. Tensile property prediction by feature engineering guided machine learning in reduced activation ferritic/martensitic steels[J]. Journal of Nuclear Materials, 2020, 529: 151823.

[76].曾宇. 滚动轴承多域退化表征指标提取及寿命预测研究[D].北京交通大学,2019.

[77].黄阳. 基于振动信号降噪与分解的轴承故障诊断研究[D].西安理工大学,2018.

[78].Ke G, Meng Q, Finley T, et al. Lightgbm: A highly efficient gradient boosting decision tree[J]. Advances in neural information processing systems, 2017, 30: 3146-3154.

[79].Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//Advances in neural information processing systems. 2017: 5998-6008.

[80].Sener O, Koltun V. Multi-task learning as multi-objective optimization[C]//Advances in Neural Information Processing Systems. 2018: 527-538.