Research On Intelligent Analysis Model Of High-speed Railway Equipment Health Condition Based On PHM Technology

Prognostic and health management have been widely studied and applied in military,nuclear power and other fields.However,in high-speed railway field,which requires high safety and reliability,further research is needed on prognostic and health management.Prognostic health management technology of high speed railway equipment by detecting and monitoring the health status of key parts and components,assessing existing faults and supporting the maintenance of high-speed railway equipment,it is of great significance to avoid catastrophic accidents,to ensure the safety of high-speed railway equipment and to maximize the benefits of the system.In this dissertation,a general theory model of prognostic and health management for high-speed railway equipment is proposed,and two important problems in the health analysis of fault prediction and health management of the high-speed EMU are studied.The two questions are feature extraction of condition monitoring and health state assessment.Finally,bearing as an example is analyzed.It is of great theoretical and practical significance to explore and construct a general theoretical PHM model for various components,systems and equipment.The result of feature extraction from condition monitoring is a compressed representation of health condition,which is often used as a preprocessing step of health condition analysis task.Health condition assessment refers to a certain method and measures to evaluate the health condition of high speed EMU as a whole or parts in order to facilitate users to make appropriate use or maintenance decisions.As the key technology of prognostic and health management,health condition assessment is the premise and foundation of health management.Based on the theory of wavelet analysis,dynamics and artificial neural network,a model of rolling element bearing fault prediction and health management for traction motor of high-speed EMU is established using traction motor rolling element bearing monitoring big data as the research object.In order to realize health condition monitoring and health condition evaluation of high speed railway equipment,some theoretical and practical problems of prognostic and health management based on big data are solved.The main contributions of this thesis are as follows:(1)From the point of view of high-speed railway,combined with the practice of developing prognostic and health management algorithm,this paper puts forward the theoretical model of prognostic and health management for high-speed railway equipment.Based on the theoretical model,the functional model and system architecture of prognostic and health management for high-speed EMU are constructed.The analysis is expected to provide theoretical reference and practical guidance for the healthy development of high-speed railway prognostic and health management in China.(2)Recently,some researchers have used combinatorial methods to improve the accuracy of condition monitoring.However,the computational efficiency and robustness are not taken into account emphatically.In this paper,a hybrid model of feature extraction based on dual-tree complex wavelet packet transform and variational mode decomposition is proposed.This model integrates the wavelet analysis technology and the variational mode decomposition into a framework.The model is especially suitable for monitoring early fault signals of rolling bearings under strong noise background.The validity of the dual-tree complex wavelet packet transform and variational mode decomposition model for monitoring early fault of rolling bearings is verified by experiments on standard CWRU database and simulation experiments.(3)This paper presents a health condition assessment combined model for online sequential extreme learning machine based on Tensor Flow.The performance of TOSELM algorithm is evaluated by health status accuracy index.In order to verify the validity of the proposed model,a series of experiments were carried out on public data sets and high-speed EMU traction motor ball bearing data sets using BP,ELM and TOSELM models.The results show that TOSELM model is an effective bearing health assessment model.The proposed model can improve the computational efficiency and the reliability of the analysis results using big data.This model solves the problem of health assessment of ball bearings of traction motors of high-speed EMUs,thus ensuring the safety of high-speed train operation.The above research results can be applied to the fault prediction and health management system of high-speed railway equipment and in-service health monitoring,and provide technical support for further optimizing the condition based maintenance strategy of high-speed EMUs.It has important engineering significance and theoretical value for improving the reliability and safety of high-speed railway equipment and reducing the cost of operation and maintenance.