Research On Intelligent Analysis Of High-Speed Railway Vehicle Communication Equipment Health Condition Monitoring Based On PHM Technology

In recent years,with the rapid development of national economy,people's demand for high-quality means of transportation is higher and higher.The development and construction of high-speed railway in China meet the growing demand of people.By the end of 2019,the high-speed railway in China has reached 35000 kilometers,ranking first in the world.Safety has become an important prerequisite for the steady development of high-speed railway in China.Health status monitoring of key components in high-speed railway plays a very important role in the safe and stable operation of high-speed railway.PHM(Prognostics and Health Management)technology can carry out state perception management on high-speed railway running equipment in transit,monitor equipment health status,fault frequency area and cycle,carry out fault diagnosis through data monitoring and analysis and make certain prediction on the occurrence of fault,so as to reduce the risk of fault.In this paper,the intelligent analysis model of health monitoring of high-speed railway vehicle communication equipment based on PHM technology is established by means of fault prediction and health management technology.The research contents include data lossless compression technology,data transmission encryption technology,fault feature extraction method,noise denoising method,key parameter optimization method and neural network method.Although the method realizes fast fault location and accurate prediction,high-performance computer is usually needed to support the calculation and analysis of data,so we will realize the key parts of high-speed railway through the railway wireless communication network,the data will be sent back to the ground analysis server for fault diagnosis and fault prediction,which effectively solves the problem that the high-speed railway on-board equipment does not have highperformance computing ability and cannot process data in real time.After receiving the data,the ground analysis server uses the high-performance computer to decrypt the data and analyze the characteristics.The neural network algorithm can locate the cause of the fault quickly,accurately and effectively,and predict the feasibility of the fault,so as to ensure the healthy operation and condition monitoring of the high-speed railway on-board equipment,the accuracy and efficiency of fault identification and fault prediction are guaranteed,and provide an important technical means to ensure the safe operation of the high-speed railway.The method proposed in this paper has been applied in the project of Lanzhou Railway Administration.In practical application,it also reduced the unrecoverable failure rate and the cost of operation and maintenance in varying degrees.It reflects the significance and value of PHM technology in the research of fault diagnosis and prediction.The main research emphasis are shown as following:(1)Aiming at the safe operation of high-speed railway,combining the theory of fault prediction and health management technology,a PHM framework based on EMU is proposed.This method combines the theory of fault prediction and health management technology,and discusses the PHM technology theory analysis,fault diagnosis technology analysis,prediction technology analysis and application analysis of highspeed Railway on-board communication equipment,and based on fault prediction Based on the theory of measurement and health management technology,the framework of EMU fault prediction and health management system is proposed,which provides important guidance value for the health monitoring and fault prediction of high-speed railway vehicle communication equipment.(2)The network resources of high-speed railway on-board equipment are limited,which can't meet the fault diagnosis and fault prediction ability of on-board equipment.Therefore,a joint algorithm of lossless compression technology and encryption technology is proposed to realize the real-time transmission of the operation data of highspeed railway on-board communication equipment in low bandwidth based on the existing railway wireless network,so that the ground server can analyze the health status of on-board equipment in real time and fault prediction.This algorithm is based on the lossless compression algorithm of travel length to reduce the compression ratio of data.At the same time,it integrates the logistic chaos theory,RSA encryption algorithm and Logistic scrambling encryption algorithm to reduce the network bandwidth occupied by network transmission,improve the security of network transmission,and provide basic guarantee for further analysis of the health status and fault diagnosis and prediction of equipment barrier.(3)Fault diagnosis usually needs to extract and analyze the fault features to remove the mixed noise in the data.Based on the in-depth analysis of the operation information of the on-board equipment transmitted in the previous chapter,this paper finds out the fault problems caused by the field strength signal in the operation of the high-speed railway,and proposes a fault diagnosis method for the on-board communication equipment of the high-speed railway based on PHM technology.This method is improved on the basis of the dual tree complex wavelet packet transform algorithm,and combines the complete empirical mode decomposition of the adaptive noise Noise reduction and feature extraction of fault noise are carried out,and then the types of feature extraction are classified by using the adaptive density clustering method in unsupervised learning.Finally,the classification results are input into the extreme learning machine for training.Experiments show that the fault diagnosis method proposed in this paper has strong feature extraction ability,fast fault recognition ability and high accuracy recognition rate,which makes a good foundation for equipment fault prediction.(4)In-depth analysis of the actual operation conditions of high-speed rail vehicle communication equipment,the high failure rate of high-speed rail vehicle communication equipment is due to the battery failure,which leads to the failure of high-speed rail vehicle communication equipment to work normally.A HA-FOSELM battery failure prediction method is proposed.In order to reduce the influence of noise on prediction,this method adopts the method of variational mode decomposition to denoise.At the same time,it uses the online sequence extreme learning machine algorithm with forgetting mechanism to learn and train the denoised data.This algorithm supports the dynamic incremental updating of data.In order to improve the recognition accuracy of the algorithm and reduce the impact of key parameters on the algorithm,the hybrid gray wolf optimization algorithm is introduced to optimize the key parameters of HA-FOSELM method adaptively,and the attention mechanism is integrated to effectively improve the prediction accuracy.Through the experimental verification,the method proposed in this paper is superior to the traditional neural network algorithm in performance,efficiency,accuracy and other aspects,effectively reducing the problem of train safety caused by battery failure.