Research On Catenary Fault Identification And Status Perception Of Electrified Railway

With the advancement of science and technology and the development of electric energy,more and more railways are upgraded to electrification,and high-voltage electric power is provided by the traction power supply system,of which the most critical equipment is the catenary.The catenary is a special power supply line of the electrified railway.Its structure composition and working mode are very complex.It is usually erected over the rail and arranged in the open air.It is vulnerable to the high-speed impact of the locomotive pantograph.It has become a weak link in the traction power supply system.Once there is a fault,it will directly lead to the interruption of current and voltage transmission between pantograph and catenary,leading to train delay or stop in serious cases,there may even be major accidents.Therefore,it is of great social significance to study the fault identification,performance degradation identification,and status perception of electrified railway catenary to ensure the normal and efficient operation of railway transportation.In this paper,the catenary system of the electrified railway is taken as the main research object,and the data-driven method is mainly used to solve the problems such as imperfect status perception system,the fault caused by single parameter and performance degradation caused by multi-parameters,the following work is completed:(1)The research status of fault identification and status perception of electrified railway catenary in China is described in detail.Based on the references in recent years,firstly,the research background and the research significance of catenary fault identification and status perception are analyzed in this paper,then,the research overview of catenary fault identification and status perception are reviewed,and finally,the main research content and full-text work arrangement of this paper are expounded.(2)The basic methods of fault identification and status perception of electrified railway catenary are reviewed and summarized in this paper.Firstly,in order to have a clear understanding of catenary,the basic structure and common faults of catenary are introduced in this paper.Secondly,the methods of catenary fault recognition are expounded,and the process of catenary fault recognition from three aspects: the method based on statistics,the method based on image vision technology and the method based on machine learning are described,respectively.Finally,the methods of catenary status perception and their principles or characteristics are expounded.(3)A method of catenary status perception based on normal fuzzy matter-element and game theory is proposed.Aiming at the problems of incomplete indicators in the current catenary status perception system and human factors in the way of indicators weighting,a method of catenary status perception is proposed,which is combined game theory with normal fuzzy matter-element.Firstly,a status perception system including catenary safety indicators,catenary ride comfort indicators,pantograph catenary current collection performance indicators,weather indicators and historical operation condition is constructed.Then,the entropy weight method and PSO-AHP method are used to solve the subjective and objective weights of quantitative indicators,and the game theory is used to give the weight coefficient,and the PSO-AHP method is used to solve the weight of qualitative indicators.Finally,in view of the failure of the maximum relevance principle,the weighted average relevance principle is used to improve,and the final catenary status perception level is determined.The results show that the proposed status perception system is more scientific and reasonable,and can accurately perceive the status of the catenary.(4)A fault identification method of catenary based on Dynamic Collaborative Update Strategy Improved PSO-ELM is proposed.In view of the current catenary parameters are numerous,and these parameters will inevitably have faults,and the current catenary fault detection method is backward,and the fault detection time is long,a method of catenary fault identification based on DCUSPSO-ELM is proposed in this paper,which is based on intelligent algorithm and using the catenary supervised learning data.Firstly,the fault data and normal data of several typical parameters in catenary are selected,and the design of the indicators of each category is completed;Then,and the fault identification model of ELM optimized by Dynamic Collaborative Update Strategy Improved Particle Swarm Optimization is constructed;Finally,the data is divided into training samples and test samples in order to complete the accurate research of catenary fault identification.(5)A performance degradation identification method of catenary based on GG clustering and improved SVDD is proposed.Aiming at the problems of low utilization rate of catenary detection data,imperfect research on catenary performance degradation caused by multiple parameters,and most of the data are unsupervised learning data,which are not conducive to research and analysis,a catenary performance degradation identification method based on GG clustering and improved support vector data description model optimized by PSO algorithm(GG-PSO-ISVDD)is proposed in this paper.Firstly,semi-supervised learning data are selected,and TSNE dimension reduction method is used to reduce dimensionality,and GG clustering method is used to cluster the unlabeled data,and the optimal number of clusters is obtained.Then,a PSO optimized ISVDD catenary performance degradation model is constructed by using labeled data,which contains as much normal data as possible and as few abnormal data as possible,and the radius of the established hypersphere model is as small as possible.Finally,the distance between the clustering sample data and the center of the hypersphere is calculated,and the distance and radius are compared to determine the degree of catenary performance degradation.The results show that the proposed method can accurately judge the performance degradation of catenary,which provides a powerful reference for the research of catenary performance degradation identification.