Research On Fault Diagnosis Method For CTCS On-board Equipment Of High-speed Railway Based On Association Rules

With the increasing speed of the High-speed train,on-board equipment plays an important role in train control system.The failure of on-board equipment would directly affect the safety of train operation.During the running of the train,the failure of on-board equipment occurred frequently.However,at present,the fault diagnosis technology of on-board equipment is over dependent on staff experience and expert knowledge,so it cannot meet the needs of High-speed running in Chinese modern railways.Therefore,it is of great significance for the intelligent research of on-board equipment fault diagnosis.In view of the characteristics of the fault data of train control on-board equipment,on the basis of in-depth research for the structure and function of on-board equipment,this thesis puts forward a fault diagnosis method of the train control on-board equipment based on the association rules.The corresponding relationship between fault characteristics and failure modes is established,which realize the intelligent analysis of on-board equipment fault diagnosis.The main work of this thesis is as follows:(1)Fault feature extraction:The characteristics of on-board equipment fault data are analyzed in this thesis.In view of the characteristics of these data,the fault feature library of the on-board equipment will be constructed,and the fault text data of the on-board equipment can be vectorized by using the TF-IDF feature extraction method.The attribute weights of fault feature words in each fault text are calculated,and it is discretized and dimensionality reduction processing.Finally,the fault decision table is generated to provide data basis for the subsequent classification of association rules.(2)Fault diagnosis:The fault diagnosis model based on association rules is constructed.Through the text mining for the historical fault text,a fault diagnosis rule base is established to provide basis for fault diagnosis.FP-Growth association rule algorithm is selected as the algorithm of fault classification in this thesis.Combined with specific examples,the implementation process of the algorithm is introduced in detail.Concrete fault diagnosis association rules are obtained,and rules are explained and illustrated.(3)Experimental verification and analysis:Through data training and data testing,the first level fault and the second level fault ratio of the model diagnosis are obtained.And the evaluation indexes of the model,such as Precision,Recall and F1-measure,are analyzed.So,the validity and accuracy of the model are verified.Finally,the fault diagnosis system of on-board equipment is designed by the mixed programming of MATLAB and C#.