Fault Diagnosis Of Locomotive Braking System Based On Data Mining

Transportation is the weapon of the country,the root of the country.As a kind of transportation,railway transportation has become an important way for people to travel and transport goods by virtue of its low cost and low environmental impact.However,it is difficult to predict the consequences of locomotive failure,which can lead to either expensive maintenance or serious death.The locomotive braking system is the core structure of the locomotive,so it is of great significance to develop a set of fault diagnosis expert system to meet the needs of the locomotive braking system.In addition,locomotives staff leaves a detailed diagnostic record after repairing the locomotive,this data is very precious,but in the real work,it is often redundant and seldom used after recording.the goal of this article is to use some data mining algorithms to help staff find fault and repair maintenance personnel faster.First introduces the safety protection system(6A system)and main faults of locomotive on-boardand braking system in detail,selects the appropriate fault features,and generates locomotive fault data after data preprocessing.Then the locomotive fault data is used to train the support vector machine(SVM)model,and the locomotive data is classified in advance,which can screen out a large number of abnormal and meaningless data.Finally,an improved Apriori algorithm based on matrix optimization is proposed to mine the correlation between locomotive features and faults,and a fault analysis tree is constructed.In order to verify the correctness and feasibility of the research content,the locomotive fault diagnosis expert system is designed and implemented,the system includes locomotive data management module,knowledge base management module and fault diagnosis module,can automatically diagnose malfunction and gives the corresponding maintenance,improve the maintenance efficiency of locomotives staff.For the common faults of locomotive braking system,the system can always correctly infer the fault conclusion,which has certain practicability.