Design And Implementation Of Vehicle Fault Detection System

The 19 th National Congress of the Communist Party of China has made a major decision to build a transportation power.As an important national strategic,overall,leading and basic facility,rail transit will play an active and indispensable role in the national strategy of building a transportation power.In recent years,China’s rail transit system has made a series of breakthroughs in core key technologies in plateau railway and high-speed railway,and the overall technical level of railway has entered the world’s advanced ranks.With the rapid development of the new generation of scientific and technological revolution and industrial change,the new enabling technologies represented by Internet and artificial intelligence are deeply integrated with rail transit,and safety and comfort have become the future development direction of rail transit system.Therefore,the vehicle fault detection system,which uses the running data of rail train for fault detection to improve the safety and reliability of train operation,arises at the historic moment.In the process of vehicle fault detection system development,the author participates in all the system development links from requirement analysis,summary design,detailed design,coding implementation to system test.In the requirement analysis stage,the author mines the user’s requirements,determines the functional requirements and non functional requirements of the system,and divides the system into three modules:algorithm management module,data management module and system setting module.In the summary design stage,the author determines the B / S architecture of the system based on the requirement analysis,and designs the function module,database and corresponding interface.In the detailed design and implementation stage,according to the requirement analysis and outline design,the author independently completed the development of all modules of the system by using My SQL database,React framework,DVA state management tool,UMI solution and Koa framework.In the algorithm management module,aiming at the mechanical fault of traction motor,the core component of high-speed train traction system,the author investigates and studies the relevant fault detection scheme and virtual sensor technology,analyzes and compares the algorithms,and determines to use stacked autoencoder,In order to realize the fault detection of the motor,the fault detection algorithm model is established based on SAE(Stacked Autoencoder)algorithm,and the improvement is made according to the specific fault characteristics Based on LSTM(Long Short-Term Memory)algorithm,the virtual sensor model is established to achieve the transformation from current data to vibration acceleration data,and the realization of virtual sensor algorithm is completed.Finally,in the test phase,the author independently completed the functional line test and nonfunctional test of each module of the system.At present,the vehicle fault detection system has been put into use in the enterprise,and the system is in good condition,which provides great convenience for vehicle fault detection.