Research Of Fault Diagnosis And Prediction On LED Lighting System Of Railway Vehicle Compartment

With the rapid development of China’s railway vehicles, People are more concerned about the reliability of its operation. Railway vehicles’LED illuminating system is a new lighting system can both extend the life of the lighting system, but also improve the resistance to shock.Meanwhile, energy saving effect is obvious.but it also increase the complexity of the railway vehicles’illuminating system.People become more and more concerned about how to protect the safety of the railway vehicles’LED illuminating system and its reliability.Fault diagnosis and forecasting techniques is to ensure the normal operation of the railway vehicles’ LED illuminating system,and to be an effective way to protect the safety and orderly traffic of passengers.In the support of national natural science foundation project "research on the maintenance of complex electromechanical system fault prediction and optimal based on Virtual Reality Technology". Researches on the fault diagnosis and prediction problems in the railway vehicles’LED illuminating system were carried out.First, LED illuminating system for railway vehicles’failure mechanism were analyzed, the working mechanism of the LED control circuit module, power driving module, lamp panel module and brightness acquisition module and common fault types and reasons wered analyzed.Analyses of possible faults of LED illuminating system and electronic components which are subject to failure, classification of LED illuminating system for railway vehicles common failures and detailed analyses of the failure mechanism of circuit failure and common electronic components and fault symptoms were carried out.Then, the nonlinear model of the whole railway vehicles’LED illuminating system was established based on features. Wavelet packet energy entropy method which is with high accuracy to classification of faults was choosed to extract feature amount of circuit fault.Combined with the railway vehicles’LED illuminating system and its nonlinear, time-varying and less prior knowledge characteristics,based on a belief rule base (belief rule base, referred to as BRB) to study the problem of the fault diagnosis and prediction. The online update algorithm BRB parameters were studied to ensure real-time systems, speed and stability. According to the prior knowledge and historical information systems, based on the confidence rule base reasoning evidential reasoning algorithm, a railway vehicles’LED illuminating system fault diagnosis and forecast system model was established.Finally, the simulation experiments were tested. In order to be able to monitor condition of railway vehicles’LED illuminating system real-time and control system failure module,fault diagnosis and prediction interactive interface of railway vehicles compartment LED illuminating system based on LabVIEW was designed.Taking a bridge rectifier filter circuit soft fault of LED power driver module on railway vehicle compartment LED illuminating system as an example,analysis of fault diagnosis was done.Fault forecast analysis of brightness sensor was also finished as an example. Simulation and experimental results showed that:the theory of belief rule base can be accurate to fault diagnosis and forecast system. Application of the belief rule base theory for the diagnosis and forecast the fault, realized the online fault diagnosis and prediction of the LED illuminating system for railway vehicles, and provided the reference for the fault diagnosis and prediction of other mechanical and electrical system.