Research On Fault Diagnosis Technology Of Energy Recovery Power Supply Device

With the rapid development of urban rail transit industry in China,energy recovery devices with energy saving and bidirectional converter functions have been widely used.This paper studies the open circuit fault of the switch tube of the energy recovery device.The corresponding fault diagnosis system is designed and implemented,and its diagnosis effect is verified.The specific work completed is as follows:First of all,the open circuit fault of the energy recovery device is simulated and analyzed.Based on the introduction of the topological structure of the energy recovery device and PWM rectifier,the simulation model of the rectifying and inverting conditions of the energy recovery device is built in MATLAB.The fault characteristics of the open circuit fault of the switch tube are classified,and the simulation experiment and analysis are carried out based on this.And the typical waveforms of normal operation and four fault states are given.Secondly,the extraction of open circuit fault eigenvalues is studied.The wavelet packet decomposition method is used to extract the fault characteristics of the three-phase current of the energy recovery device.The DB4 wavelet is selected to decompose the fault signal in three layers,and the wavelet packet norm entropy is used to extract the fault eigenvalue.Thirdly,open circuit fault diagnosis is realized by optimizing neural network with genetic algorithm.In this paper,genetic algorithm and neural network are studied.On the basis of determining the structure of BP neural network,genetic algorithm is used to optimize the weight threshold of neural network to improve the training performance of neural network.Based on this,the open circuit fault diagnosis of the switch tube of the energy recovery device is realized,and the diagnosis effect of the optimized neural network is verified by simulation.Finally,the fault diagnosis system of the energy fed power supply device is designed and implemented,and its performance is verified.The system is based on Lab VIEW environment,including wavelet packet decomposition module,BP neural network fault diagnosis module and interface display module.Through a large number of fault data obtained in the simulation of energy fed power supply device,the accuracy of fault diagnosis of the system is tested,and the test results show that the system has good diagnosis performance.