TWERD Frequency Converter Fault Diagnosis Based On Wavelet Transform

TWERD series frequency converter is a high-performance general-purpose voltage frequency converter, which can adapt to various load and low speed high torque to meet the occasion demands. This paper mainly aimed to TWERD series frequency converter for fault diagnosis, in the establishment of AC-DC-AC inverter model based on the type of fault on the various theoretical analysis and simulation results are summarized and compared various failure presented wavelet-based fault diagnosis.Frequency Control System is established mathematical model, carried out in the MATLAB simulation software. On the common voltage converter, to establish its main circuit of the simulation model the actual conditions in the inverter main circuit of several common faults Such as the SCR open, fast fuse blows, the trigger pulse caused by loss of such open faults slip.As the inverter circuit failure rate is the highest part of the inverter, the inverter circuit so this type of fault to the main objects of study the most. Inverter output voltage used as a source of fault information, for fault diagnosis. Since inverter circuit with nonlinear characteristics, can not establish accurate mathematical model. In this paper, wavelet transform is very sensitive to the singularities of the signal characteristics, design of a wavelet-based fault diagnosis method of inverter circuit.The output voltage to failure as the basis for the use of low-frequency wavelet analysis to extract the energy value as the feature vectors. Through neural networks bridge inverter circuit fault positioning arm, and finally use the same leg inverter circuit fault output signal characteristics of symmetry, using a simple logic to judge the separation of the fault components. Simulation results show that: the method can be PWM inverter circuit fault diagnosis, and the convergence speed, diagnostic accuracy.