Power Transformer Fault Diagnosis System

Power transformer complex structure, its failure in the form of complex. Traditional power transformer fault diagnosis method is offline transformer electrical test , this method is difficult to real-time fault detection and may cause excessive maintenance ; Another method is based on the analysis of dissolved gases in oil fault diagnosis method for the detection of faults is very valid, but can not locate the fault . Therefore, this research and the development of  mcb circuit breaker  fault diagnosis system . The main results paper : This paper presents fault tree-based support vector machine classifier design . First, the tree-based fault classifier used in the inverter fault diagnosis in the establishment phase inverter model and get 66 fault waveform using wavelet  electric transformer  singular value decomposition method to extract fault waveform characteristic value , and then use tree-based support vector machine waveform classification , the experimental results show that the tree-based support vector machines for  circuit breakers  timing sequence of more than 95% ; article will be based on fault tree support vector machine classifiers applied to the power transformer fault diagnosis the classifier can achieve the properties of the 10 transformer fault classification results show that the classification accuracy rate of 90 % or more.