Fault diagnosis for power transformers based on Hybrid Fuzzy Dissolved Gas Analysis

The objective of this paper is to integrate five traditional criteria of the dissolved gases analysis published in different standards into a more reliable approach of the fault diagnosis of power transformer for the power supply department of Taiwan power company. The integrated approach is designed with three steps, including the judgment of the fault factor of power transformer, the identifiable factor of fault condition and the hybrid fault factor and identifiable factor. The implementation is based on fuzzy inference system to obtain the last answer. With precision rate and recall rate, simulation results prove that the method is especially useful in cases when the methods reach different conclusion and is more accurate than each traditional method.