Research On Transformer Fault Diagnosis System Based On Vibration Signal Analysis

Transformer in electric power system carries the important task of electric energy conversion, its ability to operate safely and reliably affects the operation of power system. Therefore, it has important practical significance to study the fault diagnosis technology of transformer. The vibration signal can be changed after the transformer has failed. The fault signal contains the state information of the transformer, so the research on the vibration signal of the transformer is a feasible method to determine the fault type. Because most of the fault signal is nonlinear and non-stationary, the traditional method of processing signal can not meet the requirements of fault diagnosis. So, it is necessary to select the signal processing method which is suitable for nonlinear and non-stationary signal.This paper first introduces the common faults of the transformer and the relationship between vibration and fault. Then, the problem of false or invalid components in the vibration signal of the transformer is carried out by using the ensemble empirical mode decomposition method(EEMD). The method of using correlation coefficient method to remove invalid components is proposed. Then, we extract the effective components and study the relationship between the load current and the various intrinsic mode components(IMF). In the end, the application software of Labview is designed for fault diagnosis of transformer, and the vibration data are analyzed, and the optimum measuring point position and IMF component of fault diagnosis for transformer is presented.