Application of current fault diagnosis method based on CEEMD in transformer winding fault diagnosis

To address the issue of weak initial features in transformer inter-turn short-circuit faults, a fault current analysis method based on Complete Ensemble Empirical Mode Decomposition (CEEMD) is proposed. In this paper, a simulation model of transformer inter-turn short circuits is established on Matlab/Simulink to simulate inter-turn short-circuit faults with different severity levels. With the CEEMD algorithm, Shannon entropy, root mean square (RMS) and energy indicators, normalized criteria consisting of transient ratio and steady-state ratio are constructed. The results verify that the CEEMD algorithm can effectively extract fault features and exhibits favorable performance. Specifically, the threshold of the Shannon entropy ratio is set to 1, a ratio greater than 1 indicates the occurrence of transformer inter-turn short-circuit fault. Benefiting from low computational complexity, the proposed method can realize the detection of transformer inter-turn short-circuit faults.