Abstract: A fault localization method based on sniffing strategy slime mould algorithm is studied to address the issues of complex process and slow diagnostic speed in fault localization in AC/DC hybrid microgrids. This method utilizes a sniffing strategy consisting of the Daubechies wavelet transform and node topology optimization to perform data processing. Specifically, PowerFactory is used to build a medium and low voltage AC/DC hybrid microgrid model, and set faults on the AC and DC transmission lines to extract corresponding voltage and current fault signals separately. Secondly, the fault signal is decomposed and reconstructed using the Daubechies wavelet transform to obtain its information Shannon entropy, which is then combined to form an offset value that can characterize the fault. Then, through node topology optimization, the fault offset values of nodes and branches in the hybrid microgrid are fitted. Finally, the proposed strategy is simulated and validated using the optimized slime mould algorithm in Matlab. Simulation results show that the proposed sniffing strategy technology improves the accuracy by 21% and the convergence speed by 35.9% at the expense of a small increase in the time required for each iteration, achieving the purpose of accurately and quickly locating hybrid microgrid faults.

Key words: microgrid, short circuit fault, Slime Mould algorithm, regional location, fault localization