Distribution network differential protection scheme based on improved feature mode decomposition and adaptive similarity

Cutting off single-phase ground faults in distribution networks within the required time frame has always posed a challenging problem when selecting low-current grounding lines. With advancing communication technology, optical fiber and 5G costs continue to decrease, allowing for more convenient and quicker data transmission. As a result, various methods can be employed using signal waveforms to achieve distribution network differential protection.In this paper, we propose a new adaptive differential protection scheme that employs improved feature mode decomposition through the whale optimization algorithm to decompose zero-sequence current waveforms collected at both ends of the line. Based on the fundamental principle of current differential protection, we conduct similarity analysis using the most critical fault characteristic component. This approach effectively resolves the issue of inadequate grounding current in the distribution network, ensuring the differential protection device meets operational requirements and guarantees safe and stable system operation.Finally, simulation verification using MATLAB software validates the effectiveness of this scheme and verifies its accuracy under different fault conditions.