Fault Location Algorithm for Distribution Network Feeder Sections Based on Association Matrix

Abstract

To enhance fault location accuracy in distribution networks, the authors propose a fault location algorithm for feeder sections based on the association matrix. This method constructs a node correlation matrix using characteristic information such as voltage and current from multiple measurement points. By integrating feature selection and a weighted matching mechanism, the algorithm achieves high-precision identification of fault regions. Simulation experiments conducted on modified IEEE 13-node and IEEE 123-node test systems demonstrate that the proposed algorithm maintains stable fault location capabilities under complex scenarios, including high impedance, multiple faults, and data loss. Compared with traditional impedance methods, traveling wave methods, and machine learning approaches, the proposed algorithm demonstrates significant advantages in terms of accuracy, computational efficiency, and robustness. Particularly under high impedance fault conditions, the accuracy rate can still be maintained above 94%. The optimized version further achieves a notable reduction in computation time and memory usage, showing promising potential for engineering applications. The research results indicate that this algorithm can effectively enhance the intelligence level and practicality of fault diagnosis systems in distribution networks.

Cite this article as: H. Ma, Y. Wu and H. Jiang, “Fault location algorithm for distribution network feeder sections based on association matrix," Electrica, 26, 0103, 2026. doi: 10.5152/electrica.2026.25103.