Abstract: Objectives The current island and reef power supply system, composed of a variety of distributed power sources with numerous types and complex topologies, is located in a harsh geographical environment. If traditional distribution network fault location algorithms are used, it will be difficult to accurately and quickly determine the fault section problem. Methods This article uses an immune model to divide the island and reef power supply system into several small segments with monitoring points. The detected fault current signal determines the small segment where the fault point is located, minimizing the range of fault diagnosis and encoding the fault points of the small segment. By improving the immune algorithm to generate an antibody population, the fault section can be quickly and accurately located twice, and the computational cost can be greatly reduced. Results The Matlab simulation results show that the method is used to simulate and verify the typical island and reef power supply system constructed. The example results show that the method is suitable for rapid fault location analysis when single and multiple faults occur in island and reef power systems constructed by multiple distributed power sources. Conclusions The proposed improved immune algorithm can adapt to complex island and reef power systems with multiple distributed power sources, and has advantages such as high positioning accuracy, low computational complexity, and fast solving speed.