with the rapid improvement of Internet and information construction in all walks of life, the traditional centralized storage system is gradually replaced by the distributed storage system. Fault processing method based on distributed system is a research hotspot in the field of power system fault identification. The purpose of this paper is to integrate large-scale dynamic programming with distributed system fault handling methods. Through the research of large-scale dynamic programming and its algorithm and distributed system, the relevant fault handling methods are proposed. In the experiment, we construct the system execution flow and use the dynamic planning algorithm to study and analyze the usability and universality of the distributed system dynamic planning application.