Driven by the carbon peaking and carbon neutrality goals, the large-scale connection of new energy and new loads has created new challenges for cooperative and optimal dispatch of power systems.The distributed collaborative optimization approach in multi-intelligent systems provides innovative solutions to deal with new power system collaborative dispatch problems with its advantages of low implementation cost, high scalability, robustness, and privacy preservation.This paper provides a systematic review of the latest theoretical advances in distributed cooperative dispatch algorithms in China and abroad, and elucidates the advantages of the distributed approach over the centralized approach.In particular, the basic principles, advantages and limitations of various distributed cooperative optimization algorithms under undirected and directed graphs are analyzed from the perspective of network structure.In addition, this paper discusses the uncertainty problems of communication, renewable energy and load faced by the new power system cooperative dispatch under the distributed network and the existing solutions, as well as some non-convex cooperative scheduling models. Based on the existing research conclusions, the development trend of the new type of power system cooperative dispatch is outlooked, which provides more comprehensive information and new research perspectives for the researchers in this cross-cutting field.