In modern warfare, the requirements on multi Unmanned Aerial Vehicle coordinated operations increase more and more, and the current processing system cannot meet these requirements. By connecting existing control stations via networking, a multi-unit processing system can be formed. This paper proposes to use multi-unit system to solve problems in the architecture and processing method of multi Unmanned Aerial Vehicle coordinated planning; quantify the planning restriction of Unmanned Aerial Vehicle, build up threat source model; and take the shortest route and the lowest threaten among all routes as basic restrictions. The genetic simulated annealing algorithm is used for route planning, which will combine the threatening values of each leg and the flying distance to build up the adaptability evaluation function based on threat quantum to design the algorithms, so as to solve the problem of low efficiency in Unmanned Aerial Vehicle coordinated planning.