Abstract:Aiming at the task assignment problem of the satellite cluster game, a task assignment method based on a multi-layer dung beetle optimization (MDBO) algorithm is studied in this paper. First, three optimization criteria satellite fuel consumption, orbital transfer time, and game benefits are considered, along with constraints such as tasks, fuel, and payload, to construct the optimization mathematical model. Then, the MDBO algorithm is designed to solve the problem, introducing the "Dung Beetle Bandit Particle" to address the issue of particle swarm algorithms easily converging to local optima. A multi-layer optimization mechanism is applied to decompose and reduce the dimensionality of the complex satellite game task allocation problem. The simulation results show that compared to the DBO algorithm, the proposed method improves optimization accuracy by 3.91%, reduces response time by 19.20%, and enhances the algorithm's stability under interference by 3.84% after convergence.