基于PPO策略选择引导灰狼优化的敏捷卫星任务规划算法OA

With the continuous growth in demand for timely and high-precision earth observation,agile satellite mission planning faces significant challenges in large-scale scenarios involving highly complex constraints.To address these issues,an agile satellite mission planning algorithm based on grey wolf optimization guided by proximal policy optimization(PPO)strategy selection was proposed.In this algorithm,the grey wolf optimizer was employed to generate multiple candidate scheduling strategies and construct a scale-controlled candidate set through rapid evaluation.On this basis,PPO was introduced to adaptively select strategies,forming a hybrid optimization framework that integrates heuristic generation with reinforcement learning-based decision making.To accommodate large-scale scenarios,a sparse action space was constructed through candidate screening to reduce decision complexity while preserving effective exploration capability.Experimental results demonstrate that the proposed algorithm outperforms A-ALNS,GRILS,and RL-GA algorithms in both computational efficiency and task return rate,validating its effectiveness in large-scale agile satellite mission planning.