A multi-objective scheduling method via hybrid-strategy for multiple agile satellites observing area targets

Agile satellites cooperative observation scheduling is critical for enhancing Earth observation capability. However, when observing irregular areas, it is necessary to consider issues such as high computational complexity, numerous constraints, and difficult multi-objective trade-offs. To address these issues, this paper formulates an integer programming model centered on optimal strip selection and establishes a multi-objective optimization scheduling framework that maximizes total coverage and minimizes attitude maneuver cost. To obtain candidate strips, a discrete-projection based dynamic strip decomposition method is proposed. A planar projection coordinate system is constructed at each area’s centroid, and dynamic strip decomposition is performed based on the time varying characteristics of individual satellite overpasses. To efficiently deal with complex constraints and solve the multi-objective optimization scheduling problem, the multi-strategy parallel adaptive non-dominated sorting genetic algorithm II (MPA-NSGA-II) is proposed. It uses a probability-arbitrated greedy conflict resolution scheme that overrides greedy choices to reduce selection bias. When progress stalls, a trigger-based local search is activated to improve boundary solutions on the Pareto front. In addition, diversity driven adaptive crossover and mutation operators adjust the search intensity, and parallel generation of new populations speeds up evolution. Compared with existing multi-objective optimization scheduling algorithms, MPA-NSGA-II exhibits superior capability in locating Pareto optimal solutions characterized by higher Hypervolume and better boundary solutions on the Pareto front, while achieving the minimum calculation time. The algorithm shows better performance in both computational efficiency and solution quality.