Mission planning of agile satellites for multi-targets observation

Mission planning is an effective support and important guarantee for agile satellites to implement earth observation missions. According to traditional mission planning methods by which the mission and attitude are independently planned, the observation efficiency of agile satellites cannot be fully utilized. This paper considers the cooperative planning of mission and attitude, and establishes a multi-objective cooperative planning model of agile satellites based on dynamic traveling salesman model. In view of the complexity of the encoding method for the traditional single genetic algorithm, a novel algorithm combining genetic algorithm and polynomial fitting is proposed to reduce the encoding dimension of genetic algorithm. In this algorithm, quintic polynomials are used to estimate the attitude paths, and genetic algorithm is used to provide the cooperative planning with the observation order and the relative observation time. The numerical results demonstrate that the proposed algorithm can output the smooth attitude path and the convergence speed can be improved.