Multistage Component-Level Optimized Design Method for Satellite Constellations With Uncertainty Quantification of Multigranularity Satellite Constellation Models

The missile warning satellite constellation plays an extremely important role in the missile defense system, which is always of extremely high military value. It is significant to perform optimized design for the satellite constellation in the digital world, which can provide decision-making basis for the satellite constellation design in practical engineering. It is well known that the finer the granularity of satellite constellation models, the higher the evaluation accuracy, and the more valuable the optimized design results. Component-level satellite constellation models can accurately model all components within the satellite, which can bring more valuable optimized design results. However, the main problems of high total time consumption, huge solution space, and difficult uncertainty quantification are existed in the component-level optimized design due to the high complexity of the satellite constellation. In order to solve the above problems, the multistage component-level optimized design method for satellite constellations with uncertainty quantification of multigranularity satellite constellation models is proposed. The designed multistage optimization combining multigranularity modeling can significantly not only reduce the solution space but also improve optimization efficiency. The optimization focus and model granularity selection vary in different stages, and designed stage switching strategy can achieve on-demand switching among different stages. The designed uncertainty quantification not only can provide quantitative basis for stage switching but also quantify the evaluation accuracy of optimized design results to make them more valuable. Finally, simulation results demonstrate the effectiveness of the proposed method.