A Multi-Objective Enhanced Estimation of Distribution Algorithm for Proactive Scheduling of Agile Earth Observation Satellites

In this paper, an Enhanced Estimation of Distribution Algorithm (EDA) for the multi-objective problem is proposed for the active scheduling of agile Earth observation satellites considering external environmental disturbances and emergency mission insertion. Firstly, considering that the increased idle time can enhance the fault tolerance of the system, two metrics are proposed: schedule profit and total slack time, based on which a multiple multidimensional knapsack problem with conflicts problem is modeled. Secondly, in order to solve the problem of multi-objective model, EDA is combined with non-dominating sorting strategy module to store and screen out elite solutions during the evolution process. Finally, EDA is compared with NSGA-II by combining STK simulation data, and it is proved that the improved EDA can achieve better distribution and convergence in obtaining the Pareto optimal set. This study provides methodological support for active scheduling of agile Earth observation satellites.