Abstract:Taking the capabilities and usage restrictions of the electronic reconnaissance satellite (ERS) as well as the scheduling requirement of different tasks into account, a multi-objective mixed integer programming model for ERS cooperative reconnaissance problem (ERSCRP) was presented. Then a hybrid algorithm named MOEA-VNS was put forward. MOEA-VNS made full use of the outstanding global search capability of multi-objective evolutionary algorithm (MOEA) and the local search performance of variable neighborhood search (VNS). Taking advantage of the combination characteristic of multiple reconnaissance windows, the evolutionary operators and the neighborhood search operators were designed, which not only ensured the diversity of the non-dominated solutions acquired by MOEA-VNS, but also made them as close to the Pareto optimal solutions of the ERSCRP as possible. The experiment results show that MOEA-VNS can solve the problem effectively.