Purpose The purpose of this paper is to propose a hybrid algorithm of branch pruning and genetic algorithm to get a fast and feasible solution for satellite swarm tasks planning. Design/methodology/approach This paper proposes an algorithm in which a binary tree branch pruning method is used to create a preliminary plan for satellite tasks; this plan is then used as an initial population for optimization by the genetic algorithm. The novel combination of the branch pruning algorithm with binary trees that were simplified with coarse and detailed pruning strategies resulted in high-quality solutions in a small solution space. The proposed hybrid algorithm was compared with a genetic algorithm that used a random initial population. Simulations involving various numbers of satellites and targets were conducted. Findings The hybrid algorithm achieved superior or equal results to the traditional genetic algorithm in all scenarios in terms of solution quality and computation time. The binary tree branch pruning method is time-efficient and yields solutions that closely approximate optimality, resulting in higher-quality optimal solutions found by the genetic algorithm and accelerating convergence. Research limitations/implications This study has two main limitations. First, the algorithm assumes sufficient onboard memory; under actual constrained conditions, a post-heuristic-insertion task reordering and removal mechanism would be necessary. Second, the model is designed for single-observation point targets, limiting its direct applicability to large-area continuous coverage. Future work will involve decomposing large areas into virtual point grids and recalculating attitude transition times to enhance the algorithm’s practical scope. Practical implications This paper provides an algorithm for solving fast and high-quality solutions for swarm satellite task planning. Originality/value The proposed algorithm can surpass previous works using traditional genetic algorithm.
Ma et al. (Mon,) studied this question.