关键词: 调度算法, 时间窗口, 点目标, 侧摆模式, 演化算法

Abstract: Concerning the low efficiency of earth observation satellite, a scheduling method for multi-satellite and point target task on swinging mode was proposed. In this paper, the angle relation between satellites and ground targets and the computing method of the positive and negative swinging angle and time window were firstly analyzed. On the basis of these analyses, a scheduling model for multi-satellite and point target task on swinging mode was developed. In the model, the maximum observation was obtained, the minimum swinging number and the minimum total swinging angle were set as the optimization objectives. Based on evolution algorithm, an optimal scheduling algorithm on swinging mode was proposed. In the algorithm, the single point crossover operator and the mutation operation through the selection of time window were adopted. The fitness function based on conflict was defined. The conflicts were reduced by adjusting the actual start time of the activities. The selection strategy was designed according to the order of the optimization objectives. The conflicts were eliminated by defining the conflict cost. Finally, the scheduling result and the simulation of a practical example on 5 satellites and 100 point targets with swinging mode were given, and the scheduling performances were also analyzed with the swinging angle of 0°, 10° and 25°. The experimental results show that the observation efficiency improves by 18% when the swinging angle is 25°. The method has important application value in emergency relief and wartime rapid response.

Key words: scheduling algorithm, time windows, point target, swinging mode, evolution algorithm