基于蚁群的天基预警多目标优化算法

Multi-objective optimization algorithm of space-based early warning based on ant colony

为了解决天基预警传感器调度问题,本文提出了一种基于蚁群优化算法和R2排序算法的连续搜索空间多目标蚁群优化算法。算法中考虑传感器覆盖数量动态变化的任务约束,卫星及传感器数量的资源约束、地球遮挡、临边观测及观测距离等环境约束。针对传统蚁群算法在天基预警任务规划中存在的多目标权衡能力差以及连续搜索空间计算效率低等问题,本文对传感器调度方案进行R2排序和多目标寻优计算,权衡了目标切换次数,传感器疲劳度以及目标观测时长等优化目标。将算法与元启发式和动态蚁群算法在观测资源充足,观测资源紧缺和观测资源严重不足3种状态进行对比仿真。结果表明:该算法可以在任务、资源和环境约束下对传感器切换次数、单星观测时长和总观测时长等目标进行优化,适用于天基预警星座系统对弹道导弹等具有红外特性运动目标的跟踪方案优化问题。

In this paper, a continuous search space multi-objective ant colony optimization algorithm based on the ant colony optimization algorithm and R2 sorting algorithm is proposed to solve the scheduling problem of space-based early-warning sensors. In the algorithm, the task constraints of dynamic changes in the number of sensor coverage, the resource constraints of the number of satellites and sensors, and environmental constraints, such as earth occlusion, edge observation, and observation distance, are considered. To solve the poor multi-objective trade-off ability and low computational efficiency of continuous search spaces existing in the traditional ant colony algorithm in space-based early warning task planning, a sensor scheduling scheme is sorted, and multi-objective optimization is calculated. Optimization objectives, such as target switching times, sensor fatigue, and target observation time, are weighed. Then, the algorithm is compared with the metaheuristic algorithm and dynamic ant colony algorithm in three states: sufficient observation resources, shortage of observation resources, and a serious shortage of observation resources. The results show that the algorithm can optimize the sensor switching times, single satellite observation time, and total observation time under the task, resource, and environmental constraints and is suitable for the optimization of the tracking scheme of space-based early warning constellation systems for ballistic missile and other moving targets with infrared characteristics.