天基预警系统资源调度方法

Research on Satellite Scheduling of the Space-based Early Warning System

对面向导弹预警任务的天基预警系统中的低轨卫星资源调度问题进行了研究。对预警任务的特性进行分析,提出了一种预警任务规划与分解方法以减少任务对资源占用时间的需求;在此基础上建立了该问题的约束满足CSP调度模型;针对该模型变量多、解空间规模庞大的特点以及对求解算法时效性的要求,设计了一种基于局部解空间跳出机制的改进型遗传算法M GA。仿真表明该模型与算法在给定时间内能够得到满意的调度方案。

Scheduling of Low-Earth Orbit （LEO） satellite resources in the space-based early warning system for ballistic missile warning tasks is studied. Firstly, characteristics of missile early warning tasks are analyzed, and a task programming method including task decomposing and formalizing is designed basing on the fact that a task can be broken down to reduce its resource occupancy; Secondly, a scheduling model basing on Constraint Satisfaction Problem （CSP） is founded, whose characteristics and scale of solution are analyzed; Thirdly, a Modified Genetic Algorithm （MGA） basing on a local solution scope outrunning mechanic is presented, which meets with time requirements, is suitable for the CSP model with numerous variables and solutions, and provides capacities in fleetly searching for the best solution; Finally, an emulator is designed to validate capabilities of the model and the algorithm. The outcome of the emulator shows that satisfying scheduling schemes can be attained by the model and the algorithm during a fixed short period.