摘要
针对卫星电子侦察过程中的各种不确定因素,提出了一种静态调度与反应式调度相结合的鲁棒性调度方法.静态调度中针对具有随机分布特性的侦察数据量,并考虑卫星的使用约束和任务约束,建立了机会约束混合整数规划模型;提出了基于罚函数的进化算法,用以解决模型的非线性与非凸特性,结合蒙特卡罗抽样的罚函数用以解决模型的随机性.反应式调度中针对新任务提交、资源失效和任务执行失败等离散不确定因素,提出了一种快速启发式任务插入算法,基于初始调度方案,有效降低调度复杂度.仿真实验表明该方法适用于解决不确定因素下电子侦察卫星调度问题.
Aimied at the uncertainties in the course of satellites electronic reconnaissance,a robust scheduling method was proposed,by the combination of static scheduling with reactive one.In the process of static scheduling,taking the data quantity with stochastic distribution as well as satellites usage restriction and task restriction into account,a mixed integer programming model for electronic reconnaissance satellites scheduling problem（ERSSP）based on chance constrained programming was established.A penalty function based evolutionary algorithm was proposed.The evolutionary algorithm was resorted to resolve the non-linear and non-convex component in the model.And the penalty function combined with Monte Carlo simulation was to address the stochastic nature of ERSSP.In reactive scheduling,a fast heuristic tasks insertion algorithm （FHTIA）was proposed according to the dispersed uncertain factor,e.g.new tasks submission,satellites failure,reconnaissance plan execu- tion failure.On the basis of the excellent performance of the original schedule,the complexity of the reactive scheduling was reduced greatly.The simulation results show that the method could solve ERSSP under uncertainty effectively.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2010年第11期80-83,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家安全重大基础研究资助项目(6136101)
关键词
卫星调度
机会约束规划
罚函数
反应式调度
启发式算法
satellite scheduling
chance constrained programming
penalty function
reactive scheduling
heuristic algorithm
