摘要
为解决相控阵雷达事件调度问题中调度成功率、时间偏移率与算法时效性三者之间的矛盾,本文提出一种基于分支定界法的调度算法;该算法首先在现有调度结果上调取所有可执行事件,即获得若干“节点”,多步递推后由多“层”节点形成若干条备选“分支”,扩大解空间范围,提高算法性能;然后通过“删减”操作删减掉收益较低分支,将解空间控制在一适当范围内,减小搜索盲目性,降低算法计算量.仿真实验表明,与基于综合优先级算法的调度结果相比,利用本文算法调度时调度成功率提升了 52%;与基于时间指针算法的调度结果相比,本身算法时间偏移率降低了 61%;与基于遗传算法的调度结果相比,本文算法调度耗时仅为前者 1~2%.
To solve the contradiction of the scheduling success rate (SSR),the time shifting rate (TSR) and the algorithm timeliness in the task scheduling of phased array radar,this paper proposes a scheduling algorithm based on the branch and bound method (BBM).Firstly,this algorithm acquires all executable tasks from available scheduling results, i.e.,obtaining several “nodes” on multiple “layers”,and forms several optional “branches” with these nodes through multistep recursion.In this way,the range of solution space is extended,and the algorithm performance is improved.Then,this algorithm “discards” branches with low yield and keeps the solution space to a certain range,which improves the accuracy of the algorithm and reduces the calculation amount.The simulation experiment indicated that compared with the algorithm based on integrated priority The simulation experiment suggested that the proposed algorithm enjoyed a 52% increase in scheduling success rate(SSR) over the traditional integrated priority algorithm,experienced a 61% reduction in the time shifting rate(TSR) compared with the time pointer algorithm,and took only 1~2% of the time consumed by the genetic algorithm.
作者
段毅
谭贤四
曲智国
王红
DUAN Yi;TAN Xian-si;QU Zhi-guo;WANG Hong(Air Force Early Warning Academy,Wuhan,Hubei 430019,China;No.95174 Unit of PLA,Wuhan,Hubei 430010,China)
出处
《电子学报》
EI
CAS
CSCD
北大核心
2019年第6期1309-1315,共7页
Acta Electronica Sinica
基金
国家自然科学基金(No.61401504)
中国博士后科学基金(No.2014M562562)
关键词
相控阵雷达
事件调度
分支定界法
调度成功率
时间偏移率
综合优先级
时间指针
遗传算法
phased array radar
task scheduling
branch and bound method(BBM)
scheduling success rate(SSR)
time shifting rate(TSR)
integrated priority
time pointer
genetic algorithm
