基于不同信息素更新策略的卫星数传调度蚁群优化算法  

Ant Colony Optimization Algorithm for Satellite Data Transmission Scheduling Based on Different Pheromone Updating Strategy

作　　者：陈祥国[1] 武小悦[1] 

机构地区：[1]国防科学技术大学信息系统与管理学院,湖南长沙410073

出　　处：《运筹与管理》2009年第3期57-63,共7页Operations Research and Management Science

基　　金：国家973重点基础研究发展规划(6136101)

摘　　要：针对具有时间窗口和数传资源限制卫星数传调度问题,提出了基于解构造图模型的蚁群优化算法。借鉴精英机制,设计了绝对精英策略、相对精英策略、收益精英策略和对等精英策略等四种信息素更新策略。通过对不同规模场景的仿真试验,验证了基于不同信息素更新策略的蚁群算法是求解卫星数传调度问题的有效途径。基于信息素平衡思想的相对精英策略、收益精英策略和对等精英策略相对于绝对精英策略而言,能够避免算法过早陷入局部最优或出现退化行为,在规模较大的场景中能够收敛到比绝对精英策略更优的解。在小规模场景中,相对精英策略和收益精英策略所得解最好,而在大规模场景中对等精英策略所得解最好。For satellite data transmission scheduling problem （SDTSP） restrained with time windows and resources, ant colony optimization algorithm based on solution construction graph model is proposed. Using for reference the elitist strategy, different pheromone updating strategies are designed which include absolute elitist strategy, relative elitist strategy, income elitist strategy and opposite elitist strategy. Simulation on different size scenes shows that the ant colony optimization algorithm in the paper performs well for SDTSP. Based on pheromone balance thought, relative elitist strategy, and income elitist strategy and opposite elitist strategy can avoid getting into local optimization or appearing degradation, and can converge to better global optimization solution in bigger size scene compared with absolute elitist strategy. Relative elitist strategy and income elitist strategy can get best solution in smaller size scene, and opposite elitist strategy perform best in bigger size scene.

关 键 词：蚁群优化算法 信息素更新策略 解构造图 卫星数传 任务调度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] V57[自动化与计算机技术—计算机科学与技术]