摘要
针对以往求解多传感器联盟算法中参数多、计算复杂、实时性不强、全局搜索能力不高等问题,提出了基于改进人工蜂群算法的多传感器联盟方案。该方案对传感器联盟目标函数进行了优化,令跟随蜂采用双向轮盘赌的方式选择引领蜂,以便跳出局部最优解,提高算法收敛速度。仿真实验结果表明,所提出的算法能够有效获得传感器的最优联盟方案并保持较好的稳定性,与改进前相比,其寻优能力得到增强,进一步研究发现,适当扩大采蜜蜂在蜜源周围的搜索范围、增加蜜蜂个数、缩小未更新计数器阈值,能够进一步提高算法收敛速度,得到收益度更好的最优联盟方案。
Aiming at the problems of over-parameterized and calculably complex with poor real time and weak global searching ability in previous algorithms used to build multi-sensor coalitions are ,a coalition scheme ba-sing on the improved bee colony optimization was proposed.In this algorithm,the obj ective function was opti-mized,and the following-bee chose the leading-bee in the way of bidirectional roulette,contributing to j umping out of the locally optimal solution and improving the convergence rate of the algorithm.The simulation and ex-perimental results showed,by using the improved bee colony optimization,the most optimized scheme could be gotten easily and the algorithm could keep smooth.Comparing to the basic bee colony optimization,the im-proved one could make optimization ability enhanced.By properly expanding the searching scale of bee around the source,increasing the number of the bee and reducing the threshold of not updated counter,the convergence of the algorithm and the adaptability of the most optimized coalition scheme could be improved.
出处
《探测与控制学报》
CSCD
北大核心
2016年第5期107-111,116,共6页
Journal of Detection & Control
基金
航空科学基金项目资助(20130196004)
关键词
改进人工蜂群算法
多传感器联盟
最优联盟方案
improved bee colony optimization
multi-sensor coalition
most optimized coalition scheme
