西安电子科技大学学报 ›› 2016, Vol. 43 ›› Issue (2): 150-156.doi: 10.3969/j.issn.1001-2400.2016.02.026

• 研究论文 • 上一篇    下一篇

改进二进制粒子群优化的节点选择算法

魏声云;张静;郭虹;李鸥   

  1. (解放军信息工程大学 信息与系统工程学院,河南 郑州  450001)
  • 收稿日期:2014-12-19 出版日期:2016-04-20 发布日期:2016-05-27
  • 通讯作者: 魏声云
  • 作者简介:魏声云(1989-),男,解放军信息工程大学硕士研究生,E-mail:junyun1002@126.com.
  • 基金资助:

    国家自然科学基金资助项目(61201380);国家科技重大专项资助项目(2014ZX03006003)

Sensor selection algorithm based on modified binary particle swarm optimization

WEI Shengyun;ZHANG Jing;GUO Hong;LI Ou   

  1. (Institute of Information System Engineering, Information Engineering Univ. of PLA, Zhengzhou  450001, China)
  • Received:2014-12-19 Online:2016-04-20 Published:2016-05-27
  • Contact: WEI Shengyun

摘要:

针对无线传感器网络多目标跟踪节点选择问题,提出了一种最大化跟踪精度的二进制粒子群优化节点选择算法.该算法基于目标的预测位置,以费舍尔信息矩阵的迹为精度度量,构建节点优化选择模型,提出了二进制粒子群优化的改进形式,并用于节点选择模型的求解.改进的二进制粒子群优化算法采用矢量的二进制编码方式、约束满足的循环移位种群初始化方法,带V型转换函数的位置更新规则,并设计了引导因子引导粒子群的进化.仿真结果表明,所提出的节点选择算法能够有效地应用于多目标跟踪问题,与基本的二进制粒子群优化算法和遗传算法相比,改进的二进制粒子群优化算法能够在全局寻优和局部探索间取得平衡,且能有效地避免局部最优,对较大规模的网络具有很强的适用性.

关键词: 无线传感器网络, 节点选择, 二进制粒子群优化, 费舍尔信息矩阵

Abstract:

Considering the problem of sensor selection for multi-target tracking in wireless sensor networks(WSN),a sensor selection algorithm based on binary particle swarm optimization(PSO) is proposed to maximize the tracking accuracy. The predicted coordinate of the target and the determinant of the Fisher information matrix (FIM) is used for sensor selection. A modified form of binary particle swarm optimization(MBPSO) is proposed to solve the model, which is designed by employing the binary vector coding manner, constraint satisfaction cyclic shift population initialization method, particle position updating rules with the V-shaped transfer function and guidance factor. Simulation results show that the proposed sensor selection algorithm can be efficiently applied in the multi-target tracking problem. Compared to the basic particle swarm optimization algorithm and genetic algorithm (GA), the modified algorithm achieves a balance between global optimization and local exploration, and can effectively avoid the local optimum. Moreover, the proposed algorithm is suitable for large-scale networks.

Key words: wireless sensor networks, sensor selection, binary particle swarm optimization, Fisher information matrix