北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2018, Vol. 41 ›› Issue (4): 91-96.doi: 10.13190/j.jbupt.2017-262

• 研究报告 • 上一篇    下一篇

基于改进蚁群的BP神经网络WSN数据融合算法

余修武1,2, 刘琴1,2, 李向阳1, 张可1, 肖人榕1   

  1. 1. 南华大学 环境与安全工程学院, 湖南 衡阳 421001;
    2. 中钢集团马鞍山矿山研究院有限公司 金属矿山安全与健康国家重点实验室, 安徽 马鞍山 243000
  • 收稿日期:2017-12-29 出版日期:2018-08-28 发布日期:2018-10-09
  • 作者简介:余修武(1976-),男,副教授,硕士生导师;刘琴(1993-),女,硕士生,E-mail:lqing8008@163.com.
  • 基金资助:
    金属矿山安全与健康国家重点实验室开放基金项目(2016-JSKSSYS-04);国家自然科学基金项目(11705084);国家应急管理部安全生产重特大事故防治关键技术科技项目(hunan-0001-2018AQ);湖南省重点研发计划项目(2018SK2055)

Information Fusion Algorithm Based on Improved Ant Colony Optimization BP Neural Network in WSN

YU Xiu-wu1,2, LIU Qin1,2, LI Xiang-yang1, ZHANG Ke1, XIAO Ren-rong1   

  1. 1. School of Environment and Safety Engineering, University of South China, Hunan Hengyang 421001, China;
    2. State Key Laboratory of Safety and Health for Metal Mines, Sinosteel Maanshan Institute of Mining Research Company, Anhui Maanshan 243000, China
  • Received:2017-12-29 Online:2018-08-28 Published:2018-10-09

摘要: 为了保证无线传感器网络(WSN)在深井中能有效地工作,提出了一种改进蚁群的反向传播(BP)神经网络WSN数据融合算法(IFA-IACOBP).通过规划蚂蚁运动方向和引入节点剩余能量对蚁群算法启发因子进行改进,优化蚂蚁下一跳节点选择概率,利用改进后的蚁群算法对BP神经网络进行优化,引入井下WSN数据融合,数据经两级融合处理后,能去除大部分冗余信息.仿真实验结果表明,IFA-IACOBP算法能有效减少网络数据通信量,提高数据实时性,降低网络能耗,延长网络寿命.

关键词: 蚁群算法, 反向传播神经网络, 无线传感器网络, 数据融合, 深井

Abstract: In order to ensure the effective working of wireless sensor network (WSN) in deep mine, an information fusion algorithm based on improved ant colony optimization back-propagation (BP) neural network in WSN (IFA-IACOBP) is proposed. The heuristic factor of ant colony optimization (ACO) is improved by planning the direction of ants' motion and introducing the residual energy of nodes to improve the selection probability of the ant next-hop node. The improved ant colony algorithm is used to optimize the BP neural network, which is applied to WSN information fusion in mine. These data are processed by two-level fusion, which can remove most redundant information. Simulation results show that the IFA-IACOBP algorithm can effectively decrease the network data communication, improve data real-time performance, reduce network energy consumption and prolong network lifetime.

Key words: ant colony optimization, back-propagation neural network, wireless sensor network, information fusion, deep mine

中图分类号: