基于非均匀分簇和蚁群神经网络的WSN数据融合算法  

WSN Data Fusion Algorithm Based on Non-Uniform Clustering and Ant Colony Neural Network

作　　者：杭超 李刚[1,2,3] 谢昱卓 李雯珺 HANG Chao;LI Gang;XIE Yuzhuo;LI Wenjun(Mechatronics T&R Institute,Lanzhou Jiaotong University,Lanzhou Gansu 730070,China;Gansu Provincial Engineering Technology Center for Informatization of Logistics&Transport Equipment,Lanzhou Gansu 730070,China;Gansu Provincial Industry Technology Center of Logistics&Transport Equipment,Lanzhou Gansu 730070,China)

机构地区：[1]兰州交通大学机电技术研究所,甘肃兰州730070 [2]甘肃省物流及运输装备信息化工程技术研究中心,甘肃兰州730070 [3]甘肃省物流与运输装备行业技术中心,甘肃兰州730070

出　　处：《传感技术学报》2020年第10期1483-1488,共6页Chinese Journal of Sensors and Actuators

基　　金：甘肃省高等学校科研项目(2018C-10)。

摘　　要：为了确保无线传感器网络在列车车厢中能高效稳定地工作,提出了一种基于蚁群优化神经网络的数据融合算法(DFA-IACOBP)。该算法将无线传感器网络非均匀分簇结构与神经网络结构相结合,建立一个基于非均匀分簇路由神经网络的无线传感器网络数据融合模型。在非均匀分簇路由算法中,候选簇头根据竞争半径构造出大小不一的簇,并在每个簇中竞选出两个簇头。主簇头负责簇内信息采集和处理,副簇头承担簇间信息转发。神经网络的权值和阈值由蚁群算法优化寻得,优化后的神经网络能从存在大量冗余数据的无线传感器网络提取有效特征数据并传输至汇聚节点。仿真结果表明:DFA-IACOBP算法能大幅降低网络中冗余数据,减少网络数据通信量,提高特征数据采集效率和网络整体性能。In order to ensure efficient and stable working of wireless senor network in the train carriage,an information fusion algorithm based on ant colony optimization neural network(DFA-IACOBP)is proposed.The algorithm combines the non-uniform clustering structure of the wireless sensor network with the neural network structure.A wireless sensor network data fusion model based on non-uniform clustering routing neural network is established.In the non-uniform clustering routing algorithm,the candidate cluster heads construct clusters of different sizes according to the competition radius,and two cluster heads are elected in each cluster.The main cluster head is responsible for information collection and processing within the cluster,and the deputy cluster head is responsible for information forwarding between clusters.The weights and thresholds of the neural network are optimized by the ant colony algorithm.The optimized neural network can extract effective feature data from the wireless sensor network with a large amount of redundant data.The effective feature data is transmitted to the Sink node.Simulation results show that DFA-IACOBP can greatly decrease network redundant data,reduce network data communication,improve the efficiency of characteristic data collection and the overall performance of the network.

关 键 词：无线传感器网络 数据融合 非均匀分簇 蚁群算法 神经网络 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]