无线传感网络数据集离群目标跟踪方法仿真

Simulation of Outlier Tracking for Data Sets in Wireless Sensor Networks

对传感网络数据集离群目标跟踪,能够有效提高无线传感网络运行安全性。对数据集离群目标的跟踪,需要对目标元素采样分量和观测分量进行匹配,建立元素跟踪模型,完成离群目标跟踪。传统方法先对元素边权值计算,建立元素交叉隶属度矩阵,但忽略了加入数据元素跟踪动量项,导致跟踪精度低。提出基于改进Elman无线传感网络学习算法的无线传感网络数据集离群目标跟踪方法,分析离群目标观测方程,建立观测模型,利用拓扑序列分量,对元素采样分量和观测分量进行匹配,建立元素跟踪模型,引入无线传感网络学习算法加入元素跟踪动量项,对学习率进行调节,完成无线传感网络数据集离群目标跟踪。实验结果表明,所提方法跟踪精度高,运算时间短,高效提高无线传感网络安全性。

ABSTRACT： The outlier target tracking of data set in sensor network can effectively improve the security of wireless sensor network. The traditional method calculated the edge weight of elements, but ignores the addition of momentum term data element tracking, resulting in the low tracking accuracy. Based on improved Elman wireless sensor network learning algorithm, an outlier target tracking method of data set in sensor network is proposed. The outlier target ob- servation equation is analyzed and the observation model is established. Then, by using the topological sequence com- ponent to match element sampling component and observation component, element tracking model is established and the learning algorithm of wireless sensor network is introduced to add elements tracking momentum term. Thus, the learning rate is adjusted to complete the outlier target tracking of wireless sensor network data set. Simulation results show that the proposed method has high tracking accuracy and short computing time; which can effectively improve the security of wireless sensor network.