配电网中基于分簇定位的WSN节点故障定位研究

发布时间：2018-05-26 00:25

  本文选题：线传感器网络 + 配电网故障检测　； 参考：《传感技术学报》2017年01期

【摘要】：针对电网故障检测中使用的无线传感器网络节点定位精度较低,分簇不均问题,提出了一种基于DV-Hop算法改进均值粒子群算法(PSO),首先DV-Hop算法改进均值粒子群算法中粒子的速度与位移,使动态无线传感器网络重新定位簇头节点坐标更加接近真实值;然后递归神经网络学习算法迭代值逼近最合适的惯性权重值,优化均值PSO粒子群算法使其达到最优搜索能力。最后由Sink节点对每一次动态分簇后网络节点进行数据采集后对电能耗尽的节点进行无线充电。仿真结果表明,改进后的PSO算法比PSO算法聚类分簇误差更小,节点定位配电网故障的精确度提高12.8%,有效地延长了网络生命周期。
[Abstract]:Aiming at the problem of low location accuracy and uneven clustering of wireless sensor network nodes used in power network fault detection, An improved mean particle swarm optimization (PSO) algorithm based on DV-Hop algorithm is proposed. Firstly, the DV-Hop algorithm improves the velocity and displacement of particles in the mean PSO algorithm, so that the coordinate of cluster head nodes can be relocated in dynamic wireless sensor networks (WSNs). Then the recursive neural network learning algorithm approximates the most appropriate inertial weight value and optimizes the mean PSO particle swarm optimization algorithm to achieve the optimal search ability. Finally, the Sink node collects the data of the network nodes after each dynamic clustering. The simulation results show that the improved PSO algorithm has less clustering error than the PSO algorithm, and the accuracy of node location and distribution network fault is increased by 12.8. the network life cycle is effectively prolonged.
【作者单位】： 安徽工程大学电气工程学院;国网芜湖供电公司;
【基金】：2016年安徽高校自然科学研究项目(KJ2016A794)
【分类号】：TP18;TP212.9;TN929.5
，

本文编号：1935237