ZigBee定位系统的精度分析  

Accuracy Prediction of ZigBee Positioning System Based on BP Neural Network

作　　者：郭叶[1] 汪焰恩[1] 魏生民[1] 

机构地区：[1]西北工业大学机电学院,西安710072

出　　处：《西安工业大学学报》2014年第5期360-364,共5页Journal of Xi’an Technological University

基　　金：新疆自治区科技支疆项目;新疆科技厅资助(201191115)

摘　　要：为了对ZigBee定位系统的精度进行预测,分析引起ZigBee无线网络定位误差的四项主要参数,针对不同参考节点横向间距和纵向间距,不同A、n值,进行了852组实验,综合分析这四项因素对定位精度的影响.提取出173组稳定性强的数据组成BP神经网络的训练样本,构建3层BP神经网络,实现ZigBee定位的精度预测.通过对实验和预测结果进行分析,确定四项因素的合理取值和BP神经网络预测结构.文中方法运用到ZigBee定位时,对设置参数具有参考意义,对定位精度进行预测,根据预测结果合理的调整参数取值,从而提高ZigBee无线定位网络的可用性.This study aims to analyze the four main parameters causing ZigBee wireless network positioning errors ,and forecast the positioning precision of the ZigBee positioning system .For different reference node spacings and different A and n values ,852 groups of experiment were conducted to make a comprehensive analysis of the impact of the four factors on the positioning accuracy .The samples of the BP neural network which consists of 173 groups of high-stability data were extracted to create a three-layer BP neural network for predicting positioning precision .By analyzing the experiments and the predicted results , the reasonable values of the four factors and the BP neural network prediction structure were determined .The method results show that the proposed in this paper can predict the positioning accuracy for the setted parameters and adjust the paramenters reasonably according to the predicted values ,thus increasing the availability of the ZigBee wireless location network .

关 键 词：无线网络 ZIGBEE技术 BP神经网络 精度预测 参数设置 

分 类 号：TN92[电子电信—通信与信息系统]