摘要
通过BL4.0iBeacon技术进行室内定位,分析了最邻近值算法和贝叶斯算法的不足,提出一种将上述两者优点进行融合的算法.首先,使用最邻近值算法的优选区域限定,缩小贝叶斯算法对指纹库的搜索范围,然后,采用贝叶斯概率法计算各个iBeacon节点的信号强度值集合在指纹库中每一位置参考点对应的条件概率值,概率最大的点就是待测点的预测位置.实验结果表明,使用该算法计算的坐标准确度比较高,坐标误差较小,能明显提高室内定位的精度.
BL4.0iBeacon technology is introduced for indoor positioning.This paper analyzes the shortcoming of two traditional positioning algorithms：KNN algorithm and Bayesian algorithm,then proposes a combined algorithm of these two algorithms.First KNN algorithm is used to obtain the position set of which the distance to the position to be measured is smaller,and thus narrow search range of fingerprint database for Bayesian algorithm.After that match probability is calculated using Bayesian probabilistic algorithms for the selected position set.The biggest probability of matching point is exactly the predicted position of the target point.The experiment results verify that the accuracy of coordinate with the proposed algorithm is relatively high,and coordinate error is small,so as to improve indoor positioning accuracy significantly.
出处
《杭州电子科技大学学报（自然科学版）》
2016年第5期1-5,21,共6页
Journal of Hangzhou Dianzi University：Natural Sciences
基金
浙江省自然科学基金资助项目(LY15F030018
LY16F030004)
