射频RSS聚类与多传感器融合的室内定位算法

RF RSS-based clustering and multi-sensor integration location algorithm

为有效解决RSS定位算法中的定位点跳变问题,提出一种基于射频RSS聚类与多传感器融合的室内定位算法。将定位坐标点与MT2503传感器接收的RSS信号加权聚类均值进行绑定,形成室内定位指纹库,融合LSM303DLHC传感器勘测定位终端实时朝向,补偿RSS信号受多径效应、阴影衰落等影响造成的误差,采用欧氏距离加权平均算法解算实时位置点坐标。实验结果表明,该算法抗RSSI扰动能力强,定位精度较高。

To solve the positioning point transition problem in RSS localization algorithm effectively,an algorithm based on RF RSS clustering and multi-sensor fusion was put forward.The coordinate points were bounded to the weighted average clustering of the RSS signal which were received from Bluetooth RF.An indoor positioning fingerprint was then constituted.To mark the orientations,the terminal fused the accelerometer sensors LSM303DLHC.The errors caused by the multi-path effect and shadow fading of the RSS signal were compensated.A reverse algorithm of Euclidean best matching was used to calculate the coordinate points.The results show that the proposed algorithm has strong ability of anti-RSSI perturbation and has higher accuracy in positioning.