基于多指纹联合匹配的混合定位算法

随着信息科技的迅猛发展,室内定位技术已经成为基于位置服务LBS的研究热点之一。基于接收信号强度RSS的位置指纹与步行者航位推算PDR相结合的定位算法能有效提高定位精度,但目前已有的算法难以同时满足较高的定位精度与较小的计算量,常见的卡尔曼滤波算法精度不够,而粒子滤波算法计算量较大。提出了一种基于多指纹联合匹配的混合定位算法,有效融合惯性信息与RSS指纹信息,在较低计算量的前提下实现了高精度定位。实验结果表明,该算法80%的定位精度低于1m,平均精度高达0.77m。

A hybrid localization algorithm based onmulti fingerprint union matching

With the rapid development of information technology, indoor localization technology has become one of the most important research hotspots in location based service (LBS). Hybrid algorithms combining the received signal strength (RSS) location fingerprinting and pedestrian dead reckoning (PDR) can effectively improve the localization accuracy, but existing algorithms can hardly achieve higher localization accuracy and smaller calculation simultaneously. The common Kalman filter algorithm has limit localization accuracy while a large amount of calculation is required by the particle filter algorithm. We propose a hybrid localization algorithm based on multi fingerprint union matching, which fuses the inertial sensor and the RSS fingerprint information effectively. Under the premise of low calculation amount, it can achieve higher precision of localization. Experimental results show that, 80% of the errors remain within 1m and the average accuracy reaches as high as 0.77m.