基于互补滤波融合WiFi和PDR的行人室内定位

提出了一种基于互补滤波融合Wi Fi和PDR的行人室内定位方法。首先改善Wi Fi位置指纹定位的KNN算法,通过阈值的设定,排除相似度高但实际上不可能的点,获取动态K值;然后通过行人航位推算(PDR)初始化算法,动态轨迹概率计算,确定PDR初始位置;最后在改进的Wi Fi和PDR的定位基础上,基于互补滤波原理,根据Wi Fi和PDR定位的不同特性,利用各自的定位优点,使用Wi Fi定位修正PDR的定位结果,通过相应权重参数的调整,输出最终融合定位结果。试验过程中,选取3种不同的室内环境区域,试验结果证明了该算法可大大提高室内定位的精度和稳定性。

Integration of WiFi and PDR based complementary filtering indoor localization

This paper presents a pedestrian indoor location method based on complementary filtering fusion WiFi and PDR. Firstly, we improve the KNN algorithm of WiFi position fingerprint location. By setting threshold, we can get the dynamic K value by eliminating the points with high similarity but practically impossible. Secondly, we determine the initial position of PDR by initializing the pedestrian dead reckoning (PDR) algorithm and calculating the dynamic trajectory probability. Finally, on the basis of the improved positioning of WiFi and PDR, based on the principle of complementary filtering, according to the different characteristics of WiFi and PDR positioning, using their respective positioning advantages, using WiFi positioning to modify the positioning results of PDR, through adjusting the corresponding weight parameters, the final fusion positioning results are output. During the experiment, we choose three different indoor environment areas. The experimental results show that the algorithm can greatly improve the accuracy and stability of indoor positioning.