基于超宽带位置指纹与惯导融合的室内定位技术研究

Research on Indoor Location Technology Based on UWB Location Fingerprint and IMU Fusion

在室内定位环境复杂多变的情况下,超宽带(UWB)信号易受多径干扰和非视距环境影响,导致定位精度不高。鉴于指纹定位和惯导定位不受多径干扰和非视距环境影响,以及UWB信号具有时间分辨率高和测距精确的特点,以由非对称双边双向测距算法获得的UWB测距值为指纹库匹配数据,建立了结合K中心聚类和加权K近邻的UWB指纹定位算法;提出了利用扩展卡尔曼滤波将UWB指纹定位与惯导定位进行融合的融合定位算法。定位实验结果显示,该融合定位算法能将平均定位误差降至0.248 m,对有效提高室内定位系统的可靠性和定位精度具有重要作用。

In the complex and ever-changing indoor positioning environment,ultra-wide band(UWB)signals are susceptible to mulipath interference and non-line-of-sight environments,resulting in insufficient positioning accuracy.Considering that fingerprint localization and inertial navigation localization are not affected by multipath interference and non-line of sight environments,as well as the high temporal resolution and precise ranging characteristics of UWB signals,a UWB fingerprint localization algorithm combining K-center clustering and weighted K-nearest neighbors is established using the UWB ranging values obtained from the ADS-TWR algorithm as fingerprint library matching data.A fusion localization algorithm that fused the extended Kalman filtering and UWB position fingerprints with micro inertial units is proposed.The positioning experiment results show that the fusion positioning algorithm can reduce the average positioning error to 0.248 m,which plays an important role in effectively improving the reliability and positioning accuracy of indoor positioning systems.