采用EKF与PF的室内融合定位技术

在室内定位的复杂环境中,WiFi指纹法的定位性能易受信号波动的影响,行人航位推算方法(Pedestrian Dead Reckoning,PDR)的定位误差会随时间逐渐累积,为了解决这两个方面的问题,提出了一种采用扩展卡尔曼滤波(Extended Kalman Filter,EKF)与粒子滤波(Particle Filter,PF)的融合定位技术,该技术通过融合WiFi指纹法和优化后的行人航位推算方法来提高定位精度。在WiFi指纹法模块,该技术采用传统的加权k近邻法进行定位。在行人航位推算模块,该技术融合多种传感器进行航向估计,改进传统的非线性模型进行步长估计,使优化后的行人航位推算方法更适用于复杂环境中的实际应用。最后,结合粒子滤波和室内地图信息,该技术能够校正估计位置并进一步提高定位精度。通过实验表明,本文提出的融合校正方法能够有效提高定位精度。

Indoor Fused Positioning Technology Using EKF and PF

In a complex environment of indoor positioning,the positioning performance of WiFi fingerprinting is easily affected by signal fluctuation,the positioning error distance of Pedestrian Dead Reckoning(PDR)will gradually accumulate with time,in order to solve these two problems,this paper presents a fused positioning technology using Extended Kalman Filter(EKF)and Particle Filter(PF). This technology improves the positioning accuracy by combining WiFi fingerprinting and optimized Pedestrian Dead Reckoning. In WiFi fingerprinting module,the technology uses traditional weighted k-Nearest Neighbor to locate a user. In PDR module,the technology fuses multiple sensors for orientation estimation and improves traditional non-linear model for stride length estimation to make the optimized PDR more suitable for practical use in complex environments. Finally,combined with particle filter and indoor map information,it can correct the estimated location and further improve the positioning accuracy. Our experiment results show that the fusing and correcting method proposed in this paper can effectively improve the positioning accuracy.