对当前室内行人定位算法进行了研究.针对WiFi定位稳定性差的问题,提出了一种改进的K最近邻(Improved K-NearestNeighbor,IKNN)算法.针对行人航位推算(Pedestrian Dead Reckoning,PDR)算法中步长模型及航向估计不准确的问题,提出了一种实时更新的步长模型及基于室内环境特征的航向估计算法.在改进的WiFi定位算法与PDR算法的基础上,提出了一种基于自适应粒子滤波的室内行人WiFi与PDR组合定位算法,通过自适应因子自动调节观测量对粒子分布的影响.通过智能手机在实际室内环境中对定位方法进行了测试,实验结果表明:组合定位系统定位精度为0.66 m,高于普通的粒子滤波算法,是一种准确高效的室内行人定位算法.%In view thatWiFi localizationhas poorinstabilityin indoor pedestrianlocalization, animproved K-nearest neighboralgorithm is proposed toovercome thisproblem. A real-time updated step-length model and a heading estimation algorithm basedon indoorenvironmental featuresare proposed to improve the positioning accuracy ofpedestrian deadreckoning. In addition, aself-adaptive particle filtering algorithm is used to integrate the WiFiwiththepedestrian dead reckoning.An adaptive factoris usedtoautomatically adjustthe influence of WiFi observations on particle movements. A series ofexperiments were implemented on mobile phone,and the results show thatthe proposedintegration localization strategy achieves 0.66m location accuracywhich isbetter thanthat ofthe traditionalparticle filtering algorithm.
展开▼
