面向室内语音导航的“附近”关系协同定位方法  

A near spatial relationship collaborative indoor positioning for voice navigation

作　　者：吴志芳[1] 尹宾宾 樊红[3] 王彦坤 樊勇 WU Zhifang;YIN Binbin;FAN Hong;WANG Yankun;FAN Yong(School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,China;Tianjin Feye UAV Technology Co.,Ltd.,Tianjin 430079,China;State Key Lab for Information Engineering in Surveying Mapping and Remote Sensing,Wuhan University,Wuhan 430072,China;Key Laboratory of Urban Land Resources Monitoring and Simulation,MNR,Shenzhen 518034,China;Internet of Things Research Institute&School of Artifcial Intelligence,Shenzhen Polytechnic,Shenzhen 518055,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,武汉430065 [2]天津飞眼无人机科技有限公司,天津430079 [3]武汉大学测绘遥感信息工程国家重点实验室,武汉430072 [4]自然资源部城市国土资源监测与仿真重点实验室,广东深圳518034 [5]深圳职业技术学院物联网研究院&人工智能学院,广东深圳518055

出　　处：《导航定位学报》2024年第2期147-154,共8页Journal of Navigation and Positioning

基　　金：国家自然科学基金面上项目(41971341);国家自然科学基金青年基金项目(42001389);广东省普通高校重点领域专项(2022ZDZX3071);自然资源部城市国土资源监测与仿真重点实验室开放基金资助课题(KF-2022-07-024);深圳职业技术学院博士后出站后期资助项目(6021271017K);深圳职业技术学院项目(6022312062K)。

摘　　要：针对传统室内定位模式单一,且单一“附近”空间关系无法满足定位需求的问题,提出一种融合“附近”空间关系/无线保真(WiFi)/行人航迹推算(PDR)/地磁的室内定位方法:阐述多源融合语音交互室内定位方法框架;并针对室内环境,建立“附近”空间关系概率密度函数;然后步行构建地磁/WiFi指纹库;最后后通过“附近”空间关系约束粒子及指纹匹配空间范围,在PDR推算位置基础上,利用WiFi/地磁/“附近”空间关系更新粒子权重,实现最优位置估计。实验结果表明,基于粒子滤波融合单一“附近”空间关系/WiFi/PDR/地磁为一种低成本可行的定位方法,平均定位精度1.93 m,有90%可能性定位精度可达2.52 m。Aiming at the problem that single traditional positioning mode and single near spatial relationship cannot meet the requirements of positioning,the paper proposed an indoor positioning method based on the particle filter fusion of near relationship/wireless fidelity(WiFi)/pedestrian dead reckoning(PDR)/geomagnetic:the framework of multi-source fusion voice interactive indoor positioning method was described;and according to indoor environment,the probability density function of near spatial relation was established;then,the geomagnetic/WiFi fingerprint database was constructed by step-index;finally,the space range of particle and fingerprint matching was constrained by the near spatial relationship,and based on the location calculated by PDR,the weight of particles was updated by the WiFi/geomagnetic/near spatial relationship to obtain the optimal location estimation.Experiment result showed that the proposed method would be a low-cost and feasible localization method with the mean positioning accuracy 1.93 m,and 90%possibility of positioning accuracy up to 2.52 m.

关 键 词：室内定位 “附近”空间关系 无线保真(WiFi)指纹 行人航迹推算(PDR) 粒子滤波 

分 类 号：P228[天文地球—大地测量学与测量工程]