基于手机惯性传感器的室内轨迹跟踪技术研究

发布时间：2018-06-07 17:38

本文选题：行人航位推算 + 地图匹配　；参考：《宁夏大学》2017年硕士论文

【摘要】：信息技术的快速发展,使得室内轨迹定位跟踪技术逐渐走进人们的日常工作与生活中。常用的轨迹定位跟踪技术主要依靠全球定位系统(GlobalPositioningSystem,GPS),但由于室内环境的遮挡,卫星信号无法提供满足用户要求的室内定位服务。因此,室内轨迹定位跟踪技术逐渐成为研究与发展的热点。本文利用移动终端内部嵌入的惯性传感器开展室内环境下行人航位推算(PedestrianDead Reckoning,PDR)研究,基于手机内部的传感器进行室内行人运动轨迹的定位跟踪,具有受外界环境干扰小、使用灵活便捷和低成本等优点。根据行人在运动过程中,惯性传感器采集得到数据的变化情况,对行人进行步态检测、步长估计,航向估计。并针对行人步态探测过程中存在的伪波峰/波谷导致步态误判的现象,提出了自适应步态估计算法。针对通过陀螺仪和电子罗盘进行航向测量估计过程中,估计偏差间存在的互补特性,经过卡尔曼滤波对两种估计的航向信息进行融合,进而提高了航向估计精度。其次,针对惯性传感器进行行人运动轨迹定位跟踪过程中存在的偏差现象,本文借鉴车辆导航系统中的地图匹配理论,并根据室内环境的特点,改进了原有的匹配候选点获取规则。结合室内环境的空间拓扑关系,引入隐马尔可夫模型(Hidden Markov Model,HMM),提出了适用于室内空间结构的地图匹配算法。通过地图信息对行人航位推算的航向和坐标进行辅助校正,进一步提高运动轨迹定位跟踪精度。最后,根据本文系统模型,进行了算法软件系统的开发,并基于Android平台手机验证了所设计的室内轨迹定位跟踪系统的性能。
[Abstract]:With the rapid development of information technology, indoor location-tracking technology has gradually come into people's daily work and life. The commonly used location-tracking technology mainly relies on Global Positioning system (GPS). However, because of the occlusion of indoor environment, satellite signals can not provide indoor positioning services to meet the needs of users. Therefore, indoor locus location and tracking technology has gradually become a hot spot in research and development. In this paper, we use the inertial sensor embedded inside the mobile terminal to carry out the research of the Pede strian Dead Recognition (PDR) in the indoor environment. The sensor based on the inside of the mobile phone is used to locate and track the indoor pedestrian trajectory, which is less disturbed by the external environment. Flexible and convenient use and low cost and other advantages. According to the change of the data collected by the inertial sensor in the course of pedestrian movement, the gait detection, step size estimation and course estimation of the pedestrian are carried out. An adaptive gait estimation algorithm is proposed to solve the problem of gait misjudgment caused by pseudo wave crest / valley in the process of pedestrian gait detection. Aiming at the complementary characteristics between the estimated deviations in the course of course estimation by gyroscope and electronic compass, the course information of the two kinds of estimation is fused by Kalman filter, and the precision of course estimation is improved. Secondly, aiming at the deviation phenomenon in the course of inertial sensor track tracking, this paper draws lessons from the map matching theory of vehicle navigation system, and according to the characteristics of indoor environment, The original matching candidate point acquisition rules are improved. Combined with the spatial topological relation of indoor environment, the hidden Markov model (Hidden Markov Model) is introduced, and a map matching algorithm suitable for indoor spatial structure is proposed. By using map information, the course and coordinates of pedestrian track reckoning can be adjusted to improve the accuracy of track location and tracking. Finally, according to the system model of this paper, the algorithm software system is developed, and the performance of the indoor location-tracking system is verified based on the mobile phone of Android platform.
【学位授予单位】：宁夏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212

【参考文献】