基于移动终端的室内关键定位技术研究
发布时间：2018-11-16 07:39
【摘要】：随着位置服务的兴起和发展,作为位置服务前提的定位技术越来越受人们的重视。全球定位系统GPS作为一种广泛应用的定位技术,在开阔的室外环境中,能够提供精确的定位信息。然而,由于GPS信号无法良好地穿透建筑物,GPS在室内环境中无法实现精确定位。 随着智能移动终端的普及,基于移动终端的室内定位技术具有广泛的应用前景。目前室内定位研究领域最常见的定位技术主要包括位置指纹定位技术和航位推测技术。这两种技术分别在不同的应用背景下使用： (1)定位区域无线信号分布情况已知。基于接收信号强度(Received Signal Strength, RSS)的位置指纹定位技术因成本低、调度简单而成为研究热点。位置指纹定位技术要求在定位之前首先依据无线信号分布情况建立位置指纹数据库,定位时通过匹配当前无线信号测量值和位置指纹数据库中的记录来进行定位,因而位置指纹定位技术不能实现未知环境中的定位。为了获得精确的定位结果,位置指纹定位技术前期需要大量的数据采集和校准工作来建立指纹库,实际应用效率较低。 (2)定位区域无线信号分布情况未知。航位推测定位技术可在无线信号分布情况未知的情况下,利用移动终端的传感器测量单元进行惯性数据采集,通过惯性数据进行惯性自定位。但若惯性测量单元存在较大测量误差,定位误差会不断累积,定位精度随着时间迅速下降。 基于上述两种室内定位应用背景,本文在现有的室内关键定位技术基础上,研究基于普通移动终端的室内定位增强改进技术,以提高定位技术的精度和应用效率。本文依据空间无线信号的变化趋势规律,将高斯过程回归模型运用到位置指纹定位技术中,研究高效的位置指纹数据库构建方法,以减少指纹样本采集密度,提高位置指纹定位技术的应用效率。本文在航位推测定位技术中,引入了粒子滤波技术和位置指纹定位技术,以此来对惯性定位结果进行校正,从而解决惯性定位误差随惯性测量误差不断累积的问题。 本文对上述的定位技术方案进行了验证实验,实验结果表明：高斯过程回归模型能够很好地预测空间位置RSS,可利用高斯过程回归模型来对定位区域进行RSS预测,建立指纹库,减少信号样本采集密度；粒子滤波技术结合位置指纹技术应用到惯性定位系统中,能够显著抑制惯性导航定位的误差累积效应。
[Abstract]:With the rise and development of location service, people pay more and more attention to location technology as the premise of location service. Global Positioning system (GPS), as a widely used positioning technology, can provide accurate location information in open outdoor environment. However, because the GPS signal can not penetrate the building well, GPS can not locate accurately in the indoor environment. With the popularity of intelligent mobile terminals, indoor positioning technology based on mobile terminals has a wide range of application prospects. At present, the most common localization technology in indoor positioning research field mainly includes position fingerprint location technology and position estimation technology. The two techniques are used in different applications: (1) the distribution of wireless signals in the location area is known. Location fingerprint location technology based on received signal strength (Received Signal Strength, RSS) has become a research hotspot because of its low cost and simple scheduling. The location fingerprint location technology requires that the location fingerprint database be established according to the distribution of wireless signals before location, and the location is located by matching the current wireless signal measurement values and the records in the location fingerprint database. Therefore, the location fingerprint location technology can not realize the location in unknown environment. In order to obtain accurate location results, a large amount of data acquisition and calibration work is needed to establish fingerprint database in the early stage of location fingerprint location technology, and the practical application efficiency is relatively low. (2) the distribution of wireless signal in the location area is unknown. Under the condition that the wireless signal distribution is unknown, the mobile terminal sensor measurement unit can be used to collect the inertial data, and the inertial self-positioning can be carried out by inertial data. However, if there is a large measurement error in the inertial measurement unit, the positioning error will accumulate continuously, and the positioning accuracy will decrease rapidly with time. Based on the above two indoor positioning application background, this paper studies the indoor positioning enhancement and improvement technology based on the common mobile terminal based on the existing indoor key positioning technology, in order to improve the accuracy and application efficiency of the positioning technology. According to the changing trend of space wireless signal, Gao Si process regression model is applied to location fingerprint location technology, and an efficient location fingerprint database construction method is studied in order to reduce the collection density of fingerprint samples. Improve the application efficiency of position fingerprint location technology. In this paper, particle filter technique and position fingerprint technique are introduced to correct the result of inertial positioning, so as to solve the problem that the error of inertial positioning is accumulated with the error of inertial measurement. The experimental results show that Gao Si's process regression model can well predict the spatial location RSS, can make RSS prediction of the location area and establish the fingerprint database by Gao Si process regression model. Reducing the sampling density of signal samples; Particle filter technology combined with position fingerprint technology is applied to inertial positioning system, which can significantly suppress the error accumulation effect of inertial navigation positioning.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN92
本文编号：2334885
[Abstract]:With the rise and development of location service, people pay more and more attention to location technology as the premise of location service. Global Positioning system (GPS), as a widely used positioning technology, can provide accurate location information in open outdoor environment. However, because the GPS signal can not penetrate the building well, GPS can not locate accurately in the indoor environment. With the popularity of intelligent mobile terminals, indoor positioning technology based on mobile terminals has a wide range of application prospects. At present, the most common localization technology in indoor positioning research field mainly includes position fingerprint location technology and position estimation technology. The two techniques are used in different applications: (1) the distribution of wireless signals in the location area is known. Location fingerprint location technology based on received signal strength (Received Signal Strength, RSS) has become a research hotspot because of its low cost and simple scheduling. The location fingerprint location technology requires that the location fingerprint database be established according to the distribution of wireless signals before location, and the location is located by matching the current wireless signal measurement values and the records in the location fingerprint database. Therefore, the location fingerprint location technology can not realize the location in unknown environment. In order to obtain accurate location results, a large amount of data acquisition and calibration work is needed to establish fingerprint database in the early stage of location fingerprint location technology, and the practical application efficiency is relatively low. (2) the distribution of wireless signal in the location area is unknown. Under the condition that the wireless signal distribution is unknown, the mobile terminal sensor measurement unit can be used to collect the inertial data, and the inertial self-positioning can be carried out by inertial data. However, if there is a large measurement error in the inertial measurement unit, the positioning error will accumulate continuously, and the positioning accuracy will decrease rapidly with time. Based on the above two indoor positioning application background, this paper studies the indoor positioning enhancement and improvement technology based on the common mobile terminal based on the existing indoor key positioning technology, in order to improve the accuracy and application efficiency of the positioning technology. According to the changing trend of space wireless signal, Gao Si process regression model is applied to location fingerprint location technology, and an efficient location fingerprint database construction method is studied in order to reduce the collection density of fingerprint samples. Improve the application efficiency of position fingerprint location technology. In this paper, particle filter technique and position fingerprint technique are introduced to correct the result of inertial positioning, so as to solve the problem that the error of inertial positioning is accumulated with the error of inertial measurement. The experimental results show that Gao Si's process regression model can well predict the spatial location RSS, can make RSS prediction of the location area and establish the fingerprint database by Gao Si process regression model. Reducing the sampling density of signal samples; Particle filter technology combined with position fingerprint technology is applied to inertial positioning system, which can significantly suppress the error accumulation effect of inertial navigation positioning.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN92
【参考文献】
相关期刊论文 前10条
1 邵婷婷;马建仓;胡士峰;王超;;电子罗盘的倾斜及罗差补偿算法研究[J];传感技术学报;2007年06期
2 郭红成;罗海勇;尹浩;韩霜;;基于线性插值和动态指纹补偿的分布式定位算法[J];传感技术学报;2009年12期
3 夏克寒;许化龙;张朴睿;;粒子滤波的关键技术及应用[J];电光与控制;2005年06期
4 王俊;;浅析测绘技术的新发展——室内定位技术[J];科教文汇(下旬刊);2012年10期
5 李晓阳;;WiFi技术及其应用与发展[J];信息技术;2012年02期
6 李昊;;位置指纹定位技术[J];山西电子技术;2007年05期
7 边平勇;;粒子滤波算法在贝叶斯模型中的应用[J];统计与决策;2009年14期
8 卢志忠;杨先飞;;零速校正在惯性导航系统中的仿真研究[J];应用科技;2009年06期
9 陈岭;许晓龙;杨清;陈根才;;基于三次样条插值的无线信号强度衰减模型[J];浙江大学学报(工学版);2011年09期
10 付强文;秦永元;李四海;;速度约束辅助车载捷联惯导系统零速校正算法[J];系统工程与电子技术;2013年08期
相关博士学位论文 前1条
1 贺建军;基于高斯过程模型的机器学习算法研究及应用[D];大连理工大学;2012年
，本文编号：2334885
本文链接：https://www.wllwen.com/kejilunwen/wltx/2334885.html
