基于WIFI的隧道人员定位算法的分析与研究

发布时间：2018-04-30 21:00

本文选题：WIFI定位 + 位置指纹算法　；参考：《武汉邮电科学研究院》2017年硕士论文

【摘要】：随着科学技术的发展,人们对于获取自身位置的需求越来越多,基于位置的服务获得了越来越多的关注。目前广泛使用的GPS定位技术,可以快速精确的提供定位服务,广泛的应用于手机导航等方面,但是由于GPS信号穿透墙壁时衰减严重,因此GPS技术不能用于室内定位。近几年来WIFI技术发展迅速,WIFI覆盖率大幅上升,基于WIFI技术实现隧道人员定位,无论从组网能力、经济效益、还是定位精度上来说都是非常适合的。基于WIFI实现室内定位,一般使用RSSI(接收信号强度指示)作为信号的特征值。RSSI定位算法有两大类,一种是基于测距的算法,另一种是无需测距的算法。基于测距的定位算法是根据无线信号的自由空间中的传输损耗模型,对比目标位置RSSI值来确定位置,对环境有非常大的依赖性。无需测距的RSSI定位算法即位置指纹算法,分为离线指纹采集和在线位置估计两个阶段。离线指纹采集阶段需要采集每一个位置的RSSI值建立指纹数据库。在线位置估计阶段,取得目标位置的RSSI值,与数据库匹配,估算出目标位置。本论文选取了隧道人员定位作为研究方向,在现有的WIFI定位的基础上,分析了现有定位算法的优缺点,并作出相应的改进,主要内容可归纳如下。(1)分析了基于测距的定位算法,采用了基于传输损耗模型获得待测点与参考点的距离,采用加权质心的极大似然估计方法估算位置,提高了测距算法的定位精度。(2)为了提高传统位置指纹定位算法的定位精度,在离线指纹采集阶段,对于指纹数据进行降噪处理。使用聚类分析的方法训练指纹数据库,降低算法的时间复杂度。在在线位置估计阶段,则采用改进的最邻近算法去来估算位置。这种方法可以有效的提高传统的位置指纹算法的定位精度。(3)测试改进后的定位算法的可行性,比较改进前后的算法的定位精度。分析影响定位精度的因素。
[Abstract]:With the development of science and technology, more and more people need to obtain their own location, and more and more attention has been paid to location-based services. At present, the widely used GPS positioning technology can quickly and accurately provide positioning services, widely used in mobile phone navigation and other aspects, but because the GPS signal attenuation through the wall, GPS technology can not be used for indoor positioning. With the rapid development of WIFI technology in recent years, the coverage rate of WiFi has increased rapidly. It is very suitable for tunnel personnel positioning based on WIFI technology in terms of network ability, economic benefit and positioning accuracy. RSSI (received signal intensity indication) is generally used as the eigenvalue of the signal based on WIFI. There are two kinds of localization algorithms: one is based on ranging and the other is without ranging. The location algorithm based on ranging is based on the transmission loss model in the free space of wireless signal. The location is determined by comparing the RSSI value of the target location, which is highly dependent on the environment. The RSSI location algorithm without ranging, that is, the location fingerprint algorithm, is divided into two stages: off-line fingerprint acquisition and on-line position estimation. In the off-line fingerprint acquisition stage, the RSSI value of each location should be collected to establish the fingerprint database. In the stage of online position estimation, the RSSI value of the target position is obtained, matched with the database, and the target position is estimated. This paper selects the tunnel personnel localization as the research direction, on the basis of the existing WIFI localization, analyzes the advantages and disadvantages of the existing localization algorithm, and makes the corresponding improvement, the main content can be summarized as follows: 1) analyze the location algorithm based on ranging. Based on the transmission loss model, the distance between the point to be measured and the reference point is obtained. The maximum likelihood estimation method of weighted centroid is used to estimate the location, which improves the location accuracy of the ranging algorithm. In the off-line fingerprint acquisition stage, the fingerprint data is de-noised. The method of clustering analysis is used to train the fingerprint database to reduce the time complexity of the algorithm. In the stage of online position estimation, an improved nearest neighbor algorithm is used to estimate the position. This method can effectively improve the location accuracy of the traditional location fingerprint algorithm. It can test the feasibility of the improved location algorithm and compare the location accuracy of the improved algorithm. The factors influencing the positioning accuracy are analyzed.
【学位授予单位】：武汉邮电科学研究院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92

【参考文献】