基于ZigBee技术的室内定位方法优化研究

发布时间：2018-01-26 05:03

本文关键词： ZigBee无线传感器网络技术 Kalman滤波高斯滤波虚拟空间划分加权距离模糊聚类　出处：《南昌大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着时代的发展,特别是伴随着移动互联网的高速发展,导航与定位越来越受到人们的关注,同时与位置有关的服务成为了当前的一个热门产业。定位技术包含有室内外这两个方面。对于室外获取地理位置信息,全球卫星定位系统(GPS)和网络辅助全球卫星定位系统(A-GPS)以其高定位精度为人们所熟知,并且在目前来说已经是比较成熟的室外定位技术;然而在室内环境下,室内定位技术在目前来讲还并不很成熟,所以寻找一种有效的定位技术方案成为了目前定位服务领域的一个研究关键。随着无线传感器网络技术的出现,特别是作为一种新型的短距离无线传感器网络技术的ZigBee,正作为室内定位的一种行之有效的解决方案为越来越多的人去探索和研究。基于ZigBee的独特优势,本文选取的室内定位的研究方案是基于ZigBee的无线传感器网络技术。在阐述了研究的意义和背景的前提下,详细介绍了无线定位技术,然后与其他无线定位技术比较,ZigBee无线传感器网络技术有其优越的条件来作为室内定位系统搭建的技术基础。然而,室内环境的复杂性制约了ZigBee室内定位精度的提高,本文旨在不额外增加硬件条件的基础上,使用基于RSSI的指纹数据库室内定位方法对ZigBee室内定位精度进行优化,并且提出了可行性方案,即对信号源的处理和定位算法。通过对室内环境复杂环境的研究,本文提出了在对移动节点接收到的信号源进行滤波处理,包括Kalman滤波和高斯滤波这两种方式。通过实验可以得出,在最优的情况下,在指纹数据库定位方法中的两个阶段使用Kalman滤波方式对信号源进行处理,使用最邻近定位算法进行定位计算,最终将定位误差抑制在1.86m以内,比未做滤波处理的结果提高了0.1m,一定程度上提高了定位精度。在基于对常规定位算法研究的基础上,本文提出了三种新的室内定位算法,分别为虚拟空间划分的室内指纹库定位算法、一种加权距离室内指纹库定位算法、模糊聚类室内定位算法。在最优的情况下,虚拟空间划分的室内指纹库定位算法可以将定位误差抑制在1.50m以内,加权距离室内指纹库定位算法可以将定位误差抑制在1.57m以内,模糊聚类室内定位算法可以将定位误差抑制在1.47m以内,都在一定程度上使定位精度得到优化。
[Abstract]:With the development of the times, especially with the rapid development of mobile Internet, people pay more and more attention to navigation and positioning. At the same time location-related services have become a hot industry. Location technology includes indoor and outdoor aspects of geo-location information for outdoor access. GPS (Global Positioning system) and network-aided GPS (A-GPS) are well known for their high positioning accuracy, and are mature outdoor positioning technologies at present. However, in the indoor environment, indoor positioning technology is not very mature. Therefore, the search for an effective location technology has become a research key in the field of location services. With the emergence of wireless sensor networks (WSN) technology. Especially as a new short-range wireless sensor network technology, ZigBee. Is an effective solution for indoor positioning for more and more people to explore and research. Based on the unique advantages of ZigBee. In this paper, the indoor positioning research scheme is based on ZigBee wireless sensor network technology. On the premise of elaborating the significance and background of the research, the wireless location technology is introduced in detail. Then compared with other wireless positioning technology ZigBee wireless sensor network technology has its advantages to serve as the technical foundation of indoor positioning system. However. The complexity of indoor environment restricts the improvement of indoor positioning accuracy of ZigBee. The indoor positioning accuracy of ZigBee is optimized by using the indoor location method of fingerprint database based on RSSI, and the feasible scheme is put forward. That is the signal source processing and location algorithm. Through the study of the complex indoor environment this paper proposed to filter the signal source received by the mobile node. Including Kalman filtering and Gao Si filtering. Through experiments, we can find that in the best case. In the two stages of fingerprint database location method, the Kalman filter is used to process the signal source, and the nearest neighbor location algorithm is used to locate the signal source. Finally, the localization error is restrained within 1.86m, which is 0.1 m higher than the result without filtering, and the accuracy is improved to some extent. Based on the research of the conventional localization algorithm. In this paper, three new indoor location algorithms are proposed, they are virtual space partition indoor fingerprint database location algorithm, a weighted distance indoor fingerprint database location algorithm. Fuzzy clustering indoor location algorithm. Under the optimal condition, the location error of indoor fingerprint database can be restrained within 1.50m by virtual space partition. The location error can be restrained within 1.57m by the weighted distance indoor fingerprint database localization algorithm, and the location error can be restrained by the fuzzy clustering indoor location algorithm within 1.47m. To a certain extent, the positioning accuracy is optimized.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN92;TP212.9

【参考文献】