基于WLAN指纹和惯性测量的室内定位系统设计与研究
发布时间：2019-01-08 20:32
【摘要】：随着无线通信技术的发展和社会的进步，人们对无线定位服务的需求越来越大。卫星导航定位是一种传统的定位方式，通常能够对室外场景下的定位提供解决方案，但在室内等建筑物遮蔽严重的场景下无法完成定位，所以室内定位问题在近年来成为一个越来越热门的研究方向，其中，基于WLAN的室内定位技术应运而生，发展日趋成熟。本文提出一种基于WLAN指纹定位和惯性测量的室内定位系统设计方案，并实现了基于Android手机终端的室内定位系统。 本文提出的室内定位系统设计方法融合WLAN指纹定位的结果和惯性测量的结果，并在指纹定位技术中采用终端独立定位和网络侧定位两种模式，在惯性测量中提出了一种基于加速度向量欧式距离的静止运动判断方法，判断当前行人的静止运动状态，在静止时对指纹定位结果进行均值滤波处理，之后，，提出了基于加速度峰值检测的步态检测算法，以及步长估计算法，采用航迹推算的方式得到用户的计步状态。进一步，在指纹定位结果和惯性测量结果的融合方面，提出了一种基于扩展Kalman滤波的融合算法。最后，在系统实现方面本文给出了定位系统的总体设计架构，重点对基于Android手机终端的定位软件设计做了阐述。 本文在系统设计和实现的基础上对系统进行了测试及性能分析。测试结果表明，在本文的实验场景下，采用K值为4的WKNN算法得到的指纹定位精度较高，误差累计概率在1处的定位误差为1.3米左右，通过引入惯性测量的方法，当判断行人处于静止状态时对指纹定位结果进行均值滤波，精度在误差累计概率处有0.2米左右的提高。在运动情况下，指纹定位加入惯性测量并进行扩展卡尔曼滤波融合算法后得到的定位轨迹相比单纯使用指纹定位算法的轨迹更接近真实行人轨迹，验证了基于WLAN指纹定位和惯性测量的定位系统的可行性和系统定位性能的优越性。
[Abstract]:With the development of wireless communication technology and social progress, the demand for wireless location services is increasing. Satellite navigation and positioning is a traditional positioning method, which can usually provide a solution for outdoor location, but it can not be completed in the severely sheltered environment such as indoor buildings. Therefore, indoor positioning has become a hot research direction in recent years. Among them, the indoor positioning technology based on WLAN emerges as the times require, and the development is maturing day by day. This paper presents a design scheme of indoor positioning system based on WLAN fingerprint location and inertial measurement, and realizes the indoor positioning system based on Android mobile terminal. The indoor positioning system design method proposed in this paper combines the results of WLAN fingerprint location and inertial measurement, and adopts two modes of terminal independent positioning and network side positioning in fingerprint positioning technology. In inertial measurement, a method of judging static motion based on the Euclidean distance of acceleration vector is proposed. A gait detection algorithm based on peak acceleration detection and a step size estimation algorithm are proposed. Furthermore, a fusion algorithm based on extended Kalman filter is proposed for the fusion of fingerprint location results and inertial measurement results. Finally, in the aspect of system implementation, this paper gives the overall design framework of the positioning system, focusing on the design of location software based on Android mobile phone terminal. Based on the design and implementation of the system, this paper tests and analyzes the performance of the system. The test results show that the WKNN algorithm with K value of 4 has a high precision of fingerprint location, and the error accumulative probability is about 1. 3 meters. The method of inertial measurement is introduced. When judging that the pedestrian is in a static state, the fingerprint location results are filtered by the mean value, and the accuracy is improved by about 0.2 meters at the cumulative error probability. In the case of motion, the trajectory of fingerprint location is closer to the real pedestrian track than that of using the fingerprint localization algorithm, which is based on the inertial measurement and extended Kalman filter fusion algorithm. The feasibility and superiority of the positioning system based on WLAN fingerprint location and inertial measurement are verified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925.93
本文编号：2405083
[Abstract]:With the development of wireless communication technology and social progress, the demand for wireless location services is increasing. Satellite navigation and positioning is a traditional positioning method, which can usually provide a solution for outdoor location, but it can not be completed in the severely sheltered environment such as indoor buildings. Therefore, indoor positioning has become a hot research direction in recent years. Among them, the indoor positioning technology based on WLAN emerges as the times require, and the development is maturing day by day. This paper presents a design scheme of indoor positioning system based on WLAN fingerprint location and inertial measurement, and realizes the indoor positioning system based on Android mobile terminal. The indoor positioning system design method proposed in this paper combines the results of WLAN fingerprint location and inertial measurement, and adopts two modes of terminal independent positioning and network side positioning in fingerprint positioning technology. In inertial measurement, a method of judging static motion based on the Euclidean distance of acceleration vector is proposed. A gait detection algorithm based on peak acceleration detection and a step size estimation algorithm are proposed. Furthermore, a fusion algorithm based on extended Kalman filter is proposed for the fusion of fingerprint location results and inertial measurement results. Finally, in the aspect of system implementation, this paper gives the overall design framework of the positioning system, focusing on the design of location software based on Android mobile phone terminal. Based on the design and implementation of the system, this paper tests and analyzes the performance of the system. The test results show that the WKNN algorithm with K value of 4 has a high precision of fingerprint location, and the error accumulative probability is about 1. 3 meters. The method of inertial measurement is introduced. When judging that the pedestrian is in a static state, the fingerprint location results are filtered by the mean value, and the accuracy is improved by about 0.2 meters at the cumulative error probability. In the case of motion, the trajectory of fingerprint location is closer to the real pedestrian track than that of using the fingerprint localization algorithm, which is based on the inertial measurement and extended Kalman filter fusion algorithm. The feasibility and superiority of the positioning system based on WLAN fingerprint location and inertial measurement are verified.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925.93
【参考文献】
相关期刊论文 前2条
1 关桂霞;朱虹;关永;吴敏华;;GPS定位误差分析与状态估计[J];计算机工程;2008年01期
2 张维;李文涛;;一种基于位置指纹的室内WLAN定位方法[J];信息技术;2013年07期
本文编号：2405083
本文链接：https://www.wllwen.com/kejilunwen/wltx/2405083.html
