多传感器行人航位推算方法和UKF融合算法研究

发布时间：2018-05-29 13:14

  本文选题：行人航位推算 + 初始对准　； 参考：《南昌大学》2017年硕士论文

【摘要】：随着移动互联网快速的崛起,定位与导航技术被应用在诸多领域。在室外空旷的环境下,利用卫星信号的全球定位系统(Global Positioning System,GPS)可以较好地获得用户位置信息,然而在室内环境下,卫星信号受到阻隔很难获得准确的位置信息。目前,惯性导航系统(Inertial Navigation System,INS)依靠惯性测量单元(Inertial Measurement Unit,IMU)已成为主要的自主导航系统,但是高精度的IMU体积大且价格昂贵,很难推广使用。近些年,智能移动设备已经在人们生活中普及,且大部分都含有IMU等传感器。因此,本文利用智能移动设备的低成本传感器,提出了一种基于多传感器的室内行人航位推算方法,并且针对低成本传感器的问题,设计了相对应的误差修正模型,主要分为以下三个部分:1、初始对准:初始对准可以使INS所描述的坐标系与导航坐标系相重合,同时让计算机在正式工作的时候有正确的初始值。由于基于智能移动设备的IMU更易受到设备中其他元件的干扰。所以,本文研究在初始对准的精对准阶段引入无迹卡尔曼滤波,并且融合多传感器的数据,对多传感器误差进行修正,从而获取精确的初始信息。2、运动状态检测模型:行人运动时通过IMU获取正确的运动状态信息对于行人航位推算方法解算高精度位置、速度和姿态信息至关重要。当行人步伐状态差别较大时,仅依靠加速度计很难获取正确的行人步态信息。本文研究在智能移动设备多传感器硬件平台的基础上,利用加速度计、陀螺仪获取的运动数据,设定四种阈值条件进行步伐状态检测。3、多传感器行人航位推算方法:在初始对准、运动状态检测模型的基础上,对于多传感器工作时夹杂噪声和解算时误差累积的问题,提出一种基于无迹卡尔曼滤波的零速度更新、零角速率更新和磁力计融合的方法,有效地对航向角以及速度误差进行修正。经过多次实验以及数据分析,利用本文提出的方法得到的平均位置偏差占总路程的1.57%,可较好的满足室内定位需要。
[Abstract]:With the rapid rise of mobile Internet, positioning and navigation technology has been applied in many fields. In the outdoor open environment, the GPS (Global Positioning system) using satellite signal can obtain the user location information well, but in the indoor environment, it is difficult to obtain the accurate position information by blocking the satellite signal. At present, the inertial Navigation system (ins) has become the main autonomous navigation system depending on the inertial Measurement unit (IMU), but the high precision IMU is large and expensive, so it is difficult to be popularized. In recent years, smart mobile devices have become popular in people's lives, and most of them contain sensors such as IMU. Therefore, in this paper, a multisensor based indoor footpath estimation method is proposed by using the low cost sensor of intelligent mobile device, and the corresponding error correction model is designed to solve the problem of low cost sensor. The initial alignment can make the coordinate system described by INS coincide with the navigation coordinate system and at the same time make the computer have the correct initial value when it works. IMU based on intelligent mobile devices is more susceptible to interference from other components in the device. Therefore, in this paper, the unscented Kalman filter is introduced in the fine alignment phase of initial alignment, and the multi-sensor data is fused to correct the multi-sensor error. In order to obtain accurate initial information .2and the model of motion state detection: it is very important to obtain correct motion state information by means of IMU when pedestrian motion is used to calculate high precision position, velocity and attitude information of pedestrian dead-reckoning method. It is difficult to obtain correct pedestrian gait information by using accelerometers only when the pedestrian gait states are quite different. Based on the multi-sensor hardware platform of intelligent mobile devices, the motion data obtained by accelerometers and gyroscopes are studied in this paper. Four threshold conditions are set for step state detection. 3. Multi-sensor footpath estimation method: based on the initial alignment and motion state detection model, the problem of noise and error accumulation in multisensor operation is discussed. A method of zero velocity updating, zero angular rate updating and magnetometer fusion based on unscented Kalman filter is proposed, which can effectively correct the heading angle and velocity error. After many experiments and data analysis, the average position deviation obtained by the method proposed in this paper accounts for 1.57 of the total distance, which can better meet the indoor positioning needs.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212;TN96;TN713

【参考文献】

中国期刊全文数据库 前8条

1 杨建辉;王逸轩;刘玉红;;基于亚采样点级别的TDOA时延估计[J];无线电工程;2017年01期

2 张新喜;张嵘;郭美凤;程高峰;牛树来;;基于足绑式INS的行人导航三轴磁强计在线校准[J];清华大学学报(自然科学版);2016年02期

3 李金凤;王庆辉;刘晓梅;曹顺;张慕远;;基于MEMS惯性传感器的行人航位推算系统[J];传感器与微系统;2014年12期

4 田国会;张庆宾;丁娜娜;;基于WT-UKF的PDR/GPS组合定位算法[J];控制与决策;2015年01期

5 洪海斌;郭杭;郭俊鸽;方爽;滕长胜;;低成本传感器数据融合技术辅助iPhone导航研究[J];传感器与微系统;2014年06期

6 苏宛新;;自适应UKF滤波在SINS初始对准中的应用[J];中国惯性技术学报;2011年05期

7 黄凤荣;孙伟强;翁海娜;;基于UKF的旋转式SINS大方位失准角初始对准方法[J];中国惯性技术学报;2010年05期

8 胡宏灿;郭立;朱俊株;;卡尔曼滤波器在导航系统初始对准中的应用[J];微电子学与计算机;2006年02期

中国博士学位论文全文数据库 前2条

1 李由;基于MEMS惯性传感器、WiFi、磁场特征的移动智能终端室内行人导航算法[D];武汉大学;2015年

2 汪少初;基于PDR的定位与跟踪技术研究及系统设计[D];天津大学;2014年

中国硕士学位论文全文数据库 前3条

1 殷红;基于foot-mounted的IMU室内行人航迹推算研究[D];南昌大学;2013年

2 曾欣欣;基于单摄像头的360度旋转跟踪系统[D];昆明理工大学;2009年

3 鲁小强;INS/GPS组合导航系统初始对准滤波方法研究[D];国防科学技术大学;2006年

，

本文编号：1951004