随着室内定位需求日益增多，基于位置信息的服务发展迅速。然而由于建筑物墙体遮挡以及室内环境带来的影响，室外精准定位GPS技术不能为室内环境提供可靠的定位。目前，室内定位技术及方法包括Wi-Fi、射频识别、超宽带、蓝牙等，大多依赖于外部设备。如何在保持定位精度的同时简化定位设备的部署是当前的研究重点。鉴于智能手机的广泛普及，本文采用基于智能手机的行人航位推算(PDR)和地磁定位融合的方式进行室内定位研究。

为了提高PDR的定位精度，本文在传统的峰值检测法基础上，提出一种约束型步伐检测算法，对前后时刻的加速度变化值进行波峰波谷的判定，结合周期性、连续性、相似性和对称性约束，有效滤除无效步伐，准确率高于98%。同时，本文通过研究分析现有步长估计模型，提出基于加速度变化值的改进型Weinberg步长估计模型，实验证明本文改进的步长模型误差不超过1.23%。在航向估计问题上，通过分析单独使用陀螺仪和磁力计解算方向角的优缺点，利用卡尔曼滤波算法融合陀螺仪和磁力计进行航向估计并提出数字地图结合粒子滤波的方法解决了“位置穿墙”问题。针对地磁定位中地磁匹配时需要全局搜索的问题，本文提出结合单点匹配和基于动态时间归整(DTW)序列匹配的混合匹配方式用于地磁匹配。采取单点法采集地磁数据后通过克里金插值算法构建出双精度的地磁指纹基准库进行混合匹配，有效简化地磁匹配过程。为了进一步提高定位精度，本文提出一种基于PDR与地磁融合的室内定位算法。该算法利用PDR定位结果和地磁混合匹配方法，有效解决了PDR定位误差累积和地磁匹配搜索时间长范围过广的问题。

经过实验验证，在各类情况下，PDR与地磁融合的定位算法定位精度小于1 m的概率为62%，最大定位误差小于1.5 m，可以满足室内定位的高精度需求。

With the rapid development of location-based services, the demand for indoor positioning is increasing day by day. However, due to the blocking of building walls and the influence of indoor environment, the high-precision GPS positioning technology used for outdoor can not provide reliable positioning in indoor environment.At present, most technologies applicable to indoor positioning rely on external devices, including Wi-Fi, RFID, UWB, Bluetooth, etc. How to simplify the deployment of positioning devices while maintaining positioning accuracy is the current focus of research.In view of the widespread popularity of smart phones, this thesis adopts the method of integrating Pedestrian Dead Reckoning (PDR) and geomagnetic positioning to conduct indoor positioning research.

In order to improve the positioning accuracy of PDR, a constrained step detection algorithm based on the traditional peak detection method is proposed, which judges the peak and trough of the acceleration change value at the front and back moments, combining periodicity, continuity, similarity and symmetry Constraint, effectively filter out invalid steps, and the accuracy rate is higher than 98%. At the same time, by studying and analyzing the existing step size estimation models, an improved Weinberg step size estimation model based on the change value of acceleration proposed in this thesis. Experiments show that the error of the improved step size model in this thesis is less than 1.23%. Finally, through the analysis of the advantages and disadvantages of using gyroscope and magnetometer alone to calculate the direction Angle, the Kalman filter algorithm is used to fuse gyroscope and magnetometer to estimate the course, and the digital map combined with particle filter method to solve the problem of "position through the wall".In order to solve the problem that global search matching is needed in geomagnetic matching in geomagnetic positioning, a hybrid matching method combining single point matching and sequence matching based on DTW is proposed for geomagnetic matching. Firstly, the single point method is used to collect the geomagnetic data, and then the double precision geomagnetic fingerprint reference library is constructed by the Krigin interpolation algorithm, and the mixed matching is carried out at last.In order to further improve the positioning accuracy, the indoor positioning algorithm based on PDR and geomagnetic fusion is proposed in this thesis, and the problems of PDR positioning error accumulation and geomagnetic matching long time and wide range are effectively solved by using PDR positioning results and geomagnetic mixing matching.

Through experimental verification, under all kinds of circumstances, the positioning accuracy of the fusion positioning algorithm with 62% probability is 1m, and the maximum positioning error is less than 1.5 m, which meets the daily indoor positioning needs.

[1] Walsh P B, Sethi S A, Lake B C, et al. Estimating denning date of wolves with daily movement and GPS location fix failure[J]. Wildlife Society Bulletin, 2016.,40(4):663-668.

[2] JingBin L, RuiZhi C, YuWei C, et al. iParking: An Intelligent Indoor Location-Based Smartphone Parking Service[J]. Sensors, 2012, 12(11):14612-14629.

[3] 赵琪, 孙立双, 谢志伟. 基于改进MCNN的密度图在室内定位中的应用[J]. 测绘通报, 2020(06):12-16.

[4] 沈雪微, 冯宁, 张玲玲. 以WiFi和ZigBee联合定位的消防灭火救援系统[J]. 物联网技术, 2015, 5(01):32-35.

[5] 周晓晨. 基于iBeacon的蓝牙室内定位算法研究[J].电脑知识与技术, 2019,15(25):258-259.

[6] 刘佳玲. 基于RFID的定位系统[J]. 电子技术与软件工程, 2018, 16:29-29.

[7] 黎海涛, 齐双. 基于室内地图环境信息的多楼层WiFi定位技术研究[J]. 电子科技大学学报, 2017, 46(01):32-37.

[8] 赵红梅, 赵杰磊. 超宽带室内定位算法综述[J]. 电信科学, 2018, 34(9): 130-142.

[9] 原璟, 程承旗, 童晓冲, 等. 基于PDR和地理网格的室内外一体化定位方法[J]. 地理与地理信息科学, 2016, 32(006):32-36.

[10] 黄斌, 孙永荣, 王丽娜, 等. 地磁导航的迭代评价匹配算法[J]. 南京航空航天大学学报, 2012,44(04):565-569.

[11] Li Y, Zhuang Y, Lan H, Fet Fal. A Hybrid WiFi/Magnetic Matching/PDR Approach For Indoor Navigation with Smartphone Sensors [J], Communications Letters, IEEE, 2015, 99:1-1.

[12] Daniel B , Kyle K , John R , et al. Non-GNSS Smartphone Pedestrian Navigation Using Barometric Elevation and Digital Map-Matching[J]. Sensors, 2018, 18(7):2232-2248.

[13] Lee J H, Cho S Y, Park C G. Symmetric Position Drift of Integration Approach in Pedestrian Dead Reckoning with Dual Foot-mounted IMU[J]. Journal of Positioning, Navigation, and Timing, 2020, 9(2): 117-124.

[14] Guo Y, Ji X, Guo J, et al. Improved Heuristic Drift Elimination Algorithm Optimisation Based on a Smartphone Gyroscope[J]. The Journal of Navigation, 2020, 73(3): 581-598.

[15] Kang W , Han Y . Smart PDR: Smartphone-Based Pedestrian Dead Reckoning for Indoor Localization[J]. IEEE Sensors Journal, 2015, 15(5):2906-2916.

[16] Wang Q, Ye L, Luo H, et al. Pedestrian stride-length estimation based on LSTM and denoising autoencoders[J]. Sensors, 2019, 19(4):840-863.

[17] Pei L, Liu D, Zou D, et al. Optimal heading estimation based multidimensional particle filter for pedestrian indoor positioning[J]. IEEE Access, 2018, 6:49705-49720.

[18] Elhoushi M , Georgy J , Noureldin A , et al. Motion Mode Recognition for Indoor Pedestrian Navigation Using Portable Devices[J]. IEEE Transactions on Instrumentation & Measurement, 2015, 65(1):208-221.

[19] Wang B , Liu X , Yu B , et al. Pedestrian Dead Reckoning Based on Motion Mode Recognition Using a Smartphone[J]. Sensors, 2018, 18(6):1811-1833.

[20] Binhee, Kim, Seung-Hyun, et al. A Novel Indoor Positioning Technique Using Magnetic Fingerprint Difference[J]. IEEE Transactions on Instrumentation & Measurement, 2016, 65(9):2035-2045.

[21] Carlos G T , Juan G V , Ramon B . Magnetic Field Feature Extraction and Selection for Indoor Location Estimation[J]. Sensors, 2014, 14(6):11001-11015.

[22] Imran A , Soojung H , Yongwan P . mPILOT-Magnetic Field Strength Based Pedestrian Indoor Localization[J]. Sensors, 2018, 18(7):2283-2304.

[23] 张聪聪, 王新珩, 董育宁. 基于地磁场的室内定位和地图构建[J]. 仪器仪表学报, 2015, 36(01):181-186.

[24] Shao W, Zhao F, Wang C, et al. Location fingerprint extraction for magnetic field magnitude based indoor positioning[J]. Journal of Sensors, 2016,40:663-668

[25] Lee N, Ahn S, Han D. AMID: Accurate Magnetic Indoor Localization Using Deep Learning. Sensors (Basel, Switzerland), 2018, 18(5):1598-1613.

[26] Donghun Kim, Hyochoong Bang ,and Jae Cheul Lee. Approach to geomagnetic matching for navigation based on a convolutional neural network and normalised cross-correlation[J]. IET Radar, Sonar & Navigation, 2019, 13(8) : 1323-1332.

[27] Kok M , Hol J D, Schon T B. Indoor Positioning Using Ultrawideband and Inertial Measurements[J]. IEEE Transactions on Vehicular Technology, 2015, 64(4):1293-1303.

[28] Yuanchao, Shu, Cheng Bo,Guobin Shen, et al. Magicol: Indoor Localization Using Pervasive Magnetic Field and Opportunistic WiFi Sensing[J]. Selected Areas in Communications, IEEE Journal on, 2015, 33(7):1443-1457.

[29] Li Y , Zhuang Y , Zhang P , et al. An improved inertial/wifi/magnetic fusion structure for indoor navigation[J]. Information Fusion, 2017, 34:101-119.

[30] Ju H, Park S Y, Park C G. A smartphone-based pedestrian dead reckoning system with multiple virtual tracking for indoor navigation[J]. IEEE Sensors Journal, 2018, 18(16):6756-6764.

[31] Chung J, Donahoe M, Schmandt C, et al. Indoor location sensing using geomagnetism [C]// In Proceedings of the 9th international conference on Mobile systems, applications, and services, 2011:141-154.

[32] 万群, 郭贤生, 陈章鑫. 室内定位理论、方法和应用[M]. 北京: 电子工业出版社, 2012:154.

[33] 刘夫玉, 王新珩, 董育宁, 等. 基于粒子滤波的定位系统中累计误差消除的进化策略[J]. 南京邮电大学学报(自然科学版), 2017, 37(02):91-97.

[34] 李鑫, 程德福, 周志坚. 一种基于地磁总场梯度的匹配定位算法[J]. 传感技术学报, 2017, 30(12):1869-1875.

[35] 周家鹏, 汪云甲, 李昕, 等. 地磁室内定位技术研究[J]. 测绘通报, 2019(01): 18-22.

[36] 刘宇, 惠鸿飞, 路永乐, 等. 基于遗传算法优化神经网络的多源信息融合室内定位方法[J]. 中国惯性技术学报, 2020, 28(01):67-73.

[37] Ma Y , Dou Z , Jiang Q , et al. Basmag: An Optimized HMM-Based Localization System Using Backward Sequences Matching Algorithm Exploiting Geomagnetic Information[J]. IEEE Sensors Journal, 2016, 16(20):7472-7482.

[38] 刘维, 王明杭, 朱志宇. 基于MEMS传感器的惯导系统预处理和姿态解算[J]. 传感技术学报, 2019,32(11):1677-1681.

[39] 宋红丽. 基于航位推算的室内定位系统研究与实现[D]. 成都:电子科技大学, 2018.

[40] 王彤, 马建仓, 秦涛, 等. 基于旋转四元数的姿态解算算法[J]. 弹箭与制导学报, 2014, 34(03):15-16.

[41] 赵延鹏, 时伟, 艾明曦. 室内行人航迹推算/超声波组合定位融合算法[J]. 中南大学学报(自然科学版), 2016, 47(05):1588-1598.

[42] 余佩, 韩勇强. 几种常用步态检测典型算法的设计比较[J]. 导航定位与授时, 2018, 04:21-26.

[43] 刘雷, 慕艳艳, 刘睿鑫. 基于三轴加速度传感器的步长估算模型研究[J]. 传感器与微系统, 2017, 36(08):22-24.

[44] Sun Z, Pan S, Su Y C, et al. Headio: Zero-Configured Heading Acquisition for Indoor Mobile Devices through Multimodal Context Sensing[J]. Cheminform, 2013, 44(39):258-272.

[45] Zhao Q, Zhang B, Wang J, et al. Improved method of step length estimation based on inverted pendulum model[J]. International journal of distributed sensor networks, 2017, 13(4):1-13.

[46] J. Scarlett. Enhancing the performance of pedometers using a single accelerometer[J]. Application Note,Analog Devices, 2007.

[47] J. W. Kim, H. J. Jang, D H Hwang, et al. A step, stride and heading determination for the pedestrian navigation system [J]. Positioning, 2004, 1(08):20-31.

[48] 陈诗军, 林利成, 徐小龙, 等. 一种面向位置数据隐私保护的离线地磁定位模型[J]. 南京信息工程大学学报(自然科学版), 2017, 9(05):509-515.

[49] 王安义, 欧雪. 基于粒子滤波的PDR/地磁指纹室内定位[J]. 测绘通报, 2021(01):24-28.

[50] 黄小平, 王岩, 缪鹏程. 目标定位跟踪原理及应用—MATLAB仿真[M]. 北京: 电子工业出版社, 2018:76.

[51] 王亚娜, 蔡成林, 李思民, 等. 基于粒子滤波的多信息融合室内定位算法研究 [J]. 电子技术应用, 2017, 43(11):70-73.

[52] Ngoc-Huynh, Ho, Phuc, et al. Step-Detection and Adaptive Step-Length Estimation for Pedestrian Dead-Reckoning at Various Walking Speeds Using a Smartphone[J]. Sensors, 2016, 16(9):1423-1435.

[53] 吴雅琴, 杨硕, 师兰兰. 基于位置指纹与 PDR 融合的室内定位算法研究[J]. 矿业科学学报, 2019, 4 (5):448 - 454.

[54] 于鹏, 韦照川, 蔡成林, 等. 基于PDR和地磁融合的室内定位算法[J]. 桂林电子科技大学学报, 2018, 38(04):273-278.

[55] Wang A Y, Ou X, Wang B. Improved Step Detection and Step Length Estimation Based on Pedestrian Dead Reckoning[C]// 2019 IEEE 6th International Symposium on Electromagnetic Compatibility (ISEMC). IEEE, 2019:1-4.

[56] 邓翠婷, 黄朝艳, 赵华, 等. 地磁匹配导航算法综述[J]. 科学技术与工程, 2012, 12(24):6125-6131.

[57] 张超. 基于序列匹配的地磁/PDR融合定位算法研究[D]. 武汉:中国矿业大学, 2019.

[58] 刘晓宇, 邓平. 一种融合多方向变异性的改进克里金插值算法[J]. 传感技术学报, 2020, 33(07):991-996..

[59] 盛方园, 郭英, 冯茗杨, 等. 基于DTW的地磁指纹定位方法[J]. 传感技术学报, 2020, 33(01):84-90..

[60] 王安义, 欧雪. 基于行人航位推算的后向地磁匹配算法研究[J]. 科学技术与工程. 2021, 21(5): 1707-1712.