本文为加拿大卡尔加里大学（作者：Mohamed Ali Attia）的博士论文，共187页。

导航系统在许多重要学科中发挥着重要作用。确定用户相对于物理环境（如道路、交叉口和服务）的位置是交通服务（如车载导航、车队管理和基础设施维护）的重要组成部分。此外，还需要其他导航服务来定位用户在室内物理环境中的位置（例如机场、商场、公共建筑、大学校园）。这种基于室内的导航可以在用户导航、增强911（E911）、执法、基于位置和营销服务等多种应用中提供帮助。无论是室内还是室外导航应用，都需要一个可靠、连续的导航解决方案，以克服全球导航卫星系统（GNSS）信号不可用的挑战。为了弥补这个问题，GNSS现在通常与其他导航系统（如惯性导航系统INS）一起使用。这种双系统集成方法为解决GNSS信号中断问题提供了一种解决方案。然而，随着时间的推移，惯性导航系统会产生较大的漂移，特别是与低成本的商业传感器共同使用时。漂移对惯导系统精度的影响很大，因此需要额外的绝对辅助传感器，以降低这种漂移的影响。

本文针对GNSS信号中断的环境，提出了一种地图辅助导航方案。地图已经成为可视化用户日常旅行导航轨迹的主要媒介。本研究探讨并开发了一个辅助系统，利用地理空间数据模型，而不仅仅是一种视觉的方式。它通过为导航轨迹提供虚拟边界来丰富导航解决方案，并且仅当用户在地图上的位置符合逻辑时才限制其可能性。随后开发的算法为不同的导航解决方案集成了多个导航传感器。除了各种地图匹配算法外，还使用了室内和室外环境（例如城市、峡谷）的若干地理空间模型来匹配和投影地理空间地图上的导航位置估计，并用作导航滤波器的附加反馈。所开发的算法在室内和室外环境中进行了现场测试，得到了准确的匹配结果，并显著提高了定位精度。结果表明，所开发的地图辅助系统提高了导航轨迹在GNSS中断环境下的可靠性、可用性和准确性。

Navigation systems play an important rolein many vital disciplines. Determining the location of a user relative to thephysical environment (e.g. roadway, intersections, and services) is animportant part of transportation services such as in-vehicle navigation, fleetmanagement and infrastructure maintenance. In addition, other navigationservices are required for locating the position of a user in an indoor physicalenvironment (e.g. airports, shopping malls, public buildings, universitycampus). This indoor-based navigation can assist in several applications suchas user navigation, enhanced 911 (E911), law enforcement, location-based andmarketing services. Both indoor and outdoor navigation applications require areliable, trustful and continuous navigation solution that overcomes thechallenge of Global Navigation Satellite System (GNSS) signal unavailability.To compensate for this issue, GNSS is now commonly used in tandem with othernavigation systems such as Inertial Navigation System (INS). This dual-systemintegration method provides a solution to GNSS signal outages. However, overtime there is a significant amount of drift, characteristic of INS butespecially common with low-cost commercial sensors. The effects of drift on INSaccuracy highlight the need for additional absolute aiding sensors that cansurvive for longer periods of time. In this thesis, a map aided navigationsolution is developed for GNSS-denied environments. Maps have been the primarymedium to visualize the navigation trajectories of a user’s everyday travels.This research investigates and develops an aiding system that utilizesgeospatial data models in more than just a visual way. It assists thenavigation solution by providing virtual boundaries for the navigationtrajectories and limits its possibilities only when it is logical to locate theuser on a map. The algorithms subsequently developed integrate severalnavigation sensors for different navigation solutions. Several geospatialmodels for both indoor and outdoor environments (e.g. urban canyons) inaddition to various map matching algorithms were used to match and projectnavigation position estimates on the geospatial map and used as an additionalfeedback for the navigation filter. The developed algorithms were field testedin several indoor and outdoor environments and yielded accurate matchingresults as well as a significant enhancement to positional accuracy. Theachieved results demonstrate that the contribution of the developed map aidedsystem enhances the reliability, usability, and accuracy of navigationtrajectories in GNSS-denied environments.

1. 引言

2. 导航应用的地理空间数据模型

3. 地图匹配算法

4. 移动式多传感器组合导航系统

5. 地图辅助的导航系统

6. 现场测试、结果、讨论与分析

7. 总结、结论与建议

附录A 参考坐标系介绍

附录B 测试中传感器主要指标

更多精彩文章请关注公众号：