High availability and robustness precise GNSS positioning systems for autonomous vehicle navigation in challenging environments

自动驾驶汽车等自动驾驶汽车（AV）的开发取得了重大进展，预计将对工业、社会和我们的生活产生重大影响。目前，所有主要的汽车制造商都开发了自动驾驶汽车，并在街道上进行了测试。一款完全自动驾驶的汽车，还有许多公路法规尚待制定，最早可能在2020年上市。 AV使用各种传感器，控制硬件和软件来管理驾驶功能的各个方面，并需要非常精确的位置和方向信息，以便在没有人类输入的情况下控制和导航车辆。GNSS是自动车辆传感器套件中的此类传感器之一。虽然汽车制造商已将GNSS作为不可或缺的组件，但目前的精密GNSS技术在满足AV的严格要求方面面临着一些障碍。虽然目前正在测试的自动驾驶汽车在很大程度上依赖于天气相关的传感器，这些传感器在恶劣的天气和道路条件下表现不佳（例如，车道标记和道路标志被雨，雪或雾等覆盖或模糊），他们的成功仅限于良好的天气和道路条件。作为一种不受天气影响的传感器，也是唯一一种输出对其他机载传感器至关重要的绝对时间信息的传感器，如果能够证明其在具有挑战性的环境中的可用性和可靠性，精密GNSS可以在恶劣天气和道路条件下发挥关键作用。 本研究的主要目标是研究和开发高可用性和鲁棒性的精确GNSS定位方法和算法，以满足在具有挑战性的环境中自主车辆导航的需要。这项拟议研究的成果将包括新的流程、方法和算法，以实现高可用性和鲁棒性的精确GNSS定位，并帮助回答诸如“我可以在需要时获得精确的GNSS定位吗？"，“在GNSS校正数据丢失期间，我是否仍能获得精确的GNSS定位？"，“我可以信任GNSS的定位解决方案吗？"。还将开发一个原型系统，用于性能分析和工业示范。这些新的过程、方法和算法有望推进精确的GNSS技术，并增强AV机载其他导航传感器的性能，用于障碍物检测和环境感知，特别是在坚韧的天气条件下。

Significant progress has been made in the development of Autonomous Vehicles (AV) such as self-driving cars which are expected to greatly impact the industry, society and our lives. Currently all major automakers have developed a self-driving car and tested it on streets. A fully autonomous vehicle, with many highway regulations still to be developed, could be available to market as early as 2020. AV use various sensors, control hardware and software to manage all aspects of the driving function and require very precise position and orientation information in order to control and navigate the vehicle without input from humans. GNSS is one of such sensors in the automated vehicle sensor suite. Although automakers have adopted GNSS as an indispensable component, current precision GNSS techniques face several obstacles to meet AV's stringent requirements. While self-driving cars currently being tested rely much on weather-dependent sensors which perform poorly in adverse weather and road conditions (e.g. lane markers and road signs are covered or obscured by rain, snow, or fog etc.), their success is limited to only good weather and road conditions. As a weather-independent sensor and also the only sensor that outputs absolute time information critical to other onboard sensors, precision GNSS could play a key role in bad weather and road conditions if it can prove its availability and reliability in challenging environments. The major objective of this proposed research is to investigate and develop high availability and robustness precise GNSS positioning methods and algorithms to meet the need of autonomous vehicle navigation in challenging environments. The outcome of this proposed research will consist of new processes, methods and algorithms to enable high availability and robustness precise GNSS positioning and help answer questions such as “can I have precise GNSS positioning whenever I need?”, “can I still have precise GNSS positioning available during loss of GNSS correction data?”, “can I trust the positioning solutions from GNSS?”. A prototype system will also be developed for performance analysis and industrial demonstration. The new processes, methods and algorithms are expected to advance precision GNSS technologies and also enhance the performance of other navigation sensors onboard AV for obstacle detection and environment sensing, particularly in tough weather conditions.