Advanced Navigation Solutions for UAV Autonomous Mobility

该研究计划的目标是通过开发和测试最先进的导航技术来提高无人机（UAV）的自主移动能力。无人机是现代运输的标志，因为它们允许在不需要人为干预的情况下完成各种任务。因此，无人机正在成为各种应用任务的组成部分，包括食品和邮件递送、监视、进入危险或偏远地区、消防、基础设施检查、在偏远地区提供医疗援助、精准农业和野生动物监测。在过去的二十年里，研究界一直在积极努力，成功地解决与无人机自主机动性相关的挑战。然而，一些挑战依然存在。无人机遇到的一些具有挑战性的开放问题是：1）需要鲁棒的导航算法（即，方向、位置和线速度估计）; 2）装备有低成本传感单元的欠驱动UAV的控制;以及3）需要适合于具有不确定动态的UAV在不确定环境中行进的解决方案。廉价和紧凑的无人机通常使用机载低成本传感单元，提供不完善的数据。此外，在没有GPS信号的情况下，不能直接获得车辆的导航组件。因此，导航分量必须经由计算上简单的算法来估计，该算法对于不确定性是鲁棒的（即，偏差和噪声），例如，由包括照相机和惯性测量单元（IMU）的低成本视觉辅助惯性导航系统（VA-INS）提供的测量中所固有的。在缺乏真实导航信息的情况下，并考虑到必须使用从不完美的测量估计的数据，稳定和鲁棒控制将被证明是具有挑战性的。此外，有些任务对于单个无人机是不可行的，但可以由一组合作的无人机来完成。多无人机任务需要更复杂的导航解决方案。因此，该研究计划的总体目标是解决上述与单无人机导航和控制以及面对测量不确定性的多无人机导航相关的挑战。该研究计划将为高素质的人才提供丰富的培训环境，包括在导航，现代控制和航空航天科学方面发展强大的理论，分析和验证技能的机会。此外，拟议的计划将提供先进的自主移动解决方案，以解决多个持续存在的挑战，并为无人机应用开辟新的机遇。我们希望为加拿大的研发做出重大贡献，推进导航，现代控制，机器人和航空航天工程领域的知识。

The objective of this research program is to advance the abilities of Unmanned Aerial Vehicle (UAV) autonomous mobility through the development and testing of state-of-the-art navigation techniques. UAVs are a hallmark of modern transportation as they allow various missions to be accomplished without the need for human intervention. As a result, UAVs are becoming an integral part of a variety of application tasks, including food and mail delivery, surveillance, accessing dangerous or remote locations, firefighting, infrastructure inspection, provision of medical aid in remote areas, precision agriculture, and wildlife monitoring. Over the last two decades, the research community has been actively working on successfully addressing the challenges associated with autonomous mobility of UAVs. However, a number of challenges persist. Some of the challenging open problems encountered by UAVs are 1) the need for robust navigation algorithms (i.e., orientation, position, and linear velocity estimation) for applications involving Global Positioning System (GPS)-denied regions; 2) the control of under-actuated UAVs equipped with low-cost sensing units; and 3) the need for solutions suitable for UAVs with uncertain dynamics travelling through uncertain environments. Inexpensive and compact UAVs often use onboard low-cost sensing units which provide imperfect data. Moreover, in the absence of GPS signal, the navigation components of a vehicle cannot be obtained directly. Hence, the navigation components must be estimated via a computationally simple algorithm that is robust against uncertainties (i.e., bias and noise) inherent in measurements supplied, for instance, by a low-cost Vision-Aided Inertial Navigation System (VA-INS) composed of a camera and an Inertial Measurement Unit (IMU). In the absence of true navigation information and given the necessity of using data estimated from imperfect measurements, stable and robust control will prove to be challenging. In addition, some tasks are unfeasible for a single UAV, but can be accomplished by a team of collaborating UAVs. Multi-UAV missions require more complex navigation solutions. Hence, the overarching goal of this research program is to tackle the aforementioned challenges related to single-UAV navigation and control as well as multi-UAV navigation in the face of measurement uncertainties. This research program will provide a rich training environment for highly qualified personnel including opportunities to develop strong theoretical, analytical, and validation skills in navigation, modern control, and aerospace sciences. Moreover, the proposed program will contribute advanced autonomous mobility solutions that will address multiple persisting challenges and open new opportunities for UAV applications. We expect to make significant contributions to Canadian R&D, advancing knowledge in the areas of navigation, modern control, robotics, and aerospace engineering.