Ranging-based adaptive navigation for autonomous micro aerial vehicles

• Main Author: Nguyen, Pham Nhat Thien Minh
• Other Authors: Xie Lihua
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering::Control and instrumentation; Engineering::Electrical and electronic engineering::Computer hardware, software and systems
• Online Access: https://hdl.handle.net/10356/137097
• Institution: Nanyang Technological University

The last decade has witnessed a surge in popularity of Micro Aerial Vehicles (MAVs) in many civilian, industrial and military applications. It can be seen that most research interests on MAV revolve around two main problems, namely localization and navigation, or estimation and control in a more general sense. Most commonly we find that these two problems are addressed in a separate manner, whereas localization capability is the basis upon which different navigation strategies are developed. While this approach may facilitate convenient solutions and analysis, it also brings about several compromises, such as low adaptability to complex/cluttered environments when localization relies on an external positioning system, or estimation drift and limited cooperative capability for MAVs relying on onboard visual odometry (VO) systems. Motivated by these issues, this thesis is dedicated to studying infrastructure-free navigation schemes that can achieve a high level of flexibility, portability and practicality for autonomous operations of multi-MAV systems in GPS-denied environments. As will be shown later, rangingbased integrated estimation-control technique, i.e. adaptive navigation, is the key in our approach towards this objective. As a first expedition into this direction, a sensor fusion scheme is proposed to achieve relative positioning and tracking of a target by MAV, featuring the use of multiple Ultra-wideband (UWB) ranging sensors strategically installed on both the MAV and the target. An estimator based on Extended Kalman Filter (EKF) is developed to fuse UWB ranging measurements with data from onboard sensors such as inertial measurement unit (IMU), altimeters and optical flow. In addition, UWB-based communication capability is utilized to transfer the target’s onboard information to the quadcopter. Experiment results demonstrate the ability of the MAV to robustly control its position relative to a moving target even with uncertain velocity. While the aforementioned method of ranging to multiple UWB nodes on the target can achieve effective relative positioning and tracking at a close range, to maximize the operation range and flexibility, it would be more convenient to rely on a single beacon ranging scheme. Thus, we formulate a distance-based adaptive navigation problem where the MAV is required to approach a landmark at an arbitrary unknown location. To solve this problem, we propose an integrated estimation-control scheme to simultaneously accomplish two objectives: relative localization using only distance and odometry measurements, and navigation to a desired location under a nonlinear distance-based bounded control law. Asymptotic convergence is obtained by invoking the discrete-time LaSalle’s invariance principle in the noise-free case, and the stability under distance measurement noise is also investigated. Multiple numerical simulations and real-world experiments on quadcopter MAV in GPS-denied environments are carried out to validate theoretical findings and the efficacy of the proposed estimation-control scheme. In the next phase, a ranging-based autonomous docking operation in GPS-denied environment is conceived with the insights from previous investigations. Specifically, a method combining sequential ranging of UWB sensor with vision-based techniques is developed to achieve both autonomous approaching and landing capabilities in GPS-denied environments. In the approaching phase, a recursive leastsquares optimization algorithm is proposed to estimate the position of the MAV relative to the target by using the distance and relative displacement measurements. Using this estimate, MAV is able to efficiently approach the target until the landing pad is detected by an onboard vision system, then UWB measurements and vision-derived poses are fused with other onboard sensor information to facilitate an accurate landing maneuver. Real-world experiments are conducted to demonstrate the efficiency of the proposed method. Based on the insights from the aforementioned single-MAV localization-navigation schemes, we set out to develop new cooperative operations of multi-MAV systems. On this direction, we investigate the problem of leader-following control of multiple MAVs, which is then extended to distributed adaptive control of dynamic formation of multi-MAV system, supported by relative position estimate derived from distance and self-displacement measurements. The main challenge of the problem, which is to simultaneously fulfill both relative localization and control tasks, is resolved by different approaches to guaranteeing the persistent excitation (PE) condition, i.e the introduction of specialized agents, or by embedding the distance-based relative localization technique into a time-varying formation. By assuming that the leader is globally reachable and selecting proper parameters for the estimation and control laws, it is shown that the integrated estimation-control schemes ensure exponential convergence (EC) of relative localization error, which leads to EC of formation error when the leader’s behavior is deterministic, and bounded formation error for a nondeterministic leader. Extensive numerical simulations and real-world implementations are carried out to verify the theoretical results and demonstrate the efficacy and effectiveness of the proposed method.