Content area
This thesis presents a comprehensive control strategy for fixed-wing Unmanned Aerial Vehicles (UAVs) engaged in ground target tracking, addressing the challenges of aerodynamic constraints, unpredictable target movements, and environmental disturbances. It starts by developing a Nonlinear Model Predictive Controller (NMPC) formulated as a non-convex optimization problem, based on a nonlinear three-dimensional target tracking system model. Stability conditions for the nonlinear closed-loop system are established by analyzing a linear controller within a specified terminal region, enabling the use of convex Model Predictive Control techniques within the NMPC framework. This ensures that the UAV's controlled trajectory reaches the terminal region within a fixed prediction horizon, allowing effective tracking of the ground target.
To enhance robustness against disturbances and unknown target trajectories, the control strategy is extended by modeling the target's movement as a first-order dynamic system and deriving a comprehensive three-dimensional Dubins model that accounts for both target movement and exogenous disturbances. A bilinear time-varying disturbance model is introduced, and a Kalman filter-based estimation strategy is employed to estimate both disturbances and the target's movement. This information is integrated into the NMPC to achieve highly accurate tracking.
Further improving performance without relying on explicit disturbance models, with the integration of a phase-based Iterative Learning Control (ILC) with the NMPC. The ILC leverages data collected during each iteration of the tracking task, adjusting control actions based on the phase angle of the UAV's orbit around the target. This hybrid control scheme enhances disturbance rejection and tracking accuracy by iteratively refining control actions, without the need for precise disturbance modeling or observer-based estimation.
Extensive simulations and experimental results validate the effectiveness of the proposed control strategy. The results demonstrate that the UAV successfully tracks moving ground targets with unknown trajectories, maintaining high accuracy even in the presence of significant uncertainties and disturbances. This work contributes a robust and adaptable control framework for fixed-wing UAV ground target tracking, suitable for dynamic environments where traditional methods may be insufficient.