Robust Controller Design For A Fixed-Wing Uav Using Active Disturbance Cancellation

Autonomous aerial vehicles have become integral components of both military and civilian applications, playing crucial roles in tasks such as reconnaissance, surveillance, and inspection. Given their widespread use, extensive research contributions aim to further enhance their performance. Among the key areas of investigation for autonomous aerial vehicles, research into control system architectures used in the autopilots stands out as particularly significant. The main aim of ongoing research is to find control architectures that are not only robust but also fulfill other criteria such as being easy to implement and cost-effective, making them practical for real-world applications. This thesis focuses on the robust controller design of a fixed-wing UAV through the application of active disturbance cancellation methods in addressing the longitudinal instability of a fixed-wing UAV’s aerodynamics. Pitch stabilization is achieved through the utilization of a gain-scheduled controller based on Proportional-Integral-Derivative (PID) control. To evaluate the robustness of the gain-scheduled controller, simulations are conducted across three distinct flight stages representing different trim points of the UAV: steady-level flight at sea level, pull-up flight, and steady-level flight at a certain altitude. These simulations are carried out separately, introducing fixed disturbance and disturbance produced by the Dryden wind model. Following this, a disturbance observer is integrated into the gain-scheduled controller to further enhance its robustness. Subsequently, a Conditional Integral Sliding Mode Controller (C-ISMC) is designed and tested for both nominal and fixed disturbance scenarios, with a comparative analysis against the gain-scheduled controller. This was followed by the integration of a disturbance observer into the C-ISMC to evaluate its robustness against wind disturbance. Results obtained from the C-ISMC with a disturbance observer exhibits higher robustness than the disturbance observer based gain-scheduled controller.