Generic UAV Autopilot Prototype Based on Adaptive Modified Incremental Backstepping
Ehab Safwat, Ahmed Mohsen Kamel · AIAA Scitech 2021 Forum · 2021
View Video Presentation: https://doi.org/10.2514/6.2021-0373.vid Unmanned Aerial Vehicles (UAVs) have demonstrated their effectiveness for a huge number of civilian and military applications. However, the widely used commercial autopilots and flight controllers have to be tuned for each particular airframe. This problem is investigated in this research which focuses on the design of a generic flight control system based on the Adaptive Modified Incremental Backstepping (AMIBKS) technique. In a lot of researches, the Modified Incremental Backstepping (MIBKS) controller is used to stabilize all state variables simultaneously with partial dependency on the model parameters. MIBKS relies on estimates of the UAV angular accelerations and instantaneous control deflections measurements. However, aerodynamic coefficients include uncertainty and can lead to unexpected behavior of the flight controller in some circumstances. For this reason, an adaptive controller is designed to compensate for the effect of the aerodynamic modeling error. To design a standalone UAV flight control system, a real-time flight simulation system of Hardware-in-Loop (HIL) is constructed to facilitate the development, validation, and verification steps of flight control algorithms. HIL is implemented by using the XPC target computer to run the UAV model under test. A 32-bit microcontroller is pre-programmed with the desired reference and is used to command the UAV control surfaces. Results from HIL and numerical simulations are presented and compared to validate the proposed generic controller adequacy and evaluate its usefulness as a rapid autopilot prototyping tool. It is shown that the AMIBKS controller is a model-independent flight controller with complete dependency on the accuracy of the used UAV onboard sensors.