View in EDS

Adaptive intelligent backstepping for real-time quadrotor trajectory tracking using feedback error learning approach.

Saved in:
Bibliographic Details
Title: Adaptive intelligent backstepping for real-time quadrotor trajectory tracking using feedback error learning approach.
Authors: Baghelani, Esfandiar1 (AUTHOR) esfandiar.baghelani@email.kntu.ac.ir, Roshanian, Jafar1 (AUTHOR) roshanian@kntu.ac.ir, Teshnehlab, Mohammad2 (AUTHOR) teshnehlab@eetd.kntu.ac.ir, Georgiev, Krasin3 (AUTHOR) krasin@tu-sofia.bg
Source: AIP Conference Proceedings. 2025, Vol. 3339 Issue 1, p1-12. 12p.
Subject Terms: *BACKSTEPPING control method, *ROBOTIC trajectory control, *QUADROTOR helicopters, *ADAPTIVE control systems, *AERODYNAMIC stability, *NONLINEAR systems
Abstract: Quadrotors are underactuated and maneuverable flying robots. Their high maneuverability is due to their low stability. However, this instability makes it challenging to control these nonlinear underactuated systems, especially in the presence of uncertainty and disturbances. In the context of trajectory tracking by quadrotors, Lyapunov-based controllers, such as the backstepping method, guarantee flight stability. Furthermore, enhancing flight control with a learning-based approach, like Feedback Error Learning (FEL), provides an overall adaptive robust controller. This paper combines the backstepping method for robust control of flight altitude and adaptive control of angular attitude with the FEL method. We demonstrate the improvement of control performance in trajectory tracking by a quadrotor in the presence of uncertainty. Significant reductions in Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE) are achieved, indicating enhanced tracking performance. This method proves particularly effective in applications where precise trajectory tracking is more critical than control effort or energy consumption, such as navigating obstacles or collision avoidance. [ABSTRACT FROM AUTHOR]
Database: Academic Search Index
Description
Abstract:Quadrotors are underactuated and maneuverable flying robots. Their high maneuverability is due to their low stability. However, this instability makes it challenging to control these nonlinear underactuated systems, especially in the presence of uncertainty and disturbances. In the context of trajectory tracking by quadrotors, Lyapunov-based controllers, such as the backstepping method, guarantee flight stability. Furthermore, enhancing flight control with a learning-based approach, like Feedback Error Learning (FEL), provides an overall adaptive robust controller. This paper combines the backstepping method for robust control of flight altitude and adaptive control of angular attitude with the FEL method. We demonstrate the improvement of control performance in trajectory tracking by a quadrotor in the presence of uncertainty. Significant reductions in Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE) are achieved, indicating enhanced tracking performance. This method proves particularly effective in applications where precise trajectory tracking is more critical than control effort or energy consumption, such as navigating obstacles or collision avoidance. [ABSTRACT FROM AUTHOR]
ISSN:0094243X
DOI:10.1063/5.0297639