Development of U-model enhanced nonlinear dynamic control systems —Framework, algorithms and validation

Li, Ruobing

Abstract

This study aims to develop the classical model-based U-control design framework to enhance its robustness and reduce its dependence on model accuracy. By absorbing the design concepts of other advanced control algorithms, firstly, based on the discrete-time U-control algorithm, a continuous-time (CT) U-model based dynamic inversion algorithm is proposed. Then the CT U-control system design procedures are presented and explained step by step with numerical and simulation demonstrations of the linear and nonlinear U-control system design examples. Secondly, the U-control algorithm develops two mainstream nonlinear robust control algorithms, disturbances suppression and disturbances compensation, while maintaining its system dynamic cancellation characteristics, including two-degree-of-freedom U-model-based internal model control (UTDF-IMC), Disturbance observer-based U-control (DOBUC), sliding mode enhanced U-control (U-SMC) and U-model based double sliding mode control (UDSMC) algorithms. At the same time this study first developed and applied the U-control method to a practical industry application: robust quadrotor trajectory tracking control. The proposed UDSMC method and multiple-input and multiple-output extended-state-observer (MIMO-ESO) established the quadrotor flight control system. The difficulties associated with quadrotor velocity measurement disturbances and uncertain aerodynamics are successfully addressed in this control design. A rigorous theoretical analysis has been carried out to determine whether the proposed control system can achieve stable trajectory tracking performance, and a comparative real-time experimental study has also been carried out to verify the better effectiveness of the proposed control system than the classical SMC and built-in PID control system. This study is clearly novel as the methods and experiments it proposed have not been researched before.

• Thesis Type: Thesis
• Deposit Date: Apr 25, 2023
• Publicly Available Date: Sep 26, 2023
• Public URL: https://uwe-repository.worktribe.com/output/10708609
• Award Date: Sep 26, 2023