Finite-Time Prescribed Performance Optimal Attitude Control for Quadrotor UAV
Jiahong Xu,Lijie Wang,Yang Liu,Hong Xue
DOI: https://doi.org/10.1016/j.apm.2023.03.030
IF: 5.336
2023-04-11
Applied Mathematical Modelling
Abstract:In this paper, the optimal attitude control problem is investigated for quadrotor UAV with time-varying control effectiveness faults based on adaptive dynamic programming. Initially, a novel nonlinear conversion function is introduced to achieve the finite-time prescribed performance on tracking errors of three attitude angles of the quadrotor UAV. Furthermore, given that the controller may experience a degree of time-varying faults due to internal component aging or external environmental influences, an improved value function with a discount factor is proposed to resist the aforementioned time-varying control effectiveness faults. Then, by integrating the concurrent learning technique into adaptive dynamic programming algorithm, a valid approximate optimal control protocol is presented. Such an algorithm not only effectively improves the utilization of historical data in the history stack but also loosens the strict persistent excitation condition into a rank condition that is easy to check online. Moreover, it can be proved that the uniformly ultimately bounded stability of the attitude system is ensured even in the presence of time-varying control effectiveness faults. Meanwhile, tracking errors of attitude angles successfully converge to pre-defined symmetric or asymmetric precision intervals with an adjustable convergence rate in a user-defined finite time. Finally, a simulation example of the quadrotor UAV attitude system is given to verify the validity of the developed control scheme.
mechanics,engineering, multidisciplinary,mathematics, interdisciplinary applications