Turkish Journal of Electrical Engineering and Computer Sciences




In flight control systems, the actuators need to tolerate aerodynamic torques and continue their operations without interruption. To this end, using the simulators to test the actuators in conditions close to the real flight is efficient. On the other hand, achieving the guaranteed performance encounters some challenges and practical limitations such as unknown dynamics, external disturbances, and state constraints in reality. Thus, this article attempts to present a robust adaptive neural network learning controller equipped with a disturbance observer for passive torque simulators (PTS) with load torque constraints. The radial basis function networks (RBFNs) are employed to identify the unknown terms, providing information for the disturbance observer. Besides, the tuning parameters are chosen optimally by adopting the grey wolf optimization (GWO) algorithm. The closed-loop system stability is also proven by the barrier Lyapunov function (BLF) while the total uncertainties, including system dynamics, friction, and disturbance, are tracked by the total estimation. Thus, the predetermined performance, robust behavior, and high-precision estimation are the achievements of the presented controller for PTS. To confirm the ability of the proposed control idea, simulations are provided. Furthermore, a comparison scenario is also considered to emphasize the supremacy of the proposed control system.


Neural network, passive torque simulator, state constraint, disturbance observer, grey wolf optimization algorithm, robust control.

First Page


Last Page