This paper proposes an approximate optimal control solution for reconfigurable unmanned system in the presence of disturbance, structured around fuzzy zero-sum game. The system model, which incorporates external perturbation, is constructed via the Newton-Euler iterative method. By framing the disturbance as a competitive player, the original optimal control problem is transformed into a zero-sum game, optimized through a self-learning process inherent to the employed fuzzy logic system. To derive the approximate optimal control policy, a critic fuzzy network is implemented to approximate the associated cost function. Subsequent stability analysis, grounded in Lyapunov theory, proves that the system's tracking error achieves uniform ultimate boundedness. The practical applicability and robustness of the approach are confirmed by experimental data.
Ma et al. (Wed,) studied this question.