This paper presents a model-based controller consisting of a feedback linearization scheme and a state-dependent proportional derivative (PD) controller adapted to a parallel flight simulator Stewart mechanism. This parallel robot is considered to emulate motions of highly maneuverable aircrafts, which require well-trained pilots. The simulations are based upon a reduced-model prototype built in order to verify kinematic design aspects and control laws. Indeterminacies in the mass distribution of the system will generally affect model-based controllers, necessitating compensation or the employment of robust control methods. Through introducing the pilot's sensorial feedback of acceleration, the pilot's behavior in giving commands is emulated via an optimization process, which tunes the controller coefficients accordingly. Stability of the designed control system is guaranteed via the Lyapunov approach. To further explore the system through perilous flight scenarios, three pre-designed maneuvers are selected as test cases. It is expected that closed-loop control tasks in which a pilot tracks a target, while at the same time the controller rejects disturbances and adapts itself to the pilot's progressive skills, are ameliorated through this arrangement. Numerical results show that the proposed method is found robust in the training process in conditions of parameters indeterminacy.