No abstract

In this study the design of motion‐based flight simulators is carried out by specifying the performance required of the motion cueing mechanism, to generate translational and angular motions as a 6–3 Stewart Platform Mechanism (SPM). These motions are intended to approximate the specific forces and angular accelerations encountered by the pilot in the simulated aircraft. Firstly, the dynamics of this 6–3 SPM is given in closed form as in our earlier study. Then, for the control of obtained dynamic model, a leg‐length based PD algorithm is applied. In the optimization of the applied PD algorithm's coefficients, Real Coded Genetic Algorithms are used. So as to have faster and effective system's performance, the fitness function chosen, in Genetic Algorithms, having maximum overshoot value, settling time and steady state error which are obtained from the unit step response. The performance of the system studied is compared to the similar studies in the literature exist.

A flight simulator must be designed to generate the correct acceleration cues, attitudes and vibrations to the flight compartment to provide an extra degree of realism for the pilots. Therefore, such a system, which has six degrees of freedom (dof), should enable to produce pitch, roll, yaw, heave, forward and lateral movement simultaneously. However, such a complex dynamic system can be modeled as a Stewart platform with pneumatic actuators, having six dof. During the simulations, on the position and orientation of moving platform, motion determination parameters and their absolute, relative errors and standard deviations are outlined. Simulation results, which are obtained, when closely examined reveal that the developed motion determination algorithm for the considered parallel dynamic mechanism is highly accurate. Additionally, a technique is introduced for the motion determination and its deviation from the given task.