Stochastic versus deterministic approach to the design of the control law of moving flight simulators

摘要

In this correspondence, the optimal control of the motion of moving flight simulators is presented as an optimal tracking problem, and two approaches are compared: a stochastic approach that makes use of the past of the airplane trajectory, and a deterministic approach that makes use of both the past and the future of the airplane trajectory. The tracking error is, in both cases, the discrepancy between the motion cues measured by the vestibular sensors in real flight and in simulated flight. It is shown that the deterministic approach leads to a significantly better performance than the stochastic approach, with respect to the minimization of the root mean square (RMS) tracking error, while the improvement with respect to the decrease of the maximum tracking error is moderate. Of course, there always remains the task of verifying the quality of the design using pilot ratings.