Real-time simulation and control systems design by the Response Surface Methodology and designed experiments

Abstract

This paper examines two cases where the fitting of a model to experimental data makes possible the solution of extremely difficult design and simulation problems. In the first (aerospace) case, designed experiments were conducted on a permanent magnet AC motor which provided the motive power for a flight surface actuator in a more electric aircraft application. The Response Surface Methodology is applied to the measured data to achieve inclusion of the component in a real-time distributed aircraft simulation. In the second (automotive) case, oscillatory acceleration responses are controlled via an electronically actuated (drive by wire) throttle. Designed experiments were conducted on the test vehicle to achieve a systematic excitation of the vehicle driveline. An approximation to the measured data is achieved by the Response Surface Methodology allowing a controller to be designed extremely rapidly.