Helicopter Dynamic Model Identification by Conditional Attitude Hold Logic

Abstract

A parameter identification method was proposed for flight vehicles with dynamic models that are difficult to identify due to their highly coupled characteristics. The existing off-axis noise input method, which is used for reducing the correlation between the primary and secondary inputs, has a difficult-to-obtain convergence condition under the hold trim state. In addition, the existing method cannot easily identify high frequency dynamics due to the off-axis input noise. On the other hand, the proposed identification method relied on a computer-generated sweep in the frequency-domain, which could produce a more accurate model than the existing perturbation model. Moreover, the proposed conditional attitude hold logic described the actual pilot's motion by activating the controller when the Euler angle measurements exceeded a pre-defined angle. The proposed method was used with the Bo-105 helicopter nonlinear simulation model, and the identified coupled rotor/fuselage nine-degree-of-freedom results were reliable and comparable to those obtained with the existing perturbation model.