Algorithm for identification of helicopter open-loop transfer functions and reduced-order modeling

Abstract

This paper deals with the development and application of an algorithm aimed at the identification of openloop transfer functions from closed-loop data. Particularly suited for helicopter applications, it allows the identification of the machine transfer functions even in the presence of a controller. The identification algorithm is based on the knowledge of the time histories of arbitrary external inputs and corresponding control actuation and responses. It is successfully applied to the AW-09 helicopter for the identification of the transfer functions relating pilot inputs to vehicle attitudes and kinematics. Then, from their rational approximation the time domain representation of the helicopter dynamics is achieved. The capability of the final state-space helicopter model to simulate vehicle dynamics is assessed by comparing the provided helicopter time responses to arbitrary inputs, with those obtained through the high-fidelity nonlinear solver formerly used to obtain the database for the identification of the transfer functions