Real-time System Identification of Aircraft in Upset Condition Using Adaptive-order Zonotopic Kalman Filter

Abstract

It is essential to prevent LoC(Loss-of-Control) or upset situations caused by stall, icing or sensor malfunction in aircraft, because it may lead to the crash of the aircraft. With this regard, it is crucial to correctly identify the dynamic characteristics of aircraft in such upset conditions. In this paper, we present a SID(System IDentification) method utilizing the moving-window based least-square and the adaptive-order ZKF(Zonotopic Kalman Filter), which is more effective than the existing Kalman-filter based SID for the aircraft in upset condition at a high angle of attack with temporary sensor malfunction. The proposed method is then tested on real flight data and compared with the existing one.