Flight-Test System Identification Methodology and Hover Results for a Vectored-Thrust eVTOL Aircraft

This paper describes an ongoing aircraft system identification effort for an industry prototype electric vertical takeoff and landing (eVTOL) vehicle. Building on previous eVTOL aircraft system identification developments in wind-tunnel testing and flight simulations, an approach to modeling from flight-test data is formulated for the AIBOT 500 aircraft. The full system identification process is presented, including the experiment design, flight data collection, and model identification steps. Orthogonal phase-optimized multisine programmed test inputs are integrated into the flight control system and are applied to each control surface and propulsor simultaneously to efficiently collect informative flight data for model identification. Initial modeling results are given in hover, where an aero-propulsive model is identified using the equation-error method in the frequency domain. The presented results demonstrate the utility of the modeling approach and are compared to FLIGHTLAB predictions executed using a dynamic wake model made prior to conducting flight testing. Practical techniques and recommended improvements are discussed to inform future flight-test system identification efforts for eVTOL aircraft.