Efficient Frequency Response Identification for Small Fixed-Wing UAS Using Closed-Loop Flight Data

Abstract

In this paper, a method to identify bare-airframe, closed-loop, and broken-loop frequency responses from a single set of closed-loop flight data is presented. This approach is particularly useful for small unmanned aircraft, which face unique challenges during system identification flight tests, such as flight safety, open-loop stability, considerable pilot workload during frequency sweep maneuvers, and non-trivial disturbances from atmospheric turbulence. Closed-loop flight testing can address many of these challenges as it improves the stability, repeatability, and flexibility of flight testing, which can lead to higher data quality. The proposed method is demonstrated using simulated data of a small, flying-wing UAS operating in closed-loop in strong turbulence. Frequency responses are identified using a joint input-output method, which prevents bias errors caused by noise feedback. The bare-airframe, closed-loop, and broken-loop frequency responses are identified from one data set with excellent agreement with the simulation truth over a wide frequency range. The example demonstrates the effectiveness of this approach to improve results and reduce flight testing time for small UAS system identification.