Online Parameter Estimation for Fixed-Wing UAV based on DREM Method and Adaptive Control

The online identification of aerodynamic coefficients for fixed-wing aircraft is crucial for designing flightcontrol laws and diagnosing faults; however; this issue has not yet been sufficiently addressed. To this end, this paper presents a parameter-estimation algorithm for fixed-wing aircraft based on an improved dynamic regressor extension and mixing (DREM) method. This algorithm can accurately and efficiently determine the aerodynamic coefficients under conventional maneuvering operations that do not meet the persistent-excitation condition. Taking into account the presence of external disturbances, adaptive backstepping control laws and disturbance observers (DOs) are incorporated based on the outcomes of online parameter identification. This approach seeks to achieve precise reference tracking and effective estimation and suppression of disturbances. Simultaneously, the integration of the DO and DREM estimators synergistically enhances their impact, leading to further refinement. The stability of the system is rigorously ensured throughout the design process. Finally, two comparative simulations and a hardware-in-the-loop experiment were conducted using a small fixed-wing UAV model to validate the efficacy and real-time performance of the proposed algorithm.