Robust flight control based on a nonlinear-L1 adaptive control with modified piecewise-constant and HIL experiments

作     者：Li, Yu He, Qizhi Liu, Xiaoxiong Zhang, Weiguo Jing, Xingjian 

作者机构：City Univ Hong Kong Dept Mech Engn Hong Kong Peoples R China Northwestern Polytech Univ Coll Automat Xian 710072 Shaanxi Peoples R China Beijing Inst Technol Yangtze River Delta Res Inst Jiaxing City Zhejiang Peoples R China 

出 版 物：《NONLINEAR DYNAMICS》 (Nonlinear Dyn)

年 卷 期：2025年第113卷第9期

页      面：9839-9859页

核心收录：

学科分类：08[工学] 0802[工学-机械工程] 0801[工学-力学（可授工学、理学学位）] 

基　　金：National Natural Science Foundation of China Aeronautical Science Foundation of China Shaanxi Province Key Laboratory of Flight Control and Simulation Technology NSFC-RGC joint research scheme [N_CityU114/23] 

主　　题：NDI-based L1 adaptive control Flight control Modified piecewise-constant adaptation Center of gravity variation HIL flight experiment 

摘      要：This paper presents a novel nonlinear dynamic inversion-based L1 adaptive control (NDI-L1) for fixed-wing aircraft with strong nonlinearity, model uncertainties, and center of gravity variations. This adaptive structure can decouple the system s fast adaptation and robustness, thus reducing the oscillations caused by high adaptation gain. Moreover, compared to existing L1 adaptive control, NDI-L1 control enlarges its application range, providing satisfactory dynamic performance as well as robustness within the flight envelope. Additionally, this paper also improves the existing piecewise-constant adaptation (PCA), addressing the contradiction between estimation accuracy and sampling time. The modified PCA achieves the desired estimation accuracy while effectively reducing the computational burden. The dynamic performance of the proposed adaptive structure under disturbance is theoretically analyzed. Importantly, the robust flight controller is designed based on the proposed NDI-L1 to eliminate the influences of parameter perturbations and sudden changes in the center of gravity on flight dynamics. Simulations and hardware-in-the-loop experimental results clearly confirm the effectiveness, advantages, and robustness of the proposed NDI-L1 with modified PCA, and compare it to other existing methods.