Neural-based online model-corrected prescribed-time composite anti-disturbance control strategy for hypersonic flight vehicles

To address the attitude tracking control problem for hypersonic flight vehicles (HFVs) with multiple disturbances and significant time constraints, a neural-based online model-corrected prescribed-time composite anti-disturbance control (NOMC-PTCADC) strategy is developed in this paper. First, the HFV model is constructed, then the corrected control-oriented model is derived to decouple the inherent uncertainties and external disturbances. Second, a deep neural network-based aerodynamic coefficient identification (ACI) module is proposed to handle the inherent uncertainties online, and the NOMC mechanism is established. Third, to estimate the other disturbances primarily caused by the external disturbances, a prescribed-time extended state observer (PTESO) is developed. The proposed NOMC mechanism and the PTESOs together form the composite anti-disturbance control mechanism that can precisely handle each type of disturbance. Subsequently, a prescribed-time control (PTC) strategy is developed to meet the significant time constraints in HFVs control. The NOMC mechanism, PTESOs and PTC strategy together form the practical PTCADC framework, ensuring practical prescribed-time convergence for HFVs with multiple disturbances. Furthermore, the proposed closed-loop system exhibits excellent prescribed-time convergence performance according to Lyapunov stability analysis. Finally, a series of numerical experiments are conducted to verify the superiority of the proposed control scheme.