Safety-critical nonsingular near-optimal control of hypersonic gliding vehicle with prescribed performance

To address the problem of achieving safety-critical high-performance attitude control of hypersonic gliding vehicle (HGV), this paper proposes a safety-critical nonsingular near-optimal control method with prescribed performance. First, considering the transient and steady-state accuracy demands of HGV flight, a novel nonsingular prescribed-performance safety boundary is established, from which a safety-critical control set is derived. Then, a low-complexity single-critic neural-network-based baseline controller is developed via adaptive dynamic programming (ADP), using a composite performance index of attitude tracking error and energy consumption. This baseline controller is subsequently integrated with the safety-critical control set through quadratic programming (QP) to form the overall safety-critical near-optimal control. Moreover, by incorporating historical data and auxiliary terms, the approach reduces reliance on persistence of excitation and admissible initial control, thereby improving robustness and practical applicability. Finally, Lyapunov analysis demonstrates ultimate uniform boundedness (UUB) of the attitude tracking error, and the effectiveness of the proposed method is demonstrated via numerical simulations.