Universal Kriging control of hypersonic aircraft model using predictor model without back‐stepping

Abstract

This study describes the design of adaptive universal Kriging (UK) controller for the longitudinal dynamics of a generic hypersonic flight vehicle (HFV). The altitude subsystem is transformed into the explicit four-step ahead state predictor model, which provides the relationship between states and future outputs. Furthermore, the input--output predictor model is derived to obtain the input information for the UK estimator. The UK takes advantages of the parametric regression model and simple Kriging model to describe the systematic aspects and local fluctuations of unknown functions simultaneously. With the state prediction feedback and input--output UK estimation, the controller is synthesised. The control scheme is considerably simpler than the ones based on back-stepping since only one Kriging system is employed to approximate the lumped system uncertainty. Owing to the prediction function and the UK structure, the novel stochastic Lyapunov function-based stability analysis is conducted. The closed-loop system achieves almost surely boundedness. The effectiveness of the proposed control approach is verified by step tracking.