Robust Linear Output Feedback Control of an Airbreathing Hypersonic Vehicle

TLDR

The paper focuses on robust output‑feedback control for an airbreathing hypersonic vehicle model. The study aims to achieve robust velocity and altitude tracking under model uncertainties and varying flight conditions using limited state information, and introduces an estimation‑free design that improves performance. The authors employ robust servomechanism theory with a novel internal model, compensate unstable zero dynamics via pitch‑rate measurements, and select the plant’s output map through sensor placement to preserve favorable system structures. Simulations show that the baseline observer‑based controller lacks robustness to fuel‑consumption dynamics, while the proposed estimation‑free design achieves superior performance across a range of operating conditions.

Abstract

This paper addresses issues related to robust output-feedback control for a model of an airbreathing hypersonic vehicle. The control objective is to provide robust velocity and altitude tracking in the presence of model uncertainties and varying flight conditions, using only limited state information. A baseline control design based on a robust full-order observer is shown to provide, in nonlinear simulations, insufficient robustness with respect to variations of the vehicle dynamics due to fuel consumption. An alternative approach to robust output-feedback design, which does not employ state estimation, is presented and shown to provide an increased level of performance. The proposed methodology reposes upon robust servomechanism theory and makes use of a novel internal model design. Robust compensation of the unstable zero dynamics of the plant is achieved by using measurements of pitch rate. The selection of the plant's output map by sensor placement is an integral part of the control design procedures, accomplished by preserving certain system structures that are favorable for robust control design. The performance of each controller is comparatively evaluated by means of simulations of a full nonlinear model of the vehicle dynamics and is tested on a given range of operating conditions.