Hybrid adaptive ascent flight control for a flexible launch vehicle

Abstract

For the purpose of improving trajectory tracking performance of a launch vehicle under adverse flight conditions, this article describes the development of a hybrid adaptive ascent flight control architecture. The hybrid adaptive control approach augments linear feedback signals with contributions from both direct and indirect adaptive control elements. In addition, signal filters are incorporated into the feedback loop to prevent harmful interaction between the flight control system and structural bending modes. The performance of the hybrid adaptive flight controller is compared to that of a typical gain-scheduled linear feedback controller in a high-fidelity Ares I ascent simulator obtained from the NASA Marshall Space Flight Center. Results from simulations which consider nominal flight conditions show that the hybrid adaptive controller is able to maintain a tracking error magnitude that is comparable to or lower than its classical proportional–integral–derivative (PID) counterpart for the vast majority of the ascent phase. However, when a 10 per cent thrust reduction is applied to the first-stage rocket motor, the PID controller experiences significantly elevated error that diverges rapidly at the end of the simulation. In contrast, the hybrid adaptive controller steadily maintains a tracking error magnitude that is frequently 50 per cent or less of the corresponding PID value during the latter stages of ascent.