Robustness Analysis of Hypersonic Vehicle Controllers

Abstract

Air-breathing hypersonic vehicles, when fully developed, will offer travel in the atmosphere at unprecendented speeds. Capturing their physical behavior by analytical / numerical models is still a major challenge, still limiting the development of controls tewchnology for such vehicles. To study, in an exploratory manner, active control of air-breathing hypersonic vehicles, an analtical, simplified, model of a generic hypersonic air-breathing vehicle in fligt was developed by researchers at the Air Force Research Labs in Dayton, Ohio, along with control laws. Elevator deflection and fuel-to-air ratio were used as inputs. However, that model is very approximate, and the field of hypersonics still faces many unknowns. This thesis contributes to the study of control of air-breating hypersonic vehicles in a number of ways: First, regarding control laws synthesis, optimal gains are chosen for the previously developed control law alongside an alternate control law modified from existing literature by minimizing the Lyapunov function derivative using Monte Carlo simulation. This is followed by analysis of the robustness of the control laws in the face of system parametric uncertainties using Monte Carlo simulations. The resulting statistical distributions of the commanded response are analyzed, and linear regression is used to determine, via sensitivity analysis, which uncertain parameters have the largest impact on the desired outcome.