Adaptive control of launch vehicles in atmospheric flight

The research focuses on the development of innovative control techniques for attitude control of a launch vehicle (LV) during the atmospheric flight, and to investigate their possible benefits in terms - for instance - of improved disturbance rejection capability, as well as, a means for reducing the burden of recurrent activities of mission integration and flight program software finalization. In this respect, a complete nonlinear mathematical model of the launch vehicle dynamics, comprehensive of all relevant aspects for the attitude control problem is first developed. Next, linearized equations of motion are derived under the assumption of small deviations of the vehicle motion from a reference trajectory. The time-invariant linear model is used to synthesize a baseline controller (BC) that features two proportional-derivative (PD) components for attitude and translational motion control, plus filters to phase-stabilize and notch the bending modes by using classical, frequency-based, analysis, and control design techniques. Among several advanced algorithms referenced in the literature, the Adaptive Augmenting Control (AAC) has been selected and implemented in the LV flight control system in order to retain the functionality and proven record of success of classically designed linear control systems, while consistently and predictably improving their performance and robustness in expanded flight and/or uncertainty parameter envelopes. AAC adjusts the action of a baseline PID-type controller by means of a forward loop gain multiplicative adaptive law that, basically, on-line modulates BC output either to minimize the error with respect to a reference model or to limit undesirable high-frequency response in the control path. In order to fully exploit the AAC potentiality, an effective and reliable tuning procedure for AAC gains is developed, where a robust design optimization (RDO) problem is formulated, and the goal is to maximize a statistical metric that describes FCS performance measured over a set of LV simulations. Finally, an analysis of the effects of uncertainties on bending mode characteristics is carried out. Variations of bending mode parameters have a significant and negative impact on AAC performance and, consequently, on LV stability. In this respect, the use of adaptive filters is investigated in order to further improve flight control system robustness. An adaptive notch filter is designed, the parameters of which are updated continuously by an adaptation algorithm that uses the pitch rate sensor output so as to estimate the unknown parameters of the filter and precisely match the actual bending mode frequency.