Optimal closed-loop launch vehicle ascent guidance and control under disturbances

Abstract

This thesis presents optimal rocket-powered launch vehicle ascent guidance with disturbances. The ascent guidance system for launch vehicle determines attitude commands and engine throttle command during the ascent phase of vehicle. The optimal control theory is able to be adapted to solve optimal ascent guidance problem and that solution perform an important role to increase payload capability of launch vehicle or de-crease fuel consumption with respect to given payload. Most of the ascent guidance methods are based on the optimal open-loop guidance scheme and this conventional way of approach has been used with day-of-launch consideration. On the other hand, optimal closed-loop guidance scheme had been not preferred be-cause system model uncertainty, aerodynamics and wind disturbance make optimal closed-loop problem more difficult to be converged in short time using on-board computer. Optimal open-loop guidance method could be a good enough solution in some cases but it is too sensitive to parameter changes and uncertainties. If there is no launch vehicle heritage or data, building an open-loop guidance system becomes more difficult task. Several closed-loop based guidance systems are being developed to build a feasible algorithm to solve complex optimal guidance problem using on-board computer. The most significant aspect of closed-loop guidance system is to minimize convergence or computation time. In the closed-loop guidance system, on-board computer updates guidance command by regarding current states as initial states. During the ascent phase, path constraints and terminal condition have to be satisfied to conduct successful launch mission. A path following method could be implemented to conduct optimal closed-loop guidance. A role of path follow-ing method is to reduce position errors between flight path and nominal path. This fact leads to very small changes or perturbations between previous optimal closed-loop guidance solution and next optimal close...