Robust Design Approaches for Hybrid Rocket Upper Stage

Computational costs of robust-based design optimization methods may be very high. Evaluation of new procedures for the management of uncertainty with applications to hybrid rocket engines is here carried out. Two newly developed procedures are presented (hybrid algorithm and iterated local search), and their performances are compared with those of two previously developed procedures (genetic algorithm and particle swarm optimization). A liquid oxygen/paraffin-based fuel hybrid rocket engine that powers the third stage of a Vega-like launcher is considered. The conditions at third-stage ignition are assigned, and a proper set of parameters are used to define the engine design and compute the payload mass. Uncertainties in the regression rate are taken into account. An indirect trajectory optimization approach is used to determine a mission-specific objective function, which takes into account both the payload mass and ability of the rocket to reach the required final orbit despite uncertainties. Results show that for this kind of problem, particle swarm optimization and iterated local search outperform the genetic algorithm, but the use of a local search operator may slightly improve its performance.