The use of differential evolution (DE) for the optimization of a rocket based combined cycle (RBCC) inlet is investigated in supersonic flight. The inlet employs a semi-annular rocket-ejector exhaust profile, with the external geometry constructed around the internal rocket flow path. The inlet external flow characteristics are determined using a semi-analytical shock fitting method. The design objectives are to maximize air mass flow and total pressure recovery, while considering the rocket exhaust shear layer area and turning angle. The inlet geometry is created using 8 normalized input variables, which results in a total of 6 ∗ 109 possible configurations. By treating each combination as an individual in a population, with the input variables representing an individuals’ genes, the DE algorithm is used to indicate the optimal, or most fit, individuals. Simulations are done at a freestream Mach number of 2.5, and altitude conditions of 18.63km. The initial populations are randomly generated and showed convergence towards a global Pareto front region of non-dominated individuals. These non-dominated individuals represent geometries with the most desirable characteristics, according to the design objectives.

Additional Metadata
Persistent URL dx.doi.org/10.2514/6.2018-0409
Conference AIAA Aerospace Sciences Meeting, 2018
Citation
Jee, C. (Craig), & Etele, J. (2018). Multi-objective optimization of a supersonic rocket based combined cycle inlet through differential evolution. In AIAA Aerospace Sciences Meeting, 2018. doi:10.2514/6.2018-0409