Multidisciplinary Design Optimization of Solid Launch Vehicle Using Hybrid Algorithm

Abstract

A new approach is proposed incorporating heuristic optimization algorithm with a switch to a local search algorithm for the multidisciplinary design of solid launch vehicle at conceptual level. Heuristic optimization algorithms such as Genetic Algorithm often may locate near optimum solutions at the expense of large function evaluations. Local search algorithms including both gradient and non gradient based optimization methods are only efficient in finding the optimal solutions within convex areas of the design space but fail to determine the global optimal in multimodal design space. The hybrid optimization approach presented in this paper switches between the global and local search methods to minimize the gross liftoff weight to deliver the predetermine payload to the low earth orbit. The computational efficiency of the devised hybrid method is compared with the computational efficiency of conventional hybrid method applied to the same problem. Comparison of the developed scheme with the conventional optimization approach shows that developed approach requires less computation time. Nomenclature e A = area of nozzle exit, m 2 t A = throat area of nozzle, m 2 c ∗ = characteristic velocity, m/sec T c = coefficient of thrust 0 g = acceleration due to gravity, m/sec sp I