Multidisciplinary design optimization of a reentry vehicle using genetic algorithm

The purpose of this paper is the optimal design of a reentry vehicle configuration to minimize the mission cost which is equal to minimize the heat absorbed (thermal protection system mass) and structural mass and to maximize the drag coefficient (trajectory errors and minimum final velocity).

There are two optimization approaches for solving this problem: multiobjective optimization (lead to Pareto optimal solutions); and single‐objective optimization (lead to one optimal solution). Single‐objective genetic algorithms (GA) and multiobjective Genetic Algorithms (MOGA) are employed for optimization. In second approach, if there are n objectives (n+1) GA run is needed to find nearest point (optimum point), which leads to increase the time processing. Thus, a modified GA called single run GA (SRGA) is presented as third approach to avoid increasing design time. It means if there are n objectives, just one GA run is enough.

Two multi module function – Ackley and bump function – are selected for examination the third approach. Results of MOGA, GA and SRGA are presented which show SRGA approach can find the nearest point in much shorter time with acceptable accuracy.

GA, MOGA and SRGA approaches are applied to multidisciplinary design optimization of a reentry vehicle configuration and results show the efficiency of SRGA in complex design optimization problem.