Hybrid-GA を用いた二段式スペースプレーンの最適設計

This paper proposes a Hybrid Genetic Algorithm (Hybrid-GA) which is suitable for a large scale multidisciplinary optimization problem such as Two-Stage-To-Orbit (TSTO) spaceplane optimal design problem. The Hybrid-GA is implemented by combining Sequential Quadratic Programming (SQP) method with GA. When constructing the Hybrid-GA, there are three problems that should be solved; 1) decision of optimized variables and discrete method, 2) how to use results of local search, and 3) selection for survival method. These problems are discussed and solved in order to effectively combine SQP method with GA. In order to demonstrate the effectiveness of the proposed Hybrid-GA, the TSTO spaceplane optimal design problem, which consists of weight, aerodynamics, propulsion, and flight trajectory analyses, is investigated using the proposed Hybrid-GA, and the optimal results of Hybrid-GA are compared with that of Simple-GA, and SQP methods. Finally, strategy to achieve the TSTO spaceplane is proposed by comparing the optimal results of changing payload weight and maximum wing load factor of booster.