遺伝的アルゴリズムによる機械システムの最適化法とそのエンジン諸元の最適設計への適用

概要

論文の詳細を見る
An optimization method based on genetic algorithms is proposed for a class of design potimization problems of mechanical systems, where design variables are real numbers, objectives and constraints have rugged response characteristics, the number of obiectives is plural. While this kind of optimization problems are hard for conventional mathematical optimization techniques, genetic algorithm based approaches are promising due to their robustness against optimization hardness. This paper proposes a genetic algorithm based multi-objective optimization method, that introduces Pareto-optimality based fitness function, similarity based selection, life span limit of elite preservation and direct real number crossover. This optilnization method is also applied to the optimal design problem of an automotive engine with the design criteria on a total power train. The computational example shows the ability of the proposed method for finding a relevant set of Pareto optima.
一般社団法人日本機械学会の論文
1999-02-25