摘要
遗传算法的随机性和隐含并行性,使它能同时搜索到多个局部最优解并获得最优解集。为了发挥遗传算法群体搜索的优势,提高多目标优化设计效率和灵活性,在自适应遗传算法的基础上引入群体排序技术、小生境技术和Pareto解集过滤器,建立了一种适用于多目标优化设计的Pareto遗传算法。以Pareto前沿面的形式给出优化设计的Pareto最优解集,供设计者按设计意愿选择最优的设计结果。采用Pareto遗传算法进行跨声速翼型的多目标优化设计,设计结果表明,Pareto遗传算法是十分有效的,完全可以用来进行多目标优化设计。
Stochastic and implicitly parallel properties of genetic algorithm make it possible to search for multiple local optimal solutions and obtain optimal solution aggregate. In order to elaborate the population search advantages of genetic algorithm and improve the efficiency and flexibility of multi-objective optimization design Pareto genetic algorithm, a new method suitable for multi-objective optimization design, is established based on self-adaptive genetic algorithm which population ranking technique, niche technique and pareto solution set filter are introduced. Pareto optimal solution aggregate may be provided in the form of Pareto front from which designers may select some suitable optimization design results according to their inclination. Finally the Pareto genetic algorithm is applied to carry out multi-objective aerodynamic optimization design of transonic airfoils. The design results show that pareto genetic algorithm is effective enough to be used in multi-objective optimization design.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2003年第12期1558-1561,共4页
Systems Engineering and Electronics
