摘要
遗传算法作为一种比较成熟的智能算法,因其具有全局搜索能力和并行性得以在翼型气动优化中广范应用。本文在编码方式、种群初始化和遗传算子等方面对标准遗传算法进行了改进。其中,DNA的编码方式增加信息的丰富性;拉丁超立方抽样初始化使种群分布相对均匀;插入、删除、倒位等算子增加种群的多样性,加快收敛;感染算子加速种群摆脱停滞或早熟。计算结果表明:与标准遗传算法相比,改进后的DNA编码遗传算法收敛更快,全局性更好。
Genetic algorithm,which is a mature aptitude algorithm,has been widely applied to airfoil design.In this paper,the standard genetic algorithm is improved by encoding genetic operators and initialization,So that the DNA encoding includes more information;the Latin hypercube sampling distribution of the initialization population is more reasonable;insertion,deletion and inversion operators increase the diversity of the population to accelerate convergence;infection operator gets rid of stagnation or premature of population.Function test and airfoil design show that this means has faster convergence and better global ability compared with standard Genetic Algorithm.
出处
《航空工程进展》
2011年第2期157-162,共6页
Advances in Aeronautical Science and Engineering
基金
西北工业大学翱翔之星计划
关键词
遗传算法
翼型设计
DNA编码
genetic algorithms
airfoil design
DNA encoding
