摘要
在干式空心电抗器设计中以等电阻电压法作为计算方法,采用遗传算法进行优化。给出了设计变量的编码方式,离散变量不需要重新规划。对遗传算法的适应度线性尺度变换和最优保存策略作了改进,在遗传操作过程中离散变量和连续变量采用不同的交叉和变异方式。以一种空心电抗器为例,优化后电抗器的外径、高度、重量和损耗分别减小了9%,41%,41%,15%,表明在空心电抗器优化设计中采用改进的遗传算法取得了良好效果。
Genetic algorithm is proposed for the optimum design of air\|core reactor while equal resist\|voltage method is taken as computation method. The encode mode of design variables is given and the discrete variables needn't to be standardized. Fitness linear scaling and elitist model are modified to improve the quality of global convergence of genetic algorithm. Different methods of cross over and mutation are applied for discrete variables and continuous variable during genetic operation. The external diameter, height, weight and power loss of one kind of air\|core reactor are decreased 9%, 41%, 41% and 15% respectively. Therefore, the modified genetic algorithm is an effective method for the optimization of air\|core reactor.
出处
《电工电能新技术》
CSCD
2003年第3期45-49,共5页
Advanced Technology of Electrical Engineering and Energy
基金
国家教育部博士点基金资助项目(2000069808)
