改进遗传算法在牵引变压器优化设计中的应用

提出了一种改进的遗传算法，并将其应用到牵引变压器优化设计中，该方法采用进化前后期分别调整交叉概率和变异概率、二次交叉、对指定基因集中变异以及移民策略等以保证在整个进化过程都能进行有效的空间搜索，克服了传统算法容易陷入局部最优的缺点。并以200km／h电力机车用牵引变压器为例，分别采用传统算法和改进算法对其进行优化，对两者的效率和全局收敛性进行了比较，结果充分证明了改进遗传算法的有效性和先进性。

APPLICATION OF MODIFIED GENETIC ALGORITHM TO THE OPTIMUM DESIGN OF TRACTION TRANSFORMER

A modified genetic algorithm is proposed, and the method is applied to the optimum design of traction transformer. The method includes adjusting the probability of crossover and mutation in the evolutionary process, second crossover, focus mutation to the special gene and emigration strategies. the method can ensure space search is effective in the entire evolution process and overcome the disadvantage of tradition genetic algorithm that is easy to get a local optimum answer. Compared the optimum efficiency and the entire astringency of the tradition and modified genetic algorithm to the 200km/h traction transformer of locomotive, the validity and advantage of the modified genetic algorithm is proved.