| The development of electrical locomotive market in china will be a chance,however, it is also a challenge because of the rise of material's price and the stingingcompetition. The optimization technology and the computer assistant design can helpshorten design time and reduce cost and improve product's quality. So it will bringobvious benefit in economy and society.At First this paper elaborates the design method of electrical locomotivetransformer, establishes the mode of transformer optimization. As to the transformer oflocomotive, the optimum aim function is weight. If variables are beyond the limit, theywill be punished with penalty function.The problem of transformer's optimization is a multi variable, discrete variableand multi peak value problem. The searching technology based on differentialcoefficient is easy to be plunged in local solution, enumerating searching need toomuch time, random searching isn't able to find optimum solution. So the geneticalgorithm is selected to optimize the transformer's design and it is improved in initialcolony, fitness function, genetic operators as to conquer the blemish of geneticalgorithm and find total optimum solution.According to the optimum design method above, this paper introduces thesoftware of transformer design which is based on Visual C++ platform. This software isdivided into three parts, electromagnetism computing module, genetic algorithmmodule, input and output data management module. It has the function of transformerdesign, saving of design scheme and display of solution. Finally it designs an exampleand shows the result. |