| It's very attractive to optimize the balance transformers for their charming profits, enhanced total performance, lowered power cost, and manufacture cost, therefore more and more researcher focus on it. This dissertation tries its best to develop a satisfactory optimization design method of the impedance-matching balance transformers.What is called optimization design just extracts a satisfactory solution with the help of computer; this process should pass through several steps which employ the decisions of optimization variables, an established mathematical model, the confirmation of some initial design parameters, and optimization design method, most important of all.By description of some principles and characteristic of the genetic algorithm, aiming at several drawbacks of the traditional genetic algorithm, using modern gene engineering for reference, this dissertation proposes a gene handling genetic algorithm to implement some complex optimization problems in practice. Analyze how to ensure the smallest limit of core length, and give the formulation of laminated dimension when the given diameter limit is zero. This paper gives the gene handling genetic algorithm and the gene express genetic algorithm, separately, to accomplish the optimization transformer core section. The optimization results show the effect.Applying the gene handling genetic algorithm to the single aim function optimization of the impedance-matching balance transformer, this paper studied the influences of crossover probability and mutation probability, and chalk up the conclusion that crossover probability and mutation probability only affecting the convergence speed. Finally, via a design example, this dissertation expounds the multiple aim functions design process of the impedance-matching balance transformers. By means of the threshold limitation and chromosome selection with different probabilities, the population can evolve to the best trade-off of the premature convergence.
|
PDF Full Text Request