Optimized Design of Ideal and Actual Transformer Based on Improved Micro-Genetic Algorithm

Abstract

An optimized design method of ideal and stepped impedance actual transformer based on improved micro-genetic algorithm is presented. Firstly the optimized method is imposed on the ideal stepped impedance transformer, which is called first optimization process, then in the actual structure the first optimized result is transformed as a new initial value. Secondly the influence of high order modes nearby the steps is taking into account, a secondly optimized design process which combined with micro-genetic algorithm and mode matching method is employed. And the amount of sampled frequency in the optimization is quantificational analyzed. At last, an ideal transformer in the case of complex load and 100% relative band width is designed, and so done X-Ku band rectangular stepped impedance. The numerical values indicate that the limitation of load from which the synthesis method has, the disadvantage of local optimization which the conventional optimization method holds, the deficiency of the other genetic algorithm's convergence speed and relatively big population size are voided. The convergence speed and the precision in the two optimized design process are better than the other methods.