抄録
An optimum design technique of balanced surface acoustic wave (SAW) filters using a differential evolution (DE) is proposed. First of all, in order to evaluate the performance of balanced SAW filters based on the computer simulation, a network model of them is composed by using mixed-mode S-parameters. Then the structural design of balanced SAW filters is formulated as a function optimization problem for improving their performance in both the balance characteristics and the filter characteristics. For solving the function optimization problem, a basic DE is employed. In order to apply the basic DE to the optimization problem effectively, a distorted problem space defined by various design parameters is embedded in a regularized continuous search space. Finally, through the computational experiments conducted on a balanced SAW filter, the usefulness of the proposed optimum design technique is demonstrated. Besides, it is clarified that the non-symmetric structure of SAW filter has a possibility to improve its balance characteristics.
