Optimization of Piezoelectric Transformers Using Genetic Algorithm

摘要

In this work, we present the optimization work for piezoelectric transformers (PT) by using the elite Genetic Algorithm (GA). A finite-element solver, NTUPZE, is used to calculate the behaviors of piezoelectric transformers. The design parameters are the dimensions of the devices, and the design optimization objects are the device efficiency and gain. Since GA does not require a gradient function for projection and is capable of dealing with multi-objective optimization problems, it is a good optimization approach for this work. The optimal results of Rosen-type and Rosen-modal-type PTs are presented. The optimized efficiencies for both cases are close to 1 with a 100 kΩ electrical loading connected on the output electrodes.