Pre-Sizing of Power Converters Using Genetic Algorithms

This paper presents a computational tool to pre-size quickly and effectively static converters using genetic algorithms, with the objective function to minimize their volume, losses and cost. It was used as starting point an algorithm that performs the overall converter design as a function of the operating point. The static converter design using genetic algorithms is made from your choice of topology, the input specification data, recommendations and standards and a database of components and devices. The language used for the development was JAVA 8 and the database management system was PostgreSQL. The results obtained by the genetic algorithm are compared to the values obtained by algorithms that performs complete scan of the search space, with respect to the quality of the solutions and the processing time to get results. The use of genetic algorithms helps a lot the process static converters designing, so that it quickly offers to designer different and optimized design options, considering the topology and input parameters.