Multi-Objective Optimization of Power Converter Sizing Based on Genetic Algorithms: Application to Photovoltaic Systems
Résumé
In the development of power converters interfacing photovoltaic panels with the grid, the increasing efficiency and the decreasing of coasts and volume are the important criteria. To achieve these requirements, sophisticated computer -aided design optimization is required. Optimization techniques using genetic algorithms (GA) have received attention in power electronics optimization, since GA have the ability of handling discrete design variable. This paper provides a discrete optimization approach for the sizing of power converters. An important Pareto -based multi-objective optimization algorithm namely Non-dominated sorting Genetic Algorithm (NSGA-II) is used to obtain the design. As an example, a boost converter used for photovoltaic applications under some electrical constraints is sized for minimum costs, volume and losses.