Optimization under Constraints of Static Converters

Abstract

The paper deals with an approach to size and optimize static converters using analytical models and a gradient optimization algorithm. The aim is to carry out a compromise between the computation rapidness and the result accuracy, which allows to treat more criteria in the same optimization process than a numerical or simulation approach. This method is applied to optimize the volume of a PFC (Power Factor Correction) flyback converter under EMC (electromagnetic compatibility), loss and thermal constraints. In this way, an analytical model is developed for each optimization aspect. However, these models are often described by discontinuous functions which are built from manufacturer data (transformer winding wire diameter, magnetic circuit volume...). To make easier the use of these models in an optimization approach with a gradient algorithm under constraints, specific tools are developed and integrated in an adequate optimization environment: Pro@design. The optimization results show that the optimal volume depends on both the converter parameters to be optimized (the input parameters) and the imposed constraints (the criteria). These optimization results are validated by experiment