Pre-sizing of power converters using optimization under constraints

The aim of this paper is to present an optimization under constraints approach to pre-size power converters. In fact, the time-domain simulation is very useful to evaluate the performances of power converters when the possible design solutions of the studied converter is much reduced and the different design aspects are uncoupled. However, in a pre-sizing approach using optimization under constraints, a large solution space must be explored and a multi-physic aspect must be considered. In this case, it will be quasi impossible to obtain a reliable results within a reasonable computing time using the time-domain simulation. In this way, the paper proposes to use analytical models to carry out compromise between the model accuracy and the computing time when pre-sizing power converters under volume, electromagnetic compatibility (EMC), efficiency, thermal and control constraints. To do that, generic models are developed to take into account theses different aspects. Theses models are exploited to optimize a 1kW bidirectional buck-boost converter.