Modulating functions for modelling PWM switching converters

The aim of the work is to identify continuous average models of switching converters from input-output data. Among the various identification techniques, modulating functions were chosen due to their capability for revealing converter dynamics, even in the case of fast time constants. Due to its low-pass filtering property, such an identification technique reveals its effectiveness in the case of the evaluation of average models for switch-mode DC-DC converters. Following a brief description of the modulating functions properties, simulation and experimental examples of application are presented, first for the case of the basic boost converter and then for more complex topologies, like the half-bridge current doubler converter. The last example considers the application of such a technique to a resonant converter: a relevant result is that the technique allows an easy modelling particularly in critical cases of resonant converters where traditional averaging techniques fail.