A Response Surface Methodology Approach to Study the Influence of Specifications or Model Parameters on the Multiobjective Optimal Design of Isolated DC-DC Converters

Nowadays, the use of mathematical optimization techniques to design power converters has become more and more attractive. This paper goes a step further in studying the influence of some specifications (e. g., the input voltage of the converter) or model parameters (e. g., the transformer heat-transfer coefficient) on the multiobjective optimal design of isolated dc-dc converters. A response surface methodology (RSM) approach is used in combination with a multiobjective optimal design tool dedicated to such power converters and using genetic algorithms. At each time, two factors are studied while two response variables are of interest, namely, the power loss and the weight of the converter. The pairs of factors are: the input voltage and the output power, the current and voltage ripples (through the inductor and across the capacitor of the output filter), and, finally, the convection heat-transfer coefficient and the maximal isolation transformer temperature. It should be pointed out that the full-bridge topology is considered in this paper but the proposed methodology is easily applicable to other topologies. It can also be extended to study other factors or other responses of interest.