AN IMPROVED MEASUREMENT STRATEGY BASED ON STATISTICAL DESIGN OF EXPERIMENTS FOR A BOOST CONVERTER

This work presents a measurement strategy based on the statistical design of experiments to analyze the switching and conduction losses of the switches of a boost converter. The measurement strategy considers the measurement errors compensation and the reduction of sensors, but achieving a good quality and accuracy of the directly developed measurements. A very detailed analysis of the phenomena and parasitic elements of the circuit, which produce incorrect measurements of the variables of interest, is performed. Some recommendations are presented in order to take advantage, in an optimum manner, of the different interrupter-diode combinations that have been analyzed.