A genetic algorithm with discrete cost functions is used to design and optimise a 42 V-14 V/8 A buck converter. The use of discrete and continuous cost functions are confronted. Discrete cost functions produces better designs. Electrical time domain simulation is used for verification. Experimental measurement, electrical and thermal, validates the design
Published in:
Power Electronics and Applications, 2005 European Conference on
Date of Conference: 0-0 2005