Multi-objective optimization design for snubber in switching power amplifier
Xiaoming Sun, Hualing Long, Dichen Liu, Po Li · 2009
This paper focuses on discussing the type selection, analysis and optimization design of the snubber in switching power amplifier (SPA). Based on the principle analysis of the selected snubber, its design is summarized as a multi-objective optimization problem (MOP). Then the proposed MOP is solved by the modified nondominated sorting genetic algorithm II (MNSGA-II), which is constructed by introducing to the NSGA-II four new schemes: the pointers grouping fast nondominated sort (PGFNS), the preaccelerating crowding distance assignment (PACDA), the chaotic factor based middle recombination (CFBMR), and the reproduction without duplicates (RWD). The results obtained from simulation and practical experiment show that the optimal snubber parameters sought out by the MNS GA-II make the snubber design attain the expectant goals.