Sizing mixed-mode circuits by multi-objective evolutionary algorithms
Ivick Guerra-Gomez, Esteban Tlelo‐Cuautle, Trent McConaghy, Luis Gerardo de la Fraga, Georges E. Gielen, Gerardo Reyes-Salgado, Jesus Manuel Munoz-Pacheco · 2010
We show the behavior of the generations of two multi-objective evolutionary algorithms (MOEAs) for the optimal sizing of two mixed-mode circuits. The non-sorting genetic algorithm (NSGA-II), and the MOEA based on decomposition (MOEA/D) are used to size a second generation current conveyor (CCII+) and a current-feedback operational amplifier (CFOA). Both MOEAs take into account design constraints, and link HSPICE to evaluate the electrical characteristics of the CCII+ and CFOA. Differential evolution is used as genetic operator to show the behavior of the generations of the two MOEAs.