Hybridization Approches of Metaheuristics for Optimal Analog Circuit Design

The analog circuit design is a hard complex and time consuming task. Recently, Ant Colony Optimization algorithm (ACO) has emerged as a powerful and useful method for optimal design of analog circuits, based on the finding of the good compromise between the most important performances or merits figure of a given circuit. As a drawback it needs longer time execution when compared for instance to the genetic algorithm (GA) or to the simulated annealing (SA) one. In order to improve execution speed, we propose to combine the ACO to both the GA and SA algorithms to build up two new hybrid algorithms (SA-ACO) and (GA-ACO). The main idea is first, to use the GA or SA algorithm to explore the promising search space and secondly, to exploit the best solutions by the ACO algorithms to rapidly achieve the optimal solution. The obtained hybrid algorithms (SA-ACO) and (GAACO) are finally used for the sizing of a CMOS second generation current conveyor (CCII) and an operational amplifier (Op-Amp). The performances of the proposed algorithms will be highlighted in terms of computing time, convergence rate and the optimum quality.