Kriging Metamodel Based on Genetic Algorithms and Its Application in Analog IC Optimization

We present a method of constructing the Kriging metamodel of circuits and combining it with genetic algorithms (GA) to optimize a circuit and determine the parameters.Compared with the polynomial regression model,the Kriging model is more suitable to the circuit simulation experiments.Using GA,the problem of the circuit global optimization is solved based on the Kriging metamodel.The method is applied to the optimization design of the band-gap voltage reference,and the result shows that the presented method is effective.