Multiobjective optimization method for analog circuits design based on fuzzy logic

The paper presents a new method for optimizing analog circuit design. The method performs a multiobjective optimization, the parameter modification taking into account the degrees of unfulfilment for all the requirements. For the optimization problem formulation, it uses fuzzy sets to define fuzzy objectives and for the optimization strategy, it uses a fuzzy system to compute new parameter values. The solution of the multiobjective optimization is a set of Pareto optimal points and we try to find it by means of a population of solutions. After introducing our optimization method, we validate our method by optimizing the design of a basic two stage CMOS op-amp. As expected, the results prove that the method is efficient and supplies very good final optimal solutions.