In this paper, an accurate neuro-fuzzy-based model is proposed for efficient computer-aided design (CAD) modeling and optimization of microwave circuits and devices. The adaptive neuro-fuzzy inference system (ANFIS) approach is used to determine the scattering parameters of a microstrip filter and is applied to the optimization design of this microstrip filter. The ANFIS has the advantages of expert knowledge of fuzzy inference system and learning capability of artificial neural networks. The neuro-fuzzy model has been trained and tested with different sets of input/output data. Finally, different results, which confirm the validity of the proposed model, are reported.
