Circuit Model Parameter Extraction and Optimization for Microwave Filters

Abstract

This thesis presents a method for parameter extraction of circuit elements from microwave filters. This diagnosis method can be applied to a sufficiently large number of filters and it can also be used in conjunction with a neural network model for filter design, greatly reducing development time. This thesis is an introduction of parameter extraction and circuit modelling through use of neural networks. It also presents an implementation of the proposed method as well as numerical results and validation data. Detailed implementation code is presented in the appendix.