Methodology for Analysis of Gm-C Filters based on Statistical, Fuzzy Logic and Machine Learning Approach

Abstract

In the paper a new approach for analysis of Gm-C filters is presented that is suitable for automation of some engineering tasks and integration in CAD tools. The proposed methodology includes data gathering through simulation and circuit mathematical description, utilization of statistical experimental design technique, fuzzy logic method and machine learning algorithm. It is verified through analysis of a low pass Gm-C second order Butterworth filter and creation of several models: high level behavioral VHDL-AMS model of Gm-C active filter, statistical and fuzzy logic based model of inference and machine learning analytical model.