Fuzzy-logic-based analog design tools

Abstract

We have developed a collection of tools for the design, modeling, and test of analog circuits. Sharing a common fuzzy-logic based framework, the tools are part of FASY (Fuzzy-Logic-Based Analog Synthesis), an analog design package developed at the University of Seville. The first tool uses fuzzy logic for topology selection of analog cells. It follows decision rules directly entered by a human expert or automatically generated from its experience with earlier designs. Second, a performance-modeling tool provides a qualitative description of a circuit's behavior. Alternatively, it can use a learning process to accurately model circuit performance. Finally, an analog testing tool uses a fuzzy-neuron classifier to detect and classify faults in analog circuits.