CAFFEINE: Template-Free Symbolic Model Generation of Analog Circuits via Canonical Form Functions and Genetic Programming

TL;DR

This paper introduces CAFFEINE, a novel method that automatically generates compact, interpretable symbolic models of analog circuits from simulation data using grammar-constrained genetic programming, improving prediction accuracy over traditional methods.

Contribution

The paper presents a template-free, grammar-guided genetic programming approach for symbolic modeling of analog circuits, enabling flexible, accurate, and understandable models without prior equation templates.

Findings

Generated models are compact and interpretable.

Models outperform posynomials in prediction accuracy.

Approach effectively balances error and complexity.

Abstract

This paper presents a method to automatically generate compact symbolic performance models of analog circuits with no prior specification of an equation template. The approach takes SPICE simulation data as input, which enables modeling of any nonlinear circuits and circuit characteristics. Genetic programming is applied as a means of traversing the space of possible symbolic expressions. A grammar is specially designed to constrain the search to a canonical form for functions. Novel evolutionary search operators are designed to exploit the structure of the grammar. The approach generates a set of symbolic models which collectively provide a tradeoff between error and model complexity. Experimental results show that the symbolic models generated are compact and easy to understand, making this an effective method for aiding understanding in analog design. The models also demonstrate better prediction quality than posynomials.