206. Circuits in a Box: Multi-Objective Optimization for Analog Electronic Circuits
Invited abstract in session TD-4: Multiobjective Optimization III, stream Multiobjective optimization.
Thursday, 14:10 - 15:50Room: M:M
Authors (first author is the speaker)
| 1. | Christopher Schneider
|
| Fachbereich Grundlagenwissenschaften, Ernst-Abbe-Hochschule Jena |
Abstract
We present a method for approximating so-called performance spaces of analog electronic circuits. Performance spaces as representations for attainable image spaces of analog electronic circuits are well-suited for supporting their conventional design process. The design task can be formulated as non-convex multi-objective optimization problem.
Our approach is based on a box-coverage method for Pareto optimization with modifications to get an approximation of the complete image space of the problem. The thereby resulting models provide accurate information about all valid compromise solutions of the competing circuit performances.
Real world examples demonstrate how this approach can help analog system designers to make reliable topology decisions, even without detailed knowledge about the underlying circuits. Furthermore, we present an interactive visualization technique to explore performance spaces of dimension four and higher.
Keywords
- SS - Multiobjective Optimization
Status: accepted
Back to the list of papers