Konferensartikel

Flow Network based Diagnostics for Incorrect Synchronous Models

Hans Olsson
Dassault Systemes, Sweden

Ladda ner artikelhttp://dx.doi.org/10.3384/ecp19157743

Ingår i: Proceedings of the 13th International Modelica Conference, Regensburg, Germany, March 4–6, 2019

Linköping Electronic Conference Proceedings 157:77, s. 8

Visa mer +

Publicerad: 2019-02-01

ISBN: 978-91-7685-122-7

ISSN: 1650-3686 (tryckt), 1650-3740 (online)

Abstract

This paper will present a novel way to give diagnostics for incorrect synchronous models.

The goal is that this will ease the introduction of synchronous models, since unclear diagnostics often create a barrier for new users. In particular the case of separating the clocked and continuous parts will be considered, and shown to be equivalent to finding a “leak-flow” in a certain flow network, which can be solved using max-flow/min-cut techniques.

The result is efficient, easy-to-adapt, and gives diagnostics focused on correcting the issue.

We have not seen this idea used before in this context, even if in retrospect it seems natural and straightforward.

The methods have been implemented in Dymola 2019 (released in June 2018) and also in 3D Experience Platform 2019x.

Nyckelord

synchronous, graph theory, flow networks, minimal cut, error diagnostics

Referenser

Inga referenser tillgängliga

Citeringar i Crossref