Отправлено: 20.06.07 02:17. Заголовок: Управление знания, рабочие потоки, ИИ

Knowledge Management for Shared Earth
Modelling (SPE-107152)
L. Mastella* (Ecole des Mines de Paris/UFRGS), M. Perrin (Ecole des Mines
de Paris), M. Abel (UFRGS), J.F. Rainaud (Institut Français du Pétrole) & W.
Touari (Institut Français du Pétrole)
SUMMARY
Shared Earth Modelling (SEM) describes the process of integrating static (geological structures) and
dynamic data (hydrocarbon production) issued from several disciplines to construct a model of a
subsurface resource. We claim, however, that SEM should allow sharing and integrating not only data but
also the knowledge related to data interpretation. As a consequence, actors from different disciplines, who
have different technical skills, should be able to provide a ""Shared Earth Model"" with specific technical
knowledge concerning how to evolve a hydrocarbon prospect at the various stages of its life cycle.
The integration of various levels of scientific and technical knowledge into the modelling process requires
an overall view of the knowledge and data workflow. In order to enhance a knowledge management
approach, we first examine which types of information (data, interpretation, abstract model, symbolic or
spatial representations) should be considered at each step of the reservoir modelling chain. Moreover, in
order to allow each expert to bring his/her own contribution in an explicit way, we propose the use of a
formal vocabulary related to the different disciplines and define it through ontologies. It is thus possible to
picture how the various types of knowledge should interact for producing consistent ""Shared Earth
Models"".
In order to illustrate the above approach, the paper presents a prototype of ""Geological Knowledge
Editor"" (GKE). This module allows representing the knowledge acquired at the beginning of the SEM
process and providing services to enhance this knowledge and checking its consistency, guaranteeing thus
the quality of service (QoS) attached to the reservoir modelling process. By means of the GKE, the
geologist can display geological representations such as fault networks or stratigraphic columns. He/she
can also edit the chronological interpretation attached to the model and use it for building structural
models in an automatic way.

Работа жителей франции
Мадам Лаура Мастелла перебралась из Бразилии, где занималась применением методов анализа данных и ИИ в геологии
В статье приведены небезынтересные разветвленные рабочие потоки и рассуждения на этот счет, наряду с примерами применения технологий