Abstract
Real-time expert systems are not just fast systems but specifically designed to incorporate time constraints imposed by the application—usually a process to be monitored or controlled. This paper presents an approach by which the reasoning methods of an expert system are decoupled from the management and handling of rapidly changing input data allowing them sufficient time to complete. In a second step a cooperation layer based on the multiagent system paradigm is presented that is directly integrated into this real-time architecture.
Get full access to this article
View all access options for this article.