An approach to knowledge-based data handling in complex process

Abstract

This paper presents an approach for knowledge-based data handling in complex Process Control Systems (PCS). Special attention is paid to the specific situations taking place in the technical processes, which run under control of such PCSs. To describe those specific situations we use the concept of Active Rule (AcR). The considered PCSs operate in the real-time mode. Because of this, handling a big number of active rules may cause violations of some real-time constraints. We propose an approach of Association Rules (AsR) in order to eliminate those problems. The collection of AsR is mined from the data describing the specific situations of processes. Those data are collected in PCS as a result of executing appropriate AcR. We elaborate a collection of modification patterns for simplifying the structure of existing AcR, and the general scheme for the interaction between AcR and AsR. The results presented in the paper are based on real-life projects that we performed in the domain of Web-based PCSs for Ukrainian gas-and-oil production enterprises.