Empirical analysis on blast furnace fault diagnosis method based on ontology

Blast furnaces’ maintenance records of mechanical fault contain great amount of information which is useful to identify the similar fault，but the structural representation and the knowledge reasoning problem prevent us to use the records effectively. Aiming at the problem，a blast furnace fault knowledge representation based on ontology is proposed. Blast furnaces’ knowledge characteristics，ontology building significance and principle of fault are analyzed，and the four factors of BF cause，BF phenomenon，BF ways，Blast furnace in blast furnaces’ mechanical fault diagnosis are brought out. The fault ontology frame of BFClass，BFProperties，BFIndividuals，BFRelationships，BFDomain are defined and described in detail. The applications of fault diagnosis ontology are studied in protégé. The case study shows that the fault knowledge representation based on ontology is very intuitive and efficient，and it provides a novel way to deal with the blast furnace fault diagnosis problem.

Project Supported:

Clc Number: