Knowledge base engineering for industrial safety expertise: A model-driven development approach specialization

Abstract

Degradation of equipment in many industries is ahead the rate of its modernization and replacement. As a result, there is a need for a rapid inspection and definition of the possible hazards and appropriate measures to avoid catastrophic failures. The effectiveness of some tasks connected with this inspection (or an industrial safety expertise) can be improved by rule-based expert systems. This paper presents an end-user oriented approach for knowledge base engineering. The approach proposed is based on specialization of the MDA/MDD approach and includes application of ontologies and conceptual models to represent computation-independent models, a domain-specific notation to improve the design of logical rules, CLIPS (C Language Integration Production System) as a programming language for knowledge bases. The research software (Personal Knowledge Base Designer) implements presented models and algorithms. Approbation of the proposed approach is made in the Irkutsk Research and Design Institute of Chemical and Petroleum Engineering (IrkutskNIIhimmash) to create prototypes of knowledge bases for the industrial safety expertise tasks.