Research Of CBR-based Reasoning Framework For Automating HAZOP Analysis

Hazard and operability (HAZOP) analysis is a widely recognized process hazards analysis (PHA) technique used for hazard identification in the chemical process industry in the world. To overcome the shortcoming of the existing HAZOP expert systems with regard to "non-routine" analysis and improve the completeness and consistency, a case-based reasoning framework is proposed in this paper.The case base and the structure of cases are briefly described and the case search and matching strategies are presented. The knowledge management module, case creator and CBR reasoning module are designed and developed. The paper introduces the implementation of CBR engine in CBR reasoning module emphatically. To facilitate the management of the knowledge base, a case constructor is developed. The industrial case study demonstrates that the CBR-based HAZOP expert system has broken through the technical bottleneck of the existing HAZOP expert systems in machine learning and "non-routine" HAZOP analysis. It improves the completeness of HAZOP analysis.