논문 : 생물학적 영양염류제거 하수처리장의 설계 및 운전 최적화를 위한 지식기반 전문가시스템

장덕, 신영진, 성문성 · 2003

As there are many difficulties in the optimal design and operation of municipal wastewater treatment plant for biological nutrient removal (BNR) due to complexity of diversified influential factors and its associated requirements of extensive professional knowledge and experience, introduction of an artificial intellectual expert system to the domain of BNR was believed to be useful for problem-solving and decision-making in the design and operation of BNR processes. Research on embodiment of an integrated knowledge based expert system was carried out to provide optimized BNR expertise to process designers and operators by simulating the problem-solving methodology of the domain experts. An original powerful logical model designated as Domain Definition Box Diagram, which can clarify the problem domain of BNR simplifying and systematizing complex process parameters, was devised for structural analysis of the problem domain, and used in data mining for scenario analysis and extraction of cellularized unit boxed and blocked knowledges. A problem-solving strategy named Strategic Scenario Analysis was also designed in order to establish and analyze classified object-oriented scenarios for potential problems in the BNR processes. A number of structured elements comprising the strategic scenarios were systematically classified and relatively evaluated using Factor Matrix devised in this study. These originally designed methods ensured systematicness, standardization, accuracy, easiness, expandability, and quickness indispensable to the successful construction of BNR knowledge base and its inference rules. A hybrid knowledge base for a typical basic BNR process, A^2/O process, was able to be constructed using an expert system development tool of Blaze Expert based on the strategic scenario and its constituent unit knowledges represented as standardized objects and rules. Through the case verification process employing the integrated BNR expert system comprised of knowledge base, inference engine, graphic user interface and database, it was confirmed that the expert system could provide satisfactory logically inferred expertise, professional diagnostic advice or countermeasures, and speedy search for related information while enhancing user friendly and system efficiency.

Cited by 0 works.

Authors