一个用于化工事故分析的基于案例推理的专家系统.pdfVIP

2008 年中国过程系统工程年会会议论文集 一个用于化工事故分析的基于案例推理的专家系统 李潇，张贝克，马昕 （北京化工大学信息学院,化工安全教育部工程研究中心， 北京， 100029） 摘要：针对流程工业的安全评价，开发了基于案例推理的化工事故分析系统，为依赖人工推理的HAZOP方法提供可靠性 依据，激发专家头脑风暴。简要介绍了CBR的实现过程，描述了化工事故推理系统的结构、包括案例的表达、案例的检 索、案例的推理以及学习。其中，案例属性的表达是依据对化工事故的分类思想。案例索引采用的是分步策略，第一步 得到案例检索的子集，第二步在子集中匹配案例并排序。着重研究了数字属性和语义属性的匹配以及相似度的算法，并 结合模糊技术和语义分词技术给出了具体实现方法。实际案例的应用结果表明本系统有效的提高了事故推理的自动化程 度，具有很高的实际意义。 关键词：案例推理；化工事故推理；匹配算法；安全评价；危险可操作性分析 中图分类号：TP391.77 文献标识码：A One Expert System Using in Chemical Accident Reasoning Based on CBR LI Xiao, ZHANG Beike, MA Xin (Research Center of the Ministry of Education for Chemical Safety Engineering, School of Information Science and Technology, Beijing University of Chemical Technology, Beijing, 100029, China) Abstract: Aiming at Process Industry safety assessment, a chemical accident reasoning system oriented to safety assessment using CBR is developed, which provides reliable basis for HAZOP depending on artificial reasoning and simulates brains storm. This paper briefly introduces realization process of the CBR. The frame of the system, the representation of accident property, the index strategy and the revise method of this system are described. The representation of accident property is according to classifying thought. Index uses two-part strategy .The matching algorithm and similarity measurement of numerical and semantic properties are emphatically studied .Implementation method is presented combined with the key technology. The practical case study demonstrates that this accident analysis system for chemistry industry using CBR improves the level of automation in accident reasoning with high practical significance. Key words: case based reasoning; chem