论文部分内容阅读
自1980年以来,“专家系统”作为人工智能中一个活跃的分枝已被引入土木工程领域,开始对房屋结构的破损状态进行诊断,但至今尚未进入实用阶段。其主要困难是专家知识的获取与表达。在以往的研究工作中,可能过高地估计了从事人工智能研究的“知识工程师”的作用,而相对地低估了房屋诊断决策方法上的难度。本文根据从局部到整体和从个别到一般的认识原则,提出一个对钢筋混凝土结构进行破损评估的语言决策模型。根据这个模型建立了一个对钢筋混凝土梁进行诊断的原型系统、该系统用LISP语言写成,并成功地在IBM-PC微机上运行。通过多次考核证明,该系统可以得出和专家评估基本一致的结果。
Since 1980, the “expert system” has been introduced into the field of civil engineering as an active branch of artificial intelligence, and began to diagnose the damaged state of the housing structure, but it has not yet entered the practical stage. Its main difficulty is the acquisition and expression of expert knowledge. In past research work, the role of “knowledge engineer” engaged in artificial intelligence research may be overestimated, and the difficulty in the method of housing diagnosis and decision making is relatively underestimated. This paper presents a language decision model for evaluating the damage of reinforced concrete structures based on the principle of understanding from the part to the whole and from the individual to the general. Based on this model, a prototype system for the diagnosis of reinforced concrete beams was established. The system was written in LISP and successfully operated on an IBM-PC microcomputer. Through multiple assessments, the system can produce results that are basically the same as the experts’ assessments.