基于改进案例推理的钢结构破损原因诊断分析
Detection Analysis of Steel Structural Damage Based on CBR
-
摘要: 为了更有效地寻找钢结构发生破损的原因,快速得出钢结构事故的处理方法与建议,建立了基于改进案例推理(case-based reasoning,CBR)方法,进行钢结构事故破损原因的诊断.根据钢结构事故案例分析结果的特点,提出用统一化、标准化的破损特征与原因关系矩阵来表达事故案例分析结果的方法,矩阵参数为可能度系数,表达特征与原因的引发关系,进而归纳所有案例得出该案例库的破损特征与原因关系经验矩阵,并且根据待求事故的特征,得出事故发生的主要原因.该方法已编译成计算机人工智能软件,可作为工程事故人员分析时有利的辅助工具.Abstract: To detect the reasons for the damaged steel structure,and find the processing method and advices effectively,a method based on the improved case-based reasoning(CBR) is presented for diagnosing the damaged reason.This paper describes the analytic model of the damaged characteristics and reasons in detail,and puts forward the method expressing the analytic results using the standard damaged characteristic and reason relationship matrix.The matrix parameter is a probable index,expressing the relationship between characteristics and reasons.Then,all cases are concluded to obtain the empirical matrix,and the main reasons for the accident based on the characteristics of the actual accident.This method has been compiled to the artificial intelligence software,using as a convenient and effective aid for the engineering accident staff to analyze.