基于多属性分类的建筑物损伤案例推理方法研究

Research on case-based reasoning of building defects in multi-attribute classification

建筑物损伤信息大多以非结构化离散的方式加以储存,因此建立建筑物损伤案例推理流程已成为提高建筑运维管理水平的重要研究方向.本文根据损伤属性分类的结果设计了符合国际分类标准的结构化的损伤信息采集样例,并完成了损伤领域的本体建模.本文设计了基于案例距离推理的损伤信息多属性分组案例库构建方法,通过计算建筑物损伤等级案例与分组中心点的距离构建案例距离加权优化求解模型,简化损伤决策中多属性比较的复杂过程.本文建立了基于分组式损伤案例库的损伤案例推理流程,验证了针对案例信息的分组预处理后进行不同损伤案例相似度计算的方法,实现损伤信息的有效利用.结果表明所提方法可以较好地解决建筑物损伤多属性信息的分类问题.

Building defects information is normally stored in discrete and unstructured way, for this reason it has become an important field in related research to build up the case-based reasoning framework of building defects for enhancing the level of building maintenance management. In this paper, according to the multi-attribute building defect model, a defect collection sample corresponding to the international classification standards has been designed to capture core contents of defect information, and defect domain ontology is developed based on the defect categories. Then, a multi-attribute classified case-databased of building defect is constructed using case-based distance approach. By calculating the distance between the typical case with defect level and the case group center point, an optimization model of the multiattribute weights is proposed to simplify the complex comparison process of multi attributes in decision making related to building defects. In this paper, a defect case-based reasoning framework with classified case-databased is designed to verify the effectiveness of the calculation method of case similarity which is performed with a preprocessing of classified information of building defects. The analytical result shows that the proposed method has a good performance in solving multi-attribute classification of building defects.