针对网级评估的区域桥梁信息集成与数据挖掘

Regional bridge information integration and data mining for network-level assessment

摘要:

针对区域桥梁分布范围广、数量大、难以进行综合评估的问题,首先分析单体桥梁状态评估的信息源,进而由单体拓展至区域桥梁群评估,制订区域内检测、监测、交通量、图纸等多源信息的集成规则,按“路线—桥梁—构件”的逻辑形式进行存储和表达,数据清洗后建立路网数据库,对集成后的数据进行特征挖掘,进而对区域桥梁群进行综合网级评估. 以河北省高速公路为背景,依据气候特征划分3个子区域,评估路网总体和分区特征；分析桥梁类型、桥龄、交通量等关键变量与总体技术状态评分的相关性,开展数据挖掘,演绎退化规律,为网级评估提供数据基础和分析条件. 结果表明:所提出的信息集成与数据挖掘方法及应用模式,有效地揭示了路网内桥梁群的共性特征和退化规律,可用于区域桥梁网级评估.

Abstract:

Since regional bridges are widely distributed and large in quantity, it is difficult to comprehensively evaluate them. This paper establishes a condition assessment theory for regional bridges based on the integrated multi-source information of single bridge. The database of regional bridges is stored within a “route-bridge-component” formation, which is extracted from inspection reports, monitoring systems, traffic data, design drawings, and so on. Based on the integrated multi-source information of regional bridges, the performance evaluation of the regional bridges and network-level assessment can be carried out. Taking a partial highway in Hebei province as an example, three sub-regions were divided according to climate characteristics, and the overall and regional characteristics of the road network were evaluated. Data mining was carried out concerning key parameters such as bridge type, age, and traffic volume, and correlation analysis was conducted between these parameters and general technical scoring at network level. The proposed framework provided data basis and analysis conditions for network-level evaluation. Results indicate that the proposed data integration and mining methods can effectively reveal the common features and deterioration rules of regional bridges, and hence can be used for network-level assessment of regional bridges.