检测数据下城市桥梁群状态时空演化特征挖掘

Spatiotemporal Evolutionary Features of Urban Bridge Group State under Detection Data

摘要:

现有研究大多集中在单座桥梁的性能退化趋势分析上，较少关注由众多中小跨径桥梁构成的桥梁群体。针对区域桥梁群，构建了一个融合检测报告等多源数据的评估演绎数据库，为后续的特征提取、预测建模和网络级评估提供了坚实的数据基础。从区域层面对桥梁的跨径、桥长及技术状况分布特征进行了分析，并提出了基于机器学习的区域桥梁技术状态时空推演方法，揭示了区域桥梁群的状态特征，实现了对未来时间和未监测区域的桥梁结构状态进行预测和推演。

Abstract:

Existing research mostly focuses on analyzing the performance degradation trends of a single bridge. Less attention is paid to the bridge group composed of many small and medium span bridges. Aiming at the regional bridge group, an evaluation and inference database integrating multi-source data such as detection reports is built to provide a solid data foundation for subsequent feature extraction, predictive modeling and network-level assessment. The span, length and technical condition distribution characteristics of bridges are analyzed from the regional perspective. A spatiotemporal inference method for regional bridge technology state based on machine learning is proposed. The state characteristics of regional bridge group are revealed. The prediction and inference of the structural condition of bridges in future time periods and non-monitored areas are realized.