根据模态频率变化评估桥梁性能是结构健康监测(Structural Health Monitoring, SHM)的常用方法之一。然而实测结果表明，由温度、交通荷载等运营环境作用引起的频率正常变化比由结构轻微损伤或退化引起的频率异常变化更大。为正确确定频率的异常变化，有必要研究桥梁运营环境作用对频率变化的影响规律，以便将频率的正常变化从总变化中分离出去。利用长达6年的实测数据，本项目从周期性的角度研究桥梁正常运营时作用与频率变化的相关性，比较作用的影响大小；通过理论推导、数值分析、模型实验研究作用对频率变化的影响机理，解释实测结果；结合多元回归分析和统计过程控制，提出基于“作用-频率”关系监测的结构异常判别方法；对比公路桥与铁路桥频率随作用变化规律的异同。研究成果能使根据模态频率变化评估桥梁性能真正成为可能。

The modal frequencies are one of commonly used indices to evaluate bridge performance in Structural Health Monitoring (SHM) field. However, it is found that the measured frequency variation induced by bridges’ operational and environmental actions such as temperature, traffic loading, etc. usually exceeds the frequency variation due to small and medium structural anomalies. Therefore, it is necessary to investigate the influence of bridge actions on modal frequencies for the separation of the action-induced frequency variation so as to highlight the frequency changes caused by the anomalies. Based on the 6-year monitoring data, this study explores the periodic characteristics of the evolutions of structural frequencies as well as bridges' operational and environmental actions, revealing the dominant actions that cause frequency variation. The mechanisms of the frequency change with individual actions are analyzed through theoretical analysis, numerical simulation and model experiment to explain the measured results. According to the action-frequency relationship, a structural anomaly identification method is proposed which involves the combination of multiple regression analysis and statistical process control. Moreover, the railway bridges are compared with the highway bridges in terms of the measured frequency variation with actions. This study can lead to a more reliable evaluation of bridge performance based on modal frequency variation.