考虑运营环境不确定性的斜拉桥模态频率识别

为剔除运营环境因素（温度、车辆荷载、风等）对斜拉桥结构模态频率的影响，凸显因结构损伤引起的频率变化，对运营环境和模态频率之间的量化和传递进行研究。首先，考虑温度对材料弹性模量以及几何特性的影响，对温度影响模态频率的机理进行分析，以灌河大桥为工程背景，在较长时间尺度内对模态频率进行单因素回归分析；然后，根据匀速移动常量力作用模型推导移动荷载和结构振动强度的关系，考虑到车辆荷载及风荷载在较短时间尺度内对桥梁结构作用存在较强的耦合关系，采用非线性主成分分析方法（NLPCA）对运营环境因素进行主要特征提取和冗余信息的剔除，使用人工神经网络（ANN）模型实现两者之间的量化和传递；最后，提出基于运营环境变量分析的模态参数修正方法。结果表明：在较长时间尺度内，温度和车辆荷载与斜拉桥模态频率均有明显的负线性相关性，40℃季节温差对灌河大桥模态频率的影响为1.3%；非强风期间，主梁加速度RMS可近似反映桥上车辆荷载的变化；在短时间尺度内，温度与车辆荷载2种因素影响水平相当，一天内对灌河大桥影响一般不超过1%；基于NLPCA的ANN回归模型能较好地实现灌河大桥环境因素与模态频率的传递，剔除运营环境因素影响后的斜拉桥模态频率变异性明显降低，主要包含某些随机误差，符合正态分布。

Modal Frequency Identification of Cable-stayed Bridges Considering Uncertainties of Operational Environmental Factors

Quantification and transfer between operating environment and modal frequency were studied to eliminate the influence of operating environment factors (temperature, vehicle load, wind, etc.) on the structural modal frequency of cable-stayed bridges. Further, the frequency changes caused by the structural damage were elucidated. First, considering the influence of temperature on the elastic modulus and geometric properties of the materials, the mechanism of the influence of temperature on the modal frequency was analyzed. Subsequently, the structural vibration intensity was derived based on a certain constant force model of the relationship between the moving loads. Considering the relatively strong coupling relationship between the vehicle load and wind load on the bridge structure on a short-term scale, nonlinear principal component analysis (NLPCA) was performed to extract the main features and eliminate the redundant information of the operational environment factors. Furthermore, an artificial neural network (ANN) model was used to realize the quantification and transmission between the operating environment factors and modal frequency. Finally, a modal parameter correction method based on operational environment variable analysis was proposed. The results show that, on a long-term scale, the temperature and vehicle load have a significantly negative linear correlation with the modal frequency of the cable-stayed bridge. Furthermore, the influence of a 40℃ seasonal temperature difference on the modal frequency of Guanhe Bridge is 1.3%. During the period of a non-strong wind, the RMS acceleration of the main beam can approximately reflect the change of vehicle load on the bridge. On a short-term scale, the influence of temperature and vehicle load is approximately the same, and the influence on Guanhe Bridge in one day is generally less than 1%. The ANN regression model based on NLPCA can effectively realize the transmission of environmental factors and modal frequency of Guanhe Bridge. The modal frequency variability of the cable-stayed bridge is significantly reduced after removing the influence of operational environmental factors. Furthermore, it mainly contains some random errors and is normally distributed.