Temperature Prediction of Flat Steel Box Girders of Long-Span Bridges Utilizing In Situ Environmental Parameters and Machine Learning

Research on Deformation Analysis and Rehabilitation for a Beam–Arch Combination Bridge Suffering an Extreme Temperature Field

“…Regarding the action of the temperature gradient, the specifications of different countries regulate the form of the vertical temperature gradient curve of the To assess the effective temperature (ET) action, most bridge design codes, such as the Eurocode [25], British standard [26], AASHTO standard [27], and Chinese standard [28], recommend an indirect method, considering a conversion between air temperature and bridge temperature. Regarding the action of the temperature gradient, the specifications of different countries regulate the form of the vertical temperature gradient curve of the main girder [29]. Since the bridge was designed according to the Chinese standard [28], the deformation of temperature action was analyzed according this standard in the FE model simulation.…”

Structural temperature gradient evaluation based on bridge monitoring data extended by historical meteorological data

“…To train and validate the neural network model, the STG and meteorological parameter dataset needs to be divided into training set and test set. Due to the different mapping relationships between input and output under different meteorological conditions (such as different seasons, sunny, and rainy conditions), the correlation between meteorological parameters and STG is different 28 ; therefore, it is necessary to make the training process cover all the meteorological situations. However, the structural temperature data currently monitored by the SHM system are only 1-2 years or less, dividing the training sets and test sets chronologically or randomly may result in some data for different meteorological conditions not being effectively trained.…”

“…A lot of research tried to construct a relationship between STD and environmental factors, [23][24][25][26][27] but the following are main shortcomings of these studies for the longterm prediction of the temperature gradient: (i) Only the relationship between the maximum STD and the maximum temperature, SR, and wind speed (WS) were taken into full consideration, with other environmental factors acting on the actual bridge structure not being fully considered, and the linear regression relationship was not sufficient to meet the prediction accuracy required for bridge design; (ii) The relationship between the environmental parameters and the maximum STD of the cross-section was established, but the correlation between the STD variation of different monitoring points at different moments was ignored, which is difficult when modeling the STG; (iii) The current research is mostly based on experimental data or FEM simulations, and the results may be too idealized and less practical for actual bridges affected by environmental factors. Based on these problems, Wang et al 28 took environmental factors into consideration, adding humidity, WS, and wind direction based on SHM system monitoring, but was still reliant on long-term monitoring by the SHM system.…”

Calculation of Temperature Action of Flat Steel Box Girder of Long-Span Bridges Using a Joint Model of ARMA Mean and GARCH Variance

“…Because fat steel box girders are becoming more and more common in large-span bridge structures, it is important to study the standard values of temperature action for fat steel box girders [1][2][3][4][5]. However, current bridge researchers still have insufcient understanding of the standard values of temperature action in fat steel box girders [6][7][8][9][10][11][12]. Te China bridge design codes also do not specify the calculation method of standard values of temperature action in fat steel box girders.…”