Modeling of Temperature Effect on Modal Frequency of Concrete Beam Based on Field Monitoring Data

Abstract

Temperature variation has been widely demonstrated to produce significant effect on modal frequencies that even exceed the effect of actual damage. In order to eliminate the temperature effect on modal frequency, an effective method is to construct quantitative models which accurately predict the modal frequency corresponding to temperature variation. In this paper, principal component analysis (PCA) is conducted on the temperatures taken from all embedded thermocouples for extracting input parameters of regression models. Three regression-based numerical models using multiple linear regression (MLR), back-propagation neural network (BPNN), and support vector regression (SVR) techniques are constructed to capture the relationships between modal frequencies and temperature distributions from measurements of a concrete beam during a period of forty days of monitoring. A comparison with respect to the performance of various optimally configured regression models has been performed on measurement data. Results indicate that the SVR exhibits a better reproduction and prediction capability than BPNN and MLR models for predicting the modal frequencies with respect to nonuniformly distributed temperatures. It is succeeded that temperature effects on modal frequencies can be effectively eliminated based on the optimally formulated SVR model.

Eliminating environmental and operational effects on structural modal frequency: A comprehensive review

“…Based on the structural risk minimization principal, SVR model was introduced to improve the prediction of structural modal frequency under temperature change, and more accurate frequency variations can be achieved compared with MLR and BPNN. 38,47,48 To quantify the nonlinearity and uncertainty of modal frequency, Gaussian process regression (GPR) and relevance vector machine (RVM) based on the Bayesian inference have been applied, which frames the frequency-temperature relation with probability prediction and confidence intervals. 49,50 In practice, the optimal kernel parameters of machine learning methods should be adaptively searched using the optimization algorithm for robust frequency estimations.…”

Reducing the Training Samples for Damage Detection of Existing Buildings through Self-Space Approximation Techniques

“…This time lag can also interest the EOVs effect on the system vibrational properties [33,60,61]. Indeed, EOVs within a single element or among different elements are rather non-uniform and non-linear, especially in large-scale structures [12,35,62]. A recommended solution would be to perform the feature selection considering different measurement points and testing the correlation with properly averaged values.…”

Validation of vibration-based structural health monitoring on retrofitted railway bridge KW51

“…The simplest approach to determine an input-output model that allows removing the environmental and operational conditions is linear regression [3,7]. Alternative approaches that do not assume a linear relation between the inputs and the outputs but allow identifying a global non-linear input-output mapping are based on neural networks [8] and support vector machines [4,9]. Inputoutput modeling requires that all relevant environmental and operational variables are measured.…”