Application of machine learning methods on real bridge monitoring data

In this article, the non-linear or rather transient relationship between the air temperature and the bridge temperature is simulated by machine learning (ML) models. Based on this ML-modeling, different use cases for the application of machine learning regression methods to monitoring data are presented. The focus of the paper is to present different use cases for an already established ML method in order to show the wide range of applications of such methods. It is shown, that these methods can be used to detect and compensate sensor faults or to forecast the behavior of structures. The results show that these methods, have a great potential for the evaluation of large amounts of data since no physical models are required. For the calculations, long-term monitoring data of valley bridges from the German high-speed railroad line VDE 8 are used. •Machine learning methods are used to predict the structural behavior of bridges.•Use cases of machine learning methods on bridge measurement data are presented.•Nonlinear transient processes can be linearized due to data transformation.•Trained machine learning models are transferred to similar structures/bridges.•Sensor faults can be detected and compensated with machine learning.