Prediction of concrete dam deformation using artificial neural networks

Abstract

The goal of this paper is to assess the effectiveness of using artificial neural networks in the prediction of concrete dam deformation. The aging of dams as a concrete structures poses a significant risk for the environment, and many of them are at the stage where it is necessary to pay attention to their behaviour. Short-term term prediction of deformation is very important for rapid response in the case of any adverse events. Two examples are presented in this paper to investigate how short-term prediction of deformation can be realized by using artificial neural networks. As a comparison, a set of statistical linear regression models are established using the same data. The conclusions on the basis of the analysis and the established models are presented. The paper provides an assessment of the network structure and a comparison of neural network models for the usual concept of statistical models for the monitoring of dams.