Home J!Research Publications Novo

Prediction of Dam Behaviour using Multiple Linear Regression and Radial Basis Function Neural Network

Research Area: Uncategorized Year: 2011
Type of Publication: In Proceedings Keywords: Dam Safety, Multiple Linear Regression, Radial Basis Function Neural Networks, Dam Behaviour, Modeling
  • Rankovic, Vesna
  • Grujovic, Nenad
  • Milanovic, Goran
  • Divac, Dejan
  • Milivojevic, Nikola
Editor: Gordana Globočki-Lakić
Book title: Proceedings of the 10th Anniversary International Conference on Accomplishments in Electrical, Mechanical Engineering and Information Technology
Pages: 179-185
Organization: University of Banja Luka, Faculty of Mechanical Engineering Month: May
ISBN: 978-99938-39-36-1
The safety control of dam is supported by monitoring activities and is based on mathematical models. The variations of hydrostatic pressure and temperature are the main variables that should be taken into account when analyzing the results of the concrete dam observations. Deterministic models based on mechanical principles are often difficult to construct. Statistical procedures, such as multiple linear regression (MLR), have been applied to dam safety to determine the influence of external loads on the structure deformation. The relations between the loads and dam behavior are nonlinear. Radial basis function (RBF) neural network can be successfully applied to function approximation, forecast, and dynamic systems identification. Neural network modeling from measured data is effective tool for the approximation of uncertain nonlinear systems. This paper presents novel approach based on the use of RBF network to estimate dam behaviour. Mathematical models based on experimental data are developed. MLR and RBF neural network models for prediction of dam behaviour have been compared with the measured data on the basis of correlation coefficient.