Application of Artificial Neural Networks in Estimation of Scour Depth around the Bridge Pier with Sticky Sediments

Abstract

Most of the bridges breakdown is due to scour around the foundation of bridge during the floodwater. Accordingly, an improved model is needed to estimate the depth of scour around the bridge pier, for minimizing the possibility of failure. Since the scour in the bridge pier is a complex function of floor materials specification, fluid characteristics, flow specification and basic geometry, thus experimental equations cannot estimate the scour depth accurately. In this paper, artificial intelligence approaches have been used to estimate scour depth. In order to evaluate the performance of the mentioned method, laboratory data has been used in two cases with dimension and dimensionless data.First, an appropriate artificial neural network model is proposed and then optimized using the genetic algorithm. The mean correlation coefficient for dimensionless data were 0.97 at the testing stage. In this model, the neural network optimized by genetic algorithm, the root mean square error value, is 0.094. The results show that the recursive artificial neural network and genetic algorithm has a much better performance in estimation of scour depth around the bridge pier in substrate with sticky sediments in comparison to experimental equations.