Prediction and determination of effective parameters of local energy loss in culvert systems, using intelligent evolutionary Algorithm

A culvert is a hydraulically short conduit which conveys stream flow through a roadway or embankment. Culverts are available in many different shapes and in spite of their simple structure, the process of designing requires accurate calculations and they are controlled by many variables. Accurate estimation of the energy loss in culverts plays an important role in optimum designing of culvert systems. In this paper, in order to simulate the local energy loss coefficient in culverts with various geometries (circular and rectangular), several models with different inputs were defined. The optioned results proved the capability of GEP in predicting local energy loss coefficient in culvert systems. Also it was found that in order to estimate the bend loss coefficient, model with input parameters of (Fr, θ) and to estimate the entrance loss coefficient, models with input parameters of (Fr, D Hw ) led to more accurate outcomes. Sensitivity analyses showed that θ and D Hw had the key role in estimating bend loss and entrance loss coefficient.