Application Of Intelligent Algorithms In Providing Optimal Unit Hydrograph Using Probability Distribution Functions

The most important steps that were taken in hydrological analysis and design Hydrograph preparation, was Unit Hydrograph concept. The Unit Hydrograph was used to determine the flood, caused by the storms of duration and different intensity. The purpose of this study was the use of intelligent algorithms to provide optimal model Unit Hydrograph using the probability distribution function. For this purpose, the log-normal probability distribution function, gamma and inverse Gaussian was used. In this model, the objective function was to minimize the Sum of squared differences between the observed and predicted runoff Hydrograph at the catchment area. Computed runoff Hydrograph estimated using the proposed model by the probability distribution function. According to the values of root mean square error (RMSE), correlation coefficient (R2) and the Nash-Sutcliffe coefficient (NS), respectively 0.06,0.96,0.96 had log-normal distribution function better performance than other distribution functions at optimized Hydrograph. This distribution function had a good performance in computing peak flow. So that the calculated peak flow was near observed runoff peak flow. Also, genetic algorithms and particle swarm intelligence showed better results than other algorithms in the calibration of probability distribution functions.