Automated Simulation of Basin Characteristics Using HEC-HMS, Genetic Algorithm, and AutoIt on Observed Hydrograph Properties

The main objective in calibration of conceptual Rainfall- Runoff models is to find a set of optimal model parameters that provide the best fitness between the observed and the estimated flow hydrographs. These models are general, but when there is a lack of data, their application will be too difficult. In basins without data using the synthetic unit hydrograph for calculation of design flood is inevitable. Recently, it is shown that the use of the heuristic algorithms in combinational optimization problems is very suitable. In this research, two different algorithms, Univariate-Gradient as a classic optimizer, and Genetic Algorithm as a throughout optimizer, were used for calibrating snyder, Clark, and SCS unit hydrograph models in HEC-HMS software. The flood of Bahman 14-18, 1371 (Shamsi Calendar) in Dejgan station, Mehran River, Hormozgan Province, was used in this study. Results showed that the combination of GA and Snyder method is appropriate for forecasting the basin characteristics. The basin characteristic can be obtained using this method, observed hydrograph, and fitness function. “AutoIt” software was used for automated running of simulation and optimization.