Application of simulated annealing algorithm in calibration of area reduction method in sediment distribution of dams reservoir (Case Study: Karaj Dam)

Area reduction method is one of the most common empirical methods to determine sediment distribution in dam reservoirs. In this method reservoirs were classified into four types that for each of them the parameters c, m and n are presented. Calibration of method for a reservoir based on these parameters can increase its accuracy. In this research, two computer programs in FORTRAN language were prepared based on SA optimization algorithms and area reduction empirical method. Then, a simulation-optimization model was obtained by combination of two mentioned programs with ability to determine optimal parameters of area reduction method. Using this model, optimal parameters of area reduction method is determined so that the least difference exists between computational and actual volume. Prepared model was used for Karaj Dam with three periods of actual area-volume-height information. Using actual data from two first periods, optimal empirical area reduction parameters were obtained and capability of SA algorithm determining the global optimal was tested by asymptotic method. The optimal values of c, m and n were obtained 0.052, 0.836 and 14.234 respectively. In order to indicate the ability of optimal parameters to increase accuracy of prediction, sediment distribution in the third period of the actual data was calculated with typical and optimal parameters and the results were compared with actual values. Results showed that using optimal parameters decrease the amount of error in the prediction of sediment distribution, significantly. Difference between estimated and actual volume of Karaj dam using optimal parameters in calibration and verification periods are 54% and 20% lower than the typical parameters respectively. Finally sediment distribution of reservoir was predicted for next years based on optimal parameters.