Comparing the Performance of Genetic and Particle Swarm Algorithms in Calibration of Experimental Area Reduction Method Based on Hydrographic Results of Dez Dam Reservoir

Area-volume-elevation (AVE) curves of dams pose one of the most important tools for water resource planning and reservoir management. Area reduction method is one of the methods which was developed to modify these curves after reservoir sedimentation based on the conditions, statistics and information recorded in reservoirs abroad; however the application of the same methods for domestic dams without optimizing the coefficients is not bug-free, and associates sometimes with a great deal of error. This study aims to calibrate automatically the area reduction method based on its three influencing parameters using two genetic and particle swarm optimization algorithms and to compare the results with reservoir hydrographic studies. So that, predicting the trend of reservoir sedimentation in area reduction method with its hydrographic results at the end of the period has the least difference. Results indicate the excellence of particle swarm algorithm in calibrating the area reduction method. In this algorithm, by selecting the initial population of 50, the time of convergence and the value of the objective function (RMSE of the predicted volume-elevation curve and actual reservoir) in the last iteration was 4.7 minutes and 7 MCM, which it represents 92.6% and 48% improvement, respectively compared with the genetic algorithm. Finally, the optimal values of area reduction method parameters were used to match further the estimated and actual volume values of Dez dam reservoir. The results revealed that the predicted error value is less than 1%, which is assessed negligible according to the volume of the reservoir. Accordingly, the developed model can be employed without any change in the optimal parameters of the area reduction method by entering information of reservoir’s new hydrography to predict the sedimentation trend during the coming years. This will be very helpful in light of the importance of knowing the reservoir’s useful volume variation in the coming years and its role in future water resource planning.