SUMMARY
The study of the eighteen-year monthly operating budget of the Ghrib Reservoir from 1999 to 2016 shows an important deficit in satisfying water demand partly due to the uncertainty of stochastic water inflows. In this context, the main objective of the present research work is to overcome this shortcoming by reducing it and improving the current reservoir operation. Therefore, the optimization of this last using a genetic algorithm (GA) is carried out based on historical and simulated water inflows. The performance of the classical GA in optimizing the multipurpose reservoir operation based on the simulated water inflows using the principal component analysis (PCA) model is highly demonstrated given a significant decrease in the water deficit from 47% to 8%. The developed model could help operators make decisions for operating dam reservoirs more efficiently.