زهرا يزدان پرست

The present study has developed a simulation-optimization framework for integrated management of the watershed-reservoir system based on an alternative model in the Seymareh watershed. In this regard, an integrated model has been presented for simulating an‎d eva‎luating strategies for protecting reservoir water quality, controlling upstream point an‎d non-point pollution sources, an‎d optimizing reservoir utilization. In order to model surface runoff an‎d pollutant load transfer at the watershed level, the SWAT model was used, an‎d the CE-QUAL-W2 numerical model was used to simulate reservoir water quality. Also, the Multi-Objective Particle Swarm Intelligence Optimization (MOPSO) algorithm has been linked as an optimization algorithm with hydrological an‎d qualitative simulation models. Given the high computational cost of the CE-QUAL-W2 model, alternative data mining-based models such as artificial neural networks (ANN) an‎d support vector machines (SVR) were used to reduce the execution time. The performance comparison of these methods showed that the support vector machine (SVR) model had higher accuracy an‎d reduced the computational volume by 85.71%. In the simulation-optimization framework based on the proposed alternative model, the decision variables at the watershed level included the percentage of point pollution treatment, the dimensions of filter strips, the implementation of retention ponds, the percentage of reduction in fertilizer an‎d pesticide use in agricultural lan‎ds, an‎d the selec‎tion of critical sub-basins. In the reservoir, the key decision variable was the monthly flow release rate over a 16-year period to examine its effects on water quality an‎d optimal reservoir utilization. The optimization results showed that the implementation of the optimal management scenario, compared to the stan‎dard exploitation policy, led to an improvement in the nutrient orientation index (TSI) by 2.2%. Also, hydroelectric energy production increased by 12.4%. In the economic sector, optimization policies have improved water quality an‎d in some cases increased energy production at the cost of increased management costs. These results demonstrate the positive impact of integrating numerical models an‎d optimization methods in the combined management of the watershed an‎d reservoir. The proposed framework provides a practical solution for sustainable water resources management by improving water quality, reducing management costs, an‎d increasing energy production efficiency.

مدل جايگزين , كنترل آلودگي‌هاي نقطه‌اي و غيرنقطه‌اي , الگوريتم چند هدفه هوش تجمعي ذرات , ANN , SVR , SWAT , CE-QUAL-W2

Alternative model , Point an‎d non-point pollution control , Multi-objective particle aggregation intelligence algorithm , ANN , SVR , SAWT , CE-QUAL-W2

Zahra Yazdanparast

Dr Motahare Saadatpour