نوع مقاله : مقاله کامل پژوهشی
نویسندگان
1 پژوهشگر پسا دکتری دانشکده مهندسی و علوم محیط زیست، دانشگاه نانکای، چین
2 دانشکده مهندسی عمران، دانشگاه تبریز
3 دانشکده مهندسی عمران دانشگاه تبریز
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
To optimize the management of dam reservoir operations, accurately forecasting river inflow for the upcoming months using artificial intelligence models and metaheuristic algorithms is essential. The first phase of this study aimed to predict the next year's inflow to the Alavian Dam reservoir using an artificial neural network model. Given that optimal reservoir operation is one of the most critical management factors during the operation period, in the second phase of this research, titled "Dynamic System Modeling," the Vensim model was utilized to simulate system behavior using the predicted runoff and actual demands. In the third phase, a combination of optimization algorithms (genetic algorithm and particle swarm optimization) was employed to optimize the operation of the Alavian Dam reservoir. The comparison of results indicates that the prediction phase has shown satisfactory accuracy. To evaluate the performance of the examined algorithms in optimal reservoir operation, reservoir performance indices were used. In the short-term analysis, the proposed hybrid algorithm achieved a volumetric reliability index of 72% for the scenario with 100% agricultural demand fulfillment and 94% for the scenario with 80% agricultural demand fulfillment, while the Vensim model showed a volumetric reliability index of 75% and 83% for the respective scenarios. Therefore, using the hybrid algorithm, release rules and reservoir volume curves for the next 12 months were prepared and presented based on the predicted inflows.
کلیدواژه‌ها [English]
)