نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Side weirs play a pivotal role in flow regulation and water measurement, serving as essential components in flood management systems, urban runoff collection, and irrigation and drainage networks. Accurate prediction of the discharge coefficient is critical for the optimal design of these structures. In this study, to enhance the modeling accuracy of the discharge coefficient in sharp-crested trapezoidal, broad-crested trapezoidal, and sharp-crested triangular side weirs, Support Vector Machine (SVM) models were hybridized with two metaheuristic algorithms: Horse Optimization Algorithm (HOA) and Reptile Search Algorithm (RSA). After identifying the key dimensionless variables, multiple models were developed for each weir geometry. The results indicate that both hybrid models exhibit strong predictive capabilities; however, SVM-HOA consistently outperformed SVM-RSA in the testing phase. Specifically, for sharp-crested trapezoidal, broad-crested trapezoidal, and sharp-crested triangular weirs, the SVM-HOA model achieved NSE, RMSE, and R values of (0.913, 0.019, 0.959), (0.933, 0.014, 0.966), and (0.932, 0.015, 0.967), respectively, demonstrating its superior performance across all geometries.
کلیدواژهها English
