Particle swarm optimization-based LS-SVM for hydraulic performance of stepped spillway.

Autor: Roushangar, Kiyoumars, Akhgar, Samira
Předmět:
Zdroj: ISH Journal of Hydraulic Engineering; Aug2020, Vol. 26 Issue 3, p273-282, 10p
Abstrakt: Accurate estimation of energy dissipation and discharge coeffichine (LS-SVM) to recognize the factors affecting the discharge coefficient and energy dissipation. To this end, all data was extracted from physical models and original experiments. Particle swarm optimization (PSO) was also applied to optimize the hyper-parameters of the LS-SVM model. The results indicated that the applied methods had a high capability in modeling energy dissipation and discharge coefficient of stepped spillways. Additionally, the results of sensitivity analysis of both proposed methods showed that the Fr1 and the W/y1 parameters had the highest and the lowest important influence in modeling of the discharge coefficient, respectively. Besides, the role of y c N h parameter was very highlighted; F r 1 and B y 1 also had the lowest impact on estimating energy dissipation over stepped spillways. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index