Uncertainty assessment of kernel based approaches on scour depth modeling in downstream of ski-jump bucket spillways

Autor: Redvan Ghasemlounia, Seyed Mahdi Saghebian
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: Water Supply, Vol 21, Iss 5, Pp 2333-2346 (2021)
Druh dokumentu: article
ISSN: 1606-9749
1607-0798
DOI: 10.2166/ws.2021.063
Popis: From the hydraulic structures designer's point of view, the scour depth accurate estimation in downstream of spillways is so important. In this study, the scour depth was assessed downstream of ski-jump bucket spillways using two kernel based approaches namely Support Vector Machine (SVM) and Kernel Extreme Learning Machine (KELM). In the model developing process, two states were tested and the impacts of dimensional and non-dimensional parameters on model efficiency were assessed. The best applied model dependability was investigated via Monte Carlo uncertainty analysis. In addition, the model accuracy was compared with some available semi-theoretical formulas. It was observed that the applied models were more successful than available formulas. The sensitivity analysis results showed that q (unit discharge of spillway) variable in the state 1 and q2/[gYt3] (g is gravity acceleration and Yt is tail water depth) variable in the state 2 were the most significant parameters in the modeling process. Comparison among applied methods and one other intelligence approach showed that KELM was more successful in predicting process. The obtained results from uncertainty analysis indicated that the KELM model had an allowable degree of uncertainty in the scour depth modeling. HIGHLIGHTS The capability of two kernel based models (i.e. SVM, KELM) was investigated for scour depth assessing downstream of ski-jump bucket spillways.; The capability of applied methods was compared with some available semi-empirical equations.; The most important parameters were determined using sensitivity analysis.; Monte Carlo uncertainty analysis was applied to investigate the dependability of the applied models.;
Databáze: Directory of Open Access Journals
Externí odkaz: