We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Comparative uncertainty analysis of soft computing models predicting scour depth downstream of grade-control structures.
- Authors
Seyedian, Seyed Morteza; Riahi-Madvar, Hossien; Fatabadi, Abolhasan; Farasati, Masoomeh; Ghaznavi, Somayeh
- Abstract
The main focus of the present study is to develop GEP, ELM, LSSVM, and GMDH soft computing models predicting the scour depth of GCSs enhanced by an energy-based approach and to compare the uncertainty of these developed models. A new definition of energy parameter was employed to achieve more accurate predictions along with a scenario-based approach for input vector selection. The Monte Carlo framework was developed for the uncertainty, reliability, and resiliency analysis of the results. The results confirm the enhancement of the energy-based GEP method for the scour depth prediction in comparison with the other models. The results of robustness evaluations show the lowest uncertainty value of GEP with reliability 52.63% and resiliency 67.57% in comparison with LSSVM (reliability = 35.53%, resiliency = 40%), ELM (reliability = 34.2%, resiliency = 29.4%), and GMDH (reliability = 26.3%, resiliency = 24.6%) in the testing phase. In addition to the simplicity of the extracted equation and adaptability to a wide range of laboratory and field data, the uncertainty and robustness analysis indicate that the proposed model is more efficient than the existing equations and artificial intelligence models such as ELM, GMDH, and LSSVM in predicting the scour depth. It can be concluded that the GEP is more reliable and resilient than the other methods. The developed uncertainty analysis framework in this paper is a new approach reliably predicting the scour depth and extracting an equation that can be combined with mathematical modeling of sediment transport and scour hole geometry predictions or real-case modeling of scour around hydraulic structures.
- Publication
Arabian Journal of Geosciences, 2022, Vol 15, Issue 5, p1
- ISSN
1866-7511
- Publication type
Article
- DOI
10.1007/s12517-022-09704-0