Study of liquid flow-solid flow and sediment concentration sediment relationship using artificial intelligence methods.case of Kébir-Rhumel watershed.
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Université d'Oum El Bouaghi
Abstract
Predicting extreme concentrations of suspended sediments in arid and semi-arid regions is crucial for effective water resource management. However, the lack of a universally accepted analytical method for calculating sediment discharge or concentration in natural channels presents a significant challenge. To address this gap, our study employs Principal Component Analysis (PCA) and Wilcoxon signed-rank tests at stations (El Ancer, Athmania, and Grarem). Additionally, our research integrates seven artificial intelligence models: Kstar, LazyIBK, LazyIBKLG, RandomForest, RandomTree, RandomCommittee, and RandomizableFilteredClassifier, aimed at enhancing the accuracy of sediment discharge predictions. These advanced techniques demonstrate considerable potential in refining estimates and providing robust tools for water resource managers. Notably, RandomForest and RandomCommittee models exhibit superior precision and reliability, often surpassing conventional methods. By integrating these AI models, our research contributes to more efficient and sustainable management practices in arid and semi-arid environments. This approach enhances our ability to anticipate extreme sedimentation events, thereby improving water resource planning and management strategies
Description
Keywords
Intelegence artificial; Prediction; Transport solid; Water source management