Volume 17, Issue 2 (Summer 2023)
2023, 17(2): 228-252 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Azari T, Dadashi S, Kardel F. Prediction of Chloride in Groundwater Using Supervised Artificial Intelligence Committee Machine (SAICM) (Case Study: Sari Plain). Journal of Engineering Geology 2023; 17 (2) :228-252
URL: http://jeg.khu.ac.ir/article-1-3089-en.html
URL: http://jeg.khu.ac.ir/article-1-3089-en.html
1- 1University of Kharazmi
2- University of Mazandara
3- University of Mazandara , f.kardel@umz.ac.ir
2- University of Mazandara
3- University of Mazandara , f.kardel@umz.ac.ir
Abstract: (32 Views)
Qualitative assessment of coastal waters affected by seawater salinity can be done using the parameter of chloride in groundwater. This research proposes a supervised artificial intelligence committee machine (SAICM) method for accurate prediction of chloride concentration in groundwater of Sari plain. SAICM predicts chloride concentration as the output of the model by non-linear combination of artificial intelligence models. In this research, Principal Component Analysis (PCA) method was used to identify effective hydrochemical parameters related to chloride concentration as input components to artificial intelligence models. Based on the results of PCA, parameters (Na, K, EC, TDS, SAR) were selected as input components of artificial intelligence models. Firstly, four artificial intelligence models, Sogno fuzzy logic, Mamdani fuzzy logic, Larsen fuzzy logic and artificial neural network were designed to predict chloride concentration. Based on the modelling results, all the models showed a good fit with the chloride data in Sari Plain. Then, the combined SAICM model was built, which combines the prediction results of 4 separate AI models using the nonlinear ANN combiner and determines the chloride concentration more accurately. The results show that the proposed SAICM can estimate chloride concentration with much higher accuracy than individual methods.
Keywords: artificial neural network, committee machine, fuzzy logic, ground water quality, principal component analysis.
Type of Study: Original Research |
Subject:
Hydrogeology
Received: 2023/07/30 | Accepted: 2023/09/20 | Published: 2023/09/21
Received: 2023/07/30 | Accepted: 2023/09/20 | Published: 2023/09/21
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
