부산대학교 도서관

Estimating and predicting groundwater quality characteristics so that managers may make management decisions is one of the critical goals of water resource planners and managers. The complexity of groundwater networks makes it difficult to predict either the time or the location of groundwater. Many models have been created in this area, offering better management to preserve water quality. Most of these models call for input parameters that are either seldom accessible or expensively and laboriously measured. A better option among them is the Artificial Neural Network (ANN) Model, which draws inspiration from the human brain. This study uses Na + , Mg2 + , Ca2 + , Na%, K + , SO42 − , Cl − , pH, and HCO3 − quality parameters to estimate the Sodium Adsorption Ratio (SAR). The Shiraz Plain's groundwater quality was simulated using four optimization meta-heuristic methods, including biography-based optimization (BBO), black hole attack (BHA), sequential forward selection (SFS), and multi-verse optimization (MVO). These methods excel in adaptability, convergence speed, feature selection, diversity of solutions, and robustness to complex and noisy datasets, ultimately leading to more accurate and efficient predictive models than earlier methods. A statistical period of 16 years (2002–2018) was used to collect the groundwater quality data for the Shiraz plain to accomplish this purpose. The findings showed that the SFS-MLP approach was more accurate than the other methods with training and testing dataset values of R2 = 0.9996 and 0.99923, RMSE = 0.04929 and 0.072, and MAE = 0.039357 and 0.048968, respectively. Additionally, the findings demonstrated that the SFS-MLP approach has a high capacity and accuracy for predicting and modeling groundwater quality. This study's findings also show that intelligence models and optimization algorithms may be used to mimic groundwater quality parameters effectively.