Input attributes optimization using the feasibility of genetic nature inspired algorithm: Application of river flow...

Area of Science:

• Hydrology and Water Resources Engineering
• Computational Intelligence
• Time Series Analysis

Background:

• Streamflow patterns exhibit significant non-linearity and non-stationarity, posing challenges for accurate forecasting.
• Reliable streamflow forecasting is crucial for effective water resource management and engineering applications.
• Optimizing input variables is a key factor in enhancing the accuracy and reliability of hydrological models.

Purpose of the Study:

• To develop and evaluate an algorithm for selecting optimal input combinations for streamflow forecasting models.
• To integrate a Genetic Algorithm (GA) with a Radial Basis Function Neural Network (RBFNN) for enhanced streamflow prediction.
• To assess the performance of the proposed RBFNN-GA model in forecasting monthly streamflow at the High Aswan Dam on the Nile River.

Main Methods:

• Utilized the Genetic Algorithm (GA) for efficient selection of optimal input variables for time series forecasting.
• Employed the Radial Basis Function Neural Network (RBFNN) for its simplicity and effectiveness in modeling complex hydrological data.
• Integrated the GA with the RBFNN (RBFNN-GA) to create a hybrid model for monthly streamflow forecasting.

Main Results:

• The integrated RBFNN-GA model demonstrated high accuracy in forecasting monthly streamflow.
• The Genetic Algorithm effectively identified optimal input parameters, contributing to improved model performance.
• The proposed method proved successful in predicting streamflow at a critical location, the High Aswan Dam on the Nile River.

Conclusions:

• The GA is a powerful tool for optimizing input selection in streamflow time series forecasting.
• The RBFNN-GA hybrid model offers a robust and accurate approach for hydrological forecasting.
• This study highlights the potential of advanced computational methods in addressing challenges in water resource engineering.