article · Russian Meteorology and Hydrology
This research evaluates methods to enhance the accuracy and reliability of monthly streamflow predictions by combining artificial neural networks with meta-heuristic optimization techniques. Testing focused on the Fermatou and Bou Birek stations within the Soummam watershed in northern Algeria. Three optimization approaches were examined to refine the neural network parameters: Artificial Rabbits Optimization, Mayfly Optimization Algorithm, and a hybrid combining the Gray Wolf Optimizer with Particle Swarm Optimization. Statistical correlation functions, including autocorrelation, partial autocorrelation, and cross-correlation, were used to establish the optimal input combinations. The findings indicate that the neural network paired with Artificial Rabbits Optimization delivered the highest precision, achieving correlation coefficients between 0.981 and 0.982 and Nash Sutcliffe efficiency values from 0.960 to 0.962 across both monitoring stations, outperforming the alternative optimization methods tested.
Reliable streamflow forecasting is essential for water resource planning, flood prevention, and reservoir management. By demonstrating that the Artificial Rabbits Optimization algorithm enhances neural network performance over established techniques, this study offers hydrological practitioners a more accurate computational framework for anticipating monthly river flow variations in vulnerable watershed regions.
The model represents applied research tested on real-world watershed data, offering potential utility for hydrological services, water utility operators, and river basin managers seeking dependable forecasting tools. However, the abstract does not indicate a commercial application pathway or deployment into an operational software platform.
AI-generated from the published abstract. Always read the original work before citing.
Abstract The study aims to improve accuracy and reliability of streamflow forecasting through the optimization of artificial neural networks (ANNs) using three meta-heuristic algorithms: Artificial Rabbits Optimization (ARO), Mayfly Optimization Algorithm (MOA), and Gray Wolf Optimizer (GWO) coupled with Particle Swarm Optimization (PSO) (GWO-PSO). The study was conducted to predict monthly streamflow at the Fermatou and Bou Birek stations in the Soummam watershed, situated in the north of Algeria. Optimal inputs and parameter combinations for the hybrid ANN models were determined using the autocorrelation function (ACF), partial autocorrelation function (PACF), and cross-correlation function (XACF). The numerical results revealed the superior performance of the ANN-ARO, with correlation coefficients R and Nash–Sutcliffe efficiency ranging from 0.981 to 0.982 and from 0.960 to 0.962, respectively, for the two stations. These outcomes surpassed those achieved by the ANN-GWO-PSO and ANN-MOA. It should be pointed out that when comparing the new ARO method to existing employed meta-heuristic algorithms, it showed improved precision in results and prediction accuracy.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.3103/s1068373924600041
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
