AI TRAINING MODEL FOR WATER LEAKS PREDICTION IN UNDERGROUND NETWORK | ||
| Journal of Al-Azhar University Engineering Sector | ||
| Articles in Press, Corrected Proof, Available Online from 23 October 2025 PDF (473.47 K) | ||
| Document Type: Original Article | ||
| DOI: 10.21608/auej.2025.370367.1801 | ||
| Authors | ||
| Osama M Wahba* 1; Aeizaal A. Abdul Wahab1; Ayman T. El-Faramawy2 | ||
| 1Department of Electrical Engineering, University Sains Malaysia (USM), Malaysia | ||
| 2Earth and Space Science and Engineering, York University, Ontario, Canada | ||
| Abstract | ||
| Water supply shortages pose a significant global challenge, necessitating either an increase in water production capacity or the optimization of water utility systems to minimize losses. Efficient management of water resources is critical to meeting the escalating demand for high-quality water. Meeting the growing demands on quality water resources for people requires strategic interventions actions by governments and the private sector entities to transform water management into essential practices to preserve and control water storage and distribution facilities. This research provides a "Water Leaks Detection and Prediction model" created to enhance the efficiency of water distribution systems by identifying and mitigating leaks. The proposed model integrates advanced analytical techniques to achieve high accuracy while maintaining cost-effectiveness, offering a practical solution for sustainable water resource management. The research demonstrates the application of specific water leakage factors used as input to an Artificial Neural Network (ANN) technique to create an improved and more efficient water leakage prediction and detection model. The developed ANN technique improves the prediction levels of leakage probabilities for underground water networks and enhances the system's predictive capabilities. The prediction framework operates through a structured training process, where predefined threshold values are integrated with input parameters to generate reliable output predictions. This paper details the data acquisition, model training process, and validation of the final predictive outcomes, demonstrating the ANN model’s effectiveness in enhancing water network monitoring and management. | ||
| Keywords | ||
| Leak detection; Prediction model; Water distribution efficiency; Underground Networks; Smart water networks | ||
|
Statistics Article View: 2 |
||
