Leak detection and leak localization in a smart water management system using computational fluid dynamics (CFD) and deep learning (DL)

This research project investigates the detection and prediction of leaks in a smart water management system using a combination of Computational Fluid Dynamics (CFD) and Deep Learning (DL) techniques. The study focuses on simulating the flow of water in pipes of varying materials, including copper and steel, with different diameters. The pipes are subjected to controlled conditions, including the presence of holes at specific locations (2 m, 4 m, 6 m, and 8 m) along the length of a 10-m pipe, as well as scenarios with no leaks. To predict the leak and its position on the pipe, different deep learning models are trained and tested on the dataset generated on simulating the water flow using CFD software’s such as, Solid Works for designing the pipe model and Ansys Workbench for simulating the flow of water in the pipe and generating the data. A comparison of these models has been made to obtain the best performing model for both leak prediction and leak localization. The research outcomes hold significant implications for enhancing the efficiency and reliability of leak detection systems in smart water management infrastructures, ultimately contributing to the conservation of water resources and the prevention of potential infrastructure damage.