Affiliation:
1. Sakarya Applied Sciences University
Abstract
Abstract
The cities should improve their sustainability to meet the many objectives outlined in the sustainable development goals. For this purpose, water leakages directly affect consumers' and water companies' financial and environmentally sustainable performance, and water leakages are essential factors in drinking water. The leakage detection techniques are unpractical and inaccurate, using the traditional leakage-detection method. This study proposes that the Adaptive Neuro-Fuzzy Inference System (ANFIS) is modeled for the leakage estimation for two different districted metered areas of Sakarya in Turkey. Three different input data, minimum flow ratio, maximum flow ratio, and average flow ratio in the range of [15.4, 29.2], [41.4, 61], and [31.1, 45.2], respectively, have been determined as the input data for the ANFIS. The output of the ANFIS model is used as the leakage ratio (%) parameter. The required data for the training (174 data) and testing (173 data) processes have been gathered from the experimental study. Some parameters which affect the ANFIS training performance, such as the number of membership functions and training cycle, are investigated for different simulation cases. Consequently, it is proven that the ANFIS has a very high prediction performance of water leakage with R2 0.994 and MSE 4.63. Moreover, the performance of the ANFIS has been compared with that of the artificial neural network (ANN) in the case of leakage detection, and it has been shown that the training and testing performance of the ANFIS is better than that ANN with a ratio of 13.6% and 17.02%, respectively.
Publisher
Research Square Platform LLC
Reference43 articles.
1. Wu ZY, Farley M, Turtle D, Kapelan Z, Boxall J, Mounce S, Dahasahasra S, Mulay M, Kleiner Y (2011) Water loss reduction,Bentley Institute Press,
2. Performance optimization of a leak detection scheme for water distribution networks, IFAC-PapersOnLine;Przystałka P,2018
3. Robust leak localization in water distribution networks using computational intelligence;Quiñones-Grueiro M;Neurocomputing,2021
4. The manager’s non-revenue water handbook, A Guid;Farley M,2008
5. Leak diagnosis in pipelines using a combined artificial neural network approach;Pérez-Pérez EJ;Control Eng Pract,2021