Leakage detection and prediction of location in a smart water grid using SVM classification

A smart water grid should be capable of ensuring the 24*7 reliable water supplies with efficient distribution of water with minimum losses due to leakages in pipelines. Early Detection of water leakages is utmost important to save large amount of water from being wasted. This paper presents the SVM classification technique for detection of leakages and prediction of location in a water network. Besides the traditional methods of identifying a leakage which incurs a high cost but having a low efficiency, SVM classification technique can be used to classify the leakage and non-leakage scenarios using the pressure and flow dataset captured from the water distribution network. The dataset is obtained by simulating the water distribution network of CSIR-CEERI, Pilani into EPANET tool.