An Efficient scheme for Water Leakage Detection using Support Vector Machines (SVM) – Zig
Akshay Kothari1, M. Balamurugan2
1Akshay Kothari, M.Tech Student, Department of CSE, Christ Deemed to be University, Bengaluru (Karnataka), India.
2M. Balamurugan, Associate Professor, Department of CSE, Christ Deemed to be University, Bengaluru (Karnataka), India.
Manuscript received on 22 April 2019 | Revised Manuscript received on 04 May 2019 | Manuscript Published on 17 May 2019 | PP: 39-46 | Volume-7 Issue-6S4 April 2019 | Retrieval Number: F10080476S419/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Water is one of the most essential and valuable resources for all living beings, yet in the present day, there is a scarcity of it. Half of the water loss in large cities and industries is due to leaks and illegal lines. 10%-20% of water loss can be reduced by detecting leaks but without the presence of advanced monitoring systems, this problem is typically worsened. Monitoring the consumption and leak detection for such large areas is a challenging task. To overcome this issue a small prototype is prepared called Zig. Zig is designed for both household and industrial purposes. Its main aim is to monitor the flow and consumption of water at different levels of a building like a first-floor and so on which may represent some industrial and household situation. This work focuses on pressure/flow monitoring method to reduce the operational cost and also to detect leakage. One of the machine learning algorithms, Support Vector Machines (SVM) has been applied to detect the leakage and it is compared with Random Forest algorithm to show that proposed scheme is detecting water leakage better.
Keywords: Machine Learning, Support Vector Machines, Random Forest, Water Leakage Detection.
Scope of the Article: Machine Learning