Open AccessMonitoring and Intelligible Alert System to control Water Quality in Reverse Osmosis Plants
Author(s) -
K. Udayakumar*,
Dr.N.P. Subiramaniyam
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.f4358.049620
Subject(s) - support vector machine , water quality , quality (philosophy) , computer science , reverse osmosis , production (economics) , relevance (law) , control (management) , artificial intelligence , pace , machine learning , data mining , pattern recognition (psychology) , ecology , philosophy , genetics , macroeconomics , geodesy , epistemology , membrane , political science , law , economics , biology , geography
The classification of drinking water quality severity from RO production plant needs appropriate methods to provide intelligible alert to the operators who involve to carry out remedial action in pace with the production. The proposed technique finds more relevance to detect instantly the quality variations in plant through efficient classification system and drives to reduce the cumbersome of operators. In this paper, it is proposed a SVM based classification method to detect drinking water quality attributes temporally and then precisely classifying severity condition in order to correct quality derivations. A different control scheme is experimented to detect quality variables like pH, TDS, ORP and EC and to support production system. Thus this contributes an automated diagnosis of water quality in RO plant. For classification, SVM is trained with data obtained around 8 plants from West and North of Chennai region. This is demonstrated specifically for a top-level classification job on Quality. On the features extracted from 1280 data, the SVM is trained and achieves a sensitivity of 85% and an accuracy of 90%