Predictive potentiality of artificial neural networks for predicting the electrical conductivity (EC) of drinking water of Hyderabad city

  • Authors:

  • Niaz A. Memon;M. A. Unar;A. K. Ansari;G. B. Khaskheli;Bashir Ahmed Memon

  • Affiliations:

  • Dept: of civil Eng: Quaid-e-Awam University of Eng, Nawabshah, Pakistan;Dept/ of Computer Systems Eng: Mehran University of Eng: Jamshoro, Pakistan;Dept: of Chemical Eng: Mehran University of Eng: Jamshoro, Pakistan;Department of Civil Engineering, Mehran University of Engineering & Technology, Jamshoro, Pakistan;Department of Civil Engineering, Quaid-e-Awam Uiversity of Eng: Sciences, Nawabshah, Pakistan

  • Venue:
  • ICCOMP'08 Proceedings of the 12th WSEAS international conference on Computers
  • Year:
  • 2008

Quantified Score

Hi-index 0.01

Visualization

Abstract

Electrical Conductivity (EC) is an important parameter of drinking water. The determination of electrical conductivity provides a prompt and expedient way to measure the accessibility of electrolytes in the water. There are convinced health effects on human life through these electrolytes, like disorder of salt and water balance in infants, heart patients, individuals with high blood pressure, and renal diseases. Salty taste is one of the aesthetic effects of EC if it exceeds 150 mS/m and if greater than 300 mS/m it does not slake the thirst. The drinking water supplied to Hyderabad city is taken from River Indus and the EC of this river remains questionable. The values of EC in drinking water of Hyderabad at selected locations were recorded. From 49 samples, the average values ranged from 658 to762. In order to determine the optimal value of EC with in the distribution system, where it deteriorates, it is necessary to predict it at different locations. The use of conventional methods to predict parametric values in the distribution systems is suffered from certain precincts. To get better drinking water quality by tumbling operational costs, Advance process control and automation technologies are the tools to be used normally. The application of Artificial Neural Networks in Water Supply Engineering is enticing and more accepted because of its high predictive accuracy. In this paper Radial Basis Neural Network has been demonstrated. The data sets are prepared for training the model. It was observed that the model has high predictive potentiality to predict the values of Electrical conductivity at 07 locations of distribution system of water supply in Hyderabad city.