Prediction of Water Quality Parameters Using Artificial Intelligence

Rapid urban development s often witness deterioration of regional water quality. As part of the management process, it is important to assess the baseline characteristics of the river environment so that sustainable development can be pursued. The aim of this research is to develop a water quality prediction model in Johor River at two different stream flow level (main stream and tributary). Several modeling methods have been applied in this research including; Linear Regression Model (LRM), Multi Layer Perceptron (MLP) Neural Network and Radial Basis Function (RBF) Neural Network. In this study, the water quality parameters of interests are total dissolved solids, electrical conductivity and turbidity due to their importance when studying the water quality status of any rivers. Five years data for these three parameters have been obtained from Department of Environment (DOE). A comprehensive comparison analysis for the above modeling methods outputs have been carried out and discussed in order to achieve the appropriate model method and architecture for the current study.

Autorentext

Ali Najah he is a PhD Candidate in Department of Civil & Structural Engineering, University Kebangsaan Malaysia, Bangi, Selangor, Malaysia.His research interest is related to artificial intelligence techniques with their applications to several engineering applications giving emphasis to hydrological process, environmental and water resources.

Klappentext

Rapid urban development s often witness deterioration of regional water quality. As part of the management process, it is important to assess the baseline characteristics of the river environment so that sustainable development can be pursued. The aim of this research is to develop a water quality prediction model in Johor River at two different stream flow level (main stream and tributary). Several modeling methods have been applied in this research including; Linear Regression Model (LRM), Multi Layer Perceptron (MLP) Neural Network and Radial Basis Function (RBF) Neural Network. In this study, the water quality parameters of interests are total dissolved solids, electrical conductivity and turbidity due to their importance when studying the water quality status of any rivers. Five years data for these three parameters have been obtained from Department of Environment (DOE). A comprehensive comparison analysis for the above modeling methods outputs have been carried out and discussed in order to achieve the appropriate model method and architecture for the current study.