PREDICTION OF IRRIGATION WATER QUALITY INDEX IN RAIGARH THROUGH ANN MODELLING

Abstract

The application of conventional methods in evaluation of irrigation water quality indexes is quite cumbersome and tedious. However the application of AI models is a great substitute in such scenarios which can forecast as well as evaluate IWQI using physiochemical analysis. This project deals with the prediction of kelo river suitability for irrigation around the areas of Raigarh and Janjgir-Champa districts. The aim is to forecast the Sodium adsorption ratio(S.A.R), Kelly ratio(K.R.) and Sodium percentage(%Na) in the kelo river irrigation project, Raigarh. A total 1500 water samples are to be analysed for different physiochemical parameters as such PH, Electrical conductivity, Total dissolved solid, Calcium ions, Magnesium ions, Potassium ions, Chlorine ions and Sodium ions. These parameters are to be used as input variable in ANN. The ANN model will be developed through Neural designer and compared with MS- excel software. The best backpropagation algorithm and neuron numbers are to be determined for optimization of model architecture. The Relative mean square error, Coefficient of determination and Mean absolute percentage inaccuracy amid tentative data and prototypical output will be considered for its accuracy. The developed model will have its use in prediction of IWQI and will also be of great assistance for decision makers as well as farmers in management of irrigation water.