Water Supply and Environmental Engineering

Permanent URI for this collection

Now showing 1 - 1 of 1

Water Demand –Supply Gap Prediction and Driving Factors Analysis
(Addis Ababa University, 2024) Genet Hailemicheal; Agizew Nigussie
This study uses FFANNs to develop a predictive model for urban water demand supply gap forecasting. The model considers nine independent variables and uses gird search hypermeter tuning and found that Bayesian regularization back propagation training algorithm and hyperbolic tangent sigmoid transfer function (tansig) with 40 neurons in the hidden layer with seven input variables best performing model and concluded that MLP NN is an effective tool for understanding and simulating non linear water demand supply gap behavior, benefiting water providers and decision makers.In addition, this study decomposed water demand factors in Addis Ababa city using LDMI decomposition analysis methods. Ten influential factors were identified, including rainfall, temperature, sunshine hours, relative humidity, population, industrial growth rate, economic growth, tourist number, livestock number, and water tariff. The study found that except population and livestock factors contributing to total water demand change varied at different years, with industrial growth, rainfall, humidity, tourist number, and sunshine hours inhibiting change during 2021 2022. Generally, for driving factor analysis LMDI method could be a good factor decomposer provided that data interaction and independence is absent.