Evaluation of Artificial Neural Network Techniques for Municipal Water Consumption Modeling

Abstract

Various Artificial Neural Network techniques such as Generalized Regression Neural Networks (GRNN), Feed Forward Neural Networks (FFNN) and Radial Basis Neural Networks (RBNN) have been evaluated based on their performance in forecasting monthly water consumptions from several socio-economic and climatic factors, which affect water use. The data set including total 108 data records is divided into two subsets, training and testing. The models consisting of the combination of the independent variables are constructed and the best fit input structure is investigated. The performance of ANN models in training and testing stages are compared with the observed water consumption values to identify the best fit forecasting model. For this purpose, some performance criteria such as Normalized Root Mean Square Error (NRMSE), efficiency (E) and correlation coefficient (CORR) are calculated for all models. The best fit models are also trained and tested by Multiple Linear Regression (MLR). The results indicated that GRNN outperforms all other methods in modeling monthly water consumptions.

Identifying Prominent Explanatory Variables for Water Demand Prediction Using Artificial Neural Networks: A Case Study of Bangkok

“…temperature and rainfall (Adamowski 2008;Zhang et al 2006;Jain et al 2001; among others), relative humidity, wind speed, and dew point (Zhang et al 2006). For the medium/long-term prediction, apart from historical water demand, the use of diverse variables such as household income (Chang and Makkeasorn 2006;Lui et al 2003), population and land use (Chang and Makkeasorn 2006), water price (Firat et al 2009;Lui et al 2003) and inflation rate (Firat et al 2009) have been reported.…”

Urban Water Demand for Service and Industrial Use: The Case of Zaragoza

“…In this context, the water demand modelling is a very important tool for planning and management water resources because provides adequate understanding of this behavioural response to changes in price (Firat et al 2009). For urban water, the accurate characterization of demand needs take into account that different types of users are connected to the public water network: residential, service and industrial.…”

A Comparative Study of Daily Pan Evaporation Estimation Using ANN, Regression and Climate Based Models

“…The weights are adjusted based on a comparison of ANN output and the target until they match. Some of the recent applications include synthetic streamflow generation (Ahmed and Sarma 2009), municipal water consumption modeling (Firat et al 2009), identification of unknown pollution sources in groundwater (Singh and Datta 2007), flood management (Ahmad and Simonovic 2006), and sediment loss prediction (Sarangi and Bhattacharya 2005). ANNs have an advantage over deterministic models in respect of the data needs that are less and well suited for long-term forecasting (Sarangi et al 2006).…”