Intelligent Soft Computing Models in Water Demand Forecasting

Abstract

Given the increasing trend in water scarcity, which threatens a number of regions worldwide, governments and water distribution system (WDS) operators have sought accurate methods of estimating water demands. While investigators have proposed stochastic and deterministic techniques to model water demands in urban WDS, the performance of soft computing techniques [e.g., Genetic Expression Programming (GEP)] and machine learning methods [e.g., Support Vector Machines (SVM)] in this endeavour remains to be evaluated. The present study proposed a new rationale and a novel technique in forecasting water demand. Phase space reconstruction was used to feed the determinants of water demand with proper lag times, followed by development of GEP and SVM models. The relative accuracy of the three best models was evaluated on the basis of performance indices: coefficient of determination (R 2 ), mean absolute error (MAE), root mean square of error (RMSE), and Nash-Sutcliff coefficient (E). Results showed GEP models were highly sensitive to data classification, genetic operators, and optimum lag time. The SVM model that implemented a Polynomial kernel function slightly outperformed the GEP models. This study showed how phase space reconstruction could potentially improve water demand forecasts using soft computing techniques.

Application of Wavelet Decomposition and Phase Space Reconstruction in Urban Water Consumption Forecasting: Chaotic Approach (Case Study)

“…Two different types of variables affecting water demand: climatic (e.g., temperature, relative humidity, rainfall, etc.) and socioeconomic (e.g., population and income) [14]. Climatic variables can affect short-term and mid-term values while socioeconomic variables are useful for long-term forecasting [11,15,16].…”

Gene Expression Programming Coupled with Unsupervised Learning: A Two-Stage Learning Process in Multi-Scale, Short-Term Water Demand Forecasts

“…Traditionally, lag times have been used up to the point where using more lag times would not result in a model's improvement. Some methods like autocorrelation function (ACF) or 10 The main objective of the clustering phase was to group the days in clusters that can consider the seasonality in water consumption behaviours. In order to reach the evidence of this seasonality, a two-level clustering was performed to group the months in the first stage, followed by clustering of the days with the "month clusters".…”

“…There is a common convention among water demand forecast modellers that short term forecasts are those targeting temporal resolutions hourly, daily, or weekly that are used for operational purposes of WDS [3]. Furthermore, other researchers considered temperature, precipitation, and humidity in their analysis [6][7][8][9][10][11]. The majority of the models in the literature are data-driven techniques, using water demand with a lead time to predict the future demand.…”

A Short-Term Water Demand Forecasting Model Using a Moving Window on Previously Observed Data

“…Finally, as far as approach is concerned, many of the models recently proposed in the literature are based on data-driven techniques such as artificial neural networks (ANNs) (e.g., [12,17,[22][23][24][25], support vector machines (SVMs) (e.g., [26,27], fuzzy logic [28], projection pursuit regression (PPR), random forests (RMs) and multivariate adaptive regression splines (MARS) [12].…”