A short-term, pattern-based model for water-demand forecasting

Abstract

The short-term, demand-forecasting model described in this paper forms the third constituent part of the POWADIMA research project which, taken together, address the issue of real-time, near-optimal control of water-distribution networks. Since the intention is to treat water distribution as a feed-forward control system, operational decisions have to be based on the expected future demands for water, rather than just the present known requirements.Accordingly, it was necessary to develop a short-term, demand-forecasting procedure. To that end, monitoring facilities were installed to measure short-term fluctuations in demands for a small experimental network, which enabled a thorough investigation of trends and periodicities that can usually be found in this type of time-series. On the basis of these data, a short-term, demand-forecasting model was formulated. The model reproduces the periodic patterns observed at annual, weekly and daily levels prior to fine-tuning the estimated values of future demands through the inclusion of persistence effects. Having validated the model, the demand forecasts were subjected to an analysis of the sensitivity to possible errors in the various components of the model. Its application to much larger case studies is described in the following two papers.

Multi-Model Prediction for Demand Forecast in Water Distribution Networks

“…Donkor et al [10] report the use of different methodologies to improve the water demand prediction in the short term (from several hours to several days ahead) and the long term (one year or more ahead); nevertheless, it does not propose the use of regimes. Other statistical and machine learning modeling methodologies that deal with the water demand forecasting problem are found in [7,[11][12][13][14]. The work of Cutore et al [15] implement similar ideas about regimes associated with working and holy days to predict one day ahead water demand of daily time series using an ANN.…”

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

“…The same data were used in a case study by [10], who undertook to apply and test a model they proposed called Patt_WDF (Pattern based Water Demand Forecasting). The results of the Patt_WDF model were thus used by way of comparison to analyze the effectiveness of the model here proposed.…”

“…There are also models with multiperiodicity, in which water demand is forecast at different time steps, for example on a daily and hourly basis, as in [10].…”

“…Other models, by contrast, are based primarily on the representation of the periodic patterns that typically characterize demand [13], possibly coupled with techniques of time series analysis, as in [10].…”

“…An analogous consideration applies for pattern-based models. For example, the model proposed by [10] requires at least one year of observations in order to define long-term periodic patterns, or seasonal fluctuations, as well as to characterize daily demand patterns, which vary from season to season. From an operational viewpoint, the need to have a time series of historically observed data for the network to which the forecasting model is to be applied can be a limiting factor to be taken into account when choosing the model's structure.…”