A comparison of a few statistical models for making quantile wind power forecasts

Abstract

During the last few years, probabilistic wind power forecasts have received increasing attention because of their assumed value in decision‐making processes. In the current article, three statistical methods are described and several models based on these are compared. The statistical methods are local quantile regression, a local Gaussian model and the Nadaraya–Watson estimator for conditional cumulative distribution functions. The focus is on quantile forecasts, since these often provide the required type of information to make optimal economic decisions and are ideal for visualizing uncertainty. The statistical methods are applied to data from a wind farm in Norway and results are compared using appropriate measures for assessment of quantile forecasts and in terms of a simple model for economic value. Copyright © 2005 John Wiley & Sons, Ltd.

From probabilistic forecasts to statistical scenarios of short‐term wind power production

“…The latter ones are the most common and utilized in practice today, even though the use of prediction risk indices also comprises a promising alternative (or complementary) approach [4]. Probabilistic predictions can be either derived from meteorological ensembles [5], based on physical considerations [6], or finally produced from one of the numerous statistical methods that have appeared in the literature, see [4,7,8,9, 10] among others. Their aim is to give more information on the random variable P t+k than the simple summary statistics given by a point forecast.…”

Reliability diagrams for non‐parametric density forecasts of continuous variables: Accounting for serial correlation

“…As wind-power generation is a nonlinear and bounded process, predictive densities may be highly skewed and with heavy tails (Lange, 2005), and hence be difficult to model accurately with known parametric families of density functions (see the discussion by Pinson, 2006). This has motivated the development of a large number of non-parametric methods for wind-power density forecasting, based on statistical methods and/or ensemble forecasts (see Bremnes, 2006;Møller, et al, 2008;Nielsen, et al, 2006;Pinson and Madsen, 2009a, among others).…”

Impact of wind power forecasting error bias on the economic operation of autonomous power systems

“…Normalized prediction errors have been considered in this paper for both load and wind power forecasting, as described in equation (3). Thus, the errors take into account any loading or installed wind power capacity of the power system under study.…”

“…Many efforts have been presented in the technical literature for wind power forecasting (wpf) in power systems, [1][2][3][4][5][6][7] but few of them have been used for autonomous power systems. 7,8 Recent studies [9][10][11][12][13][14] have shown that improving wpf has signifi cant economic savings for both utilities and the wind park owners participating in the energy market.…”

“…14 Signifi cant progress has been also made in the uncertainty assessment of wind power forecasts and their application in wind power integration in electricity markets, 15 using meteorological ensembles, 16,17 physical considerations 18 or statistical methods. 3,[19][20][21] As an example, local quantile regression has been applied to optimize the income of a wind power producer by identifying which quantile of the probability distribution for wind power production should be used when bidding in the open market, 22 also stressing how the prediction uncertainty estimation can help in minimizing imbalance costs. The re-sampling method 23 has been also used to make uncertainty estimation and to identify the confi dence intervals of a wpf tool.…”