A neural network based technique for short-term forecasting of anomalous load periods

Abstract

The paper illustrates a part of the research activity conducted by authors in the field of electric Short Term Load Forecasting (STLF) based on Artificial Neural Network (ANN) architectures. Previous experiences with basic ANN architectures have shown that, even though these architectures provide results comparable with those obtained by human operators for most normal days, they evidence some accuracy deficiencies when applied to ''anomalous'' load conditions occurring during holidays and long weekends. For these periods a specific procedure based upon a combined (unsupervised/supervised) approach has been proposed. The unsupervised stage provides a preventive classification of the historical load data by means of a Kohonen's Self Organizing Map (SOM) The supervised stage, performing the proper forecasting activity, is obtained by using a multi-layer perceptron with a backpropagation learning algorithm similar to the ones above mentioned. The unconventional use of information deriving from the classification stage permits the proposed procedure to obtain a relevant enhancement of the forecast accuracy for anomalous load situations

Short-Term Load Forecasting for Microgrids Based on Artificial Neural Networks

“…While the solutions studied in the literature [35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][55][56][57][58][59] present sometimes good prediction efficiency figures (normally their MAPEs are around 2%), they deal almost exclusively with big areas, and mainly entire countries, and they are never applied to smaller environments of the size of small cities or microgrids. Therefore, they do not give any evidence of how will they behave when applied to highly variable load curves.…”

“…Alfuhaid et al [44] use a small ANN that pre-processes a data set and produces peak, valley and total load forecasts; these forecasts, in combination with other data, are used as input to a larger ANN to obtain next-day load forecast. Lamedica et al [45] present 12 ANNs-one for each month of the year-where load curves are classified using Kohonen's Self-Organized Map.…”

Forecasting the electricity load from one day to one week ahead for the Spanish system operator

“…Public holidays may be smoothed (Taylor et al, 2006), treated as Sundays (Smith, 2000), or replaced by the load observed on a similar day in the week before (Hippert et al, 2005). The number and types of special days are usually specified on a priori grounds, although some authors have used pattern recognition analysis to classify day types empirically (Lamedica et al, 1996). Within the single model approach, common practice is to use dummy variables to estimate the changes in the load due to special days.…”

“…Deterministic seasonal components include dummy variables (Ramanathan et al, 1997;Soares & Souza, 2006) and time-varying periodic splines (Harvey & Koopman, 1993). Methods for explaining shortterm time dependence include linear autoregressions (Pardo et al, 2002;Cottet & Smith, 2003), fractionally integrated processes (Soares & Souza, 2006), artificial neural networks (Lamedica, Prudenzi, Sforna, Caciotta, & Orsolini Cencelli, 1996;Hippert, Bunn, & Souza, 2005;Alves da Silva, Ferreira, & Velasquez, 2008-this issue), double seasonal Holt-Winters' exponential smoothing adjusted for error correlation (Taylor, 2003), etc.…”

Robust Smart Meter Data Analytics Using Smoothed ALS and Dynamic Time Warping

“…Over the past few years, much research has been put into load curves classification in order to solve the short-term load forecasting of anomalous days and identify different types of customers and their behaviors. Clustering methods that have been discussed so far are the self-organizing map, the k-means, the fuzzy k-means, and the average and Ward hierarchical methods [9][10][11]. These methods generally belong to pattern recognition techniques [12].…”