Abstract
Very short-term load demand forecasters are essential for power systems’ decision makers in real-time dispatching. These tools allow traditional network operators to maintain power systems’ safety and stability and provide customers energy with high reliability. Although research has traditionally focused on developing point forecasters, these tools do not provide complete information because they do not estimate the deviation between actual and predicted values. Therefore, the aim of this paper is to develop a very short-term probabilistic prediction interval forecaster to reduce decision makers’ uncertainty by computing the predicted value’s upper and lower bounds. The proposed forecaster combines an artificial intelligence-based point forecaster with a probabilistic prediction interval algorithm. First, the point forecaster predicts energy demand in the next 15 min and then the prediction interval algorithm calculates the upper and lower bounds with the user’s chosen confidence level. To examine the reliability of proposed forecaster model and resulting interval sharpness, different error metrics, such as prediction interval coverage percentage and a skill score, are computed for 95, 90, and 85% confidence intervals. Results show that the prediction interval coverage percentage is higher than the confidence level in each analysis, which means that the proposed model is valid for practical applications.
Highlights
All models in the present study were trained with the energy demand database for the whole of 2017 for 15 min resolution, whereas the whole database for 2018 with same resolution was used for the validation analysis
The input vector used to develop the FFNN, Recurrent Neural Networks (RNN), and Support Vector Machines (SVM) models was made up of: season, the time of day the prediction was made, and energy demand values during the last 24 h with 15 min resolution. Both the proposed energy demand IPIF and all of the prediction point forecasters (PPF) presented below were programmed in MATLAB®
To more compare the results obtained by each computed PPF model, Table 8 summarizes each model’s obtained error metrics and the improvement in percentage when the results are compared against the persistence model
Summary
In order to reduce the greenhouse gases associated with electricity production, some studies in the literature point out that it is necessary to decrease the energy intensity of today’s electrical-powered devices through new technological advances [3] This efficiency improvement will not be enough to balance the expected growth in demand for energy in the short term [4,5]. This energy demand increase will be based on how quickly economies develop and societies’ desire for a better way of life [6] through new electrical household equipment or full or hybrid electric transport [4,7]. Both the increase in energy demand and renewable self-supply installations will increase the volatility in energy demand, making activities carried out by the current power systems’ controllers, such as real-time dispatching or stochastic unit commitment, more uncertain [8]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
