Short‐term electric power load forecasting using feedforward neural networks

Abstract

This paper presents the results of a study on shortterm electric power load forecasting based on feedforward neural networks. The study investigates the design components that are critical in power load forecasting, which include the selection of the inputs and outputs from the data, the formation of the training and the testing sets, and the performance of the neural network models trained to forecast power load for the next hour and the next day. The experiments are used to identify the combination of the most significant parameters that can be used to form the inputs of the neural networks in order to reduce the prediction error. The prediction error is also reduced by predicting the difference between the power load of the next hour (day) and that of the present hour (day). This is a promising alternative to the commonly used approach of predicting the actual power load. The potential of the proposed method is revealed by its comparison with two existing approaches that utilize neural networks for electric power load forecasting.

Multi‐scale Internet traffic forecasting using neural networks and time series methods

“…The MAPE is a common metric in forecasting applications, such as electricity demand (Malki et al ., ; Taylor et al ., ), and it measures the proportionality between the forecasting error and the actual value. This metric will be adopted in this work, since it is easier to interpret by the network administrators.…”

“…where e t denotes the forecasting error at time t; y t the desired value;ŷ t;p the predicted value for period t and computed at period p; P is the present time and N the number of forecasts. The MAPE is a common metric in forecasting applications, such as electricity demand (Malki et al, 2004;Taylor et al, 2006), and it measures the proportionality between the forecasting error and the actual value. This metric will be adopted in this work, since it is easier to interpret by the network administrators.…”

“…there is no a priori restrictions on the type of relationship to be modeled) and non-linear learning capabilities. Indeed, the use of NNs for TSF began in the late 1980s with encouraging results and the field has been consistently growing since (Lapedes & Farber, 1987;Ding et al, 1995;Malki et al, 2004;Cortez et al, 2005).…”

“…Owing to its importance, several TSF methods have been proposed, such as the Holt-Winters (Makridakis et al, 1998), the ARIMA methodology (Box & Jenkins, 1976) and artificial neural networks (NN) (Lapedes & Farber, 1987;Ding et al, 1995;Malki et al, 2004;Cortez et al, 2005). Holt-Winters was devised for series with trended and seasonal factors.…”

A review on artificial intelligence based load demand forecasting techniques for smart grid and buildings

“…The hidden layer neuron is model with a nonlinear sigmoid activation function. However, feedforward neural network is extensively used for forecasting and pattern reorganization problem [46,47] (Fig. 15).…”