Two-hour-ahead load forecasting using neural network

In this paper, we propose a two-hour-ahead load forecasting method using the correction of similar day´s data. The proposed prediction method is an extension of a one-hour-ahead load forecasting method. Therefore, first we forecast a one-hour-ahead load and then this load is taken as the input to the neural network for the two-hour-ahead load forecasting. In the proposed prediction method, the forecasted power load is obtained by adding a correction to the selected similar day´s data. The correction is yielded from a neural network. Since the neural network yields the correction which is simple data, it is not necessary for the neural network to learn all similar day´s data. Therefore, the neural network can forecast power load by simple learning