Next day load curve forecasting using self organizing map

In this paper, we propose a new prediction scheme using self organizing map for next day load curve forecasting. In the proposed scheme, we select several similar days corresponding to forecasted day using a Kohonen network which is a representative of self organizing map, and we forecast the next day load curve by averaging selected similar days. Therefore, we do not need complex algorithm and structure such as supervised neural network, genetic algorithm (GA) and fuzzy inference, and we can forecast next day load curve easily. The suitability of the proposed approach is illustrated through an application to actual load data of the Okinawa Electric Power Company in Japan.