Nonlinear Load Forecasting during Heat wave via Feed-Foreword Neural Network

Abstract

The paper describes nonlinear load forecasting during the summer season by using feed-foreword neural network (FFNN). Load forecasting during summer season is very important because the sensitivity of load pattern to temperature is rapidly high. The power load depends on the weather condition, social environment, economic situation, and so on. It is impossible to completely define mutual influence of those factors on the power demand. Therefore, methods for nonlinear pattern estimation have been used to forecast power demand. Feed-forward neural network (FFNN) is one of the generally used artificial neural network (ANN) algorithms. Especially, the paper focuses on the nonlinear power demand forecasting by summer season in South Korea.