Volume 2: Innovative Solutions for Energy Transitions: Part I

Short-Term Load Forecasting Aiming at Chinese Festivals Using Trend-Bias Prediction and Virtual Load Replacement Y. Yu , M. S. Li , Z. G. Li , Q. H. Wu

Abstract

This paper proposes a novel load forecasting algorithm aiming at improving the accuracy of prediction near the holiday. The improvement can be divided into two aspects: during the holiday, and a period after the holiday. Trend-Bias Prediction (TBP) algorithm is applied for predicting the load during the holiday, while Virtual Load Replacement (VLR) algorithm is used after a period of the holiday. The data in this experiment is from an industrial part in China. Comparing with the benchmark, both proposed TBP and VLR are efficient and have better performances.

Keywords short-term load forecast, local predictor, Chinese festivals