关键词: 灰色模型, 背景值, 马尔可夫链, 检验指标, 负荷预测

Abstract: With the rapid development of the electric power industry, mid- and long-term power load forecasting has become a very important fundamental work. Based on the study of a traditional mid- and long-term grey model, two methods to restructure the background values of the model are introduced and the improvement of the model structure is made based on the Markov chain and then the fitting accuracy index and the prediction accuracy index are tested to verify the modification from different aspects. Finally, through quantitive comparison between two scenarios, the effectiveness of the improved model is proved with the real power load data.

Key words: grey model, background value, Markov chain, test index, load forecasting