Medium and long term load forecasting is important for power system planning and optimization. To solve the problems of extra-long time span and heavy fluctuations in mid-long term load forecasting, a new daily load forecasting method is proposed in this paper, which can make fully use of the big data of economy, meteorology and electricity. Firstly, to address the issue of inaccuracy during holidays, a new method to depict the Spring Festival effect on a daily scale is proposed. Then, the quarterly GDP is expanded to daily level by Boot-Feibes and Lisman disaggregation (BLF), so that the time scale of economy and daily load is consistent. Finally, a support vector machine-based forecasting model is established to predict daily electricity consumption. The model is tested using the load data of a certain province in China. The results show that the proposed model outperforms other existing models, which is suitable for mid-long term daily load forecasting with complex influential factors.

Spring Festival effect,Boot-Feibes and Lisman disaggregation,support vector machine