Day-ahead tariff prediction method for electricity spot market based on VMD and integrated learning

With the deepening of the power system reform, electricity price prediction is crucial to the decision-making of power market participants. In order to further improve the accuracy of electricity price prediction, this paper proposes an electricity spot market day-ahead electricity price prediction method based on VMD and integrated learning, which combines the variational modal decomposition (VMD) and integrated learning algorithms, and constructs a day-ahead electricity price prediction model for the electricity spot market. First, the supply-demand ratio and bidding space features describing the supply-demand relationship in the market are constructed by analyzing the formation mechanism of the electricity price; second, the influencing factors of the electricity price are analyzed and the original feature sequence is decomposed using the VMD to reduce its non-stationarity and volatility, and extract more regular modal components; then, the features with strong correlation with the electricity price are screened by the importance of the features as inputs to the prediction model; Finally, the integrated learning method is used to construct the day-ahead electricity price prediction model, and the final prediction results are obtained through weighted average. The experimental results show that the method is better than a single model in prediction accuracy, with better prediction precision and generalizability.