A day-ahead electricity price forecasting algorithm considered with multidimensional factors

In light of the increasing number of participating members in the electricity market, the surge in the volume of electricity trading data and the rising demand for accuracy in electricity price forecasting, this paper proposes to construct a day-ahead electricity price forecasting method that considers multi-dimensional electricity price influencing factors. Firstly, the factors influencing the formation of electricity prices are analysed and the factors with high correlation are filtered out using the correlation coefficient method; secondly, the filtered factors and historical electricity prices are simultaneously used as inputs to the LSTM model to forecast the day-ahead electricity prices for the coming day; finally, the historical data of the PJM market is used to verify the prediction accuracy of the proposed model. The results show that the day-ahead electricity price forecasting model can be effectively applied to the actual electricity market, and also show that the short-term electricity price forecasting accuracy can be effectively improved by taking into account the electricity price influencing factors.