Research on Day-ahead Marginal Electricity Price Prediction Based on WOA-LSTM

Abstract

Accurately predicting the day-ahead market marginal electricity price is vital to both sides of the electricity market transaction. However, with a large proportion of new energy sources connected to the grid, new energy generation’s stochastic and volatile nature has become more pronounced, resulting in substantial fluctuations in the day-ahead marginal electricity price. To solidify the guiding role of the day-ahead marginal electricity price on the quotation of the electricity trading market and to solve the problem of difficult parameter selection and high randomness of the Long Short-Term Memory (LSTM) network, a day-ahead market marginal electricity price prediction model based on the whale optimization algorithm (WOA) to optimize the network parameters is proposed. The model is applied to the New-England day-ahead market, and the results show that the model is effective in improving the accuracy of day-ahead price forecasting compared to traditional forecasting models.