Study on Electricity Market Price Forecasting with Large-Scale wind Power Based on LSTM

Abstract

In the deregulated electricity market, accurate knowledge of electricity price trend helps maximize the profit of participants in the electricity market. But with the increasing proportion of clean energy, it brings new challenges to price forecast. This paper mainly studies how to forecast the electricity price more accurately in the power market which has large proportion of wind power. A new feature called wind load ratio is introduced, which is not only used as a candidate input of the predicted model, but also an important indicator to distinguish day and night. The electricity price model is established according to the selected characteristics, and the actual data of the Danish electricity market are used for simulation. The results show that the time series LSTM electricity price model with wind load ratio has the highest accuracy, which proves the feasibility of the proposed model.