1.中核核电运行管理有限公司，浙江省嘉兴市 3143001;2.上海大学机电工程与自动化学院，上海市 200444
1.School of Mechatronic Engineering and Automation, Shanghai University, Shanghai 200444, China;2.CNNC Operation and Management Co., Ltd., Jiaxing 314300, China
This work is supported by National Natural Science Foundation of China (No. 61876105) and State Grid Corporation of China.
In the deregulated electricity market, the accurate forecasting of electricity price is helpful for all parties to participate in the market operation and management. The increase of clean energy penetration rate brings new challenges to the accuracy of electricity price prediction. By choosing different input characteristic variables and combined with the characteristics of long-short term memory (LSTM) network, the electricity price prediction model for the electricity market with high proportion of wind power is built to predict the electricity price of the electricity market with wind power. The results show that the ratio of wind power to load is the key input parameter of the electricity price prediction in the electricity market with high proportion of wind power. LSTM has the characteristic of time delay memory, so it has better ability to predict the electricity market price in time series. Based on the actual data of DK1 electricity market in the Nordic market, three models are used for the comparative analysis. The results show that the LSTM model with the ratio of wind power to load and considering multi-time information input can greatly improve the prediction accuracy of electricity price duing slack time.
YAO Zilin,ZHANG Liang,ZOU Bin,et al.Electricity Price Prediction for Electricity Market with High Proportion of Wind Power[J/OL].Automation of Electric Power Systems,http://doi.org/10.7500/AEPS20190614002.