Day-ahead electricity price prediction applying hybrid models of LSTM-based deep learning methods and feature selection algorithms under consideration of market coupling

• Published in: Energy (Oxford) Vol. 237; p. 121543
• Main Authors: Li, Wei, Becker, Denis Mike
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 15.12.2021; Elsevier BV
• Subjects: Algorithms; Artificial neural networks; Deep learning; Economic forecasting; Electricity; electricity costs; Electricity market coupling; Electricity price forecasting (EPF); Electricity pricing; Empirical analysis; empirical research; energy; Feature selection; Long short-term memory; Long short-term memory (LSTM); Machine learning; markets; Mathematical models; Neural networks; prediction; Predictions; prices; The Nord Pool system price; time series analysis; Trade liberalization; Deep learning; Electricity price forecasting (EPF); Feature selection; The Nord Pool system price; Electricity market coupling; Long short-term memory (LSTM)
• ISSN: 0360-5442; 1873-6785
• DOI: 10.1016/j.energy.2021.121543

The availability of accurate day-ahead electricity price forecasts is pivotal for electricity market participants. In the context of trade liberalisation and market harmonisation in the European markets, accurate price forecasting becomes difficult for electricity market participants to obtain because electricity forecasting requires the consideration of features from ever-growing coupling markets. This study provides a method of exploring the influence of market coupling on electricity price prediction. We apply state-of-the-art long short-term memory (LSTM) deep neural networks combined with feature selection algorithms for electricity price prediction under the consideration of market coupling. LSTM models have a good performance in handling nonlinear and complex problems and processing time series data. In our empirical study of the Nordic market, the proposed models obtain considerably accurate results. The results show that feature selection is essential to achieving accurate prediction, and features from integrated markets have an impact on prediction. The feature importance analysis implies that the German market has a salient role in the price generation of Nord Pool. •Propose three LSTM-based deep learning hybrid architectures for electricity price forecasting, considering market coupling.•Apply a broad set of explanatory variables from the Nord Pool and its six integrated markets.•Conduct comprehensive comparisons for various feature selection algorithms.•Detect the feature importance from integrated markets by applying a Shapley value-based approach.•Conclude that the influence of feature selection is significant on the forecasting accuracy of LSTM-based hybrid models.