Detection of spike imbalance prices in the French electricity market using machine-learning methods
| Autor: | Barbero, Mattia, Guillain, Pierre, Corchero García, Cristina|||0000-0002-8465-0830, Perroy, Edouard |
|---|---|
| Přispěvatelé: | Universitat Politècnica de Catalunya. Doctorat en Estadística i Investigació Operativa, Universitat Politècnica de Catalunya. Departament d'Estadística i Investigació Operativa |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | UPCommons. Portal del coneixement obert de la UPC Universitat Politècnica de Catalunya (UPC) |
| Popis: | Electricity price is a key factor in determining short¿term operating schedules and bidding strategies in competitive electricity markets for retailers, Balance Responsible Parties and Aggregators. However, forecasting spike prices in imbalance markets may prove particularly challenging, due to the nature of the service. This study proposes a new day-ahead classification method to detect winter spike prices occurrences using an ensemble of Support Vector Machine, Random Forest and Extreme Gradient Boosting. Results over 2019 imbalance prices show that the method can correctly forecast almost half of the days in which a spike price occurred, with just 25% of false positives. Economic results are even more encouraging as the real value of the correctly predicted spike days is about six times higher than that of the days wrongly predicted, having almost the same number of days. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|