kth.sePublications KTH
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Multi-Horizon Forecasting of Aggregated Electricity Consumption and Production Volumes for Balance Responsible Portfolios: A Benchmark Study using Transformers, XGBoost and LSTM Models
KTH, School of Electrical Engineering and Computer Science (EECS).
2025 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Prognostisering av aggregerad elanvändning och produktion för portföljer med balansansvar : En jämförande studie med Transformer-, XGBoost- och LSTM-modeller (Swedish)
Abstract [en]

This thesis addresses the challenge of accurately forecasting day-ahead electricity consumption and production for bidding purposes in Nord Pool’s Day-Ahead Market. In this highly interconnected environment, even minor forecast errors can result in suboptimal bidding strategies and elevated imbalance costs for Balance Responsible Parties in the electricity system, highlighting the importance of robust and reliable prediction methods. In this study, the performance of Autoregressive Integrated Moving Average (ARIMA) as a baseline model is compared against more advanced approaches: Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM) networks, and the Temporal Fusion Transformer (TFT). These models are evaluated based on their accuracy—measured through MAE, nMAE, MAPE, and RMSE—as well as interpretability, with the goal of identifying the most effective approach for day-ahead consumption and production forecasting. The results indicate that the Temporal Fusion Transformer demonstrates strong potential, achieving the lowest error metrics across the board among the models evaluated. Moreover, by leveraging Quantile Regression, TFT natively produces probabilistic prediction intervals, providing valuable uncertainty estimates without the need for external post-processing—a critical advantage for operational bidding under uncertainty. For the more challenging production case, a hybrid approach combining an XGBoost-based classifier with the TFT model further improved performance. The classifier outputs a probability of production per hour, which is then passed as a known real-valued input to the TFT model. This setup enhances the model’s ability to anticipate sparse and volatile production events, especially under high variance and zero-heavy regimes. Further, a simple theoretical framework for bidding optimization is proposed leveraging linear programming, in order to try to maximize revenue and minimize cost in the day-ahead market while staying risk-averse. By offering a clearer picture of future energy supply and demand, this forecasting framework can support more effective bidding strategies, reduce imbalance risks, and enhance overall market efficiency. This thesis makes a valuable contribution to the growing body of work in data-driven energy forecasting and lays the foundation for future research in hybrid modeling and real-time energy operations.

Abstract [sv]

Den här examensarbetet tar sig an utmaningen att prognostisera elförbrukning och elproduktion för budgivning i Nord Pools day-ahead-marknad. Även små prognosfel kan leda till sämre budstrategier och ökade obalanskostnader för balansansvariga parter i elsystemet, vilket gör robusta och tillförlitliga prognosmetoder extra viktiga. I studien jämförs en baslinjemodell, Autoregressive Integrated Moving Average (ARIMA), med mer avancerade metoder: Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM) och Temporal Fusion Transformer (TFT). Modellerna utvärderas baserat på noggrannhet (mätt i MAE, nMAE, MAPE och RMSE) samt tolkbarhet, med målet att hitta den bästa metoden för dag-för-dag-prognoser av förbrukning och produktion. Resultaten visar att Temporal Fusion Transformer presterar bäst, med de lägsta felen i alla mätningar. Tack vare kvantilsregression kan TFT dessutom direkt generera probabilistiska prognosintervall, vilket innebär att osäkerheten i prognoserna fångas utan att behöva lägga till extra beräkningar i efterhand— en viktig fördel vid budgivning under osäkra förhållanden. För den mer utmanande produktionsdatan förbättrades resultaten ytterligare med en hybridmodell, där ett XGBoost-baserat klassificeringslager förutser sannolikheten för överproduktion varje timme. Dessa sannolikheter används som kända indata i TFT-modellen och ger en bättre grund för att hantera sparsamma och volatila produktionsmönster. Utöver detta så framhävs en simpel och risk-avers budgivningsstrategi baserad på linjärprogrammering för att maximera intäkter och minimera kostnader från day-ahead marknaden. Genom att ge en tydligare bild av framtida elproduktion och efterfrågan kan det här prognosramverket bidra till bättre budstrategier, minskad obalans och effektivare marknadsdeltagande. Rapporten bidrar med nya insikter inom datadriven energiprognostisering och lägger en grund för fortsatt forskning kring hybridmodeller och operativ energihantering.

Place, publisher, year, edition, pages
2025. , p. 80
Series
TRITA-EECS-EX ; 2025:628
Keywords [en]
Electricity forecasting, Time series prediction, Day-ahead market, BRP portfolios, Solar production, Temporal Fusion Transformer, XGBoost, Hybrid modeling, Prediction intervals, Multi-horizon forecasting, Feature importance, Machine learning, Deep learning, Smart grids, Renewable energy, Energy markets
Keywords [sv]
Elprognoser, Tidsserieprediktion, Dayahead-marknad, BRP-portföljer, Solöverskott, Temporal Fusion Transformer, XGBoost, Hybridmodellering, Prognosintervall, Flerhorisontprognoser, Funktionsvikt, Maskininlärning, Djupinlärning, Smarta elnät, Förnybar energi, Energimarknader
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:kth:diva-368315OAI: oai:DiVA.org:kth-368315DiVA, id: diva2:1988429
External cooperation
Greenely AB
Supervisors
Examiners
Available from: 2025-08-18 Created: 2025-08-11 Last updated: 2025-08-18Bibliographically approved

Open Access in DiVA

fulltext(3201 kB)492 downloads
File information
File name FULLTEXT01.pdfFile size 3201 kBChecksum SHA-512
ad5726da286e95669f77b9742921482b90309c08fdc1cffaed18c13e67307efe88d68f039a414899f885634c05537af6dfa9f7294ac12864fdf1da834078ef08
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 493 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 1010 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf