DG 0027/DIA/2020/49

Decision support for electrical energy market participants: point and probabilistic forecasting using simulation methods and statistical learning

Grant no.: 0027/DIA/2020/49
Funding agency: Ministry of Science and Higher Education, Poland
Funding scheme: Diamond Grant
Funding period: 15.10.2020 – 14.10.2024 (48 months)
Budget: 180 000 PLN
Title in Polish: Wspomaganie podejmowania decyzji uczestników rynku energii elektrycznej: prognozowanie punktowe i probabilistyczne z wykorzystaniem metod symulacyjnych oraz uczenia statystycznego

Research team:

Principal Investigator (Kierownik):

Supervisor (Opiekun naukowy):

Collaborators (Współpracownicy):

– Ph.D. / M.Sc. / B.Sc. student

Aims and scope:

This project’s goal is to develop statistical methods for forecasting the fundamental variables in order to support the decision processes of energy market participants. Each of the three tasks planned within this project concerns generation planning, which is an issue important for the stability of a country’s energy grid. Forecasting fundamental variables and residual demand aggregated on a regional or national level is a field with a major potential for development. The obtained results will be able to help energy market participants in the decision making process. Practical, large-scale use of the expected results of the project can contribute to reducing the uncertainty related to balancing renewable energy generation and demand, with the added benefit of decreasing the extreme price spikes, predominantly in the intraday and balancing markets. This can in turn reduce the operational risk of energy generators and traders, enabling a more dynamic growth of the renewable energy field, or even reduction in prices.

Tasks:

  1. Employing regression and statistical learning methods for point forecasting of renewable energy generation and electrical load.
  2. Analysis and comparison of simulation and bootstrap methods for probabilistic forecasting of the level and structure of electrical energy generation.
  3. Comparing accuracy of probabilistic forecasts based on uni- and multivariate models of residual demand in decision support.

Publications:

Peer-reviewed articles in JCR-listed journals

2024 (0+), 2023 (1), 2022 (0), 2021 (1), 2020 (0)

  1. W. Nitka, R. Weron (2023) Combining predictive distributions of electricity prices. Does minimizing the CRPS lead to optimal decisions in day-ahead bidding?, Operations Research and Decisions 33(3), 103-116  (doi: 10.37190/ord230307). Working paper version available from arXiv: https://arxiv.org/abs/2308.15443
  2. K. Maciejowska, W. Nitka T. Weron (2021) Enhancing load, wind and solar generation for day-ahead forecasting of electricity prices, Energy Economics 99, 105273 (doi: 10.1016/j.eneco.2021.105273).

Conference papers

  1. J. Nasiadka, W. Nitka, R. Weron (2022) Calibration window selection based on change-point detection for forecasting electricity prices. In: D. Groen et al. (eds.), Computational Science – ICCS 2022, Lecture Notes in Computer Science 13352, pp. 278-284, Springer (DOI: 10.1007/978-3-031-08757-8_24). Working paper version available from arXiv: https://doi.org/10.48550/arXiv.2204.00872
  2. W. Nitka, T. Serafin, D. Sotiros (2021) Forecasting electricity prices: Autoregressive Hybrid Nearest Neighbors (ARHNN) method. In: M. Paszynski et al. (eds.), Computational Science – ICCS 2021, Lecture Notes in Computer Science 12745, pp. 312-325, Springer (DOI: 10.1007/978-3-030-77970-2_24). Working paper version available from RePEc: https://ideas.repec.org/p/ahh/wpaper/worms2106.html