Affiliation:
1. University of Natural Sciences and Humanities , Computer Science Institute , 3 Maja 54, 08-110 Siedlce , Poland
Abstract
Abstract
The purpose of the work, presented in this article, was to obtain a price model for the Day-Ahead Market of the Polish Power Exchange (PPE). The resulting proposed models are based on Artificial Neural Networks (ANN), and the involved suggested improvement concerns the proper selection of both the type of network and the factors used in model construction. The article also proposes a new approach to the ANN with the implemented quantum learning model. The purpose of the research was to analyze factors, which exert influence on the quality of the model, like weather or economic factors, or the type of neural network used. The model determines the relationship between the price and the volume of electricity for a given hour of the day.
The mean square error and the coefficient of determination were used to measure the quality of the obtained models. The results from the experiments performed indicate the possibility of developing improved models of the Day-Ahead Market.
Reference30 articles.
1. Adamowski, J. (2019) Podstawy obliczeń kwantowych [Foundations of quantum calculations; in Polish]. Lectures at the Faculty of Physics and Applied Informatics, AGH University, Kraków, http://www.ftj.agh.edu.pl (access: 14.08.2019)
2. Alaminos, D., Esteban, I. et al. (2020) Quantum Neural Networks for Forecasting Inflation Dynamics. Journal of Scientific & Industrial Research, 79, 2, doi: 10.56042/jsir.v79i2.68439 (access: 12.12.2022)
3. Bai, J. and Ng, S. (2002) Determining the Number of Factors in Approximate Factor Models. Econometrica, 70, 1, 191–221.
4. Bernhardt, C. (2020) Obliczenia kwantowe dla każdego [Quantum calculus for everyone; in Polish]. PWN, Warszawa.
5. Bissing, D., Klein, M. T. et al. (2019) A Hybrid Regression Model for Day-Ahead Energy Price Forecasting. IEEE Access, 7, 36833-36842.