Electricity price forecasting on the day-ahead market using artificial intelligence algorithms

This article presents the application of AI algorithms to predict hourly product prices in the uniform price auction system of the Polish Power Exchange. It involves analyzing the electricity price determination process, identifying factors shaping price curves, reviewing literature on AI methods for electricity price prediction, proposing a research methodology, and modeling solutions using artificial neural networks. Nine variants of multilayer perceptrons with backpropagation were optimized and compared using most common indicators. Results were compared with forecasts from foreign articles for other European markets to evaluate the effectiveness of using AI in predicting electricity prices in the Polish Day-Ahead Market.