Developing Two RNN-Based Algorithms for Electricity Price Forecasting in Markets with High Penetration of Renewable Energy Resources

Precisely predicting electricity prices can greatly benefit various participants in the energy market. This is a challenging endeavor due to the inherent volatility of electricity prices. Studies have shown that as renewable energy generation increases, electricity markets and prices tend to become more unpredictable. To tackle this problem effectively, different ML-based algorithms have been presented. This paper proposes two novel electricity price forecasting algorithms based on Recurrent Neural Networks-Long Short-Term Memory (RNN-LSTM) and Recurrent Neural Networks-Gated Recurrent Unit (RNN-GRU). Two algorithms perform well in terms of MAE and RMSE scores, indicating their suitability for electricity price forecasting tasks. These results align with similar research findings and highlight their effectiveness.