Predicting electricity prices is useful if you are selling electricity to the grid. Day ahead prices are sometimes provided by suppliers but predicting 2 day ahead prices can be useful if you need longer lead times.
Reviewing the literature reviews we find that the Lasso Estimated AutoRegressive (LEAR) model is one of the best available.
I have used a variety of datasets:
- Forecast demand (1 day ahead)
- Forecast Demand (7 days ahead)
- Wind forecast (1 day ahead)
- Wind Forecast (14 days ahead)
These datasets are correlated with the trend in electricity prices.
The price chosen is the price paid for export of solar energy in South East England (Available here)