Hi, I have a question about making spot price forecasts (Day ahead) using AMIRIS. How exactly can this be done?
The instructions I found on github allow to model the pricing of power plants (‘market agents’). I would like to get some information on how to model spot prices.
Hi Sylwia,
you may want to have a look at our AMIRIS Examples repository, which features a full scenario parameterization for 2019.
Running this scenario will make you familiar with the functionality of AMIRIS and its simulation inputs & outputs.
A main challenge for making day-ahead price forecasts is the input data quality (forecasted load, RES generation, powerplant availabilities, costs, etc.).
However, if you’re happy to have access to such data, you may parameterize AMIRIS to simulate prices for the near-future.
Best,
Felix
I have access to data that affects spot prices, but unfortunately I have not found price forecasting on your github. I have only found scenarios to create pricing for the work of agents (the work of specific types of power plants). If you could send me a link to the examples of the spot price modelling strikcie I would be very grateful. It is possible that I just did not find it.
Thank you in advance for your help!
I’m not sure if I understand your question correctly.
Electricity prices (generated by the EnergyExchange agent) are one of the main model results by AMIRIS.
However, we did not yet use AMIRIS to predict real-world day-ahead electricity prices (i.e., running AMIRIS today for tomorrow’s prices).
If this is your goal, we unfortunately cannot provide you with any up-to-date model parameterization or data.
Although, we think that AMIRIS would be capable of achieving such forecasts if correctly parameterized.
In our current research we typically use AMIRIS to i) characterize energy systems in scenarios of the more distant future (e.g., year 2030 and beyond), or ii) running historical scenarios for benchmarking purposes.
Perhaps there is also some confusion from the PriceForecaster agent, which forecasts electricity prices and provides those forecasts to respective traders (e.g., StorageTrader) within the model. This does not resemble real-world electricity price forecasts but only “in-model” forecasts.
Hope this helps!