Help in understanding a set up

[tomissg]

Good Evening

I am trying to use a framework to simulation for a specific battery storage unit.
My I started to understand the market but I do not know if it is possible to simulate the prifits of a specific unit in a target year.
My understanding is lacking about the usage of the data input. Do i need for example to ad a demand side having a time series of the demand share that corresponds to the unit or do i need to train a whole "world" with all the possible participants to including the one in interest and train it and then make a simulation.

Best regards Thomas Dogkas

[mthede]

Hi Thomas,

We are sorry for the delayed reply. Unfortunately, we did not see your question until now.

To answer your question, I need more information on what you are planning to simulate.

• Could you describe the setting of interest in more detail? (Market, clearing mechanism, other participants, expected behavior of your simulated unit, what results you'd like to analzye, ...)
• What are your data sources? Is it a stylzed scenario or do you want to use e.g. historical data?
• Did I understand correctly that you are interested in a simulation with Reinforcement Learning (RL) strategies for this specific battery storage unit? Check out for example our example_02e where two storage units are learning agents. We usually "learn" the bid price and direction (charge/discharge), so the actual dispatch will be determined based on the submitted bids and market clearing. What is your goal?
• The storage unit can also apply rule-based strategies like in example_01c.
• Otherwise, we have optimization strategies for DSM units based on a provided demand profile. But that is without RL. Check example_01h for a residential household with storage. Is that rather what you need?

You can run the mentioned examples by selecting it ("small_learning_with_storage", "small_with_vre_and_storage", "small_household") in the examples.py file or via our command line interface.

Looking forward to hear more details on your simulation in mind!

Best,
Marie