Electricity Price for Electric battery storage
opened this issue · comments
Hi Adam,
I am just trying your demo on Electric battery storage, but there is something strange with my results.
The Electricity price in "observation.csv" are different from results in your demo.
your demo:
results with "observation.csv":
Can you also help me to understand what represent the diffrent columns (
C_forecast_electricity_price_hh_0 [$/MWh] | C_forecast_electricity_price_hh_1 [$/MWh] | C_forecast_electricity_price_hh_2 [$/MWh] | C_forecast_electricity_price_hh_3 [$/MWh])?
Are forecast for the following 1,2,3 hours?
Best regards,
Luca
Hi Luca,
I've made changes to the electricity price data - mostly to make it more interesting (i.e. more spikes). You do raise an interesting point about consistently of package data though - I'll have a think about this.
The columns are the forecasts for the next half hours - 0 being the current, 1 being the next etc.
Let me know if you have any more questions - Adam.
Thanks Adam.
In this moment I would like to use your implementation of DPL Agent, but with a custom environment.
I would like to model a HVAC system and I need to develop this enviroment.
Luca
Hi Adam,
executing demo with battery example, I have got the follwing issue:
I think there is a problem with line 90 in "env_info.py":
output.index = env.state_space.episode.index[:-1]
Is it possible?
Regards,
Hi Luca - I've fixed this, the commit is here.
Regarding HVAC - I've done some work on electric chillers. I would suggest looking for some open source simulation models, which we can wrap energy-py around.
Hi @ADGEfficiency ,
when you talk about open source simulation models, do you mean like Energy Plus software?
I would like to create a variable that is controlled by my action. Action 1 sum a value, action 2 decrease a value. How can I constrain this variable in a maximum and minimum range?
If i pass my reward, the algorithm tends to always choose a single action.