Hello everyone,
I am building a reinforcement learning platform which runs EnergyPlus simulations via BCVTB, leveraging ExternalInterface for the communication. The idea is to observe what is going on in the simulation at each timestep and then let the AI decide the best possible action for the given simulation state.
Right now, my EnergyPlus simulation is a very simple one-thermal-zone building with only one heating/cooling furnace. So far I have been controlling the furnace by changing the value of heating/cooling setpoint. This heating setpoint is inside ZoneControl:Thermostat which has ThermostatSetpoint:SingleHeatingOrCooling as the control method. Please see the attached pictures.
https://ibb.co/mqQfkYg
https://ibb.co/YRtbHHf
With this method everything seems to be working well: if the AI chooses to raise the heating setpoint, the furnace uses more power to heat the zone. The opposite happens when the AI lowers the heating setpoint.
My question is: is there a way to control the current power of HVAC equipment (that is the furnace in this case) directly using a variable/schedule? So, instead of setting the heating/cooling setpoints and letting the furnace get the zone to the correct temperature, I would like to directly control the heating/cooling power of the furnace.
I hope I was clear in my question. Please do not hesitate to ask clarifying questions. Thanks a lot for your help!
