I compiled an EP prototype building model (Small office, ASHRAE 90.1, 2019, Seattle) into an FMU and studied its behavior. Sometimes, there are strong oscillations in the room temperature when the heat pump/chiller go through repeated on/off cycles. I noticed that this mostly occurs during unoccupied hours.

In the attached plot, you see a few days of operation of one room during summer. On top is the room temperature (setpoint and actual), center-left is the electric and thermal power of the chiller, bottom-left is the power of the fan and bottom-right is the occupancy. Temperatures are in °C, powers are in W and the x-axis is in days.

Is there a way to make this smoother? As far as I understand, EP usually sets the setpoint temperature and then retroactively calculates the necessary heating/cooling power.

C:\fakepath\TemperatureOscillations.png