Enhancing generic_thermostat: more presets, temperature extrapolation for underfloor heating

Hello,

the generic thermostat could profit from more presets and from a temperature extrapolation for heatings with a high inertia, such as underfloor heating.

I am looking into these features. Feedback on how this should be done and what makes sense is welcome!

Boost preset: To eat up to a higher temperature, potentially for a limited time. E.g. for a towel heating in the bathroom. The boost_temp and boost_time can be configured.

Any other meaningful presets?

Temperature extrapolation: Heatings with high inertia continue to heat the room for e.g. 1-3h after they have been switched off. The temperature increase could be extrapolated for extrapolation_time from the change e.g. over the past 60min. Example: With an extrapolation_time of 2h, a measured change of 0.3°C/h and a target of 22°C, we could switch off at 21.4°C.

Does this make sense for heating controllers? Is there any insight how commercial solutions do it?

Best Regards

Multiple, configurable pre-sets would be a huge improvement for my use case. I need different settings for different parts of the day. Pre-sets could include Boost, Night, Morning, Afternoon, Evening etc.

@irxy, closely related to underfloor heating with high thermal inertia are storage heaters. We have them, they are a nightmare to control because they only charge up over night irrespective of what the weather will be like the next day (so some days we roast, some days we shiver). If I could take inputs from weather forecasts as well as temperature ramps due to thermal inertia I could probably create a more optimised charging profile for the heaters.

I recommend you look at advanced control strategies such as PID - Proportional Integral Derivative. Very common closed loop feedback strategies. I have no idea how HA would implement such approaches though. There is a community discussion here.