Feature requests...

Tips, tricks and frequently asked questions
Post Reply
Dadde
Posts: 7
Joined: 15 Apr 2015, 09:24

Feature requests...

Post by Dadde » 15 Apr 2015, 13:36

Hello to all the team netatmo,
it would be nice if the internal module weather station could interact with the thermostat, so you could also monitor the internal temperature of another room (perhaps with a different thermal coefficient), and the thermostat could to know even better the thermal characteristics of the house to get even more specific ...
What do you think ???


Thanks a lot !!!

IN ITALIAN:
"Ciao a tutto il team netatmo,
sarebbe carino se il modulo interno della stazione meteo potesse interagire con il termostato, in questo modo si potrebbe monitorare anche la temperatura interna di un'altra stanza (magari con un coefficiente termico differente), e il termostato potrebbe conoscere ancora meglio le caratteristiche termiche della casa per diventare ancora più preciso...
cosa ne pensate ???

Grazie mille !!!"

wonderiuy
Posts: 26
Joined: 21 Nov 2014, 09:58

Re: Feature requests...

Post by wonderiuy » 15 Apr 2015, 15:52

Mhhh, and which of the two thermostat should be taken into consideration for turning the boiler on and off?
This will lead to a room with a greater temperature than the desired one

Dadde
Posts: 7
Joined: 15 Apr 2015, 09:24

Re: Feature requests...

Post by Dadde » 15 Apr 2015, 17:44

The data of the inside module of the weather station should serve to the PID algorithm to better understand the heat loss of the apartment. more would be nice to see the application of the thermostat also the graph of the temperature of the indoor weather station. but it should always be the thermostat to turn on or off the boiler. or alternatively add an option to decide which of the two temperatures used for switching on or off the boiler .... I believe that in this way you could get a more even temperature of the house ...

TypeC
Posts: 289
Joined: 30 Dec 2014, 21:35

Re: Feature requests...

Post by TypeC » 16 Apr 2015, 10:49

This seems like a good idea, but I'm not sure it's likely to work very well.

It is simple (and not inaccurate) for Netatmo to talk about learning about the thermal inertia, insulation and so on of the building, but that's not really what happens. The PID control algorithm is more abstract than this.

PID is a control system which (if you think in graphical terms) tries to keep a line on a graph as straight as possible when there are other things working to destabilise it. It's not specific to heating systems: it was originally designed to keep a straight line literally, as a purely mechanical system to keep a ship going in a straight line. The variables necessary to run it are not directly those that might be measured by examining all the thermal characteristics of a house and its heating system in detail, they are those that make a wobbly line less wobbly. There's a relationship between the two in this case, obviously, but the variables are chosen by examining the wobbly line, not measuring the thickness of your loft insulation.

So, if you want to introduce a second wobbly line, with slightly different wobbles (and probably varying around a different temperature), you have a problem. It may be possible to come up with a new tuning algorithm that works to minimise the fluctuations in both lines, but there are only 3 variables and I would be surprised if reducing the fluctuation in the second line didn't simply increase that in the first line by the same amount. In other words, you need to design new mathematics, test it exhaustively and still hope it doesn't suddenly misbehave in unusual circumstances, for what is probably no real gain.

Post Reply

Return to “General questions”