Climate Control

We have already captured our thoughts on climate control design. This project will be started in 2014 and will implement the software controller, to enable our designs and implementation to be tested. Our plans are to test the concepts and software in a number of zones, where we can control heaters at a zones level.

Watch this space!

Design

Our design is based upon the following assumptions:

  • We model both 'Heating' and 'Cooling' in our home. Our assumption is that both can't be on at the same time.
  • Our system will automatically exercise the climate control hardware regardless of the on/off settings.

  • Climate control is controlled at an individual zone level.
  • A zone is considered open if one of the doors into the zone is open. It is considered closed if all of the doors into the zone are closed. Our model will also take the status of windows in each zone but, we don't have the ability to monitor this in our current home.
  • We already model zones and know if they (and all doors) are open or closed. Our climate control system will adjust its behaviour based on this zone attribute.
  • Our model knows about zone relationships and takes account of neighbour zones and their characteristics and status (e.g. temperature).
  • We model zone area and height (and thus the volume). This is used by our model to calculate the required energy input.
  • We model heaters and coolers in each zone in terms of their power rating. This is an 'effective power rating' and will be estimated and refined by through testing.
  • We model the insulation level in each zone relative to neighbour zones, e.g. some zones have external walls and some don't. We use this to model rate of heat escape.
  • We model both binary (on/off) and variable (0% to 100%) heaters and coolers and control them accordingly. This hleps to avoid 'overshoot', whilst keeping responsiveness.
  • We assume one or more temperature sensors in each zone and know about their height. This is used to calculate average temperature and temperature gradient from floor to ceiling. It is assumed these sensors are installed in optimal locations.
  • We are assuming that our temperature sensors are accurate and have low latency.
  • We model external temperature and humidity and this is an input to our controllers.
  • We model the some zones having ceiling fans and other methods of moving air, to reduce temperature gradients.
  • We model some zones having humidity sensors.
  • Each zone has a controllers that takes account of the house status and presence and occupancy
  • The controllers must model multiple heat sources in the zone, such as towel rails in the bathrooms.
  • The controllers must model devices such as extractor fans in the bathrooms.

  • The controllers must provide frost protection.

Features & Interfaces

We are not great advocates of 'toy' thermostats that provide (remote) control via apps. The objective of this project is to actually avoid any physical thermostats being visible on walls and to reduce the amount of user interaction required. We envisage a climate control system that will basically be set up once and then will adjust its behaviour automatically based on our automatic presence and occupancy detection. One a year we will switch the heating on and sometime later switch it back on again. If we think we need cooling we will switch it on or off. Everything else will just happen seamlessly in the background.

Any control will be done by the generic interfaces that our smart home already supports.

Implementation

Climate is modelled as a static Java class (there is only one instance).

'Heating' and 'Cooling' are both objects in our home and have a status of 'On' or 'Off'. The current status is stored the database. The class ensures that both heating and cooling are not on at the same time. Changes in status of these two objects result in an whole house audio announcement.

We have been using controllers in our home for some time now. This project will extend their capability much further.

Testing

Our conservatory and study are going to be our initial test cases. Our testing involves capturing detailed temperature information within each zone to see how they react to heat input over time.

Share ...
We are on ...
Facebook Twitter
YouTube Flickr Follow us on Pinterest