An Algorithm for Smarter Heating
Abstract: This report aims to examine the use and development of a more intelligent heating process in a house. By controlling switches integrated into a fuse, the techniques that can be used to allow an algorithm to predict how long the heating process would take are evaluated. Currently users have to, themselves, estimate how long the heating process will take. The algorithm developed aims to alleviate this problem and to allow the heating to be turned off as long as possible. With this kind of automation, the total efficiency of the heating process aims to be increased. Using linear regression, an algorithm was developed based on ON/OFF regulation. The developed algorithm attempts to estimate how fast a house can be heated. The system was then tested in a simple test facility in order to determine its functionality. Several mathematical functions for handling the required matrices for a least squares analysis in conjunction with other operations were implemented. It was found that the algorithm performs acceptably when the heating system is well set up. However, it is slow to adjust its prediction when the thermal properties of the house change. Nevertheless, it was discovered that the tests used in this project were biased. This was due to the fact that the testing facility used had no isolation and the radiator was too weak to heat the facility adequately.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)