Jim Kangerud MCS-2005-07, pp. 22. TEK/avd. för interaktion och systemdesign, 2005.
Many machines in these days have sensors to collect information from the world they inhabit. The correctness of this information is crucial for the correct operation. However, at times sensors are not so reliable since they are sometimes affected of some type of noise and thus give
incorrect information. Another drawback might be lack of information due to shortage of existing sensors. Sensor fusion is trying to overcome these drawbacks by integrating or combining information from multiple sensors.
The heating of a building is a slow and time consuming process, i.e. either the flow or energy consumption are object to drastically changes. On the other hand, the tap water system, i.e. the heating of tap water can be the source to severe changes in both flow and energy consumption. This because of that the flow is stochastic in the tap water system, at any given time a tap may be opened or closed and therefore drastically change the flow.
The purpose of this thesis is to investigate if is it possible to use sensor fusion to get accurate continuous flow values from a district heating substation. This is done by integrating different sensor fusion algorithms in a district heating substation simulator.