Predicting the desired thermal comfort conditions for shared offices

Ensuring thermal comfort is an important goal in the operation of any office building. Often these buildings are therefore controlled according to the predicted mean vote (PMV) model which is also adopted as the ISO 7730 norm. This model predicts the mean thermal preferences of an average group of people that would satisfy the thermal comfort needs of about 80% of the occupants. It is derived from a number of environmental and person-dependent variables that are either difficult to measure in practice or require the placement of many sensors which is very costly. Furthermore, this model produces precise results for well defined environmental conditions only. This paper presents two new case-based reasoning algorithms that predict the average comfort vote based on previously recorded votes. In order to determine how relevant the latter are to a new state a distance measure is defined that quantifies the similarity between two states. Based on that similarity the previous votes are weighted and the expected comfort vote for the new state is determined. The two algorithms differ in the definition of their distance function. One depends only on the temperature of the state and the other one depends on all the parameters of the PMV model. The paper concludes with an experimental evaluation using real field study data. The results reveal that the algorithm, which requires temperature sensors only, outperforms two existing PMV based approaches. It shows that the goals of increasing user comfort and reducing sensor costs can be achieved simultaneously when considering the comfort votes that occupants have previously provided. The experiments also show that for hot arid, wet equatorial, and temperature marine climate zones the measurement of the PMV value, rather than that of temperature alone, can lead to increased user comfort when considering previously recorded comfort votes.