Qualification of indoor natural lighting: application of artificial neural networks and 3DSkyView

Abstract

Environmental comfort in public school buildings is not often seen as a project requirement, due to the difficulty in considering all comfort-related architectural variables at once. This paper focuses on the level of illuminance in classrooms, by defining the importance and the relationship among visual comfort variables. The research method consisted of the observation of the users’ behaviour under different conditions of illuminance levels in the environment. Two methodological tools have been applied: the 3DSkyView extension, which was used for determining sky view factors, and artificial neural networks, which were applied for modelling the relationships between variables. The results indicate that students are used to develop tasks, either under too high, or under too low illuminance levels. The sky view factor (SVF) has presented the highest relevance for visual comfort among all variables investigated, assuming an importance level of 21% in May and 37,1% in November.