Dynamic Artificial Daylight Simulation

Dynamic Artificial Daylight Simulation

Bollen, B.

Heynderickx, I. (mentor)

Electrical Engineering, Mathematics and Computer Science

Interactive Intelligence

2012-07-25

The purpose of this study is to find an appropriate sky model for further development of an artificial daylight solution. The current concept for the artificial daylight solution, developed at Philips Research, contains a skylight with a static sky view. To make this concept more realistic and appealing Philips is interested in investigating how to convert the current skylight solution to one that provides ‘daylight’ dynamics. These dynamics should mimic the real changes in the position of the sun in the sky and should change the lighting in the room accordingly. To implement appropriate dynamics two models widely used in literature for the luminance values of the sky over daytime were studied and based on those models two additional, simplified models for the sky luminance were developed. One of the simplified models used a homogeneous luminance value, whereas the other interpolated the luminance linearly between a minimum and maximum value. The two models from literature and the two simplified versions were implemented in Matlab and their effect on the illumination in a real room was rendered in images. For the renderings HDR environmental maps of the skies were made and used as the light source for a 3D scene. Three different scenes all of the same room with a view outside were used: one contained only windows, one only skylights and the third one contained both. The renderings were compared in two experiments: the first experiment determined which model performed best in terms of realism and appeal of the daylight, and the second experiment determined which time of day was most realistic and appealing. The results of the experiments showed that overall the models from literature performed best, while the simplified linear model performed worst. However, for the scene with only skylights there was no significant difference between the simplified models and the models from literature, and so the simplified models were considered as realistic and appealing as the more complicated ones. The results of the second experiment showed that somehow the earliest time of the day was preferred over noon and later times of the day. From the results it can be concluded that a simplified sky model can equally well be used in an artificial daylight solution than a more complex model from literature, but only for a scene with only skylights. When there is a view to the sky near the horizon the simplified models scored significantly worse than those from literature. So, because Philips Research currently develops an artificial skylight solution the author advices to use a simple model of the sky luminance and to examine more complicated alternative models only when needed for window-type of artificial lighting solutions.

daylight simulation
artificial daylight

http://resolver.tudelft.nl/uuid:4789c552-5dad-4d91-8aad-d47608097947

Student theses

master thesis

(c) 2012 Bollen, B.