Parametric regression model for lighting calibration

This article aims to assess some of the main lighting software programs currently available for architecture. It compares the daylight simulation results from programs, Flamingo nXt rendering, DIVA and Lightscape with the physical measurements carried out on a flat plate surface and in a model box. Identical parameters of geometry and lighting conditions were set. Two strategies were used to investigate the difference and correlation between the measurements and simulations. One test simulated several angles which were identical to the real-time sun position during daytime by rotating the box for both the simulation and field test, and the other test conducted every hour during daytime simulated three typical days (spring equinox, summer solstice and winter solstice). The output from simulation programs was conducted on a set grid of identical surface locations. The simulation programs and physical measurements were used to generate a parametric model. The results suggested that the accuracy of all these inspected simulation programs was acceptable for design exploration and that the parametric model could be used to calibrate the programs to give more accurate, real-time illuminance and luminance.
Source: Indoor and Built Environment - Category: Occupational Health Authors: Tags: Original Papers Source Type: research