Radiance analysis and its application to real time dynamic lighting control

There is a great deal of research being conducted on the use of LED lighting systems in buildings. Their potential benefits derive from both the high levels of control they offer and increased energy efficiency as the technology develops. Rather than cluster many LEDs into large light sources as a replication of existing technologies, some of this research is looking at more distributive models where multiple LEDs are embedded into each ceiling tile and distributed more evenly over large office spaces. From a daylight-linked perspective, such lighting systems offer unprecedented levels of control, with the potential to maximise both comfort and energy savings. However, the ability to accurately and efficiently control such a multitude of individual light sources is significantly limited by existing sensor technologies required for dynamic feedback within such a system. This paper proposes that, instead of relying on real-time measurements from in-room sensors, the control system could be based on the results of a large number of pre-computed lighting simulations using daylight and fixture coefficients for multiple control points over each floor in a building. This would involve generating reasonably accurate geometric computer models of those areas of the building to be subject to this level of lighting control and using an application such as Radiance to perform the analysis