Optimizing Costs and Quality of Interior Lighting by Genetic Algorithm

This paper proposes the use of multi-objective optimization to help in the design of interior lighting. The optimization provides an approximation of the inverse lighting problem, the determination of potential light sources satisfying a set of given illumination requirements, for which there are no analytic solutions in real instances. In order to find acceptable solutions we use the metaphor of genetic evolution, where individuals are lists of possible light sources, their positions and lighting levels. We group the many, and often not explicit, requirements for a good lighting, into two competing groups, pertaining to the quality and the costs of a lighting solution. The cost group includes both energy consumption and the electrical wiring required for the light installation. Objectives inside each group are blended with weights, and the two groups are treated as multi-objectives. The architectural space to be lighted is reproduced with 3D graphic software Blender, used to simulate the effect of illumination. The final Pareto set resulting from the genetic algorithm is further processed with clustering, in order to extract a very small set of candidate solutions, to be evaluated by the architect.