Perfect Matching: Comparison of methods for deriving energy-optimized occupancy schemes

In applications with occupancy-dependent, zoned lighting, a zone is used in an energy-efficient manner if there are large overlaps in occupancy patterns at energy-relevant times of day. The distribution of users in the room thus has a significant impact on energy consumption. However, deriving optimal distributions proves to be a combinatorial problem that is NP-hard, even with just a few people.

Therefore, various approximation methods were compared in terms of energy savings, computing time and computing power. The results lead to energy savings of 20-25% depending on the algorithm. All algorithms use partially matching user pairings in the optimum.