Intelligent Sectorization Method for Lighting Networks Based on Clustering Algorithm

In order to balance energy saving and safety requirements, the intelligent lighting system divides the streetlight network into multiple sectors so that only the streetlights in the corresponding sector are activated when traffic elements such as pedestrians and vehicles pass by, thereby achieving traffic element-based sectorized lighting control. This strategy requires the manual pre-configuration of neighbor relationships among streetlights so that when a traffic element passes a streetlight, only the other streetlights within its sector are activated. However, in complex and large-scale roadway lighting networks, manually configuring neighbor relationships for a vast number of streetlights is extremely tedious and prone to errors. To address this issue, a method for modeling the streetlight network as a social network and performing sectorization through clustering algorithms is proposed. In this method, traffic element events detected by streetlights and network advertising are used to construct a probabilistic graph of neighbor relationships, and the ACO-CUG algorithm is employed to mine the neighbor relationships among streetlight nodes, with the clustering result thus obtained regarded as the sectorization of the streetlight network. Experimental results on a simulation dataset demonstrate that the proposed method can effectively partition streetlights in different areas, thereby achieving intelligent sectorization of the streetlight network.