Estimation of illuminance sensor positions and improvement of energy efficiency in the distributed control lighting system

Abstract

We propose a distributed control lighting system (hereafter, Intelligent Lighting System) for achieving personal illuminance. The Intelligent Lighting System changes the luminances for individual lights based on a self-distribution algorithm. It learns the influence of lights on illuminance sensors through regression analysis and controls the lighting to achieve the required illuminance in the required places. When the numbers of lights and illuminance sensors increase, however, an error in the regression coefficient becomes larger; accordingly, the influence of the lights on illuminance sensors may not be correctly estimated. To solve this problem, this study proposes a method for correcting the results of the learning of the influence of the lights on illuminance sensors by estimating the illuminance sensor positions. With the proposed method, we estimate the illuminance sensor positions and determine the distances between the individual lights and illuminance sensors to correct the results of the learning according to the distances. The method proposed in this study enables the illuminance sensor positions to be estimated with an error of less than 1 m. Moreover, the method makes it possible to appropriately control lighting luminance compared with conventional method, achieving an improvement in illuminance convergence speed as well as a 5% reduction in power consumption.