Lighting System Control in Office Building Using Occupancy Prediction Based on Historical Occupied Ratio

Abstract

Intelligent control has attracted more attention in the operation of commercial buildings, and the control of lighting system is one significant part. The way of lighting system control in most of commercial buildings is commonly based on occupancy data. The passive infrared (PIR) sensor is mostly chosen to obtain the occupancy data. Compared with the traditional lighting system control method based on constant schedule, the proposed control methods are adjusted by actual occupancy. The problem lies in the inaccuracy of PIR sensor, which causes "false off" when room is still occupied. To solve this problem, the current control method uses delay period as the criteria. In this research, we process the occupancy data, and improve the control method combining occupancy prediction method. The improved prediction method mainly correlates the prediction with historical occupied ratio. Higher accuracy and lower rate of "false off" are achieved.