Data Analytic Models for Lighting Energy Sensitivity Analysis of Building

Various factors affecting building lighting control include natural factors, construction parameters, occupancy, and the system factors including window blinds and luminaires. Lighting energy consumes about 20 to 40% of total electricity in large office buildings. For improving the lighting control strategies, it is inevitable to understand the energy pattern. This paper focuses on deriving dependency of lighting energy consumption by various climatic factors and building design. The relevant attributes segregated from the data set obtained using Energy Plus simulation are used to carry out the correlation test. The data analytics model developed based on regression, helps in predicting lighting consumption of the building. The model was verified using the analytic tools. This derived model tested for south and north regions of India. The classification based on window transmitted radiation, solar heat gain, solar altitude, outside temperature found to have a high impact on window blinds and luminaire dimming control. Sensitivity analysis of building performance data and climate data is significant for data based building modelling and monitoring. This preliminary study is leading to pre-emptive control of building lighting and HVAC system using machine learning.