Adaptive Occupancy-based Lighting Control via Grey Prediction

This paper describes adaptive occupancy-based lighting control via Grey prediction. Determining the timing of automatically switching off the lights after the occupied space is vacated is quite challenging when trade-off between occupant's comfort and energy saving is concerned. The occupancy behavior is modeled by the well-known Pareto distribution for determining such delay time based on probability of no false off-switch. Since the occupancy behavior for determining the delay time at controlling time cannot be known in advance and dynamically changes, Grey prediction model is thus applied in order to determine the delay time according to the knowledge of that occupant's activity in the past. The real-time control implementation to lighting an office room is used to illustrate a viability of the proposed methodology in practical use for energy saving.