Comparative Study of Predictive and Rule-Based Lighting Control Algorithms for Context-Aware Energy Optimization

This study presents a comparative analysis of rule-based and predictive lighting control algorithms in commercial building environments, focusing on their responsiveness, accuracy, and energy efficiency. A full-scale indoor testbed comprising nine lighting zones was constructed, and both algorithms were applied under identical conditions over a 28-day period. The predictive control method employed a lightweight first-order Markov model to forecast user movement paths and proactively control lighting activation. Quantitative performance indicators—including CV(RMSE), NMBE, Time to Responsive Dimming (TRD), and Redundant Activation Count (RAC)—were utilized in accordance with ASHRAE Guideline 14-2023. The experimental results showed that the predictive algorithm outperformed the rule-based approach, reducing lighting error by 39.4%, improving TRD by 68%, and achieving a 15.2% reduction in energy consumption. These findings suggest that predictive control offers significant advantages in user-centric lighting management, particularly for retrofit commercial spaces and edge-device environments. This study provides a practical reference for selecting intelligent lighting control strategies that balance energy savings and user comfort.