Predictive control of internal curtains for balancing energy efficiency and visual comfort in office buildings

Balancing visual comfort and building energy efficiency remains challenging for existing curtain adjustment methods in office buildings, especially under practical implementation constraints. This study proposes an energy-saving curtain adjustment method (ECAM) that includes an energy-oriented baseline mode (ECAM-B) and an extended comfort mode (ECAM-C). Specifically, ECAM-B fully deploys or retracts the curtain according to the trade-off between air-conditioning and lighting energy consumption. ECAM-C is built upon ECAM-B by searching discretized candidate curtain retraction rates under daylight glare probability (DGP) and daylight illuminance constraints to determine the control command. A coupled luminous-thermal prediction model, validated against full-scale experimental data, was applied to estimate DGP, indoor daylight illuminance, and cooling load. ECAM was then evaluated on a representative summer day using these predicted luminous-thermal indicators and the derived energy consumption indicators. Across all transmittance cases in the south-facing room with low lighting power density, ECAM reduced daily total energy consumption relative to the traditional curtain adjustment method (TCAM), while ECAM-C maintained the predicted DGP below 0.35. At the optimal curtain transmittance of 20%, ECAM-C achieved daily energy-saving rates of 20.8% and 19.5% compared with the no-curtain and TCAM cases, respectively. In the simulated west-facing office with high lighting power density, ECAM-C further improved the indoor luminous-thermal environment. Moreover, ECAM-C provided an additional daily energy saving of 184.6 Wh compared with ECAM-B, while improving the indoor visual environment. These results may inform practical occupancy-responsive curtain control and retrofit-oriented shading automation in office buildings.