Development of a blind control occupant behaviour model (BC-OBM) based on contextual and time-related factors through supervised machine learning approaches

Abstract

Blind use patterns significantly impact visual comfort and energy efficiency in office buildings. To quantify this impact on building performance, blind control occupant behaviour models (BC-OBMs) are typically integrated into building performance simulation (BPS) during the building design. Existing BC-OBMs mainly focus on environmental factors, and integrating the complexities of these BC-OBMs into BPS requires familiarity with BPS tools. However, especially in developing countries, integrating BC-OBMs into BPS poses challenges due to the limited exposure to the BPS tools within the industry, leading to an energy performance gap. Moreover, existing BC-OBMs overlook contextual and time-related factors due to data collection and quantification challenges. To address these gaps, this study proposes a novel BC-OBM grounded solely on contextual and time-related factors using mixed effect binary multivariate logistic regression (BMLR) and random forest (RF) models. A six-month longitudinal field study was conducted in forty-two offices in Tiruchirappalli, India, collecting 935 blind state observations from 87 participants. The analysis revealed a preference for closed blind (73.8 %) and identified statistically significant relationships between blind/shade use and key contextual and time-related factors. The proposed BMLR and RF models exhibit substantial predictive power, with accuracies of 0.847 and 0.734, respectively, underscoring their efficacy. This will facilitate architects and building engineers to make informed early-stage building design decisions without relying on BPS tools. This study also contributes significantly to window blind research by defining novel relationships between contextual factors and blind use, which stand unprecedented within academic research to the best of the author's knowledge.