Real-time prediction of workplane illuminance distribution for daylight-linked controls using non-intrusive multimodal deep learning

TL;DR

This paper introduces a multimodal deep learning framework that predicts indoor workplane illuminance distributions in real time using non-intrusive images, enhancing daylight-linked control systems for energy efficiency.

Contribution

It presents a novel deep learning approach that uses only window-side images to accurately predict indoor illuminance, applicable in dynamic indoor environments.

Findings

Achieved R2 > 0.98 and RMSE < 0.14 on same-day test data.

Achieved R2 > 0.82 and RMSE < 0.17 on unseen-day data.

Demonstrated high accuracy and temporal generalization in real-world experiments.

Abstract

Daylight-linked controls (DLCs) have significant potential for energy savings in buildings, especially when abundant daylight is available and indoor workplane illuminance can be accurately predicted in real time. Most existing studies on indoor daylight predictions were developed and tested for static scenes. This study proposes a multimodal deep learning framework that predicts indoor workplane illuminance distributions in real time from non-intrusive images with temporal-spatial features. By extracting image features only from the side-lit window areas rather than interior pixels, the approach remains applicable in dynamically occupied indoor spaces. A field experiment was conducted in a test room in Guangzhou (China), where 17,344 samples were collected for model training and validation. The model achieved R2 > 0.98 with RMSE < 0.14 on the same-distribution test set and R2 > 0.82 with RMSE < 0.17 on an unseen-day test set, indicating high accuracy and acceptable temporal generalization.