Learning personalized visual comfort: A data-driven adaptive shading control framework using artificial intelligence

doi:10.1016/j.enbuild.2025.116801

Learning personalized visual comfort: A data-driven adaptive shading control framework using artificial intelligence

2025-12-02

0

PRE

AI

Jean‐Louis Scartezzini

DOI：10.1016/j.enbuild.2025.116801

Original

Original request for help

Save

Abstract

En 中文

• Personal WPI/DGP ranges learned from blind override behavior. • 4-day spatiotemporal CNN outperforms LSTM and RF models • 70 % reduction in behavioral mismatch vs default control. • Occupant acceptability rate increased from 68 % to 83 %. • Pilot deployment showed rapid adaptation to a real occupant.

Learning personalized visual comfort: A data-driven adaptive shading control framework using artificial intelligence

Abstract

Journal

Organization