Supplementary Research Materials: Multi-Algorithm Surrogate-Based Optimization Framework for Office Space Daylighting Design Across Diverse Climates
Year: 2026
Type: supplementary-materials
Abstract: The research explores how to develop and evaluate a multi-objective optimization framework to improve the reliability and broad applicability of surrogate models for office-space daylighting design across diverse sky conditions. The proposed framework couples a validated machine learning daylighting model with four optimization algorithms to maximize spatial Daylight Autonomy (sDA) and minimize Annual Sunlight Exposure (ASE).Unlike existing studies that typically adopt a single optimization algorithm and fixed algorithmic settings, optimization was conducted across seven cities with varying clearness indices, where the control parameters of each algorithm were systematically fine-tuned. The optimal solutions from each tuned algorithm were synthesized into a compiled Pareto front, thus capturing a more comprehensive decision space.The framework was integrated into an online daylighting tool, providing designers with the optimal window-to-wall ratio, room depth, and height for each cardinal direction within a selected geographical context. This workflow provides architects with an accessible tool during early-stage design to create climate-responsive office spaces across diverse sky conditions and geographical locations.
Source: Figshare
Author: Mariam Amer
Language: English
Subfield: Building and Construction
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225072875
