Author Information
Shuhan Liang, Cornell University, United StatesLiya Xia, Peking University, China
Junnan Xie, Beihang University, China
Abstract
With rapid urbanization, industrial parks have become critical spatial carriers for economic development; however, large-scale construction often leads to degraded outdoor environmental performance, including thermal discomfort, poor ventilation, and excessive solar exposure. Traditional performance-driven design workflows rely on post-evaluation simulations, which are computationally expensive and inefficient in early design stages (Attia et al., 2013; Negendahl, 2015). This study proposes an integrated generative design framework that couples morphology generation, machine learning-based performance prediction, and multi-objective optimization to improve environmental performance in industrial park design. At the layout level, a Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is applied to optimize Voronoi-based land subdivision under multiple constraints (Hansen, 2006). At the building level, a Rectangular Voronoi Diagram (RVD)-based method is introduced to generate flexible and modular building morphologies (Dillenburger, 2010). A dataset of 5,000 samples is constructed using parametric generation and simulated using Ladybug Tools, with the Universal Thermal Climate Index (UTCI) as the primary performance metric (Bröde et al., 2012). Machine learning models, particularly XGBoost, are trained to establish mappings between morphological parameters and environmental performance (Chen & Guestrin, 2016). Finally, a multi-objective optimization framework based on NSGA-II is implemented (Deb et al., 2002). Results demonstrate that the proposed method significantly improves design efficiency and enables early-stage performance feedback, providing a scalable and data-driven approach for environmentally responsive industrial park design.
Paper Information
Conference: AGen2026Stream: Entrepreneurship/Silver Economy
This paper is part of the AGen2026 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window
To cite this article:
Liang S., Xia L., & Xie J. (2026) CMA-ES-Guided RVD-Based Layout and Building Generation Method for Aging-Friendly Industrial Parks ISSN: 2432-4183 The Asian Conference on Aging & Gerontology 2026: Official Conference Proceedings (pp. 85-93) https://doi.org/10.22492/issn.2432-4183.2026.7
To link to this article: https://doi.org/10.22492/issn.2432-4183.2026.7
Comments
Powered by WP LinkPress