Improved Non-dominated Sorting Genetic Algorithm III: An innovative multi-objective optimization framework for sustainable building design

Abstract

This study presents an innovative multi-objective optimization framework for sustainable building design, addressing the intricate balance between building morphology and building performance in the early design phase. By integrating four critical attributes—energy consumption, cost, functionality, and thermal comfort—into a cohesive model, this framework offers a comprehensive evaluation approach. The proposed methodology combines the Penalty Function (PF) method, Generative Adversarial Networks (GAN), and the Non-dominated Sorting Genetic Algorithm III (NSGA-III) to navigate the complex trade-offs and interconnections among these objectives. Through a series of numerical experiments and multi-scenario case studies, the framework demonstrates its ability to achieve high-quality Pareto-optimal solutions, effectively balancing multiple objectives. The results confirm the framework’s effectiveness and reliability, highlighting its potential for practical application in sustainable building design. This research contributes to the field by proposing a more holistic optimization model, enhancing the optimization process through advanced algorithms, and providing a robust tool for sustainable building design.