Leveraging data-driven artificial intelligence in optimization design for building structures: A review

Abstract

Applying optimization methods to design, referred to as optimization design, is a widely adopted approach in structural design of buildings. Conventional optimization methods primarily focus on enhancing the performance or reducing the cost of buildings, while ensuring that they satisfy certain structural design requirements. However, these methods often suffer from low optimization efficiencies and struggle to satisfy implicit design constraints. Recent rapid advances in data-driven artificial intelligence (AI) approaches enable the extraction of implicit design knowledge from extensive datasets and efficient handling of complex optimization tasks, thereby introducing new possibilities for optimization design. The integration of data-driven AI methods into structural optimization has led to the growth of research on intelligent optimization design for building structures, demonstrating significant potential for generating initial designs, simplifying optimization problems, solving the related models, and evaluating the results. This study systematically reviews data-driven intelligent optimization design for building structures, with the aim of classifying various optimization techniques, and summarizing the distinct roles of data-driven AI methods in intelligent optimization design. The findings indicate a significant upward trend in the application of intelligent optimization methods, while the emergence of novel AI techniques presents both opportunities and challenges. This study also aims to provide a comprehensive reference for methods and application scenarios of intelligent optimization design for building structures; this helps designers leverage the learning capabilities of data-driven AI approaches alongside the quantitative-analysis strengths of optimization methods to enhance the quality and efficiency of building structures.