Title The Multi-objective Optimization Using Evolutionary Algorithm to Design Architectural Layouts
Authors 장동국(Chang, DongKuk) ; 박주희(Park, Joohee)
DOI https://doi.org/10.5659/JAIK.2022.38.11.37
Page pp.37-48
ISSN 2733-6247
Keywords Architectural layout; genetic algorithm; multi objective optimization
Abstract This research aims to propose an efficient genetic algorithm model that generates a high-quality set of alternatives in architectural design where various objectives interact and compete. By integrating a novel location-based genotyping expression approach into an architectural design domain, an automated model would generate architectural layout forms using a genetic algorithm. Depending on the degree of fitness to the architectural layout form, the initialization and crossover method based on adjacent nodes proposed in this study exhibited different morphological characteristics. However, both quickly accomplished the desired result. The evolutionary algorithm and the fitness function for evaluating architectural layouts provided the opportunity to rapidly produce the best alternatives out of a large pool of options by evaluating user requirements and properties as used during the preliminary stages of architectural design. In a generating environment where many degrees of fitness are applied simultaneously and that contribute to fitness, the Pareto optimal method was utilized to provide balanced alternatives between multiple user requirements.