Performance Management for Parametric Building Design Generated by a Genetic Algorithm

Abstract

The paper analyzes the results of applying a genetic algorithm to generate building designs. The genetic algorithm developed by the authors uses shadow length as fitness function value to examine the impact on the building design. The input of the genetic algorithm for the creation of building designs is the volume of the building, the number of building blocks, as well as the area in which the building is constructed within. The model proposed in the paper serves to illustrate to the reader how parameters can be added, and the model can be extended to apply it to different use cases. Four parameters are chosen for the analysis: shadow, transportation, outdoor area, and view. The parameters are demonstrated in simplified mathematical equations and are introduced with the necessary basic information to understand the concept of formulating the equations. The results show that the model is useful for engineers in their decision-making process and can be extended by adding other parameters. From the perspective of building design, the study emphasizes the adaptability and changeable features of the model for different case studies, which can prove useful in terms of feedback.