Addressing the high-rise form finding problem by evolutionary computation

摘要

This paper aims to examine the application of evolutionary algorithms to the form finding problem of high-rise buildings. In the light of mentioned purpose, this study concentrates on the conceptual phase of the design process due to the importance of early design decisions. In this respect, multiobjective real-parameter constrained optimization is considered as the method of this study in order to solve high-rise design problem. From the point of evolutionary computation, we compare two evolutionary algorithms (NSGA-II and DE) focusing on their computational performance and architectural features of the resulting alternatives. Two objective functions are formulated that specifically focus on structural displacement minimization and construction cost per square meter minimization, which are clearly conflicting. As a conclusion, we discuss in the context of the high-rise design problem, the solutions identified by the NSGA-II and DE algorithms.