Optimization of Structural Form using a Genetic Algorithm to Search Associative Parametric Geometry

Different methods using evolutionary computation (EC) have been successfully applied to structural form optimization for many years. A well known limitation of these methods is the computational effort required to analyze significantly large and complex systems. As a result most examples have been limited to simple truss forms or smaller aspects of overall structural systems. This is because the number of variables used to describe the structural geometry limits the size and complexity of problems that can be explored using EC methods. One solution to this limitation is the use of associative parametric models. In recent years use of parametric modeling software has become more widespread in architectural design. Parametric approaches to geometric modeling have the advantage that they are able to describe large, complex systems with very few variables. The reduced number of variables needed to describe geometric systems makes parametric models ideal for use with EC to explore optimal forms. This paper shows examples of how large, complex structural systems can be optimized for basic parameters like weight and number of components by combining a genetic algorithm with a finite element analysis and parametric modeling software. The procedure is outlined with specific applications. Examples show the use of this technique used to explore complex spatial structures. Results are shown with specific documentation of computational effort required.