Multiobjective optimization of structure and visual qualities

Abstract

The paper describes and demonstrates an optimization tool that combines methods developed independently in the fields of structural engineering and fine art. Optimization has a long history in structural engineering. Most commonly, structural optimization seeks a configuration that minimizes material volume, subject to constraints on strength and stiffness. Metaheuristic search methods, such as genetic algorithms, have been widely applied to structural optimization problems. Such methods have also been applied in the field of evolutionary art. This field seeks to produce aesthetic images or objects through an iterative search, guided by visual, rather than functional objectives. The paper presents the Variable Evolutionary Strategy for Pareto Optimization (Vespo), and demonstrates its use to integrate visual objective functions like those in evolutionary art with technical objective functions and constraints in structural optimization. The goal is to aid a designer in developing an efficient structure with desired visual qualities. The work is motivated in part by the prominent trend in contemporary architecture toward randomized, free-form structures.