Augmented Morphological Designs Enabled Through Network Based Genetic Algorithms

This paper describes a computational protocol that generates architectural form via internal logic rather than the creation of objects that resemble natural forms produced with advanced geometry engines. Taking cues from the neo-Darwinian process of natural selection, the proposed algorithm augments the standard genetic algorithm. Rather than optimizing existing forms, it generates new form by evolving hierarchical networks rather than scalar parameters. The emulation of non-linear regulatory mechanisms, as in DNA, allows for the modeling of complex hierarchical architectural morphologies, responsive to internal forces and epigenetic factors, fashioned after the construction of natural organisms. The far-reaching conclusion is the potential of developing a computational protocol that is capable of negotiating multiple types of independent information in the designing of the architectural object.