Using generative representations for structural design

When using an evolutionary algorithm, to undertake the search for good topological solutions to a design problem, the choice of representation has a significant affect on the ability of the algorithm to search. Generative representations have been heralded as having the power to allow evolutionary algorithms to search widely for topologies but as the authors have shown in a previous paper, there are many challenges to successfully implementing generative representations. This paper shows how some of these challenges have been overcome and describes further work in testing the abilities and functionality of generative representations.