In evolutionary design algorithms, the evolutionary procedures have a major impact on the quality of the genotype-fitness mapping, which in turn impacts the performance of the algorithm. Two key issues affecting the quality of mapping are the size of the genotype space and the locality of the mapping. In order to systematically investigate the role that genotype space and locality have on evolutionary performance, a set of experiments are conducted using benchmark test cases consisting of simple LEGO structures. Three different developmental procedures are implemented and tested. The results confirm that locality is critical in achieving good performance and in some cases may have a greater impact than genotype length.

View Full Paper