Hybrid gene regulatory network for product styling construction in interactive evolutionary design

Interactive evolutionary design (IED) systems, based on interactive evolutionary algorithms, have been the hot topic in both computer science and design recently. Due to the difference in both designer preferences and design capabilities for product styling, there are still many challenges that need to be addressed in practical application of IED. In order to tackle these challenges, this paper proposes a hybrid complex network model for product styling. The model incorporates both design knowledge and personal preferences, through a so-called gene regulatory network (GRN). First, we construct a GRN for product styling using design knowledge with both explicit and implicit knowledge, and then a GRN for product styling based on personal preferences using Boolean evaluation. In the next, combining the design knowledge and personal preferences, we construct a hybrid GRN for product styling. Then, according to the knowledge findings from GRNs, we develop an application strategy of the analytic knowledge in an IED. Taking the product styling SUV as an example, validation work is carried out to evaluate the evolutionary efficiency and design quality. The experimental results demonstrate that the proposed strategy, which embeds expert knowledge in an IED, can greatly improve the efficiency and quality of product styling designs.