Research On Interactive Evolutionary Design For Product Styling Based On Gene Regulatory Network

Intelligent design methods based on interactive genetic algorithms,known as Interactive Evolutionary Design(IED),have received a great deal of attention in the fields of computing and design.IED is considered to be an intelligent design method that can effectively combine human preferences and computer intelligence.For product styling,the diversity of user preferences and the differences in knowledge backgrounds make the practical application of IED challenging.Therefore,this thesis proposes the construction of a Gene Regulatory Network(GRN)model for product styling that integrates design knowledge and user preferences from the perspective of product styling design,and apply it to IEDs.The details of the work are as follows:Firstly,a GRN of product styling with design knowledge aggregated by emplicit and implicit knowledge is constructed for the attributes of product modelling design knowledge;a GRN of product modelling with user preferences based on Boolean evaluation method is constructed for the uncertainty of user preferences;and a GRN of product modelling fusing design knowledge and personal preferences is finally constructed by studying the cognitive processing characteristics of product styling.Then,in order to effectively plan the IED process,a topological analysis is carried out for the product styling GRN incorporating design knowledge and user preferences to obtain the parsed knowledge,i.e.important genes and gene association groups;and the relevant genetic operations were adjusted accordingly to propose a strategy for the application of the product styling GRN in IED.Finally,the SUV side profile styling line was used as the application object,the curve control method was used for gene coding,and a database of typical case samples was established;based on the proposed strategy,an interactive evolutionary design system for SUV side profile modelling was developed using MATLAB software.Moreover,a comparative experiment was carried out to verify the evolutionary efficiency and design quality.The experimental results show that the strategy can improve the efficiency and quality of product styling design by embedding expert knowledge in the IED while quickly obtaining a certain level of individual preference.The thesis contains 29 figures,20 tables and 163 references.