Research On The Deep Generative Design Method Of Sail Yacht Driven By Case Knowledge

With the popularization of intelligent design research,the market has put forward higher requirements for the rapid iteration of the design.However,at present,intelligent design that relies on the random generation of the algorithm usually have largely uncontrollable factors.Previous research on sail yacht design is weak,inefficient,long cycle and high labor cost,so this paper proposes an intelligent design method based on case knowledge,which aims to use deep learning technology to realize customized editing and generation of design attributes.Firstly,from the perspective of knowledge driven to analyze the design elements,brand,concept and type,function and structure of sailing yacht.Then,based on the survey of a large number of sailing yacht brands,build a relevant case base.After acquiring and defining the explicit features and implicit perception of the sailing yacht,the ontology construction tool Protégé is used to construct the sailing yacht case knowledge base.Secondly,through pre-processing the image data information,the sailing yacht design attribute database is established.The attribute generation antagonism network is used to train the sail yacht design attribute data set,so as to establish the sail yacht design attribute generation model,which can be use to generate the design properties for sailing yachts,and automatically generate a large number of design schemes.Finally,use perceptual needs for case retrieval,after obtaining the characteristic attributes that meet the requirements and the case graph with high similarity,the attribute generation can be carried out.Selecting the five-point scale selection scheme,and express the three highest score schemes in three dimensions.Also combined with entropy weight method and VIKOR to sort the optimal design schemes.In this paper,dominant modeling features and recessive user perceptual images are encoded and summarized,and the ontology structure is used to express them,aiming at providing a structured knowledge basis for intelligent design and building a bridge between designers and computers.This generative design method,which integrates computer intelligence and human intelligence,can expand innovation on the basis of designers’ deep creative thinking,effectively improve the design efficiency of scheme generation,reduce uncertainty risks,promote man-machine collaborative design,and provide a certain reference for the research and development of the intelligent design.