Research On Evolutionary Design Method Of Product Form Based On Computational Aesthetics

In the era of aesthetic economy,users pay more and more attention to the aesthetic quality and emotional experience of products.In the field of matured technology,the gap between product functions is becoming smaller and smaller,and it homogenizes product quality.At this time,the aesthetic quality of product form becomes one of the key factors in consumers' decision-making.The traditional design process usually depends on the designer's individual talent,judgment and aesthetic experience,and lacks objective and reliable scientific means.In order to meet the needs of enterprise design innovation and user aesthetic,and adapt to the development of advanced manufacturing technology in modern manufacturing industry,this research takes product form as the object,and explores the effective methods of product form aesthetic evaluation and evolutionary design from the quantitative level according to the aesthetic cognitive and computational aesthetic theories.The main works and results of this study are as follows:(1)From the perspective of many disciplines,the relevant theories of aesthetic evaluation of product form was analyzed.The essence of aesthetic was probed,the physiological and psychological mechanisms of product aesthetic cognition were analyzed,the aesthetic characteristics and formation reasons of the aesthetic principles and the Gestalt Principles were expounds.All above work laid a theoretical foundation for the research of aesthetic evaluation from the quantitative level.(2)From the point of objective and quantitative view,a nonlinear information dynamics model of comprehensive evaluation on product form aesthetic was constructed.According to the aesthetic principles and the Gestalt Principles,the evaluation system composed of 15 aesthetic measures was established.The 15 aesthetic measures were described quantitatively and structurally with the method of computational aesthetics to propose the equations for aesthetic measures and express explicitly the tacit aesthetic cognitive knowledge.According to Maximum Flux Principle,aesthetic measures were taken as the information elements of system evolution.The system became stable after the appropriate weightes of aesthetic measures were given by self-organized and self-evolved.Then the comprehensive evaluation of product form aesthetic was realized.This model reveals the formation mechanism of stable aesthetic structure pattern.This method is not affected by subjective factors such as artificial weighting,and has good adaptability.The values of the aesthetic measures and connection weight in the model can directly reflect the aesthetic performance of each scheme,and effectively guide the product form design.(3)From the point of subjective and objective view,a variety of evidence information of aesthetic evaluation were obtained.Using Dempster-Shafer evidence theory,the credibility of information was fused to establish an uncertainty reasoning model of subjective and objective evaluation on product form aesthetic.Compared with the Maximum Flux evaluation,the results show that the accuracy of the two methods is close.This method has some advantages in calculation efficiency and is more suitable to be the fitness function of product form evolution design system.(4)From the perspective of aesthetic evaluation of design style,aiming at problems of aesthetic standards on different styles,aesthetic measures were moderately standardized using experimental methods.A model of comprehensive evaluation on interface stylistic beauty was built based on moderately standardized of measures.The experiment shows that the stylistic beauty evaluation function is a multimodal function formed by the coupling of several functions of moderation aesthetic measures,and verifies the aesthetic phenomena that beauty has various forms and each form has its own merits.(5)By simulating the information processing of image cognition and introducing "form aesthetic measures",a grey box model,design features-form aesthetic measures-perceptual images,was proposed.The model reveals the mechanism of image visual cognition to a certain extent,and expresses explicitly the tacit image cognitive knowledge.The experiment shows that the model takes into account the potential relationship between image outputs,and the calculation is simple,efficient and accurate.The model can be used as the fitness function of product form evolution design system.(6)According to the multi-peak characteristic of product form aesthetic evaluation function,from the perspective of aesthetic evaluation and image evaluation,the multimodal optimization design system was realized utilizing Niche Artificial Fish-Swarm Algorithm to simulate design thinking.This system can prevent effectively the design solidification and meet the diversity needs of design.The quantitative aesthetic evaluation and image evaluation model are used as fitness function to make the evolutionary system have a clear optimization direction.To a certain extent,the system simulates the design thinking mode,and is conducive to improving the intelligent level and calculation efficiency of the evolutionary system.