Research On Car Side Elevation Feature Line Aided Design Method Based On Genetic Algorithm

With the development of China’s automobile industry as well as the improvement ofpeople’s living standards in the recent years, consumers are not only concerned about thecars’quality, power performance economic applicability, etc., but also concerned about thebeautifulness, diversity, andpersonalization of automobile styling. Whether Automobilestyling is beautiful has become an important factor which determines its market performance.Domestic automobile companies have gradually realized this fact, and have increasedinvestment in product development at the modeling stage, which could improve thecompetitiveness and brand imageof their products.The development of automobile modeling in China has only be10years, it is muchbackward compared to foreign countries with lack of skills and experience. The cardesigners usually do their work according to the feedback of market research, the overallarrangement, as well as the brand’s styling language.Although there have been a lot ofconstraints, the car designers also need to depend on their inspiration, intuition andsubjective experience speaking of the detailed innovative design and the control of the linerproportion. In view of the lack of effective theory and method supporting the design process,it has much uncertainty and meanwhile, the designing products are often difficult to meet theneed of the consumers accurately.This article explores a rational design method in which the genetic algorithm has beenintroduced, which quantifies the emotional design. It can provide the designers withauxiliary design main frame on the basis of the psychological needs of consumers, helpingdesigners grasping of the features of the design elements and the user’s demand properly, andreducingdesign risk.This paper focused on the side elevation feature line，which is the most important line inautomobile design. Using Kansei engineering theory and genetic algorithm optimizationmethod，this paper studied the car side elevation feature line-aided design method. The study is mainly divided into three parts:(1) Through the internet, magazines and other means, the C-class models on the marketwere collected, according to sales andstyles,10representative sample models were selected,and the side elevation feature line evaluation maps were made.(2) Adjectives describing the user intention were collected through the network,magazines and expert consulting method. Using questionnaire the respondents were asked tochoose the adjectives which can describe the perceptual intention of car side lines mostproperly, and eventually five adjectives were selected.The second part is the questionnaire survey and statistical analysis, the model for therelationship between user intention adjectives and line shape variable was established.(1) On the one hand, the questionnaire about the relationship between user intentionadjectives and C-class side elevation feature line was made, before the formal hand out ofthe survey questionnaire, it was modified according to the assessment of the reliability andvalidity through the preliminary research. On the other hand, to analyze the side elevationfeature lines, the side elevation feature lines from10kinds of cars were described usingBézier curve, the hard point coordinates and control point coordinates were defined as theshape design variables and the value of each variable were respectively recorded.(2) The correlation analysisand multiple regression analysis had been done to the resultsof the questionnaire through SPSS multiple regression method and five models for therelationship between user intention adjectives and line shape variable were established.The third part is to obtain the optimal side elevation feature line based on the userintention through genetic algorithm.(1) Using genetic algorithm optimization toolbox,generational genetic optimization wasdone and the ten line shape variable was set as initial population. In the process ofoptimization, the model for the relationship between user intention adjectives and line shapevariableserved as a fitness function or the optimization goal in another word. Ten samples ofthe models of form design variable values for the initial population. Genetic optimizationoperation was done for five user intention adjectives resulting out the optimal side elevation feature line corresponding to each adjective respectively.(2) Finally, several examples were introduced about how to use the side elevationfeature line-aided design methodto help the car designers to do the relative design work.