The Study Of Kansei Image Model Based On Product Appearance Form Deconstruction

The attraction of the product appearance modeling is one of matter to the success ofa product design, intention and communication between users in conveying productsplay an important role. Now consumers of perceptual demand, has gradually perceptualengineering research methods in the study of modeling design has developed into adesigner an important tool to understand the consumer feeling. In previous studies whileit is possible to establish high precision intentions of perceptual model prediction, butfor modeling element evaluation method is still not perfect. Research methods, thisstudy will use Kansei Engineering in smart phones, for example to deconstruct itsexterior shape analysis, research category of product modeling design and the project,and to explore the Kansei features of design elements and the relevance of its intentionto Kansei image.First conducted research into more effective pre-evaluation modeling elements.Through investigation and analysis, select the target product sample, by using themethod of morphological analysis to deconstruct of product form, determine the productmodeling elements and its categories, to understand the main factors affecting samplesto attract consumers. Combined with orthogonal test method, a number of differentdesign elements or components for ranking, replacement, portfolio changes producenew modeling, to determine the typical representative sample.Intention to collect a lot of emotional vocabulary, create a database; usingmultidimensional scaling method and cluster analysis to select representative Kanseivocabulary; combination of product modeling samples for perceptual evaluation tests,the use of semantic differential method to people for subjective evaluation of theproduct morphology transformed into quantitative evaluation value to determine theemotional needs and consumers.In the pre-study data, product modeling is intended to build on the number ofclasses I multivariate linear regression analysis model. Thus understanding the impactforecast from between variables and the dependent variable, the establishment of rules on product design modeling elements emotional intent. Established model and validatedusing statistical methods to analyze the reliability of the test model.Quantitative values based on product form elements and product form subjectiveevaluation of quantitative values, using BP artificial neural network modeling methods,choose appropriate input and output data, set the network structure, training the networkto the intended target. Products shape elements and image of prediction network modelis set up, clear the quantitative relationship. Through comparing the rules to validate theestablished product form error evaluation, the pros and cons of different modelingmethods. Using the model to predict results analyzed the affect law of the perceptualintention of model elements.