Research On Blade Electric Vehicles Shape Image Evaluation Based On ERP And Improved CRITIC-LRA Method

The function of Blade Electric Vehicles is changing from modern transportation to intelligent travel space,so users have more perceptual needs for its shape design.In the process of obtaining users’ perception evaluation of pure electric vehicles,Event-Related Potential(ERP)can deeply excavate users’ implicit perceptual needs and accurately obtain users’ perceived automobile modeling images,which makes it possible to explore the relationship between automobile modeling images and users’ implicit cognition.Users’ perceptual needs for automobile modeling design tend to be diversified,and automobile products also show a comprehensive performance of multiple images.In view of the lack of quantification and objectivity in the process of automobile modeling image evaluation,this thesis carries out the cognitive experiment of Blade Electric Vehicles modeling image matching,obtain user perception ERP data,constructs the sample image evaluation model,and objectively evaluates the samples by combining the improved Criteria Importance Though Intercriteria Correlation(CRITIC)and Linear Regression Analysis(LRA),realize the goal of quantifying and objectifying the sample evaluation scheme.The main research contents and innovative achievements of this thesis are as follows:(1)from the perspective of brain cognition,it is found that relevant perceptual image words can induce obvious P300 components when used as stimuli,and irrelevant perceptual image words can induce obvious N400 components when used as stimuli;(2)By constructing the difference wave,each experiment includes two different types of stimulation sequences.The waveform difference is calculated to obtain a relatively pure ERP component,and then a sample image evaluation model composed of brain electrophysiological data is constructed;(3)The improved CRITIC-LRA method is proposed to objectively evaluate the automobile modeling image.In view of the problems that the traditional CRITIC method requires the data to be continuous variables with normal distribution,there is negative correlation coefficient,ignoring the internal correlation of indicators,this thesis introduces Spearman correlation coefficient to calculate the conflict between evaluation indicators,reduce the requirements for normal distribution of data,and use LRA to modify the initial weight to reduce the correlation between evaluation indicators.This thesis creatively applies ERP technology and improved CRITIC-LRA method to the process of automobile modeling image evaluation.The improved CRITIC-LRA method is used to evaluate the image of Blade Electric Vehicles modeling.The volatility of sample evaluation scheme data and the correlation between image evaluation indicators are fully considered,which can effectively improve the differentiation of image evaluation indicators and solve the ambiguity problem in the image evaluation process,it can provide quantitative,objective and valuable evaluation results for designers.