Automobile appearance beauty degree metering method based on shape beauty degree and energy efficiency beauty degree

Abstract

The invention discloses an automobile appearance beauty degree metering method based on form beauty degree and energy efficiency beauty degree, and relates to the technical field of automobile design. The method comprises the steps of data collection and preprocessing, including automobile 3D model data collection and preprocessing and energy efficiency comment data collection and preprocessing; calculating the automobile form beauty degree based on the automobile form complexity and the automobile form order; calculating the automobile energy efficiency beauty degree based on the automobile energy efficiency complexity and the automobile energy efficiency order; based on weighted summation of the automobile form beauty degree and the automobile energy efficiency beauty degree, the automobile appearance beauty degree is calculated; and verifying the appearance beauty of the automobile through a nine-level Like scale and a Spearman rank correlation coefficient. Through the scientific calculation method and verification means, numerical value quantification of the automobile appearance beauty is achieved, the efficiency and quality of automobile appearance design are improved, and innovative development of the automobile design field is promoted.

Description

Technical Field

[0001] The invention relates to the technical field of automobile design and evaluation, in particular to a method for measuring automobile appearance beauty based on form beauty and energy efficiency beauty. Background Art

[0002] With the continuous advancement of science and technology, especially the rapid development of artificial intelligence technology, the field of automobile design has also ushered in a new era of intelligent design. Traditional automobile appearance design mainly relies on the designer's experience and subjective aesthetics, and lacks scientific and quantitative evaluation standards.

[0003] At present, some studies have attempted to apply aesthetic measurement theory and information entropy to the evaluation of automobile appearance, but most of these studies remain at the theoretical level and lack practical application verification. In addition, the existing evaluation methods mostly focus on a specific aspect, such as morphological beauty or energy efficiency beauty, and lack a comprehensive evaluation of automobile appearance beauty. Therefore, establishing a scientific and objective automobile appearance beauty evaluation system has become an urgent problem to be solved in the current automobile design field. Summary of the invention

[0004] The purpose of this invention is to propose a method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty, to achieve numerical quantification of automobile appearance beauty through scientific calculation methods and verification means, and to promote innovative development in the field of automobile design.

[0005] To achieve the above object, the present invention proposes a method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty, and the specific steps are as follows:

[0006] S1. Data collection, including automobile 3D model data and energy efficiency review data, and preprocessing the automobile 3D model data and the energy efficiency review data;

[0007] S2, calculating the beauty of the automobile form based on the complexity and order of the automobile form;

[0008] S3, calculate the beauty of automobile energy efficiency based on automobile energy efficiency complexity and automobile energy efficiency order;

[0009] S4, calculating the appearance beauty of the automobile based on the weighted sum of the automobile form beauty and the automobile energy efficiency beauty;

[0010] S5. The appearance beauty of the car was verified by using a nine-level Likert scale and Spearman rank correlation coefficient.

[0011] Preferably, in S1, the automobile 3D model data preprocessing includes: view mode switching, background graying processing and normalization processing; the energy efficiency review data preprocessing includes: data cleaning, text normalization, word segmentation, word frequency statistics and LDA topic induction.

[0012] Preferably, in S2, the calculation formula of the automobile shape beauty is as follows:

[0013]

[0014] Among them, M s For the beauty of the car, s is the car morphology order, C s is the complexity of the car shape.

[0015] Preferably, the vehicle shape complexity C s The rank of the system, the dissipation of energy, and the slow growth rate present a complex state under the complex interaction of the system. The calculation formula is as follows:

[0016] C s =r s ×E×t;

[0017] Among them, r s is the dimension of the vehicle appearance structure line vector space, E is the energy dissipation of the vehicle appearance, and t is the slow growth rate;

[0018] Using information entropy theory, calculate the order of the car shape O s , the formula is as follows:

[0019]

[0020] Among them, O s is the car morphology order, f j (x i ) is the car appearance structure line function, W j is the weight of structure line j, m is the number of points on the curve with different absolute values ​​of derivatives, k = (ln m) -1 , which can satisfy 0≤S j ≤1; a=1, 2, ···, 5, corresponding to the outline, front view, side view, rear view, top view respectively Figure 5 car exterior views, j is an integer, i is an integer.

[0021] Preferably, the weight W of the structure line j is j The calculation formula is as follows:

[0022]

[0023] Among them, w fj is the weight of the eye movement data of the number of fixations of the jth curve, w tj is the weight of the fixation duration eye movement data of the jth curve.

[0024] Preferably, in S3, the automobile energy efficiency beauty calculation formula is as follows:

[0025]

[0026] Among them, M e For car energy efficiency, e For the automobile energy efficiency order, C e The complexity of automobile energy efficiency.

[0027] Preferably, the vehicle energy efficiency complexity C e The calculation formula is as follows:

[0028] C e =r r ;

[0029] Among them, r e is the dimension of the energy efficiency vector space;

[0030] Using the Gibbs free energy function, the energy efficiency order of the vehicle is calculated. e , the formula is as follows:

[0031]

[0032] Where p is energy efficiency, O e is the energy efficiency order, S p is the energy efficiency entropy, W p is the weight of energy efficiency p, q is an integer;

[0033] The weight W of the energy efficiency p p The calculation formula is as follows:

[0034]

[0035] Among them, f p is the word frequency.

[0036] Preferably, the energy efficiency entropy in, is the vehicle energy efficiency experience intensity value, and its calculation formula is as follows:

[0037]

[0038] in, is the pth energy efficiency experience positive polarity word; is the pth energy efficiency experience negative polarity word; C δ is the δth degree level, δ is an integer; ε δ is the frequency of the degree adverbs of the δth degree level of the positive polarity words of energy efficiency experience; θ δ The frequency of the degree adverbs of the δth degree level of the negative energy efficiency experience words; ΔG pis the sum of the products of positive energy efficiency experience words and their sentiment intensity in the review data; ΔH p It is the absolute value of the sum of the products of the positive polarity words of energy efficiency experience and their sentiment intensity in the comment data plus the sum of the products of the negative polarity words of energy efficiency experience and their sentiment intensity.

[0039] Preferably, in S4, the formula for calculating the beauty of the automobile appearance is as follows:

[0040]

[0041] Among them, M is the beauty of the car appearance, W s is the beauty weight of the car shape, W e Midea weighting for car energy efficiency.

[0042] Preferably, the automobile appearance beauty weight W s and the weight of the car's energy efficiency e The weighting is based on the G1 expert weighting method, which quantitatively scores the relative importance of adjacent indicators to determine the importance weight of each indicator. The most important indicator is U1, and the importance order is recorded as U2, U3, ..., U n , and its calculation formula is as follows:

[0043]

[0044] Where τ is an integer, is an integer, r n For U n-1 with U n Importance ratio.

[0045] Therefore, the present invention proposes a method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty, and its beneficial effects are as follows:

[0046] (1) The present invention combines aesthetic measurement theory, information entropy theory and complexity theory, and systematically analyzes the automobile appearance aesthetic evaluation theory and method practice, and condenses the automobile appearance beauty measurement model that combines morphological beauty measurement with energy efficiency beauty measurement. It overcomes the shortcomings of traditional evaluation methods that are highly subjective and difficult to standardize, and improves the scientificity and objectivity of the evaluation results.

[0047] (2) The present invention combines eye tracking experiments, information entropy calculation and G1 expert weighting method to propose a scientific and systematic beauty calculation method with high practical value.

[0048] (3) The beauty measurement method based on morphological beauty and energy efficiency beauty of the present invention can also be extended to the appearance design evaluation of products in other fields, and has wide applicability and promotion value.

[0049] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of a method for measuring the appearance beauty of an automobile based on the morphological beauty and the energy efficiency beauty of the present invention;

[0051] Figure 2 A position distribution diagram of automobile appearance design elements of a method for measuring automobile appearance beauty based on form beauty and energy efficiency beauty of the present invention;

[0052] Figure 3 This is an eye movement experiment flow chart of a method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty of the present invention;

[0053] Figure 4 A schematic diagram of a questionnaire on automobile appearance beauty based on a method for measuring automobile appearance beauty based on form beauty and energy efficiency beauty of the present invention;

[0054] Figure 5 It is a schematic diagram of the calculation results of the noun participle set perplexity of a method for measuring the appearance beauty of an automobile based on the morphological beauty and the energy efficiency beauty of the present invention;

[0055] Figure 6 A schematic diagram of the order of the appearance of each view of 20 sample automobiles according to a method for measuring the appearance beauty of an automobile based on the appearance beauty and the energy efficiency beauty of the present invention;

[0056] Figure 7 This is a score distribution diagram of 20 sample cars according to a method for measuring the appearance beauty of cars based on morphological beauty and energy efficiency beauty. DETAILED DESCRIPTION

[0057] In order to make the technical solutions, advantages and purposes of the present invention clearer, the technical solutions of the embodiments of the present invention are clearly and completely described below. The described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of this application.

[0058] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0059] like Figure 1 As shown, the present invention provides a method for measuring the appearance beauty of an automobile based on the morphological beauty and the energy efficiency beauty, and the specific steps are as follows:

[0060] 1. Data collection and preprocessing

[0061] 1.1 Automobile Model Data Collection

[0062] The research sample set is constructed based on the car sales criterion, and the sample set of the top 100 car models in sales on the "Autohome" evaluation website is constructed to ensure the sufficiency of the research data. For the collection of specific research data, 20 car models are randomly selected from the sample set to form a research data set. The content of the data set includes: car 3D model data, energy efficiency review data (user review data from the Autohome website and the Autohome website are crawled by web crawlers). First, the car 3D model data is used as the research data for the calculation of car morphological complexity and morphological order. Secondly, the energy efficiency review data is used as the research data for the calculation of car appearance energy efficiency complexity and energy efficiency order.

[0063] 1.2 Data Preprocessing

[0064] 1.2.1 Car 3D model data preprocessing steps are as follows:

[0065] (1) Import the sample car 3D proportional simulation model from www.dongchedi.com into Rhino software, switch the view mode to pen mode, and generate a 3D line drawing model.

[0066] (2) Use the perspective switching function of Rhino software to generate front view, side view, back view, and top view. Figure 4 Line drawing of each view.

[0067] (3) As shown in Table 1, according to the automobile appearance design element criteria, the automobile appearance design elements are mapped to the corresponding views; Figure 2 As shown, the line drawings of the four views are imported into Adobe Illustrator software for standardization, wherein the background in the view is grayed (R: 172, G: 172, B: 172), the line type of the vehicle line drawing in the view is black line (R: 0, G: 0, B: 0), and the vehicle image itself is filled with white (R: 255, G: 255, B: 255), so as to eliminate the influence of factors such as color and material, so as to achieve standardized line drawings of the car appearance elements. The line drawings are shown in Table 2.

[0068] Table 1 List of automobile exterior design elements extracted in this paper

[0069]

[0070] Table 2 Pre-processed line drawing of automobile appearance

[0071]

[0072]

[0073] (4) According to the logic of calculating the order of the car's form, the line drawing is divided into five parts: the first part is the outer contour view, the second part is the front view, the third part is the side view, the fourth part is the rear view, and the fifth part is the top view.

[0074] 1.2.2 Preprocessing of energy efficiency evaluation data, the steps are as follows:

[0075] (1) Data cleaning, use the Excel data filter to remove data noises such as special characters, non-semantic related symbols, duplicate comments, invalid comments, etc. in the text.

[0076] (2) Text normalization, standardize the energy efficiency evaluation data, and perform related stop word removal processing. Specifically, use the micro word cloud software to perform stop word removal processing on the massive comment data, including removing meaningless prepositions, verbs (such as "de", "shi", "zai"...), brand institution nouns and other stop words.

[0077] (3) Word segmentation, perform word segmentation on the energy efficiency evaluation after normalization processing. Specifically, use the jieba word segmentation algorithm of the micro word cloud software to split the relevant text data into individual words. Finally, construct a noun word segmentation collection and an adjective word segmentation collection of the energy efficiency comment data, further count the number of adjective feature words and group similar feature words.

[0078] (4) Word frequency statistics, use the word frequency statistics tool of the micro word cloud software to perform word frequency statistics on the grouped adjectives. As shown in Table 3, sort according to the frequency from high to low to prepare for the subsequent calculation of the energy efficiency order.

[0079] Table 3 Adjective Word Frequency Statistics (Top 50 Adjectives)

[0080]

[0081] (5) LDA topic induction, based on the LDA topic model, cluster and induce the topic feature word collection of the feature words; first, use the perplexity model to calculate the number of topics in the noun word segmentation collection of the comment data. As Figure 3 shown, when the number of topics increases, the perplexity gradually decreases and tends to be stable, so as to determine the number of noun topics in the comment data; then determine the general parameter settings of the LDA model; finally, according to the LDA model combined with the Gibbs sampling iterative algorithm, calculate the probability of the noun topic words in the comment data, and determine the energy efficiency complexity according to the calculated noun topic words.

[0082] 2. Calculation of the beauty degree of the car appearance

[0083] 2.1 Calculation of the beauty degree of the car form

[0084] Based on the automobile shape beauty formula, the automobile shape beauty calculation is carried out. The calculation formula is as follows:

[0085]

[0086] (1) Calculate the vehicle shape complexity C s =r s ×E×t,r s is the dimension of the vector space of the car's exterior structure line. Based on the principle of the simplest order of complexity, r s =3; since this study only involves the measurement of the beauty of the car's appearance and does not involve the iteration and improvement of the car's appearance, E=1, t=1.

[0087] (2) Calculate the order of automobile shape

[0088] First is W j The sample car line drawing is divided into the outer contour view, front view, side view, rear view and top view, and the W of each view is calculated separately. j Eye movement experiments were conducted on different views. The eye movement experiment data statistics were collected based on the number of fixations and fixation duration. The experimental steps are as follows:

[0089] 1) Selection of experimental subjects. College students were randomly selected as subjects, regardless of gender and age. They were required to have myopia less than 500°, no color blindness or color weakness, and no neurological diseases.

[0090] 2) Experimental environment. This experiment was conducted in the Human-Computer Interaction Laboratory of Shenyang Aerospace University. The laboratory environment can fully meet the environmental requirements of the experiment: the overall color tone of the experimental space environment is gray; the experimental table is clean and free of clutter; the illumination of the experimental space working surface is set to 120-150 lux, three-primary color light source, color temperature set to 5000K, and there is no natural light; the laboratory has a noise shielding function, and the indoor noise is within 30db.

[0091] 3) Experimental equipment and materials. The equipment uses the Tobii Pro Lab integrated platform, which includes the Tobii ProGlasses 3 wearable eye tracker (wearable myopic lenses, up to 500°, sampling frequency of 100Hz, and the error range of the measurement result from the actual eye movement position is 0.5 degrees) and a PC test computer (DELL Latitude 3420 laptop with an external monitor, the monitor model is PHL 2751, the screen size is 27 inches, the screen resolution is 2560×1440, and the operating system is Windows 10). The car views after the preprocessing of the car model data in 3.1.1 are used as experimental materials.

[0092] 4) Experimental test procedures. Figure 4As shown, first, debug the experimental equipment, input the experimental materials into the PC test computer, fill in the subject information and read the experimental instructions, emphasize the standard sitting posture of the subjects, standard wearing of the eye tracker, and limit the distance between the subject's eyes and the test display to 70cm. Car outline experiment stage: The subject observes the test card for one-point calibration. After gaze correction, the subject first observes a specific observation point in the black view (observation time is 5000ms), and then observes the test material (observation time is 10000ms). The page switching speed is set to 50ms, and the above observation process is repeated until the observation of the sample experimental material is completed. Car four-view experiment stage: Except for the observation time of the test material of 20000ms, the rest remains the same.

[0093] 5) Data collection. After the experiment, the experimental data on the SD card on the eye tracker was imported into the PC test computer. The eye movement index data (number of fixations and fixation duration) of all subjects were collected through the Pro Lab software pre-installed in the PC test computer, and further data analysis was carried out in combination with Excel.

[0094] 6) Data statistical analysis: First, calculate the weight of each eye movement data of each structure line in the view: where w fj The weight of the eye movement data represented by the number of fixations of the jth curve; Where W tj It is expressed as the weight of the fixation duration eye movement data of the jth curve; then the importance weight of each structural line in the view is calculated: Where W αj Denoted as the importance weight of the j-th curve of the α-th view.

[0095] Eye tracking experimental data and W of the outline view, front view, side view, rear view and top view of the sample car line drawing j The weights are shown in Tables 4 to 8.

[0096] Table 4 Car outline eye tracking experimental data and W j Empowerment

[0097]

[0098] Table 5 Car front view eye tracking experimental data and W j Empowerment

[0099]

[0100]

[0101] Table 6 Car side view eye tracking experimental data and W j Empowerment

[0102]

[0103]

[0104] Table 7 Car rear view eye tracking experimental data and W j Empowerment

[0105]

[0106] Table 8 Car top view eye tracking experimental data and W j Empowerment

[0107]

[0108]

[0109] The second step is to calculate the information entropy of the j-th structural line of the car appearance. The steps are as follows:

[0110] 1) Import the pre-processed line drawing of the car appearance into Rhino 8 software, use the curve tool to construct the car appearance structure line in proportion (the model size is in millimeters) (because the car appearance is highly symmetrical, only half of the appearance structure lines are selected in the front view, rear view and top view of the car appearance according to the central symmetry axis), and use the derivative calculation method based on the curve parameterization representation (Grasshopper plug-in) to calculate the derivative of the j-th structure line. The specific method steps are: select and import the j-th structure line into the Curve component; use the Divide Distance component to differentiate the structure line at the limit equidistant to obtain discrete points; use the Derivatives component to calculate the derivatives of each point on the structure line; use the Deconstruct Vector component to split the derivative vector into x and y components (two-dimensional curve); use the Division component to calculate the derivatives of each node, and import the data into Excel software.

[0111] 2) After importing the derivative data into Excel software, use the data analysis tool in the software according to the formula Calculate the information entropy of the j-th structural line of the car appearance.

[0112] like Figure 5 As shown, the information entropy redundancy of each structural line is weighted and summed using Excel software to calculate the appearance order of the car. The calculation formula is as follows:

[0113]

[0114] (3) Calculate the beauty of the car's shape

[0115] According to the importance weight value and information entropy redundancy value of the car's exterior structure line, the car's appearance beauty is calculated using the data analysis tool of Excel software. The calculation formula is as follows:

[0116]

[0117] The car appearance beauty is ranked, and the results are shown in Table 9.

[0118] Table 9 Ranking of the beauty of the 20 sample cars

[0119] Ranking Sample car Form beauty Ranking Sample car Form beauty 1 hq* 0.9835 11 pst* 0.9376 2 bm* 0.9738 12 sy* 0.9353 3 xp* 0.9667 13 xr* 0.9337 4 bc* 0.9613 14 yg* 0.9317 5 wl* 0.9543 15 zj* 0.922 6 wew* 0.9512 16 han* 0.9122 7 ad* 0.9483 17 hb* 0.8966 8 jk* 0.9464 18 ylt* 0.8907 9 tsl* 0.9435 19 st* 0.8751 10 xm* 0.9415 20 kmr* 0.8419

[0120] 2.2 Calculation of Automobile Energy Efficiency

[0121] First, calculate the energy efficiency complexity of the car C e =r e , r e is the dimension of the energy efficiency experience vector space during vehicle operation, that is, the number of energy efficiency experiences is equal to r e (determined according to the results of LDA noun topic induction).

[0122] Secondly, calculate the energy efficiency of the car The specific steps are as follows:

[0123] 1) Assignment of energy efficiency experience word pairs (adjective word pairs). Corresponding to the dimension of the energy efficiency experience vector space, the energy efficiency experience word pairs are determined according to the frequency ranking of the adjective word set, and the determined energy efficiency experience word pairs are assigned values; as shown in Table 10, a set of energy efficiency experience word pairs contains a positive emotional word and a negative emotional word, the positive polarity word is assigned a value of +1, and the negative polarity word is assigned a value of -1. This is used as a criterion for further measuring the emotional intensity of energy efficiency experience.

[0124] Table 10 Energy efficiency word pairs

[0125]

[0126] 2) Measurement of the emotional intensity of energy efficiency experience. The emotional intensity of energy efficiency experience word pairs is measured based on the degree adverb data. Degree adverbs are adverbs that modify adjectives in degree, and are generally placed before or after the modified adjectives, such as "very", "very", "very". Different words express different tone intensities, and their intensity is assigned by the degree adverb level and sense of language in the How Net dictionary, as shown in Table 11.

[0127] Table 11 Relationship between degree adverbs and intensity assignment

[0128]

[0129]

[0130] 3) Calculate the vehicle energy efficiency experience intensity value The formula is:

[0131]

[0132] in represents the pth energy efficiency experience positive polarity word; represents the pth negative energy efficiency experience word; C δ represents the δth degree level, δ=1,2,…,7; ε δ The frequency of adverbs of degree at the δth degree level of positive energy efficiency experience words, θ δ The frequency of occurrence of degree adverbs of the δth degree level of negative energy efficiency experience words. ΔG p It is represented as the sum of the products of positive energy efficiency experience words and their sentiment intensity in the review data; ΔH p It is expressed as the absolute value of the sum of the products of the positive polarity words of energy efficiency experience and their sentiment intensity in the review data plus the sum of the products of the negative polarity words of energy efficiency experience and their sentiment intensity.

[0133] 4) Calculate the weight W of automobile energy efficiency word pairs p As shown in Table 12, based on the frequency of the energy efficiency experience word pairs, the energy efficiency experience word pairs are weighted by frequency, and the result is equal to the weight W of each energy efficiency word pair. p (There are q energy efficiency experience word pairs, and the word frequency of word pair p is f p ,but ).

[0134] Table 12 Weights of Energy Efficiency Experience Word Pairs

[0135]

[0136]

[0137] Finally, calculate the car's energy efficiency beauty. Determine the car's exterior energy efficiency experience intensity value and energy efficiency word pair weight W p Then, through the formula The car energy efficiency beauty is calculated and ranked, and the results are shown in Table 13.

[0138] Table 13 Automobile Energy Efficiency Beauty Ranking

[0139] Ranking Car Model Energy efficiency Mido Ranking Car Model Energy efficiency Mido 1 kmr* 0.2944 11 hq* 0.2809 2 sy* 0.2854 12 zj* 0.2798 3 st* 0.2850 13 ylt* 0.2787 4 wl* 0.2848 14 pst* 0.2779 5 xm* 0.2842 15 tsl* 0.2776 6 hb* 0.2828 16 han* 0.2720 7 bm* 0.2828 17 bc* 0.2713 8 xr* 0.2825 18 yg* 0.2710 9 jk* 0.2824 19 xp* 0.2668 10 ad* 0.2820 20 wew* 0.2598

[0140] 2.3 Weighted calculation of automobile appearance beauty and automobile energy efficiency beauty

[0141] The weight of the car's appearance beauty (Ws ) and the weight of automobile energy efficiency (W e ) to assign weights. The principle of this method is to quantify the relative importance of adjacent indicators and thus determine the importance weight of each indicator. Suppose the most important indicator is U1, then the importance ranking is recorded as U2, U3, ..., U n , r n For U n-1 with U n The importance ratio and the indicator importance rating scale are shown in Table 14.

[0142] Table 14 Indicator importance rating scale

[0143]

[0144]

[0145] In order to ensure the rationality and scientificity of the weighting, 15 experts in five fields, including automotive design experts, vehicle engineering experts, industrial design experts, ergonomics experts, and market management experts, were invited to conduct comprehensive weighting. The weighting results are shown in Table 15. The final calculation of the morphological beauty weight W s The energy efficiency beauty weight is 0.6269. e is 0.3731; W s and W e The calculation formula is as follows:

[0146]

[0147] where τ=1,2,…, represents the number of experts;

[0148] Table 15 GI Expert Empowerment

[0149]

[0150] 2.4 Calculation of car appearance beauty

[0151] The weighted sum of the appearance beauty and energy efficiency beauty of the car's appearance is used to calculate the appearance beauty of each sample car. The calculation formula is as follows:

[0152]

[0153] And sort them from large to small according to the beauty value to obtain the research sample beauty evaluation data set, as shown in Table 16.

[0154] Table 16 Ranking of the overall beauty of the appearance of 20 sample cars

[0155]

[0156] 3. Verification of Mido Calculation

[0157] The questionnaire survey data verifies the results of the beauty calculation and the research hypothesis of this article, reflecting the significance of this study. The specific steps are as follows:

[0158] (1) Questionnaire design. Figure 6 As shown, the 20 sample cars were uniformly constructed with the Likert 9-level scale to construct a questionnaire, and the beauty level was designed from small to large: extremely unsightly, very unsightly, unsightly, somewhat unsightly, neutral, somewhat beautiful, beautiful, very beautiful, and extremely beautiful;

[0159] (2) Questionnaire distribution. This study used the “Credamo” platform (www.credamo.com) to manually distribute questionnaires for a fee. A total of 400 questionnaires were distributed over three days. The subjects of the questionnaire survey were car enthusiasts on various Internet platforms (Autohome forum users, DoCar Forum users, Pacific Automotive Network forum users, and Weibo Auto Super Topic users). This ensured that the surveyed groups had a full understanding of cars to ensure the accuracy of the questionnaire data.

[0160] (3) Questionnaire collection and organization. Figure 7 As shown, after collecting all the questionnaires, it is necessary to screen out invalid questionnaires (including missing data, unified answers, not filled in according to the requirements, and too short filling time), count the valid questionnaire data into Excel software, and draw a score distribution chart.

[0161] (4) Questionnaire reliability analysis. First, the Cronbach's Alpha value of the valid questionnaire data was calculated. Based on the characteristics of item independence in the questionnaire of this study and the requirements for global consistency detection, the classic Cronbach's Alpha formula was used to determine the reliability of the questionnaire score. 0 means unreliable; 1 means completely reliable; less than 0.6 means poor reliability; greater than 0.7 means good reliability. The analysis results are shown in Table 17. The classic Cronbach's Alpha formula is:

[0162]

[0163] Where: α represents Cronbach's Alpha value, N represents the number of items in the questionnaire, is the variance of the ith item in the questionnaire, is the variance of the total score of all items for each respondent.

[0164] Table 17 Questionnaire data analysis of 20 models of cars

[0165]

[0166] (5) Verification of beauty calculation. The average of the beauty scores of each car's appearance is calculated using a credible questionnaire, which is used as the comprehensive beauty value of the car appearance questionnaire survey. The comprehensive beauty values ​​of the questionnaire survey are ranked and compared with the ranking of the appearance beauty values ​​of each sample car calculated by the beauty measurement model to complete the verification of the beauty calculation. In the process of comparative verification, the Spearman rank correlation coefficient is used to test the verification results (-1 indicates a complete negative correlation, 0 indicates a complete irrelevance, and 1 indicates a complete positive correlation). The verification results are as follows:

[0167] As shown in Table 18. The formula of Spearman rank correlation coefficient is:

[0168]

[0169] Among them, ρ s is the Spearman rank correlation coefficient, d i is the rank difference of each pair of observations (i.e., for each i, the difference in its rank in the two variables is calculated), and n is the total number of observations.

[0170] Table 18 Questionnaire ranking and Midu calculation ranking analysis

[0171]

[0172] Therefore, the present invention proposes a method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty, which combines aesthetic measurement theory, information entropy theory and complexity theory, and combines a systematic analysis of automobile appearance aesthetic evaluation theory and method practice, and condenses a measurement model for automobile appearance beauty that combines morphological beauty measurement with energy efficiency beauty measurement, overcomes the shortcomings of traditional evaluation methods that are highly subjective and difficult to standardize, and improves the scientificity and objectivity of the evaluation results.

[0173] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty, characterized in that: The specific steps are as follows: S1. Data collection, including automobile 3D model data and energy efficiency review data, and preprocessing the automobile 3D model data and the energy efficiency review data; S2, calculating the beauty of the automobile form based on the complexity and order of the automobile form; S3, calculate the beauty of automobile energy efficiency based on automobile energy efficiency complexity and automobile energy efficiency order; S4, calculating the appearance beauty of the automobile based on the weighted sum of the automobile form beauty and the automobile energy efficiency beauty; S5. The appearance beauty of the car was verified by using a nine-level Likert scale and Spearman rank correlation coefficient.

2. The method for measuring the appearance beauty of an automobile based on the beauty of form and the beauty of energy efficiency according to claim 1, characterized in that: In S1, the car 3D model data preprocessing includes view mode switching, background graying processing and normalization processing; The energy efficiency review data preprocessing includes: data cleaning, text normalization, word segmentation, word frequency statistics and LDA topic induction.

3. The method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty according to claim 1, characterized in that: In S2, the calculation formula of the automobile shape beauty is as follows: Among them, M s For the beauty of the car, s is the car morphology order, C s is the complexity of the car shape.

4. The method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty according to claim 3 is characterized in that: The vehicle shape complexity C s The rank of the system, the dissipation of energy, and the slow growth rate present a complex state under the complex interaction of the system. The calculation formula is as follows: C s =r s ×E×t; Among them, r s is the dimension of the vehicle appearance structure line vector space, E is the energy dissipation of the vehicle appearance, and t is the slow growth rate; Using information entropy theory, calculate the order of the car shape O s , the formula is as follows: Among them, O s is the car morphology order, f j (x i ) is the car appearance structure line function, W j is the weight of structure line j, m is the number of points on the curve with different absolute values ​​of derivatives, k = (lnm) -1 ; a is an integer, j is an integer, and i is an integer.

5. The method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty according to claim 4 is characterized in that: The weight W of the structure line j j The calculation formula is as follows: Among them, w fj is the weight of the eye movement data of the number of fixations of the jth curve, w tj is the weight of the fixation duration eye movement data of the jth curve.

6. The method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty according to claim 1, characterized in that: In S3, the formula for calculating the automobile energy efficiency is as follows: Among them, M e For car energy efficiency, e For the automobile energy efficiency order, C e The complexity of automobile energy efficiency.

7. The method for measuring the beauty of automobile appearance based on morphological beauty and energy efficiency beauty according to claim 6 is characterized in that: The vehicle energy efficiency complexity C e The calculation formula is as follows: C e ＝r e ； Among them, r e is the dimension of the energy efficiency vector space; Using the Gibbs free energy function, the energy efficiency order of the vehicle is calculated. e , the formula is as follows: Where p is energy efficiency, O e is the energy efficiency order, S p is the energy efficiency entropy, W p is the weight of energy efficiency p, q is an integer; The weight W of the energy efficiency p p The calculation formula is as follows: Among them, f p is the word frequency.

8. The method for measuring the beauty of automobile appearance based on form beauty and energy efficiency beauty according to claim 7, characterized in that: Energy efficiency entropy in, is the vehicle energy efficiency experience intensity value, and its calculation formula is as follows: in, is the pth energy efficiency experience positive polarity word; is the pth energy efficiency experience negative polarity word; C δ is the δth degree level, δ is an integer; ε δ is the frequency of the degree adverbs of the δth degree level of the positive polarity words of energy efficiency experience; θ δ The frequency of the degree adverbs of the δth degree level of the negative energy efficiency experience words; ΔG p is the sum of the products of positive energy efficiency experience words and their sentiment intensity in the review data; ΔH p It is the absolute value of the sum of the products of the positive polarity words of energy efficiency experience and their sentiment intensity in the comment data plus the sum of the products of the negative polarity words of energy efficiency experience and their sentiment intensity.

9. The method for measuring the beauty of automobile appearance based on form beauty and energy efficiency beauty according to claim 1, characterized in that: In S4, the formula for calculating the beauty of the car appearance is as follows: Among them, M is the beauty of the car appearance, W s is the beauty weight of the car shape, W e Midea weighting for car energy efficiency.

10. The method for measuring the beauty of automobile appearance based on form beauty and energy efficiency beauty according to claim 9, characterized in that: The car shape beauty weight W s and the weight of the car's energy efficiency e The weighting is based on the G1 expert weighting method, which quantitatively scores the relative importance of adjacent indicators to determine the importance weight of each indicator. The most important indicator is U1, and the importance order is recorded as U2, U3, ..., U n , and its calculation formula is as follows: Where τ is an integer, is an integer, r n For U n-1 with U n Importance ratio.

Images