New energy vehicle styling design methods, systems, electronic devices and storage media

Abstract

The application discloses a new energy vehicle modeling design method and system, electronic equipment and a storage medium, and belongs to the technical field of vehicle industrial design. The method comprises the following steps: based on an interval 2-type trapezoidal fuzzy Kano model, the weight of the user emotional vocabulary obtained through the evaluation construction diagram is calculated and sorted to identify key emotional needs; based on a limit gradient boosting model optimized by a hippo optimization algorithm, a nonlinear mapping relationship between the key emotional vocabulary and the new energy vehicle form design features is established; the mapping relationship is used to predict and generate an optimal form feature combination with the highest emotional evaluation value; the generated design scheme is comprehensively evaluated subjectively and objectively by combining eye tracking technology and a discrete information data fluctuation weighting method, and a final design scheme is screened out. The method can accurately quantify and sort the subjective and fuzzy emotional needs of users, establish a nonlinear mapping relationship between the emotional needs of users and specific product design features, and improve the efficiency of new energy vehicle modeling design.

Description

Technical Field

[0001] This invention relates to the field of vehicle industrial design, and in particular to a new energy vehicle styling design method, system, electronic device and storage medium based on user emotional needs analysis and artificial intelligence technology. Background Technology

[0002] With increasing global emphasis on environmental protection and sustainable development, new energy vehicles (NEVs) have become a core development direction for the automotive industry. In a highly competitive market, consumers' car-buying decisions are no longer limited to hard indicators such as vehicle performance and range. The emotional resonance and perceived appeal evoked by the vehicle's exterior design—its "attractive form"—are playing an increasingly important role. How to accurately translate consumers' vague and subjective emotional needs (Kansei) into concrete and objective design elements is a major challenge currently facing the automotive design field.

[0003] Traditional automotive design processes rely heavily on designers' personal experience and subjective judgment, lacking a systematic and scientific approach to quantify and analyze user emotional preferences. The Kano model has been introduced for user needs classification, but traditional Kano models, when processing evaluative information, use discrete scales (such as "like" and "should be"), making it difficult to effectively address the ambiguity, uncertainty, and individual differences in user feelings. Furthermore, establishing the relationship between emotional needs and product design features typically employs linear models or simple statistical methods, failing to capture the complex, non-linear mapping between the two, potentially leading to designs that deviate from true user preferences.

[0004] Therefore, there is an urgent need for an innovative method that can effectively handle the uncertainty of user emotions, accurately establish the mapping relationship between emotions and design features, and scientifically evaluate design schemes, so as to improve the efficiency and success rate of attractive form design for new energy vehicles. Summary of the Invention

[0005] Based on this, the present invention provides a new energy vehicle styling design method, system, electronic device and storage medium, which aims to accurately quantify and sort users' subjective and vague emotional needs, establish a non-linear mapping relationship between users' emotional needs and specific product design features, and finally design a scientific and comprehensive evaluation system.

[0006] In a first aspect, the present invention provides a styling design method for new energy vehicles, comprising the following steps:

[0007] Based on a pre-set sample set of new energy vehicles and user Kano questionnaire survey data targeting emotional words, an interval type 2 trapezoidal fuzzy Kano model was used to calculate and determine the comprehensive importance weight of multiple emotional words, and to select a pre-set number of key emotional words with the highest weight.

[0008] The morphological features of the new energy vehicle sample set are deconstructed to obtain multidimensional morphological feature data; based on the multidimensional morphological feature data and the user evaluation values ​​for the key emotional words, an extreme gradient boosting model optimized by the Hippo optimization algorithm is trained to establish a mapping model between morphological features and key emotional words.

[0009] Using the mapping model, the perceptual evaluation value is predicted for all morphological feature combinations within a preset range, the optimal morphological feature combination corresponding to the key perceptual words is determined, and at least one candidate design scheme is generated based on the optimal morphological feature combination.

[0010] Eye-tracking technology is used to acquire visual perception data of the candidate design schemes, and discrete information data fluctuation weighting method is used to perform objective quantitative analysis on the candidate design schemes. Combining the visual perception data and the results of objective quantitative analysis, the final new energy vehicle styling design scheme is determined from the candidate design schemes.

[0011] As an optional implementation of the first aspect of this application, the step of calculating and determining the comprehensive importance weight of multiple emotional words includes: calculating the distribution differences of each emotional word under different Kano categories based on the improved CRITIC weighting method, using the Gini coefficient instead of the standard deviation, to determine its objective importance; determining the satisfactory importance of each emotional word based on the minimum deviation method, combined with the fuzzy evaluation of the importance of different Kano categories by experts; and weighting and integrating the objective importance and the satisfactory importance to obtain the comprehensive importance weight.

[0012] As an optional implementation of the first aspect of this application, the step of employing the interval-2 trapezoidal fuzzy Kano model includes: quantizing the language scale set in the Kano questionnaire into interval-2 trapezoidal fuzzy numbers; the user uses the interval-2 trapezoidal fuzzy numbers to answer positive and negative questions about the sensory words, obtaining fuzzy evaluation values; by calculating the distance between the fuzzy evaluation values ​​and the fuzzy numbers corresponding to each language scale, determining its membership degree to adjacent language scales, and then generating a Kano category membership vector to classify and calculate the weights of the sensory words.

[0013] As an optional implementation of the first aspect of this application, the step of deconstructing the morphological features of the new energy vehicle sample set includes: using morphological analysis to divide the morphology of the new energy vehicle into multiple design parts, the design parts including windows, headlights, air intake grilles, hubs and rearview mirrors; defining multiple types of features for each design part, and encoding the type features to form multidimensional morphological feature data.

[0014] As an optional implementation of the first aspect of this application, the step of training an extreme gradient boosting model optimized by the hippo optimization algorithm includes: treating the hyperparameters of the XGBoost model, including the number of trees, the depth of the trees, and the learning rate, as individuals in a hippo population; iteratively updating the hyperparameter combination by simulating the swimming and walking phases of the hippo population, wherein the swimming phase is used for global exploration and the walking phase is used for local development; evaluating the performance of the XGBoost model corresponding to each hyperparameter combination using the mean squared error as the fitness function until a preset number of iterations is reached or the model performance converges, thereby determining the optimal hyperparameter combination.

[0015] As an optional implementation of the first aspect of this application, the step of acquiring the user's visual perception data of the candidate design scheme using eye-tracking technology includes: recording the user's eye movement trajectory when observing the candidate design scheme, and statistically analyzing eye movement indicators including at least total fixation duration, average fixation duration, fixation count, and first fixation duration; and determining the user's visual perception priority order for each candidate design scheme based on the standardized average value of the eye movement indicators.

[0016] As an optional implementation of the first aspect of this application, the step of objectively quantitatively analyzing the candidate design schemes using the discrete information data fluctuation weighting method includes: collecting scores from multiple evaluators on each candidate design scheme for the key perceptual terms; calculating the coefficient of deviation and independent information ratio of each candidate design scheme under each key perceptual term score; calculating the pure information content of each candidate design scheme by combining the coefficient of deviation and the independent information ratio, and normalizing it to obtain the objective weight of each candidate design scheme; and ranking each candidate design scheme based on the objective weight.

[0017] Secondly, embodiments of this application provide a new energy vehicle styling design system, including:

[0018] The module for quantifying emotional needs is configured to use a pre-set sample set of new energy vehicles and user Kano questionnaire survey data targeting emotional terms. It employs an interval type 2 trapezoidal fuzzy Kano model to calculate and determine the comprehensive importance weight of multiple emotional terms and to select a pre-set number of key emotional terms with the highest weight.

[0019] The mapping model construction module is configured to deconstruct the morphological features of the new energy vehicle sample set to obtain multi-dimensional morphological feature data; based on the multi-dimensional morphological feature data and the user evaluation values ​​for the key perceptual words, a limit gradient boosting model optimized by the Hippo optimization algorithm is trained to establish a mapping model between morphological features and key perceptual words.

[0020] The optimal solution generation module is configured to use the mapping model to predict the perceptual evaluation value of all morphological feature combinations within a preset range, determine the optimal morphological feature combination corresponding to the key perceptual words, and generate at least one candidate design solution based on the optimal morphological feature combination.

[0021] The comprehensive evaluation and screening module is configured to use eye-tracking technology to acquire the user's visual perception data of the candidate design schemes, and to use the discrete information data fluctuation weighting method to perform objective quantitative analysis on the candidate design schemes. Combining the visual perception data and the objective quantitative analysis results, the final new energy vehicle styling design scheme is determined from the candidate design schemes.

[0022] Thirdly, embodiments of this application provide an electronic device, which includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the steps of the method described in the first aspect.

[0023] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. Improved the accuracy of demand analysis: IT2Tr-FKM was applied to NEV form design for the first time. The fuzzy and uncertainties in user's emotional evaluation were effectively handled by the interval type 2 trapezoidal fuzzy number, making the weight ranking of emotional needs more scientific and reliable.

[0026] 2. Intelligent mining of design elements was achieved: Utilizing the HO-XGBoost model, a complex nonlinear mapping relationship between user sentiment and specific design features was successfully established. The introduction of the HO algorithm significantly improved the prediction accuracy and generalization ability of the XGBoost model, enabling it to efficiently and accurately discover the optimal combination of design features.

[0027] 3. Enhanced the scientific rigor of solution evaluation: It innovatively combines eye-tracking technology and the DIDF weighting method. Eye-tracking captures users' subconscious visual preferences, while the DIDF method provides objective weights for users' subjective ratings. The combination of these two elements constructs a multi-dimensional evaluation system that integrates subjective and objective factors, making the final solution selection more comprehensive and reliable.

[0028] 4. Improved overall design efficiency and quality: The entire framework transforms the original subjective and fragmented design process into a data-driven and intelligent decision-making system, which significantly shortens the design cycle and produces new energy vehicle designs that better meet the emotional preferences of target users and have stronger market competitiveness. Attached Figure Description

[0029] Figure 1 This is a flowchart of a new energy vehicle styling design method according to an embodiment of the present invention;

[0030] Figure 2 This is a technical roadmap of a new energy vehicle styling design method according to an embodiment of the present invention;

[0031] Figure 3 This is a schematic diagram of the disassembly of a new energy vehicle in an embodiment of the present invention;

[0032] Figure 4 This is a structural schematic diagram of a new energy vehicle styling design system provided in an embodiment of the present invention.

[0033] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0035] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0036] Example 1

[0037] Please see Figure 1 The flowchart illustrates a new energy vehicle styling design method provided in this embodiment of the invention, and can be found in the reference section. Figure 2 This is a technical roadmap for a new energy vehicle styling design method proposed in this invention. The method may include the following steps:

[0038] S1: Based on a pre-set sample set of new energy vehicles and user Kano questionnaire survey data targeting emotional words, the interval type 2 trapezoidal fuzzy Kano model (IT2Tr-FKM) is used to calculate and determine the comprehensive importance weight of multiple emotional words, and to select a pre-set number of key emotional words with the highest weight.

[0039] Specifically, addressing the shortcomings of existing fuzzy Kano models (FKM) in handling the subjectivity, uncertainty, and individual differences of evaluation information in user demand classification, this step uses Interval Type 2 Trapezoidal Fuzzy Numbers (IT2TrFN) to handle the uncertainty of evaluation information in the process of perceptual evaluation, and constructs an IT2Tr-FKM integrated model. This process includes two main stages: classification and weighting of perceptual terms.

[0040] 1. Definition and application of Interval Type 2 Trapezoidal Fuzzy Number (IT2TrFN)

[0041] Interval Type 2 Fuzzy Sets (IT2FS) introduce the uncertainty of the membership function itself on the basis of traditional fuzzy sets, allowing decision-makers to consider the uncertainty of the shape / position of the membership degree and the membership function. IT2TrFN describes the evaluation information through primary membership degree and secondary membership degree, as shown in Equation (1).

[0042] Definition 1:

[0043]

[0044] In the formula, Represent an interval type-2 trapezoidal fuzzy set. express The lower bound of the trapezoidal fuzzy number, express The upper bound of the trapezoidal fuzzy number, , It serves as the starting point of the trapezoid with both upper and lower bounds; , , , The inflection points of the trapezoid with upper and lower bounds; , The endpoint of the trapezoid with upper and lower bounds; , express The upper and lower bounds of the trapezoidal height. These ten parameters can be freely adjusted according to actual conditions, providing a high degree of flexibility.

[0045] Definition 2: Distance Calculation Formula

[0046] Let set and set Let the two non-negative IT2TrFN on the domain X be as shown in formulas (2)-(3).

[0047]

[0048]

[0049] Represents a set The height of the upper bound trapezoid, Represents a set The height of the lower bound trapezoid;

[0050] To measure the difference between two IT2TrFNs, fuzzy sets and Distance between The calculation formula is defined as shown in formula (4):

[0051]

[0052] In the formula: , Indicates the height of the lower bound trapezoid. This indicates the height of the upper bound trapezoid.

[0053] 2. Classification of Emotional Vocabulary

[0054] First, extract the key information through interviews and methods such as the KJ method. N A word of feeling E i ( i=1,2,...,N ) 。 Then, the language scale set of the Kano questionnaire (such as dissatisfied, acceptable, indifferent, expected, satisfied) was quantified into IT2TrFN and used standard language scales. express.

[0055] The user's positive and negative answers to each emotional word are represented using IT2TrFN. The distance between the user's fuzzy evaluation value and each standard linguistic scale C is calculated using formula (4), and the membership degree of the evaluation value to the left and right adjacent linguistic scales is calculated according to formula (5).

[0056]

[0057] In the formula: Represents fuzzy evaluation value x left adjacent standard language scale membership function, Represents fuzzy evaluation value x right adjacent standard language scale membership function, Indicates the use of calculation The distance function, Indicates the use of calculation In the classification of sensory vocabulary, the distance function, the standard language scale C={C1,C2,C3,C4,C5}, is quantized into an interval type II trapezoidal fuzzy number. Indicates the first i The quantization value of a standard language scale. This indicates the first number in the Kano questionnaire. One scale, The sum of the membership degrees of the evaluation value to its adjacent left and right scales is... The membership values ​​for other scales are all 0. d(˙) represents the function used to calculate the membership, and d represents the set of memberships for the problem.

[0058] The membership sets of the positive and negative problems are calculated using formula (5), and then respectively... P i and U i This is indicated. Next, a 5×5 fuzzy evaluation matrix is ​​generated based on the preference information provided in the survey questionnaire. L i As shown in formula (6):

[0059]

[0060] Repeat the above steps to count. H The Kano category membership vector of each participant in the evaluation of this emotional word, and the overall category membership vector. As shown in formula (7):

[0061]

[0062] In the formula: Indicates the first h The Kano category membership vector of each evaluator for the emotional word i. Representing emotional words i The Kano ratios corresponding to the six different Kano categories.

[0063] The Kano category corresponding to the highest membership value is the category to which the sensory word belongs. The set of category membership vectors constitutes the category distribution matrix of the sensory word. D As shown in formula (8):

[0064]

[0065] In the formula, N represents the total number of emotional words, and K represents the total number of Kano categories.

[0066] 3. Empowering Emotional Vocabulary

[0067] First, objective importance is calculated using a modified CRITIC weighting method. The Gini coefficient is used instead of the standard deviation to measure the distributional differences (contrast) of emotional words across different Kano categories, as shown in formula (9). Combining the conflict between words, the objective importance of the i-th emotional word is calculated using formula (10). .

[0068] In this study, the Gini coefficient is used instead of the standard deviation to measure the distribution of class membership information. express:

[0069]

[0070] In the formula, Indicates the first i The first emotional word belongs to the first p Membership degree of each Kano category, Indicates the first The closer the Gini coefficient of a word is to 1, the more unbalanced the information distribution and the more information it contains.

[0071] The first can be calculated using formula (10). i The objective importance of a single emotional word :

[0072]

[0073] In the formula: Indicates the first i The information capacity of a single emotional word .

[0074] Therefore, the objective importance of sensory vocabulary is... .

[0075] Secondly, the importance of satisfaction was calculated: Experts were invited to conduct a fuzzy evaluation of the importance of the Kano category itself, considering the different contributions of different categories of emotional words to customer satisfaction. Using the minimum deviation method, combined with the Kano category membership vector of the emotional words, the importance of satisfaction for the emotional words was calculated according to formulas (11)-(12). .

[0076] Kano category importance is expressed as a vector. This indicates that an expert fuzzy evaluation matrix is ​​established to determine the importance of Kano categories, and then the satisfaction importance of emotional words is calculated using the minimum deviation method.

[0077] Based on the Kano category membership vector of sensory vocabulary To obtain the absolute satisfaction importance of emotional vocabulary :

[0078]

[0079] The absolute satisfaction importance of emotional words is normalized, as shown in formula (12):

[0080]

[0081] Therefore, the satisfaction importance of emotional vocabulary is: , Indicates the first i The importance of satisfaction for each emotional word.

[0082] Finally, the overall importance is calculated by weighting and integrating the objective importance and the satisfaction importance, and obtaining the overall importance according to formula (13). .

[0083]

[0084] The overall importance vector of emotional words is represented as follows: .

[0085] Finally, all emotional words were ranked according to their overall importance, and the words with the highest weight were selected as key emotional words.

[0086] For example, to construct a representative sample set of new energy vehicles (NEVs), 221 images of different brands and models of NEVs were collected from mainstream automotive websites, professional automotive magazines, and other multimedia channels. To ensure data quality, all images were screened, removing those with low resolution, poor angles, or severe interference from complex lighting environments, ultimately retaining 110 high-quality, multi-angle NEV sample images. These 110 images were standardized using Adobe Photoshop, including uniform size (15cm x 15cm), background removal, and color balance adjustment, creating standardized sample cards. Next, to explore deeper emotional needs of users, 10 participants from diverse backgrounds (including designers, students, and ordinary consumers) were invited to participate in an in-depth interview based on the Evaluative Construction Map (EGM) method. During the interviews, the participants were shown the aforementioned NEV sample cards, and by comparing different vehicle forms, they were guided to freely express their likes and dislikes, as well as the original reasons for these feelings. After the interviews, the KJ method (Affinity Map Method) was used to systematically organize the large amount of raw data collected. By summarizing and merging similar or related descriptions, a three-tiered EGM evaluation framework was constructed: the upper tier consists of 52 evaluation items (i.e., emotional terms), the middle tier consists of 16 evaluation items (original reasons), and the lower tier consists of 200 evaluation items (specific morphological conditions). After further refinement and selection, 14 of the most representative upper-tier emotional terms were identified, including "simple," "unique," "dynamic," "technological," "advanced," and "reserved," which form the basis for subsequent quantification of emotional needs.

[0087] S2: Deconstruct the morphological features of the new energy vehicle sample set to obtain multidimensional morphological feature data; based on the multidimensional morphological feature data and the user evaluation values ​​for the key perceptual words, train an extreme gradient boosting model (XGBoost) optimized by the Hippo Optimization Algorithm (HO) to establish a mapping model between morphological features and key perceptual words.

[0088] Specifically, this step uses the HO-XGBoost model to establish a non-linear mapping relationship between design features and user subjective evaluation.

[0089] 1. XGBoost model

[0090] XGBoost makes predictions by integrating multiple decision trees. For input samples Its predicted value is the sum of all L decision trees. The output is accumulated, then the first... The predicted value for each sample is given by formula (14):

[0091]

[0092] XGBoost's optimization objective includes a loss function and a regularization term, and is approximated using a second-order Taylor expansion. In the... In the next iteration, the objective function is optimized. for:

[0093]

[0094] in The first derivative of the loss function. The loss function is expressed as follows: The first-order partial derivative, It is a loss function that represents the difference between the actual value and the predicted value. Indicates the first t Tree samples x i The predicted value, The second derivative of the loss function. The loss function is expressed as follows: The second-order partial derivatives of . It is the first The number of leaf nodes in a tree. The score is for the leaf nodes. and It is the regularization coefficient that controls the complexity of the model.

[0095] 2. Hippo Optimization Algorithm (HO)

[0096] The HO algorithm is used to optimize the hyperparameters (such as the number of trees, depth, learning rate, etc.) of the XGBoost model. It finds the optimal combination of hyperparameters by simulating the "swimming phase" (global exploration) and "walking phase" (local exploration) of a hippopotamus population.

[0097] Swimming phase: Simulate hippo group behavior to generate candidate hyperparameter combinations, as shown in formula (16). Individuals learn from the globally optimal combination, while introducing randomness to explore new spaces.

[0098] Initial generation was achieved by simulating hippopotamus group behavior. Group of hyperparameter combinations This ensures that each dimension of each hyperparameter set is within a preset range. Each hyperparameter combination is analogous to an individual hippopotamus; the current hyperparameter combination... Refer to the globally optimal combination ,formula This reflects the behavior of individuals learning from successful individuals and approaching the optimal solution, while Then, random factors are introduced to simulate the hippopotamus's autonomous exploration of new directions, so that the algorithm has both directionality and exploration when searching the hyperparameter space, preventing it from getting trapped in local optima. Then, candidate hyperparameter combinations are generated by simulating the behavior of the hippopotamus group, and calculated as in formula (16).

[0099]

[0100] in Indicates the first j New candidate values ​​for each hyperparameter It indicates that it is the first j The current values ​​of each hyperparameter. Indicates the first j The global optimal value of each hyperparameter.

[0101] a and b The value range is generally between [0,1]. This value is chosen to balance the algorithm's exploration and utilization capabilities, controlling the search direction. An empirical value is taken as... a =0.5, b =0.3. , These are uniformly distributed random numbers.

[0102] Model evaluation: The performance of each hyperparameter combination is evaluated using mean squared error (MSE) as the fitness function, as shown in Equation (17).

[0103] For the generated candidate hyperparameter combinations An XGBoost model is constructed and trained on the training set. The model performance is then evaluated by calculating the Mean Squared Error (MSE) on the validation set. Assume the validation set contains... There are 10 samples, and the true value of each sample is... The predicted value is The mean square error is calculated as shown in formula (17):

[0104]

[0105] If the new candidate hyperparameter combination makes If it is better, then update the global optimal solution.

[0106] Walking phase: Fine-tuning is performed near the current global optimum for a more refined search, as shown in formula (18).

[0107] During the walking phase, the global optimal solution is fine-tuned according to the following formula to generate a new combination of hyperparameters. .

[0108]

[0109] in, Indicates the first hyperparameter in the updated hyperparameter combination The values ​​of the hyperparameters. This represents the value of the j-th hyperparameter in the current globally optimal hyperparameter combination. Is The first uniformly distributed random variable within the range Each possible value. Step size. As a dynamic parameter, its value typically ranges from (0,1). U represents a uniform distribution. This represents a continuous, uniform distribution in the interval [-1, 1]. Its value decays with the number of iterations. Assigning a larger value in the early stages of iteration allows the algorithm to quickly search the solution space with large step sizes, accelerating the convergence process. Gradually decreasing the value as iterations progress facilitates fine-tuning of hyperparameters to improve solution accuracy. For example, ,in The total number of iterations, initial value .

[0110] This process is repeated iteratively, continuously executing the swimming phase, model evaluation, and walking phase of the HO algorithm, until the maximum number of iterations is reached. Convergence. Finally, the optimal combination of hyperparameters is determined, and a high-precision XGBoost prediction model is trained.

[0111] For example:

[0112] IT2Tr-FKM Questionnaire Design and Data Collection: Based on the 14 subjective terms selected in the previous example, a questionnaire based on the Interval Type 2 Trapezoidal Fuzzy Kano Model (IT2Tr-FKM) was designed. Unlike traditional Kano questionnaires, this questionnaire precisely quantifies the user's five subjective feelings—"Dissatisfied (D)", "Acceptable (W)", "Indifferent (N)", "As It Should Be (M)", and "Satisfied (L)"—as interval Type 2 trapezoidal fuzzy numbers (IT2TrFN). These fuzzy numbers can simultaneously capture the hesitation and uncertainty of the evaluation. The questionnaire includes positive questions ("If this car has XX characteristic, how would you feel?") and negative questions ("If this car does not have XX characteristic, how would you feel?") for each subjective term. The questionnaire was distributed to 60 graduate students majoring in industrial design, who have a deeper understanding of automotive design. 49 valid questionnaires were ultimately collected.

[0113] Kano Attribute Classification of Emotional Vocabulary: The data from 49 collected questionnaires were processed. Taking the emotional vocabulary "simple" as an example, if a respondent's answer to a positive question has the following fuzzy values: [(7,7.8,7.8,9.2;0.53),(5.2,7.4,8.6,9.8;1)], and the answer to a negative question has the following values: [(0.2,1.8,1.8,3.2;0.53),(0,1.2,2.6,4.8;1)]. First, the distance between the fuzzy value of the answer and the fuzzy values ​​of the five standard language scales is calculated according to formulas (4) and (5), thereby determining the membership degree of the answer to the adjacent language scales. Then, the membership degree is transformed into a 5x5 fuzzy evaluation matrix according to formula (6). By comparing this matrix with the pre-set Kano category evaluation table, the membership degree of the word "simple" to the attractive attribute (A), one-dimensional attribute (O), essential attribute (M), indifferent attribute (I), and reverse attribute (R) can be calculated, forming a membership degree vector. The membership degree vectors of all 49 respondents are summed and averaged to obtain the overall Kano category membership degree vector of the word "simple" according to formula (7). This process is repeated to finally obtain the distribution matrix of 14 emotional words on the six Kano categories of A, O, M, I, R, and Q (Q is a questionable result). According to the principle of highest membership degree, the main Kano attributes of each word are determined. For example, "introverted" is classified as an essential attribute, and "refined" is classified as a one-dimensional attribute.

[0114] Weighting of the overall importance of emotional vocabulary:

[0115] Objective Importance Calculation: To measure the importance of each emotional word in the user's demand structure, an improved CRITIC method (ICRITIC) is adopted. This method is based on the Kano category distribution matrix mentioned above. According to formula (9), the Gini coefficient is used to replace the standard deviation in the traditional CRITIC method to more accurately measure the imbalance (i.e., contrast) of the distribution of each emotional word across the six Kano categories. Subsequently, the conflict between each word is calculated according to formula (10), and the contrast and conflict are combined to calculate the objective importance vector Wo of the 14 emotional words.

[0116] Satisfaction Importance Calculation: To incorporate the impact of user satisfaction, 49 experts were invited to conduct fuzzy evaluations of the importance of the five Kano categories: Aesthetic (A), One-Dimensional (O), Essential (M), Indifferent (I), and Reverse (R). Then, based on the minimum deviation method defined in formulas (11) and (12), the experts' fuzzy evaluations were combined with the Kano category membership degrees of each emotional term to calculate the satisfaction importance vector W for the 14 emotional terms. S .

[0117] Overall Importance Calculation: Finally, according to formula (13), the objective importance Wo and the satisfaction importance W are combined by weighted integration. S The results are fused to obtain the final comprehensive importance vector W. The calculation results show that "introverted (K7)" and "refined (K7)" are important. 11 The three emotional terms "(K6)" and "advanced" scored the highest in overall importance. Therefore, these three terms were identified as key emotional terms for subsequent design and modeling.

[0118] S3: Using the mapping model, perform perceptual evaluation value prediction on all morphological feature combinations within a preset range, determine the optimal morphological feature combination corresponding to the key perceptual words, and generate at least one candidate design scheme based on the optimal morphological feature combination.

[0119] For example:

[0120] Morphological feature deconstruction and training data preparation: Refer to Figure 3 A morphological analysis method was used to systematically dissect the appearance of NEVs. The overall vehicle form was divided into five key design parts: X1 - windows, X2 - headlights, X3 - grille, X4 - wheels, and X5 - rearview mirrors. Multiple specific type features were defined for each part and digitally encoded. For example, there are five types of windows, encoded as 1 to 5. The features of each of the five parts were analyzed and encoded for each of the 110 NEV samples from step one. Simultaneously, 100 participants were invited to rate these 110 sample images using a seven-order Likert scale based on the three key perceptual terms "reserved," "refined," and "high-end." The average of all ratings was used to construct a training dataset containing 110 records. The input for each record was a 5-dimensional morphological feature encoding (e.g., [3,2,4,1,5]), and the output was a 3-dimensional perceptual evaluation value (e.g., [5.8, 6.2, 6.5]).

[0121] Training the HO-XGBoost Mapping Model: To capture the complex nonlinear relationship between morphological features and sensory evaluations, this embodiment employs an Extreme Gradient Boosting (XGBoost) model optimized by the Hippo Optimization (HO) algorithm. The dataset constructed in the previous step is divided into training and test sets in an 8:2 ratio. The population size of the Hippo Optimization algorithm is set to 5, and the maximum number of iterations is set to 100. The optimization objective of the HO algorithm is to find a set of hyperparameters that minimizes the mean squared error (MSE) of the XGBoost model prediction. These hyperparameters include the number of trees, the maximum tree depth, and the learning rate. During the iteration process, the HO algorithm simulates the "swimming phase" (global exploration, searching for new possible solution spaces) and the "walking phase" (local development, fine-tuning the search near the current optimal solution) of the hippo population, efficiently updating the hyperparameter combination. After training, an optimized, high-precision HO-XGBoost prediction model for each key sensory word is obtained.

[0122] Optimal solution generation: Using three pre-trained HO-XGBoost models, a comprehensive perceptual score prediction is performed for all possible combinations of morphological features. Since each part has 5 types, there are a total of 5... ^5 =3125 possible combinations. Using these 3125 combinations as input, predict their scores on three dimensions: "concise," "refined," and "advanced." The prediction results determine the morphological feature combination with the highest score in each dimension, for example:

[0123] The combination that scored highest on the "introverted" dimension was [5, 5, 1, 3, 2].

[0124] The combination that scored highest on the “refined” dimension was [5, 5, 3, 4, 5].

[0125] The combination that scored highest in the "advanced" dimension was [5, 2, 5, 1, 4].

[0126] Based on the optimal combination of features driven by these three sets of data, combined with their own design experience and aesthetic knowledge, the designers creatively drew sketches of multiple candidate design schemes, and used generative AI tools such as Midjourney to quickly render and generate high-quality visual candidate design schemes.

[0127] S4: Eye-tracking technology is used to obtain the user's visual perception data of the candidate design schemes, and the Discrete Information Data Fluctuation Weighting Method (DIDF) is used to perform objective quantitative analysis on the candidate design schemes. Combining the visual perception data and the results of the objective quantitative analysis, the final new energy vehicle styling design scheme is determined from the candidate design schemes.

[0128] Specifically, this step combines subjective and objective evaluation methods to achieve a comprehensive and scientific assessment of candidate solutions.

[0129] 1. Eye-tracking technology (subjective subconscious evaluation)

[0130] This study uses eye-tracking experiments to verify whether products align with customer emotional preferences and to identify the focal areas of user visual attention. By recording and analyzing eye-tracking metrics (such as total fixation time and first fixation time) when users observe design options, we can understand users' "first visual response" and determine which design options quickly attract users' attention. However, a high visual perception sequence does not necessarily equate to a superior design, as it can be influenced by non-core design factors such as color and location. To address this limitation, an objective DIDF weighting method is introduced.

[0131] 2. Discrete Information Data Fluctuation Weighting Method (DIDF) (Objective Quantitative Analysis)

[0132] DIDF is an objective weighting method that combines indicator independence and data volatility to process expert or user ratings of candidate solutions. Its core steps are as follows:

[0133] Step 1: Calculate the coefficient of deviation for each indicator The coefficient of variation reflects the consistency or volatility of the evaluation data.

[0134]

[0135] In the formula: Standard deviation, This is the average value.

[0136] Step 2: Calculate the independent information ratio for each indicator.

[0137] Calculate the independent information ratio of each indicator By performing multiple linear regression with a certain indicator as the dependent variable and the other indicators as independent variables, the coefficient of determination of that indicator is obtained. ,but:

[0138]

[0139] Here Indicates the first The goodness of fit of each index. In the limiting case, if the goodness of fit... A value of 0 indicates that the indicator provides complete information; if A value of 1 indicates that the indicator cannot provide any independent information and can be deleted.

[0140] Step 3: Calculate the pure information content of each indicator .

[0141] Each , Standardization is performed to obtain standardized data. , Then calculate the pure information content. .

[0142]

[0143]

[0144]

[0145] in It is the first Coefficient of variation of each indicator before standardization It is the first The ratio of independent information for each indicator before standardization. Standardization eliminates differences in the dimensions of different indicators, and then the two ratios are multiplied together to obtain the pure information content.

[0146] Step 4: Calculate the weight of each indicator .

[0147]

[0148] Step 5: Standardization and weighted summarization of evaluation indicators.

[0149] The evaluation indicators are standardized, and then the weights obtained in step four are used as a basis. The results are then weighted and aggregated to obtain the final evaluation result.

[0150] For example:

[0151] Eye-tracking experiment (subjective subconscious evaluation): From the numerous schemes generated in the previous step, the top 4 designs based on their comprehensive performance across the three key sensory vocabulary dimensions were selected, totaling 12 schemes (marked as A1 to A). 12 Five users with NEV driving experience were recruited to participate in an eye-tracking experiment. In the experiment, renderings of 12 scenarios were integrated into a single large image, and users were required to observe freely for 3 minutes. The Tobii Pro Glasses 2 wearable eye tracker was used to record the users' eye movements in real time. After the experiment, the eye-tracking data was analyzed using Tobii Pro Lab software, extracting seven key eye-tracking metrics, including total fixation duration, average fixation duration, fixation count, and first fixation duration. The seven metrics for each scenario were standardized and averaged to quantify the user's subconscious visual attention to each scenario. The results showed that the user's visual perception priority was: A3 > A. 10 >A12 >A9>... indicates that schemes A3 and A... 10 It is most likely to attract the user's visual attention.

[0152] DIDF Weighting (Objective Quantitative Evaluation): To obtain a more objective evaluation, 32 designers with industrial design backgrounds were invited to score the "introversion," "refinement," and "sophistication" of 12 candidate schemes using a seven-point Likert scale. After collecting all the scoring data, the Discrete Information Data Fluctuation Weighting (DIDF) method was used for processing. According to formulas (19)-(24), the coefficient of variation (reflecting the consistency of evaluation) and the ratio of independent information (reflecting the uniqueness of evaluation data) were calculated for the scoring data of each scheme under each sensory term. The pure information content of each scheme was calculated by combining these two indicators and normalized to obtain the objective weight of each scheme under the three sensory dimensions. The weights of the three dimensions were summed to obtain the comprehensive evaluation score and ranking of the 12 schemes. The results showed that the comprehensive ranking was: A3>A 10 >A9>A 12 >... indicates that schemes A3 and A 10 It also possesses the highest overall quality in the eyes of experts.

[0153] The final solution was determined by combining the results of the two evaluation methods: eye-tracking experiments revealed users' subconscious preferences, while the DIDF weighting method provided an objective quantitative ranking based on expert judgment. Both methods consistently ranked schemes A3 and A... 10 Ranked first and second, after final discussion and deliberation by the design team, the A3 scheme, with the best overall performance, was selected as the final design scheme for the attractive form of the new energy vehicle. Finally, professional 3D modeling software such as Rhino 3D was used to accurately digitally model the A3 scheme, laying the foundation for subsequent engineering development.

[0154] To verify the superiority of the HO-XGBoost model used in this invention, its performance was compared with four other commonly used artificial intelligence prediction models, including GRU-Attention, CNN, RBF, and LSTM. All models used the same training and test sets. The comparison results show that, on both the training and test sets, the HO-XGBoost model has the highest coefficient of determination (R²) (approximately 0.97-0.99) and the lowest root mean square error (RMSE) when predicting the three subjective evaluation values ​​of "Reserved," "Exquisite," and "High-class." This result strongly demonstrates that the HO-XGBoost model has the highest prediction accuracy and best generalization ability, and can most effectively establish the complex mapping relationship between NEV morphological features and user subjective needs.

[0155] Example 2

[0156] Please see Figure 4 The diagram shown is a structural schematic of a new energy vehicle styling design system according to the second embodiment of this application. The system includes the following key modules:

[0157] The Emotional Needs Quantification Module 100 is configured to use a preset new energy vehicle sample set and user Kano questionnaire survey data targeting emotional words. It adopts an interval type 2 trapezoidal fuzzy Kano model to calculate and determine the comprehensive importance weight of multiple emotional words, and selects a preset number of key emotional words with the highest weight.

[0158] The mapping model construction module 200 is configured to deconstruct the morphological features of the new energy vehicle sample set to obtain multidimensional morphological feature data; based on the multidimensional morphological feature data and the user evaluation values ​​for the key perceptual words, a limit gradient boosting model optimized by the Hippo optimization algorithm is trained to establish a mapping model between morphological features and key perceptual words.

[0159] The optimal solution generation module 300 is configured to use the mapping model to predict the perceptual evaluation value of all morphological feature combinations within a preset range, determine the optimal morphological feature combination corresponding to the key perceptual words, and generate at least one candidate design solution based on the optimal morphological feature combination.

[0160] The comprehensive evaluation and screening module 400 is configured to use eye-tracking technology to acquire the user's visual perception data of the candidate design schemes, and use the discrete information data fluctuation weighting method to perform objective quantitative analysis on the candidate design schemes. Combining the visual perception data and the objective quantitative analysis results, the final new energy vehicle styling design scheme is determined from the candidate design schemes.

[0161] The new energy vehicle styling design system in this application embodiment can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, in-vehicle electronic devices, wearable devices, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), etc. This application embodiment does not impose specific limitations.

[0162] The new energy vehicle styling design system in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit it.

[0163] The new energy vehicle styling design system provided in this application embodiment can achieve... Figure 1 The various processes of implementing a new energy vehicle styling design method in the method embodiments are not described in detail here to avoid repetition.

[0164] Optionally, this application embodiment also provides an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the various processes of the above-described embodiment of a new energy vehicle styling design method and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0165] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described embodiment of a new energy vehicle styling design method and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0166] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0167] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0168] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0169] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A new energy vehicle modeling design method, characterized in that, The method comprises the following steps: Based on the preset new energy vehicle sample set and the user Kano questionnaire survey data for the emotional vocabulary, an interval 2-type trapezoidal fuzzy Kano model is used to calculate and determine the comprehensive importance weight of multiple emotional vocabulary, and a preset number of key emotional vocabulary with the highest weight are screened out; The step of using the interval 2-type trapezoidal fuzzy Kano model comprises: quantifying the language scale set in the Kano questionnaire into interval 2-type trapezoidal fuzzy numbers; the user uses the interval 2-type trapezoidal fuzzy numbers to answer the positive and negative questions of the emotional vocabulary, and obtains fuzzy evaluation values; the distance between the fuzzy evaluation values and the fuzzy numbers corresponding to each language scale is calculated to determine the membership degree of adjacent language scales, and then a Kano category membership degree vector is generated to classify and calculate the weight of the emotional vocabulary; The step of calculating and determining the comprehensive importance weight of multiple emotional vocabulary comprises: based on the improved CRITIC weight method, the Gini coefficient is used to replace the standard deviation to calculate the distribution difference of each emotional vocabulary in different Kano categories to determine the objective importance; based on the least deviation method, the fuzzy evaluation of the importance of different Kano categories by experts is combined to determine the satisfaction importance of each emotional vocabulary; the objective importance and the satisfaction importance are weighted and integrated to obtain the comprehensive importance weight; The morphological characteristics of the new energy vehicle sample set are deconstructed to obtain multi-dimensional morphological characteristic data; based on the multi-dimensional morphological characteristic data and the user evaluation values for the key emotional vocabulary, an extreme gradient boosting model optimized by the Hippocampus optimization algorithm is trained to establish a mapping model between the morphological characteristics and the key emotional vocabulary; The step of training an extreme gradient boosting model optimized by the Hippocampus optimization algorithm comprises: taking the hyperparameters of the XGBoost model, including the number of trees, the depth of the tree and the learning rate, as individuals in the Hippocampus population; the hyperparameter combinations are iteratively updated by simulating the swimming stage and the walking stage of the Hippocampus population, wherein the swimming stage is used for global exploration and the walking stage is used for local development; the mean square error is used as the fitness function to evaluate the performance of the XGBoost model corresponding to each hyperparameter combination until a preset iteration number is reached or the model performance converges, thereby determining the optimal hyperparameter combination; Using the mapping model, the emotional evaluation values of all morphological characteristic combinations within a preset range are predicted to determine the optimal morphological characteristic combination corresponding to the key emotional vocabulary, and at least one candidate design scheme is generated according to the optimal morphological characteristic combination; Eye tracking technology is used to obtain the visual perception data of the user for the candidate design scheme, and a discrete information data fluctuation weighting method is used to objectively quantify and analyze the candidate design scheme, and the final new energy vehicle styling design scheme is determined from the candidate design scheme based on the visual perception data and the objective quantitative analysis result.

2. The method of claim 1, wherein, The step of deconstructing the morphological characteristics of the new energy vehicle sample set comprises: Adopting morphological analysis method, the new energy vehicle form is divided into multiple design parts, the design parts include windows, headlights, air intake grille, hub and rearview mirror; Define multiple types of features for each design part, and encode the type features to form multi-dimensional form feature data.

3. The method of claim 1, wherein, The step of obtaining visual perception data of the candidate design scheme by using eye tracking technology includes: Record the eye movement trajectory of the user when observing the candidate design scheme, and count the eye movement indicators including total fixation time, average fixation time, fixation count and first fixation time; Determine the visual perception priority order of each candidate design scheme based on the standardized average value of the eye movement indicators.

4. The method of claim 1, wherein, The step of objectively quantifying and analyzing the candidate design scheme by using discrete information data fluctuation weighting method includes: Collect the scores of multiple evaluators on each candidate design scheme for the key emotional words; Calculate the deviation coefficient and independent information ratio of each candidate design scheme under the score of each key emotional word; Calculate the pure information quantity of each candidate design scheme by combining the deviation coefficient and independent information ratio, and normalize it to obtain the objective weight of each candidate design scheme; Sort each candidate design scheme based on the objective weight.

5. A new energy vehicle modeling design system, characterized in that, It includes: The emotional demand quantification module is configured to adopt the interval 2-type trapezoidal fuzzy Kano model based on the preset new energy vehicle sample set and the user Kano questionnaire survey data for emotional words, calculate and determine the comprehensive importance weight of multiple emotional words, and select a preset number of key emotional words with the highest weight; the interval 2-type trapezoidal fuzzy Kano model includes: quantifying the language scale set in the Kano questionnaire into interval 2-type trapezoidal fuzzy numbers; the user uses the interval 2-type trapezoidal fuzzy numbers to answer the positive and negative questions of emotional words, and obtains fuzzy evaluation value; by calculating the distance between the fuzzy evaluation value and the fuzzy number corresponding to each language scale, the membership degree of adjacent language scale is determined, and then the Kano category membership degree vector is generated to classify and weight the emotional words; the calculation and determination of the comprehensive importance weight of multiple emotional words includes: based on the improved CRITIC weight method, the Gini coefficient is used instead of the standard deviation to calculate the distribution difference of each emotional word in different Kano categories, so as to determine its objective importance; based on the least deviation method, combined with the fuzzy evaluation of experts on the importance of different Kano categories, the satisfaction importance of each emotional word is determined; the objective importance and the satisfaction importance are weighted and integrated to obtain the comprehensive importance weight; The mapping model construction module is configured to deconstruct the shape features of the new energy vehicle sample set to obtain multi-dimensional shape feature data; based on the multi-dimensional shape feature data and the user evaluation value for the key perceptual words, an extreme gradient boosting model optimized by a hippo optimization algorithm is trained to establish a mapping model between the shape features and the key perceptual words; the training of the extreme gradient boosting model optimized by the hippo optimization algorithm includes: taking the hyperparameters of the XGBoost model, including the number of trees, the depth of the tree and the learning rate, as individuals in a hippo population; the hyperparameter combinations are iteratively updated by simulating the swimming stage and the walking stage of the hippo population, wherein the swimming stage is used for global exploration and the walking stage is used for local development; the mean square error is used as the fitness function to evaluate the performance of the XGBoost model corresponding to each hyperparameter combination until a preset iteration number is reached or the model performance converges, thereby determining the optimal hyperparameter combination; The optimal scheme generation module is configured to use the mapping model to predict the perceptual evaluation value of all shape feature combinations within a preset range, determine the optimal shape feature combination corresponding to the key perceptual words, and generate at least one candidate design scheme according to the optimal shape feature combination; The comprehensive evaluation and screening module is configured to use eye tracking technology to obtain visual perception data of the user on the candidate design scheme, and use a discrete information data fluctuation weighting method to objectively and quantitatively analyze the candidate design scheme, and combine the visual perception data and the objective and quantitative analysis result to determine the final new energy vehicle modeling design scheme from the candidate design scheme.

6. An electronic device, comprising: A processor, a memory, and a program or instructions stored on the memory and executable on the processor are included, and the program or instructions are executed by the processor to implement the steps of the new energy vehicle modeling design method according to any one of claims 1-4.

7. A readable storage medium characterized by, The readable storage medium stores programs or instructions, and the programs or instructions are executed by the processor to implement the steps of the new energy vehicle modeling design method according to any one of claims 1-4.

Images