Evaluation of electric car styling based on analytic hierarchy process and Kansei engineering: A study on mainstream Chinese electric car brands

Abstract

To evaluate the fuzzy and uncertain factors brought by human's emotional sensibilities during the design of new energy vehicle models, a scientific design evaluation method was adopted. This study combines the Analytic Hierarchy Process (AHP) and Kansei Engineering to assess the design of new energy vehicle models. It uses both objective criteria and subjective user perceptions. First, it quantified the design imagery vocabulary through Kansei Engineering and identified the target imagery vocabulary. Secondly, the AHP method was used to establish a relationship between design features and imagery vocabulary. Thrid, the study created an evaluation index system for design imagery, and adopted the fuzzy comprehensive evaluation matrix to rank the representative vehicles and brands in the Chinese new energy vehicle market. Based on the weight analysis results, the vehicle functionality (whether it is safe and reliable), the elegance of design, and the intricacy of vehicle structure were identified as the top three aspects that affect the impression of new energy vehicles. Therefore, we need to focus on these aspects during the design of new energy vehicles. Additionally, by integrating objective evaluation and emotional analysis results, we assessed the design for each new energy vehicle brand. The results of this study are important for developing the new energy vehicle market.

1. Introduction

Global energy system faces an unprecedented crisis and there has been a growing increase in concern about environmental issues. Electric vehicles (EVs) have gained wider attention and importance as an eco-friendly and sustainable mode of transportation. The Chinese government has introduced various policies to promote the development of the new energy industry. Driven by policy support and market demand in 2022, the market penetration rate of new energy vehicles reached nearly 26% for the entire year [1] (Fig. 1). While China has become a significant market for the production and sales of new energy vehicles globally. The development of the new energy vehicle industry has yet achieved substantial breakthroughs [2]. In the study, Chen, Y identified the challenges faced by the new energy vehicle industry, including a low willingness to popularize, a short path to popularization, and poor popularization benefits [3]. To address these issues, Jiang, Z. found that the Chinese government has made preferential policies in pilot cities for new energy vehicles and increased subsidies. It has a significant positive impact on the quantity of technological innovations in new energy vehicle enterprises through partial mediating effects [4]. Intelligent vehicle testing has been a focal point in smart vehicles. Zhu, B proposed an intelligent vehicle testing critical scenario search method based on the Social Cognitive Optimization (SCO) algorithm. The test results indicate that this method can enhance search efficiency and coverage of critical scenarios [5]. Furthermore, automotive driving safety remains a focus in automotive research. J., X introduced a hybrid EFL (Eye-Tracking, Facial Expression Analysis, and Lidar) recommendation strategy to enhance driving safety. By extracting visual features from human dynamic vision, the proposed method demonstrated its potential value in various collected driving tasks [6,7]. However, as an integral part of electric vehicles, the car's exterior will influence consumer's purchasing decisions. The aesthetic appeal of the vehicle affect consumers' first impression and is closely related to user experience, brand image, and market competitiveness [8]. New energy vehicles, as emerging products, have distinct differences in their internal and mechanical structures compared to traditional fuel-powered cars. These differences inherently influence their exterior designs, resulting in variations between the aesthetics of traditional and new energy vehicles. Hence, we need to follow the trends and design new energy vehicles that cater to consumer demands [9]. As the functions and performance of the product becomes homogenized, the design technology for product exterior quality has garnered significant attention from both the academic and industrial sectors. To meet the growing demands of consumers for diverse and personalized. Designs, new technologies are used. For companies, product design is a valuable tool to gain a competitive edge. Therefore, design analysis helps brands to increase competitiveness. Scholar Hyun [10] employs a mixed-quantitative approach to assess certain design elements' weight factors from four aspects: style analysis, mixed quantification, data structure configuration, and the identification of unique design elements. Liu [11] has developed a car styling emotional design method that combines Kansei Engineering with online review analysis. This method integrates Kansei Engineering with online comment analysis to predict vehicle models of interest to users. Some scholars have conducted research on the front-end design of automobiles using Gestalt psychology. They have conducted a positioning analysis of the emotional image of Geely automotive brand's front-end design [12]. Chinese scholar Wang has conducted design research on electric car styling concepts. He proposed a series of strategies and design styles of new energy vehicles. It analyzed consumers' aesthetic preferences for electric car styling design [13]. However, as China's new energy industry started late, there is little in-depth exploration into whether the design of new energy vehicles aligns with the aesthetic preferences of Chinese consumers. Therefore, conducting an objective evaluation and optimization study of the styling of new energy vehicles holds significant theoretical and practical significance. Many scholars primarily focus on positive design research during the study of product aesthetics. Moreover, research on the evaluation of the styling of new energy vehicles is relatively scarce, especially concerning mainstream Chinese new energy automotive brands. Scholars mainly investigate users' sensory responses during new product development. In such research, designers often predict users' sensory needs but they find it hard to identify these needs. Additionally, designers' thoughts can influence product design and development. Existing methods for evaluating automotive styling primarily rely on subjective opinions of designers and feedback from consumers. They lack systematic and objective approaches to evaluate user reactions to products. To build Chinese new energy automotive brands, we need to adopt effective evaluation method that comprehensively assess the mainstream styles in both quantitative and qualitative ways (see Fig. 2).

Fig. 1.

Trend of penetration rate in the Chinese new energy market.

Fig. 2.

Evaluation model combining KE and AHP.

Products need to be improved to adapt to the diverse needs of users [14]. Based on the products of the company, this study adopts a reverse engineering approach to explore user sensibilities more accurately. Its main focus is to investigate the sensibility demands of Chinese users for products from emerging automotive companies in the new energy market. It aims to uncover designs that better align with user preferences and provide insights for developing new energy vehicles in China. Additionally, the research incorporates both user surveys and expert evaluations to ensure a diverse perspective and avoid a singular focus on the design. This study will serve as a vital reference and guidance for the design and marketing of Chinese electric car brands. It also contributes to the development and expansion of research methods to evaluate new energy vehicle aesthetics. Product reference and alignment with consumer preferences are essential for the aesthetic design of new energy vehicles (NEVs). Analytic Hierarchy Process (AHP) and Kansei Engineering are used in this process. By building an evaluation system and allocating weights, AHP systematically ranks and compares elements of new energy vehicle design [15]. Kansei Engineering, on the other hand, emphasizes user's subjective feelings and emotional responses [16]. Through methods such as surveys, experiments, and user interviews, the study seeks to reveal the distinctive design features of different brands of NEVs and determine consumer preferences and feelings about these features. This study only covers a limited sample of representative models from prominent Chinese companies, which may limit the scope of its findings. Due to the emerging nature of the Chinese new energy industry and unclear development strategies of companies, detailed quantitative research is not be feasible at this stage.

2. Theoretical research and analysis

In this section, we explore theoretical frameworks and conduct in-depth analyses about the styling of new energy vehicles. The research integrates Analytic Hierarchy Process (AHP) principles and Kansei Engineering methodologies. The theoretical investigation in this section will lay a solid foundation for the subsequent empirical evaluation of mainstream Chinese electric car brands. By synthesizing key concepts from both fields, the author expresses its understanding of the aesthetic considerations and design principles that shape the visual identity of these vehicles.

2.1. Kansei engineering theory research

Kansei Engineering was first proposed by Chairman of Mazda Motor Corporation, a Japanese multinational automotive manufacturer, in October 1986 [17]. Subsequently, Japanese scholar Mitsuo Nagamachi conducted a practical research on Kansei Engineering, paving the way for its further development. Professor Yohei Harada further explored the concept of Kansei Engineering and evaluated its potential. Kansei Engineering uses qualitative and quantitative analysis to reflect the relationship between “human sensibility” and “object characteristics”. It transforms vague feelings into quantitative data [18]. This method uses surveys, questionnaires, and other approaches to capture consumers' brand preferences about features of new energy vehicles. It uses surveys to collect user feedback and conducts experiments to observe user reactions under different design scenarios. In this way, it accumulates sensory vocabulary and match the words with users to uncover their sensory needs. The sensory engineering approach emphasizes emotions, experiences, and intuition. Therefore, in surveys, we focus particularly on the emotions generated when users see the exterior of new energy vehicles. The concept of Kansei Engineering has been proposed for over 50 years and scholars around the world have conducted detailed and in-depth research about it. Based on emotional analysis from Kansei Engineering, Professor Kong quantifies subjective evaluations of users to explore the effects of different design elements on the modeling of new energy vehicles and its combustion engine vehicles. Finally, he validates the feasibility of this approach [19]. Kansei Engineering is a consumer-oriented technology for a new product development developed in Japan. It becomes a popular product development technology in Japan and around the world [20]. Ju Qinghui explored the conceptual design of Renault cars with Asian culture by combining Kansei Engineering and hierarchical analysis and he extracted the styling-related perceptual vocabulary through Kansei Engineering. It used hierarchical analysis to derive the relationship between the features and styles of cars and the perceptual vocabulary to reflect the indexes that affect the evaluation of styling images [21]. Patrick Jordan is an expert in Kansei Engineering who has made significant contributions to exterior design and user experience of cars. His research reveals users' perceptual and emotional responses to the appearance of automobiles and how design elements can induce specific emotions [22]. Scholars Jindo improved the interior design of cars with semantic differential method. Based on subjective evaluation data, this method conducts multivariate analysis to find out the relationship between user perceptions and styling features [23]. According to the literature review, Kansei engineering has been widely applied in automotive design. Sensory Engineering can help designers understand user's emotional needs, but its results may be subjective and difficult to interpret, so we need to have more conclusive quantification about the product's aesthetics sense.

2.2. Theoretical research on the analytic hierarchy process (AHP)

The Analytic Hierarchy Process (AHP) is a multi-criteria decision-making method introduced by American operations researcher Saaty in 1970 [24]. This method combines both quantitative and qualitative analysis. Before making decisions, it compares the weights of various factors. The fundamental idea of AHP is to break down complex problems into elements and build a hierarchical structure model, as shown in Table 1.

Table 1.

Content of the hierarchical analysis modeling.

Sequences	Element
1	Determine the decision-making objectives and scope of the problem, and decompose the problem into a number of levels
2	Construct a hierarchical model that identifies the goal, the guideline, and the program
3	A two-by-two comparison of the factors at the guideline level was performed to obtain the relative importance between the factors
4	Calculate the weights of each criterion for the objectives and perform a consistency test
5	A two-by-two comparison of the program hierarchy was made to obtain the relative importance of the programs to each other
6	Calculate the score for each option and the option with the highest score will become the final decision [25].

Park et al. used the AHP method to assess consumer preferences for automobile appearance in different countries and regions. They investigated the opinions of consumers in the United States, China, and South Korea, and applied the AHP method to compare and analyze consumer preferences in each country and region [26]. Jae Yoon et al. used the AHP method to assess the impact of different automobile exterior designs on consumers' attractiveness and identification in their 2013 study. Through a questionnaire surveys, they collected consumers' opinions and used the AHP method to analyze the weights of different design elements [27]. The Analytic Hierarchy Process (AHP) is commonly applied in the design of product appearance and organizational innovation. During the product design, it enhances the scientific and objective evaluation and improves the design solutions and decision-making [28]. AHP is used for static decision problems and has limitations in considering users' specific needs. With iterative product development, it is necessary to capture users' requirements in real time.

2.3. Evaluation model

The design of car styling includes evaluating subjective factors such as brand and appearance, as well as objective factors such as technology and function. The hierarchical analysis method and the Kansei Engineering theory have their own strengths and weaknesses in evaluating the styling of electric cars [29]. Kansei Engineering focuses on subjective factors such as user perception, emotion and cognition, and it reveals users' feelings and preferences for electric car styling [30]. The AHP method can quantify each evaluation factor through hierarchical structure and weight assisnment. It takes into account consider multiple criterias to make a comprehensive evaluation of car styling. Scholars such as F Syaifoelida combine Kansei Engineering with the AHP to identify customers' feelings about a product and develop requirements based on customers' needs. Studies showed that appearance will invoke emotions and is correlated with product selection [31]. Therefore, the evaluation of products is crucial for product development. Scholar Zuo proposed another research method that combines Kansei Engineering with AHP. The method categorizes evaluation indicators into subjective ones and objective ones [32]. After combining Kansei Engineering with AHP, we can obtain user opinions and feedback through questionnaires or focus group discussions. With more user participation, judgment matrices and criterion weights can be established at the same time. In this way, the evaluation results will better meet users’ expectations [29].

This study is based on the following five aspects:

• 1)Identify representative new energy vehicle models and determine the features that affect styling imagery in the Chinese market.
• 2)Select the vocabulary of styling imagery using the semantic difference method in Kansei Engineering, and use the KJ method to filter the vocabulary and determine the final vocabulary for the perceptual imagery.
• 3)Build the relationship between modeling features and imagery vocabulary with hierarchical analysis, and calculate the weight of each evaluation index.
• 4)Conduct fuzzy and comprehensive evaluation by analyzing China's representative new energy vehicle modeling, and adopt five levels of evaluation.
• 5)Use the fuzzy and comprehensive evaluation and assign values to each evaluation to derive the scores of each modeling.

3. Screening of mainstream new energy vehicles and related sensory vocabulary in China

As sales of China-made vehicles are rising at notable rates, Chinese automakers have increased their share of the overseas market, expanding into Europe and elsewhere through export and cooperation projects [33]. Many Chinese electric vehicle brands and automakers have emerged, such as “Li Auto Inc”, “Nio Inc”, and “XPENG Motors”. These local brands keep making technological innovation and improving product quality to compete with traditional automakers. To meet the needs of different consumers, various types of new energy vehicles appear in the market, including pure electric vehicles, plug-in hybrids and fuel cell vehicles.

The primary data collected by the study is mainly from automotive websites. It is necessary to choose reliable and well-known websites and compare the top three websites based on their number of downloads, word-of-mouth ratings, and the number of comments. By comparing the three indexes, the most representative website is selected as the source of product researching [34].

SevenMac Data was used to comprehensively compare the top three automotive websites in China, and the results are shown in Table 2.

Table 2.

Comparison of data from various automobile websites.

Branding	Downloads (billions)	Popularity rating	Number of comments (million)
ERNIE Bot	45.28	4.9	261
DCar	23.92	4.8	38
BITAUTO	8	4.6	23

According to data analysis, “autohome” is a leading and credible online destination for automobile consumers in China (www.autohome.com.cn) with a large and engaged user base. Therefore, it is adopted as an information source to compare different new energy vehicles. According to the results, “Li Auto Inc” (Fig. 3), “XPENG Motors” (Fig. 4) and “Nio Inc” (Fig. 5) are three brands of new energy vehicles that have the highest reputation and popularity in China. Moreover, we select three popular cars with similar prices from the three car brands to be the object of this research, as shown in Table 3.

Fig. 3.

“Li Auto Inc” L8.

Fig. 4.

“XPENG Motors” G9.

Fig. 5.

“Nio Inc” Es6.

Table 3.

Vehicle-by-vehicle data.

Brand name	Price (million/RMB)	Attributes
“Li Auto Inc” L8	33.98–39.98	SUV
“Nio Inc” Es6	33.80–55.40	SUV
“XPENG Motors” G9	30.99–46.99	SUV

3.1. Determination of styling features

This study tries to deconstruct car styling and exmplain how the look of a vehicle will affect a user's cognitive state. It uses automotive styling terms to describe a car's appearance and its relation with the function and brand [35].

Car styling features, appearance features, and functional features are decomposed to form the criterion level in the hierarchical analysis structure. The criterion level is then divided into sub-criterion levels. The sub-criterion levels adopt Kansei Engineering to derive vocabulary from users’ perceptual experience of car styling. The vocabulary is then used by the criterion level. The framework of the hierarchical analysis is shown in Fig. 6.

Fig. 6.

Hierarchical analysis to extract modeling feature.

3.2. Vocabulary from perception

First, we extracted a total of 115 perceptual vocabularies about the electric vehicle from promotional materials, interviews, related literature, Internet reviews etc. The vocabularies focused on the “brand,” “look” and “function” of the vehicles and we eliminated synonyms and words with minimal lexical significance. After group discussions with designers, we chose 60 perceptual (Table 4) vocabularies and use the semantic differential technique to further filter these words.

Table 4.

Car styling words.

Serial number	Elements of identity	Imagery vocabulary	Serial number	Elements of identity	Imagery vocabulary
1	Human-machine	Sensible-Confusing	16	Modeling	Feminine-Robust
2	Style	Fashionable-vulgar	17	Innovation	Unique-Universal
3	Innovation	Futuristic-Backward	18	Modeling	Geometric-Rounded
4	Style	Sophisticated-humble	19	Style	Modern-traditional
5	Design	Beautiful-Ugly	20	Color	Bright-Dark
6	Styling	Beautiful-ugly	21	Proportion	Harmonized-dissonant
7	Function	Reliable-Doubtful	22	Structure	Stable-Variable
8	Color	Elegant-Vulgar	23	Styling	Harmonious-Dissonant
9	Proportion	Balanced-unbalanced	24	Structures	Dynamic-smooth
10	Structure	Flexible-awkward	25	Design	Dynamic-indifferent
11	Color	Lively-dull	26	Proportion	Convenient-Contrary
12	Design	Individualistic-communal	27	Function	Convenient-inconvenient
13	Man-Machine	Friendly-Resistant	28	Color	Soft-strong
14	Structure	Delicate-crude	29	Design	Stylish-outdated
15	Function	Safe-Dangerous	30	DETAILS	Sophisticated-brutal

Following the three criteria layers of the hierarchical analysis model, the filtered perceptual terms were grouped based on “brand features”, “appearance features”, and “functional features”. Based on the Likert scale, which is a 5-point evaluation scale, a questionnaire created, in which the perceptual words were categorized into “extremely important”, “very important”, “moderately important”, “slightly important”, “not at all important”. They were assigned scores of 5, 4, 3, 2, 1, respectively [36]. We invited fifteen interviewees to fill the questionnaires and mark the perceptual words, including five professors, designers and users of related industrial design. Finally, we sorted out perceptual words based on the average scores, and picked up the most representative perceptual words within the three sub-criteria levels. These perceptual words were used as the evaluation rules for the sub-criteria layers. The hierarchical analysis is shown in Table 5.

Table 5.

Styling design evaluation index system X.

Brand characteristics	Evaluation indicator A	Appearance	Evaluation indicator B	Functional characteristic	Evaluation indicator C
		Modeling	Soft-Rigid B1
Innovation	Unique-Universal A1	Proportion	Harmonized-Disordered B2	Functions	Reliable-Doubtful C1
Style	Fashionable-vulgar A2	Color	Soft-Strong B3	Man-machine	Reasonable-Confusing C2
Technology	Advanced-backward A3	Design	Stylish-outdated B4	Details	Precise-Rude C3
		Structures	Delicate-rough B5

4. Evaluation of stylistic imagery for new energy vehicles

This section evaluates the stylistic imagery for new energy vehicles. The systematic assessment adopts both the Analytic Hierarchy Process (AHP) and Kansei Engineering to reflect the aesthetic trends of mainstream Chinese electric car brands. It provides valuable insights into the visual perceptions and preferences in the styling of electric vehicles.

4.1. Evaluation indicators and weighting

Hierarchical analysis requires experts to compare each demand and calculate the weight of each indicator to obtain a strong quantitative value and a weak quantitative value for the evaluation indicator. When CR < 0.1, the consistency of the judgment matrix is within an acceptable range. Therefore, to prove is correct, the numerical calculation of the hierarchical analysis should pass the consistency test. Otherwise, the analysis may be wrong [37].

Calculate the weights:

$${\mathrm{w}}_{\mathrm{j}}=\frac{{\mathrm{a}}_{\mathrm{i}}}{\sum _{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{a}}_{\mathrm{i}}}$$ (1)

Calculate the maximum eigenvalue:

$$\mathrm{\lambda }\max =\frac{1}{\mathrm{n}}\sum _{\mathrm{i}=1}^{\mathrm{n}}\frac{\left[\text{AW}\right]\mathrm{i}}{\text{nWi}}$$ (2)

Consistency test formula:

$$\text{CR}=\text{CI}/\text{RI}$$ (3)

$$\text{CI}=\frac{\mathrm{\lambda }\max -\mathrm{n}}{\mathrm{n}-1}$$ (4)

The consistency index (CI) determines whether the pairwise comparisons made by the decision makers are performed. The value of RI (evaluation scale value) in AHP is calculated and analyzed by experts (Table 6) as a reference. Table 7 describes the specific evaluation of importance.

Table 6.

RI value.

Order (n)	1	2	3	4	5	6	7	8	9	10	11	12	13
RI	0	0	0.52	0.89	1.12	1.26	1.36	1.41	1.46	0.49	0.52	1.54	1.56

Table 7.

Judgment matrix scale.

Scale	Rating	Hidden meaning
1	Equally important	Indicator $\mathrm{i}$ is equally important as $\mathrm{j}$
3	Slightly important	Indicator $\mathrm{i}$ is slightly more important than indicator $\mathrm{j}$
5	Significantly important	Indicator $\mathrm{i}$ is significantly more important than indicator $\mathrm{j}$
7	Strongly important	Indicator $\mathrm{i}$ is more strongly important than indicator $\mathrm{j}$
9	Extremely important	Indicator $\mathrm{i}$ is more important than indicator $\mathrm{j}$
2/4/6/8	Intermediate	is the middle of the above values
Reciprocal	Inverse Comparison	The inverse of indicator i over indicator $\mathrm{j}$, $\mathrm{i}$.e. 1/ ${\mathrm{a}}_{\text{ij}}$

To correctly calculate the weights for multiple criteria in the evaluation system, this study invited 9 individuals from the new energy vehicle industry to take part. The evaluators included 3 automotive industry executives, 3 exterior designers, and 3 professors. They provided pairwise comparisons and marked the numerical values of each criterion both in the criteria layer (Table 8) and in the sub-criteria layer (Table 9, Table 10, Table 11). Finally, the judgment matrix of each level was obtained.

Table 8.

Criterion level judgment matrix and weights.

Criteria Laye (X)	A	B	C	${\mathrm{w}}_{\mathrm{x}}$
A	1	1/4	1/3	0.123
B	4	1	2	0.557
C	3	1/2	1	0.320
$\text{CI}=$ 0.009$;\text{RI}=0.520$; $\text{CR}=0.018$

Table 9.

Brand identity evaluation indicators and weights.

Criteria Laye (A)	A1	A2	A3	${\mathrm{w}}_{\mathrm{A}}$
A1	1	3	2	0.539
A2	1/3	1	1/2	0.164
A3	1/2	2	1	0.297
$\text{CI}=$ 0.005$;\text{RI}=0.520$; $\text{CR}=0.009$

Table 10.

Appearance evaluation indicators and weights.

Criteria Laye (B)	B1	B2	B3	B4	B5	${\mathrm{w}}_{\mathrm{B}}$
B1	1	2	4	2	4/3	0.329
B2	1/2	1	3	3/2	1	0.210
B3	1/4	1/3	1	1/2	1/3	0.076
B4	1/2	2/3	2	1	1/2	0.143
B5	3/4	1	3	2	1	0.241
$\text{CI}=$ 0.007$;\text{RI}=1.120$; $\text{CR}=0.007$

Table 11.

Function Evaluation indicators and weights.

Criteria Laye (C)	C1	C2	C3	${\mathrm{w}}_{\mathrm{C}}$
C1	1	3	4	0.608
C2	1/3	1	3	0.272
C3	1/4	1/3	1	0.120
$\text{CI}=$ 0.037$;\text{RI}=0.520$; $\text{CR}=0.071$

Based on the scores assigned to each level, the weights of each sub-criterion layer are obtained. The rankings are determined based on the weight of each evaluation indicator, as shown in Table 12.

Table 12.

Ranking of the evaluation indicator.

	A	B	C	$\mathrm{w}$	Ranking
A1	0.539	–	–	0.066	7
A2	0.164	–	–	0.020	11
A3	0.297	–	–	0.037	10
B1	–	0.329	–	0.183	2
B2	–	0.210	–	0.117	4
B3	–	0.076	–	0.042	8
B4	–	0.143	–	0.079	6
B5	–	0.241	–	0.134	3
C1	–	–	0.608	0.195	1
C2	–	–	0.272	0.087	5
C3	–	–	0.120	0.038	9

When assessing the design of electric cars, the functionality of a car plays a more significant part. Furthermore, the aesthetic appeal of the vehicle also carries a significant weight in how it is generally perceived. To better understand the three electric car brands in China and determine whose design is the best, it is necessary to grasp the styling trends in Chinese new energy vehicle market. Therefore, the fuzzy comprehensive evaluation method is used for scoring and provide ideas for future development.

4.2. Fuzzy comprehensive evaluation

Fuzzy set theory was proposed by Zadeh in 1965 as an extension of the classical notion of a set (Zadeh, 1965). With the proposed methodology, Zadeh introduced a mathematic method with which decision-making using fuzzy descriptions of some information becomes possible. This method allows for clear and systematic resolution of problems that cannot be precisely quantified [38,39]. Based on the rankings of the evaluation indicators, the fuzzy comprehensive evaluation was performed to quantify the evaluation indicators in an objective way. Therefore, the evaluation indicators of the selected new energy vehicles can be transformed into:

Organize the weight vectors of indicators at each level in the evaluation system: ${\mathrm{w}}_{\mathrm{x}}=\left(0.1230.5570.320\right)$, ${\mathrm{w}}_{\mathrm{A}}=\left(0.5390.1640.297\right)$, ${\mathrm{w}}_{\mathrm{B}}=\left(\left(0.3290.2100.0760.1430.241\right)\right)$, ${\mathrm{w}}_{\mathrm{C}}=\left(0.6080.2720.120\right)$.

Determine the evaluation levels and criteria for fuzzy comprehensive evaluation: Very Good, Good, Bad, Very Bad (Table 13). Assign values to each level using the notation $M={\left(9080706050\right)}^T$.

Table 13.

Evaluation levels and criteria.

Rating	Very good	Good	Okay	Bad	Very bad
Scores	90	80	70	60	50

During the investigation, 4 industry experts, 3 users, and 3 relevant designers were invited to evaluate the samples of the three selected electric vehicles. Taking “Nio Inc” car as an example, the evaluators scored each indicator based on the fuzzy evaluation and the criteria of the Analytic Hierarchy Process (AHP). For each evaluation indicator level, the number of evaluations was summarized. If the sub-criterion layer A1 is evaluated as “Very Good” for five times, it is recorded as 0.5 (assuming that there was a total of 10 evaluators, the five “Very Good” evaluations was then normalized to 0.5). In this case, matrix D₁ represented the fuzzy evaluation matrix for indicators in criterion layer A; matrix D₂ represents the fuzzy evaluation matrix for indicators in criterion layer B; and matrix D₃ represents the fuzzy evaluation matrix for indicators in criterion layer C. The final results are as follows: $D_1=\{\begin{array}{c}0.50.20.20.10\\ 0.30.40.20.10\\ 0.20.20.40.20\end{array},D_2=\{\begin{array}{c}0.40.20.30.10\\ 0.20.30.40.10\\ 0.50.30.10.10\\ 0.40.20.20.20\\ 0.30.10.40.20\end{array}{,D}_3=\{\begin{array}{c}0.40.20.30.10\\ 0.30.30.20.20\\ 0.20.30.30.20\end{array}$.

Calculate the evaluation weights of the criterion layer for the “Nio Inc” car using the formula: ${\mathrm{P}}_{\mathrm{n}}={\mathrm{W}}_{\mathrm{n}}*{\mathrm{D}}_{\mathrm{n}}$, the results are as follows: ${\mathrm{P}}_1={\mathrm{W}}_1*{\mathrm{D}}_1=\left(0.3780.2330.2590.1300.000\right)$, ${\mathrm{P}}_2={\mathrm{W}}_2*{\mathrm{D}}_2=\left(0.3410.2040.3150.1380.000\right)$, ${\mathrm{P}}_3={\mathrm{W}}_3*{\mathrm{D}}_3=\left(0.3490.2390.2730.1390.000\right)$.

Based on the results, the second-level evaluation matrix for fuzzy comprehensive evaluation is established:

$$\mathrm{P}=\{\begin{array}{c}{\mathrm{P}}_1\left(0.3780.2330.2590.1300.000\right)\\ {\mathrm{P}}_2\left(0.3410.2040.3150.1380.000\right)\\ {\mathrm{P}}_3\left(0.3490.2390.2730.1390.000\right)\end{array}\text{.}$$

Based on the evaluation matrix and the criterion layer weight ratios, the percentage values for the scheme can be summarized as $\mathrm{W}={\mathrm{W}}_{\mathrm{X}}$*P = (0.348 0.219 0.295 0.137). The final score for the “Nio Inc” car is (0.348 0.219 0.295 0.137) (90 80 70 60) T = 77.71. Following the same process, the final scores for the other two brands are as follows: “Li Auto Inc”: 76.54 and “XPENG Motors”: 75.32. With the highest score among the three brands, the “Nio Inc” car is considered to have the best exterior design.

5. Conclusion

To evaluate designs, the methods of Analytic Hierarchy Process (AHP) and Kansei Engineering have their own advantages and limitations. Integrating the two methods makes the evaluation indicators more redible and accurate. With the assistance of both methods, evaluation results can better reflect the intuition of consumers and can be conveniently compared. In this way, designers can more efficiently manage resources by knowing which elements will significantly affect the user experience. This study adopts a more scientific approach to select samples of emerging new energy car brands in the Chinese market:“XPENG Motors,” “Li Auto Inc,” and “Nio Inc.” Next, the method of Kansei Engineering was adopted to categorize car styling features and select representative perceptual words. Then it used hierarchical analysis to determine the weight ratios of each feature. According to the weight analysis, the top three factors that determine the impression of an electric car included the car's functionality (whether it is safe and reliable), the car styling, and car structure. When developing electric cars, these aspects should be given particular attention. Besides, this study used the fuzzy comprehensive evaluation to evaluate the market trends, based on which it ranked the three cars brands. The results showed that the “Nio Inc” car has a good styling in the current Chinese new energy vehicle market. The study also assessed the differences among the selected brand models. Due to time and resource constraints, this study mainly focused on mainstream electric car brands in China and may miss some new energy vehicle brands. Future research can expand the sample pool to include more brands for more comprehensive results and better understanding of aesthetic preferences in the Chinese market. Moreover, the study may not cover the latest changes in the industry because it develops in a stunning speed. Therefore, subsequent research could focus on new technologies and keep up with the design trends and market dynamics. Further research could also delve into how user involvement will affect the exterior design of new energy vehicles and how user-centric approaches will improve decision-making. In conclusion, the combination of both AHP and Kansei Engineering provides a more objective evaluation for products. This study validates its feasibility and provides a reference for future research.

Ethics statement

The research has been reviewed and approved by the Faculty Ethics Committee to ensure the ethicalsoundness and safety of the research. The research team started to recruit students from December 7, 2022 andfinish all questionnaire collection in May 6, 2023. The students who agreed to participate in theresearch completed the corresponding learning tasks and simulation exercises according to the teachingrequirements, and filled in the anonymous questionnaire at the end of the course. Meanwhile, the researchteam asked the consent of some students and conducted personal interviews.

Data availability statement

The data used to support the findings of this study are all in the manuscript.

CRediT authorship contribution statement

Yuzhe Qi: Writing – original draft, Formal analysis, Data curation, Conceptualization. Kiesu Kim: Writing – review & editing, Methodology, Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article: