© Wuhan University 2025

0 Introduction

Since the 18th National Congress of the Communist Party of China, the construction of a new energy system in China has accelerated, with the foundation of energy security continuously strengthened, providing strong support for economic and social development. On February 29, 2024, the Political Bureau of the Communist Party of China (CPC) Central Committee held the 12th group study session on new energy technologies and China's energy security, emphasizing that energy security is crucial to the overall development of the economy and society. China has already achieved global leadership in several new energy technologies and equipment manufacturing sectors, and its new energy vehicle (NEV) industry has become highly competitive in international markets. China is now a key driver in the global energy transition and in addressing climate change^[1]. With the increasing prominence of global climate change and environmental pollution, the Chinese government is actively promoting the development of new energy vehicles (NEVs) to reduce dependence on traditional fuel vehicles, lower carbon emissions, and improve air quality. At the same time, protecting and improving the ecological environment has become one of China's key development goals^[2,3]. The development of NEVs is closely linked to environmental improvement. On one hand, NEVs, with their zero or low emissions, can effectively reduce greenhouse gas and harmful exhaust emissions, mitigating air pollution and climate change. On the other hand, environmental improvements create favorable conditions for the development of NEVs, such as improve air quality that can extend battery life. Therefore, there is a mutually reinforcing relationship between NEVs and the ecological environment. Researching the coupling coordination between China's NEVs and the ecological environment is of significant theoretical and practical importance.

At present, scholars worldwide have conducted extensive research on the development of NEVs, which can be divided into two main perspectives: (i) The first perspective is the connotation and evaluation of NEV development. Scholars primarily focus on the theoretical and evaluative research surrounding NEV development, technological innovation, policy incentives, and sustainable development. For example, Liu^[4] analyzed the reasons and prospects for the development of the NEV industry using BYD (a Chinese automotive company) as a case study. Lee et al^[5] examined the impact of the development of technology-intensive industries on environmental protection through the example of South Korea's NEV industry. Zhang et al^[6] conducted a systematic analysis of the technological dissemination driving the promotion of NEVs in China using content and semantic network analysis methods. Patchell et al^[7] investigated the rise of the NEV industry in China's Greater Bay Area using a mixed-method approach of policy inventory and analysis. Na^[8] constructed a collaborative innovation mechanism model based on grounded theory to clarify the four key factors influencing the collaborative innovation of the NEV industry and their interrelationships. Zhao et al^[9] analyzed the impact of NEV promotion on carbon emissions in China from 2010 to 2020 using a spatial econometric model. Lv et al^[10] introduced the innovation value chain theory and constructed a data envelopment analysis (DEA) model to analyze the innovation efficiency of China's NEV enterprises across two stage— technology research and development, and result transformation. Yang^[11] similarly used a network DEA model to analyze the level of technological innovation in China's NEV industry across the two stages of technological development and innovation transformation to assess the impact of government subsidies on the innovation capabilities of the NEV industry. (ii) The second perspective focuses on the coordinated development of the NEV system. Current research on the coordinated development of the NEV system mainly focuses on the coupling coordination of industrial policies and promotion application within the NEV industry. For instance, Guo et al^[12] divided NEV industrial policies into four subsystems— planning guidance, fiscal and tax support, technical standards, and administrative supervision— for coupling and coordinated development. He et al^[13] added other subsystems on this basis to explore the coupling and coordination across five segments of the NEV industrial chain. Ye et al^[14] analyzed the coupling coordination of foundational resources, demand conditions, supporting policies, and model innovation within the NEV promotion application system to drive the "endogenous development" transformation of NEVs. However, there is relatively limited literature on the coordination between the NEV system and other systems. For example, Pan et al^[15] explored the degree of coupling coordination between the NEV industry and the "Beautiful China" initiative and conducted a predictive study on its development prospects. Despite this, research on the interaction between the NEV system and the ecological environment system remains scarce.

Currently, improving the high-quality development of NEVs and the ecological environment is a strategic priority for China. Investigating the coordination between the development of NEVs development and environmental improvement is crucial for promoting the high-quality development of China's new energy and NEV industries. In recent years, the Chinese government has placed significant emphasis on the development of the NEV industry, issuing a series of preferential policies and support measures, which have helped China maintain its leadership position in global NEV production and sales. However, questions remain regarding whether the rapid development of the NEV industry has been well-coordinated with environmental improvements and whether there are imbalances in development. What is the interaction mechanism between the NEV subsystem and the ecological environment subsystem within a coupled system, and what are the key drivers of their coupling coordination? These issues require further research and discussion. To address these questions, this study constructs an evaluation index system for NEV development and ecological environment, using a game-theoretic combined weighting model, particle swarm optimized projection pursuit evaluation model, coupling coordination degree model, and machine learning algorithms to analyze the coupling coordination development level and driving factors between the two areas. The research findings aim to provide theoretical support and policy recommendations for promoting the coordinated development of China's NEVs and the ecological environment, enhancing the understanding of the interaction between the development of China's NEV market and environmental protection. This will not only contribute to the sustainable development of the NEV industry but also help address global challenges such as climate change and ecological degradation, playing a significant role in promoting the high-quality development of China's new energy sector.

1 Materials and Methods

1.1 Analysis of the Coupling Coordination Mechanism

Coupling coordination theory examines the relationships of coordination, feedback, and development mechanisms between two or more systems, or between subsystems within a single system. It is mainly applied in fields such as physics, geography, and ecology. The concept of coupling originates from physics and refers to the mutual interactions and influences between two or more systems, reflecting the degree of interdependence and constraint between them. Coordination refers to the degree of positive coupling in these interactions, representing a harmonious and beneficial relationship between systems, which indicates the quality of the coordination. Together, these concepts reflect the internal harmony of a system, showing how the elements within a system transition from disorder to order through their interactions^[16]. The development of NEVs and the state of the ecological environment are two broad, complex, and structurally intricate systems, which exhibit a relationship of coupling and coordinated development, as illustrated in Fig. 1.

Fig. 1 Coupling coordination mechanism model between NEV and ecological environment

The sustainable development of NEVs improves the ecological environment, while an healthier ecological environment, in turn, supports and accelerates the transition of China's automotive industry from traditional fuel to new energy sources. The rapid development of NEVs brings several benefits, including reduced exhaust emissions, lower energy consumption, and the promotion of renewable energy. First, NEVs, which are powered by electricity or hydrogen, do not produce exhaust emissions, effectively reducing air pollution, improving air quality, and protecting the ecological environment. Second, compared to traditional fuel vehicles, NEVs utilize energy more efficiently, reducing energy consumption and alleviating the pressure on natural resource extraction and utilization. Third, the promotion and application of NEVs facilitate the development and use of renewable energy sources, such as wind and solar power, reducing reliance on fossil fuels and protecting the environment. Thus, there exists a strong coupling and coordinated relationship between the development of China's NEVs and the ecological environment.

1.2 Indicator Construction

This article, grounded in an extensive literature review^[3-15] and adhering to principles of systematization, scientific rigor, measurability, and operability, constructs an evaluation index system for assessing the development level of NEVs. This system includes 15 indicators across five dimensions: economic development, technological progress, market demand, infrastructure, and policy support. Additionally, it establishes an evaluation index system for ecological environment levels, comprising 20 indicators across six dimensions: energy, water, land, biology, air, and environment. These indicators are shown in Table 1.

1.3 Optimal Weight Based on Game Theory

1.3.1 Order relationship analysis method

The order relationship analysis (ORA) method is a subjective weighting method based on the analytic hierarchy^[17]. The expert's judgment regarding the relative importance of a specific indicator compared to all others is represented by assessing the elements in each row of the matrix. The specific procedure is as follows: (i) Determine the importance of the nine evaluation indicators. (ii) Assess the relative significance of neighboring indications. (iii) Using order synthesis, calculate the structural weight of each indicator.

1.3.2 Entropy weight method

Information entropy is a measure of information uncertainty. The entropy weight (EW) method is a weighting technique that calculates the indicator weight coefficients based on the impact of each indicator's relative variation on the overall system^[18]. The procedure is as follows: (i) Construct a decision matrix X. (ii) Indicator standardization: homogeneity of diverse indicators. (iii) Calculate the characteristic proportion of i evaluation object under indicator j. (iv) Determine the entropy value of indication j. (v) Calculate the coefficient of variation in indicator j. (vi) Determine the weight coefficient of indication j.

1.3.3 Game theory combination weighting mothod

The game theory combination weighting (GTCW) method is a technique for determining index weights in multi-attribute evaluation problems. This method combines subjective and objective weighting approaches to determine the final weights. The basic idea of the GTCW method is to find a compromise or consensus among different weights, minimizing the deviation between each weight and the optimal weight, thereby achieving a balanced and coordinated combined weight vector^[19]. This approach takes into account the competitive and cooperative relationships between different evaluation methods, using Nash equilibrium as the target for coordination, and introduces game theory into the field of comprehensive evaluation research.

Let the weight vectors obtained from the analytic hierarchy process and the EW method as ${\mathit{w}}_i(i=\mathrm{1,2})$. We can perform an arbitrary linear combination of the two weight vectors, denoted as $\mathit{w}={\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{2}}}{\alpha }_i{\mathit{w}}_i^{\mathrm{T}}({\alpha }_i>\mathrm{0})$, using linear combination coefficients denoted as ${\alpha }_i$. The goal is to optimize ${\alpha }_i$ to minimize the deviation between $\mathit{w}$ and ${\mathit{w}}_i$:

$\mathrm{m}\mathrm{i}\mathrm{n}{\Vert {\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{2}}}({\alpha }_i{\mathit{w}}_i^{\mathrm{T}}-{\mathit{w}}_i)\Vert }_{\mathrm{2}}$(1)

Taking the first-order derivative of the equation (1) based on the different co-efficient, we can derive the optimal values:

$\left(\begin{array}{cc}{\mathit{w}}_{\mathrm{1}}{\mathit{w}}_{\mathrm{1}}^{\mathrm{T}}& {\mathit{w}}_{\mathrm{1}}{\mathit{w}}_{\mathrm{2}}^{\mathrm{T}}\\ {\mathit{w}}_{\mathrm{2}}{\mathit{w}}_{\mathrm{1}}^{\mathrm{T}}& {\mathit{w}}_{\mathrm{2}}{\mathit{w}}_{\mathrm{2}}^{\mathrm{T}}\end{array}\right)\left(\begin{array}{l}{\alpha }_{\mathrm{1}}\\ {\alpha }_{\mathrm{2}}\end{array}\right)=\left(\begin{array}{l}{\mathit{w}}_{\mathrm{1}}{\mathit{w}}_{\mathrm{1}}^{\mathrm{T}}\\ {\mathit{w}}_{\mathrm{2}}{\mathit{w}}_{\mathrm{2}}^{\mathrm{T}}\end{array}\right)$(2)

Solving these equations will yield the optimal coefficients $({\alpha }_{\mathrm{1}},{\alpha }_{\mathrm{2}})$, which can be normalized to obtain the optimal weight distribution ratio coefficients, denoted as $({\alpha }_{\mathrm{1}}^{\mathrm{*}},{\alpha }_{\mathrm{2}}^{\mathrm{*}})$. Using equation (3), we can calculate the optimal combination weights.

${\mathit{w}}^{\mathrm{*}}={\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{2}}}{\alpha }_i^{\mathrm{*}}{\mathit{w}}_i^{\mathrm{T}}$(3)

1.4 Particle Swarm Optimized Projection Pursuit Evaluation Method

The projection pursuit evaluation (PPE) method is an efficient statistical approach for dealing with multi-factor complicated problems, and it can be directly applied to nonlinear and non-normal issues. Its primary premise is to project high-dimensional data into low-dimensional space using a specific combination, and by determining the optimal projection value, the evaluation value becomes more referential. The particle swarm optimized (PSO) algorithm is primarily based on adaptive weight computation, integrating historical and current ideal individual positions to gradually converge toward the optimal solution. We employ a novel algorithm, the PSO-PPE method^[20], which combines the PSO algorithm and PPE method. This approach not only accelerates the convergence of PSO but also solves the problem of discrete random variables. The specific procedure is as follows: (i) Construct a linear projection function. (ii) Develop the projected indicator function. (iii) Optimize the projection direction. (iv) Establish the solution model based on the PSO algorithm.

The objective function of the PPE method is as follows:

$\begin{array}{l}\mathrm{m}\mathrm{a}\mathrm{x}Q(a)=S(a)\cdot B(a)\\ \mathrm{s}.\mathrm{t}.\{\begin{array}{l}a>\mathrm{0}\\ {\displaystyle \sum _{j=\mathrm{1}}^{\mathrm{9}}}{a}_j^{\mathrm{2}}=\mathrm{1},-\mathrm{1}<a_j<\mathrm{1}\end{array}\end{array}$(4)

$\{\begin{array}{l}S(a)=\sqrt[]{\frac{\mathrm{1}}{n-\mathrm{1}}{\displaystyle \sum _{i=\mathrm{1}}^n}(z_i{-\overline{z})}^{\mathrm{2}}}\\ B(a)={\displaystyle \sum _{i=\mathrm{1}}^n}{\displaystyle \sum _{j=\mathrm{1}}^m}(R-r_{ij})\cdot I(R-r_{ij})\end{array}$(5)

where $S(a)$ is the sample standard deviation of the projected eigenvalue$z_i$, and $\overline{z}$ is the average value. The larger $S(a)$ indicates a more even distribution of values. $B(a)$ represents the local density of the projected eigenvalue. $r_{ij}$ is the distance between projected eigenvalues, and its value reflects the degree of dispersion. The density window width parameter R depends on the sample data structure. $I(R-r_{ij})$ is the unit step function.

1.5 Coupling Coordination Degree Model

Coupling coordination theory is an important research tool for analyzing the interrelationships, interactions, and cooperation between subsystems or elements within a system^[16]. The coupling degree indicates the mutual influence between two or more systems, reflecting the dynamic relationship required for coordinated development. It reveals the degree of interdependence and mutual constraint between systems. The coordination degree indicates the level of positive coupling in the interaction, reflecting the quality of coordination. Together, they reflect the internal harmony of the system, showing the transition from disorder to order among internal elements after interaction. Drawing from the capacity coupling coefficient model in physics, we calculate the coupling relationship and coordination degree between China's NEV development and ecological environment using the following formulas:

$C=\frac{\sqrt[]{N\times E}}{\frac{\mathrm{1}}{\mathrm{2}}(N+E)}$(6)

$T=\alpha N+\beta E$(7)

$D=\sqrt[]{C\times T}$(8)

where C is the coupling degree between the two subsystems. $N$ and $E$ represent the indices of NEV development and ecological environment, respectively, calculated using the particle swarm optimized projection pursuit evaluation model. A higher C value indicates a smaller degree of dispersion between subsystems and a higher coupling degree. T is the comprehensive coordination index of the two systems. $\alpha$ and $\beta$ are the importance weights of the two systems, calculated using the game theory combination weighting model. D is the coupling coordination degree between the two subsystems, ranging from 0 to 1. A higher D value indicates a higher degree of coupling coordination and more balanced development between the two systems. This paper divides the coupling coordination degree into 5 stages, as shown in Table 2.

1.6 Driving Factor Identification Model

Clarifying the driving factors affecting the coupling coordination development of China's NEVs and ecological environment helps to understand the reasons behind the differences in coupling coordination degrees, thereby facilitating more targeted suggestions to promote the high-quality development of China's NEVs. To evaluate the degree of influence of different indicators on the coupling coordination degree, this paper uses the normalized indicator importance calculated by two algorithms, Random Forest and Categorical Boosting (CatBoost), to detect the determining factors of coupling coordination degree and their importance.

1.6.1 Random Forest algorithm

Random Forest is a highly flexible ensemble machine learning algorithm based on decision trees as base learners, proposed by American scientist Leo^[21]. In the Random Forest algorithm, we quantify the importance of each indicator in the two systems as the average contribution of each variable across all decision trees in the random forest. By comparing the contribution magnitude of each variable, we can determine the importance of variables and assess the significance of each indicator's impact on the coupling coordination degree. The main steps are as follows: (i) Calculate the contribution of variables using the Gini index. (ii) Train the random forest based on Bagging and feature subspace concepts. (iii) Calculate feature weights and rank them.

1.6.2 CatBoost algorithm

CatBoost algorithm is a gradient boosting tree algorithm that excels in handling categorical features by employing a mean encoding-based technique to convert categorical features into numerical form and optimize feature encoding during training^[22]. The model incorporates regularization terms, including L1 and L2 regularization, to control model complexity and improve generalization ability. Unlike XGBoost, CatBoost uses an "Ordered Boosting" strategy to automatically adjust learning rates and gradually reduce them during training. The objective function of CatBoost typically consists of three parts: loss function, regularization term, and auxiliary function, expressed as follows:

$O={\displaystyle \sum _{i=\mathrm{1}}^N}L(y_i,{\widehat{y}}_i)+{\displaystyle \sum _{k=\mathrm{1}}^K}\Omega (f_k)+{\displaystyle \sum _{j=\mathrm{1}}^J}\Gamma (g_j)$(9)

where $L(y_i,{\widehat{y}}_i)$ is the loss function. $\mathrm{\Omega }(f_k)$ is the regularization term used to control the complexity of tree $f_k$, including L1 and L2 regularization. $\mathrm{\Gamma }(g_j)$ is the auxiliary objective function used to optimize model performance and robustness.

1.7 Data Sources

Based on scientific validity, rationality, and accessibility, this paper selects the period from 2010 to 2021 as the research timeframe. The main data sources include China Statistical Yearbook, Energy Conservation and New Energy Vehicle Statistical Yearbook, China Automotive Industry Statistical Yearbook, World Bank, World Development Indicators, 2030 Agenda for Sustainable Development, China Ecological Environment Status Bulletin, China Environmental Statistical Yearbook, Environmental Performance Index (EPI), and Quality of Life Index published by Numbeo for the relevant years. The extreme difference standardization method^[19] is used to eliminate the impact of dimensionality on various indicators.

2 Results

2.1 Evaluation of Indicator Weight Results

This paper first uses the order relation analysis method to calculate the subjective weights of each indicator, then uses the EW method to calculate the objective weights, and finally employs the game theory combination weighting method to integrate the order relation analysis and EW methods to obtain comprehensive weights. This approach enhances the scientific validity and rationality of the indicator weight assignment. The results are shown in Fig. 2.

Fig. 2 Evaluation indicator weight results

As shown in Fig. 2, in the NEV system, the indicator weights are ranked as follows: NEV sales volume (N₉, 0.173 8), NEV invention patent grants (N₆, 0.106 6), power battery installation volume (N₅, 0.095 4), public charging pile ownership (N₁₃, 0.084 3), relative market share (N₁₀, 0.081 2), NEV-related enterprise registrations (N₁₂, 0.076 4), maximum range of pure electric vehicles (N₈, 0.069 9), and the number of government positive policies for NEVs (N₁₅, 0.058 2). In the ecological environment system, the indicator weights are ranked as follows: number of invasive alien species (E₁₁, 0.071 4), water consumption per unit of GDP (E₃, 0.069 5), proportion of local breeds at risk of extinction (E₁₀, 0.067 7), energy consumption per unit of GDP (E₁, 0.065 6), biological carrying capacity (E₁₃, 0.064 7), CO₂ emissions per capita (E₁₅, 0.064 4), environmental performance index (E₁₈, 0.055 6), and water and soil coordination degree (E₄, 0.053 8).

2.2 Development Level Evaluation Results

By inputting the relevant data for the evaluation indicators and weight coefficients into the particle swarm optimized projection pursuit evaluation model, we can obtain China's NEV development index and ecological environment level index. The results are shown in Fig. 3.

Fig. 3 China's NEVs and ecological environment development level evaluation results from 2010 to 2021

The results show that China's NEV development and ecological environment levels both demonstrated a year-on-year upward trend from 2010 to 2021. Specifically, the NEV development index steadily increased from 0.085 in 2010 to 0.634 in 2021, while the ecological environment level index rose from 0.170 to 0.884, also showing a significant increase. This reflects China's strong emphasis and continuous investment in NEVs and ecological environment issues. The development of NEVs not only reduces dependence on traditional oil resources but also lowers environmental pollution and greenhouse gas emissions, contributing positively to the achievement of sustainable development goals. Meanwhile, the improvement in ecological environment levels indicates that China has made significant progress in ecological protection and environmental enhancement, providing people with a cleaner and healthier living environment.

2.3 Coupling Coordination Degree Results

By inputting relevant data into the coupling coordination model of NEVs and ecological environment constructed in this paper, we can obtain the coupling coordinated development relationship between the two systems. The results are shown in Fig. 4.

Fig. 4 China's NEVs and ecological environment coordinated development level from 2010 to 2021

Figure 4 clearly shows that the coupling degree C, coordination index T, and coupling coordination degree D between China's NEV development and ecological environment systems generally exhibit an upward trend year by year. Specifically, the coupling degree (blue dashed line) slowly declined from 0.667 in 2010 to 0.607 in 2014, and then slowly rose to 0.797 in 2021. The coordination index (red dash-dot line) rapidly increased from 0.127 in 2010 to 0.859 in 2021. As shown in equation (8), the coupling coordination degree (black solid line) comprehensively considers both the coupling degree and coordination index, reflecting the overall "effectiveness" and "synergy" effects between the two systems. In Fig. 4, the coupling coordination degree increased from 0.292 in 2010 to 0.468 in 2012, and then rose to 0.859 in 2021. Combined with Table 2, the development of China's NEV and ecological environment systems can be divided into three phases: a recession period from 2010 to 2012, a transition period from 2013 to 2016, and a coordinated development period from 2017 to 2021. Specifically, 2010-2011 belonged to the mild imbalance type (D = 0.292 and 0.321), 2012-2016 belonged to the primary coordination type (D = 0.468, 0.485, 0.502, 0.549, and 0.586), 2017-2020 belonged to the intermediate coordination type (D = 0.634, 0.675, 0.703, and 0.731), and 2021 belonged to the good coordination type (D = 0.828). This demonstrates that the coupling coordination between China's NEV development and ecological environment is showing a continuous improvement and a positive development trend year on year.

2.4 Driving Factors Identification Results

This paper uses Random Forest and CatBoost algorithms to detect the driving factors behind the coupling coordinated development of China's NEVs and ecological environment. By calculating the average importance of indicators based on the normalized indicator importance of the two algorithms, the results are shown in Fig. 5.

Fig. 5 Identification results of driving factors for China's NEVs and ecological environment coordinated development

In the figure, the vertical axis represents the importance of indicators, while the horizontal axis shows 35 variables arranged in descending order of importance. As shown in the figure, the number of NEV invention patent grants (N₆, 0.056 5) is the most important driving factor for the coupling coordinated development of China's NEVs and ecological environment. The top 10 driving factors following this are: water consumption per unit of GDP (E₃, 0.048), energy consumption per unit of GDP (E₁, 0.046 5), power battery installation volume (N₇, 0.046 5), biological carrying capacity (E₁₃, 0.043), SO₂ emissions per capita (E₁₆, 0.039 5), environmental protection investment index (E₁₉, 0.039 5), number of government positive policies for NEVs (N₁₅, 0.038 5), proportion of cultivated land area (E₆, 0.038 0), gross domestic product (N₁, 0.037 5), and installed power generation capacity (N₁₄, 0.035 5).

3 Discussion

Combining the results of Fig. 3 and Fig. 4, it can be observed that there are differences in the growth rates of China's NEV development index and ecological environment index during different periods, leading to varying contributions of the two systems, which in turn affect the coupling degree, coordination index, and coupling coordination degree. (i) From 2010 to 2014, the growth rate of the ecological environment index significantly outpaced that of the NEV development index, reducing the degree of mutual influence between the two systems and resulting in a decline in the coupling degree. However, the substantial improvement in ecological environment levels enhanced the coordination between the two systems, driving a rapid increase in the coordination index, which subsequently elevated the coupling coordination degree. (ii) The years 2015 to 2016 marked a turning point, as the NEV development index began to accelerate, leading to a convergence in the growth rates of both systems, which caused the coupling degree to rise again. The continuous increase in the coordination index further enhanced the coupling coordination degree. (iii) From 2017 to 2021, both systems maintained rapid development, with the coupling degree and coordination index consistently rising, reflecting an increasing mutual promotion and coordination, significantly boosting the coupling coordination degree to a high level. Therefore, only when the growth rates of the NEV development and ecological environment systems converge can the coupling degree, coordination index, and coupling coordination degree improve synchronously, achieving positive coordinated development.

The results of Fig. 2 and Fig. 5 show that the main driving factors for the coupling coordination degree between the NEV development and ecological environment systems in China are correlated with the weight rankings of their respective indicators, but also exhibit notable differences. (i) In the NEV system, driving factors such as the number of NEV invention patents authorized, installed capacity of power batteries, and the number of government policies supporting NEVs all have high weights, which aligns with the indicator weight ranking. This indicates that technological innovation, industrial strength, and policy support play significant roles in promoting the coordinated development of NEVs and the ecological environment. (ii) In the ecological environment system, factors like water consumption per unit of GDP, energy consumption per unit of GDP, biological carrying capacity, and per capita SO₂ emissions have high weights, but there are discrepancies with the weight rankings of the system's own indicators. Although factors like the number of invasive species have a greater impact on the ecological environment system itself, energy conservation and emission reduction, as well as resource utilization efficiency, may be more crucial for promoting the coordinated development of NEVs and the ecological environment. Overall, factors such as technological innovation, industrial strength, policy support, energy conservation and emission reduction, and resource utilization efficiency are all important driving forces affecting the coupling coordination degree of the two systems. However, the specific degree of influence varies due to the characteristics of each system. Identifying these differences can help formulate more targeted policy measures to promote the positive coordinated development of NEVs and the ecological environment.

Based on the research findings, the following recommendations are proposed to achieve alignment between the development of NEVs and the level of the ecological environment in China: (i) Increase efforts in technological innovation. The high ranking of indicators like the number of NEV invention patents authorized and installed capacity of power batteries in the driving factors highlights the importance of technological innovation. Therefore, it is essential to continue increasing R&D investment in key technological areas of NEVs, enhance independent innovation capabilities, and promote collaborative innovation along the industrial chain to accelerate the application of new technologies and improve overall industrial competitiveness. (ii) Optimize industrial policies. It is necessary to further improve industrial policies such as fiscal incentives and subsidies, continuously release policy dividends, increase investment in infrastructure construction, and accelerate the layout of supporting facilities like charging stations to create a favorable environment for industrial development. At the same time, promote deeper integration of NEVs with energy and transportation sectors to cultivate new economic growth points. (iii) Advance energy conservation and emission reduction. Factors such as energy consumption per unit of GDP and per capita SO₂ emissions indicate that energy conservation and emission reduction are crucial for achieving coordination between the two systems. Efforts should focus on accelerating the transformation of the energy structure, vigorously develop new and renewable energy sources, and strengthen energy conservation and control in key areas such as industry, transportation, and construction, while continuously reducing energy consumption intensity and pollution emissions. (iv) Improve resource utilization efficiency. Driving factors such as water consumption per unit of GDP and biological carrying capacity reflect significant pressure on resource and environmental sustainability. Comprehensive utilization policies for resources should be improved, water resource efficiency enhanced, ecological protection strengthened, and biodiversity and ecosystem stability maintained. In summary, leveraging the synergistic effects of key factors like technological innovation, industrial support, policy guidance, energy conservation, and resource utilization can promote positive interactions and coordinated development between NEVs and the ecological environment. This will not only help advance the sustainable development of the NEV industry but also contribute to addressing global challenges such as climate change and environmental improvement, which is of great significance for promoting the high-quality development of new energy in China.

4 Conclusion

This paper selected 35 indicators to construct an evaluation index system for the development of NEVs and the ecological environment. Using game theory combination weighting model, particle swarm optimized projection pursuit evaluation model, coupling coordination degree model, and machine learning algorithms, we measured and analyzed the coupling coordinated development level of China's NEVs and ecological environment from 2010 to 2021 along with its driving factors. The following conclusions can be drawn: (i) The results of the simulated annealing optimized projection pursuit evaluation model show that the development of China's NEV and the ecological environment demonstrated a year-on-year upward trend from 2010 to 2021. Specifically, the NEV development index increased from 0.085 to 0.634, while the ecological environment index rose from 0.170 to 0.884, with two distinct stages of growth: an initial period of gradual increase followed by a phase of rapid growth. These trends reflect China's continuous development in both NEVs and ecological environment. (ii) The results of the coupling coordination degree model show that China's NEV development and ecological environment systems were in a recession period from 2010 to 2012, a transition period from 2013 to 2016, and entered a coordinated development period from 2017 to 2021, showing a trend of continuous positive improvement. By 2021, it had reached a good coordination level. (iii) The driving factors identification results show that indicators such as the number of NEV invention patent grants, water consumption per unit of GDP, energy consumption per unit of GDP, power battery installation volume, and biological carrying capacity are the main driving factors affecting the coupling coordinated development of China's NEVs and ecological environment.

References

All Tables

All Figures

	Fig. 1 Coupling coordination mechanism model between NEV and ecological environment
In the text

	Fig. 2 Evaluation indicator weight results
In the text

	Fig. 3 China's NEVs and ecological environment development level evaluation results from 2010 to 2021
In the text

	Fig. 4 China's NEVs and ecological environment coordinated development level from 2010 to 2021
In the text

	Fig. 5 Identification results of driving factors for China's NEVs and ecological environment coordinated development
In the text