Research on the decision-making of high-tech manufacturing enterprises for energy conservation and emission reduction driven by industrial green microgrids under the green credit regulatory system Authors: Hao Qin, Xiao Zhong, Maidina Aihemaiti, Haiwei Zhang, Zhongping Cui, Yuanlei Chen Published: 2026 ABSTRACT AbstractUnder the circumstances where the green credit regulatory system has been implemented and the large-scale application of industrial green microgrids has occurred, the institutional constraints and clean energy technologies jointly reshape the emission reduction choices of high-tech manufacturing enterprises. This study constructs an evolutionary game model involving the government, research and development institutions, and high-tech manufacturing enterprises. Through this model, the stable equilibrium of the system’s evolution is sought, and the evolution patterns of key parameters are explored through numerical simulation. The conclusions are as follows: (1) The various input costs of the three parties constitute the threshold for emission reduction costs. The increase in costs significantly hinders green behaviors such as supervision, research and development, and active emission reduction. (2) Market-based benefits serve as the intrinsic driving force for collaborative emission reduction. The increase in technology transfer benefits and the improvement in enterprises’ green operation benefits can lead to a spontaneous shift towards a low-carbon economy. (3) The government’s policy tools exhibit differentiated effects, and research subsidies facilitate the breakthroughs in green microgrids technology. Excessive rewards can lead to a policy-dependent situation within the enterprise. Appropriate fines can rely on green credit constraints to compel enterprises to comply with emission reduction regulations. This study fills the analytical gap regarding the technical supply side in multi-party collaborative emission reduction. The conclusions provide a theoretical basis and practical reference for improving the collaborative emission reduction system, promoting the green transformation of high-tech manufacturing enterprises, and facilitating regional green and low-carbon development.Subject terms: Energy science and technology, Engineering, Environmental social sciences, Mathematics and computing Introduction With global warming, resource depletion, and environmental degradation, a low-carbon transition has become a global consensus1. The green transformation of the industrial sector, particularly high-tech manufacturing enterprises, is crucial for countries to achieve their goals of efficient energy conservation and emission reduction2. According to the International Energy Agency, the industrial sector accounts for over one-third of global final energy consumption. In the industrial sector, compared with other manufacturing enterprises, high-tech manufacturing enterprises are the fastest-growing sector and also the one with the most significant increase in energy consumption and emissions. During the production process, such enterprises have higher requirements for power supply quality and a clean environment. And their products are deeply integrated into the global supply chain. The carbon footprint directly affects export competitiveness and the cost of complying with international regulations. This underscores the critical importance of fostering green development among high-tech manufacturing enterprises3. Although high-tech manufacturing enterprises are often labeled "low-pollution, high-value-added", the industry’s scale expansion and deepening division of labor within the global industrial chain have increasingly highlighted energy consumption and emissions issues. Consequently, the strategic urgency of energy conservation and emission reduction continues to rise4. Against this backdrop, the government is compelling enterprises to reduce emissions through policies such as green credit regulation. The Guidelines for the Construction and Application of Industrial Green Microgrids (2026–2030) also explicitly designate industrial green microgrids as a key vehicle for energy conservation and carbon reduction. High-tech manufacturing enterprises serve as the core implementers of this technology, whose deployment directly supports companies in reducing carbon emissions at the production end and represents a key lever for achieving corporate emission reduction targets. No single entity can successfully implement such technologies or achieve emission-reduction outcomes. Tripartite collaborative decision-making among governments, research institutions, and high-tech manufacturing enterprises has become the core mechanism. For instance, BYD, as a typical high-tech manufacturing enterprise, has accelerated the achievement of its energy conservation and emission reduction goals through green innovation. Leveraging distributed photovoltaic and new energy storage technologies to build industrial green microgrids, the company has benefited from local government initiatives such as vehicle purchase subsidies and dual-credit policies, as well as technological support from R&D institutions5. The green credit regulatory system represents an iterative upgrade of the government’s green information disclosure framework, serving as an incentive-compatible institutional innovation grounded in a relational contract network involving multiple stakeholders. Its core function lies in linking the environmental performance of high-tech manufacturing enterprises to financing costs and policy support6, imposing hard constraints, such as financing restrictions on energy-intensive enterprises, while offering soft incentives, such as policy preferences for green, emission-reducing enterprises7. The green credit regulatory system aims to guide enterprises from passive “regulatory compliance” toward proactive "green value creation", while also serving as a crucial tool for governments to regulate the environmental conduct of market entities. Industrial green microgrids serve as core supporting infrastructure for corporate green emissions reduction. Their value lies in integrating clean energy to lower corporate carbon emissions, reduce reliance on high-carbon energy sources, and enhance energy supply stability and utilization efficiency8. Industrial green microgrids provides a critical practical pathway for enterprises to gain green credit advantages, thereby fully realizing the incentive and regulatory effectiveness of the green credit regulatory system. The synergistic effects of the green credit regulatory system and industrial green microgrids are jointly driving energy conservation and emission reduction in high-tech manufacturing. The green credit regulatory system mainly influences technological research decisions through policy guidance and expected returns. Su et al. conducted an analysis of the optimal contract design for green credit regulation from the perspective of the bilateral relationship between the government and enterprises9. Gao et al. explored the mechanism by which the credit system enables the green transformation of the manufacturing industry through a dual-subject research framework10. Wu et al. studied the challenges in the construction of the integrated energy market and policy adaptation through focusing on the market systems design11. In conclusion, the existing studies are mostly limited to the bilateral game between the government and enterprises, lacking the analysis of the three-party interaction involving research and development institutions, and ignoring the intermediary transmission effect of research and development transformation. Previous studies have explored the multiple effects of various measures on the construction and operation of industrial green microgrids from aspects such as technical architecture, scheduling optimization, and benefit calculation. Gutiérrez-Oliva et al. defined the microgrids technical architecture from the perspective of technical characteristics12. Chen et al. designed a low-carbon dispatching strategy for microgrids13, but they disconnected the connection between policy systems and technological applications. Rajamand modeled and calculated the cost reduction effect of demand response through technical and economic analysis14. In conclusion, the current studies fail to clarify the transmission path through which policy promotes research innovation and enables the implementation of results to enterprises empower microgrids for emission reduction, and it also lacks the analysis of multi-party interest linkage and the coordination of institutional and technical aspects. High-tech manufacturing enterprises possess the characteristics of energy conservation and emission reduction as well as resource utilization efficiency, and are also influenced by various complex factors. Zhang et al. empirically examined the emission reduction effect of high-tech industrial clusters15. Sun et al. explored the path of enterprise innovation16. Wang et al. calculated the contribution of innovation efficiency to emission reduction17. All three adopted static measurement models, neither integrating the dual variables of green credit and microgrid, nor capable of depicting the long-term dynamic evolution of the parties. In conclusion, the existing research mainly focuses on static analysis, lacking the adjustment processes of each party over time as well as the influence of key parameters. Based on the existing studies results, the innovation points of this study lie in: First, this study expands the scope of the game participants and constructs an evolutionary game model involving the government, R&D institutions, and high-tech manufacturing enterprises. This addition provides a strategy interaction analysis for the technical supply parties. Second, this study conducts a study on the coupling of systems and technologies, integrating green credit regulation and industrial green microgrids into the same analytical framework, and revealing the internal mechanism of enterprises’ emission reduction under the collaborative effect of the two. Third, this study conducts multi-parameter sensitivity analysis through numerical simulation, dynamically depicting the influence patterns of variables such as cost, benefit, rewards, and punishments on the evolution of the three-party decision-making process. Based on this, this study explores the following question: How do green credit regulations and industrial green microgrids influence energy conservation and emission-reduction decisions in high-tech manufacturing enterprises? What are the dynamic evolutionary patterns of strategic interactions among these three elements? Can the introduction of industrial green microgrids optimize this synergistic effect? To address these issues, this study constructs a tripartite evolutionary game model involving governments, R&D institutions, and high-tech manufacturing enterprises. By analyzing the stable equilibrium conditions for strategic choices among entities and validating the impact effects of key parameters through numerical simulations, it systematically reveals the intrinsic logic of corporate energy conservation and emissions reduction under the green credit regulatory system and R&D support. This study provides theoretical foundations and practical references for improving collaboration among innovation entities, advancing the green transformation of high-tech manufacturing enterprises, and facilitating regional green and low-carbon emissions reduction. Literature review Research on the interaction between green credit regulatory system and multi-entity strategies The green credit regulatory system is a mechanism whereby governments evaluate the creditworthiness of green behaviors based on entity-specific green conduct information, using prescribed indicators, methodologies, and procedures. It assigns credit ratings and publicly discloses them for use by relevant departments, institutions, and organizations, with multi-entity participation18. Su et al. argued that the green credit regulatory system aims to create an environment conducive to incentives and disincentives under low information thresholds, thereby seeking optimal contractual arrangements9. Building upon this, Gao and Ma further clarified that the system integrates market entities’ environmental behaviors into the credit evaluation system, combining incentive and constraint mechanisms to guide entities toward decisions aligned with energy conservation and emission reduction objectives10. Furthermore, Lv et al. contended that this system is essentially a Chinese-style green institution grounded in laws and regulations, focused on information sharing and application19, and aimed at consolidating the foundations of a modern market. The green credit regulatory system represents an incentive-compatible institutional innovation based on a relational contractual network involving multiple entities. For R&D institutions, green credit regulation primarily influences their technological R&D decisions through two key aspects: policy guidance and profit expectations. On the one hand, Chen et al. noted that governments promote technological innovation by incorporating the green technology conversion efficiency of R&D institutions into credit evaluations20. Simultaneously, Tang et al. noted that governments also provide subsidies, rewards, and credit bonuses to R&D institutions that contribute to corporate energy conservation and emissions reduction21. These incentive policies guides R&D institutions to focus on areas such as low-carbon technologies and energy-saving processes urgently needed by high-tech manufacturing enterprises. On the other hand, R&D institutions must make decisions based on market demand and the costs of technological development. With the intensification of green credit regulation and the growing demand for emissions reduction among high-tech manufacturing enterprises, Peng and Zhang argued that R&D institutions should increase investment in industry-academia-research collaboration and shorten technology transfer cycles22. For high-tech manufacturing enterprises, the core of decision-making lies in balancing compliance costs against credit gains. Multiple studies by Tian et al. indicated that the intensity of green credit regulation exhibits an inverted U-shaped relationship with corporate emission reduction costs23. Moderate regulatory pressure encourages voluntary emission reductions by enterprises. However, Lin et al. found that overly stringent regulation causes firms to prioritize short-term compliance over long-term technological innovation24. Meanwhile, the transparency of credit evaluations can also influence corporate decision-making. Lin and Pan proposed that a transparent evaluation system can reduce firms’ information costs associated with compliance and enhance their willingness to participate in green transformation25. For instance, synergies between digital finance and green credit can increase corporate transparency of environmental information by over 30%, significantly improving decision-making efficiency. Research on the technical characteristics and emission reduction mechanisms of industrial green microgrids As a key energy supply model supporting the green transformation of industry, the definition and technical characteristics of industrial green microgrids have become a focal point of academic attention. Gutiérrez-Oliva et al. explicitly stated that industrial green microgrids are microgrids within industrial sites or parks whose primary power supply originates from low-carbon sources, constituting a vital component of developing industrial green infrastructure12. Mehrabadi and Sathiakumar emphasized that industrial green microgrid technology, by integrating distributed power sources, energy storage devices, and energy conversion equipment into compact generation and distribution system, enables both grid-connected and standalone operation. It facilitates the development of green, low-carbon, and high-quality energy, playing a crucial role in advancing the green and low-carbon transition26. Green microgrids simultaneously fulfill multiple demands, including power supply, heating, cooling, humidity control, and domestic water use, thereby enhancing energy utilization efficiency. Furthermore, Kaur et al. noted that integrating energy storage units facilitates internal energy regulation within the system, enabling efficient and rational energy utilization27. Industrial green microgrids serve as key vehicles for reducing industrial emissions, and their mechanisms have drawn significant academic attention. To achieve carbon peak and carbon neutrality goals28, Chen et al. conducted extensive research on low-carbon dispatch and operation of the power system. Existing scholars found that applying low-carbon integrated energy microgrid models in the power sector not only reduces system carbon emissions more effectively but also enhances energy utilization efficiency13. Bustamante et al. indicated that microgrids integrate distributed generation, energy storage systems, controllable loads, and, often, combined heat and power units. Microgrids can operate in both grid-connected and islanded modes, enhancing resilience, energy autonomy, and operational flexibility29. Xi et al. further noted that, driven by the global carbon neutrality goal, industrial green microgrids (IGMs) have become a key approach to accelerating the low-carbon transformation of the industrial sector30. Research on driving factors and dynamic evolution of energy conservation and emission reduction in high-tech manufacturing enterprises High-tech manufacturing enterprises are companies that rely on advanced technology as their core driver, engaging in research, development, design, and production of high-tech products. Energy conservation, reduced emissions, and efficient resource use characterize them. Zhang et al. constructed a panel threshold econometric model. Existing scholars found that high-tech enterprises exhibit green-industry characteristics in terms of resource conservation and environmental protection, and that their clustering has a significant impact on the development of the green economy15. Wang et al. proposed that high-tech manufacturing enterprises utilize advanced technologies for development and production. Due to their substantial R&D investments, high value-added nature, favorable international market prospects, and profound influence on other industries, high-tech manufacturing enterprises have become a critical domain in international economic and technological competition31. Meanwhile, Sun et al. found that, as key drivers of corporate transformation and upgrading, high-tech enterprises possess strong innovation capabilities and high levels of innovation. High-tech manufacturing enterprises serve as vital pillars of national economic development and can lead to national innovation and progress16. Additionally, Chen et al. noted that high-tech manufacturing enterprises possess robust research and development capabilities and technological reserves, enabling them to achieve innovations that enhance resource utilization efficiency and reduce environmental pollution32. A complex interplay of multiple factors is influencing Energy conservation and emission reduction efforts in high-tech manufacturing enterprises, encompassing both internal technological and managerial capabilities and external regional characteristics and their positions within the global value chain. Wang et al. employed the DEA-Malmquist index model and panel econometric models to analyze that innovation efficiency in high-tech industries significantly contributes to carbon emission reduction, with one mechanism being the optimization of energy consumption structure17. Wei et al. argued that optimizing the energy consumption structure is a crucial mechanism for high-tech manufacturing enterprises to achieve carbon emission reductions, with the energy-saving bias of technological progress being key to lowering energy intensity33. Meanwhile, Wang et al. established an evaluation framework based on the DEA method. Existing scholars found that China’s high-tech manufacturing enterprises exhibit significant regional disparities in green capacity utilization rates, with higher rates in the west and lower rates in the east. This phenomenon indicated that regional heterogeneity leads to differences in emission-reduction performance34. Furthermore, Huang and Zhang employed a high-dimensional fixed-effects model to demonstrate further that digitalization exhibits an inverted U-shaped relationship with the embedded carbon emissions of exports, and that advancing one’s position in global value chains can significantly reduce carbon emissions35. Insights and limitations of existing research Existing research findings provide crucial insights for advancing this study: First, at the research perspective level, existing paradigms focusing on single subjects or elements offer a reference for multi-entity, multi-factor correlation analysis from a holistic viewpoint. Second, at the methodological level, the application experience with various quantitative empirical methods guides the optimization of research approaches and enhances the scientific rigor of conclusions. Third, at the theoretical level, Existing research clarifies the core logic linking green credit regulatory systems, industrial green microgrid technologies, and emissions reduction in high-tech manufacturing enterprises, laying a solid foundation for constructing a multi-entity interactive analysis framework. While drawing extensively on the value of existing research, the current literatures still have the following shortcomings: First, the limitations of the game-theoretic perspective, which primarily focuses on the two-way interaction between government and enterprises, fail to incorporate R&D institutions adequately. Existing literature lack of a systematic analysis of the dynamic three-party interaction makes it difficult to reveal the mediating role of technological R&D and transformation in policy transmission and corporate emissions reduction. Second, the dynamic transmission logic among multiple entities remains inadequately explained. Existing literature fails to clarify the pathways through which policy instruments guide research institutions in tackling technological challenges and how technological achievements empower enterprises to reduce emissions. Consequently, the intrinsic logic of multi-entity collaborative emission reduction remains insufficiently elucidated. Third, research methods primarily rely on static analysis, lacking simulations of the dynamic evolution of multi-party strategies, technologies, and institutions. Static approaches struggle to capture the adjustment processes of entity’s strategies over time and the marginal effects of key parameters, thereby failing to comprehensively evaluate long-term emission reduction outcomes and policy optimization pathways. Problem description and model construction Problem description During the green and low-carbon transformation process of high-tech manufacturing enterprises, the industrial green microgrids, as the core technology carrier, can not only reduce the consumption of fossil energy and carbon emissions at the source, but also promote the total factor productivity improvement of high-tech manufacturing enterprises. The application of this technology can achieve the coordinated development of green benefits and production efficiency. For instance, an integrated microgrid system combining power generation, distribution, load and storage specifically designed for industrial scenarios can achieve coordinated scheduling of photovoltaic power, energy storage and flexible loads. The system ensures the reliability of production, achieves an increase in the utilization rate of renewable energy, a significant reduction in carbon emission intensity, and provides a replicable practical path for the green transformation of high-tech manufacturing enterprises36. This study focuses on three parties: the government, R&D institutions, and high-tech manufacturing enterprises. And this study forms a dynamic interaction system centered around the common goal of "reducing emissions and carbon, and green development". Furthermore, this study constructs a mechanism through which the three parties can promote the industry green transformation in a closed-loop process of "policy guidance—technical support—practical implementation", as shown in Fig. 1. Energy conservation and emission reduction mechanism for high-tech manufacturing. Governments, R&D institutions, and high-tech manufacturing enterprises each play distinct roles and are closely interconnected in the green and low-carbon transition. As the regional environmental regulator and policy maker, the government’s core role is to establish institutional frameworks and provide policy safeguards. The green credit regulatory system serves as the primary vehicle for fulfilling this role. Through a dual mechanism of incentives and constraints, this system links corporate environmental performance to financing costs and policy support. The green credit regulatory system also guides research institutions to focus on low-carbon technology fields through targeted subsidies and other measures. R&D institutions serve as independent third-party technical entities separate from government and enterprises. Their core responsibility is to tackle and promote key technological breakthroughs. Industrial green microgrids, as core supporting infrastructure for corporate emissions reduction, represent the primary focus of these R&D institutions. By overcoming challenges in distributed photovoltaics, novel energy storage, and coordinated control of generation, the grid, load, and storage, they provide enterprises with technical solutions that simultaneously achieve energy conservation, reduced consumption, enhanced energy efficiency, and a stable power supply. High-tech manufacturing enterprises, as the primary implementers of energy conservation and emission reduction, serve as the core vehicles for policy implementation and technological application. Their emission-reduction outcomes directly determine the achievement of transformation goals and serve as the pivotal link between policy requirements and technological value. The three parties have formed a close strategic linkage in the research and application of industrial green microgrids, with mutual influence and restraint. How to fully leverage the carbon reduction enabling effect of industrial green microgrids through tripartite strategy collaboration, avoid falling into the governance predicament of "weak government regulation, R&D institutions low investment, and enterprises negative compliance", and achieve the active evolution of the tripartite strategy, are the core issues that this study urgently needs to address. The logical relationship among the three parties is illustrated in Fig. 2. Tripartite entity relationship diagram. Model assumptions Hypothesis 1: This model has three parties, which are denoted by Ω to represent the set of game participants. Ω = {Government, R&D institutions, high-tech manufacturing enterprises}. The three parties together constitute a complete game system. Based on the fundamental difference between evolutionary game theory and traditional game theory, this study assumes that all the participating parties are of limited rationality. The participating parties have differences in their analytical, judgmental and learning abilities. At the initial stage, none of the parties can predict the optimal strategy. They learn and imitate during the game process, thereby making the most appropriate behavioral decisions in the current environment. The equilibrium state of the system is the result of the long-term interaction and evolution of various strategies, rather than the outcome of a one-time static game. This study separates R&D institutions from high-tech manufacturing enterprises as separate parties, and separates technological research from production and emission reduction activities37. This is a classic simplified assumption used in the research on the three-party evolution game of government, enterprise and R&D institution. This assumption is in line with the division of labor model in reality where R&D institutions focus on tackling green technology issues, while manufacturing enterprises concentrate on implementing production38. This helps to clearly depict the interaction mechanism among policy incentives, technological supply and emission reduction implementation. Hypothesis 2: The government has two core strategies: “Strict supervision” and “Lenient supervision”. Among them, the probability that the government implements a strict supervision strategy is x, and the probability that it implements a lenient regulatory strategy is 1 - x, with . “Strict supervision” indicates that the government focuses on green credit assessment, emission reduction achievements verification, and penalties for violations enforcement. “lenient regulation” indicates that the government focuses on basic process management and market-oriented guidance. R&D institutions have two core strategies to choose from when it comes to energy conservation and emission reduction technologies: "High investment R&D" and "Low investment R&D". The probability that a R&D institution adopts a high investment R&D strategy is y, and the probability of adopting a low investment R&D strategy is 1 - y, with ."High investment R&D" focuses on key technologies such as distributed photovoltaics, new energy storage, and coordinated control of power generation, grid, load and storage, in order to ensure the technology transformation efficiency and implementation adaptability. "Low investment R&D" indicates that the research institution merely maintains basic scientific research activities. High-tech manufacturing enterprises have two core strategies for energy conservation and emission reduction: "Green value-driven emission reduction" and "Supervision compliance-driven emission reduction". The probability that high-tech manufacturing enterprises choose the green value-driven emission reduction strategy is z, and the probability that they choose the regulatory compliance-driven emission reduction strategy is 1 - z, with ."Green value-driven emission reduction" refers to the situation where high-tech enterprises actively adopt industrial green microgrids technology, taking into account both the emission reduction effect and market value. "Supervision compliance-driven emission reduction" indicates that enterprises merely meet basic environmental protection standards and respond passively to supervision requirements. Hypothesis 3: The cost of strict government supervision is . This cost mainly includes expenses such as green credit reviews, employee salaries, and environmental inspections. The cost of choosing “lenient regulation” is . This option only covers the basic regulatory process costs, and . This indicates that the cost of strict supervision investment is even higher. If high-tech manufacturing enterprises choose the "Green value-driven emission reduction" approach and achieve the emission reduction targets, at this point, the government will obtain positive benefits due to the improved environment. Conversely, if high-tech manufacturing enterprises choose the "Supervision compliance-driven emission reduction" approach and fail to meet the standards, they will have to bear the environmental damage cost . Hypothesis 4: R&D institutions, as the main providers of industrial green microgrid technology, exhibit significant differences in costs and benefits under various strategies. When R&D institution adopts the "Active research and transformation" strategy, it needs to incur costs such as technological research, employee training, and mid-term testing, which shall be denoted as . If the R&D institution opts for "Active research and transformation ", and the high-tech manufacturing enterprises adopt its technology to achieve emission reduction, the institution can obtain benefits such as technology transfer fees and patent usage fees. If the enterprise does not adopt their technology, the R&D institution will only receive the special subsidy from the government for high-investment research and development, and . When R&D institution adopts the "negative research and transformation" strategy, the basic research costs it needs to bear are , and . At this point, since there are no technological transformation benefits or additional market returns, the R&D institutions only receive the fixed research funding . Hypothesis 5: High-tech manufacturing enterprises, as the main implementers of energy conservation and emission reduction, exhibit significant differences in costs and benefits under different strategies. For enterprises to adopt the "green value-driven emission reduction" strategy, they need to incur cost , which includes expenditures such as green technologies modernization, clean energy substitution, and comprehensive environmental protection reforms. By adopting this strategy, enterprises can gain opportunities such as green supply chain cooperation and brand image enhancement, and thereby obtain market revenue . When high-tech manufacturing enterprises opt for "Supervision compliance-driven emission reduction", they only need to invest in the basic governance cost . Such as the construction and operation of waste gas and wastewater treatment facilities, and . Under this strategy, the enterprise only has the basic benefit , with no additional market benefit. If enterprise is strictly monitored by the government and found to have failed to meet the emission reduction targets, in addition to paying fine , it will also have to bear the financing costs resulting from the downgrade of its green credit rating. Hypothesis 6: When high-tech manufacturing enterprises adopt the "Green value-driven emission reduction" strategy, they utilize the core technology of industrial green microgrids, which is the "high investment R&D" approach of the R&D institutions. This technology has a minimum application scale: it requires that the installed capacity of distributed photovoltaic power should account for a reasonable proportion of the enterprise’s daily average electricity consumption, and the storage capacity should cover the main part of the peak-valley load difference39. The conversion efficiency of this technology is determined by the investment intensity of the R&D institution. The higher the value of , the stronger the power regulation accuracy of the microgrid. Under the strict supervision of the government, if enterprises adopt this technology, their green credit rating will be improved, and thus they can receive an additional credit reward from the government, denoted as . Here, . For the government, when the efficiency of technology transfer is insufficient, high-tech manufacturing enterprises will not adopt this technology. Due to the lack of advanced technological support, enterprises have reduced their emission reduction capabilities, which in turn has led to environmental pollution. At this point, the government needs to incur additional costs for environmental monitoring, denoted as , where (Table 1). Table 1Key model parameters and their meanings.Main bodyParametersMeaningGovernmentThe government imposes strict regulatory costsThe government has lenient regulatory costsWhen the government enforces strict regulations, high-tech manufacturers’ poor emission-reduction performance forces them to bear additional environmental monitoring costsThe government’s gains after green emission reduction by high-tech manufacturing enterprisesWhen the government strictly supervises, additional incentives provided by the government to high-tech manufacturing enterprisesHigh-tech manufacturing enterprises’ compliance-driven emission reduction, and government-borne environmental damage costsR&D institutionsHigh investment costs for R&D institutionsLow investment costs for R&D institutionsBenefits when high-tech manufacturing enterprises adopt the developed technologyWhen R&D investment is low, the fixed research funding is obtainedGovernment subsidies specifically allocated for technologies not adopted by high-tech manufacturing enterprisesHigh-tech manufacturing enterprisesThe cost of green value-driven emissions reductionCosts of regulatory compliance-driven emissions reductionWhen the government strictly regulates, enterprises bear the financing costs resulting from the downgrade of green ratingsMarket benefits of choosing green value-driven emissions reduction strategiesFoundational benefits of selecting regulatory compliance-driven emissions reduction strategiesFines are imposed when emissions reduction targets are not met under strict government oversight Key model parameters and their meanings. Payoff matrix Based on the above assumptions, the government, R&D institutions, and high-tech manufacturing enterprises make independent decisions. The strategic behaviors of these three parties interact within the market, jointly determining the game’s outcome. The payoff matrix for the game is shown in (Table 2). Table 2Payoff matrix.GovernmentR&D institutionsHigh-tech manufacturing enterprisesGreen value-driven emission reduction zSupervision compliance-driven emission reduction (1- z)Strict supervision (x)High- investment R&D (y)Low- investment R&D (1- y)Lenient supervision(1- x)High investment R&D (y)Low investment R&D (1- y) Payoff matrix. Analysis of strategy stability among game participants Strategic stability analysis of government Government’s expected returns under strict supervision strategy and lenient supervision strategy and , respectively, and the average expected return, are shown in Eqs. (1)–(3). The government’s replicating dynamic equation is shown in Eq. (4). Strategic stability analysis of R&D institutions The expected returns for high-investment and low-investment R&D strategies, and , and the average expected return, , are shown in Eqs. (5)-(7). Similarly, the replicating dynamic equation for the R&D institution is shown in Eq. (8). Strategic stability analysis of high-tech manufacturing enterprises The expected returns and for high-tech manufacturing enterprises adopting green value-driven emission reduction strategies versus regulatory compliance-driven emission reduction strategies, along with the average expected return , are shown in Eqs. (9)–(11). Similarly, the replicating dynamic equation for high-tech manufacturing enterprises is shown in Eq. (12). After conducting individual equilibrium strategy analyses for the government, R&D institutions, and high-tech manufacturing enterprises respectively, this study further conducts an overall analysis of the stability of the game system. Based on Eqs. (4), (8), and (12), by combining these three formulas, the replication dynamic equation system of the government – R&D institutions—high-tech manufacturing enterprises can be obtained as shown in Eq. (13). Equilibrium point analysis In the game system, when the three entities choose pure strategies, eight equilibrium points (E1–E8) are formed, along with one mixed-strategy equilibrium point (E9). Only pure-strategy Nash equilibria are asymptotically stable. Since E9 belongs to the mixed strategy Nash equilibrium, it does not satisfy the conditions for asymptotic stability. This section determines equilibrium points by solving for fixed points in the replicator dynamics system. Combined with stability criteria, it clarifies the system’s evolution toward stable states. The Jacobian matrix derived from the replicator dynamics equations F(x), F(y), and F(z) are shown in Eqs. (14)–(23). The eigenvalues of the potential equilibrium points are shown in Table 3. Table 3Eigenvalues of each potential equilibrium point.Potential equilibrium pointλ1λ2λ3E1(0,0,0)E2(0,1,0)E3(0,0,1)E4(0,1,1)E5(1,0,0)E6(1,1,0)E7(1,0,1)E8(1,1,1) Eigenvalues of each potential equilibrium point. The stability of each equilibrium point is affected by the parameter values, and the evolution path of the system changes dynamically with the adjustment of parameters. Therefore, this study sets up 8 different parameter scenarios and analyzes the system stability one by one. According to Lyapunov’s first criterion40, the stability of the equilibrium point can be determined by the signs of the eigenvalues of the Jacobian matrix: When the real parts of all the eigenvalues of the matrix are less than 0, the equilibrium point is an evolution stable point. If the real parts of all eigenvalues are greater than 0, then it is an unstable point. If there are 1–2 positive real eigenvalues, it is a saddle point. The results are shown in Table 4. Table 4Eigenvalues of the Jacobian matrices corresponding to each equilibrium point and stability analysis.Equilibrium pointReal part symbolStable conditionsStability results( + , + , + );;Instability point( + , - , + );;Saddle point( - , + , - );;Saddle point( - , - , - );;ESS( - , + , * );;Saddle point( - , - , - );;ESS( + , + , + );;Saddle point( + , - , + );;Saddle point*Indicates uncertain symbol. Eigenvalues of the Jacobian matrices corresponding to each equilibrium point and stability analysis. ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; *Indicates uncertain symbol. Scenario 1: When , and , the eigenvalues of the Jacobian matrix corresponding to E4 (0,1,1) are all less than 0, indicating an evolutionarily stable point. At this point, the evolutionarily stable strategy for the three parties is {lenient supervision, high investment R&D, and green value-driven emissions reduction}. In this situation, the incremental benefits from the high investment in R&D by the R&D institutions exceed the incremental costs. Even without the pressure of mandatory regulation, they still have the motivation to continuously carry out green technology research. After high-tech manufacturing enterprises adopt industrial green microgrids technology, the incremental benefits from green emission reduction can cover the cost difference. Choosing green value-driven emission reduction is more economically reasonable. The government can reduce administrative costs by implementing more lenient supervision. Moreover, the mutually beneficial strategies between R&D institutions and high-tech manufacturing enterprises can already release the positive externalities of emission reduction through the market mechanism, and the system stability can be maintained without additional government intervention. This equilibrium indicates that industrial green microgrids technology is the core technical support for achieving market-driven low-carbon emission reduction, and it is also a key condition for achieving the dual goals of enterprises green transformation and high-quality development. In conclusion, the government should establish a market-oriented governance framework, and R&D institutions should enhance their efforts in technological research and the transformation of research outcomes. This can jointly encourage high-tech manufacturing enterprises to incorporate green microgrids technology into their long-term development strategies. All parties can achieve mutual benefits and win–win results, ultimately achieving coordinated development. Scenario 2: When , , and, the Jacobian matrix of equilibrium point E6 (1,1,0) yields three negative eigenvalues, indicating that equilibrium point E6 is an evolutionarily stable point. At this point, the evolutionarily stable strategy for all three parties is {strict supervision, high investment R&D, and supervision compliance-driven emission reduction}. In this situation, the government implements strict supervision. The additional supervision costs and the environmental monitoring costs are lower than the fines imposed on enterprises that fail to meet the emission reduction targets. Therefore, the government implements strict supervision is economically feasible. Meanwhile, the total sum of the government’s special subsidies and the difference between the benefit and cost obtained by the R&D institutions through high investment R&D is higher than the total cost of their investment. Even if the technology is not adopted by enterprises, it can still obtain stable subsidy income. Therefore, R&D institutions have abundant innovation motivation under the policy incentives. For enterprises, the supervision compliance-driven emission reduction measures they adopt represent a negative and reactive approach to reducing emissions. They merely meet the minimum environmental protection standards and do not undertake in-depth green transformation. Under the intense pressure of strict government supervision, enterprises will be confronted with hefty fines and rising financing costs. However, the high cost of green value-driven emission reduction and the market benefit uncertainty make its net benefit lower than that of negative compliance-based emission reduction. At this point, enterprises, due to the fact that the overall benefits of supervision compliance cover the costs of non-compliance , tend to opt for supervision compliance-driven emission reduction. In conclusion, the government should establish a sound green credit supervision and reward-punishment linkage mechanism, and strengthen precise subsidies and technological transformation incentives for R&D institutions with high investment R&D. R&D institutions should focus on conducting research on low-cost and easily adaptable industrial green microgrids technologies, providing technical support for enterprises to achieve green emissions reduction. The three parties jointly guide high-tech manufacturing enterprises to shift from negative compliance to proactive green value creation, and systematically enhance the carbon reduction and emission reduction efficiency. Simulation analysis To verify the stability of the equilibrium points in the aforementioned theoretical model, and to further explore the influence of each key parameter on the evolution path of the system. Based on the research results of Liu et al.41, this study adopts the numerical simulation method to conduct a dynamic simulation of the evolutionary game among the government, R&D institutions, and high-tech manufacturing enterprises. All simulations were carried out in the MATLAB environment. The simulation process adopts the method of controlling variables, that is, only a single core variable is changed to analyze its impact on the system’s stability strategy, while the values of the other variables remain constant. The parameter values in this study are determined based on the proportional allocation method commonly used in authoritative literature, and the real data are converted to a dimensionless form for hierarchical determination. The benchmark parameters for government supervision costs, rewards, penalties, etc. are based on the official statistics and policy standards in the "China Environmental Statistical Yearbook", "Industrial Green Microgrid Construction and Application Guidelines (2026–2030)", and "2024 Green Development Report". This study extracts the original data related to environmental protection supervision budget, green special subsidies, and industrial pollution control investment, converts them into relative proportions, and performs dimensionless scaling to complete the initial assignment. The parameters such as R&D investment, technological benefits, and enterprise emission reduction costs and benefits were combined with the data on green technology transformation, R&D investment, and operating income in BYD’s annual reports and sustainable development reports. Furthermore, this study uses the average input–output levels of the high-end manufacturing industry as a reference for calibration, ensuring that all parameters are in line with reality. The specific parameter values are shown in Table 5. Table 5Parameter assignment.ParameterAssignmentParameterAssignmentParameterAssignment1237868151012461051432016 Parameter assignment. Simulation analysis of initial probability Figure 3 shows the impact of simultaneous changes in the initial intentions of the government, R&D institutions, and high-tech manufacturing enterprises on the evolution trajectories of the three parties, under the condition that other parameters remain constant. This study assumes that the initial intentions of the three parties are all (0.5, 0.5, 0.5), meaning that the government, R&D institutions, and high-tech manufacturing enterprises have an equal probability of adopting each strategy at the initial moment. The horizontal axis in the figure represents the passage of time, while the vertical axis reflects the probability of the participants’ strategy choices. The curve depicts the process of behavioral evolution. As can be seen from the figure: The threshold for the simultaneous change in the initial intentions of the three parties is 0.5. As time goes by, when the value is less than this threshold, the government will adopt a stricter supervision strategy at a faster pace. When the level exceeds this threshold, high-tech manufacturing enterprises will rapidly evolve towards a green value-driven emission reduction strategy. The speed at which R&D institutions achieve balance is relatively slow. Finally, it evolved into a low investment R&D strategy. The system eventually converges to E4(0,1,1), which is consistent with the theoretical analysis. E4 (0,1,1) evolution simulation results. Sensitivity analysis of three parties cost expenditure In the game system of tripartite collaborative emission reduction, the cost of each party choosing an active strategy is a key factor influencing their decision-making intention. This study conducts a sensitivity analysis on the government strict supervision costs , R&D institutions high investment costs , and green value-driven emission reduction costs of high-tech manufacturing enterprises. Based on this, the study explores and concludes the impact of changes in participation costs on the system’s evolution path, as shown in Fig. 4. When equals 10, the probability of the evolutionary strategy adopted by the R&D institution is 1. As the high investment costs of R&D institutions rose to 12, 14, and 16, the strategy shifted from being proactive to being defensive. And the rate of evolution gradually increased and eventually stabilized at 0. When equals 12, the enterprise tends to adopt green value-driven emission reduction measures, which leads to the rapid convergence probability of the three-party active strategy towards 1. When the value of is 14, 16, or 18, the “cost threshold” for emission reduction increases. The evolution rate for enterprises to reduce emissions has slowed down, and the probability of adopting proactive strategies has dropped to 0.2. When equals 6, the probability of the government evolutionary strategy is 1. As supervision costs rose, the government’s strategy shifted from active to negative, and finally stabilized at 0. Sensitivity analysis of three parties cost expenditure. Further analysis reveals that all three parties have a relatively high sensitivity to their respective costs. For the government, an increase in supervision costs will reduce the intensity of its environmental protection constraints, lower the cost for enterprises to violate supervision, and diminish the internal driving force for green emission reduction. At the same time, the demand for green technologies in enterprises has declined, and government research subsidies have also decreased. As a result, the economic efficiency of high investment R&D by R&D institutions has dropped. Ultimately, the strategies of both the R&D institutions and the enterprises have tended to be negative. At the level of R&D institutions, high R& costs will reduce the cost-effectiveness of technological research and development. The profits from patent transfer and technology implementation are insufficient to cover the high R&D costs, which has led R&D institutions to adopt a low investment strategy. Meanwhile, the supply of industrial green microgrids technologies has shrunk. Enterprises find it difficult to reduce their green emission costs by relying on mature technologies. Therefore, they will adopt supervision compliance-driven strategies. Poor emission reduction performance by enterprises will increase environmental pollution and the pressure for environmental governance. This will increase the burden on the government for comprehensive management. At this point, the government tends to adopt a lenient supervision strategy. At the level of high-tech manufacturing enterprises, when the cost of green value-driven emission reduction increases, the emission reduction cost is significantly higher than the benefit. This will reduce the enthusiasm of enterprises to take active measures for emission reduction. The willingness of enterprises to adopt green technologies has decreased, and the market demand for green technologies has shrunk. As a result, R&D institutions will reduce their investment in technological research. Therefore, the government can reduce the marginal cost of strict supervision through digital supervision and cross-departmental collaboration, and ensure the supervision impetus. R&D institutions can utilize technology sharing platforms, collaborative research and industry-university cooperation to share the R&D costs, thereby enhancing the cost-effectiveness of technological innovation. Enterprises reduce their emission reduction costs by applying for green technology subsidies and participating in carbon trading, thereby promoting the transition from “negative compliance” to "active emission reduction". Sensitivity analysis of the three parties benefit mechanism The benefit mechanism serves as the intrinsic driving force that prompts each party to adopt active strategies. Through conducting a sensitivity analysis on the R&D institutions technologies adoption benefits and manufacturing green emission reduction market benefits , this study explores and concludes the impact of benefits changes on the evolution path of the system, as shown in Fig. 5. When equals 10, the probability of the strategy evolution of the R&D institution is 0. As increases to 12, 14, and 16, the strategy shifts from negative to active, and shows a stepwise growth trend, eventually stabilizing at 1. As the value of increases from 14 to 20, the curve of high-tech manufacturing enterprises shows a clear positive correlation, and eventually approaches 1. As the values of and increase, the government evolves from an active strategy to a negative one. Sensitivity analysis of the three parties benefit mechanism. Further analysis reveals that R&D institutions are highly sensitive to the direct benefits of technology adoption. Higher investment benefits can effectively strengthen their willingness to invest heavily in R&D. High market benefits can directly prompt high-tech manufacturing enterprises to adopt green value-driven emission reduction strategies. Under the joint efforts of the R&D institutions and enterprises to actively reduce emissions, the environmental pollution problem has been improved, and the government’s pressure for environmental governance has been significantly reduced. The marginal benefit of strict government supervision has decreased. From a cost–benefit perspective, the government no longer needs to maintain such high-intensity control. Therefore, the strategy has shifted from strict supervision to lenient supervision. R&D institutions can establish technology exchange markets and improve intellectual property protection mechanisms to enhance the commercial returns on technological achievements. This ensures that high-investment R&D receives commensurate returns, thereby stimulating enthusiasm for R&D. Enterprises can amplify the market benefits of green emissions reductions by cultivating green consumer markets and expanding carbon-trading channels. This approach drives companies to transition from a “negative compliance” model to an "active profit-seeking" approach to reducing green emissions. Sensitivity analysis of government subsidy mechanisms Subsidies are an important policy tool used by the government to guide market behavior. The subsidy mechanism includes two components: is the government specific subsidy that is awarded to R&D institutions when their high investment R&D technology are not adopted by high-tech manufacturing enterprises; is the additional reward given by the government to high-tech manufacturing enterprises. The influence of the subsidy mechanism on the evolution path of the system is shown in Fig. 6. As and increase, the government gradually evolves towards a more active state, and the evolution rate becomes faster. The evolution rate of the R&D institutions has changed, but their stable strategies remain the same. High-tech manufacturing enterprises have evolved to a stable state of 1, and the evolution rate has been increasing rapidly. Sensitivity analysis of government subsidy mechanisms. Further analysis reveals that when the special subsidies increase, government fiscal expenditures also rise. Implementing strict supervision not only incurs supervision costs but also requires continuous expenditure on high-level R&D subsidies. At this point, the government’s overall fiscal burden has increased and its net income has decreased. Therefore, the government is accelerating its shift towards a negative strategy of lenient supervision. For R&D institutions, they cannot obtain the benefit from technology transfer because enterprises do not adopt the technologies developed by these institutions with high investment. Choosing a negative strategy does not require incurring high R&D costs. Research institutions can obtain stable income merely through government subsidies, and under economic rationality, their stable strategy remains unchanged. Furthermore, even if the technology cannot be implemented, the R&D institutions can still receive considerable subsidies, and the losses risk in the research process is somewhat mitigated. As a result, the evolution rate of the R&D institutions towards a negative stable state has slowed down. The government has increased the additional incentives for high-tech manufacturing enterprises. The costs for the government are higher than the benefits. Therefore, the trend of the government moving towards more lenient supervision remains unchanged, and the convergence speed has accelerated. The government’s additional incentives directly increase the green emission reduction benefits for enterprises, thereby motivating them to undertake green transformation. Therefore, the government can optimize the subsidy structure, moderately increase the subsidy amount, and precisely support the technological innovation of R&D institutions. The government should avoid excessive incentives and should link the rewards to the emission reduction performance, avoiding “inefficient rewards”, and guiding enterprises to focus on enhancing green value. The R&D institutions utilize the subsidy funds to optimize their R&D efforts, and increase the green technology implementation and transformation rate. High-tech manufacturing enterprises should fully leverage the green incentive benefits provided by the government and continuously increase their green emission reduction investment, achieving a mutually beneficial win–win situation for both enhancing enterprise efficiency and generating benefit from R&D. Sensitivity analysis of punishment mechanism The punishment mechanism constrains the behavior of the parties by increasing the cost of violations. Through a sensitivity analysis of the fines that high-tech manufacturing enterprises have to pay when they fail to meet emission reduction standards due to strict government supervision, this study explores the changes impact in the fines on the system’s evolution path, as shown in Fig. 7. As continues to increase, the government’s evolving strategy changes, but its stable strategy remains the same. For R&D institution, when = 4, the probability of the R&D institution adopting the strategy is 0. When reaches a certain threshold, that is, when , the strategy of the R&D institution changes from negative to active and eventually stabilizes at 1. For high-tech manufacturing enterprises, when equals 4 and 8, the probability of the enterprise’s strategy is 1. As increases, the enterprise’s strategy shifts from active to negative, and the evolution rate becomes increasingly rapid. Sensitivity analysis of punishment mechanism. Further analysis reveals that as the punishment standards for enterprises failing to meet emission reduction targets have increased, the three-party evolution has shown a differentiated feature. Increasing fines boosts government supervision benefit, but harsh penalties dampen the overall transformation momentum of the industry and raise the long-term cost of environmental governance. The active and negative effects cancel each other out, and the government’s strategy shifts from a more negative 0.2 to a neutral range of 0.4. From the perspective of enterprises, most manufacturing companies have invested more in industrial green microgrids technology throughout the entire process than they would have spent on continuous violations and penalties. From the perspective of economic benefits, most enterprises will choose to regularly pay fines and abandon the initiative to reduce emissions. Their overall strategy shifts to a negative approach. However, there has been a divergence in industry-wide demand. Some leading enterprises, in order to maintain their production qualifications and avoid large penalties, have a strong demand for purchasing emission reduction technologies, which has driven R&D institutions to make significant investment in R&D. This evolutionary pattern is in line with the market stratification phenomenon observed in reality, where severe penalties lead to a need for major technological upgrades among the leading enterprises, and smaller entities comply with the penalties to discharge pollutants. Eventually, this results in a differentiated outcome where the overall enterprises tend to be more negative while R&D becomes more active. Therefore, the government can, based on the actual situation of the industry, set reasonable punishment standards. The punishment funds can be specifically allocated for the R&D of emission reduction technologies and subsidies for enterprises to achieve green emission reduction, thereby enhancing the efficiency of collaborative emission reduction. Conclusions and implications Main conclusions By constructing a three-party evolutionary game model involving the government, R&D institutions, and high-tech manufacturing enterprises, this study systematically analyzed the strategic interactions and evolutionary paths of the three parties under the collaborative effect of the green credit regulatory system and industrial green microgrid technology. Then, this study conducted numerical simulations on the influence of changes in parameters such as costs, benefits, rewards, and punishments on the evolutionary path of the game system. The main conclusions are as follows: First, cost barriers significantly constrain third parties’ strategic choices. When governments impose strict cost controls, research institutions incur high investment costs, and enterprises face excessively high costs for green emission reductions, these factors suppress stakeholders’ willingness to adopt proactive strategies. Particularly when a company’s green emission reduction costs exceed the combined value of its market gains and policy incentives, the system tends to converge toward a passive state driven solely by regulatory compliance. This makes it difficult to achieve genuine green-value-driven emission reductions. Second, the revenue mechanism serves as the core endogenous driver propelling the system’s green and collaborative evolution. High-level returns can generate com-prehensive positive incentive effects. The direct benefits from adopting R&D institutions’ technologies and the market gains from corporate green emissions reductions can effectively incentivize their proactive R&D investments and voluntary emissions reductions. When the benefits of technology transfer and market returns become sufficiently significant, even with lenient government regulation, the system will naturally gravitate toward a virtuous cycle of high R&D investment and substantial emissions reductions, making a decisive contribution to sustained green emissions reduction. Third, the design of government reward and punishment tools directly determines the effectiveness of the system’s regulatory measures. Specialized R&D subsidies effectively incentivize research institutions to pursue high-investment innovation, with subsidy levels positively correlated to innovation enthusiasm. Conversely, substantial rewards for enterprises may foster incentive dependency, thereby dampening their intrinsic motivation to reduce emissions. Furthermore, increasing fines for violations by a reasonable amount can significantly strengthen the constraints on corporate emissions reduction. However, this must be balanced with regulatory costs and corporate affordability to avoid disproportionate penalties. Theoretical implications Based on the model construction, evolutionary analysis, and simulation validation results of this study, the following three theoretical insights can be drawn: First, this study fills the analytical gap regarding technology providers in multi-entity emissions reduction governance. Existing research on energy conservation and emissions reduction in manufacturing has primarily focused on the interaction between government and enterprises, without sufficiently incorporating research institutions as the core technology providers. This study constructs a tripartite evolutionary game framework involving "government-R&D institutions-high-tech manufacturing enterprises", addressing oversight in traditional dual-subject analyses of the technology supply chain. It responds to the research call by Ding et al. (2025) to "strengthen the theoretical positioning of technology entities within low-carbon governance system42 ". Second, this study responds to the theoretical call for institutional and technological synergies to drive green transformation. Existing research has primarily examined the isolated effects of either the green regulatory system or the industrial green microgrid technologies. This study reveals the interactive mechanism through which these two elements mutually reinforce each other-policy guiding technological R&D while technology supporting policy implementation-demonstrating that the synergistic alignment of institutional frameworks and technological innovations is crucial for driving deep corporate emissions reductions. This conclusion aligns with the United Nations Environment Program’s theoretical advocacy that "integrating institutional tools and technological innovation is essential for achieving low-carbon industrial transformation", thereby enriching the theoretical framework of cross-dimensional collaborative emission reduction. Third, this study expands the scope of application of evolutionary game theory in the field of low-carbon technologies. Existing applications of evolutionary game theory in energy conservation and emissions reduction primarily focus on static strategic equilibria. This study employs dynamic simulation to analyze the three-party strategic interactions across the entire chain of industrial green microgrid technology R&D, transformation, and application. It quantitatively assesses the marginal impact of various parameters on the evolutionary trajectory of the game system, thereby over-coming the limitations of simplified variables and static equilibrium inherent in traditional evolutionary game theory research. Practical implications Based on the aforementioned theoretical breakthroughs, this study has drawn the following practical implications: First, rely on the green credit grading supervision to optimize management costs and implement a tiered disciplinary mechanism. Based on the current situation of the environmental protection credit platform in our country and the cross-departmental credit information linkage, promote the interconnection of ecological environment, industry and information technology data, and use digital rating to reduce regulatory expenditures such as on-site inspections by the government. A tiered penalty system based on the environmental credit rating of enterprises is implemented. The fines are collected and allocated to the local green industry special fund, which is used to subsidize the implementation of micro-grid projects. By increasing the cost of non-compliance, the problem of enterprises’ passive compliance is solved. At the same time, the cost barriers for the collaborative efforts of the three parties are reduced from the institutional perspective. Second, implement differentiated and precise subsidies based on different targets, optimize the incentive structure to avoid incentive distortion. In combination with the domestic special support policies for industry-university-research cooperation, targeted research and development subsidies will be provided for R&D institutions focusing on core technologies of microgrids such as energy storage, power grid, load and storage. For projects whose technologies cannot be industrialized and implemented in the short term, transitional period guarantee subsidies will be set up. Link the green credit ratings of high-tech manufacturing enterprises with the dynamic adjustment of the reward amounts for actual emission reduction achievements, and eliminate the generous subsidies for general enterprises. At the same time, prevent enterprises from overly relying on fiscal subsidies and losing their own internal motivation for emission reduction, in order to adapt to the refined reform orientation of domestic fiscal subsidies. Third, improve the market-based system for green electricity and carbon trading within the country, and rely on the generated revenue to encourage enterprises to actively build micro-grids. Based on the actual implementation of green electricity market trading, the national carbon market, and green supply chain policies in our country, we will connect multiple revenue channels such as green electricity export, carbon quota monetization, and green loan interest subsidies, thereby increasing the operational income of enterprises in setting up green micro-grids in their factory areas. Enterprises can further reduce the financing costs of their projects by leveraging the green credit premium. By achieving market-oriented benefits such as cost reduction and revenue increase, they can be encouraged to break away from passive compliance and voluntarily adopt green value-based emission reduction solutions. Research limitations and future directions To facilitate analysis, the model constructed in this study includes only the government, R&D institutions, and high-tech manufacturing enterprises, with certain simplifications applied to relevant parameters. Therefore, subsequent research can be enhanced in the following four aspects: (1) Incorporate diverse entities such as financial institutions and industry associations to expand the analytical dimensions of multi-party collaborative games. (2) Refine parameter settings by considering the heterogeneous characteristics and dynamic adjustment patterns of enterprises across different regions and scales. (3) Incorporate external variables such as carbon market volatility and technology iteration risks to enhance the model’s adaptability to complex real-world scenarios. (4) Conduct empirical tests using field survey data and panel data to strengthen the model’s practical applicability and persuasiveness. Acknowledgements This work was supported by 2025 Liaoning Provincial Department of Education Project "Research on the Mechanism and Path of Enabling Data Elements to Improve the Quality and Efficiency of Green Finance in Liaoning Province" (Project number: LJ112510142004). Author contributions Conceptualization, H.Q. and X.Z.; methodology, X.Z.; software, H.Q.; validation, Z.C.; formal analysis, X.Z.; investigation, Y.C.; resources, H.Z.; data curation, X.Z.; writing—original draft preparation, M.A.; writing—review and editing, H.Q.; visualization, H.Z.; supervision, M.A.; project administration, Z.C.; funding acquisition, H.Q. All authors have read and agreed to the published version of the manuscript. Funding This work was supported by 2025 Liaoning Provincial Department of Education Project "Research on the Mechanism and Path of Enabling Data Elements to Improve the Quality and Efficiency of Green Finance in Liaoning Province" (Project number: LJ112510142004). Data availability The datasets generated during and the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets generated during and the current study are available from the corresponding author on reasonable request.