In the context of China's transition from "dual control of energy consumption" to "dual control of carbon emissions," understanding the synergistic mechanisms among carbon emission trading (CET), energy use rights trading (EURT), and electricity markets is critical for achieving the nation's dual carbon goals. This study develops a system dynamics (SD) model to examine the coupled interactions within this "carbon-electricity-energy" ternary market system, focusing on thermal power enterprises as the primary analytical subject. The model reveals that the ternary market framework drives energy conservation and emission reduction through three key mechanisms: price signal transmission, dual regulatory constraints, and mutual quota recognition. These mechanisms propagate low-carbon incentives throughout the industrial chain by transmitting cost signals to end-users via electricity prices. Compared to binary market structures, the ternary framework achieves superior outcomes, it facilitates higher renewable energy consumption, maintains more stable price levels, enhances market liquidity for both carbon and energy rights, and improves resource allocation efficiency alongside environmental-economic performance. However, the simulation also exposes critical inefficiencies under the current "dual control of energy consumption" regime. The parallel operation of EURT and CET markets creates functional overlap and duplicated compliance burdens. This redundancy increases enterprise costs without commensurate environmental gains, validating the necessity of transitioning to carbon-focused dual control. Further analysis demonstrates that a mutual recognition mechanism between carbon and energy rights effectively alleviates dual compliance pressures and improves enterprise profitability. Optimal market performance emerges when the recognition ratio is appropriately calibrated. Additionally, gradually increasing the share of auctioned quotas while maintaining appropriate levels of free allowances can drive emission reductions without compromising enterprise profitability. This research provides both theoretical foundations and practical policy recommendations for building an efficient multi-market coordination mechanism, facilitating the policy transition, and advancing low-carbon transformation in China's power sector.

To enhance the precision of short-term load forecasting at regional electric vehicle (EV) charging stations, it is essential to address both the intrinsic patterns within historical load data and the variability caused by weather conditions. This study presents a forecasting framework combining an improved triangulation topology aggregation optimizer (ITTAO), a bidirectional gated recurrent unit (BiGRU), and a convolutional neural network (CNN). First, a BiGRU-CNN model is first set up and the ITTAO algorithm is then employed for fine-tuning the BiGRU layers. Second, the BiGRU neural network is employed to fully capture the forward and backward correlations within historical load data; the CNN is used to carry out local optimization on this historical load data, and the forecasting process is implemented through the fully connected layer. In addition, a multi-head attention mechanism is integrated to further strengthen the model's effectiveness. Lastly, given that weather data possesses weak internal regularity, the BiGRU-CNN neural network is adopted to predict and rectify the load fluctuations induced by weather data. By taking an EV charging station in a specific area of Hebei Province, China, as a case study, the EV load was predicted for forecast horizons of 4 h and 24 h, respectively. Experimental findings demonstrate the superiority of the proposed framework over mainstream baselines. Specifically, in the 4-h working day forecast, the proposed model achieves a Mean Absolute Percentage Error (MAPE) of 2.21%, significantly outperforming the state-of-the-art Transformer model (4.35%) and the XGBoost model (5.89%). Furthermore, the Root Mean Square Error (RMSE) is reduced by over 60% compared to the uncorrected baseline, thereby validating the effectiveness of the methodology presented in this research.

Accelerating new energy vehicle (NEV) adoption is critical for transport decarbonization worldwide, particularly as policy-driven diffusion increasingly shifts toward market-oriented and household-demand mechanisms. Drawing on institutional configuration theory, this study develops a government-market-society (GMS) framework and examines 31 Chinese provinces from 2011 to 2024, divided into three stages aligned with China’s Five-Year Plans: 2011–2015, 2016–2020, and 2021–2024. By integrating necessary condition analysis (NCA) and multi-period fuzzy-set qualitative comparative analysis (fsQCA), we identify necessary constraints, sufficient configurational pathways, and the temporal reconfiguration of core conditions for high provincial NEV adoption. Results show that no single factor is necessary or sufficient; high adoption arises from conjunctural causation and equifinality with multiple concurrent pathways. Specifically, Period 1 is dominated by a market-society-driven pattern; Period 2 shows coexisting market-consumption and government-market synergy configurations; and Period 3 presents both a household-income-led and a government-market-society collaborative configuration. Over time, the core conditions evolve from “market competition + public environmental concern” to “market competition,” and finally to “household income,” indicating a dynamic reconstruction of institutional logics. These findings clarify the causal and temporal complexity of NEV adoption and provide implications for stage-specific diffusion strategies and adaptive policy mixes.

Artificial intelligence (AI) is commonly assessed through the electricity used by models and data centres, leaving the carbon transferred through the wider production system insufficiently resolved. We disaggregate China’s AI industry from the broader information and communication technology sector and construct comparable 31-sector environmentally extended input–output accounts for 2015, 2017, 2020 and 2023. Embodied-emission accounting is integrated with linkage analysis, structural path analysis and structural decomposition analysis to quantify the footprint, locate critical supply-chain pathways and explain temporal change. AI-related embodied CO2 emissions increased from 140.03 Mt in 2015 to 324.37 Mt in 2023. Indirect emissions reached 187.01 Mt in 2023 and remained higher than direct emissions, while backward linkages consistently exceeded forward linkages. Approximately half of the footprint was concentrated within the first three production tiers, with other ICT, electricity and heat, transport, metals and non-metallic minerals forming the largest short paths. In the current-price SDA, domestic final demand was associated with a 138.82 Mt increase between 2020 and 2023, outweighing the 20.24 Mt reduction from production-structure change. The results recast AI decarbonisation as a systems-governance problem: operational efficiency must be coupled with demand management, low-carbon electricity, hardware circularity and targeted upstream procurement.

Under the strategic goals of ecological protection and high-quality development, whether the Yellow River Basin (YRB) can efficaciously enhance its carbon emission efficiency (CEE) via green technology innovation (GTI) bears significant practical implications for attaining a win-win situation of “green” and “wealth.” Using panel data from 9 provinces in the YRB from 2007 to 2021, this study employs standard deviation ellipse analysis to examine the spatial pattern and temporal evolution of CEE in the region. And it also applies the spatial Durbin model to explore how GTI influences CEE from the perspectives of inventive green technology innovation (INGTI) and improved green technology innovation (IMGTI). Additionally, the study explores the moderating role of industrial agglomeration (IA) in this relationship. The results reveal that the CEE in the YRB is distributed from southwest to northeast, with clustering in the east-west direction. CEE’s center of gravity generally moves towards the northeast of the basin, with an overall stable evolutionary trend. In the YRB, both INGTI and IMGTI can effectively enhance the CEE of neighboring provinces while promoting local CEE. However, their impacts are spatially heterogeneous. In upstream provinces, INGTI has a stronger effect on CEE than IMGTI, whereas in lower-middle stream provinces, the enhancing effects are consistently well demonstrated. Additionally, IA exhibits significant moderating effects, which vary according to the type of GTI. It exerts a significant negative linear moderating effect on the CEE enhancement from INGTI, while demonstrating a U-shaped moderating effect on the CEE improvement from IMGTI. Furthermore, this study presents corresponding policy recommendations to the above results.

Prior research on tour guides' influence on tourists' pro-environmental behavior has largely emphasized informational content (e.g., interpretation) and communication tactics (e.g., humor), while leaving the social-influence role of guides' emotional displays underexamined, especially with respect to simultaneous affective and cognitive mechanisms. Drawing on Emotions-as-Social-Information (EASI) theory, we develop a dual-path model in which tour guides' environmental passion affects tourists' pro-environmental behavior via an affective-reaction pathway (positive emotions) and an inferential pathway (self-protection motivation), with tourists' self-construal moderating the first-stage effects. Using a seven-day experience sampling (intensive longitudinal) survey (873 day-level observations nested within 159 tourists) and estimating a 1-1-1 multilevel structural equation model with Monte Carlo confidence intervals, we find that guides' environmental passion predicts tourists' pro-environmental behavior both directly and indirectly through the two mediators, and these indirect effects are stronger among tourists with a more interdependent self-construal. The study extends EASI theory to guide-tourist interactions and advances tourism sustainability research by clarifying how emotional displays operate as social information in shaping tourists' daily pro-environmental responses.

To address the issues of limited interpretability and low predictive accuracy in traditional photovoltaic forecasting models, this paper proposes a hybrid forecasting model named HO-Transformer-KAN-PINN. First, Maximal Information Coefficient (MIC) is used to select the key meteorological features: irradiance and temperature. Then, the grey relational analysis combined with cosine similarity is applied to identify similar days. The prediction framework is then constructed. The Transformer-KAN model provides high predictive accuracy and strong interpretability, while embedding physics-informed neural network (PINN) constraints enforces compliance with the underlying physical laws, yielding the Transformer-KAN-PINN framework. Simultaneously, the Hippopotamus Optimization (HO) algorithm is used to optimize the model hyperparameters. Finally, the photovoltaic power combination prediction model of HO-Transformer-KAN-PINN is constructed. This model has achieved excellent results in short-term photovoltaic power forecasting in Yunnan, Gansu, and Australia. Taking winter in Yunnan Province as an example, the forecasting results of this model yield an MAE of 0.3204 MW, an RMSE of 0.4197 MW, a MAPE of 4.9561%, and an R2 of 0.9986. Therefore, the hybrid forecasting model proposed in this paper demonstrates a certain degree of advancement and effectiveness. Therefore, it provides reliable technical support for accurate prediction of photovoltaic output.

Amid global climate change and environmental degradation, effective environmental governance has become a central concern for scholars and policymakers. This study proposes a pressure-state-response-environment (PSRE) framework that links external pressures (public and peer pressure), regional states (population, affluence, and technology), and organizational responses (government regulation and environmental fiscal expenditure) to assess environmental performance (EP). Using multi-period fuzzy-set qualitative comparative analysis (fsQCA) on panel data from 31 Chinese provinces (2018-2023), this study provides a replicable model for regional governance research. The results show that no single factor is indispensable for either high or non-high EP; instead, multiple equifinal configurations exhibit distinct temporal patterns. Comparing high and non-high EP configurations shows causal asymmetry, as pathways to non-high EP do not mirror those leading to high EP. For high EP, three dominant models emerged in Period 1: external pressure-regional state, external pressure-regional state-organizational response, and external pressure-organizational response. In Period 2, four models emerged: regional state-organizational response and external-pressure pathways were added, whereas the external pressure-organizational response model disappeared. Among the high-EP configurations, peer pressure is pervasive in both periods, while the dominant policy lever shifts from environmental fiscal expenditure in Period 1 to government regulation in Period 2, consistent with China's move from financial support to stricter enforcement and institution building. Moreover, China's eastern, central, and western regions exhibit distinct environmental governance pathways. These findings advance environmental governance theory and offer policymakers actionable insights for designing context-sensitive strategies that account for regional heterogeneity and evolving governance pathways.

To address renewable-output volatility, complex device mechanisms, and uncertainty-aware dispatch challenges in park-level integrated energy systems (IESs) with the high renewable penetration, an integrated ‘forecasting-simulation-optimization’ framework is proposed. Firstly, a unified CSGW-PI framework is developed, where task-adaptive deep networks are constructed for photovoltaic and wind power forecasting, as well as surrogate simulation of energy output in gas turbine and electrolyzer. The proposed device surrogate models significantly correct the bias of conventional mechanistic equations, reducing the MAE and RMSE of GT power-output prediction by 58.42% and 48.11%, respectively. Secondly, forecasted renewable outputs are used as upper bounds of renewable energy availability, while GT-DLNN and EL-DLNN for the equipment replace fixed-efficiency formulations through deep integration of equipment simulation and schedule optimization, thereby establishing operational optimization model of IES integrating CHP, P2G, and multiple energy storage units. The results provided by deterministic optimization model show that the fully configured scenario achieves the lowest dispatch cost, 14.61% lower than the benchmark, with wind and photovoltaic curtailment cost limited to only 130.04 CNY. Referring to the deterministic optimal scenario, a multi-weight IGDT framework is further introduced to generate both opportunity-seeking and risk-averse strategies under multi-source uncertainties. The results indicate that opportunity-seeking strategy improves the cost performance by expanding opportunity space, whereas risk-averse strategy enhances robustness at the higher operating costs. The multi-weight mechanism captures the differentiated effects of wind and photovoltaic uncertainties and enables adaptive allocation of uncertainty margins. These findings provide a surrogate-assisted and uncertainty-aware modeling basis for more reliable dispatch of IES with the high renewable penetration.

The utilization of coal-based solid waste materials (CSW) in road engineering is an important pathway for reducing stockpiling pressure, mitigating environmental risks, and promoting resource recycling. However, their large-scale diffusion is still constrained by residual engineering risk, misaligned cost and risk allocation between upstream and downstream actors, and imperfect regulatory and incentive mechanisms. To address these issues, this study develops a tripartite evolutionary game model involving the regulator, the waste producer, and the waste utilizer. The model incorporates pretreatment investment, residual engineering risk, government rewards and penalties, and green collaborative benefits to examine the evolutionary dynamics of the three parties and the stability of the system under different conditions. The results show that deep pretreatment by waste producers is a key prerequisite for the diffusion of CSW materials, as it reduces material instability and downstream engineering risk and increases the utilizer's willingness to adopt such materials. The effects of rewards and penalties are differentiated across actors: effective penalties play a stronger role in constraining low-cost disposal by waste producers, whereas rewards are more effective in encouraging adoption by waste utilizers. The interaction analysis further shows that residual engineering risk significantly constrains the positive effect of green collaborative benefits, indicating that benefit enhancement cannot substitute for risk governance. In addition, the total amount of green collaborative benefits and their release and distribution structure jointly affect behavioral convergence and system stability. The system is more likely to evolve toward a stable state characterized by deep pretreatment, active adoption, and routine regulation when benefit sharing is consistent with the costs and risks borne by each party. Based on these findings, this study suggests that differentiated policy design is needed, including stronger source pretreatment and quality control, a coordinated reward-penalty mechanism for different actors, more targeted incentives and acceptance requirements for waste utilizers, and an improved governance framework featuring quality standards, full-process traceability, and risk warning mechanisms. These measures are essential for promoting the stable and large-scale utilization of CSW materials in road engineering. By translating model results into staged regulatory, quality-control, and supply-chain actions, the findings also support broader sustainable development goals, including responsible consumption and production, resilient infrastructure, climate action, and ecosystem protection.

To address renewable-output volatility, complex device mechanisms, and uncertainty aware dispatch challenges in park-level integrated energy systems (IESs) with high renewable penetration, an integrated ‘forecasting-simulation-optimization’ framework is proposed. Firstly, a unified CSGW-PI framework is developed, where task-adaptive deep networks are constructed for photovoltaic and wind power forecasting, as well as surrogate modeling of energy output in gas turbine and electrolyzer. The proposed device surrogate models significantly correct the bias of conventional mechanistic equations, reducing the MAE and RMSE of GT power-output prediction by 58.42% and 48.11%, respectively. Secondly, forecasted renewable outputs are used as upper bounds of renewable energy availability, while GT-DLNN and EL-DLNN replace fixed-efficiency formulations through deep integration of equipment simulation and schedule optimization, thereby establishing an operational optimization model of IES integrating CHP, P2G, and multiple energy storage units. Deterministic optimization results show that the fully configured scenario achieves the lowest dispatch cost, 14.61% lower than the benchmark, with wind and photovoltaic curtailment cost limited to 130.04 CNY. Based on this scenario, a multi-weight IGDT framework is further introduced to generate opportunity-seeking and risk-averse strategies under multi source uncertainties, revealing differentiated impacts of wind and photovoltaic uncertainty and supporting adaptive uncertainty-margin allocation for more reliable renewable-rich IES dispatch.

Urbanization, while driving economic growth, has engendered urban maladies typified by frequent waterlogging, constraining urban operational safety and sustainable development. Taking the Yangtze River Delta urban agglomeration as a case, this study focuses on the critical period from 2006 to 2024 when rapid urbanization and climate variability were intertwined, integrates urbanization and flood disaster into a unified analytical framework, employs a modified coupling coordination degree (CCD) model to measure their coordination status, and uses necessary condition analysis and dynamic qualitative comparative analysis to reveal the configurational driving paths of high CCD. The results show that: (1) During the study period, the urbanization level cumulatively increased by 152.7%, while flood disaster risk exhibited a fluctuating upward trend with an increase of 10.96%. (2) CCD rose by 69.71%, upgrading from "Mild Disorder" to "Moderate Coordination", and displayed a differentiation pattern of "high in the south and low in the north". (3) The strong pull of economic urbanization and the rapid growth of social urbanization were the dominant driving forces boosting CCD, whereas the persistent deterioration of the hazard formative environment exerted a suppressive effect, and vulnerability improvement exerted a facilitative effect. (4) High CCD did not stem from a single necessary condition but resulted from the synergistic concurrence of multiple conditions, characterized by "multiple conjunctural causation" and "equifinality"; the digital economy and green innovation constituted the most prominent bottleneck conditions. (5) The eight identified configuration paths can be summarized into three patterns: digital economy-led, green innovation-led, and government intervention-openness synergy types.

Achieving the low-carbon transformation of energy consumption structure is central to carbon-neutral development, yet how artificial intelligence (AI) contributes to sustained structural change remains insufficiently understood. Addressing this gap, this study examines not only whether AI facilitates the green and low-carbon transformation of energy consumption structure in China, but also through which mechanisms and under what conditions this effect is sustained. Using provincial panel data for 2013-2022, we estimate the direct effect of AI and test the mediating role of green innovation resilience and the moderating role of green finance. The results show that AI significantly promotes the low-carbon transformation of energy consumption structure. Green innovation resilience serves as an important transmission mechanism, while green finance strengthens this relationship. These findings deepen understanding of the AI-energy transition nexus and provide policy implications for effectively coordinating digital technology, green innovation, and green finance.

To address low-carbon economic dispatch challenges in integrated energy systems (IES) under dynamic carbon emission factors (CEFs), this study proposes a hierarchical bi-level optimization framework for hydrogenenriched IES (H2-IES). The core innovation establishes an upper-level dynamic CEFs calculation framework based on carbon emission flow theory, enabling node-and-time-resolved carbon tracing to guide real-time energy pricing and market clearing. This framework dynamically quantifies carbon intensity in electricity/hydrogen procurement, reflecting temporal grid decarbonization variability and spatial energy conversion differences. The lower-level model employs a Stackelberg game with four stakeholders (producers, storage providers, users, retailers) to align energy activities with CEF-driven price signals. Two technical innovations enhance applicability: (1) a two-stage power-to-gas (P2G) process with adaptive hydrogen blending (5 %-20 %) for fuel flexibility, leveraging hydrogen/methane byproducts; (2) a P2G-CCS-CHP coupled system closing carbon loops via waste heat recovery, improving energy efficiency by 18.7 %. A dual-participant ladder carbon trading mechanism incentivizes collaboration by sharing carbon costs across consumption tiers, bridging technical optimization and behavioral adjustment. The multi-strategy enhanced dung beetle optimization algorithm (IDBOA) achieves 40 % faster computation than conventional methods. Case studies show the framework reduces total carbon emissions from 12.72 to 7.06 tons (22.1 % reduction vs. static CEF models), cuts operational costs by 6.8 %, and lowers carbon trading costs by 5.1 % through synergized carbon pricing, flexible hydrogen utilization, and collaborative cost-sharing. Results highlight how macro-level carbon accounting, meso-level energy technology, and microlevel stakeholder incentives form a self-reinforcing low-carbon dispatch mechanism, demonstrating systemic innovation for carbon responsibility allocation and economic-environmental synergy.

Accurate photovoltaic (PV) output forecasting is essential for secure grid operation, yet most existing approaches fail to capture long-term variability or physical monotonicity inherent in PV systems. This study develops a hybrid forecasting framework that integrates CGCE-SDS, VMD, RIME, PINN, and a TCN-LSTM-Attention network to enhance both accuracy and robustness. First, four representative weather patterns are identified using DPC clustering. A coupled similarity-day selection strategy based on GRA and CEEMDAN is then introduced to jointly extract short-term fluctuations and long-horizon trends. The RIME algorithm is further applied to optimize the VMD penalty coefficient and key network hyperparameters. By embedding physically grounded monotonicity constraints into a PINN-enhanced TCN-LSTM-Attention architecture, the proposed method transforms the traditional black-box paradigm into a constrained and physics-guided predictive model. Ablation and comparative experiments are conducted at PV plants in Yunnan and Gansu, China, and Alice Springs, Australia. Results show that the proposed model consistently surpasses baseline methods under diverse climatic scenarios. Under four weather conditions, the Yunnan site achieves AVGMAPE, AVGMAE, and AVGRMSE of 0.0599, 0.8734 MW, and 1.1099 MW, respectively; the Gansu site obtains 0.3187, 1.2699 MW, and 1.5782 MW; and the Australian plant records 0.3522, 0.1356 kW, and 0.2106 kW. Parallel cross-validation further confirms the robustness and transferability of the developed model across different regions and operational environments.

As a key piece of equipment enabling the cascade utilization of energy in the power-generation sector, the gas turbine exhibits the strong nonlinear characteristics during the operational process. To overcome the limitations inherent in traditional white-box, black-box, and gray-box modeling approaches, a brand-new predictive simulation method for gas turbine is innovatively proposed in this study, in which an intelligent forecasting algorithm is creatively integrated with prior physical-law constraints. Firstly, on the basis of the operating mechanism analysis of main gas-turbine modules, namely compressor, combustor and turbine, prior monotonic physical relationships are derived, including those between natural-gas consumption and electricity generation, as well as between opening degree of IGV and waste-heat flue-gas flow volume. Subsequently, a customized regression layer is constructed, such that the physical constraints are innovatively embedded into the model-training process by aid of the loss function; Meanwhile, the optimal combination of critical model parameters is determined by an improved Northern Goshawk Optimization algorithm, where multiple enhancement strategies are firstly employed. Finally, a novel INGO-CNN-GRU gas-turbine simulation model fusing with prior physical-law constraints is established. Taking a gas turbine of steel mill, China as the research subject, the comparative simulation experiments are conducted, by which the scientific validity and reliability of the proposed model are verified. The results indicate that, compared with other benchmark models, the proposed approach achieves the best predictive performance, whereby the high prediction accuracy is ensured while the conformity to prior physical laws is maintained. It is further demonstrated that under the context of physical-monotonicity constraints, the dependence on data quality can be markedly reduced, the extrapolation capability and interpretability can be improved, and a new modeling paradigm for gas-turbine simulation can thereby be provided.

The rapid growth of the digital economy has transformed the tourism industry, yet the industrial linkages and environmental impacts of this integration remain underexplored. This study employs an input-output framework to examine the interactions between the digital economy and tourism and their carbon footprint effects in China. Multi-year digital economy-tourism input-output tables for 2017, 2018, 2020, 2022, and 2023 are constructed using sectoral disaggregation and the RAS updating method. Results indicate increasing integration, with tourism more dependent on the digital economy sectors and both industries exerting the strongest influence on the secondary sector. The digital economy shows a gradual shift from hardware manufacturing to information services. Structural decomposition analysis and structural path analysis reveal that technological progress significantly reduces emissions, whereas population growth drives increases. These findings offer empirical evidence for guiding digital-tourism integration and supporting low-carbon strategies in the tourism sector.