网络首发
Application of grey clustering and cloud model method in the evaluation of lock emergency management capacity
Yang E;To address the significant issues of insufficient rationality in the subjective and objective weight allocation of the indicator system， as well as the challenges in accurately quantifying the uncertainty of management capability in evaluating ship lock emergency management capabilities， this study aims to establish a scientific， reliable， and practically applicable evaluation model for emergency management capabilities. This model provides robust technical support for the safe， efficient， and seamless operation of inland waterway transportation systems. Based on the four core dimensions of emergency management—namely "prevention， early warning， response， and recovery"—a comprehensive evaluation system was constructed， consisting of 11 secondary indicators and 31 tertiary indicators that cover all key aspects of ship lock emergency management. The improved entropy weight method was employed to optimize the allocation of indicator weights， reducing subjective biases. Additionally， the grey clustering method was integrated with the cloud model to comprehensively quantify the inherent randomness and fuzziness associated with emergency management risks. This approach allows for the scientific and accurate determination of emergency management capability levels through multi-method result comparison and verification. In the case application， the improved entropy weight method revealed significant differences in the importance of various evaluation indicators， highlighting that "emergency disposal and rescue" and "prevention and emergency preparedness" are the primary driving factors for emergency management evaluation， with their combined weight exceeding 60%. The combined use of the grey clustering method and the cloud model effectively overcame the limitations of single-method approaches in determining critical level values， allowing for precise identification of the system's weak links. Based on detailed case evaluation results， this study proposes targeted improvement strategies for emergency capabilities: "strengthening weak points and consolidating core competencies" and "stabilizing advantages and securing boundary capabilities" for two specific ship locks. These strategies provide a valuable theoretical foundation and practical operational guidance for the targeted enhancement of inland waterway ship lock emergency management capabilities and the rational optimization of emergency resource allocation.
Mechanisms and planning strategies for the impact of spatial configuration of centralized payment facilities on carbon emissions
Zhang He;Liu Fanfei;Li Yantong;Yang Xingyuan;Under the “Dual Carbon” goals， the mechanisms through which the spatial distribution of public service facilities influences residents’ behavior and carbon emissions require further elucidation. This study focuses on centralized payment facilities in Nankai District， Tianjin， constructing an analytical framework of “spatial distribution—payment behavior—travel-related carbon emissions.” Employing a combination of online and offline methods， a resident questionnaire survey was conducted， yielding 2244 valid responses that collected data on residents’ payment behavior characteristics， modes of travel， travel distances， and travel frequencies. Based on this data regarding actual travel modes， distances， and frequencies， travel-related carbon emissions were calculated. An indicator system for the spatial configuration of centralized payment facilities was developed across three dimensions: activity associations， usage characteristics， and travel characteristics. Building on this foundation， an XGBoost model was utilized to identify the nonlinear relationship between spatial configuration variables and residents’ travel-related carbon emissions. The SHAP method was then applied to analyze the relative importance of each variable on carbon emissions， further revealing low-carbon effects under different levels of spatial configuration. The research results indicate that: (1) Residents’ payment behaviors exhibit distinct “high-frequency， short-distance” characteristics; spatial configuration affects travel-related carbon emissions by influencing these payment behaviors， with behavioral responses serving as a key mediating factor between spatial configuration and carbon emissions; (2) A significant nonlinear relationship and threshold effect exist between spatial configuration variables and travel-related carbon emissions， with road classification showing the strongest explanatory power， followed by accessibility to facilities of the same brand; the contributions of different variables to carbon reduction vary markedly across different value ranges; (3) SHAP analysis further indicates that functional diversity， facility type heterogeneity， and walkability are critical spatial dimensions influencing residents’ travel-related carbon emissions; enhancing these dimensions can reduce the demand for short-distance motorized travel. Based on these findings， this study proposes a coordinated optimization strategy focused on enhancing functional mix， advocating for the synergistic allocation of facility types， and improving walkable environments to enhance the low-carbon efficiency of the spatial allocation of public service facilities. This study provides empirical evidence to support the transition of spatial planning for centralized payment facilities from a scale-oriented approach to one that is responsive to behavior and emphasizes low-carbon performance.
Mechanical properties of Q420C steel after fire induced high temperature
Yang Liguo;Lu Yuhao;Su Zhe;To clarify the evolution of mechanical properties in Q420C steel after fire exposure， tensile tests were conducted on specimens subjected to temperatures of 20 ℃， 200 ℃， 400 ℃， 600 ℃， and 800 ℃， followed by either natural cooling or water quenching. A room-temperature group served as the reference， resulting in a total of nine test conditions. The specimens were heated in a KSL–1700X furnace at a rate of 10 ℃/min， held at the target temperature for 30 minutes， and then cooled to room temperature. Random speckle patterns were sprayed on the specimen surfaces， and full-field strain development was recorded using a VIC–3D digital image correlation system during monotonic tensile loading. Engineering stress-strain curves and characteristic parameters， including yield strength， tensile strength， elastic modulus， yield-to-tensile strength ratio， and elongation after fracture， were obtained for each condition. Finite element simulations were then performed in Abaqus by assigning the measured material parameters to the corresponding post-fire conditions. The results indicate that the mechanical properties of Q420C steel change only slightly after exposure to temperatures not exceeding 600 ℃. In this range， both yield strength and tensile strength increase modestly， the elastic modulus remains nearly unchanged， and elongation after fracture varies within a narrow band. When the exposure temperature reaches 800 ℃， the impact of the cooling mode becomes pronounced. After natural cooling， strength decreases， but ductility improves significantly. In contrast， after water quenching， yield strength drops markedly， and the yield-to-tensile strength ratio is sharply reduced， while tensile strength remains relatively high. Surface oxidation becomes progressively severe with increasing temperature， and noticeable scale spalling occurs once the exposure temperature exceeds 600 ℃， particularly for water-quenched specimens. A comparison with published results for Q235， Q355， Q460， and Q690 steels shows that Q420C steel follows a similar overall temperature-dependent trend， but exhibits a more pronounced strength increase around 600 ℃ and greater degradation of yield strength after exposure to 800 ℃ followed by water cooling. The simulated stress-strain curves align well with the experimental results， with prediction errors for yield strength and tensile strength remaining within 10%. These findings can serve as a basis for post-fire performance assessment and provide guidance for the repair or strengthening design of Q420C steel components.
Scenario-based dynamic allocation and optimal scheduling method for terminal fire emergency resources
Li Hang;Feng Yuhang;He Shuya;Chong Xixiao;Hu Xiaobing;To address the issues of inadequate dynamics in resource allocation and poor targeting in resource scheduling for existing terminal fire emergency responses， this paper proposes a scenario-based dynamic allocation and optimal scheduling method for emergency resources in the context of terminal fires. Considering the complexity and variability of fire occurrences， the proposed method first identifies the scenario elements relevant to terminal building fires， analyzes the temporal accessibility of fire information， and constructs a comprehensive fire scenario. Building on this foundation， we analyze and establish potential quantitative relationships between the characteristics of fire occurrences and the corresponding emergency response resources. This leads to the development of a dynamic allocation method for emergency resources tailored to specific fire scenarios. Subsequently， focusing on aspects of fire rescue and medical care， we establish a dynamic allocation and optimization scheduling method aimed at minimizing emergency response time， maximizing rescue capacity， and reducing emergency response costs. We then formulate a multi-objective optimization model for emergency resource dispatching， which generates a targeted dynamic configuration and dispatching scheme for terminal fire emergency resources. The objectives include achieving the shortest emergency response time， maximizing rescue capability， and minimizing emergency response costs. A case study is conducted at Tianjin Binhai International Airport， where seven firefighting stations and seven medical rescue stations are selected. Simulations are carried out under three fire-evolution scenarios with burning areas of 50.24 m2， 100 m， and 1 000 m， along with the corresponding casualty scenarios. For burning areas of 100 m and 1 000 m， the optimized schemes demonstrate actual firefighting capacities of 180 m and 1 010 m， respectively. When transporting 17 and 315 injured individuals， the medical rescue capacities reached 42 and 335 persons， respectively， thereby meeting the anticipated demands. The simulation results indicate that compared to traditional hierarchical response strategies for emergency resource allocation， the method proposed in this study significantly enhances both the targeting and efficiency of resource allocation for terminal fires， ultimately improving the emergency response capabilities for airport fires.
Study on the bandgap characteristics of single-point defect states in continuous periodic wave-impeding block barriers
Wang Haiyang;Yu Junjie;Zhai Chunxin;Wang Zhu;Li Shoulong;Defect states are a critical factor affecting the bandgap characteristics and vibration isolation performance of continuous periodic wave-impeding block barriers. To investigate the evolution of bandgap characteristics in periodic wave-impeding blocks under single-point defect states， a 5×5 supercell lattice model was first established. The finite element method was employed to examine the effects of central material absence， core filling ratio， and core shape variation on the bandgap characteristics， with an analysis of the generation mechanism of defect bands from the perspective of the wave vector. Furthermore， an indoor model test was conducted for the core-absence case to validate the findings. The results indicate that specific single-point defects can generate defect bands within the complete bandgap， thereby reducing the vibration isolation performance of the continuous barrier. In particular， central core absence creates seven defect bands within the bandgap range of 41.1–50.9 Hz， weakening the vibration suppression effect at these frequencies. As the core filling ratio increases， the defect bands gradually diminish and eventually disappear， leading to the formation of a complete bandgap. Additionally， the core shape significantly influences the generation of defect bands; a circular core produces no defect bands within the bandgap range of 41.1–51.1 Hz， whereas a cross-shaped core generates a single defect band at 43.2 Hz. Mechanism analysis shows that single-point defects alter the equivalent mass， stiffness， and symmetry of the local unit cell， facilitating the formation of defect bands within the bandgap. Indoor model test results demonstrate that after core absence， the insertion loss difference of the continuous barrier is generally positive， with an average value of 0.98 dB. Moreover， near 50.0 Hz， the acceleration response of the core-absence barrier exceeds that of the intact barrier， aligning with the overall trend observed in the numerical simulation results. These findings provide valuable insights for the optimal design and construction quality control of continuous periodic wave-impeding block barriers.
Research on enterprise fire safety credit evaluation and application under the background of new quality productive forces
Chen Xin;Liu Shiliang;Zeng Mingrong;Shen Xinyao;To effectively enhance corporate fire safety credit levels and mitigate fire incidents， this study develops a risk-structured database for fire safety misconduct based on the three elements of new quality productive forces. We construct the "452" conceptual model underlying fire safety. By analyzing influencing factors from a system integration perspective， we establish an evaluation indicator system comprising three dimensions—cycle， review， and calculation—organized into three categories: cycle， common， and industry-specific， which quantitatively characterizes the basic weights. A Similarity Comparison Ranking (SCR) method is proposed to adjust weights for small sample sizes， based on four evaluation objectives: data sufficiency， data timeliness， similarity of influencing factors， and similarity of accident consequences. An improved Technique for Order Preference by Similarity to Ideal Solution-Grey Relational Analysis (TOPSIS–GRA) method is utilized to construct a fire safety credit evaluation model. Results indicate that the entire sector and the industrial-commercial sector are similar to the light industry sector， with weight correction coefficients of 0.375， 0.291 and 0.334， respectively. This correction effectively addresses statistical biases arising from limited sample data. Thirteen enterprises were selected as demonstration sites for evaluation. The research findings reveal that common indicators exhibit the typical characteristic of "unified structure and hierarchical performance，" while industry-specific indicators significantly influence the fire safety credit levels of light industry enterprises. A multi-dimensional capability analysis of fire safety credit was conducted across different types of enterprises. Four common weak indicators were identified: enterprise safety fund investment， talent reserve， material performance， and material disposal. Additionally， six industry-specific weak indicators were identified: innovation capability， implementation of safety education and training， facility and equipment management， facility and equipment safety performance， working environment， and external environmental risks. Finally， an innovation-driven strategy is proposed to enhance fire safety credit by reshaping the workforce， upgrading means of production， and optimizing labor resources. This strategy provides decision-making support for targeted fire safety credit policies and fire risk prevention and control.
Attenuation characteristics and prediction modeling of explosion pressure waves in nuclear power plant buffer spaces
GAO Chunyan;LIU Zhenyi;Hydrogen explosions and steam explosion accidents represent critical hazards to nuclear power plants due to the generation of intense explosion-induced pressure waves， which may threaten the integrity of safety-related structures and systems. Buffer spaces are commonly incorporated into nuclear facility layouts as passive protective measures to mitigate such pressure waves. However， a quantitative understanding of the relationship between buffer space volume and pressure attenuation capacity remains limited， and existing theoretical models often exhibit notable prediction errors across different spatial scales. In this study， the attenuation characteristics of explosion pressure waves in nuclear power plant buffer spaces are systematically investigated using a combined numerical simulation and theoretical analysis approach. A three-dimensional computational model is established， comprising an incident pressure wave region and a buffer space connected through a rectangular opening that represents typical venting or pressure relief structures. Identical triangular pressure input curves with a peak overpressure of 20 kPa and a duration of 300 ms are applied to buffer spaces with volumes ranging from 1800 m³ to 18000 m³. The spatiotemporal evolution of pressure waves and the variation of internal overpressure levels are analyzed in detail. The results demonstrate that increasing buffer space volume leads to a pronounced reduction in peak overpressure and a significant extension of pressure duration， indicating effective pressure wave mitigation through volumetric expansion and wave modulation. In small buffer spaces， strong interactions between incident waves and multiple reflections result in complex pressure superposition， with the maximum overpressure occurring during the reflection-dominated stage. In contrast， in large buffer spaces， wave reflection and accumulation effects are substantially suppressed， and the maximum overpressure tends to occur during the initial transmission stage of the pressure wave. A theoretical model based on an idealized adiabatic expansion assumption is developed to describe the relationship between minimum overpressure and buffer space volume. While the model agrees well with numerical results for medium-sized buffer spaces， systematic deviations are observed for small and large volumes. By introducing a volume-dependent effective adiabatic index， the theoretical model is corrected， significantly improving prediction accuracy across the full volume range. The proposed prediction model provides a reliable quantitative basis for buffer space volume selection and explosion pressure wave protection design in nuclear power plants.
From risk control to green growth: the pathways through which agricultural insurance empowers the low-carbon transformation of agriculture
Feng Lan;Chang Yunfei;Bao Shuangbao;Under the dual pressures of climate change and China's "dual carbon" goals， the agricultural sector faces an urgent need for a low-carbon transition. Utilizing panel data from 30 provincial-level regions in China， comprising provinces， autonomous regions， and municipalities， from 2009 to 2023 ， this paper employs a two-way fixed effects model as the baseline specification to identify the emission reduction effect of agricultural insurance. To address endogeneity concerns， the first-order lag term of agricultural insurance penetration is used as an instrumental variable， followed by two-stage least squares estimation. A mediation effect model， estimated via a stepwise regression approach， is applied to unravel the transmission channels， with a series of robustness checks conducted to validate the reliability of the findings. The results demonstrate that agricultural insurance significantly reduces agricultural carbon emissions， and this conclusion remains valid after endogeneity is addressed. The mediation analysis reveals two distinct pathways. The first is the capital allocation pathway: agricultural insurance enhances credit availability by acting as a collateral substitute and sending positive signals to financial institutions， thereby alleviating financing constraints on green technology adoption. The second is the factor allocation pathway: agricultural insurance facilitates land transfer and scaling operations by stabilizing income expectations and reducing production uncertainty， promoting farmland reallocation to more efficient operators and curtailing extensive production practices. Both indirect effects are statistically significant. Heterogeneity analysis indicates that the emission reduction effect is significant in the eastern region and non-major grain-producing areas， but becomes positive in the central region and major grain-producing areas， while no significant effect is found in the western region. Overall， these findings suggest that agricultural insurance does not automatically yield environmental benefits; rather， its low-carbon enabling function is contingent upon aligning institutional design with regional contexts. When complemented by credit linkage mechanisms， well-functioning land markets， and supportive green technology policies， agricultural insurance can evolve into an effective institutional vehicle for agricultural green transition. This paper provides empirical evidence for designing regionally differentiated green agricultural insurance products and targeted policy interventions.
Study on differential responses and moderating effects of distracted behaviors on driving risk
He Fei;Hao Chaoyu;Zhang Chunhua;Zhang Le;Liu Wenzhi;Hong Lihao;Yao Mingyuan;This study developed a five-dimensional experimental paradigm based on the "Behavior-Individual-Environment" triadic interaction risk analysis framework. Utilizing a driving simulator platform， we conducted an in-depth experimental investigation into the differences in responses and moderating effects associated with distracted driving behaviors. The experiment involved five driving scenarios: checking navigation， making phone calls， engaging in driver-passenger conversations， WeChat messaging， and normal driving. Data were collected from participants while navigating both familiar and unfamiliar roads， encompassing three physiological indicators (heart rate， heart rate variability， and skin conductance) alongside three categories of driving performance (hard braking， rapid acceleration， and lane deviation). After preprocessing the data in the R programming environment， we performed analyses such as time-series variation and feature aggregation. The findings revealed that distracted driving behaviors significantly triggered sympathetic activation and parasympathetic suppression， with these physiological responses showing a marked dependence on road familiarity. Under familiar conditions， all types of distracted behaviors resulted in degraded driving performance and increased risk. Conversely， unfamiliar road environments induced a persistent defensive driving state in drivers， reflecting an evident risk transfer phenomenon characterized by improved longitudinal control but worsened lateral control. Among the various distractions， two compound behaviors—WeChat messaging and checking navigation—elicited stronger physiological stress responses and had more pronounced negative effects on driving performance. Gender was found to significantly moderate driving risk， highlighting distinct individual differences. Male drivers exhibited smaller fluctuations in physiological indicators and more stable driving operations during distracting tasks compared to female drivers， resulting in a substantially lower overall driving risk. This study constructs a “Behavior-Individual-Environment” triadic interaction risk model， which systematically elucidates the internal mechanisms and risk output pathways associated with this interaction. It further suggests that strategies for preventing and controlling distracted driving should transition from a single-dimensional approach to a dynamic response mode informed by the coordination of driver， vehicle， and road factors.
Research progress on source apportionment of atmospheric particulate matter based on multi-source data fusion
Su Changqing;Ding Yu;Ma Chenchen;Liu Hongyu;Lin Zuhong;Source apportionment of atmospheric Particulate Matter (PM) is increasingly hindered by limitations in spatiotemporal resolution and uncertainty characterization， a challenge that is becoming more pronounced as air quality management shifts towards precision-oriented and regionally coordinated control. This review examines the progress and methodological landscape of multi-source data fusion in PM source apportionment， based on 4，642 publications retrieved from the Web of Science and the China National Knowledge Infrastructure (CNKI) for the period 2000–2025. The study combines bibliometric analysis with semantic mining using Large Language Models (LLMs). Keyword co-occurrence networks were constructed with VOSviewer employing a minimum occurrence threshold of 20， while Python scripts were used to extract titles， abstracts， and author keywords from highly cited publications， forming the input corpus for subsequent semantic analysis. The LLM GPT-4o-mini was prompted to distill core research questions， technical pathways， and application scenarios within each keyword cluster， resulting in the identification of nine research subfields. The extracted semantic features were manually compared with objective high-frequency keywords to verify the reliability of the semantic analysis. Three developmental stages of source apportionment research were identified. The period from 2000 to 2009 represented an initial phase with fewer than 20 publications per year. In contrast， the period from 2010 to 2019 experienced rapid expansion， with annual output exceeding 110 publications. Since 2020， the field has entered a stable high-output phase， producing no fewer than 280 publications annually. During this timeframe， the research paradigm has shifted from static and unidirectional source allocation to dynamic apportionment， wherein numerical simulation and data-driven methods operate in tandem. Notably， multi-source data fusion research surged after 2020， with annual output surpassing 250 publications and reaching 618 by 2025. The LLM-based semantic analysis indicates that multi-source data fusion and source apportionment are converging， particularly in complex nonlinear modeling and environmental decision support. This convergence is organized around a three-tier architecture of data， feature， and decision fusion. Within this framework， research is advancing towards continuous spatiotemporal fields， extended source attribution for health-risk assessment， and explicit quantification of uncertainty， bolstered by data assimilation and physics-informed deep learning. Overall， the integration of multi-source data fusion with source apportionment provides a methodological foundation for regional joint prevention and precise pollution control. The review also identifies significant gaps that remain， including model interpretability， control of error propagation across scales， and the translation of validated methodologies into operational environmental management.