China Safety Science Journal

Research on fundamental theory of macrosecurisafetyology

WU Chao

2025, 35(8): 1-13. doi:10.16265/j.cnki.issn1003-3033.2025.08.1814

Asbtract ( 127 )

HTML ( 0)

PDF (6082KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to construct the fundamental theory of macrosecurisafetyology (which combines the meanings of security & safety science for the large-scale issues), combined methods of literature review, data statistics, logical deduction and model analysis, etc. were used in this investigation. First, the research process and characteristics on the relevant theory of macrosecurisafety (which means the integrated security & safety for the large-scale issues) in the past decade were reviewed. Then, definitions, connotation, and characteristics of macrosecurisafety were given from multi perspectives, and a group of the basic concepts and their systems of macrosecurisafetyology were put forward. The disciplinary group that constitutes macrosecurisafetyology was built, and a research paradigm and principles of macrosecurisafetyology were summarized. Two macrosecurisafety models and their related definitions were built according to two scenarios. The research results clarify the scope and identification criteria of macrosecurisafety, extract 10 sets of core concept groups supporting 10 macrosecurisafetyology subjects, including macrosecurisafety state studies, macrosecurisafety meta principle studies, macrosecurisafety internal cause studies, macrosecurisafety system studies, macrosecurisafety natural disaster studies, macrosecurisafety man-made disaster studies, macrosecurisafety external cause studies, macrosecurisafety prevention and control studies, macrosecurisafety comprehensive studies, and macrosecurisafety marginal studies, providing their definitions and scopes for these macrosecurisafetyology subjects, form a research paradigm and classification for macrosecurisafetyology principles, and obtain a modeling method and two typical models for macrosecurisafetyology. The above results play an important role in consolidating the foundation of macrosecurisafetyology.

Evolution model and mechanism of work safety governance system

ZHAO Yigui, WANG Mengyao, DENG Qinglu, GUO Hailin

2025, 35(8): 14-21. doi:10.16265/j.cnki.issn1003-3033.2025.08.0277

Asbtract ( 82 )

HTML ( 0)

PDF (3151KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to promote the development of theories on the modernization of work safety governance, a four-dimensional model of the work safety governance system—comprising governance actors, governance rules, governance tools, and governance resources—was constructed based on systems theory and governance theory. Using CAS theory, seven fundamental characteristics—aggregation, nonlinearity, flow, diversity, tagging, internal models, and building blocks—were analyzed. Furthermore, by integrating fitness landscape theory and the NK fitness landscape model, an NK landscape model for the evolution of the work safety governance system was developed, revealing 16 state combinations of the system's genetic morphology (A-R-T-S) and their fitness distributions. The results indicate that the system's evolution follows a three-stage mechanism: Initial Stage: Random walk and low-state locking → Development Stage: Adaptive walk and local coordination limitations → Maturity Stage: Adaptive leap and deep coordination. Empirical analysis indicates that the evolution of China's work safety governance system encompasses three periods with corresponding mechanisms: the System Initial Establishment Period (1978—2002), the System Improvement Period (2003—2012), and the System Modernization Advancing Period (2012—2024).

A SEM-based study of influence of miners'Big Five personality traits and emotions on unsafe behaviors

TIAN Shuicheng, NIE Yurong, CHEN Lei, ZHANG Xiaoya

2025, 35(8): 22-32. doi:10.16265/j.cnki.issn1003-3033.2025.08.0723

Asbtract ( 80 )

HTML ( 0)

PDF (2086KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to further prevent coal mine safety accidents triggered by miners'personality traits and psychological factors, this study adopts the Big Five personality model to assess personality traits. From the perspective of emotions, this study explores the predictive role of the Big Five personality and emotions in miners'unsafe behaviors, while also examining the moderating effect of emotions on the relationship between the Big Five personality traits and unsafe behaviors. First, SPSS 26.0 and AMOS 26.0 were used to analyze the reliability and validity of the questionnaire. Then, a SEM was established to verify the hypotheses. On this basis, the hierarchical regression analysis method was employed to test the moderating effect of emotions. The results show that: the three personality traits of extraversion, neuroticism, and openness, and negative emotions positively influence unsafe behaviors. The two personality traits of agreeableness and conscientiousness, and positive emotions negatively affect unsafe behaviors. Positive emotions play a moderating role in the relationship between the openness personality trait and unsafe behaviors. Negative emotions have a moderating effect in the relationships between the four personality traits of neuroticism, openness, extraversion, and agreeableness and unsafe behaviors. Therefore, coal mining enterprises can identify potential safety hazards and develop strategies to intervene in employees'unsafe behaviors based on workers'personality traits and emotional states, thereby enhancing the on-site safety level.

Digital twin system for early warning of regional intrusion behavior of personnel in deep foundation pit projects

GONG Peisong, SHEN Xinyun, HAN Bowen, GUO Shengyu, PENG Guanqing, ZHONG Junhao

2025, 35(8): 33-39. doi:10.16265/j.cnki.issn1003-3033.2025.08.1471

Asbtract ( 54 )

HTML ( 0)

PDF (5782KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To address the dynamic risks in hazardous areas and the lack of timely warning of personnel intrusion within deep foundation pit, a digital twin system integrating BIM and UWB technology was developed. Initially, a virtual scene representing the deep foundation pit was constructed using the OurBIM 3D graphics engine. Second, a method for calibrating hazardous areas was proposed, and UWB technology was utilized to locate construction personnel and hazardous sources in real-time. Third, a real-time warning mechanism for personnel intrusion behavior was established, and a digital twin system was designed for regional intrusion warning of personnel in deep foundation pit, incorporating functionalities such as scene visualization, hazardous area definition, and warning data logging. Finally, the system was applied to a specific deep foundation pit in Wuhan. The results demonstrate that the system is capable of monitoring personnel intrusion behavior in two distinct types of hazardous areas within the deep foundation pit, achieving a comprehensive warning success rate of 97.4%. This system validates the feasibility and effectiveness of digital twin technology in the construction safety domain.

Identification model of unsafe behaviors among operators in machining workshops

HU Xiaofeng, TENG Teng, HU Jinming, WEN Jiajun

2025, 35(8): 40-47. doi:10.16265/j.cnki.issn1003-3033.2025.08.0176

Asbtract ( 64 )

HTML ( 0)

PDF (9227KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To enhance the safety management of operators in machining workshops, an identification model based on YOLOv11 was constructed. The YOLOv11 model was improved by integrating the MetaFormer architecture, Mixed Aggregation Network (MANet) module, and Adaptive Feature Grid Convolution Attention (AFGC Attention) mechanism. A video dataset captured in a real workshop environment was established to validate the identification model. The results show that the improved YOLOv11 model can identify three types of behaviors, namely unattended operation, operating without a face shield, and operating without protective clothing, with F₁scores exceeding 0.93 for all categories. The improved model demonstrates a significant enhancement in identifying small-sized targets, with the F₁ score for identifying glove-wearing behavior increasing from 0.684 to 0.708, and the mAP@0.5 value rising from 0.604 to 0.651. The research findings may provide technical support for the identification and early warning of unsafe behaviors among operators in machining workshops.

Analysis of causes of chemical fire accidents based on SIF and social network

WANG Pei, LI Xiaoting, YANG Rui, ZHENG Lina

2025, 35(8): 48-53. doi:10.16265/j.cnki.issn1003-3033.2025.08.0164

Asbtract ( 53 )

HTML ( 0)

PDF (6088KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To effectively identify and prevent the risks of chemical fire accidents, a comprehensive research framework was proposed, integrating text mining, SIF model hierarchical analysis, and social network analysis based on rich accident investigation reports in the era of big data. Firstly, the key causes of the accidents were systematically extracted through text preprocessing, term frequency-inverse document frequency (TF-IDF) keyword extraction, and latent dirichlet allocation (LDA) topic modeling, combined with 75 representative chemical fire accident investigation reports from 2000 to 2024. Then, based on the SIF model, the extracted causes were classified into the micro, meso, and macro levels. Subsequently, the accident cause network was constructed using social network analysis methods. Core nodes and key influence paths in the accident cause network were identified through frequency statistics, centrality analysis, and key relationship mining. The research results show that the micro-causes accounted for 56.3%, representing the most crucial factors contributing to chemical fire accidents, with individual factors and environmental equipment risks being dominant. At the meso level, insufficient safety supervision has the highest degree centrality. The key cause path for chemical fire accidents is: inadequate organizational management → insufficient safety supervision → insufficient personal preparation → violation of regulations.

Prediction method for process safety management indicator data of refining and chemical enterprises

LIU Yang

2025, 35(8): 54-60. doi:10.16265/j.cnki.issn1003-3033.2025.08.1054

Asbtract ( 60 )

HTML ( 0)

PDF (1754KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To improve the level of process safety management in refining and chemical enterprises and enhance dynamic monitoring and trend warning capabilities for key safety indicators, a multi-model fusion method for process alarm data prediction was proposed. This approach integrated three time series forecasting models: Autoregressive Integrated Moving Average (ARIMA), double exponential smoothing, and particle swarm optimization(PSO)-based support vector regression (SVR). The method effectively modeled and predicted diverse safety indicators by addressing their trend, autocorrelation, and nonlinear characteristics. Initially, outliers in the raw indicator data were processed. Three types of forecasting models were then constructed and their prediction results were computed. The optimal model for trend prediction was automatically selected based on error comparison. Finally, an empirical analysis was conducted using the time-averaged alarm count indicators from a refinery enterprise over one year. The results show that the proposed method dynamically adapts to varying data characteristics, with selected model prediction errors consistently remaining below 0.1, significantly outperforming the existing requirements for alarm magnitude accuracy. This method effectively enhances the accuracy and flexibility of safety indicator prediction in refining and chemical enterprises and enables the timely identification of potential risk indicators.

Analysis of human causative factors in urban underground space construction accidents based on PAR modeling

YANG Yujiang, WANG Yibao, LI Chong

2025, 35(8): 61-69. doi:10.16265/j.cnki.issn1003-3033.2025.08.0999

Asbtract ( 39 )

HTML ( 0)

PDF (4360KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to reduce the occurrence of accidents during construction in urban underground spaces, a combined methodology of content analysis and social network analysis was employed to deconstruct human factors in accidents during urban underground space construction for risk reduction. Firstly, based on PAR accident causation model, a tri-dimensional analytical framework of "Subject-Factor-State" was developed for human-induced accidents in urban underground construction. Secondly, authentic accident cases were utilized to encode and extract causal factors, while causal subjects and states were systematically identified. Finally, a structural model of human-induced causation was constructed through network co-occurrence, centrality analysis, and correlation analysis to reveal inherent causal mechanisms. Key findings reveal: Collapse, poisoning, and suffocation constitute the predominant accident types, with 24 high-frequency human factors identified, including 9 core contributing factors. The structural model demonstrates that decision-making managers bear critical safety responsibilities among causal subjects. The core factor "inadequate safety technical disclosure", functions as a pivotal risk hub significantly influencing the causal network. Causal states exhibit strong interdependencies, particularly highlighting the pronounced interaction between information deficiency and other states.

Research on collision risk of parallel air routes under autonomous aircraft operation

ZHANG Honghai, TANG Sijia, WAN Junqiang

2025, 35(8): 70-83. doi:10.16265/j.cnki.issn1003-3033.2025.08.0813

Asbtract ( 45 )

HTML ( 0)

PDF (9349KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In the scenario of air-ground collaborative autonomous operations, the collision risk of aircraft was assessed to ensure operational safety. The collision risk of parallel routes under autonomous flight was studied based on human factors and the performance of CNS systems. Firstly, an in-depth analysis of the characteristics of autonomous operation scenarios was performed. The Cognitive Reliability and Error Analysis Method(CREAM) model was improved by incorporating cognitive behavior and interactions between pilots and air traffic controllers to assess human reliability in autonomous operations. Subsequently, combined with the CNS performance and human factors, a collision risk model for autonomous operation aircraft was established. The lateral, longitudinal, vertical, and overall collision risks of the aircraft were evaluated separately. The feasibility of this method was validated through a parallel route example, and the impact patterns of human reliability and CNS performance on the collision risk of parallel routes were analyzed. The results indicate that the collision risk values in all three directions and the overall risk remain below the target safety level (5×10^-9). Furthermore, it is demonstrated that improving human cognitive reliability and CNS performance contributes to reducing the spacing between parallel routes, thereby enhancing the safety of autonomous operations.

Multimodal information fusion decision-making strategy for personnel behavior in industrial scene

WANG Haiquan, YU Haowei, YANG Yueyi, XU Xiaobin, BU Xiangzhou, KURKOVA P

2025, 35(8): 84-92. doi:10.16265/j.cnki.issn1003-3033.2025.08.0084

Asbtract ( 54 )

HTML ( 0)

PDF (13698KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to reduce the accidents in industrial scenarios which were caused by workers'unsafe operation behaviors, meanwhile improve the performance of visual-based action recognition methods in industrial scene with poor lighting, limited field of view and occlusions, an improved decision-making strategy based on self-adaptive ER (S-ER) was introduced in this paper. This strategy could integrate video information and inertial measurement unit (IMU) information effectively. It firstly analyzed video information and IMU information with attention mechanism-based multi-task convolutional 3D (M-C3D) model as well as one-dimensional convolutional neural network (1D-CNN) fused with attention mechanism, then ER theory was introduced to achieve decision-level fusion, where the set of evidence weights and reliability under different environmental conditions was optimized through the firefly optimization algorithm for improving the recognition accuracy and robustness of the model. The effectiveness of the proposed algorithm was verified on the public dataset Multimodal Human Action Dataset from University of Texas at Dallas(UTD-MHAD) and the self-built dataset Multimodal Human Action Dataset from Zhongyuan University of Technology(ZUT-MHAD). The results show that the identification results of S-ER for workers'unsafe behaviors in complex industrial scenarios can reach up to 98.53%, which is 17.52% higher than the maximum value of traditional multimodal fusion methods and single-modality recognition methods.

Study on characteristics of Chinese liquor pool fire under action of liquid CO₂

ZHANG Yutao, BIAN Jing, GUO Qiang, BAI Kexin, LI Xing, ZHANG Yuanbo

2025, 35(8): 93-99. doi:10.16265/j.cnki.issn1003-3033.2025.08.1319

Asbtract ( 35 )

HTML ( 0)

PDF (12672KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To explore more efficient fire extinguishing techniques for Chinese liquor warehouses, based on the similarity theory, an experimental bench for liquid CO₂ fire extinguishing with dimensions of 5 000 mm×2 400 mm×2 900 mm was constructed. An in-depth study was conducted on the changes in Chinese liquor quality, temperature distribution, and gas volume fraction during the process of extinguishing Chinese liquor pool fire using liquid CO₂ fire extinguishing systems with different combustion disk diameters. The research results indicate that liquid CO₂ can effectively extinguish Chinese liquor fires. The larger the diameter of the combustion disk, the more significant the change in mass. The larger the diameter of the combustion disk, the higher the temperature that can be reached by the Chinese liquor pool fire, up to 340 ℃. The extinguishing time under three different combustion disc diameters is 32, 36, and 38s respectively, and there is no re-ignition phenomenon. The highest gas volume fraction of CO₂ in the fire scene space can reach 26%. The distribution of CO₂ content at measurement points G₁-G₄ under different combustion disc diameters is as follows: G₄ > G₃ > G₂ > G₁. The distribution of CO content is: G₃ > G₄ > G₁ > G₂. The distribution of O₂ content is G₁ > G₂ > G₃ > G₄.

Fracture development and evolution features of fractured roof in Xindeng mine and zoned support technology

CHEN Ying, BAO Shiji, SHI Mingzhe, YANG Hongtao, ZHANG Zikai

2025, 35(8): 100-109. doi:10.16265/j.cnki.issn1003-3033.2025.08.1713

Asbtract ( 33 )

HTML ( 0)

PDF (9164KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To solve the challenges in supporting fractured roof coal seam roadways, a rock layer detection recorder was used to conduct tests on the coal seam roadways at Xindeng mine. The analysis focused on the evolution characteristics of surrounding rock fractures. The R value for the direct roof rock and the β coefficient for roof fragmentation were introduced as evaluation indicators for roadway surrounding rock stability. The surrounding rock was categorized into four types: easy to support, relatively easy to support, relatively difficult to support, and difficult to support. Specific support parameters were provided for each category. Zoned support technology was applied in the gas drainage chamber of +90 North wing of Xindeng mine, achieving excellent support results. The study shows that the state of the roof surrounding rock is zoned, with the shallow fractured zone, followed by the fractured-fracture mixed zone, the developed fracture zone, and the intact surrounding rock zone. Through fracture analysis, it is found that the width of the roof fracture zone gradually expands to the deep surrounding rock over time, and the number of fractures also increases. The existing support cannot effectively control the continuous deterioration of the surrounding rock integrity.

Experimental study on spontaneous water infiltration and absorption characteristics in different coal grades

YANG Ming, JIANG Shuwei, XU Jing, ZHOU Ziheng, WANG Mingyang, GAO Jianliang

2025, 35(8): 110-117. doi:10.16265/j.cnki.issn1003-3033.2025.08.1322

Asbtract ( 28 )

HTML ( 0)

PDF (8507KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to understand the characteristics of water transport during spontaneous infiltration and absorption of coal of different rank, coal samples from Mengtai Manlailiang mine, Jiulishan mine, Tiandiwangpo mine, and Pingdingshan No.10 mine were taken as the research objects, The T₂ spectrum of water in the coal sample at different times during the imbibition process was measured by using the low-field NMR experimental system, and the variation of water in the coal sample with time and space during the imbibition process was explored based on the change of water quality in the coal sample. The results show that the wettability of low-rank coal is better than that of high-rank coal and medium-rank coal, and the imbibition capacity of low-rank coal is the largest under the same experimental conditions, followed by high-rank coal, and the imbibition capacity of medium-rank coal is the smallest. According to the change of water in the process of coal sample imbibition with time, the coal sample imbibition process can be divided into three stages: the initial stage of imbibition, the middle stage of imbibition and the late stage of imbibition. In the early stage of imbibition, the sensitivity of low-rank coal to hydraulic action is higher than that of high-rank coal and medium-rank coal, and the imbibition velocity of low-rank coal is the largest, followed by medium-rank coal, and the imbibition velocity of high-rank coal is the smallest. In the middle and late stages of imbibition, the imbibition rate of high-rank coal is larger than that of medium-rank coal. In the process of imbibition and sorption, water will be transported from the pores with smaller pore size to the pores with larger pore size due to capillary force in the micropores/mesopores of the same magnitude, and the resistance of water in the lower rank coal is less than that in the higher rank coal.

Development and coupling relationship analysis of downhole real-time identification experimental device for coal-rock drilling parameters under load

FAN Chaojun, SUN Qiwang, LUO Mingkun, ZHANG Jiaji, LAI Xinfeng

2025, 35(8): 118-128. doi:10.16265/j.cnki.issn1003-3033.2025.08.1440

Asbtract ( 29 )

HTML ( 0)

PDF (14039KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

Aiming at the problem of non-homogeneity and difficulty in obtaining the mechanical parameters of surrounding rock in the process of coal mining, the experimental device for determining the parameters of loaded coal rock with drilling was developed independently to realize the real-time perception of strength and structure of rock layer. The effects of circumferential pressure, drilling speed, rotational speed and rotational torque on strength of simulated rock specimens were investigated and analyzed by coupling the gray correlation of collected parameters with drilling. Under loaded conditions, stratified specimens of different combinations of types were drilled in order to restore the actual effect of drilling into coal and rock seams in real environments, and to verify the accuracy of identification of compressive strength of specimens by the following drilling parameters. It was shown that the strength of the specimens was positively correlated with the circumferential pressure, rotating torque and drilling speed. The strength of the specimen remained almost constant when the rotational speed varied. The correlations between rotational torque, circumferential pressure, drilling speed, and rotational speed and compressive strength were 0.996, 0.831, 0.739, and 0.347, respectively. The strength of the specimen was linearly related to the rotational torque. By analyzing the features of change in the rotating torque curve, the strength of the rock formation can be identified and thus the location of the stratum boundary can be located.

Stability analysis of deep tunnel faces under pipe shed support

ZHANG Jiahua, YANG Xiaoli, YANG Yongxiang, LU Yi, ZHANG Daobing

2025, 35(8): 129-138. doi:10.16265/j.cnki.issn1003-3033.2025.08.0974

Asbtract ( 36 )

HTML ( 0)

PDF (4991KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To investigate the stability of tunnel faces under pipe shed support in deep environments characterized by high stress and high-water pressure, a failure model for the deep tunnel face was established by the variational method. High water pressure and surrounding rock stress were incorporated into the mechanical model, and the external power and internal energy dissipation rates were calculated using the upper limit theorem. The analytical solution for the potential fracture surface of the tunnel faces was derived based on the principle of virtual work. The safety factor of the deep tunnel was solved by Matlab software. The influence of diverse parameters on the failure surface curve was analyzed, and the variation characterstics of safety factor under high stress and high-water pressure environment were discussed. Additionally, the effect of pipe shed support was evaluated. The results indicate that when the surrounding rock strength is low, the failure zone is larger and the longitudinal depth is greater. The high-water pressure and the surrounding rock stress have remarkable influence on the stability of tunnel face. Without considering them, the safety of tunnel face will be overestimated, and the relative error of safety factor can reach more than 60%. Pipe shed support can enhance the stability of deep tunnel face, increasing the safety factor by 79%. Meanwhile, high stress and high-water pressure environment will significantly reduce the effectiveness of pipe shed support. These findings provide theoretical guidance for the support design of deep tunnel, high-stress, and high-water pressure environments.

Discussion on main controlling factors of air leakage in shallow coal seam goaf

LIU Bing, ZHENG Xiawen, LUO Zhenyan, TANG Chao, ZHANG Limin

2025, 35(8): 139-147. doi:10.16265/j.cnki.issn1003-3033.2025.08.1049

Asbtract ( 31 )

HTML ( 0)

PDF (6959KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to conduct a comprehensive analysis of the air leakage characteristics in the goaf of shallow coal seams, five factors were selected, including the average length and width of overlying rock fractures, the porosity and thickness of overlying rock fractures, and the pressure difference above and below the well, to establish an index system for the impact of air leakage in the goaf of shallow coal seams. The dominance-based rough set and response surface method were combined to construct an analysis model for the air leakage characteristics and their factor effects in the goaf. Taking a coal mine as an example, empirical analysis was conducted to generate preference class rules for air leakage in the goaf and extract preference class features. Using the 3-factor 3-level response surface method, regression function fitting was performed on the air leakage volume in the goaf of shallow buried coal seams. The results indicate that the porosity of overlying rock layers, thickness of overlying rock layers, and pressure difference above and below the well are the core influencing factors of air leakage in goaf, with a single factor F-value of 96.06-226.82. The F values for the interaction factors of "interaction between overlying rock porosity and overlying rock thickness" and "interaction between overlying rock porosity and wellbore pressure difference" are 62.34 and 24.66, respectively. Compared with a single factor, in the interaction of multiple factors, the interaction between the porosity and thickness of overlying rock layers has a significant impact on air leakage in goaf. Therefore, when conducting analysis and prevention of air leakage in shallow coal seam goaf, it is important to pay attention to the role of individual factor indicators, as well as the interaction of factors that have a significant impact on air leakage.

Social risk assessment of Three Gorges lock scheduling mode based on SNA-BN

LI Rong, LIU Qing, WANG Lei, ZHONG Yue, LAN Yufeng, NAN Hang

2025, 35(8): 148-155. doi:10.16265/j.cnki.issn1003-3033.2025.08.1600

Asbtract ( 24 )

HTML ( 0)

PDF (5195KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To enhance the intelligent level and risk-bearing capacity of three gorges locks, first, SNA was used for stakeholder identification and extraction for Three Gorges lock reservation scheduling. Stakeholder network characteristics were characterized by three centrality indicators: degree centrality, betweenness centrality, and closeness centrality. An assessment index system was construct from four dimensions: legality, rationality, feasibility and controllability. Then, based on the potential coupling relationship between the indicators, a BN was used to construct a social risk assessment model for the Three Gorges lock reservation and scheduling mode to quantify the direction and intensity of the role of each indicator. Finally, key factors affecting social stability were identified through a sensitivity analysis. The results show that the social risk level under the scheduling mode of the Three Gorges lock is low. The order of influence of the four first-level indicators on the overall social risk is legality, controllability, feasibility, and rationality. The compliance with rule revision, approval, and issuance, the vulnerability of negative public opinion, the vulnerability of mass incidents, the success rate of appointment, and the coverage of security management strategy are the key factors affecting the overall social risk.

Key risk identification of flight accidents based on multi-source text mining

TIAN Ze, LUO Fan

2025, 35(8): 156-163. doi:10.16265/j.cnki.issn1003-3033.2025.08.1120

Asbtract ( 51 )

HTML ( 0)

PDF (9441KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to enhance the risk control efficiency of civil aviation flight safety in China and accurately identify the key risk factors that cause flight accidents, multi-source text mining of flight accident risks was conducted using Weibo information, news reports, and aviation accident investigation reports as samples. Flight accident risk factors were identified using the bidirectional encoder representations from transformers for topic modeling(BERTopic). The semantic correlation of risk factors was analyzed using the Word2Vec model and complex network, from which the key risk factors were determined. The bidirectional encoder representations from transformers(BERT) was adopted to mine the personnel risk factors that trigger the most serious negative emotions among the public. The key personnel risk factors in low-altitude airspace were identified through word frequency statistics. The results indicate that personnel risk is the key risk factor leading to flight accidents. Among them, failure to strictly follow operating procedures, bird strikes, and sickness of flight crew are the key risk factors affecting flight safety. The non-strict implementation of operation procedures by flight crew not only arouses the most negative public emotions but also constitutes the key personnel risk factor leading to low-altitude airspace accidents.

Dam anomaly detection model based on improved Prophet-LSTM-PSO

GE Dalong, DING Yong, LI Denghua

2025, 35(8): 164-170. doi:10.16265/j.cnki.issn1003-3033.2025.08.0155

Asbtract ( 25 )

HTML ( 0)

PDF (4576KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to improve the anomaly detection performance of dam monitoring data, a dam abnormal data detection method based on the improved Prophot-long short term memory-particle swarm optimization Prophet-LSTM-PSO was proposed. Firstly, by improving the Prophet method, the trend component features obtained from the decomposition of abnormal data points were clearly visible. Secondly, the decomposed trend, periodic, and residual components were represented in a three-dimensional space, where the original time series data was substituted with the mean distance of the nearest neighbors in this space. Finally, abnormal data points were identified precisely by combining the LSTM network and PSO algorithm to set and optimize anomaly thresholds. The results show that the method proposed in this paper significantly improves detection performance and exhibits high stability compared with traditional methods. Notably, while maintaining a stable recall rate exceeding 95%, both accuracy and precision surpass 95%, thereby validating the effectiveness and practicality of the proposed method.

Trajectory planning for robotic arm transfer of pulverized coal material in coal dust environment based on SEKOA

DONG Xiangjie, SHI Yan, LIN Chunsong, LUO Yi, LI Peihua, SHUI Xiaoye

2025, 35(8): 171-179. doi:10.16265/j.cnki.issn1003-3033.2025.08.1517

Asbtract ( 20 )

HTML ( 0)

PDF (6368KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to address potential safety hazards of dust explosions caused by the dispersion of combustible dust during the intelligent transfer of medium and heavy flammable materials (such as coal) in mines, triggered by the movement of robotic arms and materials, this study takes 6-axis industrial mechanical arm as the subject. Referencing the technical parameters of the arm's joint motion and the wind speed threshold for coal dust dispersion as motion constraints, the "4-5-4-4-5-4" polynomial pose interpolation method and the proposed SEKOA were employed as the motion trajectory planning model for the robotic arm. This model analyzed the nonlinear combinatorial engineering optimization problem for achieving safe and efficient operation of the robotic arm. Under the established motion constraints, the SEKOA algorithm demonstrates higher efficiency compared to other algorithms in safely transferring coal powder materials, achieving the fastest time of approximately 9.5 seconds. The material movement is stable, with no tilting or collision. The motion trajectory planned by the "4-5-4-4-5-4" polynomial interpolation method is smooth and continuous. During the material lowering phase, the peak velocities of the main drive joints 1 and 2 are approximately 0.72 rad/s and 0.47 rad/s, respectively, with peak accelerations around 0.5 rad/s². After decelerating for about 2.5 seconds to 0 rad/s, the robotic arm can gently place the material at the designated position. This approach effectively prevents secondary dispersion of coal dust on the sealing container.

Research on influence of shaft incline angle and ambient pressure on smoke plug-holing

TANG Junlei, LI Yuanzhou

2025, 35(8): 180-187. doi:10.16265/j.cnki.issn1003-3033.2025.08.1341

Asbtract ( 24 )

HTML ( 0)

PDF (9601KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To better design the smoke exhaust system for tunnel shafts in high-altitude areas, numerical simulation methods were used to calculate and analyze the effects of environmental pressure and shaft inclination angle on the smoke plug-holing effect during tunnel fires. The temperature fields under different working conditions were obtained using FDS simulations. The plug-holing height was introduced to determine the degree of plug-holing for each working condition. Combined with the degree of plug-holing, the dimensionless Richardson number (Ri') was redefined, and a theoretical calculation model for the plug-holing height was proposed. The critical Ri' number (Ri_c') for the occurrence of plug-holing was derived. The results show that the plug-holing height increases with the increase of the angle and environmental pressure. The Ri' number decreases with the decrease of environmental pressure and increases with the increase of the shaft inclination angle. Under the same environmental pressure, Ri_c' number decreases with the increase of the shaft inclination angle. The relationship between Ri' number and the Ri_c' number can be used to determine whether a complete plug-holing effect occurs.

ADS-B abnormal data detection model based on WGAN-XGBoost

LI Huaiqian, CHEN Yuhao, FU Yuxiang, SHEN Jiayi

2025, 35(8): 188-195. doi:10.16265/j.cnki.issn1003-3033.2025.08.1566

Asbtract ( 41 )

HTML ( 0)

PDF (2982KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To enhance aviation operation safety, improve airspace management efficiency, and enhance the defense capabilities of system against spoofing and interference, an anomaly data detection model was proposed based on WGAN-XGBoost. Firstly, WGAN was utilized to learn the intrinsic distribution of the preprocessed ADS-B data, generating abnormal data for augmenting and balancing the training dataset. Then, XGBoost algorithm was employed to train the mixed dataset, building the final abnormal classification detector. Finally, the performance comparisons were conducted through experiments with benchmark models such as Naive Bayes, Logistic Regression, and Perceptron. The results show that the performance of XGBoost is superior to that of all comparison models including accuracy, precision, recall, and F₁ score, with accuracy and precision both exceeding 0.999. The total detection time for 243 792 data points is 2.070 2 s, with an average detection time of 0.008 5 ms per data point. It achieves the optimal balance between detection performance and time cost and has been validated by real abnormal events, demonstrating good practicality and applicability.

Fire prediction in urban villages based on improved grey wolf optimized BP network

LYU Shuran, TIAN Jiangxue, DANG Xinyu

2025, 35(8): 196-204. doi:10.16265/j.cnki.issn1003-3033.2025.08.1426

Asbtract ( 35 )

HTML ( 0)

PDF (7515KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to prevent fires in urban villages, IGWO and BP neural network were used to predict the risk of fires in urban villages. By introducing nonlinear convergence factors and mutation operators, the traditional grey wolf optimizer (GWO) was improved to enhance its global search capability, convergence speed, and stability. Furthermore, a fire risk prediction model for urban villages based on IGWO optimized BP neural network (IGWO-BP) was constructed. Taking into account the complexity and specificity of urban village fire risk factors, an indicator system was developed to predict fire risk, and an empirical study was conducted for verification. The results show that IGWO has significantly improved global search ability, convergence speed, and stability compared to traditional GWO, particle swarm optimization (PSO), and the Great Wall construction algorithm (GWCA). The IGWO-BP model can predict fire risk in urban villages by processing fire risk indicators.

Analysis of coupling mechanisms and prevention strategies for community fire risk factors

ZHAO Pengxia, LI Tie, TONG Ruipeng, DAI Baoqian

2025, 35(8): 205-212. doi:10.16265/j.cnki.issn1003-3033.2025.08.0376

Asbtract ( 34 )

HTML ( 0)

PDF (4625KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

Community fires exhibit high incidence rates and interconnected risk factors. To establish an effective fire prevention and control system aligned with community risk profiles, a hybrid SNA/N-K model was constructed to calculate risk coupling values and network centrality metrics. By adjusting the out-degree values of closeness centrality using coupling values, critical risk factors were identified, providing evidence-based decision-making insights for community fire prevention. Key findings reveal that management deficiencies dominate risk coupling dynamics. An increase coupling factors significantly elevates community fire risk levels. The top three critical risk factors are failure to investigate fire hazards as required, inadequate safety awareness and knowledge of residents and absence of regular fire safety education. Based on this research, a "three-dimensional four-stage" community fire safety management model is proposed. From the dimensions of institutional optimization, technological empowerment and behavioral intervention, a comprehensive strategy is formulated to decouple and interrupt risk coupling chains in community fire prevention and control, encompassing the full cycle of prevention, early warning, response, and recovery. This study contributes a theoretical framework and empirical intervention pathways for community fire risk mitigation, offering actionable guidance for policymakers and practitioners.

Fire temperature field prediction in commercial buildings based on FDS

CAO Yanxi, MA Hongyan, WANG Shun

2025, 35(8): 213-218. doi:10.16265/j.cnki.issn1003-3033.2025.08.0125

Asbtract ( 29 )

HTML ( 0)

PDF (3404KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To address the complexity of fire environment and the difficulty in predicting the temperature field in modern commercial buildings, a fire temperature field prediction model was constructed by combining CNN with SVM. Firstly, FDS was used to construct a commercial building fire model, and the sequence data received by the temperature measurement points were obtained. The temperature, position coordinates, and fire duration were used as input parameters to build the dataset. Secondly, the Rime Optimization Algorithm (RIME) was introduced to optimize the number of hidden layer nodes, regularization coefficient, and learning rate in the CNN-SVM, and then the prediction model was established. Finally, experiments were conducted based on the established dataset and prediction model, and the anti-interference ability of the model under different sensor failure rates was discussed. The results show that the model performs optimally in the prediction of the temperature field plane, with an average absolute percentage error of 5.6% and a maximum relative temperature error not exceeding 25%. The anti-interference performance is the best under three working conditions, and the maximum error does not exceed 15% under extreme conditions.

Comprehensive risk evaluation of interchange ramp area traffic based on real vehicle driving data

ZHANG Yue, ZHANG Shi, DUAN Weijian, ZHU Xinglin, XU Jin

2025, 35(8): 219-226. doi:10.16265/j.cnki.issn1003-3033.2025.08.0852

Asbtract ( 30 )

HTML ( 0)

PDF (9690KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to evaluate comprehensive driving risk in interchange ramp areas, a real-vehicle test was conducted on Chongqing interchange groups, focusing on three typical ramps: right-turn directional, left-turn semi-directional, and small-radius loop ramps. Drivers' electrocardiographic data and vehicle operation status were collected by PhysioLAB and Speedbox, respectively, when passing through ramps. Psychological load and vehicle operation risks of the three ramps were analyzed, and an improved entropy-weighted Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS)method was used to construct and evaluate a comprehensive driving risk model. Results show four HRI change patterns: convex curve, continuous increase, continuous decrease and concave curve. Heart Rate Increase (HRI) in right-turn directional ramps first decreases, then increases, in left-turn semi-directional ramps, it first increases, then decreases near diverging/merging points, in small-radius loop ramps, it fluctuates significantly there. Vehicle operation risk is highest in ramp sections, increasing with smaller radii. The level of vehicle operation risk is small-radius loop ramp greater than left-turn semi-directional ramp, greater than right-turn directional ramp. Comprehensive driving risk peaks in ramp sections, widely distributed in split/confluence areas and peaking shortly after diversion points.

Risk assessment of public opinion on extreme rainstorm disasters based on three-way decisions under event system theory

NIE Ruxin, WANG Luo, ZHANG Ziyu

2025, 35(8): 227-235. doi:10.16265/j.cnki.issn1003-3033.2025.08.1750

Asbtract ( 32 )

HTML ( 0)

PDF (2959KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to realize scientific risk assessment of public opinion on extreme rainstorms, this paper proposed a public opinion risk assessment method of extreme rainstorm disasters based on three-way decisions under the event system theory. From the perspective of event system theory, this method determined the life cycle of public opinion and built indicators for evaluating the risk of public opinion on extreme rainstorm disasters by integrating three dimensions: time, space, and strength. Meanwhile, this paper integrated review helpfulness with hesitant intuitionistic fuzzy sets, effectively quantifying public opinion preferences regarding risk assessment indicators of public opinion. On this basis, the paper extended the three-way decision method to classify public opinion risk into five levels, achieving a scientific assessment of public opinion risk towards extreme rainstorm disasters. Using actual cases and comparative analyses, the effectiveness and adaptability of the risk assessment model of public opinion towards extreme rainstorm disasters were verified. The results show that the proposed method not only successfully realizes the scientific selection of public opinion risk assessment indicators with respect to extreme rainstorm disasters and the effective quantification of public opinion preferences, but also alleviates the decision-making errors caused by the information and cognitive limitations of existing assessment methods to a certain extent.

Research on critical path of emergency response to public health emergencies based on Bayesian networks

ZHAO Zebin, CHENG Rui

2025, 35(8): 236-243. doi:10.16265/j.cnki.issn1003-3033.2025.08.1142

Asbtract ( 39 )

HTML ( 0)

PDF (9127KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To enhance the emergency response capacity of public hospitals in dealing with public health emergencies, the Bayesian network approach was adopted in this study. From the perspective of public hospitals, scenario elements, tasks, and capabilities of public health emergencies were extracted and an emergency response network was constructed. The network was centred around the emergency leading group as the core node, the emergency expert group, medical departments and other core departments as well as pharmacy and other supporting departments were collaborated internally. And it linked with local disease control, public security, communities and other administrative agencies externally, thus forming an emergency response structure with internal and external collaboration. Through diagnostic analysis, the priority of response objectives was clarified. The results show that "preventing nosocomial infection incidents" is the primary objective (accounting for 40%). Three key paths are identified: establishing the emergency leading group-establishing the emergency expert group, establishing the emergency leading group - strictly implementing various prevention and control systems, and establishing the emergency leading group - virus testing.

Study on enterprise-led emergency supply reserve system using Bayesian persuasion theory

CHU Hongrui, HUANG Bin

2025, 35(8): 244-252. doi:10.16265/j.cnki.issn1003-3033.2025.08.1376

Asbtract ( 24 )

HTML ( 0)

PDF (4620KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to effectively address the escalating challenges of natural disasters and enhance the efficiency of the enterprise-led emergency supply reserve system, an economic information disclosure mechanism was proposed to boost enterprise participation. Firstly, an enterprise-led emergency supply reserve model was constructed based on an option contract to clarify the intrinsic driving mechanism of disaster occurrence probability on enterprises' reserve decisions. Then, the Stackelberg sequential game framework was further employed to evaluate the effectiveness of government subsidy policies in increasing supply reserve volumes and improving government-enterprise objectives. Furthermore, Bayesian persuasion theory was introduced to design a government information disclosure mechanism, aimed at guiding enterprises' reserve decisions and improving the overall efficiency of relief supply reserves. Finally, a numerical analysis was conducted with flood disasters as a specific application scenario. The research findings indicate that the volume of enterprise-led emergency supply reserves is positively correlated with disaster occurrence probability, reflecting enterprises' sensitivity to disasters. Although government subsidies can increase reserve quantities and reduce government costs, they have limitations in improving enterprises' profits. The information disclosure mechanism designed based on Bayesian persuasion theory outperforms both the baseline model and the subsidy model in terms of reserve quantities, government costs, and enterprises' profits. Through effective disaster information disclosure, the government can enhance the relief efficiency of the enterprise-led emergency supply reserve system.

Comprehensive evaluation of intelligent emergency capability of metal mines on Qinghai-Tibet plateau based on ANP-VIKOR

WANG Guodong, HU Jiayu

2025, 35(8): 253-262. doi:10.16265/j.cnki.issn1003-3033.2025.08.1530

Asbtract ( 27 )

HTML ( 0)

PDF (3592KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

In order to enhance the emergency management capabilities of metal mines on Qinghai-Tibet Plateau, a comprehensive intelligent emergency capability evaluation model was developed based on the application of intelligent technology. By analyzing the region's unique geographical conditions and complex mining operations, a five-dimensional evaluation index system was proposed, encompassing intelligent technology, behavioral safety, resource allocation, institutional mechanisms, and community relations. This study employed ANP method to determine indicator weights and applied the VIKOR multi-criteria decision-making method to systematically assess the intelligent emergency capabilities of three representative metal mines (Mines Y, V, and J) on the Qinghai-Tibet Plateau. The findings indicate that the comprehensive evaluation system of smart emergency capability based on ANP-VIKOR covers 5 dimensions, including "intelligent technology, behavior safety, resource allocation, mechanism, and community relationship", with a total of 20 evaluation indicators. It fully integrates the complex geographical environment, extreme climatic conditions, and mining operation characteristics of the region. It is helpful to make up for the lack of quantitative research on smart emergency capability in plateau areas. In the case, the mine J exhibits the strongest intelligent emergency management capabilities, particularly excelling in intelligent technology and emergency response. Mine Y demonstrates a moderate level of technological application with notable improvements in emergency management over time, while Mine V shows room for improvement in resource allocation and emergency response speed. The significant differences in emergency capabilities reflect the uneven development of intelligent emergency systems in plateau mining areas. The effectiveness and adaptability of the constructed model are verified.

Musculoskeletal fatigue assessment method based on biomechanics

HE Jiehua, WANG Yiyan, MAO Ying, TONG Ruipeng

2025, 35(8): 263-270. doi:10.16265/j.cnki.issn1003-3033.2025.08.0401

Asbtract ( 36 )

HTML ( 0)

PDF (2753KB) ( 0 )

Figures and Tables | References | Related Articles | Metrics

To effectively prevent WMSDs caused by cumulative fatigue, this study proposes a fatigue quantification method that integrates computer vision and biomechanical modeling. The method extracts 3D posture data of workers using the MMPose model and builds individualized musculoskeletal models in OpenSim. It reconstructs complete work cycles and simulates muscle loading. By calculating joint reaction forces and muscle activation levels, fatigue indices for joints and muscles are developed. The method is applied to analyze fatigue progression in the lumbar spine during a typical lifting task. Results show that the L5-S1 joint and associated multifidus muscles exhibit significant fatigue accumulation in a typical bending and lifting task. Some muscles show lateral asymmetry in fatigue, and both deep stabilizing and core muscle groups are involved in posture maintenance and load response. The method can dynamically track fatigue changes during work tasks and identify high-risk areas of fatigue, and can be applicable to fatigue risk warning in dynamic, high-load operation scenarios.

Table of Content