In order to effectively prevent the occurrence of civil aviation maintenance related events, a sample of 2 687 voluntary reports from the US Aviation Safety Reporting System (ASRS) database (2001-2023) was examined. Fourteen risk factors, including work environment and human factors, were analyzed using CiteSpace. Strong couplings between these factors were identified through UCINET and Gephi. The results show the following findings: Among the nine single risk factors, key risks are identified as landing gear system malfunctions, poor lighting, and rain (with frequencies of 103, 64, and 45, respectively). Among the five coupled risk factors, key risks are found to be airframe maintenance, communication breakdown, and human factors (with co-occurrence frequencies of 1 646, 1 448, and 1 206, respectively). The analysis of all risk factor couplings reveals that the strongest coupling exists between airframe maintenance and powerplant maintenance (with a weight value of 3 568). Additionally, a significant coupling relationship is observed between procedures and improper landing gear system operations (with a weight value of 118).

CBR can effectively address the issue of unclear elements in safety risk management. However, existing research lacks systematic reviews of its application status and development trends. To systematically summarize the current state of research on the application of CBR in construction safety risk management and to identify future research directions, this paper comprehensively reviewed 45 Chinese and 50 English papers, employing bibliometric analysis and content analysis. The results show that the number of publications shows a fluctuating upward trend and tends to stabilize, spanning multiple disciplines such as environmental science and computer science. Key scholars demonstrate both commonalities and differences in collaboration themes, with China ranking first in publication volume. The research themes have evolved from knowledge construction and methodological exploration toward integration and intelligent convergence. CBR has been widely applied in safety risk identification, analysis and assessment, response, and knowledge management, yet limitations remain regarding control targets, application scenarios, knowledge management, data integration, and workflow optimization. Future research should focus on expanding control targets, developing whole-process management scenarios, establishing collaborative knowledge management mechanisms, integrating multi-source heterogeneous data, and improving workflows to advance the development of intelligent safety risk management in construction projects.

In order to enhance the emergency response efficiency of fire rescue zones, a block accessibility-based method for dividing fire rescue territories was proposed. Basic data were obtained through block boundary extraction, as well as the localization of fire stations and demand points. Multi-period real-time traffic data were acquired via an online map to calculate the integrated travel time from each fire station to every block unit, which served as the basis for division. Blocks were assigned to the corresponding fire station based on the shortest accessibility time. Finally, an optimized zoning scheme and an evaluation of the overall regional accessibility were output. A case study was conducted in the HN-FR area, involving 8 fire stations and 10 067 demand points. A total of 542 block units were extracted, and analysis was performed using 618 483 data samples across 29 scenarios. Results indicate that the overall accessibility time for the HN-FR region is 435.96 s, reflecting relatively poor fire rescue accessibility. Furthermore, significant disparities are observed in the areas of the eight fire station territories, ranging from 0.37 to 12.36 km². This variation is primarily attributed to the proximity of some fire stations to the boundary of HN-FR, resulting in minimal coverage capacity.

In order to enhance the intelligent diagnosis level of safety problems in complex construction environments, an intelligent question-answering model for construction safety hazards based on vision-language multimodality was proposed. A dataset of image-text pairs related to construction safety hazards was constructed. A visual encoder was used to complete the visual encoding of safety hazard images, and a language model was employed to encode the question-answering texts about safety hazards. A multimodal feature fusion module was adopted to achieve effective interaction between image and text information. A specific input template for visual question answering adapted to the scenario of construction safety hazards was constructed. The model was fine-tuned based on matrix low-rank decomposition, and multi-round prompts were used to guide the model in generating accurate answers. The results show that compared with existing contrastive models, the intelligent question-answering model for construction safety hazards performs better in automatic evaluation metrics, Generative Pre-trained Transformer(GPT)-4 evaluation, and expert evaluation, with significantly improved fluency and semantic relevance of the generated texts. Ablation experiments further verify the effectiveness of each sub-module, confirming that the synergistic effect of matrix low-rank decomposition fine-tuning and multi-round reasoning is the key for the model to achieve optimal performance, and that reasonably setting the rank parameter of the low-rank matrix can effectively avoid the overfitting problem.

In order to improve the ability of accident prevention and risk control of urban drainage networks, first of all, based on the indicators from the four aspects of "human-material-management-environment", 19 assessment indicators were obtained to construct the risk assessment index system of urban drainage pipe networks, and a risk evaluation index system for urban drainage networks was established accordingly. Subsequently, a model of urban drainage networks was constructed by integrating fuzzy theory and BN. Triangular fuzzy numbers were introduced to quantify the scores of experts. Weights were assigned according to differences in professional titles and working years, and α-weighted valuation method was adopted to transform fuzzy evaluations into clear probabilities. Thus, a risk evaluation model for urban drainage networks was obtained, and forward and reverse reasoning of BN was conducted to calculate the posterior probabilities of key nodes. Finally, taking the drainage network of a certain urban area as an example, the risk level of the drainage network in a specific area of the city was evaluated, and investigations were conducted for verification. The results indicate that the model effectively handles uncertainties and subjectivity in the risk assessment process, achieves probabilistic characterization of drainage network risks, and improves the accuracy of evaluation outcomes. External pipe protection is identified as the most critical factor affecting drainage network safety, followed by anti-corrosion measures and joint methods. The overall safety performance of the urban drainage network is found to be satisfactory, with risks remaining within controllable limits. Comparative analysis with historical monitoring data and fault records confirms the practicality and reliability of the model.

In order to enhance driving safety in highway tunnels and to reduce visual distraction and burden caused by improperly designed visual guiding devices, a research framework for the design and evaluation of tunnel visual guiding devices was proposed. A systematic review of domestic and international literature has been conducted to analyze the effectiveness and design parameters of various types of visual guiding devices, and to clarify their key functions in delineating spatial right-of-way, ensuring adequate sight distance, and enhancing visual comfort. In addition, current studies were reviewed to summarize data acquisition methods and core evaluation indicators, thereby constructing a scientific evaluation system. The limitations of existing research were also identified, and future research directions were proposed. The results show that visual guiding devices can optimize the tunnel visual environment by clarifying spatial right-of-way, increasing effective sight distance and visual fields, and mitigating visual illusions. The type and spacing of devices should be determined based on spatial right-of-way, curvature perception, speed perception, and visual comfort. Real-vehicle experiments and indoor simulations serve as key approaches for exploring design parameters and optimization strategies, while field observations and accident data analysis can further validate their effectiveness. Indicators such as visual perception, visual characteristics, physiological responses, and driving behavior are suggested as core evaluation metrics. Moreover, multi-indicator interaction analysis may help reveal the mechanisms by which visual guiding devices influence driver behavior.

Safety management of long-distance oil and gas pipeline enterprises has been regarded as essential for ensuring energy security and maintaining social stability. As a key approach to enhancing safety management efficiency, audits of QHSE management systems are increasingly emphasized. In this study, traditional audit modes—including international safety rating audits, full-factor quantitative audits, and system certification audits—were systematically compared and analyzed. Additionally, the features and applicable contexts of emerging audit modes, such as data-driven audits and AI-based audits, were further explored. It was found that each audit mode presented distinct advantages and limitations, depending on the enterprise's stage of development, risk management demands, and resource capabilities. Accordingly, a comprehensive audit framework integrating conventional methods and intelligent technologies was constructed. Optimization strategies were proposed, including phased audit mode selection, risk-oriented focus, capacity enhancement, digital and intelligent transformation, and the establishment of a closed-loop rectification mechanism. The findings show that a multi-mode integrated audit system significantly improves the safety performance of long-distance oil and gas pipeline enterprises. The conclusions provide theoretical guidance and practical reference for the selection and optimization of QHSE audit modes in the oil and gas industry.

To study the impact of cognitive bias on energy safety engineering decision-making, the Fukushima nuclear power plant accident was taken as the research object. A full-cycle decision-making analysis chain, covering risk identification (earthquake and tsunami assessment), crisis disposal (cooling system failure treatment), and aftermath management (information disclosure decision-making), was built through retrospective analysis of the accident timeline and key decision points. Cognitive psychology theory was used to analyze decision-making bias phenomena in energy engineering safety management, and the specific mechanisms of these biases in emergency response and risk assessment were revealed. Results show that six typical cognitive biases are present in the emergency decision-making during this accident, including confirmation bias, anchoring effect, representative heuristics, framing effect, loss aversion, and overconfidence. This analysis demonstrates that cognitive bias identification can enhance energy engineering safety management, and improve the effectiveness of safety emergency responses.

To decouple the structural complexity of industrial and supply chains, address systemic risks, and enhance supply chain resilience, bidding transaction big data was employed. Taking the vaccine sector as an example, a framework for constructing a supply chain knowledge graph was designed, and a systematic supply chain knowledge graph was established. On this basis, complex network techniques were applied to examine the vulnerability and potential security risks of China's vaccine industry supply chain network from 2011 to 2023. The research encompassed complex knowledge queries of the industrial chain, an analysis of city degree distribution patterns, and simulations and analyses of supply chain risks. The study shows that the vaccine industry chain exhibits spatial imbalance, particularly between eastern and western regions. The production structure is highly dependent, with approximately 61.3% of vaccine varieties relying on a single manufacturer or overseas agent. Manufacturers with high centrality constitute potential risk points within the vaccine supply chain network, where disruptions to about 33 enterprises significantly hinder vaccine supply. Compared with core cities, the cumulative effects of cities with lower network status, such as Chongqing, Dalian, Shenzhen, and Shenyang, have a more pronounced impact on the efficiency of vaccine circulation.

In order to clarify the research hotspots, development trends and accident risk characteristics of alcohol-based fuels, and point out the direction for the development of alcohol-based fuels, CiteSpace software was used to collect 3 951 relevant journal papers from 1978 to 2024 in China National Knowledge Infrastructure. Bibliometrics and knowledge map visualization analysis methods were used to systematically analyze the annual publication volume and subject distribution in this field, and analyze research hotspots from keyword co-occurrence, clustering and emergence. The results show that the research on alcohol-based fuels in China has experienced three stages of continuous growth, fluctuation adjustment and steady development. At present, driven by the development of national energy strategy and technology, the research attention has been steadily improved. The research involves many fields such as electric power, chemical industry, automobile, power and fuel. The research hotspots are highly concentrated in the four major themes of methanol fuel preparation technology, fuel cell application, power and environmental impact assessment of methanol fuel. Existing research significantly focuses on basic material properties and technology applications, and systematic research on safety risks is relatively scarce. Therefore, this paper further clarifies the bottleneck of current risk prevention and control, and puts forward that future research should focus on fuel intrinsic safety technology, construction of intelligent supervision system of the whole chain and improvement of emergency response capacity, so as to provide theoretical reference and direction for the safe development of alcohol-based fuel.