In order to prevent safety accidents caused by unsafe behaviors of underground personnel and to ensure their safety, by utilizing advanced machine vision and computer technologies, the traditional YOLOv5s algorithm and OpenPose algorithm target detection models were improved, and a dual-model coupled algorithm for identifying unsafe behaviors of underground personnel was proposed. Through statistical analysis of the most common unsafe behaviors in current underground coal mines, the unsafe behaviors of miners were classified, including item-related, action-related, and area-related unsafe behaviors. According to the characteristics of miners' unsafe behaviors, the improved YOLOv5s algorithm and the OpenPose algorithm were coupled for recognition, and training and verification were conducted on public datasets and self-built datasets. The results show that compared with the current mainstream methods, the dual-model coupled recognition method has a significant improvement in recognition accuracy on self-built datasets and public datasets, with an increase of 5% to 10%, and can quickly and effectively identify unsafe behaviors of underground personnel.

In order to reduce traffic safety incidents, the mechanism by which negative emotions affect drivers' CL was investigated, and the influence of driving scenarios on emotional intensity and multi-channel CL levels was analyzed. A simulated driving experiment was designed, integrating emotion-inducing materials with a driving simulator, and young drivers were recruited to complete driving tasks. The interaction characteristics between negative emotions and CL were systematically examined through the Self-Assessment Manikin (SAM) scale, the VACP multidimensional assessment model, and retrospective interviews. The results indicate that negative emotions significantly increase CL. Specifically, anger and fear tend to trigger transient fluctuations in load, whereas anxiety is associated with the highest average load level. Furthermore, stressors embedded in driving scenarios induce negative emotions of varying intensities: aggressive cut-ins and sudden lane changes commonly elicited high levels of anger. Traffic accidents and pedestrians running red lights predominantly evoke intense fear, and unfamiliar routes primarily trigger heightened anxiety. Finally, emotional intensity is significantly positively correlated with CL level, highly arousing emotions lead to a substantial increase in the demand for visual, cognitive, and motor resources.

In order to address the managerial complexities arising from the multi-causal and dynamic nature of unsafe behaviors in high-risk industries such as mining and construction, and to overcome the "one-size-fits-all" limitations of traditional behavior-based safety management in the era of informatization and intelligence, this paper introduces the concept of behavioral risk profiling based on the developmental trajectory of behavioral safety and risk profiling theory. By employing literature review methodology and integrating interdisciplinary perspectives, this study systematically examines the constituent elements and construction methods of behavioral risk profiling. The findings reveal that behavioral risk profiling effectively addresses the growing limitations of traditional behavior-based safety management, providing a powerful analytical and managerial tool for the new era. Through a four-layer labeling system comprising work scenarios, behavioral manifestations, risk factors, and risk assessment, it systematically reveals the mapping relationships of unsafe behaviors from phenomena to their underlying essence. On this basis, a construction methodology is developed following the logic of hierarchical data collection, risk-based group assessment, and personalized intervention.

To improve the safety of aircraft landing phase, this study investigates the prediction of pilot human error during the carrier landing phase complex task environments based on CREAM. Based on the pilot operational workflows during the carrier landing phase, this work recalibrates the original Common Performance Condition (CPC) factors in CREAM, and proposes an improved approach for predicting the probability of human error in pilots based on the improved CREAM method, describing the situational environment in which the pilot is located during the carrier landing phase. It introduces an Environmental Impact Index and Effect Impact Index to characterize the influence of various task environments on pilots' cognitive functions. A simulated carrier landing assessment experiment is designed and conducted based on the actual situational environment during the carrier landing phase. By analyzing characteristic indicators such as electroencephalogram (EEG), eye-tracking, electrocardiogram (ECG), and electromyography (EMG) data, and National Aeronautics and Space Administration Task Load Index(NASA-TLX) scale data of the subjects, the load level of the subjects in different environments is assessed, and then the effect impact index of different cognitive functions is calculated to predict the probability of human error. The results demonstrate that the proposed method for predicting probability of human error in pilots can improve the accuracy of prediction results by using objective data assessment, which instead of experts' subjective evaluations. This method can predict the probability of human error in pilots under the influence of 20 different environmental factors.

The situation of road traffic safety of non-motorized vehicles remains severe, and extensive research has been conducted in China and abroad. To comprehensively analyze the research status and prospect, 448 Chinese papers and 860 English papers from 2000 to 2025 were selected from China National Knowledge Infrastructure (CNKI) and Web of Science (WOS). Subsequently, the papers were analyzed for keyword co-occurrence and cluster by CiteSpace, and on this basis, the research hotspots and prospect were summarized. The results indicate that current research shows a significant growth trend, and primarily focuses on four areas: risk of riding behaviors, road traffic accidents, mixed traffic flow conflicts, and traffic management strategies. Future studies should expand into diverse scenarios and deepen the application of multimodal data, with methodological support of computer vision technologies and machine learning algorithms to enhance research capabilities.

To address the current issues of low intelligence level in coal mine working faces and insufficient research on the performance monitoring of shaft structures, a digital twin-based performance monitoring method for vertical shafts is proposed. Firstly, a five-dimensional framework for the digital twin of vertical shafts is proposed based on the operational mechanism and performance monitoring requirements of the shafts. Secondly, a digital twin of the shaft is established by combining virtual-real mapping technology with a finite element surrogate model for grid dimensionality reduction. The structural performance of the vertical shaft is predicted online through artificial neural network technology, where the predicted data is the real-time prediction of shaft structure performance data obtained during the shaft operation process using a shaft structure performance prediction model. The prediction model for the structural performance of the vertical shaft adopts the RBF surrogate model, and the Unity3D virtual engine platform is built to integrate the above functions and achieve online prediction of the structure performance of the vertical shaft. The results indicate that during the operation, by simulating 120 sets of stress and strain data under different working conditions, the average coefficient of determination between predicted and simulated values is 0.995 5, indicating a high correlation between the predicted strain and simulated strain, thus verifying the feasibility of the digital twin framework for vertical shafts. This provides an effective reference for the digital improvement of vertical shafts.

In order to address the safety and environmental risks associated with aqueous ammonia leakage accidents in the chemical industry, based on the rich pore structure and physicochemical properties of porous materials, AC and zeolite were selected as the basic raw materials for detergents. The powder detergents AC@AlCl₃ and Zeolite@AlCl₃ were prepared by impregnating with an AlCl₃ solution of varying mass fractions. Using aqueous ammonia as the target contaminant, systematic experiments were conducted involving both individual decontamination and combined powder-liquid decontamination, employing the aforementioned modified powdered decontaminants along with various surfactants. By evaluating indicators such as decontaminant dosage, volume fraction of volatile gas, and the morphology of decontamination products, the effectiveness of individual versus combined decontamination was compared, and the practical performance of the powder-liquid synergistic decontamination strategy was comprehensively assessed. The results demonstrate a significant enhancement in the decontamination capacity of the modified powders, with decontamination efficiencies for aqueous ammonia exceeding 70%. Among the samples, AC@20%AlCl₃, combined with a 3% NaHCO₃ powder solution, forms a more complete gel-like solid during the decontamination process, and the decontamination efficiency reaches 80%. This approach successfully realizes the goals of high efficiency, cost-effectiveness, and safety in aqueous ammonia decontamination.

To address the problems of state-space explosion and unclear causal relationships in the safety analysis of IMA system under the future SPO mode, multi-phase mission profiles for the SPO mode were first defined, and the resident applications and fault criteria for different mission phases were identified. The concept of CEA was introduced, and a quantitative computational model for cascading fault propagation was developed. Methods for evaluating the probability of cascading faults and the cascade risk coefficient were designed to assess the propagation speed, scope, and impact severity of faults within the IMA system. Finally, a comparative analysis was conducted under an SPO scenario to examine how cascading faults affect the IMA system across different mission phases and initial fault conditions. The results show that the proposed model can characterize both the direct and indirect impacts of faults in the IMA system. When primary flight control and integrated navigation are initial fault events, cascading faults are more likely to occur, indicating that targeted backup, monitoring, or reconfiguration measures for key resident applications are required.

To improve the accuracy of airworthiness assessment for aviation lithium batteries, this paper evaluated the feasibility of equivalent substitution of propane mixed gas in the explosion containment test of thermal runaway gases of aviation lithium batteries. Based on airworthiness standards, the explosion characteristics of standard volume fraction propane and in-situ thermal runaway gases from lithium batteries at different states of charge (SOC) were experimentally compared and analyzed. Ternary lithium batteries were employed, and a dedicated experimental platform was constructed to systematically measure the explosion temperature, maximum explosion overpressure, and pressure rise rate of the thermal runaway gases under different SOCs. The results indicate that the explosion limit range of the lithium battery thermal runaway gases significantly widens with increasing SOC, with a maximum explosion overpressure of 0.519 8 MPa and an explosion power index of 1.093 9. For propane within the standard volume fraction range (3.85% to 4.25%), the maximum explosion overpressure reached 0.822 5 MPa, with an explosion power index of 1.501 7, its explosion potential being higher than that of thermal runaway gases under most SOC conditions. Directly adopting the standard propane concentration for explosion containment verification may lead to over-testing, resulting in an excessively high safety margin. Propane can serve as a preliminary assessment medium, but for equivalent substitution, the test concentration needs to be optimized in combination with the actual SOC state.

This article proposes a nuclear environment monitoring method based on ionizing radiation response characteristics in visual images. Through experimental analysis of the response characteristics of ionizing radiation to visual images, including statistical parameters such as mean, variance, skewness, and kurtosis of pixel values, the relationship between feature data and radiation dose rate was quantified, and the high-precision linear correlation of the fitting was verified. A two-dimensional wavelet packet decomposition is used to analyze high-frequency components in video images, and an algorithm is proposed to extract radiation response signals from complex backgrounds, achieving accurate monitoring of gamma ray radiation dose rates in nuclear environments. The experimental results indicate that within the gamma ray radiation dose rate range of 51.61 Gy/h to 479.24 Gy/h, there is a significant linear relationship between the number of response events and the dose rate. The correlation coefficients of the fitted curves are 0.998 9 and 0.999 3, respectively. In terms of pixel response characteristics, the mean and variance significantly increase with increasing dose rate, while skewness and kurtosis show an exponential downward trend. In addition, the experiment verified the influence of setting the pixel value threshold on radiation dose rate measurement. When the pixel value threshold is 130, the linearity of the fitting results is optimal. By using two-dimensional wavelet packet decomposition, the statistical analysis of high-frequency components in the image has further improved the accuracy of dose rate characterization, especially achieving maximum linearity in the diagonal components. This method can efficiently extract radiation response information in complex environments, achieve precise monitoring of gamma ray dose rates in nuclear environments, and provide technical support for emergency response to nuclear leaks and assessment of radioactive areas.