To have a deep understanding of the causal relationship between crane drivers' situation awareness information and behavior，the situation awareness and behavior response model of crane drivers was proposed by combining the ENDSLEY situation awareness model with the DEMATEL-AISM method. Firstly，the situation awareness theory was used to analyze the driver's behavioral response process and obtain information factors during the crane operation task. Secondly，the DEMATEL method was used to quantitatively analyze the association between the factors and determine a comprehensive influence matrix. Moreover，the attributes and characteristic values of the factors were analyzed to identify the key factors. Finally，a stable hierarchical structure of cause-effect attributes obtained by the AISM was used to propose the situation awareness and behavior response model of crane drivers. The results revealed that a five-layer information model consisted of 22 elements and influence relationships such as key elements of trajectory prediction and planning，and collision avoidance. Furthermore，the proposed model clarified the attribute characteristics，influence relationships，and influence degree among the information elements，offering a deep understanding of the crane driver's situation awareness and behavior response.

In order to address the severe consequences of lithium battery vehicle fires during ship transportation，a vehicle fire simulation model of a ro-ro ship deck lithium battery was constructed. Firstly，the accuracy of the fire simulation model of single and dual lithium battery vehicles was verified by using the full-scale fire experiment data of lithium battery vehicles. Then，the verified simulation model was applied to simulate the fire of a lithium battery vehicle on the deck of a ro-ro ship，and the variation rules of fire heat release rate，temperature and radiant heat under different vehicle spacing were analyzed. The results show that in the initial stage of the lithium battery vehicle fire，the flame first spreads to the side-parked vehicles，especially the side of the vehicle affected by the jet fire，and then expands to the front and rear vehicles. The front vehicle suffers significantly more fire damage than the rear vehicle. At the same time，jet fire aggravates the spread and development of lithium battery vehicle fires. The temperature of the area with jet fire is significantly higher than that of the area without jet fire，and the adjacent vehicles affected by jet fire are ignited first. The numerical simulation method can effectively simulate the fire development and spreading process of the vehicle carrying lithium battery on a ship.

In order to explore the damage characteristics of coal under different ultrasonic excitation time，the ultrasonic excitation test system，nuclear magnetic resonance imaging system，stereo microscope and automatic permeability test system of coal core were used to analyze the changes of pore structure，surface crack and permeability characteristics of coal under different ultrasonic excitation time. The results show that with the increase of ultrasonic excitation time，the change rates of T2 spectral peak area，total porosity，surface fracture area and permeability of small，medium and large pores in coal are linearly increased. When the coal body is excited by ultrasonic for 30 to 150 min，the change rate of T2 peak area of micro，medium and large pores increase from 14.60%，52.50% and 24.90% to 61.30%，145.50% and 235.70%，respectively. The total porosity change rate，surface crack area change rate and permeability change rate increased from 5.04 %，47.27 % and 41.67 % to 24.93 %，127.91 % and 208.33 %，respectively. Under the excitation of different ultrasonic time，the change rate of permeability of the coal and the change rate of total porosity and the change rate of surface cracking area are in line with the linear increasing relationship，with the increase of ultrasonic excitation time，the pore and crack modification effect of coal is enhanced，and the permeability can be improved.

In order to identify the different dimensions of social justice in the risk communication of public crisis events for the promotion of sustainable social development，the theoretical mechanism between the four dimensions of social justice: procedural，interactive，distributive，and corrective justice and sustainable social，ecological，spatial，and temporal characteristics were explained by constructing a theoretical framework of risk justice. Then，we defined risk justice as the fair and reasonable quality of stakeholders' right to participate in the governance and management of public crisis events. To validate the constructed risk justice frameworks，we conducted a detailed content analysis of the risk management guideline，the Sendai Framework for Disaster Risk Reduction 2015-2030，which was used to elucidate the prospective potential of the risk justice framework. The results suggest that this risk management guideline emphasizes the social and spatial issues of distributive and procedural justice，and less consideration in interactivity，corrective justice，and temporal and ecological issues，which may lead to the negative impacts of public crisis events risk communication on the sustainable development of society.

To accurately identify the entities of hydropower engineering construction safety specification，the named entity recognition model of hydropower engineering construction safety specification was constructed. The rich semantic information in the text was mined by the BERT. The semantic features of the specification were extracted by using BILSTM. The dependency relationship between entities was analyzed by relying on CRFs. The Technical Specification for Safety Protection in Construction of Water Conservancy and Hydropower Projects (SL714-2015) was taken as an example to calculate the named entity recognition model accuracy rate. The results show that the accuracy rate of the BERT-BILSTM-CRF model is 94.21%. Compared with the three traditional methods，the accuracy is significantly improved. The research will effectively assist in the intelligent management of safety regulations knowledge for hydropower engineering construction，and provide important support for the intelligent identification of construction safety hazards.

To optimize operator fatigue risk management and reduce the impact of fatigue on operator performance，the monitoring tasks before and after fatigue induction were conducted. Firstly，the fatigue-inducing task of the 2-back paradigm and the monitoring task of the oddball paradigm were designed with the digital main control room of a pressurized-water reactor nuclear power plant，and the control interfaces of low，medium，and high complexity were used. Then，human trials with 23 participants were performed to obtain subjective fatigue ratings，behavioral data，and EEG signals. Moreover，the effectiveness of fatigue induction was verified using relevant fatigue indicators. Finally，the participator's behavioral data and event-related potential P3 components under three different complexity interfaces before and after fatigue induction were analyzed. The results showed that the shortest 2-back task lasting 30 to 60 min induced fatigue. Fatigue or interface complexity increase resulted in a decrease in the monitoring behavior performance，and the maximum performance difference was observed between high-complexity and medium-complexity interfaces under fatigue conditions. Moreover，a 4.9% decrease in accuracy and a 10.4% increase in reaction time were observed. The trend of P3 latency was consistent with that of reaction time，and P3 amplitude increased significantly only under high-complexity interfaces. Based on the performance analysis and event-related potential data，it was concluded that the interface complexity increased the negative impact of fatigue on operators' mental workload.

To prevent occupational injuries of national mine rescuers，the epidemiological characteristics and risk factors of occupational injuries were investigated. Firstly，the injuries of the rescuers of the National Mine Emergency Rescue Team in Datong were investigated through field study，expert interviews，and questionnaire surveys. Then，a national mine rescuer injury causal model was proposed to determine the occupational injury risk factors. Finally，all risk factors were used in binary Logistic regression equation to determine the key risk factors. The results indicated that occupational injuries occurrence of the national mine rescuers was high (54.9%). Furthermore，training injuries (80.3%)，pulling and picking force devices (29.1%)，and weighted running (15.2%) were the most common training items for injuries. The waist (37.2%)，knee joints (11.2%)，hands (9.0%)，and ankle joints (7.6%) were the most vulnerable body segments. Work-related injuries (28.7%)，sprains (25.1%)，and strains (22.4%) were the most common injury types.The major internal risk factors affecting occupational injuries of national mine rescuers were training years，and the external factors were drinking alcohol，weekly training frequency，training with injuries，and lack of sleep. The major inducing events for injuries were insufficiency warm-up，recurrence of old injuries，and incorrect technical movements.

In order to reduce the pedestrian safety problems caused by the lag of the emergency warning system in urban rail transit stations under large passenger flow conditions，the YOLOv5 algorithm was selected to predict passenger flow information. The artificial neural network (ANN) model was used to construct the urban rail transit emergency warning perception system. Firstly，the YOLOv5 algorithm was improved by optimizing the model training hyperparameters and prior frame parameters. Then，the emergency warning perception system was designed by selecting warning indicators，weight analysis and threshold definition. Finally，the self-organizing competitive network emergency warning model based on ANN was constructed by using Matlab software. The data collected by the optimized YOLOv5 algorithm were substituted into the emergency warning perception system through calculation，and the emergency warning perception system was verified by experiments. The results show that the optimized YOLOv5 algorithm can improve the accuracy of pedestrian target monitoring under large passenger flow conditions of urban rail transit by 7.04%. The judgment results obtained by substituting the pedestrian data collected by the optimized YOLOv5 algorithm into the constructed emergency warning perception system are consistent with the actual warning level，which proves the feasibility and effectiveness of the system and helps to improve the emergency warning level of urban rail transit.

In order to improve the safety analysis of old buildings，a method was introduced for quantitatively predicting their structural safety based on digital twins. Identifying risk factors in the buildings' structures，setting up an intelligent monitoring system，and using monitoring data in numerical simulations to assess the safety of critical structural elements were included. The method provided quantitative predictions for damaged areas and the extent of destruction in the buildings. This study validates the proposed method using an old building in Xi'an with a history of over 70 years. A two-year structural safety monitoring was conducted following an actual engineering design monitoring plan. The numerical model of the building was created using Revit software and imported into ABAQUS. Monitoring data was used as external parameters in numerical simulations，with the assumption that monitored values exceeding settlement thresholds indicate a quantitative prediction of building damage. The outcome demonstrates that the old building is currently in a safe condition. However，if the settlement of the building exceeds the safety threshold，the likelihood of structural damage occurring at the front side of the stairs will be increased. The study verifies the high feasibility of a quantitative prediction method that integrates real-time monitoring data with numerical simulations based on digital twin technology for analyzing and assessing the overall and local structural safety of aging buildings.

In order to solve the problem of travelling obstacle detection in the context of complex open pit mines，a mining road obstacle detection algorithm based on improved cross-scale feature fusion is proposed. Firstly，to address the problem of unbalanced small target sample categories in the original mine dataset，a data enhancement method based on geometric transformation and weighted Poisson fusion is used to expand the number of samples.Secondly，a cross-stage connectivity network that is more suitable for obstacle detection is proposed in the feature extraction stage to increase the detection scale and improve the algorithm's learning ability of the small target features，and then a 3D parameterless attention (SimAM) and de-weighted Bi-directional feature fusion pyramid network (Bi-FPN) are proposed in the feature fusion stage to improve the multi-scale detection performance by enlarging the predicted feature map and feature receptive field. Finally，to address the problems of sample imbalance and imprecise obstacle bounding box localisation in the training，the quality focal loss function (QFL) and the scalable Intersection and combination ratio loss function (SIoU)，which combines the classification score with the quality prediction of the position to improve the localisation accuracy for dense occlusion targets. The results show that the improved method can effectively identify unstructured road obstacles in open pit mining area under complex background，and in practical application，the detection accuracy reaches 91.88% and the detection speed reaches 68.7 f/s，which has a better performance of small-target and multi-scale detection compared with the mainstream detection methods，and it can satisfy the requirements of obstacle safety detection in the travelling of unmanned mine cards in open pit mining area.