To mitigate the seismic collapse risks of elevated silo structures, this study adopted base isolation devices to enhance their seismic performance. A photovoltaic-grade ethylene-vinyl acetate (EVA) packaging project was selected as the research object. A finite element model incorporating base isolation was developed using IDA, and seismic fragility analysis was conducted. The maximum inter-story drift ratio was selected as the Engineering Demand Parameter (EDP), while Peak Ground Acceleration (PGA) was adopted as the Intensity Measure (IM). The seismic performance and failure probability of the original structure and the base-isolated structure were compared and analyzed. The results demonstrate that under the rare earthquake (PGA = 0.22 g), peak inter-story drift ratios are reduced by 65.8% (X-direction) and 51.07% (Y-direction) through base isolation. Furthermore, the progression of plastic hinges is effectively suppressed, and the soft-story collapse mechanism is eliminated. Under the rare earthquake, the collapse probability of the base-isolated structure is significantly reduced from 3.63% to 0.173%. Compared to the original structure, the base-isolated structure exhibits significantly better capability in resisting collapse risks induced by low-probability seismic events. This research provides a theoretical foundation and practical guidelines for the seismic isolation design of elevated silo structures.

To address the issues of an incomplete indicator system and oversimplified weight determination in current lifting machinery safety evaluation, a safety evaluation indicator system for lifting machinery was constructed based on Man, Machine, Milieu, Management(4M) factors theory and 95 lifting machinery accident investigation reports from 2019 to 2023. The BWM was employed to determine subjective weights, while an CRITIC method was introduced to determine objective weights. By applying game theory combination weighting, the subjective and objective weights were integrated to form composite weights, thereby enabling the weighting and evaluation of the established indicator system. The results reveal that human factors are the primary cause of lifting machinery accidents. During safety inspections of operators, particular attention should be paid to their safety awareness and technical proficiency. In safety inspections of lifting machinery, the effectiveness of safety protection devices and the structural integrity of key components should be emphasized. The model outputs are highly consistent with key safety indicators in the on-site inspection records, confirming the effectiveness of the proposed model in the safety evaluation of lifting machinery.

In order to study the accident risks caused by the simultaneous explosions of multiple storage tanks in chemical industrial parks, a research method for the overpressure damage evolution process of multiple tank explosions was proposed. First, the explosion damage probability of a single storage tank and the explosion damage probability distribution of a single storage tank were calculated. Then, considering the simultaneous explosions of multiple storage tanks, the overpressure damage evolution process of multiple tank explosions was studied. Finally, the regional damage probability distribution of the simultaneous explosions of multiple storage tanks was obtained. The results show that the evolution result is mainly based on the regional damage probability distribution of Tank o₂, and the subsequent regional damage probability distribution decreases gradually. Through the research, the tank explosion sequence set with the maximum damage probability, the corresponding damage probability set and the tank damage probability distribution set are obtained, and the calculation of the regional damage probability distribution of the simultaneous explosions of multiple storage tanks is realized. The complexity of the storage tank explosion process in chemical industrial parks is caused by the differences in the energy provided during the initial explosion and the differences in the overpressure damage evolution process of storage tank explosions.

To enhance operational efficiency and ensure flight safety, a multi-objective flight slot optimization model was developed by comprehensively considering the requirements of airlines, air traffic control authorities, slot coordination authorities, and airports, subject to airport capacity, terminal waypoint capacity, and safety requirements. The model aimed to achieve balanced and efficient slot allocation within airport clusters, and its effectiveness was validated using actual flight data from Beijing Capital International Airport, Beijing Daxing International Airport, and Tianjin Binhai International Airport in the Beijing-Tianjin-Hebei airport cluster. The results show that the optimized slot distribution becomes more balanced, with peak-hour flight numbers reduced by 7%, the average adjustment per airline decreased by 18.2%, and the total adjustment across the airport cluster reduced by 10.4%. At Beijing Capital Airport, capacity utilization efficiency during peak hours is improved, with redundancy increasing from 2.9% to 8.3%. At Beijing Daxing Airport, slot distribution becomes more balanced, with redundancy increasing from 1.6% to 6.5%. At Tianjin Binhai Airport, the optimized flight volume is maintained within capacity limits, improving safety margins, with redundancy rising from -13.9% to 8.3%. The model effectively enables the rational allocation of flight schedules, significantly enhancing operational safety and resource utilization efficiency.

To balance stope stability with mining economy and to inform stope span design in similar engineering projects, a comprehensive investigation was conducted on the stope span and its stability, using the -200 m level of a copper mine as a case study. The Mathews stability chart method was adopted to evaluate the stope stability by calculating the rock mass stability index and hydraulic radius, thereby defining a reasonable range of stable stope spans. Based on the analysis, four stope span scenarios were designed and numerically simulated using finite difference software to study the stope responses. Comparative analyses were performed on roof displacements and the characteristics of plastic zones in the surrounding rock. The results show that as the stope span increases from 12 m to 18 m, vertical displacement of the roof increases gradually, and the volume of the plastic zone fluctuates slightly. However, when the span reaches 20 m, a significant increase in roof displacement and a sudden expansion of the plastic zone which leads to through-going failure is observed, severely compromising roof stability. An optimal stope span of 18 m is recommended for the -200 m level of the copper mine. The stope span design of mining should consider the influence of factors such as orebody occurrence conditions, safety and efficiency.

In order to curb the occurrence of bottled LPG accidents and ensure the safety production, firstly, a multi-agent accident causation model for bottled LPG was established based on 50 typical cases in mainland China from 2013 to 2024. Secondly, seven core accident causal factors were extracted from it as conditional variables. Finally, fsQCA method was applied to explore the configuration paths of accidents. The analysis results indicate that the bottled LPG accidents are not only characterized by distinctive multi-factorial complications, but also have diverse formation mechanisms. In total, there are eight configuration paths of accidents, which are further categorized into six types, namely the multiple risk coupling, coupling of illegal applied behaviors and equipment failure, illegal applied behaviors of bottled LPG, organizational deficiencies, illegal operation of bottled LPG, and equipment failure. The organizational factors, such as deficiencies in safety production management systems, inadequate safety inspections and rectification of hidden dangers, illegal operation of bottled LPG, and ineffective safety supervision and law enforcement play key roles in the emergence and evolution of accidents. The conclusion indicates that three configuration paths with relatively the highest unique coverage, configuration 1a, configuration 5, and configuration 6a, should collectively be regarded as priorities for accident prevention and control.

To address wellbore instability caused by differences in the physical-mechanical properties of coal-rock and gangue seams, this study established a numerical model for borehole collapse in deep horizontal wells with composite coal-rock layers. The model leveraged triaxial compression test data on coal-rock and gangue mechanics, integrating composite material failure criteria with poroelastic coupling equations. This model was used to simulate how coal-gangue configurations affect peripheral collapse pressure in horizontal wells and analyze the influence of interface positioning and dip angles on collapse pressure distribution. The results show that significant differences in mineral composition ratios between coal and gangue create compromised cementation strength at interfaces, where coal samples fracture more readily due to lower strength. Disparities in physical-mechanical properties and weak interfacial bonding cause pronounced pressure discontinuity at contacts when drilling traverses varied composites. While interface position dictates coal-gangue collapse pressure differentials, dip angle variations not only modulate peak collapse pressure magnitudes but also trigger migration of extreme-value locations along the borehole wall.

To achieve high-precision recognition of wall hollowing sound signals and improve multi-category detection accuracy, a multi-feature fusion method for wall hollowing detection based on BO-SVM was proposed. First, the collected knocking sound signals from different wall types were preprocessed by pre-emphasis, framing, and windowing, and both MFCC and MSC were extracted. The two acoustic features were concatenated at the frame level and normalized to construct a fused feature dataset. Then, a BO-SVM classification model was developed, and the kernel function penalty and parameters were optimized using five-fold cross-validation to establish the MFCC+MSC-BO-SVM model. Finally, classification experiments were conducted using hollow and non-hollow from multiple wall types, including cement, coating, marble, and ceramic tile walls. The results show that the fused features outperform single features in terms of accuracy, recall, and F₁-score. The MFCC+MSC-BO-SVM model achieves an overall recognition accuracy of 96.36%, representing improvements of 6.61%, 9.58%, 15.27%, 13.90%, and 5.02% compared with standard SVM, Random Forest, K-Nearest Neighbor, Grid Search-optimized SVM, and Chaos Particle Swarm Optimization SVM respectively. Furthermore, the BO method can obtain the optimal parameter combination with fewer iterations, showing superior convergence and classification stability.

In order to reduce tunnel lining diseases and extend the safe service life, a diagnostic framework for tunnel lining diseases was constructed, and maintenance and reinforcement methods for diseased tunnels were explored from the perspectives of diagnosis and reinforcement. The application status of technologies for diagnosing tunnel diseases and maintaining/reinforcing tunnels was reviewed. The advantages and disadvantages of various detection technologies were systematically identified, and disease diagnosis methods and reinforcement measures were summarized. The results show that non-destructive testing technologies have evolved from single/semi-automatic to integrated/intelligent systems. Acoustic wave, ground-penetrating radar, and laser scanning technologies have different efficiency and accuracy characteristics. Although automatic detection systems integrating photogrammetry and multi-sensor fusion have gradually been applied, they face challenges in the synchronous detection of multiple diseases. Traditional maintenance and reinforcement technologies are difficult to meet the requirements of low cost and high safety. Although new technologies such as Fiber Reinforced Polymer (FRP) are advantageous, they have issues such as high cost and insufficient environmental adaptability.

An adaptive spatio-temporal multi-head graph attention network was proposed for traffic-flow prediction to improve urban traffic safety and alleviate congestion. A three-stage architecture was designed to achieve refined prediction. Firstly, a spatio-temporal information from traffic flow data was encoded by a spatio-temporal encoding and a traffic-flow graph was constructed. Subsequently, the traffic-flow graph is adaptively enhanced at both the flow attribute and graph structure levels. Then, a two-branch multi-head spatial self-attention mechanism was employed to mine intrinsic associations at global and local spatial scales, thereby decoupling complex spatial dependencies. By temporal heterogeneity and a multi-head temporal attention mechanism, the complex dynamic relationships in time-series data were capture from multiple perspectives. A multilayer perceptron (MLP) was incorporated to explore deep connections between present and future traffic flows. The model's effectiveness was validated using real-world datasets. The results demonstrate that the proposed model outperforms existing methods in key metrics for traffic flow prediction, with reductions in Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) of up to 5.12% and 19.11%, respectively. Furthermore, a high degree of agreement was observed between the predicted and true values, particularly in scenarios with sudden traffic fluctuations. The predictions reflect the actual changes effectively. By optimizing spatio-temporal feature extraction capabilities, prediction errors are effectively reduced and the performance of traffic flow prediction is improved, which demonstrates the effectiveness of the model.