To respond and alleviate the sudden large passenger flow of metro lines in time，a subway passenger flow control optimization model was proposed. Firstly，with the goals of minimizing the total waiting time of passengers and maximizing passenger flow through the interval，permitted inbound passenger flow was used as a decision-making variable to propose a RDM considering constraints such as the supply side，demand side，and passenger flow control intensity. Moreover，the volatility of passenger flow demand was analyzed，and a RM was developed by combining robust optimization theory. The volatility of passenger flow demand was analyzed，and an RM was developed combined with robust optimization theory. Secondly，the robust equivalent transformation theory was used to linearize the nonlinear constraints in RM and solved by the Lingo optimization solver. Finally，a metro line was taken as an example for analysis and verification. The results showed that the RDM model using capacity balance coefficients to decide the permissible inbound passenger flow effectively alleviated the pressure of passenger congestion and improved the efficiency of interval transport. When dealing with uncertain passenger demand，robustness coefficients were introduced in the RM model to adjust fluctuations range of passenger flow demand，thereby reducing the aggregation passenger flow risk and improving the reliability of the passenger flow control scheme.

There is a risk of collision in the batch operation of logistics UAVs. To evaluate the collision probability between UAVs in the air，a model was proposed to analyze the collision risk of logistics UAVs in the aerial operation stage. A UAV collision probability model with the double errors of positioning and velocity was proposed based on the conflict zone theory，and the logistics UAV collision probability variations are analyzed from different angles. The key risk factors in small civil UAV accident/incident data were analyzed and identified based on the data from the Federal Aviation Administration of the United States. The UAV safety flight interval model was proposed based on the logistics UAV failure data of 20 thousand flight hours to determine the shortest flight interval under the acceptable safety level. Furthermore，the effects of track angles and wind directions on the collision probability between UAVs under constant flight intervals were investigated. The results showed that the risk for all flight scenarios was acceptable when the shortest safety interval between logistics UAVs exceeded 90.71 m. When the track angle was 30，60，and 90°，the collision probability varied little with the wind direction. Specifically，the collision probability was maintained constantly when the track angle was 90°. The collision probability varied a lot when the angle was 140°，thereby the safe flight of the UAV was sensitive to environmental factors.

In order to ensure the safe operation of bag dedust system and prevent the occurrence of dust explosion accident，a real-time quantitative risk evaluation model of bag dedust system was proposed based on deviation degree-FCE. Firstly，based on the monitoring data of the pressure difference between inlet and outlet of the dust，box temperature and lock-in ash discharge fault signal collected by the industrial Internet of Things sensor，the deviation degree was introduced to characterize the risk status of the monitoring parameters of the bag dedust system. Then，FCE was used to calculate the risk status of the bag dedust system. The analytic hierarchy process(AHP) and variable weight theory were combined to assign weights to the evaluation indicators. Based on the membership function and weighted average principle of membership degree，the risk of the bag dedust system was quantified by deviation degree，and the risk evaluation result for the bag dedust system was obtained. Finally，the monitoring data of a certain type of bag dedust system was used to verify the effectiveness of the model. The findings demonstrate that when the monitoring value of the evaluation indicators of the bag dedust system gradually approaches the alarm threshold，and the number of indicators approaching the alarm threshold increases，the risk level is higher. The evaluation results are correlated with the actual operation situation，thus validating the efficacy of the model.

In order to prevent transformer fire accidents，short-chain fluorocarbons (FS-50) were combined with non-ionic (Alkyl Polyglucoside(APG)-0810)，zwitterionic (Disodium Cocoamphodiacetate(CAD)-40) and anionic (Sodium Dodecyl Sulfate(SDS) )three kinds of foam solutions，APG-0810/FS-50，CAD-40/FS-50 and SDS/FS-50 were prepared. The basic properties，foamability，foam and liquid film stability of the foam solution in the presence of transformer oil were systematically studied. The curves of foam drainage，volume decay，oil volume and film thickness were obtained over time. The results show that oil has a significant influence on foam properties. Under the action of transformer oil，the foaming ability of the foam solutions decreased significantly，and the foaming ability of APG-0810/FS-50，CAD-40/FS-50 and SDS/FS-50 were reduced by 16.9%，18.4% and 18.8% respectively. The volume of drainage in 60 min increased by 13.6%，74.6% and 15.5%，respectively. Foam volume decay increased by 2.5%，9.8% and 17.9%; The stability of liquid film decreased，and the film thickness decreased by 276.5，414.7 and 552.9 nm，respectively，at 100 s. Among the three foam solutions，the mixture of APG-0810/FS-50 has the best inhibition ability to transformer oil，and the stable foam of this system has the best oil resistance to transformer oil. The research results of this paper can provide a theoretical basis for the development and selection of special extinguishing agent for transformer oil fire.

In order to reduce conflicts between heterogeneous traffic flows during overtaking，a non-motorized lane width design method that considers avoidance maneuver in heterogeneous traffic was proposed. This method represents an improvement over the traditional design approach. Firstly，the conflict levels of non-motorized vehicles were analyzed through the deceleration group proportions and pedaling cadence. Secondly，a vehicle-to-vehicle force model was established to calculate the forces between different types of vehicles. The additional safety gap was determined based on the relationship between these forces and lateral width. The recommended non-motorized lane width was the sum of the traditional lane width and the additional safety gap. Then，the rationality of the proposed design method was evaluated based on the safety levels of non-motorized lanes with different widths，which was obtained from the safety evaluation model developed in this research. Finally，data from four non-motorized lanes in Xi'an city were analyzed as case studies. The research results show that the deceleration group proportion is 0.93 times，pedaling cadence increases by 0.07 revolutions per second，and the road safety value is 1.07 times compared to traditional non-motorized lanes in non-motorized lanes accommodating heterogeneous traffic. These findings demonstrate that lanes meeting the design width have lower conflicts and greater safety.

To improve the applicability and scalability of the joint reserve model for social emergency supplies，and reduce the reserve risks of the government and manufacturers，raw material suppliers were used to propose a government-led Stackelberg three-tier joint reserve model of government and enterprises under a quantity flexibility contract. In this model，the government and manufacturer engaged in joint a stockpile of physical emergency supplies，while the suppliers were responsible for stockpiling raw materials for the production of emergency supplies. The inverse-order solution method in game theory was used to analyze the optimal reserve quantity of the three parties to balance the expected profit and cost. Moreover，the advantages of the proposed three-tier supply chain model involving government and enterprises were evaluated. Finally，the model performance was validated by numerical simulation and sensitivity analysis. The results demonstrated that the three-tier joint government-enterprise stockpile model not only improved the overall emergency stockpile volume under specific market conditions but also reduced the government's storage risk. Furthermore，it can enhance the profitability of high-performing enterprises and reinforce the structure and efficacy of the emergency stockpile system.

In order to investigate the expansion law of high flow grouting material and the creep mechanical properties and pore structure of coal after grouting，a new nonlinear viscoplastic element was created and a creep mechanical model was constructed. Firstly，the high flow grouting material was prepared and injected into the coal body to simulate the coal body grouting around the borehole. Secondly，mercury injection experiment and uniaxial fractional loading creep test were used to test the pore structure of the coal after grouting and the creep performance of the coal under different stress levels，and the strain variation law of the coal at different creep stages was obtained. Finally，based on the creep test results，a new nonlinear viscoplastic element was constructed and introduced into the Boedin-Thomson model. Finally，the relevant parameters of the creep model were obtained by inversion，the reliability of the model was verified，and the creep instability failure law of the coal after grouting was obtained. The results show that the expansion rate of high flow grouting material is 92% higher than that of ordinary cement material，which effectively improves the shrinkage problem of cement-based grouting material after curing. After the high flow grouting material is injected into the coal body，the critical load of instability failure of the coal body is 15.3 MPa，which improves the stability of the coal body to some extent.

In order to address the "fragmentation" issue in emergency collaboration networks and enhance the effectiveness of emergency management in response to accident disasters，three typical accident disasters were selected. Based on theory of collaborative governance，a "network density-average path distance" analytical framework was constructed using SNA to map the collaborative relationships among various emergency response entities. At the macro level，indicators such as network size，density，shortest path，cohesion，and centralization were employed to portray the overall characteristics of the emergency collaboration networks. At the meso level，cohesive subgroups were analyzed to explore the clustering patterns of emergency collaboration，with the criterion for subgroup division being the professional nature of emergency matters in all three cases. At the micro level，the characteristics of nodes within the networks were evaluated through the triple indices of degree centrality，closeness centrality，and betweenness centrality. The results indicate that emergency collaboration networks of Changsha self-built building collapse，Xiangshui explosion，and Xinjia hotel collapse accidents exhibited tight-collaboration，tight-centralization，and loose-collaboration structures，respectively. It is recommended that the emergency collaboration networks be optimized from three dimensions: clarifying the rights and responsibilities of emergency response stakeholders，adjusting heterogeneous interest demands，strengthening the leadership of core nodes and leveraging their pivotal roles，and promoting the deep embedding of emergency culture to empower endogenous collaborative governance.

In order to prevent hard landing overrun events of civil aircraft，first，data including kinematics，system performance and other engineering parameters was collected from QAR. Then QAR data processing activities such as the airport segment clustering，sample balancing and statistical feature extraction were carried out. Subsequently，LightGBM model was used to predict the hard landing events of civil aircraft，and compared with extreme gradient boosting (XGBoost)，decision tree (DT) and long short-term memory (LSTM) models. Finally，the shapley additive explanation (SHAP) algorithm was employed to identify the causal mechanisms of hard landing events and to analyze the impact of various flight parameters on the model's prediction results. The result demonstrates that the proposed model not only exhibits high accuracy and precision in predicting hard landing events (accuracy，correctness and recall reaching 99%，92% and 88%，respectively)，but also provides quantitative and visual explanation information for the decision-making process of hard landing prediction for specific flight segments.

In order to explicit the distribution characteristics of polluted gases in factory buildings and realize efficient control of industrial building environments，the study analyzed the differences and simplification conditions between the distribution characteristics of multiple gaseous pollutants and single gaseous pollutants，clarified the transportation characteristics of multi-component mixed gases emitted simultaneously from the same source，revealed the feasibility range and principles to be followed for mass transfer simplification of multi-component mixed gases in full scale，and verified the validity of the conclusion through numerical simulation. The results show that the applicable range of the multi-component gas mass transfer simplification model is as follows: when the mixed gas emitted from the same source has no phase change and chemical reaction，and its flow parameter Reynolds number Re is above 10³ and Mach number Ma < 0.3，and the physical property parameters are within the range of Knudsen number Kn < 0.01 and temperature between 0.3 Characteristic vibrational temperature θ_v and T'，the components can be simplified for studying the distribution characteristics of pollutants. The simplified deviation stems from the difference in volume force per unit mass of the mixed gas before and after simplification (ΔF). When simplifying，it is necessary to ensure that the ratio of ΔF to the main driving force is less than 0.01. The theoretical model's application range and simplification principle agree with the simulation results. It is proved that a variety of gaseous pollutants can be simplified into a single type under the condition that the simplification principle is met within the applicable scope to explore the distribution characteristics of pollutants.