In order to solve the heat damage issue brought by deep mining in coal mines，based on ANSYS simulation software，the roadway heat insulation model was constructed. The influence of basic parameters of different heat insulation layers and laying methods on the roadway wall structure was analyzed，and the cooling effect of the heat insulation roadway was further investigated. The results of the study show that the laying of the heat insulation layer could reduce the radius of the peripheral rock heat-regulating circle and slow down the heat dissipation of the peripheral rock，and the three ways of laying the heat insulation layer have little effect on the temperature field of the peripheral rock. Compared with the temperature of the non-insulated support layer，the average temperature of the support layer is reduced significantly by laying of the heat insulation layer in the rock-adhering type，whereas the opposite is true for wall-adhering type and the sandwich type. The increase in the thickness of the heat insulation layer and the decrease in the heat conduction coefficient of heat-insulating material could increase the heat-insulating effect，and 7cm is the optimal thickness of heat insulation layer. Excessive inlet wind speed or air temperature will affect the cooling effect of heat-insulated roadway. Heat-insulated roadway is suitable for high temperature roadway with low wind speed，low inlet air temperature and long ventilation distance.

To improve the efficiency of urban sewage pipeline defect detection，reduce resource wastage resulting from indiscriminate inspection methods，and mitigate environmental safety risks，the XGBoost model was used to predict the probability of urban sewage pipeline defects. Firstly，the causes of sewage pipe defects were statistically analyzed to determine key indicators that can characterize the pipeline defects as the inputs of the XGBoost model. Secondly，appropriate objective functions and base learner parameters were selected. Then the model training and optimization were performed by a grid search algorithm to determine the key parameters of the base learner. Finally，the XGBoost model prediction performance was validated against an area of the sewage pipeline network in Zhongshan，Guangdong province. Moreover，the main factors and paths affecting defect probability were investigated based on the model output，and the defect probability of the sewage pipe network in the area was divided into 4 different levels for visualization.The results indicated that the average area under curve (AUC) of the XGBoost model was 0.97 under 10-fold cross-validation with a prediction accuracy of 93%. Pipeline depth，slope，and length had the greatest impact on the probability of pipeline defect. As the pipe length increases，the sewage pipe defect probability will increase if the slope becomes greater and the buried depth becomes shallower.

Traffic conflicts caused by illegal riding of e-bikes are a great challenge and negative impact on the safety management and operation efficiency of signalized intersections. In this paper，three indexes including post-encroachment time (PET)，time to collision (TTC) and deceleration-to-safety time (DST) were selected from the two aspects of the number of collision objects and motion state. The k-means clustering was adopted to divide the severity of traffic collisions caused by illegal riding of e-bikes into three categories: general，serious and potential collision. Secondly，the Poisson function was used to fit the distribution characteristics of conflict frequency，random variables were introduced to describe the mixed effects of heterogeneity among traffic conflicts，and a prediction model for traffic conflicts of illegal riding e-bikes based on GLMM was built to predict the frequency of traffic conflicts of multi-grade severity. Combined with the data of 996 e-bike traffic conflicts obtained by video investigation，the empirical study shows that the proportion of e-bike traffic conflicts with different severity has nothing to do with the types of violations. The constructed GLMM model is better than generalized linear model (GLM) in fitting the traffic conflict data of illegal cycling e-bikes，and has the best prediction effect on the common conflict frequency. By strengthening the management of e-bike occupation of motor vehicle lanes and the waiting behavior of crossing the line，adding escort officers and adjusting the signal phase，the incidence rate of e-bike conflict can be reduced.

In order to timely and effectively detect the explosives hidden in luggage，packages and individuals，multiple explosives such as trinitrotoluene(TNT)，hexahydro-1，3，5-trinitro-1，3，5-triazine(RDX)，triacetone triperoxide(TATP)，ammonium nitrate(AN)，etc. were detected by FQ and IMS instruments by wiping and aspiration sampling methods. The comparison was mainly made from two aspects: alarm time and recovery time. The experimental results show that under wiping sampling，the average alarm time of FQ is about 2 seconds less than that of IMS，and the average recovery time is about 30 seconds less than that of IMS，which has higher detection efficiency. In the case of aspirated sampling，FQ instruments can detect TNT and TATP，while it is difficult for IMS instruments to detect explosives.

To improve the evacuation efficiency and ensure the safety of passengers on an open deck in the event of an accident，a heterogeneous evacuation model considering passenger heterogeneity and group effect was constructed using AnyLogic software. Four simulated evacuation scenarios，including individuals，two-person，three-person and mixed groups，were constructed，and the evacuation process of heterogeneous passengers under different group sizes was simulated. The results indicate that with the increase of the group size，the total evacuation time increases nonlinearly. Specifically，the evacuation time for passengers in two-person，three-person and mixed-group scenarios is 11.8%，19.6% and 15.5% longer than that in individual evacuations，respectively. Additionally，the peak of passengers' arrival at the terminal occurs in the early and middle stages of evacuation. In group evacuations，the higher expected passenger speed correlates with more dispersed arrival time distributions. These findings highlight the interaction effect between heterogeneous passengers and group effect in the evacuation process.

To effectively prevent and control the risk of subway waterlogging accidents，a quantitative analysis method integrating the AcciMap model，FAHP，and triangular fuzzy DEMATEL-ISM method was proposed to investigate the causes of urban subway waterlogging accidents and their coupling associations. Accident causation factors were selected from the system level through AcciMap model. FAHP was used to integrate expert opinions and determine expert weights. The DEMATEL method was used to calculate the importance of accident causes，ISM method was used to reveal the relationship between factors，and a multi-level ladder stepwise graph was constructed. The model was validated by waterlogging accidents on the Zhengzhou subway. The results showed that the core contributing factors leading to the incident were inadequate awareness and prevention measures of major disasters by the management department，the lack of diligence in supervision and inspection duties by regulatory authorities，and ineffective response and handling by operational departments. These factors can be paid special attention to prevent system failures.

To improve the efficiency of disaster reconnaissance，a prioritized multi-UAV cooperative scheduling optimization model was proposed for the UAV reconnaissance scheduling problem in disaster emergency responses. An improved tabu search algorithm was proposed by using a mixed integer nonlinear programming model. Moreover，specific neighborhood operators and repair operators were used to enhance the algorithm's search capability and solution quality. Two cases were designed to validate the model and algorithm performance. The first case was based on the actual scenario of the Zhengzhou flood in 2021，and the second case simulated examples of different scales. The results indicated that the model and algorithm were effectively validated. The effects of flight speed，UAV quantity，model，and endurance on reconnaissance capabilities were investigated through sensitivity analyses. The reconnaissance efficiency can be significantly improved by increasing the number of UAV models and optimizing them. UAVs with strong reconnaissance capabilities should be given priority when resources are limited. However，flight speed is more critical when resources are sufficient.

In order to improve the durability of the construction machine and the working environment in deep buried tunnels，based on the theory of gas-solid two-phase flow，CO and dust were selected as the main objects of study，and a physical model of deep buried tunnels was established by Fluent software. The effects of different surrounding rock temperatures and the outlet speed of the wind pipe on the transport process of soot in deep tunnels were investigated through simulation. The results show that after blasting，CO is uniformly distributed in the throwing area. With the increase of ventilation time，the CO transport shows two modes of translation and diffusion. CO is discharged out of the tunnel in the form of a mass，and the CO transport speed at the tunnel wall is larger than that at the center of the tunnel. At the moment of blasting，a large amount of dust gathers near the working face，and with the increase of ventilation time，the dust is continuously discharged out of the tunnel. Among them，the temperature of the surrounding rock has a certain effect on the transportation of CO. The higher the temperature of the surrounding rock，the faster the transportation of CO. However，the effect of the surrounding rock temperature on the transportation of dust is relatively small. The outlet speed of the wind pipe has a greater impact on the transportation of CO and dust. The greater the outlet speed of the wind pipe，the faster the transportation of CO and dust. The field application should be combined with the actual conditions and economic budget to select the relevant equipment.

To address the issue of predicting heat stress risks for police officers engaged in outdoor operations under high-temperature conditions，a test dataset for monitoring core temperature of police officers under different environmental working conditions，levels of physical exertion and clothing scenarios was constructed. First，features such as height，weight，age，gender，body fat percentage，physical activity ratio (PAR)，clothing insulation (CI)，environmental temperature and relative humidity were extracted. Then，machine learning methods，including K-nearest neighbors (KNN)，random forest (RF) and gradient boosting decision trees (GBDT)，were used to establish predictive models of core temperature and heat stress risk for outdoor police officers. These models were subsequently validated. The results indicate that for the predictive model of core temperature for outdoor police officers working in high-temperature environments，the goodness-of-fit measure R² exceeds 0.9 for KNN，RF and GBDT. In terms of error，the KNN model has the smallest prediction error，with a root mean square error (RMSE) of 0.053 ℃. For the heat stress prediction model for police officers engaged in outdoor operations under high-temperature conditions，the predictive performance of RF，GBDT and KNN models is significantly better than that of other models.

To explore the changing characteristics of driving load under the construction environment of underground tunnels，a test platform was built，and a driving simulation test was carried out in the construction environment of underground tunnels to obtain the driver's eye movement and ECG data. Taking the time domain index of heart rate，heart rate variability(HRV) and average number of blinks frequency(BF) as parameters，a comprehensive evaluation model of driving load based on factor analysis and entropy was constructed，and a classification method of driving load based on K-means clustering algorithm was proposed. The results show that in the relatively monotonous environment of non-construction sections in underground tunnels，psychological pressure could more accurately reflect the driving load compared to visual pressure. However，in the complex and changeable environment of construction section，psychological pressure is prone to being influenced by driving operations. Moreover，when line of sight is limited，a single visual pressure indicator tends to overestimate the driving load. The comprehensive evaluation model of driving load based on eye movement and ECG exhibits high sensitivity and good stability，and can correct the evaluation results of a single indicator and effectively quantify driving load.