Traffic accidents occur frequently, causing significant losses to people’s lives and property safety. Accurately predicting the severity level of traffic accidents is of great significance. Based on traffic accident data, this study comprehensively considers various influencing factors such as the geographical location, road conditions, and environment. The data are divided into accident-related categories, weather-related categories, and road- and environment-related categories. The machine learning method is improved through integration for the accident level prediction. In the experiment, effective preprocessing measures were taken for problems such as data imbalance, missing values, the encoding of categorical variables, and the standardization of numerical features. The unbalanced distribution of “Severity” was improved through under-sampling and over-sampling techniques. Firstly, we adopted a multi-stage fusion strategy. A multi-layer perceptron (MLP) was used for the preliminary prediction, and then its result was combined with the original features to form a new feature. Decision tree, XGBoost, and random forest algorithms, respectively, were applied for the secondary prediction. The analysis results show that the improved machine learning model is significantly superior to a single model in the overall performance. The “MLP + random forest” model performs well in evaluation indicators such as the accuracy, recall rate, and F1 value. The accuracy rate is as high as 94%. In the prediction of different traffic accident severity levels (minor, moderate, and severe), the improved machine learning model also generally shows better performance and stability. The research results of this study have broad prospects in the field of intelligent driving. It can realize real-time accident prediction and early warnings, and provide decision support for drivers and autonomous driving systems. The research also provides a scientific basis for traffic planning and management departments to improve driving conditions and reduce the probability and losses of traffic accidents.

Densified wood possesses outstanding mechanical properties and serves as a desired construction material for modern timber buildings. However, the limited research on its flame behavior hinders its broader applications. The authors of this paper experimentally and analytically investigated the influence of wood species and density on horizontal flame spread behavior. Densified oak and densified fir were tested. The flame spread rate decreased with wood density in both densified wood types. Their values were close at the same density. The mass loss rate ($\dot{\mathit{m}}$) of the densified wood decreased with the increase in wood density. The densified oak had higher $\dot{\mathit{m}}$ due to its lower lignin content. Dimensionless correlations between the $\dot{\mathit{m}}$ and density were obtained which agree with the experiments. The flame heights (L_f) of the densified wood also decreased with the increase in wood density. The densified oak had higher L_f due to its higher $\dot{\mathit{m}}$. As the densified wood density increased, the radiation (and conduction) was reduced (and enhanced), but the convection remained constant. The densified oak had lower convection, lower conduction, and higher radiation than the densified fir at the same density. Gas-phase heat transfer was dominant in the flame spread of the densified wood, but conduction was also significant as its contribution can be as high as 70% of gaseous heat transfer.

The emergence of inclined tunnels under natural ventilation has brought many new fire safety issues. The smoke movement in the tunnel is affected by the chimney effect induced by the shaft and the downstream tunnel. The characteristics of temperature distribution in inclined tunnels are different from horizontal tunnels, which is worthy of further study. A series of conditions were carried out in an inclined model tunnel with a single shaft to investigate the temperature distribution characteristics. In this study, the longitudinal air inlet velocity is used to replace the longitudinal ventilation wind velocity. Results showed that the variation of fire source location L_f,, shaft height L_s, and the tunnel slope φ have obvious effect on the air inlet velocity. Based on the previous theories and the non-dimension analysis, the formulas of the dimensionless longitudinal inlet air velocity and the distribution of the maximum smoke temperature under the ceiling are proposed, which show good consistency with the simulation results. The reduced-scale experiments were conducted to validate the results of numerical simulation. The error range between the theoretical results and the simulation results is less than 20%.