Frequent extreme rainstorms have significantly increased the flooding risk, threatening the security and stability of electrical substations. The process of flood-induced substation damage is complex and nonlinear, challenging traditional predictive methods. Therefore, a novel predictive framework is proposed for flood-induced substation damage. This framework uses a generative adversarial network (GAN)-based model to capture complex data relationships and generate realistic samples, which mitigates training data imbalance. A multivariate predictive Transformer network (MPformer), integrating three improved modules: time embedding, multi-factor fusion encoding, and attention-based encoder, is proposed to capture temporal dependencies and complex interactions between influencing factors and flood-induced damage. Based on MPformer and sensitive cost learning, a twostage integrated model is designed to reduce the problem of sample imbalance further and realize the simultaneous prediction of the substation damage probability, severity, and time. The experimental results show that the GAN-based method is superior to the traditional method in terms of sample balancing, and the MPformer-based two-stage model outperforms the mainstream model, with a 12.30 % average increase in F1 score for probability prediction and reductions of 38.56 % and 45.31 % in RMSE for severity and time predictions, respectively. A case study shows that the proposed method can offer reliable pre-disaster prediction.

The mechanical properties of the tunnel concrete lining degrade over time during operation, and under fire conditions, its durability and reliability can be further compromised, potentially rendering it nonfunctional, suggesting that evaluating the mechanical properties and reliability of the tunnel under fire is essential for its maintenance and long-term sustainability. Therefore, this study develops fragility curves to assess fire-induced damage to highway tunnels, considering uncertainties such as service time, heat release rate, burning duration, and lining thickness. The ratio of the actual bending moment to the ultimate bending moment of the lining under fire is used as the evaluation index, and the lining damage is corrected by considering the concrete spalling phenomenon at high temperatures. Finally, the effect of ventilation on tunnel lining damage is considered, and the damage to the lining is evaluated. The specific conclusions are as follows: (1) The impact of damage from fire on the lining is small for tunnels with less than 50-year service time, beyond which the impact of damage from fire on the lining increases significantly. (2) A concrete spalling model is proposed in this paper, which differs by only 1 mm from the test results and 10% from the actual fire tunnel spalling thickness, confirming the accuracy of the concrete spalling model. The method is used to correct the friability curves to increase the accuracy of the results. (3) Repair advice under different damage conditions is given.

The harsh environment in tunnels with high geothermal temperatures and humidity can adversely impact machinery, personnel, and construction. The main causes of specific problems are the unknown mechanisms of local geothermal formation, inappropriate temperature control measures, and insufficient systematic safeguards. In this study, three work sections relating to a high geothermal tunnel are: the tunnel face, middle-of-tunnel section, and outside-of-tunnel section. A cooling strategy is proposed to offer technical support in achieving comprehensive cooling, overall as well as for each of the sections. First, a comprehensive geological survey explores the mechanism and exact location of the heat source. Secondly, grouting and centralized drainage measures are used to control the heat release of hot water. Enhanced ventilation, ice chillers and other applicable measures are used to control the ambient temperature. Finally, a monitoring and early warning system is established to prevent accidents. This cooling strategy has been applied in the field with good results.

The change trend of the temperature field in high geothermal tunnels directly affects the ambient temperature and the internal force of the lining. Currently, the experimental insulation is poor and the thermal conductivity of similar materials is much smaller than that of the prototype material. This results in a temperature difference of 20-30°C between the boundary temperature and the liner temperature at the initial condition. So the results of the tunnel temperature field have a large bias. Moreover, the experiment has been carried out at lower temperatures, and the temperature field of the ultra-high geothermal tunnel has not been addressed. To eliminate this experimental bias, a 1:40 model experiment was designed in this study. 10% aluminum powder was added to the similar material to make the thermal conductivity the same as the prototype material. In order to minimize heat loss, 5 cm of insulation was placed on both the inner and outer sides of the model experiment. The initial surrounding rock temperature is set to 40°C, 60°C and 80°C, which is favorable for exploring the change rule of temperature field of different initial surrounding rock temperatures. The effects of surrounding rock temperature, wind speed and airflow temperature on the temperature field were considered, and the experiment was started only when the bias between the boundary temperature and lining temperature was less than 5°C. This greatly improves experimental accuracy. The specific conclusions are as follows: Comparing the model experiment results with the numerical simulation, the bias is less than 3°C; thus, experimental results are reliable. From the inside of the second lining to the boundary of the surrounding rock, the temperature change decreases, and the surrounding rock temperature does not basically change after a distance of 26 cm from the lining. The surrounding rock temperature, wind speed and airflow temperature all affect the temperature field of the tunnel; the wind speed and surrounding rock temperature have the greatest influence, and the airflow temperature has the least influence.

Tunnels with surrounding rock temperatures exceeding 28°C are called high geothermal tunnels. During the construction of high geothermal tunnels, construction personnel health, machine operation efficiency and second lining safety are affected by the heat tunnel. Therefore, calculating the precise heat release characteristics of high-geothermal temperature tunnels is a prerequisite for ensuring the smooth passage of the tunnel. The guideline that states that the wall temperature is constant in the longitudinal direction of the tunnel is not realistic. This constant temperature results in large heat release calculations and leads to increased energy wastage while cooling high geothermal tunnels. Therefore, a method is proposed for calculating the heat release along the longitudinal direction of a tunnel based on a three-dimensional implicit finite-difference technique considering the ventilation time and average wind speed. This calculation method is used to determine the heat release values of the surrounding rock, initial support and second lining and to compare these findings with the guideline algorithm. The results show that the difference between the internal surrounding rock temperature calculated by the three-dimensional implicit finite-difference method and the measured value is 0.94 °C, verifying the theory accuracy. The heat release of the initial support section is much higher than that of the surrounding rock and the second lining; all heat release values are smaller than the normative heat release. Therefore, the heat release calculation method should be extended and used as a basis for calculating ventilation and cooling in other high geothermal tunnel construction projects.

Tunnels with surrounding rock temperatures exceeding 28 degrees C are called high geothermal tunnels. During the construction of high geothermal tunnels, construction personnel health, machine operation efficiency and second lining safety are affected by the heat tunnel. Therefore, calculating the precise heat release characteristics of high-geothermal temperature tunnels is a prerequisite for ensuring the smooth passage of the tunnel. The guideline that states that the wall temperature is constant in the longitudinal direction of the tunnel is not realistic. This constant temperature results in large heat release calculations and leads to increased energy wastage while cooling high geothermal tunnels. Therefore, a method is proposed for calculating the heat release along the longitudinal direction of a tunnel based on a three-dimensional implicit finite-difference technique considering the ventilation time and average wind speed. This calculation method is used to determine the heat release values of the surrounding rock, initial support and second lining and to compare these findings with the guideline algorithm. The results show that the difference between the internal surrounding rock temperature calculated by the three-dimensional implicit finite-difference method and the measured value is 0.94 degrees C, verifying the theory accuracy. The heat release of the initial support section is much higher than that of the surrounding rock and the second lining; all heat release values are smaller than the normative heat release. This method of calculating heat release is more realistic and can effectively lay the foundations for high geothermal tunnel cooling while saving costs. Therefore, the heat release calculation method should be extended and used as a basis for calculating ventilation and cooling in other high geothermal tunnel construction projects.

Weathered rocks typically exist in the portal section of mountain tunnels where the tunnels in the portal section often cross soft and hard rock strata. The interface between the soft and hard rocks is the control point in the seismic design of tunnels. However, there is currently no seismic design method based on dynamic methods that consider tunnels crossing soft and hard rock strata along their longitudinal direction. In this study, two elastic foundation beams in different rock strata were used to model the behaviour of a tunnel crossing soft and hard rock strata. Based on the Timoshenko beam vibration response theory, the higher-order differentials of the beam dynamic governing equation are transformed into algebraic equations using the Laplace transform. The closed-form solution of the dynamic response of the entire beam is derived by combining the boundary and deformation coordination conditions of two elastic foundation beams. Subsequently, the dynamic response of any point along the longitudinal direction of the elastic foundation beam (tunnel) crossing the soft and hard rock strata was obtained using Green's function. The validity of the method was verified by comparing it with a finite difference model. Finally, a parametric analysis was performed based on the analytical method derived in this study to investigate the influences of parameters such as the liner elastic modulus, liner thickness, modulus of elasticity ratio of hard rock to soft rock, and axial forces, on the seismic response of tunnels crossing soft and hard rock strata. The results suggest that with an increase in lining thickness, the bending moment and shear force of the tunnel become increasingly significant, but the transverse displacement response of the tunnel was slightly affected. With an increase in the elastic modulus of the lining, the difference in the peak values of the bending moment and shear force increases, whereas the transverse displacement response is slightly diminished. With an increase in the elastic modulus ratio of hard rock to soft rock, the difference in the peak values of the bending moment, shear force, and transverse displacement of the tunnel, varied more along the tunnel axis. With the increase in the axial force, the seismic responses from the bending moment and shear force on the tunnel cross-section were significantly weakened.

Because of the harsh environment around city subways and the typically poor geological conditions, it is often difficult to ensure the stability of a tunnel, which makes it prone to collapses and other accidents. The influence of multiple factors on tunnel stability was analyzed under the same geological conditions and environmental factors by using a three-dimensional numerical simulation. Taking into account the objective environment and manual operation and other factors, construction sequence, support closure time, and reinforcement area and parameters were chosen as the main factors affecting the tunnel stability. Finally, based on ground deformation, tunnel convergence deformation, and surrounding soil stress as evaluation indicators, a fuzzy analytic hierarchy process (FAHP)-gray correlation degree-technique for order performance by similarity to ideal solution (TOPSIS) model was established. It was found that the higher the relative closeness of the construction parameter combination, the greater the tunnel stability. The best combination of construction parameters was applied to the Tiantongyuan Station project of Beijing Metro Line 17. All geological conditions being equal, reasonable selection of construction parameters can ensure optimal tunnel construction stability, reduce construction risks, and support sustainable tunnel development. In addition, the FAHP-gray correlation degree-TOPSIS method achieved multi-index evaluation of tunnel stability, which provides practical guidance for the construction of similar projects.

为了对近海砂性土层地铁深基坑进行安全风险评估,建立了地铁车站施工安全风险评估体系,运用模糊层次分析法确定基坑施工风险因素指标的权重,在此基础上,结合灰色关联度法,计算出风险评估系统中总风险层与第一层次因素的关联度,以及第一层次因素与其对应第二层次因素的关联度.实例分析结果表明,基于模糊层次分析法的灰色关联度评价方法能有效降低主观因素的影响,且该方法的运用不受样本量大小的影响,数学处理计算过程简单,为深基坑安全风险评价与风险管理提供参考.