Collapse is a major engineering hazard in open-cut foundation pit construction, and risk assessment is crucial for considerably reducing engineering hazards. This study aims to address the ambiguity problem of qualitative index quantification and the failure of high-conflict evidence fusion in risk assessment. Thus, a fast-converging and high-reliability multi-source data fusion method based on the cloud model (CM) and improved Dempster–Shafer evidence theory is proposed. The method can achieve an accurate assessment of subway pit collapse risks. First, the CM is introduced to quantify the qualitative metrics. Then, a new correction parameter is defined for improving the conflicts among evidence bodies based on conflict degree, discrepancy degree and uncertainty, while a fine-tuning term is added to reduce the subjective effect of global focal element assignment. Finally, the risk assessment result is obtained according to the maximum affiliation principle. The method is successfully applied to Luochongwei Station, where the difference between the maximum value and the second largest value of the basic probability assignment is 0.624, and the global uncertainty degree is 0.087. Both values satisfy the decision evaluation condition; however, values of other methods only satisfy one or neither condition. In addition, the proposed method requires only four cycles to reach the steady state by fusing data of the same index, which has faster convergence compared with that of other methods. The proposed method has good universality and effectiveness in subway pit collapse risk assessment.

Collapse is the main engineering disaster in tunnel construction when using the drilling and blasting method, and risk assessment is one of the important means to significantly reduce engineering disasters. Aiming at the problems of random decision-making and misjudgment of single indices in traditional risk assessment, a multi-source data fusion method with high accuracy based on improved Dempster–Shafer evidence theory (D-S model) is proposed in this study, which can realize the accurate assessment of tunnel collapse risk value. The evidence conflict coefficient K is used as the identification index, and the credibility and importance are introduced. The weight coefficient is determined according to whether the conflicting evidence is divided into two situations. The advanced geological forecast data, on-site inspection data and instrument monitoring data are trained by Cloud Model (CM), Gradient Boosting Decision Tree (GBDT) and Support Vector Classification (SVC), respectively, to obtain the initial BPA value. Combined with the weight coefficient, the identified conflict evidence is adjusted, and then the evidence from different sources is fused to obtain the overall collapse risk value. Finally, the accuracy is selected to verify the proposed method. The proposed method has been successfully applied to Wenbishan Tunnel. The results show that the evaluation accuracy of the proposed multi-source information fusion method can reach 88%, which is 16% higher than that of the traditional D-S model and more than 20% higher than that of the single-source information method. The high-precision multi-source data fusion method proposed in this paper has good universality and effectiveness in tunnel collapse risk assessment.

Considering the complexity of the metro pit construction environment, the existing risk early-warning methods cannot ensure high-precision early warning. A high-accuracy metro pit collapse risk fusion early-warning method is proposed in present study. The main contributions include (1) presenting a new input to the fusion model by optimizing the machine learning model through a multi-step rolling method, and then using the basic probability assignment values obtained from the cloud model as input to the fusion model and (2) developing an improved methodology to address the paradoxical results of the fusion of traditional Dempster–Shafer evidence theory when there is a high level of conflict in multi-source risk prediction data. The proposed method is successfully applied to the Guangzhou Metro station project. By analyzing the early-warning results of 240 moments in 6 monitoring points, compared with the single information source method and the traditional D-S method, the early-warning accuracy of this method is increased by 15.8% and 10.8% respectively, the false alarm rate is reduced by 6.3% and 5.5%, respectively, and the missed alarm rate is reduced by 9.5% and 5.3%, respectively. The high-accuracy fusion early-warning method proposed in this paper has good universality and effectiveness in the early warning of subway foundation pit collapse risk.

The backfill area of tunnel projects may deform or collapse due to the cyclic disturbance of groundwater and train loads. Hence, the anti-deformation and crack resistance performance of backfill materials under cyclic disturbance is critical to engineering safety. In this paper, concrete was produced by mixing 0.85 mm, 1–3 mm and 3–6 mm rubber particles instead of 10% sand, and tested to discuss the effect of rubber particle size on the deterioration of concrete material properties (compressive characteristics and energy dissipation) after bearing cyclic loading. The stress–strain curve and various parameters obtained through the uniaxial compression test and cyclic load test were used to explore the optimal grain size that can be applied to the tunnel engineering backfill area, and numerical simulation was adopted to calculate the deformation of the surrounding rock and the structural stress of different backfill materials. Research shows that the increase in particle size lessens the compressive strength, deformation resistance and cracking resistance of specimens, but after the cyclic loading test, the concrete material deterioration analysis indicates that rubber concrete has lesser and more stable losses compared to ordinary concrete, so the optimum rubber particle size is 0.85 mm. Numerical calculations show that RC-1 reduces the arch top displacement by 0.4 mm, increases the arch bottom displacement by 0.6 mm and increases the maximum principal stress by 11.5% compared to OC. Therefore, rubber concrete can ensure the strength and stability requirements of tunnel structures, which can provide a reference for similar projects.

Collapse is the main engineering hazard in the construction of open-cut foundation pit, and risk assessment is one of the important ways to reduce it. Aiming at the high conflicting evidence fusion failure, a fast convergence and highly reliable multi-data fusion method based on cloud model (CM) and improved Dempster-Shafer evidence theory is proposed, which can achieve accurate assessment of subway pit collapse risk. Firstly, CM is introduced to quantify the qualitative metrics. Then, A new correction parameter is defined for improving the conflicting among evidence bodies based on conflict degree, discrepancy degree and uncertainty, while a fine-tuning term is added to reduce the subjective effect of global focal element assignment. Finally, the risk level is obtained by the maximum affiliation principle. The method has been successfully applied to Luochongwei Station. The difference between the maximum value and the second largest value of the basic probability assignment is 0.624, and the global uncertainty degree is 0.087, both of which satisfy the decision evaluation condition, but other methods can only satisfy one or neither of them. And it requires only 5 cycles to reach the steady state by fusing data of the same index, which has faster convergence compared with other methods. The proposed method has good universality and effectiveness in subway pit collapse risk assessment.

To address the complex mechanical properties of coastal clay encountered during artificial ground freezing construction in coastal regions, this study investigates coastal clay by conducting mechanical tests under various freezing temperatures and confining pressures. The variations in peak strength, elastic modulus, and shear strength parameters of the frozen coastal clay were systematically analyzed. The results indicate that the peak strength and elastic modulus of the coastal clay exhibit a clear upward trend with approximately linear increases in confining pressure and temperature. A damage constitutive model for frozen soil was established, accounting for the coupled effects of temperature and confining pressure. Furthermore, a Physics-Informed Neural Network (PINN), incorporating the constitutive equations as physical constraints, was introduced to achieve the inversion and prediction of model parameters. The results demonstrate that the PINN model exhibits excellent predictive accuracy, with the coefficient of determination (R²) reaching 0.8513 under uniaxial compression and 0.9689 under triaxial compression. Compared with traditional empirical models, the PINN model reduces the prediction error of the elastic modulus to 8.96% (at 0 MPa) and 4.79% (at 0.8 MPa), demonstrating superior generalization capability. This study provides a reliable theoretical basis and parameter guidance for frozen soil construction in coastal regions.