Given the flammability of power cables and the high cost of utility tunnel construction, power cable fires cause serious economic losses and are associated with a negative social impact. In the study, a weighted fuzzy Petri net and an event tree are combined to propose a quantitative evaluation method to mitigate cable fire risks in a utility tunnel. First, cable fire risk factors are analyzed. Given the lack of utility tunnel cable fire historical data, fuzzy theory is used to calculate the failure probability of the primary event. Second, a weighted fuzzy Petri net is used for fuzzy reasoning, and an event tree is used to analyze all possible consequences. Subsequently, the numerical simulation method is used to quantify the loss from the cable fire and thereby quantify the risk of cable fire. Finally, the effect of different risk factors on a cable fire is analyzed to determine the main factors that affect cable fires. Simultaneously, the control ability of different control measures with respect to the fire is analyzed to determine key control measures. A case study of a utility tunnel cable cabin in Liupanshui in Guizhou is employed to validate the utility of the proposed method.

This paper studies the limitations of binocular vision technology in monitoring accuracy. The factors affecting the surface displacement monitoring of the slope are analyzed mainly from system structure parameters and environment parameters. Based on the error analysis theory, the functional relationship between the structure parameters and the monitoring error is studied. The error distribution curve is obtained through laboratory testing and sensitivity analysis, and parameter selection criteria are proposed. Corresponding image optimization methods are designed according to the error distribution curve of the environment parameters, and a large number of tests proved that the methods effectively improved the measurement accuracy of slope deformation monitoring. Finally, the reliability and accuracy of the proposed system and method are verified by displacement measurement of a slope on site.

Intelligent safety management based on machine vision has become indispensable in reducing collision safety accidents during construction. To prevent collisions between workers and machines in excavation site construction, a real-time intelligent evaluation system to reflect worker–machine safety status was developed. The system included: (1) determination of the key factors affecting the safety of the interactive operation between workers and machines; (2) extraction of precursor semantic information related to the safety assessment for each object in the construction site based on machine vision; and (3) assessment of the safety state of a monitored object using a fuzzy neural network. A case study of excavation site construction is presented to illustrate and verify the entire process of safety assessment using the developed framework. The results show that the proposed model achieves high detection rates: 96% and 94% for tracking accuracy and 91.67% for prediction accuracy.

This paper proposes a systemic framework for dynamic safety risk analysis for natural gas pipeline leakage in a utility tunnel, which merges the bow-tie model (BT), Bayesian network (BN), and fuzzy set theory (FST) combined with monitoring data. Firstly, the hazard factors that cause natural gas leakage in a utility tunnel are identified to establish a BT model for simulating the causal relationship under specific accident scenes. Subsequently, the BT model is mapped to the BN model via a logic relationship for probabilistic reasoning, in which the prior probability of basic event is obtained through multiexpert scoring and FST. Then, the critical events can be obtained by using sensitivity analysis. Finally, combined with the monitoring data, the prior probabilities are updated to obtain the dynamic analysis result. A case study relating to safety risk analysis of overhead natural gas pipelines in a utility tunnel in the city of Liupanshui, China, is used to verify the feasibility of the approach as well as its application potential. The proposed method not only reduces the subjectivity of expert estimations, but also provides dynamic guidance for the risk analysis of utility tunnel gas pipelines.

Working at heights causes heavy casualties among workers during construction activities. Workers’ unsafe action detection could play a vital role in strengthening the supervision of workers to avoid them falling from heights. Existing methods for managing workers’ unsafe actions commonly rely on managers’ observation, which consumes a lot of human resources and impossibly covers a whole construction site. In this research, we propose an automatic identification method for detecting workers’ unsafe actions, considering a heights working environment, based on an improved Faster Regions with CNN features (Faster R-CNN) algorithm. We designed and carried out a series of experiments involving five types of unsafe actions to examine their efficiency and accuracy. The results illustrate and verify the method’s feasibility for improving safety inspection and supervision, as well as its limitations.

The measurement of the saturated internal friction angle of landslide sliding zone soils during the geological investigation could assist in engineering design, as it is a crucial parameter for landslide control. This research developed a comprehensive method to predict the saturated internal friction angle based on the support vector machine (SVM), after investigating the relationship between the saturated internal friction angle and the factors that affect the saturated internal friction angle. The procedure mainly consists of exploratory data analysis, influencing factors analysis, and prediction model set-up. Among the 18 factors, the evolution stage, gravel content, plasticity index, and dry unit weight were the most relevant factors predicting the saturated internal friction angle. The proposed method was illustrated and verified by a model accuracy test with the mean absolute error of the radial basis functions (RBF) kernel function was 0.90°, and the mean relative error was 8.77%. Therefore, this method could provide a reliable saturated internal friction angle in assisting engineers during landslide control.