Prediction of Shield Construction Risks in Subway Tunnelling Based on Fault Tree and Bayesian Network
(1 China Railway Engineering Equipment Group Mechanical and Electrical Engineering Co., Ltd., Chengdu 610083; 2 National Engineering Laboratory of Integrated Transportation Big Data Application Technology, Chengdu 611756; 3 School of Economics and Management, Southwest Jiaotong University, Chengdu 610031; 4 Sichuan Key Laboratory of Service Science and Innovation, Chengdu 610031; 5 Chengdu Tianfu International Airport Branch, Sichuan Airport Group Co., Ltd., Chengdu 641419)
Abstract：
At present, in the process of subway shield construction in China, risks and accidents occur frequently, and it has become important tasks in shield construction risk management to predict the occurrence trend of shield construction risks and improve the accuracy of risk response. This paper designs an effective method for prediction of shield construction risks and its development trends according to the alarm data of shield machines. Firstly, the fault tree method is used to identify the correlation between major shield construction risks and the fault alarm data of shield machines; then, on the basis of this, a risk prediction model based on Bayesian network is established, and the backward diagnostic inference technique of Bayesian network is applied to calculation of the maximum possible path to identify key shield failures that lead to risk events, so that subsequent risks could be strictly controlled in case of such failures; finally, a subway project is used as an example to verify the reliability and applicability of the method. The studied method and model can be embedded in various types of shield construction risk management systems to provide risk warning support to construction sites.
ZHANG Longguan1 LIU Suimei1 LI Kaifu1 XU Jin2,
3,
4 WANG Shengnan3 etc
.Prediction of Shield Construction Risks in Subway Tunnelling Based on Fault Tree and Bayesian Network[J] MODERN TUNNELLING TECHNOLOGY, 2021,V58(5): 21-29
2021, Vol. 58 