In order to ensure the safety and reliability of railway tunnels during service, a health diagnosis and ear-ly warning technology scheme for railway tunnel linings in seasonally frozen areas is proposed. Firstly, a tunnel lin-ing health diagnosis index system including 7 first-level indicators and 13 second-level indicators is constructed in combination with the actual tunnel lining. Then, the freezing season weight and non-freezing season weight of each index are determined based on the improved group G2 method. Finally, the SA-BP combined prediction model is selected to realize comprehensive diagnosis and early warning of tunnel lining health by comparing the prediction accuracy and stability of SARIMA prediction model, BP neural network model and SA-BP combined prediction model. The research results show that the early warning values in January and February 2022 are 0.4734 and 0.4720, respectively, and the early warning level is level 2, but there is a trend of deterioration to level 3. It is recommended to strengthen maintenance and not call the police for the time being.
关键词
季冻区/铁路隧道/健康诊断/改进群组G2法/神经网络
Key words
seasonal frozen area/railway tunnel/health diagnosis/improved group G2 method/neural network
维普
万方数据