| More and more railway tunnels are built in cold regions with unique seasonal climates and complex terrain conditions with the maturity of China’s rail transit construction technology.To strengthen the safety prevention of tunnel slopes along the railway in cold regions and ensure the safety of transportation,the research was carried out based on the special project of the Liaoning Provincial Science and Technology Department " Service Status Monitoring and Safety Status Assessment and Early Warning of High-speed Railway Tunnels."Intelligent processing was carried out on the horizontal displacement and elevation displacement data obtained by the Beidou high-precision sensing module on the slope of railway tunnel entrance.By combining the real-time monitoring data with the prediction data,the variation trends of the slope displacement in two directions were found.According to this trend,the safety evaluation and risk warning of slope region were studied in order to provide technical and decision support for the safety protection of the railway tunnel entrance slope in cold regions.(1)Selection of slope monitoring point and area location;monitoring data preprocessing and denoising.Firstly,the location of the measuring points and the safety evaluation area were determined according to the actual situation of the specific railway tunnel slope project in the cold region.Secondly,for real-time monitoring of horizontal and elevation displacement data,the Rhineda rule was used to eliminate the gross error of the original monitoring data.Then,different interpolation methods were used to complete the eliminated or missing monitoring data.After that,the nonlinear wavelet threshold denoising method was used for denoising.The comparison results showed that the denoising method with linear interpolation,the db7 wavelet base and the 3 denoising scale have the best error processing effect.Finally,the optimal denoising method was used to obtain the data series which can restore the actual values of horizontal and elevation displacement.(2)Intelligent learning of slope data and analysis of monitoring points displacement evolution trend.Firstly,the particle swarm optimization algorithm was used to optimize the hyperparameters of the LSTM model,and the number of PSO populations and iterations were determined by comparing the prediction effect evaluation indicators.Then the hyperparameters were input into LSTM to predict the horizontal and elevation displacement after processing.Finally,the PSO-LSTM model was compared with the LSTM model.The results showed that the RMSE and MAE of PSO-LSTM are less than 0.5%,and the effect was better.The maximum horizontal and elevation displacement in the next 14 days was 4mm and6 mm,respectively.(3)Regional evaluation and protection of railway tunnel entrance slope in cold region.A hierarchical model of safety evaluation was constructed.Firstly,the influencing factors of the complex railway tunnel entrance slope system in cold region were analyzed,and the evaluation index system of 4 first-level indexes and 18 second-level indexes was established and divided into 5 evaluation grades.Secondly,the entropy weight method was used to calculate the index weights according to the quantitative standards.Then,the variable fuzzy set model was used to evaluate the safety level of the slope monitoring area from the local measuring point to the whole slope.Finally,the safety grade of the slope monitoring area was Grade IV.And the corresponding suggestions and protective measures were put forward.(4)Real-time monitoring and early warning system for railway tunnel slope in cold region.JAVA technology was used to record and integrate the monitoring information of the monitoring points.The system includes the monitoring information display of each piece of equipment and displacement dynamic prediction;the slope safety evaluation grade was updated regularly.When predicting the possible instability of the slope,the emergency plan and protective measures will initiate.The system can monitor slope slide and rock fall in real time and accurately locate,form risk maps;and share the information synchronously to various departments for closed-loop management.Thus,the complete disease disaster prevention system at railway tunnel entrance slope in cold region was realized. |
PDF Full Text Request