A tunnel perimeter convergence monitoring and early warning method

Abstract

The present invention provides a tunnel perimeter convergence monitoring and early warning method, which is applicable to the field of tunnel construction monitoring, including monitoring point arrangement, monitoring data acquisition, data analysis and data early warning; the tunnel perimeter convergence monitoring and early warning method proposed in the present invention has the advantage of high accuracy and can be widely used in the field of tunnel construction monitoring.

Description

Technical Field

[0001] The invention relates to a tunnel perimeter convergence monitoring and early warning method, which is applicable to the field of tunnel construction monitoring. Background Art

[0002] As an important part of modern transportation infrastructure, tunnel engineering plays a key role in many fields such as urban rail transit, highways, and water conservancy projects. With the continuous deepening of underground space development and utilization, the scale and complexity of tunnel engineering are increasing, and its construction safety issues are becoming increasingly prominent. During tunnel construction, the stability of the surrounding rock and soil is one of the key factors to ensure engineering safety. Tunnel perimeter convergence, that is, the displacement of the surrounding rock and soil into the tunnel, is an important indicator for evaluating tunnel stability.

[0003] Traditional tunnel perimeter convergence monitoring methods mainly rely on contact measurement equipment, such as convergence meters and total stations. These methods have limitations such as low monitoring efficiency, inconvenient data collection, and inability to monitor in real time. With the development of information technology and intelligent sensing technology, the field of tunnel construction monitoring urgently needs a more efficient and intelligent monitoring and early warning method to achieve real-time monitoring, data analysis, and risk early warning of tunnel perimeter convergence.

[0004] However, in the current analysis of peripheral convergence data, the actual situation of the peripheral convergence data is not considered, and some algorithms are blindly used to build models to obtain unrealistic prediction values, which weakens its ability to guide construction; the main reason for this situation is that the peripheral convergence data has a positive and negative growth alternating trend, which is the result of the mutual adaptation between the tunnel surrounding rock and the support. Therefore, there is an urgent need for a tunnel peripheral convergence monitoring and early warning method that can effectively consider the positive and negative growth alternating trend of peripheral convergence data. Summary of the invention

[0005] The purpose of the present invention is to solve the problem that a tunnel perimeter convergence monitoring and early warning method that can effectively consider the positive and negative growth alternating trend of peripheral convergence data is urgently needed, and a tunnel perimeter convergence monitoring and early warning method is proposed.

[0006] The purpose of the present invention can be achieved by adopting the following technical solutions:

[0007] S101, monitoring point arrangement;

[0008] The arrangement of the monitoring points includes arranging peripheral convergence monitoring points at the specified section of the tunnel, wherein the peripheral convergence monitoring points are arranged in pairs, and the deformation of the tunnel section is studied by measuring the change of the value of the connecting line over time;

[0009] The arrangement of the monitoring points also includes that the arrangement of the monitoring points is determined according to the excavation method of the tunnel, and the spacing between the sections where the peripheral convergence monitoring points are arranged is determined according to the topographic and geological conditions of the tunnel and the tunnel construction method;

[0010] S102, monitoring data acquisition;

[0011] The acquisition of the monitoring data includes, after the arrangement of the peripheral convergence monitoring points is completed, measuring with a total station or a steel ruler convergence meter at a prescribed time interval to obtain the peripheral convergence data of the section, and arranging the peripheral convergence monitoring data obtained at each peripheral convergence monitoring point according to the monitoring time into a data set C i ;

[0012] S103, data analysis;

[0013] The data analysis includes constructing a prediction model based on the data set C using a gated recurrent unit neural network. i Obtaining the predicted value, the steps of the method of using the gated recurrent unit neural network to construct a prediction model to obtain the predicted value are:

[0014] a) Data preprocessing, wherein the data preprocessing includes converting the data set C i Adjust the data set to have consistent time intervals, and use polynomial interpolation for missing values.

[0015] b) data reshaping, reshaping the data set according to the gated recurrent unit neural network to obtain a reshaped data set,

[0016] c) dividing the reshaped data set into a training set and a test set,

[0017] d) constructing a prediction model using the gated recurrent unit neural network,

[0018] e) using the data of the training set to carry out model training on the model,

[0019] f) using the data of the test set to perform model testing on the model and obtain model testing results,

[0020] g) evaluating the model performance and adjusting it according to the model test results to obtain an adjusted model,

[0021] h) using the adjusted model to perform prediction to obtain a predicted value;

[0022] S104, data warning;

[0023] The early warning of the data includes comparing the obtained predicted value with the established early warning standard to determine whether to issue an early warning, thereby achieving the purpose of guiding construction. The established early warning standard is related to the topographic and geological conditions of the section where the surrounding convergence monitoring points are located and the excavation construction conditions;

[0024] Furthermore, in the above S103, the expression for constructing the polynomial P(x) in the polynomial interpolation method is formula (1),

[0025] P(x)=a 0 +a 1 x+a 2 x 2 +...+a n x n (1) In the formula, a 1 ~a n are the polynomial coefficients;

[0026] Furthermore, in the above S103, in the above 3), the data preprocessing method further includes data normalization, and the normalization method adopts the maximum-minimum value method, and the maximum-minimum value method is expressed as shown in formula (2),

[0027]

[0028] In the formula, X is the value before normalization, X MAX is the maximum value, X MIN is the minimum value, X N is the normalized value;

[0029] Furthermore, in the above S103, in order to be suitable for the surrounding convergence monitoring data with alternating positive and negative growth trends, the prediction model uses an adaptive piecewise linear function as an activation function, and the expression of the adaptive piecewise linear function is formula (3).

[0030]

[0031] The beneficial effects of the present invention are:

[0032] The beneficial effect of the present invention is: a tunnel perimeter convergence monitoring and early warning method that comprehensively considers the positive and negative growth alternating trend of perimeter convergence data. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 : A flow chart of a tunnel perimeter convergence monitoring and early warning method of the present invention;

[0034] Figure 2 : Monitoring value curve image of the peripheral convergence monitoring section C2H04+692 of the embodiment of the present invention;

[0035] Figure 3 : The monitoring value and prediction value curve image of the peripheral convergence monitoring section C2H04+692 of the embodiment of the present invention. DETAILED DESCRIPTION

[0036] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings; it should be understood that the specific embodiments given here are only used to illustrate and explain the present invention and cannot be used to limit the present invention.

[0037] The following is a specific embodiment of a tunnel perimeter convergence monitoring and early warning method.

[0038] like Figure 1 FIG. 1 is a flow chart of a tunnel perimeter convergence monitoring and early warning method according to the present invention.

[0039] S101, monitoring point arrangement;

[0040] The arrangement of the monitoring points includes arranging peripheral convergence monitoring points at the specified section of the tunnel, wherein the peripheral convergence monitoring points are arranged in pairs, and the deformation of the tunnel section is studied by measuring the change of the value of the connecting line over time;

[0041] The arrangement of the monitoring points also includes that the arrangement of the monitoring points is determined according to the excavation method of the tunnel, and the spacing between the sections where the peripheral convergence monitoring points are arranged is determined according to the topographic and geological conditions of the tunnel and the tunnel construction method;

[0042] In this embodiment, the project is the 8th sub-step of the East Coast Rail Link Project, C2 Line, Serendah Tunnel 1#. Serendah Tunnel 1# is located in the central part of the Malaysian Peninsula. The central part of the Malaysian Peninsula is a number of mountain ranges parallel to the direction of the peninsula, dividing the peninsula into three geographical areas: the eastern coastal belt, the central mountain belt, and the western coastal belt. The terrain is high and low on both sides. This project is mainly located in the eastern coastal belt and the central mountain belt. The area near the end belongs to the transition area from the central mountain belt to the western coastal belt. The lowest point along the line is about 3m, near the eastern coastal belt, and the highest point is about 1060m, near the central mountain belt. In this tunnel, the surrounding convergence monitoring points are arranged according to the requirements. The arrangement principle of the surrounding convergence monitoring points is arranged according to the clearance change section spacing table in Table 1. A total of 167 surrounding convergence sections are arranged in the Serendah Tunnel 1#.

[0043] Table 1 Clearance change section spacing

[0044] Surrounding rock level Section spacing Q≤0.1 5m 0.1＜Q≤1 10m 1＜Q≤10 30m

[0045] S102, monitoring data acquisition;

[0046] The acquisition of the monitoring data includes, after the arrangement of the peripheral convergence monitoring points is completed, measuring with a total station or a steel ruler convergence meter at a prescribed time interval to obtain the peripheral convergence data of the section, and arranging the peripheral convergence monitoring data obtained at each peripheral convergence monitoring point according to the monitoring time into a data set C i ;

[0047] In this embodiment, the peripheral convergence monitoring points are monitored in accordance with the specifications, and the instrument used is a steel ruler convergence meter, and the parameters are shown in Table 2. Taking the data of a peripheral convergence monitoring section C2H04+692 of the tunnel as an example, it is organized into a data set C 1 , the dataset C 1 The changing trend is Figure 1 shown.

[0048] Table 2 Steel Ruler Convergence Meter Parameters

[0049] project parameter Instrument Model Steel ruler convergence meter / JY-YT-SLJ-1 Specification 20 / 30 / 50 Standard range 20、30、50M Minimum reading 0.01mm Systematic Error ≤0.2m

[0050] S103, data analysis;

[0051] The data analysis includes constructing a prediction model based on the data set C using a gated recurrent unit neural network. i Obtaining the predicted value, the steps of the method of using the gated recurrent unit neural network to construct a prediction model to obtain the predicted value are:

[0052] a) Data preprocessing, wherein the data preprocessing includes converting the data set C i Adjust the data set to have consistent time intervals, and use polynomial interpolation for missing values.

[0053] b) data reshaping, reshaping the data set according to the gated recurrent unit neural network to obtain a reshaped data set,

[0054] c) dividing the reshaped data set into a training set and a test set,

[0055] d) constructing a prediction model using the gated recurrent unit neural network,

[0056] e) using the data of the training set to carry out model training on the model,

[0057] f) using the data of the test set to perform model testing on the model and obtain model testing results,

[0058] g) evaluating the model performance and adjusting it according to the model test results to obtain an adjusted model,

[0059] h) using the adjusted model to perform prediction to obtain a predicted value;

[0060] Furthermore, in the above S103, the expression for constructing the polynomial P(x) in the polynomial interpolation method is formula (1),

[0061] P(x)=a 0 +a 1 x+a 2 x 2 +...+a n x n (1) In the formula, a 1 ~a n are the polynomial coefficients.

[0062] Furthermore, in the above S103, in the above 3), the data preprocessing method further includes data normalization, and the normalization method adopts the maximum-minimum value method, and the maximum-minimum value method is expressed as shown in formula (2),

[0063]

[0064] In the formula, X is the value before normalization, X MAX is the maximum value, X MIN is the minimum value, X N is the normalized value.

[0065] Furthermore, in the above S103, in order to be suitable for the surrounding convergence monitoring data with alternating positive and negative growth trends, the prediction model uses an adaptive piecewise linear function as an activation function, and the expression of the adaptive piecewise linear function is formula (3).

[0066]

[0067] In this embodiment, a prediction model is constructed using a gated recurrent unit neural network based on the data set C 1 The predicted value is obtained, and the change trend between the predicted value and the measured value is as follows Figure 2 As shown, from Figure 2 It can be seen that the predicted value and the measured value are basically consistent in the changing trend, which can better reflect the situation of the measured value.

[0068] S104, data warning;

[0069] The early warning of the data includes comparing the obtained predicted value with the established early warning standard to determine whether to issue an early warning, thereby achieving the purpose of guiding construction. The established early warning standard is related to the topographic and geological conditions and excavation construction conditions of the section where the surrounding convergence monitoring points are located.

[0070] In this embodiment, the warning standard is formulated as described in Table 3. The predicted values ​​of the peripheral convergence monitoring section C2H04+692 all exceed the warning standard, and no warning is issued, indicating that the change in the peripheral convergence clearance displacement of the section is under control.

[0071] Table 3 Reference for early warning standard values

[0072]

[0073] In the above embodiments, the present invention discloses a tunnel perimeter convergence monitoring and early warning method, including monitoring point arrangement, monitoring data acquisition, data analysis and data early warning; the tunnel perimeter convergence monitoring and early warning method proposed in the present invention has the advantage of high accuracy and can be widely used in the field of tunnel construction monitoring.

[0074] The above description is a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A tunnel perimeter convergence monitoring and early warning method, characterized in that: The steps of the tunnel perimeter convergence monitoring and early warning method are as follows: 1) Layout of monitoring points; 2) Acquisition of monitoring data; 3) Data analysis; 4) Data early warning; The arrangement of the monitoring points includes arranging peripheral convergence monitoring points at a specified section of the tunnel, wherein the peripheral convergence monitoring points are arranged in pairs, and the deformation of the tunnel section is studied by measuring the change of the value of the connecting line over time; The arrangement of the monitoring points also includes that the arrangement of the monitoring points is determined according to the excavation method of the tunnel, and the spacing between the sections where the peripheral convergent monitoring points are arranged is determined according to the topographic and geological conditions of the tunnel and the tunnel construction method; The acquisition of the monitoring data includes, after the arrangement of the peripheral convergence monitoring points is completed, measuring with a total station or a steel ruler convergence meter at a prescribed time interval to obtain the peripheral convergence data of the section, and arranging the peripheral convergence monitoring data obtained at each peripheral convergence monitoring point according to the monitoring time into a data set C i ; The data analysis includes constructing a prediction model based on the data set C using a gated recurrent unit neural network. i Obtaining the predicted value, the steps of the method of using the gated recurrent unit neural network to construct a prediction model to obtain the predicted value are: a) Data preprocessing, the data preprocessing includes converting the data set C i Adjust the data set to have consistent time intervals, and use polynomial interpolation for missing values. b) data reshaping, reshaping the data set according to the gated recurrent unit neural network to obtain a reshaped data set, c) dividing the reshaped data set into a training set and a test set, d) constructing a prediction model using the gated recurrent unit neural network, e) using the data of the training set to carry out model training on the model, f) using the data of the test set to perform model testing on the model and obtain model testing results, g) evaluating the model performance and adjusting it according to the model test results to obtain an adjusted model, h) using the adjusted model to perform prediction to obtain a predicted value; The early warning of the data includes comparing the obtained predicted value with the established early warning standard to determine whether to issue an early warning, thereby achieving the purpose of guiding construction. The established early warning standard is related to the topographic and geological conditions and excavation construction conditions of the section where the surrounding convergence monitoring points are located.

2. A tunnel perimeter convergence monitoring and early warning method according to claim 1, characterized in that: In the above 3), the expression for constructing the polynomial P(x) in the polynomial interpolation method is formula (1), P(x)=a0+a1x+a2x 2 +...+a n x n (1) In the formula, a1~a n are the polynomial coefficients.

3. A tunnel perimeter convergence monitoring and early warning method according to claim 1, characterized in that: In the above 3), the data preprocessing method further includes data normalization, and the normalization method adopts the maximum-minimum value method, and the maximum-minimum value method is expressed as shown in formula (2), In the formula, X is the value before normalization, X MAX is the maximum value, X MIN is the minimum value, X N is the normalized value.

4. A tunnel perimeter convergence monitoring and early warning method according to claim 1, characterized in that: In the above 3), in order to adapt to the positive and negative growth alternating trend of the surrounding convergence monitoring data, the prediction model adopts an adaptive piecewise linear function as the activation function. The expression of the adaptive piecewise linear function is formula (3): Where f(x) is an adaptive piecewise linear function.

Images