BOTDA-based monitoring and early warning method during tunnel operation

Abstract

A monitoring and early warning method for tunnel operation based on BOTDA is proposed. Since BOTDA mainly monitors the temperature and strain distribution information of each measuring point in the tunnel, this invention performs certain data processing on the strain measured at the measuring points. The steps are as follows: BOTDA optical fiber is deployed in the tunnel; strain data of the grating monitoring points are extracted; based on the longitudinal strain vector angle between the measuring points obtained in different tunnel surrounding rock grade sections, the monitoring data of each tunnel surrounding rock section is clustered; the threshold for uneven settlement of the tunnel structure is determined; based on the healthy monitoring data, the data to be diagnosed is judged by threshold judgment to analyze whether uneven settlement has occurred in the tunnel.

Description

Technical Field

[0001] This invention relates to the field of monitoring technology during the operation of underground tunnels, and particularly to a tunnel monitoring and early warning method based on BOTDA. The innovation lies in using Mahalanobis distance as the vector angle calculation method, and using the small error of the sensitivity matrix as a method to determine whether uneven settlement has occurred in the tunnel. BOTDA refers to Brillouin Optical Time Domain Analyzer. Background Technology

[0002] With the rapid development of society and economy and the continuous construction of underground tunnels, more and more tunnels need maintenance. However, manual maintenance is time-consuming, labor-intensive and ineffective. Therefore, there is an urgent need for a monitoring and early warning technology for tunnel operation.

[0003] Based on BOTDA, the relationship between Brillouin scattering spectrum and temperature, strain, etc., can be used to measure the temperature and strain of optical fibers by observing changes in the Brillouin scattering spectrum. Since the strain value at each measuring point is extremely small in actual tunnel operation, and uneven settlement often occurs between different tunnel sections, this invention proposes using the strain angle between different tunnel sections as the unit, and using the change in the strain angle between sections to determine whether uneven settlement has occurred. Summary of the Invention

[0004] The purpose of this invention is to solve the problems existing in the background art and to provide a monitoring and early warning method for tunnel operation based on BOTDA.

[0005] The BOTDA-based monitoring and early warning method for tunnel operation includes the following steps:

[0006] 1) Install BOTDA fiber optic sensors inside the tunnel;

[0007] 2) Extract strain data from fiber optic measuring points. Since the strain of the tunnel varies under different surrounding rock grades, the GMM algorithm is used to divide the tunnel into different intervals based on the number of different surrounding rock grades (k=N).

[0008] 3) Using the strain of adjacent tunnel segments as vectors, and combining it with Mahalanobis distance, the vector angle between different intervals is calculated;

[0009] 4) Based on the divided angle, establish the diagnostic factors for uneven settlement, and use low-probability events to determine the threshold of uneven settlement in different sections of the tunnel.

[0010] 5) Using the determined threshold, after processing the strain data under the condition to be diagnosed as described above, compare it with the diagnostic threshold to determine whether uneven settlement has occurred in the tunnel.

[0011] The GMM mentioned refers to the Gaussian Mixture Model algorithm.

[0012] The beneficial effects of this invention are:

[0013] This invention is applicable to health monitoring of tunnels during tunnel operation. It not only performs temperature-compensated preprocessing on the monitoring data to improve accuracy, but also processes the monitoring data using different rock grades in different sections of the tunnel and improves calculation efficiency by utilizing the included angles of different sections. This ensures the safe operation of the tunnel and improves monitoring efficiency. Attached Figure Description

[0014] Figure 1 A schematic diagram of the structure for laying BOTDA optical fibers inside a tunnel;

[0015] Figure 2 It is an early warning diagram of the strain angle between tunnel sections based on different surrounding rock grades in the invention. Detailed Implementation

[0016] A monitoring and early warning method for tunnel operation based on BOTDA includes the following steps:

[0017] 1) Two BOTDA optical fibers are laid along the entire length of both sides inside the tunnel to form a loop, such as... Figure 1 As shown.

[0018] 2) Extract strain data from each measuring point on the optical fiber, eliminate the influence of ambient temperature on strain, and then use the GMM algorithm to divide the measured data into N clusters, where N represents the number of different surrounding rock grades passed along the tunnel direction. The data in each cluster have similar tunnel surrounding rock grades and surrounding soil conditions.

[0019] 3) After the tunnel section classification is completed, the structural measuring points of a certain cluster interval are divided into N segments on average, with p measuring points in each segment. The specific classification diagram is shown below. Figure 2 As shown.

[0020] 4) The strain values ​​of each tunnel segment are treated as a vector. Based on the cluster division, there are N vectors, denoted as X1, X2…X… N This represents the strain value vector of a tunnel section, calculated by taking the longitudinal strain vectors of adjacent tunnel sections as the vector angle. Figure 2 As shown, the calculation formula is:

[0021] θ i =sub(X) i X i+1 i = 1, 2, N-1

[0022] Therefore, the angle between the vectors of the first two tunnel sections can be obtained as follows: The angle between the vectors at time n is defined as follows, and it is defined as a random variable:

[0023]

[0024] Therefore, the set of vector angles between all tunnel structural sections can be obtained as follows:

[0025] X = [X1, X2, ..., X] m-1 ]

[0026] The mean and covariance matrix of the angle between the vectors are as follows:

[0027]

[0028]

[0029] Since X is defined as a random variable in the above paragraph, the sample estimation set for each measurement point can be f(X) = [f(X1), f(X2)...f(X...]. N Combining the Taylor first sequence expansion, the sample estimation set f(X) can be expanded as follows:

[0030]

[0031] In the formula It is the strain values ​​of the sample set f(X) for different tunnel surrounding rock intervals. deviation, This is called the sensitivity matrix, which can be used to adjust for uneven settlement in tunnel sections. Using a first-order Taylor expansion, the covariance matrix of the sample set f(X) can be transformed into:

[0032]

[0033] To fully consider the distribution of sample data among tunnel vectors, Mahalanobis distance is used as the distance between vector angle samples. The resulting Mahalanobis distance is used as a diagnostic factor for uneven settlement in the tunnel section.

[0034]

[0035] The Mahalanobis distance between the strain vectors at n time points under healthy conditions is expressed by the following formula:

[0036]

[0037] Therefore, the diagnostic factor for uneven settlement of the tunnel structure under this cluster can be expressed by the following formula:

[0038] Φ = [Φ1, Φ2, ... Φ N-1 ]

[0039] The diagnostic threshold for uneven settlement of tunnel structures is determined based on low-probability events. Theoretically, the vector values ​​of adjacent clusters should not change unless structural damage such as uneven settlement occurs locally in the tunnel structure. The n diagnostic factors for uneven settlement of tunnel structures, calculated from the angle between the strain vectors of measuring points under the initial healthy state, are rearranged in ascending order:

[0040]

[0041] In the above formula This represents the minimum value of the Φ1 diagnostic factor. The maximum value of Φ1

[0042] By assigning 0.95 as the ranking order of the diagnostic factor, the diagnostic factor can be diagnosed through tunneling at the 0.95n position. This serves as a diagnostic threshold for uneven settlement in tunnels. To improve the accuracy, tolerance, and efficiency of the diagnostic process, this paper multiplies it by a... Damage diagnosis is performed using coefficients.

[0043]

[0044] 5) Finally, the extracted tunnel strain data is processed in the above steps, and the obtained uneven settlement diagnostic factor is compared with the diagnostic threshold for the healthy state of the tunnel. If the diagnostic factor exceeds the diagnostic threshold determined above, it is considered that the tunnel has experienced uneven settlement. If it does not exceed the diagnostic threshold determined above, it is considered that the tunnel has not experienced uneven settlement.

[0045] This invention is applicable to health monitoring of tunnels during tunnel operation. It not only performs temperature-compensated preprocessing on the monitoring data to improve accuracy, but also processes the monitoring data using different rock grades in different sections of the tunnel and improves calculation efficiency by utilizing the included angles of different sections. This ensures the safe operation of the tunnel and improves monitoring efficiency.

Claims

1. Early warning method based on BOTDA monitoring during tunnel operation, characterized in that: Comprising the following steps: 1) arranging BOTDA optical fiber sensor in the tunnel: Two BOTDA optical fibers are arranged on both sides of the tunnel hole to form a loop; 2) extracting strain data of each measuring point on the optical fiber, eliminating the influence of ambient temperature on strain, and then using GMM algorithm to divide the measured data into N clusters, N representing the number of different surrounding rock grades passed along the tunnel direction, and each cluster has similar tunnel surrounding rock grade and surrounding soil condition; 3) After the tunnel section classification, the structure measuring points in a certain cluster interval are evenly divided into N sections, and there are p measuring points in each interval; 4) The strain value of the tunnel section is taken as a vector, and there are N vectors according to the division of the cluster. Each vector represents the strain value vector of the corresponding tunnel section. The vector angle between the strain value vectors of adjacent tunnel sections is calculated to obtain the vector angle set X of all tunnel structure sections. The mean and covariance matrix of the vector angle are calculated, and the definition is X is a random variable, and the sample estimate set of each measuring point is The sensitivity matrix is introduced, and the first-order Taylor expansion is performed on the sample estimate set The sensitivity matrix is used to convert the covariance matrix of the sample estimate set The sensitivity matrix is used as a small error judgment, and the Mahalanobis distance is used as the distance of the vector angle sample to obtain the Mahalanobis distance as the diagnosis factor of the uneven settlement of the tunnel section. 5) The extracted tunnel strain data is processed by the above steps, and the obtained uneven settlement diagnosis factor and the diagnosis threshold of the health state are compared. If the diagnosis factor exceeds the diagnosis threshold, it is considered that the tunnel has uneven settlement. If the diagnosis threshold is not exceeded, it is considered that the tunnel has not occurred uneven settlement.

Images