Monitoring and early warning method for grouting reinforcement construction of roadbed in goaf area with complex multi-boundary conditions

Abstract

The present invention provides a monitoring and early warning method for grouting reinforcement construction of goaf roadbed in complex multi-boundary conditions, including the installation of monitoring points, acquisition of monitoring data, analysis of image monitoring photos, analysis of ground displacement monitoring data and comprehensive analysis. It has better safety and comprehensiveness and can be widely used in the field of roadbed construction monitoring.

Description

Technical Field

[0001] The invention relates to a monitoring and early warning method for grouting reinforcement construction of a goaf area with complex multi-boundary conditions, and is applicable to the field of roadbed construction monitoring. Background Art

[0002] With the development of economy and the continuous advancement of infrastructure construction, the construction demand of roadbed engineering under various complex geological conditions is increasing. Especially in special geological environments such as goaf areas, the roadbed is affected by underground cavities or soft layers, and its stability and bearing capacity are often difficult to meet normal use requirements. In order to ensure the safety and reliability of roadbed engineering, grouting reinforcement technology came into being. By injecting slurry into the foundation, filling the gaps, and enhancing the mechanical properties of the roadbed material, the overall stability of the roadbed is improved.

[0003] However, there are many uncertain factors in the grouting reinforcement construction process, such as the size, shape, and distribution of underground cavities, the selection of grouting materials, and the control of grouting pressure, which may affect the reinforcement effect. In addition, the complex boundary conditions that may be encountered during the construction process, such as adjacent buildings, pipelines, and environmental protection areas, also put forward higher requirements for construction safety. Therefore, traditional construction monitoring methods can no longer meet the needs of current projects, and there is an urgent need to develop more scientific and accurate monitoring and early warning technologies.

[0004] At present, with regard to the safety monitoring of grouting reinforcement construction of goaf areas under complex multi-boundary conditions, the focus is on "point" monitoring, that is, monitoring points are arranged at a certain distance to obtain monitoring data, and the displacement changes of the goaf areas are monitored through monitoring data analysis; however, this method has great disadvantages. If there are large displacement changes or collapses in the goaf areas, and these dangerous areas happen to be outside the monitoring points, the current safety monitoring methods cannot effectively monitor the changes in the goaf areas, and safety accidents are prone to occur, especially in goaf areas under complex multi-boundary conditions. Monitoring points are often not arranged well due to complex boundary conditions; in view of this, we need a more comprehensive and safer safety monitoring method for grouting reinforcement construction of goaf areas under complex multi-boundary conditions. Therefore, we propose a "point-surface combination" monitoring and early warning method for grouting reinforcement construction of goaf areas under complex multi-boundary conditions. Summary of the invention

[0005] The purpose of the present invention is to solve the current urgent need for a more comprehensive and safer safety monitoring method for grouting reinforcement construction of goaf roadbed under complex multi-boundary conditions. Therefore, a monitoring and early warning method for grouting reinforcement construction of goaf roadbed under complex multi-boundary conditions is proposed.

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

[0007] The method for monitoring and early warning of grouting reinforcement construction of goaf roadbed in complex multi-boundary conditions is characterized in that the steps of the method for monitoring and early warning of grouting reinforcement construction of goaf roadbed in complex multi-boundary conditions are as follows:

[0008] S101, installation of monitoring points;

[0009] The installation of the monitoring points includes the installation of ground displacement monitoring points and the installation of image monitoring points. The number of the ground displacement monitoring points is m, and the number of the image monitoring points is n. The image monitoring points are composed of high-definition cameras fixed in a safe position. The ground displacement monitoring points are arranged at the edge of the goaf;

[0010] S102, acquisition of monitoring data;

[0011] The acquisition of the monitoring data includes the acquisition of ground displacement monitoring data and the acquisition of image monitoring photos, wherein the ground displacement monitoring data is acquired based on the ground displacement monitoring points at certain time intervals, and the image monitoring photos are acquired based on the image monitoring points at certain time intervals;

[0012] The acquisition of the monitoring data includes synchronously starting the acquisition of the monitoring data when the grouting work of the roadbed in the goaf area begins;

[0013] S103, analysis of image monitoring photos;

[0014] The analysis of the image monitoring photos includes using the optical flow method to analyze the obtained image monitoring photos to obtain the predicted value of the displacement of the feature points. The steps of using the optical flow method to analyze and obtain the displacement of the feature points are: a) photo preprocessing, preprocessing the obtained image monitoring photos, the preprocessing method is denoising, contrast enhancement and image alignment, b) optical flow calculation, using the Lucas-Kanad optical flow algorithm, selecting obvious feature points in the image, calculating the optical flow of the adjacent photos of the image monitoring photos according to the acquisition time, and obtaining the displacement vector of the feature points, c) recording the displacement data of each feature point at different time points, and generating a displacement time series; d) using a recurrent neural network RNN ​​to construct a prediction model according to the obtained displacement time series to obtain the predicted value of the feature point displacement for the next time interval;

[0015] S104, analysis of ground displacement monitoring data;

[0016] The analysis of the ground displacement monitoring data includes using a long short-term memory neural network to construct a prediction model for the ground displacement monitoring data to obtain a predicted value of the ground displacement monitoring data. The method for obtaining the predicted value of the ground displacement monitoring data comprises the following steps: a) arranging the displacement data and checking for missing values; b) normalizing the data; c) constructing a long short-term memory neural network prediction model; d) using the displacement data to perform model training and model testing on the prediction model to finally obtain a final prediction model; and e) obtaining the predicted value of the ground displacement monitoring data according to the prediction model.

[0017] S105, comprehensive analysis;

[0018] The comprehensive analysis includes formulating characteristic point displacement warning standards and ground displacement monitoring data warning standards based on the actual topographic and geological conditions of the goaf, and issuing a warning when the characteristic point displacement prediction value reaches the characteristic point displacement warning standard or the ground displacement monitoring data prediction value reaches the ground displacement monitoring data warning standard.

[0019] Furthermore, in the above S102, the acquisition of the ground displacement monitoring data includes obtaining the data through observation using a total station or a level.

[0020] Furthermore, in the above S103, the denoising method adopts Baidu AI image denoising method.

[0021] Furthermore, in the above S103, the parameters of the Lucas-Kanade optical flow algorithm are set as follows: the window size is set to (15, 15), the maximum level is set to 1, the maximum number of iterations is set to 20, and the convergence threshold is set to 0.01.

[0022] Furthermore, in S105, the method for obtaining the feature point displacement warning standard is as follows: a) monitoring all the feature points for a certain period of time using the Lucas-Kanade optical flow algorithm to obtain the displacement vector of the feature point; b) calculating the average displacement value s and the standard deviation σ of the feature point according to the displacement vector of the feature point; c) calculating the threshold value θ according to formula (1), and the threshold value θ is the feature point displacement warning standard;

[0023] θ=s+ησ (1)

[0024] Wherein, η is the safety factor which is related to the geological conditions and size of the goaf, and takes a value of 1.5 to 2.0.

[0025] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention adopts high-definition camera technology and advanced image processing technology, and combines it with ground displacement monitoring to more accurately obtain small changes in the road surface, thereby improving the accuracy of monitoring data; appropriate early warning standards are formulated according to the topographic and geological conditions of the goaf, and it has a wide range of applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 : Flow chart of the monitoring and early warning method for grouting reinforcement construction in goaf areas with complex multiple boundary conditions of the present invention.

[0028] Figure 2 : The actual displacement time series values ​​and predicted values ​​of the monitoring points in the embodiment of the present invention.

[0029] Figure 3 : The actual displacement monitoring value and predicted value of the monitoring point of the embodiment of the present invention. DETAILED DESCRIPTION

[0030] 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.

[0031] The following is a specific implementation example of a method for monitoring and early warning of grouting reinforcement construction in goaf areas with complex multi-boundary conditions.

[0032] like Figure 1 As shown, it is a flow chart of the method for monitoring and early warning of grouting reinforcement construction in goaf areas with complex multi-boundary conditions according to the present invention.

[0033] The method for monitoring and early warning of grouting reinforcement construction of goaf roadbed in complex multi-boundary conditions is characterized in that the steps of the method for monitoring and early warning of grouting reinforcement construction of goaf roadbed in complex multi-boundary conditions are as follows:

[0034] S101, installation of monitoring points;

[0035] The installation of the monitoring points includes the installation of ground displacement monitoring points and the installation of image monitoring points. The number of the ground displacement monitoring points is m, and the number of the image monitoring points is n. The image monitoring points are composed of high-definition cameras fixed in a safe position. The ground displacement monitoring points are arranged at the edge of the goaf;

[0036] In this embodiment, the name of the engineering project on which the goaf area relies is a new construction project of the Xingyang to Xinmi section of the Jiaozuo to Pingdingshan Expressway. There are 3 coal seam goaf area treatments within the main line of the project. Among them, the treatment projects with pile numbers K22+331 to K22+476 are equipped with monitoring points, and the monitoring points are equipped with more than 210 ground displacement monitoring points, and the image monitoring points are more than 80;

[0037] S102, acquisition of monitoring data;

[0038] The acquisition of the monitoring data includes the acquisition of ground displacement monitoring data and the acquisition of image monitoring photos, wherein the ground displacement monitoring data is acquired at a certain time interval based on the ground displacement monitoring points, and the image monitoring photos are acquired by taking photos at a certain time interval based on the image monitoring points; the acquisition of the monitoring data includes synchronously starting the acquisition of the monitoring data when the grouting work of the roadbed in the goaf area begins; further, the acquisition of the ground displacement monitoring data includes obtaining the data through observation using a total station or a level;

[0039] In this embodiment, the ground displacement monitoring data is obtained based on the ground displacement monitoring points at a certain time interval, the time interval is set to one day, and the data is collected by total station observation. The performance parameters of the total station used are shown in Table 1; the image monitoring photos are obtained by taking photos based on the image monitoring points at a certain time interval, and the time interval is every 2 hours;

[0040] Table 1 Total Station Performance Parameters

[0041] project parameter Instrument Model Leica TS60 Ultra-High Precision Total Station Accuracy 0.5 seconds Range 3500M Magnification 30X Origin Switzerland Distance measurement accuracy 0.6mm+1ppm

[0042] S103, analysis of image monitoring photos;

[0043] The analysis of the image monitoring photos includes using the optical flow method to analyze the obtained image monitoring photos to obtain the predicted value of the displacement of the feature points. The steps of using the optical flow method to analyze and obtain the displacement of the feature points are as follows: a) photo preprocessing, preprocessing the obtained image monitoring photos, the preprocessing method is denoising, contrast enhancement and image alignment, b) optical flow calculation, using the Lucas-Kanad optical flow algorithm, selecting obvious feature points in the image, calculating the optical flow of the adjacent photos of the image monitoring photos according to the acquisition time, and obtaining the displacement vector of the feature points, c) recording the displacement data of each feature point at different time points to generate a displacement time series; d) using a recurrent neural network RNN ​​to construct a prediction model according to the obtained displacement time series to obtain the predicted value of the displacement of the feature points in the next time interval; further, the denoising method adopts the Baidu AI image denoising method; further, the parameters of the Lucas-Kanad optical flow algorithm are set as follows: the window size is set to (15, 15), the maximum level is set to 1, the maximum number of iterations is set to 20, and the convergence threshold is set to 0.01;

[0044] In this embodiment, taking a certain image monitoring point (P-13) of the K22+348 section of the engineering project as an example, the optical flow method is used to generate a displacement time series, and then the recurrent neural network is combined to generate a predicted value. Finally, within a certain period of time, the actual displacement time series value of the monitoring point and the predicted value are as follows: Figure 2 As shown;

[0045] S104, analysis of ground displacement monitoring data;

[0046] The analysis of the ground displacement monitoring data includes using a long short-term memory neural network to construct a prediction model for the ground displacement monitoring data to obtain a predicted value of the ground displacement monitoring data. The method for obtaining the predicted value of the ground displacement monitoring data comprises the following steps: a) arranging the displacement data and checking for missing values; b) normalizing the data; c) constructing a long short-term memory neural network prediction model; d) using the displacement data to perform model training and model testing on the prediction model to finally obtain a final prediction model; and e) obtaining the predicted value of the ground displacement monitoring data according to the prediction model.

[0047] In this embodiment, a ground displacement monitoring point (P-13) at the K22+355 section of the engineering project is taken as an example to obtain the ground displacement prediction value, and the long short-term memory neural network model is used to generate the prediction value. Finally, within a certain period of time, the actual displacement monitoring value and the prediction value of the monitoring point are as follows: Figure 3 As shown;

[0048] S105, comprehensive analysis;

[0049] The comprehensive analysis includes formulating characteristic point displacement warning standards and ground displacement monitoring data warning standards based on the actual topographic and geological conditions of the goaf, and issuing a warning when the characteristic point displacement prediction value reaches the characteristic point displacement warning standard or the ground displacement monitoring data prediction value reaches the ground displacement monitoring data warning standard.

[0050] Furthermore, the method for obtaining the feature point displacement warning standard is as follows: a) monitoring all the feature points for a certain period of time using the Lucas-Kanad optical flow algorithm to obtain the displacement vector of the feature point; b) calculating the average displacement value of the feature point according to the displacement vector of the feature point and standard deviation σ, c) calculating a threshold value θ according to formula (1), wherein the threshold value θ is the characteristic point displacement warning standard;

[0051]

[0052] Wherein, η is the safety factor which is related to the geological conditions and size of the goaf, and takes a value of 1.5 to 2.0.

[0053] In this embodiment, the Lucas-Kanad optical flow algorithm is performed on all eigenvalues ​​to obtain the displacement vector of the feature point. Then, the average displacement value is calculated to be 0.82 mm, and the standard deviation is 0.26. The threshold is calculated according to Formula 1 as follows:

[0054] θ=0.82+1.5×0.26=1.21mm

[0055] The threshold value calculated according to the above formula is 1.21 mm, which is used as the warning standard for feature point displacement. That is, when the daily change value of the feature point displacement prediction value is less than 1.21 mm, we consider it safe, otherwise an early warning needs to be issued.

[0056] In the above embodiment, the present invention discloses a monitoring and early warning method for grouting reinforcement construction of roadbed in goaf areas with complex multi-boundary conditions, including the installation of monitoring points, acquisition of monitoring data, analysis of image monitoring photos, analysis and comprehensive analysis of ground displacement monitoring data. It has better safety and comprehensiveness and can be widely used in the field of roadbed construction monitoring.

[0057] 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 monitoring and early warning method for grouting reinforcement construction in goaf areas with complex multi-boundary conditions, characterized in that: The steps of the monitoring and early warning method for grouting reinforcement construction in goaf areas with complex multi-boundary conditions are as follows: 1) Installation of monitoring points; The installation of the monitoring points includes the installation of ground displacement monitoring points and the installation of image monitoring points. The number of the ground displacement monitoring points is m, and the number of the image monitoring points is n. The image monitoring points are composed of high-definition cameras fixed in a safe position. The ground displacement monitoring points are arranged at the edge of the goaf; 2) Acquisition of monitoring data; The acquisition of the monitoring data includes the acquisition of ground displacement monitoring data and the acquisition of image monitoring photos, wherein the ground displacement monitoring data is acquired based on the ground displacement monitoring points at certain time intervals, and the image monitoring photos are acquired based on the image monitoring points at certain time intervals; The acquisition of the monitoring data includes synchronously starting the acquisition of the monitoring data when the grouting work of the roadbed in the goaf area begins; 3) Analysis of image monitoring photos; The analysis of the image monitoring photos includes using the optical flow method to analyze the obtained image monitoring photos to obtain the predicted value of the displacement of the feature points. The steps of using the optical flow method to analyze and obtain the displacement of the feature points are: a) photo preprocessing, preprocessing the obtained image monitoring photos, the preprocessing method is denoising, contrast enhancement and image alignment, b) optical flow calculation, using the Lucas-Kanad optical flow algorithm, selecting obvious feature points in the image, calculating the optical flow of the adjacent photos of the image monitoring photos according to the acquisition time, and obtaining the displacement vector of the feature points, c) recording the displacement data of each feature point at different time points, and generating a displacement time series; d) using a recurrent neural network RNN ​​to construct a prediction model according to the obtained displacement time series to obtain the predicted value of the feature point displacement for the next time interval; 4) Analysis of ground displacement monitoring data; The analysis of the ground displacement monitoring data includes using a long short-term memory neural network to construct a prediction model for the ground displacement monitoring data to obtain a predicted value of the ground displacement monitoring data. The method for obtaining the predicted value of the ground displacement monitoring data comprises the following steps: a) arranging the displacement data and checking for missing values; b) normalizing the data; c) constructing a long short-term memory neural network prediction model; d) using the displacement data to perform model training and model testing on the prediction model to finally obtain a final prediction model; and e) obtaining the predicted value of the ground displacement monitoring data according to the prediction model. 5) Comprehensive analysis; The comprehensive analysis includes formulating a characteristic point displacement warning standard and a ground displacement monitoring data warning standard according to the actual terrain and geology of the goaf. When the characteristic point displacement prediction value reaches the characteristic point displacement warning standard or the ground displacement monitoring data prediction value reaches the ground displacement monitoring data warning standard, a warning is issued. The characteristic point displacement warning standard is obtained by: a) monitoring all the characteristic points for a certain period of time using the Lucas-Kanad optical flow algorithm to obtain the displacement vector of the characteristic point; b) calculating the average displacement value of the characteristic point according to the displacement vector of the characteristic point and standard deviation σ, c) calculating a threshold value θ according to formula (1), wherein the threshold value θ is the characteristic point displacement warning standard; Wherein, η is the safety factor which is related to the geological conditions and size of the goaf, and takes a value of 1.5 to 2.

0.

2. The method for monitoring and early warning of grouting reinforcement construction in goaf areas with complex multi-boundary conditions according to claim 1 is characterized in that: In the above 2), the acquisition of the ground displacement monitoring data includes obtaining it through observation using a total station or a level.

3. The method for monitoring and early warning of grouting reinforcement construction in goaf areas with complex multi-boundary conditions according to claim 1 is characterized in that: In the above 3), the denoising method adopts Baidu AI image denoising method.

4. The method for monitoring and early warning of grouting reinforcement construction in goaf areas with complex multi-boundary conditions according to claim 1 is characterized in that: In the above 3), the parameters of the Lucas-Kanad optical flow algorithm are set as follows: the window size is set to (15, 15), the maximum level is set to 1, the maximum number of iterations is set to 20, and the convergence threshold is set to 0.01.

Images