Large semi-filled karst cave plum blossom-shaped reinforcing structure and filling method

Abstract

The invention provides a large semi-filled karst cave plum blossom-shaped reinforcing structure and a filling method, which are suitable for the field of underground engineering karst cave treatment and comprise karst cave geological investigation and modeling, drill hole arrangement design, grouting construction and karst cave filling effect detection. Through a recurrent neural network prediction model fusing a parameterized activation function and a multi-stage dynamic pressure regulation and control system, integration of accurate hole arrangement and positioning, intelligent pressure prediction and self-adaptive regulation and control of large semi-filled karst cave reinforcement construction is realized; the method has remarkable engineering applicability and large-scale popularization value in the field of karst cave treatment of infrastructure construction such as tunnels and bridges.

Description

Technical Field

[0001] The invention relates to a large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and a filling method, which are applicable to the field of underground engineering karst cave treatment. Background Art

[0002] With the rapid development of infrastructure construction in my country, the reinforcement of large semi-filled caves has become a key technical problem when carrying out tunnels, bridges and other projects in karst landform areas. Traditional cave reinforcement methods mostly use disordered grouting or uniform hole distribution, which has problems such as extensive grouting pressure control, uneven distribution of filling materials, and difficulty in quantifying and evaluating the reinforcement effect. In the existing technology, grouting pressure control mainly relies on manual experience adjustment and lacks real-time dynamic prediction capabilities, which leads to slurry crossflow caused by sudden pressure rise or loose filling caused by insufficient pressure during the grouting process. Especially in the plum blossom-shaped hole distribution scenario, due to the mutual influence of the grouting pressure between the center hole and the edge hole, the traditional PID control method is difficult to adapt to the complex working conditions of multiple equipment collaborative operation.

[0003] In recent years, although some studies have attempted to apply machine learning to grouting pressure prediction, there are still problems such as insufficient data adaptability and delayed dynamic response. Therefore, there is an urgent need to develop a cave reinforcement technology system that integrates intelligent prediction and dynamic regulation. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of dynamic regulation of grouting pressure in plum blossom-shaped reinforcement filling of semi-filled karst caves, and propose a large-scale plum blossom-shaped reinforcement structure and filling method for semi-filled karst caves.

[0005] The purpose of the present invention can be achieved by adopting the following technical solutions: A large semi-filled karst cave plum blossom-shaped reinforcement structure and filling method, comprising the following steps: S101 Cave geological survey and modeling; The karst cave geological survey and modeling includes determining the spatial morphology of the karst cave through three-dimensional geological radar scanning and establishing a three-dimensional model of the karst cave; S102 drilling arrangement design; The drilling arrangement design includes arranging grouting drilling holes in a plum blossom shape on the surface of the cave roof according to the three-dimensional model of the cave. The plum blossom-shaped grouting drilling holes include a center hole and adjacent drilling holes. The spacing L between the center hole and the adjacent drilling holes is obtained by formula (1). (1) In the formula, k is the adjustment parameter, with a value of 0.4~0.6, and H is the cave height; S103 grouting construction; The grouting construction includes using a layered grouting technology to perform cave grouting, wherein the layered grouting technology includes dividing the cave grouting work into n grouting sections for layered grouting, wherein n is 3 to 5, and the cave grouting includes a grouting pressure control method to control the cave grouting pressure within a reasonable range; S104 forms a large semi-filled cave plum blossom-shaped reinforcement structure; The forming of a large semi-filled karst cave plum blossom-shaped reinforcement structure comprises using the layered grouting technology to perform karst cave grouting to finally form a large semi-filled karst cave plum blossom-shaped reinforcement structure; S105 cave filling effect detection; The cave filling effect detection includes performing cave filling effect detection on the large semi-filled cave plum blossom-shaped reinforcement structure.

[0006] Furthermore, in the above step S103, the implementation steps of the grouting pressure control method are: a) There are n grouting equipments, marked as S1~S n , in the grouting equipment S i Install the grouting pressure sensor P i , obtain the pressure sensor P at a fixed time interval t i The pressure sensing data P ij , sort out the pressure sensor data P of the pressure sensor Pi for a period of time ij , forming the data set C i ; b) For the dataset C i The data in the data set C is preprocessed, and the preprocessing includes i The missing values ​​of the data in the C are queried and interpolated, and the interpolation of the missing values ​​adopts the cubic spline interpolation method. The preprocessing also includes the data set C i The data in the data set C are normalized. The preprocessing includes normalizing the data set C i The data in is divided into training set and test set; c) constructing a recurrent neural network model prediction model, wherein the recurrent neural network prediction model includes an input layer, a hidden layer and an output layer, wherein the activation function used by the hidden layer is a parameterized hyperbolic tangent function, and the expression of the parameterized hyperbolic tangent function is shown in formula (2): (2) In the formula, It is a trainable parameter with a value range of [0.2, 0.4]; d) Using the data of the training set to carry out model training on the recurrent neural network model prediction model, and using the data of the test set to carry out model testing and optimization on the recurrent neural network model prediction model, and finally obtaining a deployable prediction model; e) Using the prediction model to perform model prediction to obtain pressure sensing data P ij The future prediction value W ij+1 , f) Based on the obtained future prediction value W ij+1 Evaluate the grouting pressure and take corresponding actions to achieve the purpose of controlling the grouting pressure.

[0007] Furthermore, the future prediction value W obtained according to ij+1 The following steps are included to evaluate the grouting pressure and make corresponding treatment actions: a) Based on the obtained future prediction value W ij+1 And get the future prediction value W ij+1 Time T ij+1 First, determine which stage it is in, and use formula (3) to make a judgment: (3) Where, T is the planned grouting time; b) When the judgment result is the initial stage, the future prediction value W is determined by formula (4) ij+1 Whether the initial stage pressure requirements are met, if they are met, no action is taken, if they are not met, an early warning is issued; (4) c) When the judgment result is the final pressure stage, the future prediction value W is determined by formula (5) ij+1 Whether the pressure requirement of the final pressure stage is met, if it is met, no action is taken, if it is not met, an early warning is issued; (5) d) Regardless of the judgment result at any stage, it is necessary to use formula (6) to determine the future prediction value W ij+1 Whether the pressure drop requirements are met, (6) If it meets the requirements, no action will be taken; if it does not meet the requirements, an early warning will be issued.

[0008] Furthermore, in the above step S104, the large semi-filled cave plum blossom-shaped reinforcement structure consists of an unexplored cave boundary portion outside the cave, an explored cave boundary portion inside the cave, a sand, gravel, and soft mud filling portion, a double-slurry stirred mixture grouting curtain wall portion, a cement mortar multiple normal pressure grouting portion, a cement multiple high-pressure filling grouting portion, and a high-pressure grouting portion for filling gaps and cracks to form a sand and gravel vein slurry mixture portion.

[0009] Furthermore, in the above step S105, the method for detecting the filling effect of the cave includes a filling effect calculation method based on weight calculation, and the main steps of the filling effect calculation method based on weight calculation are: a) Divide the filling area into n grids, the size of the grid is the positive direction of the side length a, and the value of a is 5~15m, b) Obtain volume filling rate using geological radar scanning on the nth grid , volume filling rate The calculation formula is shown in formula (7): (7) Where V1 is the void volume after 28 days of grouting, and V2 is the cave volume before grouting; c) Calculate the volume filling rate using the drilling method in the nth grid area , volume filling rate The calculation formula is shown in formula (8): (8) Where L1 is the sum of the lengths of the cement mixture and the voids in the coring section of the borehole, and L2 is the total length of the borehole; d) Calculate the weighted filling rate and use the weighted calculation method to calculate the comprehensive filling rate of the nth grid , the calculation formula is shown in formula (9), (9) f) Filling effect grading judgment, when ≥95%, it is judged as the first-level filling effect, and no further grouting is required. ≥90%, it is judged as the secondary filling effect, and grouting needs to be carried out in combination with the geological radar scanning results. When it is less than 90%, it is judged as the third-level grouting effect, and the grouting plan needs to be redesigned.

[0010] The beneficial effects of the present invention are as follows: (1) Spatially adaptive plum blossom-shaped hole arrangement design: Based on the three-dimensional model of the cave, the borehole spacing is dynamically calculated by the formula L=k⋅H (k=0.4∼0.6), and spatially adaptive hole arrangement is realized in combination with the cave height H. (2) Intelligent layered grouting pressure control system: The parameterized hyperbolic tangent function is innovatively used as the hidden layer activation function of the recurrent neural network. The function adaptively adjusts the nonlinear response intensity through trainable parameters. In the grouting pressure prediction task, the average absolute error is lower than that of the standard tanh function. (3) Dynamic determination of multi-stage pressure thresholds: A three-stage pressure evaluation mechanism is designed to determine the grouting state through the dual dimensions of time and pressure. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a flow chart of a large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method of the present invention; Figure 2 This is a diagram of a plum blossom-shaped reinforcement structure of a large semi-filled karst cave according to an embodiment of the present invention; 1- double liquid grouting area, 2- double grout stirring mixture grouting curtain wall, 3- unexplored cave boundary on the outside, 4- filling materials such as sand, gravel, and soft mud, 5- cement grout grouting, 6- high pressure grouting to fill gaps and cracks, 7- cement mortar grouting, 8- explored cave boundary on the inside; Figure 3 Detailed diagram of the plum blossom-shaped reinforcement structure of an embodiment of the present invention; 9 - cement mortar normal pressure grouting, 10 - cement high pressure filling grouting. DETAILED DESCRIPTION

[0012] 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.

[0013] The following is a specific embodiment of a large semi-filled cave plum blossom-shaped reinforcement structure and filling method.

[0014] S101 Cave geological survey and modeling; The karst cave geological survey and modeling includes determining the spatial morphology of the karst cave through three-dimensional geological radar scanning and establishing a three-dimensional model of the karst cave; S102 drilling arrangement design; The drilling arrangement design includes arranging grouting drilling holes in a plum blossom shape on the surface of the cave roof according to the three-dimensional model of the cave. The plum blossom-shaped grouting drilling holes include a center hole and adjacent drilling holes. The spacing L between the center hole and the adjacent drilling holes is obtained by formula (1). (1) In the formula, k is the adjustment parameter, with a value of 0.4~0.6, and H is the cave height; S103 grouting construction; The grouting construction includes using a layered grouting technology to perform cave grouting, wherein the layered grouting technology includes dividing the cave grouting work into n grouting sections for layered grouting, wherein n is 3 to 5, and the cave grouting includes a grouting pressure control method to control the cave grouting pressure within a reasonable range; Furthermore, in the above step S103, the implementation steps of the grouting pressure control method are: a) There are n grouting equipments, marked as S1~S n , in the grouting equipment S i Install the grouting pressure sensor P i, obtain the pressure sensor P at a fixed time interval t i The pressure sensing data P ij , sort out the pressure sensor data P of the pressure sensor Pi for a period of time ij , forming the data set C i ; b) For the dataset C i The data in the data set C is preprocessed, and the preprocessing includes i The missing values ​​of the data in the C are queried and interpolated, and the interpolation of the missing values ​​adopts the cubic spline interpolation method. The preprocessing also includes the data set C i The data in the data set C are normalized. The preprocessing includes normalizing the data set C i The data in is divided into training set and test set; c) constructing a recurrent neural network model prediction model, wherein the recurrent neural network prediction model includes an input layer, a hidden layer and an output layer, wherein the activation function used by the hidden layer is a parameterized hyperbolic tangent function, and the expression of the parameterized hyperbolic tangent function is shown in formula (2): (2) In the formula, It is a trainable parameter with a value range of [0.2, 0.4]; d) Using the data of the training set to carry out model training on the recurrent neural network model prediction model, and using the data of the test set to carry out model testing and optimization on the recurrent neural network model prediction model, and finally obtaining a deployable prediction model; e) Using the prediction model to perform model prediction to obtain pressure sensing data P ij The future prediction value W ij+1 , f) Based on the obtained future prediction value W ij+1 Evaluate the grouting pressure and take corresponding actions to achieve the purpose of controlling the grouting pressure.

[0015] Furthermore, the future prediction value W obtained according to ij+1 The following steps are included to evaluate the grouting pressure and make corresponding treatment actions: a) Based on the obtained future prediction value W ij+1 And get the future prediction value W ij+1 Time T ij+1 First, determine which stage it is in, and use formula (3) to make a judgment: (3) Where, T is the planned grouting time; b) When the judgment result is the initial stage, the future prediction value W is determined by formula (4)ij+1 Whether the initial stage pressure requirements are met, if they are met, no action is taken, if they are not met, an early warning is issued; (4) c) When the judgment result is the final pressure stage, the future prediction value W is determined by formula (5) ij+1 Whether the pressure requirement of the final pressure stage is met, if it is met, no action is taken, if it is not met, an early warning is issued; (5) d) Regardless of the judgment result at any stage, it is necessary to use formula (6) to determine the future prediction value W ij+1 Whether the pressure drop requirements are met, (6) If it meets the requirements, no action will be taken; if it does not meet the requirements, an early warning will be issued.

[0016] S104 forms a large semi-filled cave plum blossom-shaped reinforcement structure; The forming of a large semi-filled karst cave plum blossom-shaped reinforcement structure comprises using the layered grouting technology to perform karst cave grouting to finally form a large semi-filled karst cave plum blossom-shaped reinforcement structure; Furthermore, in the above step S104, the large semi-filled cave plum blossom-shaped reinforcement structure consists of an unexplored cave boundary portion outside the cave, an explored cave boundary portion inside the cave, a sand, gravel, and soft mud filling portion, a double-slurry stirred mixture grouting curtain wall portion, a cement mortar multiple normal pressure grouting portion, a cement multiple high-pressure filling grouting portion, and a high-pressure grouting portion for filling gaps and cracks to form a sand and gravel vein slurry mixture portion.

[0017] In this embodiment, the large semi-filled cave plum blossom-shaped reinforcement structure is formed as follows: Figure 2 The details of the plum blossom-shaped reinforcement structure formed at the same time are shown in Figure 3 As shown, the composition of this reinforced structure is: the outer side of the cave has not been explored, the inner side of the cave has been explored, filling materials such as sand, gravel, and mud, double slurry stirring mixture grouting curtain wall, multiple normal pressure grouting of cement mortar, multiple high-pressure filling grouting of cement, high-pressure grouting to fill gaps and cracks to form a sand and gravel vein slurry mixture, etc.

[0018] At the same time, the composition advantages of this structure are: plum blossom grouting can achieve full grouting, cement mortar common grouting and cement slurry high-pressure grouting are combined, so that sand, mud, water and poisonous gas in the cave are stirred through grouting, water and other toxic gases are removed, mud and cement slurry react chemically to form a vein-like mixture, sand and gravel and cement slurry form a concrete-like mixture, so as to achieve the purpose of filling and reinforcing large caves.

[0019] S105 cave filling effect detection; The cave filling effect detection includes performing cave filling effect detection on the large semi-filled cave plum blossom-shaped reinforcement structure.

[0020] Furthermore, in the above step S105, the method for detecting the filling effect of the cave includes a filling effect calculation method based on weight calculation, and the main steps of the filling effect calculation method based on weight calculation are: a) Divide the filling area into n grids, the size of the grid is the positive direction of the side length a, and the value of a is 5~15m, b) Obtain volume filling rate using geological radar scanning on the nth grid , volume filling rate The calculation formula is shown in formula (7): (7) Where V1 is the void volume after 28 days of grouting, and V2 is the cave volume before grouting; c) Calculate the volume filling rate using the drilling method in the nth grid area , volume filling rate The calculation formula is shown in formula (8): (8) Where L1 is the sum of the lengths of the cement mixture and the voids in the coring section of the borehole, and L2 is the total length of the borehole; d) Calculate the weighted filling rate and use the weighted calculation method to calculate the comprehensive filling rate of the nth grid , the calculation formula is shown in formula (9), (9) f) Filling effect grading judgment, when ≥95%, it is judged as the first-level filling effect, and no further grouting is required. ≥90%, it is judged as the secondary filling effect, and grouting needs to be carried out in combination with the geological radar scanning results. When it is less than 90%, it is judged as the third-level grouting effect, and the grouting plan needs to be redesigned.

[0021] In the above embodiment, the present invention discloses a large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method; including karst cave geological survey and modeling, drilling layout design, grouting construction and karst cave filling effect detection. By integrating the recurrent neural network prediction model with parameterized activation function and the multi-stage dynamic pressure control system, the present invention realizes the integration of precise hole layout positioning, intelligent pressure prediction and adaptive control in the reinforcement construction of large-scale semi-filled karst caves, and has significant engineering applicability and large-scale promotion value in the field of karst cave treatment in infrastructure construction such as tunnels and bridges.

[0022] 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 principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A large semi-filled karst cave plum blossom-shaped reinforcement structure and filling method, characterized in that: The following steps are involved: S101 Cave geological survey and modeling; The karst cave geological survey and modeling includes determining the spatial morphology of the karst cave through three-dimensional geological radar scanning and establishing a three-dimensional model of the karst cave; S102 drilling arrangement design; The drilling arrangement design includes arranging grouting drilling holes in a plum blossom shape on the surface of the cave roof according to the three-dimensional model of the cave. The plum blossom-shaped grouting drilling holes include a center hole and adjacent drilling holes. The spacing L between the center hole and the adjacent drilling holes is obtained by formula (1). （1） In the formula, k is the adjustment parameter, with a value of 0.4~0.6, and H is the cave height; S103 grouting construction; The grouting construction includes using a layered grouting technology to perform cave grouting, wherein the layered grouting technology includes dividing the cave grouting work into n grouting sections for layered grouting, wherein n is 3 to 5, and the cave grouting includes a grouting pressure control method to control the cave grouting pressure within a reasonable range; S104 forms a large semi-filled cave plum blossom-shaped reinforcement structure; The forming of a large semi-filled karst cave plum blossom-shaped reinforcement structure comprises using the layered grouting technology to perform grouting in the cave to finally form a large semi-filled karst cave plum blossom-shaped reinforcement structure; S105 cave filling effect detection; The cave filling effect detection includes performing cave filling effect detection on the large semi-filled cave plum blossom-shaped reinforcement structure.

2. A large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method according to claim 1, characterized in that: In step S103, the implementation steps of the grouting pressure control method are: a) There are n grouting equipments, marked as S1~S n , in the grouting equipment S i Install the grouting pressure sensor P i , obtain the pressure sensor P at a fixed time interval t i The pressure sensing data P ij , sort out the pressure sensor data P of the pressure sensor Pi for a period of time ij , forming the data set C i ; b) For the dataset C i The data in the data set C is preprocessed, and the preprocessing includes i The missing values ​​of the data in the C are queried and interpolated, and the interpolation of the missing values ​​adopts the cubic spline interpolation method. The preprocessing also includes the data set C i The data in the data set C are normalized. The preprocessing includes normalizing the data set C i The data in is divided into training set and test set; c) constructing a recurrent neural network model prediction model, wherein the recurrent neural network prediction model includes an input layer, a hidden layer and an output layer, wherein the activation function used by the hidden layer is a parameterized hyperbolic tangent function, and the expression of the parameterized hyperbolic tangent function is shown in formula (2): (2) In the formula, It is a trainable parameter with a value range of [0.2, 0.4]; d) Using the data of the training set to carry out model training on the recurrent neural network model prediction model, and using the data of the test set to carry out model testing and optimization on the recurrent neural network model prediction model, and finally obtaining a deployable prediction model; e) Using the prediction model to perform model prediction to obtain pressure sensing data P ij The future prediction value W ij+1 , f) Based on the obtained future prediction value W ij+1 Evaluate the grouting pressure and take corresponding actions to achieve the purpose of controlling the grouting pressure.

3. A large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method according to claim 2, characterized in that: The future prediction value W obtained according to ij+1 The following steps are included to evaluate the grouting pressure and make corresponding treatment actions: a) Based on the obtained future prediction value W ij+1 And get the future prediction value W ij+1 Time T ij+1 First, determine which stage it is in, and use formula (3) to make a judgment: (3) Where T is the planned grouting time; b) When the judgment result is the initial stage, the future prediction value W is determined by formula (4) ij+1 Whether the initial stage pressure requirements are met, if they are met, no action is taken, if not, an early warning is issued; （4） c) When the judgment result is the final pressure stage, the future prediction value W is determined by formula (5) ij+1 Whether the pressure requirement of the final pressure stage is met, if it is met, no action is taken, if it is not met, an early warning is issued; （5） d) Regardless of the judgment result at any stage, it is necessary to use formula (6) to determine the future prediction value W ij+1 Whether the pressure drop requirements are met, （6） If it meets the requirements, no action will be taken; if it does not meet the requirements, an early warning will be issued.

4. A large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method according to claim 1, characterized in that: In step S104, the large semi-filled cave plum blossom-shaped reinforcement structure consists of an unexplored cave boundary portion outside the cave, an explored cave boundary portion inside the cave, a sand, gravel, and soft mud filling portion, a double-slurry stirring mixture grouting curtain wall portion, a cement mortar multiple normal pressure grouting portion, a cement multiple high-pressure filling grouting portion, and a high-pressure grouting to fill gaps and cracks to form a sand and gravel vein slurry mixture portion.

5. A large-scale semi-filled karst cave plum blossom-shaped reinforcement structure and filling method according to claim 1, characterized in that: In step S105, the method for detecting the filling effect of the cave includes a filling effect calculation method based on weight calculation, and the main steps of the filling effect calculation method based on weight calculation are: a) Divide the filling area into n grids, the size of the grid is the positive direction of the side length a, and the value of a is 5~15m, b) Obtain volume filling rate using geological radar scanning on the nth grid , volume filling rate The calculation formula is shown in formula (7): (7) where V1 is the void volume 28 days after grouting, and V2 is the cave volume before grouting; c) Calculate the volume filling rate using the drilling method in the nth grid area , volume filling rate The calculation formula is shown in formula (8): (8) where L1 is the sum of the lengths of the cement mixture and the voids in the coring section of the borehole, and L2 is the total length of the borehole; d) Calculate the weighted filling rate and use the weighted calculation method to calculate the comprehensive filling rate of the nth grid , the calculation formula is shown in formula (9), （9） f) Filling effect grading judgment, when ≥95%, it is judged as the first-level filling effect, and no further grouting is required. ≥90%, it is judged as the secondary filling effect, and grouting needs to be carried out in combination with the geological radar scanning results. When it is less than 90%, it is judged as the third-level grouting effect, and the grouting plan needs to be redesigned.

Images