A method and system for fine prediction of rock burst grade in drill-and-blast tunnel construction stage

Abstract

The present application relates to the field of tunnel engineering, and specifically discloses a method and system for fine prediction of rock burst grade in tunnel construction stage by drill and blast method, comprising the following steps: constructing a basic sample library; establishing a mapping relationship between drilling parameters and first principal stress, second principal stress, third principal stress, elastic modulus and Poisson's ratio and other surrounding rock mechanical parameters based on energy method; constructing an extended sample library based on the mapping relationship and the basic sample library; constructing an intelligent rock burst grade prediction model based on deep learning, and bringing the extended sample library into the intelligent rock burst grade prediction model for training; and fine prediction of rock burst grade in front of the tunnel face according to the constructed intelligent rock burst grade prediction model. The present application can effectively reduce the input of construction personnel and prevent the impact of rock burst on the safety of construction personnel, and guide the intelligent construction of tunnels.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tunnel intelligent construction, in particular to a fine prediction method and system for rock burst grade in tunnel construction stage by drill and blast method. BACKGROUND

[0002] With the gradual increase of tunnel depth and cross section size, the problem of tunnel rock burst is increasingly prominent, which seriously affects the safety and efficient construction of the tunnel. At the same time, with the gradual improvement of the data collection capability of tunnel construction machinery, the full computer three-arm drill rig has the functions of real-time collection and transmission of drilling parameters, and the artificial intelligence technology has shown strong robustness in the fields of massive data analysis and intelligent prediction.

[0003] Due to the randomness and uncertainty of the geological conditions of underground engineering, the rock burst grade predicted in the geological exploration stage is greatly different from the actual rock burst grade of the working face surrounding rock in the construction stage. The traditional rock burst prediction method in the construction stage of the drill and blast method consumes a lot of manpower and physical resources, and it is difficult to meet the current demand for intelligent construction of tunnels. For example, by using artificial intelligence technologies such as drilling parameters, machine learning and deep learning, the rock burst grade of the surrounding rock in front of the working face can be quickly and finely predicted, and the rock burst prone area in front of the working face can be warned. According to the prediction results, the construction scheme and the supporting structure parameter design are adjusted in time, which will greatly ensure the safety of intelligent construction of tunnels.

[0004] Therefore, there is an urgent need for a method that can finely predict the rock burst grade of the surrounding rock in front of the working face. SUMMARY

[0005] To solve the problems in the prior art, the present application provides a fine prediction method for rock burst grade in tunnel construction stage by drill and blast method, which can finely predict the rock burst grade of the surrounding rock in front of the working face and solve the problems mentioned in the background.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a fine prediction method for rock burst grade in tunnel construction stage by drill and blast method, comprising the following steps:

[0007] S1, constructing a basic sample library, the basic sample library including drilling parameters, rock hardness and corresponding rock burst grade of the working face and other information;

[0008] S2, establishing a mapping relationship between drilling parameters and first principal stress, second principal stress, third principal stress, elastic modulus and Poisson's ratio of the surrounding rock mechanical parameters based on energy method;

[0009] S3, constructing an extended sample library based on the mapping relationship between drilling parameters and surrounding rock mechanical parameters and the basic sample library;

[0010] S4, construct a deep learning-based rock burst grade intelligent prediction model, and bring the expanded sample library into the rock burst grade intelligent prediction model for training;

[0011] S5, according to the constructed rock burst grade intelligent prediction model, finely predict the rock burst grade of the surrounding rock in front of the working face.

[0012] Preferably, the step S1 constructs a basic sample library, and according to the automatic acquisition of each borehole feeding speed, impact pressure, pushing pressure, and rotation pressure raw data of the full-computer three-arm rock drilling rig, the average value of each drilling parameter is used as the representative value of the drilling parameter of the target working face; the rock hardness and the rock burst grade of the working face are determined through the working face geological sketch, the rock hardness includes five types of extremely hard rock, hard rock, relatively soft rock, soft rock, and extremely soft rock, and the rock burst grade includes five types of no rock burst, slight rock burst, medium rock burst, strong rock burst, and extremely strong rock burst.

[0013] Preferably, the step S2 establishes a mapping relationship between the drilling parameters and the first principal stress, the second principal stress, the third principal stress, the elastic modulus, and the Poisson's ratio and other surrounding rock mechanical parameters based on the energy method, and specifically includes the following:

[0014] S21, according to the impact-pushing-rotation rock breaking mode of the full-computer three-arm rock drilling rig, the elastic deformation energy of the surrounding rock is calculated;

[0015] S22, based on the energy method and numerical simulation or laboratory test method, the nominal first principal stress, the nominal second principal stress, the nominal third principal stress, the nominal elastic modulus, and the nominal Poisson's ratio and other surrounding rock mechanical parameters in each borehole unit area of the target working face are calculated;

[0016] S23, according to the nominal first principal stress, the nominal second principal stress, the nominal third principal stress, the nominal elastic modulus, and the nominal Poisson's ratio in each borehole unit area of the target working face, the first principal stress, the second principal stress, the third principal stress, the elastic modulus, and the Poisson's ratio of the target working face are calculated.

[0017] Preferably, the step S3 constructs an expanded sample library based on the mapping relationship between the drilling parameters and the surrounding rock mechanical parameters and the basic sample library.

[0018] Preferably, the deep learning-based rock burst grade intelligent prediction model constructed in the step S4 has 6 layers, i.e. 1 input layer, 4 hidden layers, and 1 output layer; and the bringing of the expanded sample library into the rock burst grade intelligent prediction model for training means that the average value of the feeding speed, the impact pressure, the pushing pressure, and the rotation pressure of the target working face, the rock hardness, the first principal stress, the second principal stress, the third principal stress, the elastic modulus, and the Poisson's ratio are used as the input parameters of the input layer, and the rock burst grade corresponding to the working face is used as the output parameter of the output layer.

[0019] Preferably, the step S5 finely predicts the rock burst grade of the surrounding rock in front of the tunnel face according to the constructed rock burst grade intelligent prediction model, and specifically includes the following:

[0020] S51, according to the spatial coordinate information of the blast hole, block in two-dimensional plane, along the tunneling direction with 20cm granularity, block on the basis of longitudinal segmentation;

[0021] S52, according to the two-dimensional block and longitudinal segmentation results of the surrounding rock of the tunnel face, using the rock burst grade intelligent prediction model, finely predicting the rock burst grade of the surrounding rock in front of the tunnel face.

[0022] Preferably, the step S21 specifically includes: the unit volume of rock mass unit is deformed under the action of machinery, it is assumed that there is no heat exchange with the outside world in this process, the energy output by the drill jumbo in the impact-advancing-rotating rock breaking process is all converted into the internal energy of the rock mass, mainly including the dissipation energy of the surrounding rock and the elastic deformation energy of the surrounding rock, then

[0023]

[0024] In the formula: K is the conversion coefficient of the energy output by the drill jumbo converted into the elastic deformation energy of the surrounding rock, which is determined according to the numerical simulation or laboratory test results; P c is the impact oil cylinder inlet oil pressure monitored by the pressure sensor, Pa; P t is the impact oil cylinder inlet oil pressure monitored by the pressure sensor, Pa; V d is the flowmeter monitored drill bit feed speed, m / s; D cA is the impact oil cylinder piston rear diameter, m; D cB is the impact oil cylinder piston front diameter, m; S c is the piston design stroke, m; m c is the impact piston mass, kg; D is the drill hole diameter, m; P t is the rear cavity pressure of the piston advancing oil cylinder, which is also the advancing oil cylinder inlet pressure, that is, the monitored drilling parameter advancing pressure, Pa; D t is the advancing oil cylinder piston diameter, unit m; P r is the inlet pressure of the motor rotation, that is, the monitored drilling parameter rotation pressure, Pa; V r is the drilling tool rotation speed, unit r / s; q r is the hydraulic motor displacement, that is, the flow rate discharged by the motor per rotation, unit ml / r or cc; i r is the rotation motor reduction ratio; σ1 is the first principal stress, Pa; σ2 is the second principal stress, Pa; σ3 is the third principal stress, Pa; E is the rock mass elastic modulus, Pa; υ is the rock mass Poisson's ratio.

[0025] Preferably, the step S22 specifically comprises: adopting numerical simulation or indoor test method to simulate the drilling process of the full computer three-arm rock drilling jumbo under different rock burst grades, and determine the conversion coefficient K of the energy conversion from the mechanical output to the elastic deformation energy of the surrounding rock; grid the blast hole in the drilling direction with 8 cm as a unit; the full computer three-arm rock drilling jumbo can automatically collect 5 groups of drilling parameters in the 8 cm area, and the nominal first principal stress, the nominal second principal stress, the nominal third principal stress, the nominal elastic modulus and the nominal Poisson's ratio of each unit area of the target blast hole are analyzed by using the 5 groups of drilling parameters;

[0026] The analytic formulas of the nominal first principal stress, the nominal second principal stress, the nominal third principal stress, the nominal elastic modulus and the nominal Poisson's ratio of each unit area of the blast hole are as follows:

[0027]

[0028] In the formula, P t,i is the starting advancing pressure in the unit area in the drilling direction, Pa; P r,i is the rotation pressure recorded by the i number data point in the unit area in the drilling direction, Pa; P c,i is the impact pressure recorded by the i number data point in the unit area in the drilling direction, Pa; V d,i is the feeding speed recorded by the i number data point in the unit area in the drilling direction, m / s; P t,i+1 is the advancing pressure recorded by the i+1 number data point in the unit area in the drilling direction, Pa; P r,i+1 is the rotation pressure recorded by the i+1 number data point in the unit area in the drilling direction, Pa; P c,i+1 is the impact pressure recorded by the i+1 number data point in the unit area in the drilling direction, Pa; V d,i+1 is the feeding speed recorded by the i+1 number data point in the unit area in the drilling direction, m / s; P t,i+2 is the advancing pressure recorded by the i+2 number data point in the unit area in the drilling direction, Pa; P r,i+2 is the rotation pressure recorded by the i+2 number data point in the unit area in the drilling direction, Pa; P c,i+2 is the impact pressure recorded by the i+2 number data point in the unit area in the drilling direction, Pa; V d,i+2 is the feeding speed recorded by the i+2 number data point in the unit area in the drilling direction, m / s; P t,i+3 is the advancing pressure recorded by the i+3 number data point in the unit area in the drilling direction, Pa; P r,i+3 is the rotation pressure recorded by the i+3 number data point in the unit area in the drilling direction, Pa; P c,i+3 is the impact pressure recorded by the i+3 number data point in the unit area in the drilling direction, Pa; V d,i+3is the feed speed recorded at the i+3 data point in the unit area along the drilling direction, m / s; P t,i+4 is the thrust pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; P r,i+4 is the rotation pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; P c,i+4 is the impact pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; V d,i+4 is the feed speed recorded at the i+4 data point in the unit area along the drilling direction, m / s; σ 1,j,i is the first principal stress of the surrounding rock in the i blasthole area along the drilling direction of the j blasthole, Pa; σ 2,j,i is the second principal stress of the surrounding rock in the i blasthole area along the drilling direction of the j blasthole, Pa; σ 3,j,i is the third principal stress of the surrounding rock in the i blasthole area along the drilling direction of the j blasthole, Pa; E j,i is the elastic modulus of the surrounding rock in the i blasthole area along the drilling direction of the j blasthole, Pa; v j,i is the Poisson's ratio of the surrounding rock in the i blasthole area along the drilling direction of the j blasthole.

[0029] Preferably, the step S23 specifically comprises: on the basis of the analysis result of the mechanical parameters of the surrounding rock in the unit area of the blasthole, performing weighted average calculation on the initial ground stress analysis result of the surrounding rock in each unit area of the single blasthole, and taking the weighted average calculation result as the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock of the working face, and the specific calculation formula is:

[0030]

[0031] In the formula: σ1 is the first principal stress of the target working face, Pa; σ2 is the second principal stress of the target working face, Pa; σ3 is the third principal stress of the target working face, Pa; E is the elastic modulus of the surrounding rock of the target working face, Pa; v is the Poisson's ratio of the surrounding rock of the target working face; m is the number of blastholes of the target working face; and n is the number of region division of the blasthole.

[0032] Preferably, the step S51 specifically comprises: according to the section form and the blasthole arrangement, dividing the working face into 5 layers and 4 columns, a total of 18 blocks, with 2m as the unit; and further segmenting the surrounding rock in front of the working face longitudinally based on the two-dimensional segmentation, with 20 cm as the granularity, according to the interval of the hole support structure.

[0033] Preferably, the step S52 specifically comprises: assuming that the rock hardness in each region (M block, N section) in front of the working face does not change significantly in any blasting cycle, the rock hardness in each region (M block, N section) in front of the working face is consistent with the geological profile result of the working face; according to the three-dimensional division rule of the surrounding rock in front of the working face, the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each region (M block, N section) are calculated respectively according to the drilling parameters; and the rockburst grade in front of the working face is finely predicted by using the rockburst grade intelligent prediction model.

[0034] In addition, in order to achieve the above-mentioned purpose, the present application also provides the following technical scheme: a rockburst grade fine prediction system for drill-and-blast method tunnel construction stage, the system specifically comprises:

[0035] The basic sample library construction module: a basic sample library is constructed, and the basic sample library includes drilling parameters, rock hardness and rockburst grade information corresponding to the working face;

[0036] The mapping relationship establishment module: the mapping relationship between the drilling parameters and the surrounding rock mechanical parameters of the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio is established based on the energy method;

[0037] The extended sample library construction module: based on the mapping relationship between the drilling parameters and the surrounding rock mechanical parameters, the extended sample library is constructed based on the basic sample library;

[0038] The rockburst grade intelligent prediction model construction module: a rockburst grade intelligent prediction model based on deep learning is constructed, and the extended sample library is brought into the rockburst grade intelligent prediction model for training;

[0039] The prediction module: according to the constructed rockburst grade intelligent prediction model, the rockburst grade in front of the working face is finely predicted.

[0040] The present application has the beneficial effects that: the present application finely predicts the rockburst grade in front of the working face according to the drilling parameters generated in the drilling process of the drill-and-blast method tunnel jumbo and the rock hardness of the working face geological profile, and the position of the rockburst in front of the working face can be further inferred according to the fine prediction result. The present application patent ensures the safety of the drill-and-blast method tunnel construction, reduces the investment of manpower and material resources in the rockburst prediction process, and helps the intelligent construction of the drill-and-blast method tunnel. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 It is the flowchart of the rockburst grade fine prediction method for drill-and-blast method tunnel construction stage in embodiment 1;

[0042] Figure 2 It is the schematic diagram of the blast hole unit division and drilling parameter recording rule in embodiment 1;

[0043] Figure 3 Figure for the deep learning-based rock burst grade intelligent prediction model in Example 1;

[0044] Figure 4 Figure for the two-dimensional block diagram of the working face in Example 1;

[0045] Figure 5 Figure for the longitudinal segmentation rule in Example 1;

[0046] Figure 6 Figure for the fine prediction result of the rock burst grade of the working face in front of H3DK2+879.3 in Example 1;

[0047] Figure 7 Figure for the comparative analysis details of the rock burst prediction result and the photo of the rock burst working face of H3DK2+877.5 in Example 1, (a) is the photo of the working face, and (b) is the rock burst prediction result;

[0048] Figure 8 Figure for the fine prediction system of the rock burst grade during the drilling and blasting tunnel construction stage in Example 2;

[0049] In the figure, 110 is a basic sample library construction module, 120 is a mapping relationship establishment module, 130 is an extended sample library construction module, 140 is a rock burst grade intelligent prediction model construction module, and 150 is a prediction module. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0051] Example 1

[0052] With the increase of the buried depth and length of the tunnel construction site, the rock burst problem of the tunnel is increasingly prominent, which directly affects the safety of the drilling and blasting tunnel construction. The prediction result in the geological exploration stage is greatly different from the actual situation in the construction stage. The traditional seismic monitoring method, rock mass stress monitoring method, rock mass sound wave monitoring method, and rock mass vibration monitoring method in the construction stage need to invest a lot of manpower and material resources, and it is difficult to guarantee the safety of the monitoring personnel in the monitoring process.

[0053] To this end, the inventors propose, through long-term research and practice, a method for fine prediction of rock burst grade in tunnel construction stage by drilling and blasting method, which uses the drilling parameters collected by the full-computer three-arm rock drilling rig, the rock hardness degree recorded by the geological sketch of the tunnel face, and the artificial intelligence technology to fine predict the rock burst grade of the surrounding rock in front of the tunnel face. The method aims to provide a fast and convenient rock burst prediction method to ensure the safety of tunnel construction and improve the level of intelligent construction of tunnels.

[0054] The embodiment provides a technical scheme: a method for fine prediction of rock burst grade in tunnel construction stage by drilling and blasting method, and a step flowchart is shown in Figure 1 The method specifically comprises the following steps:

[0055] Step S1: Construct a basic sample library, which includes drilling parameters, rock hardness degree, and rock burst grade corresponding to the tunnel face, etc. The full-computer three-arm rock drilling rig automatically collects the original data of the hole feeding speed, impact pressure, pushing pressure, and rotation pressure, and uses the average value of each drilling parameter as the representative value of the drilling parameter of the target tunnel face. The drilling tool rotation speed is usually 269 r / min. The rock hardness degree and the rock burst grade of the tunnel face are determined by the geological sketch of the tunnel face. The rock hardness degree includes five types of extremely hard rock, hard rock, relatively soft rock, soft rock, and extremely soft rock. The rock burst grade includes five types of no rock burst, slight rock burst, moderate rock burst, strong rock burst, and extremely strong rock burst. The data of the constructed basic sample library is shown in Table 1.

[0056] Table 1: Data example of the basic sample library

[0057]

[0058] Step S2: Establish the mapping relationship between the drilling parameters and the first principal stress, the second principal stress, the third principal stress, the elastic modulus, and the Poisson's ratio, etc. based on the energy method.

[0059] 1) According to the impact-pushing-rotation rock breaking mode of the full-computer three-arm rock drilling rig, the elastic deformation energy of the surrounding rock is calculated. The unit volume of rock mass unit is deformed under the action of the machine. It is assumed that there is no heat exchange with the outside world in this process. The energy output by the rock drilling rig in the impact-pushing-rotation rock breaking process is all converted into internal energy in the rock mass, mainly including the dissipation energy of the surrounding rock and the elastic deformation energy of the surrounding rock. Then,

[0060]

[0061] In the formula, K is the conversion coefficient of the energy output by the rock drilling rig converted into the elastic deformation energy of the surrounding rock, which is determined according to the results of numerical simulation or laboratory test; P c is the impact oil cylinder inlet oil pressure monitored by the pressure sensor, Pa; P t is the impact oil cylinder inlet oil pressure monitored by the pressure sensor, Pa; Vd It is the drill bit feed rate monitored by the flow meter, in m / s; D cA It is the diameter of the piston in the impact cylinder, in meters (m); D cB It is the diameter of the front end of the impact cylinder piston, in meters (m); S c It refers to the piston design stroke, in meters (m). c D is the mass of the impact piston (kg); D is the borehole diameter (m); P t D is the pressure in the rear chamber of the piston propulsion cylinder, which is also the pressure at the inlet of the propulsion cylinder, i.e., the propulsion pressure, a monitored drilling parameter, in Pa. t It refers to the diameter of the hydraulic cylinder piston, in meters (m); P r The inlet pressure is the pressure at the oil inlet when the motor rotates, i.e., the rotary pressure of the monitored drilling parameters, in Pa; V r q is the drill bit rotation speed, measured in r / s; r The displacement of the hydraulic motor is the flow rate discharged per revolution of the motor, measured in ml / r or cc; r σ is the reduction ratio of the rotary motor; σ1 is the first principal stress, Pa; σ2 is the second principal stress, Pa; σ3 is the third principal stress, Pa; E is the elastic modulus of the rock mass, Pa; υ is the Poisson's ratio of the rock mass.

[0062] 2) Based on energy methods and numerical simulation or laboratory testing, calculate the nominal first principal stress, nominal second principal stress, nominal third principal stress, nominal elastic modulus, and nominal Poisson's ratio, among other surrounding rock mechanical parameters, within each borehole unit region of the target face. Using numerical simulation or laboratory testing, simulate the drilling process of a fully computerized three-arm drilling rig under different rockburst levels to determine the conversion coefficient K between the mechanical output energy and the elastic deformation of the surrounding rock. In this embodiment, the conversion coefficient determined using laboratory testing is 0.6.

[0063] like Figure 2 As shown, the boreholes are gridded in 8cm units along the drilling direction. Within the 8cm area, the fully computerized three-arm drilling rig can automatically collect 5 sets of drilling parameters. These 5 sets of drilling parameters are used to analyze the nominal first principal stress, nominal second principal stress, nominal third principal stress, nominal elastic modulus, and nominal Poisson's ratio of the surrounding rock in each unit area of ​​the target borehole.

[0064] The analytical formulas for the nominal first principal stress, nominal second principal stress, nominal third principal stress, nominal elastic modulus, and nominal Poisson's ratio of the surrounding rock in each unit area of ​​the borehole are as follows:

[0065]

[0066] In the formula: P t,i It is the initial thrust pressure within the unit region along the drilling direction, in Pa; P r,i It is the rotary pressure recorded at data point i in unit region along the drilling direction, in Pa; P c,iis the recorded push pressure of the unit region along the drilling direction, Pa; P d,i is the recorded feed speed of the unit region along the drilling direction, m / s; P t,i+1 is the recorded push pressure of the unit region along the drilling direction, Pa; P r,i+1 is the recorded rotation pressure of the unit region along the drilling direction, Pa; P c,i+1 is the recorded push pressure of the unit region along the drilling direction, Pa; V d,i+1 is the recorded feed speed of the unit region along the drilling direction, m / s; P t,i+2 is the recorded push pressure of the unit region along the drilling direction, Pa; P r,i+2 is the recorded rotation pressure of the unit region along the drilling direction, Pa; P c,i+2 is the recorded push pressure of the unit region along the drilling direction, Pa; V d,i+2 is the recorded feed speed of the unit region along the drilling direction, m / s; P t,i+3 is the recorded push pressure of the unit region along the drilling direction, Pa; P r,i+3 is the recorded rotation pressure of the unit region along the drilling direction, Pa; P c,i+3 is the recorded push pressure of the unit region along the drilling direction, Pa; V d,i+3 is the recorded feed speed of the unit region along the drilling direction, m / s; P t,i+4 is the recorded push pressure of the unit region along the drilling direction, Pa; P r,i+4 is the recorded rotation pressure of the unit region along the drilling direction, Pa; P c,i+4 is the recorded push pressure of the unit region along the drilling direction, Pa; V d,i+4 is the recorded feed speed of the unit region along the drilling direction, m / s; σ 1,j,i is the nominal first principal stress of the surrounding rock in the i-th blasthole region along the drilling direction of the j-th blasthole, Pa; σ 2,j,i is the nominal second principal stress of the surrounding rock in the i-th blasthole region along the drilling direction of the j-th blasthole, Pa; σ 3,j,i is the nominal third principal stress of the surrounding rock in the i-th blasthole region along the drilling direction of the j-th blasthole, Pa; E j,i is the nominal elastic modulus of the surrounding rock in the i-th blasthole region along the drilling direction of the j-th blasthole, Pa; v j,i is the nominal Poisson's ratio of the surrounding rock in the i-th blasthole region along the drilling direction of the j-th blasthole.

[0067] 3) On the basis of the analytical results of the mechanical parameters of the surrounding rock in the blasthole unit area, the initial ground stress of each unit area of the surrounding rock of a single blasthole is calculated by weighted average, and the weighted average calculation results are used as the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock of the working face, and the specific calculation formula is:

[0068]

[0069] In the formula: σ1 is the first principal stress of the target working face, Pa; σ2 is the second principal stress of the target working face, Pa; σ3 is the third principal stress of the target working face, Pa; E is the elastic modulus of the surrounding rock of the target working face, Pa; v is the Poisson's ratio of the surrounding rock of the target working face; m is the number of blastholes of the target working face; and n is the number of regional divisions of the blastholes.

[0070] Taking a drill-and-blast tunnel project as an example, the calculation results of the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock of each typical working face are shown in Table 2.

[0071] Table 2 Analytical results of the surrounding rock parameters of each typical working face of a drill-and-blast tunnel project

[0072]

[0073] Step S3: According to the mapping relationship between the drilling parameters and the mechanical parameters of the surrounding rock, an extended sample library is constructed on the basis of the basic sample library.

[0074] The extended sample library stores, in units of working face, the average values of four drilling parameters of the target working face, the calculation values of five mechanical parameters of the surrounding rock, i.e., the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio, and two geological information, i.e., the rock hardness and the rock burst grade of the geological sketch record of the working face. The details of the information recorded in the extended sample library are shown in Table 3.

[0075] Table 3 Details of the information recorded in the extended sample library

[0076]

[0077] Step S4: Construct a deep learning-based rockburst grade intelligent prediction model, and bring the expanded sample library into the rockburst grade intelligent prediction model for training. The deep learning-based rockburst grade intelligent prediction model network layer is 6 layers, that is, 1 input layer, 4 hidden layers and 1 output layer; the bringing of the expanded sample library into the rockburst grade intelligent prediction model for training refers to that the average values of the four drilling parameters of the target face feeding speed, impact pressure, advance pressure and rotation pressure, the rock hardness, the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio are taken as the input parameters of the input layer, and the rockburst grade corresponding to the face is taken as the output parameter of the output layer. Please refer to Figure 3 .

[0078] In the sample library construction process, in order to ensure the accuracy of rockburst grade prediction, the expanded sample library constructed should contain different rockburst grade characteristics, and the data quantity should not be less than 500 groups, that is, 100 groups for each of the five rockburst grades of no rockburst, slight rockburst, moderate rockburst, strong rockburst and extremely strong rockburst. The samples of each rockburst grade in the sample library are brought into the model for training in the proportion of training set: prediction set = 8:2, and the accuracy rate of each rockburst grade in the prediction set is > 85%, as shown in Table 4.

[0079] Table 4 Prediction accuracy rate of each rockburst grade in the prediction set

[0080] Serial number Rock burst grade Accuracy rate / % 1 No rock burst 87.6 2 Mild rock burst 88.6 3 Medium rock burst 89.1 4 Strong rock burst 90.5 5 Extremely strong rock burst 92.3

[0081] Step S5: According to the constructed rockburst grade intelligent prediction model, the rockburst grade of the surrounding rock in front of the face is finely predicted. Specifically, the following steps are included:

[0082] S51, according to the spatial coordinate information of the blast hole, the face is divided into blocks on a two-dimensional plane, and the blocks are segmented longitudinally along the tunneling direction with a granularity of 20 cm;

[0083] S52, according to the two-dimensional block and longitudinal segmentation results of the face surrounding rock, the rockburst grade intelligent prediction model is used to finely predict the rockburst grade of the surrounding rock in front of the face.

[0084] 1) According to the section form and blast hole arrangement, the face is divided into 5 layers and 4 columns with a unit of 2 m, a total of 18 blocks. Taking a typical drill-and-blast method face section as an example, please refer to the two-dimensional block diagram of the face Figure 4 .

[0085] According to the spacing of the tunnel support structure, the surrounding rock in front of the face is further segmented longitudinally on the basis of two-dimensional block division with a granularity of 20 cm. Please refer to the longitudinal segmentation rule diagram Figure 5 .

[0086] 2) Assuming that the hardness of the rock does not change significantly within any blasting cycle, the hardness of the rock in each region (M block, N section) in front of the working face is consistent with the geological profile of the working face; according to the three-dimensional division rule of the surrounding rock in front of the working face, the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each region (M block, N section) are calculated according to the drilling parameters; the rockburst grade intelligent prediction model is used to finely predict the rockburst grade of the surrounding rock in front of the working face.

[0087] Taking a typical drill-and-blast method project as an example, during the drilling process of H3DK2+879.3~H3DK2+876.9, the fine prediction results of the rockburst grade of the surrounding rock in front of the working face H3DK2+879.3 are shown in Figure 6 . A slight rockburst occurred on the left side of the H3DK2+877.5 working face, and the surface of the surrounding rock burst and fell off. The fine prediction results of the rockburst grade are basically consistent with the actual situation of the rockburst of the working face. For details of the comparison and analysis of the H3DK2+877.5 rockburst working face photo and the rockburst prediction results, please refer to Figure 7 .

[0088] Example 2

[0089] A fine prediction system for rockburst grade in drill-and-blast method tunnel construction stage, as shown in Figure 8 , the system specifically comprises:

[0090] The basic sample library construction module 110: constructs a basic sample library, and the basic sample library includes drilling parameters, rock hardness and rockburst grade information of the corresponding working face;

[0091] The mapping relationship establishment module 120: establishes the mapping relationship between the drilling parameters and the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock mechanical parameters based on the energy method;

[0092] The extended sample library construction module 130: based on the mapping relationship between the drilling parameters and the surrounding rock mechanical parameters, the extended sample library is constructed based on the basic sample library;

[0093] The rockburst grade intelligent prediction model construction module 140: constructs a rockburst grade intelligent prediction model based on deep learning, and brings the extended sample library into the rockburst grade intelligent prediction model for training;

[0094] The prediction module 150: according to the constructed rockburst grade intelligent prediction model, the rockburst grade in front of the working face is finely predicted.

[0095] The present application can finely predict the rock burst grade in front of the tunnel face by using the drilling parameters collected by the full-computer three-arm rock drilling jumbo, the rock hardness degree recorded by the geological sketch of the tunnel face, and the artificial intelligence technology. The drilling parameters generated in the drilling process of the drilling and blasting tunnel face are used to finely predict the rock burst grade in front of the tunnel face, which can effectively reduce the input of construction personnel, prevent the influence of rock burst on the safety of construction personnel, and guide the intelligent construction of the tunnel.

[0096] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can modify the technical solutions recorded in the foregoing embodiments or make equivalent replacements for part of the technical features, and any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for refined prediction of rockburst severity during the drill-and-blast tunnel construction stage, characterized in that, The method comprises the following steps: S1, constructing a basic sample library, the basic sample library comprising drilling parameters, rock hardness and corresponding rockburst grade information of a working face; S2, establishing a mapping relationship between the drilling parameters and the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock based on the energy method; specifically comprising the following: S21, calculating the elastic deformation energy of the surrounding rock according to the impact-pushing-rotation rock breaking mode of the full-computer three-arm drill rig; S22, calculating the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each blasthole unit area of the target working face based on the energy method and numerical simulation or laboratory test method; specifically comprising: The numerical simulation or laboratory test method is used to simulate the drilling process of the full-computer three-arm drill rig under different rockburst grades, and the conversion coefficient K of the energy output by the machine into the elastic deformation energy of the surrounding rock is determined; the blasthole is gridded in the drilling direction with 8 cm as a unit; in the 8 cm area, the full-computer three-arm drill rig can automatically collect 5 groups of drilling parameters, and the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each unit area of the target blasthole are analyzed by using the 5 groups of drilling parameters; The analytical formulas of the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each unit area of the blasthole are as follows: In the formula: K is the conversion coefficient of the energy output by the rock drilling jumbo converted into the elastic deformation energy of the surrounding rock, which is determined according to the numerical simulation or laboratory test results; P c is the oil pressure at the inlet of the impact oil cylinder monitored by the pressure sensor, Pa; P t is the oil pressure at the inlet of the impact oil cylinder monitored by the pressure sensor, Pa; V d is the feed speed of the drill bit monitored by the flowmeter, m / s; D cA is the rear diameter of the impact oil cylinder piston, m; D cB is the front diameter of the impact oil cylinder piston, m; S c is the designed stroke of the piston, m; m c is the mass of the impact piston, kg; D is the diameter of the drill hole, m; P t is the rear cavity pressure of the piston push cylinder, which is also the inlet pressure of the push cylinder, that is, the monitored drilling parameter push pressure, Pa; D t is the diameter of the push cylinder piston, m; P r is the inlet pressure of the motor when rotating, that is, the monitored drilling parameter rotation pressure, Pa; V r is the rotation speed of the drilling tool, r / s; q r is the displacement of the hydraulic motor, that is, the flow rate discharged by the motor per revolution, ml / r or cc; P t,i is the push pressure at the starting point in the unit area along the drilling direction, Pa; P r,i is the rotation pressure recorded at the data point No. i in the unit area along the drilling direction, Pa; P c,i is the impact pressure recorded at the data point No. i in the unit area along the drilling direction, Pa; V d,i is the feed speed recorded at the data point No. i in the unit area along the drilling direction, m / s; P t,i+1 is the push pressure recorded at the data point No. i+1 in the unit area along the drilling direction, Pa; P r,i+1 is the rotation pressure recorded at the data point No. i+1 in the unit area along the drilling direction, Pa; P c,i+1 is the impact pressure recorded at the data point No. i+1 in the unit area along the drilling direction, Pa; V d,i+1 is the feed speed recorded at the data point No. i+1 in the unit area along the drilling direction, m / s; P t,i+2 is the push pressure recorded at the data point No. i+2 in the unit area along the drilling direction, Pa; P r,i+2 is the rotation pressure recorded at the data point No. i+2 in the unit area along the drilling direction, Pa; P c,i+2 is the impact pressure recorded at the data point No. i+2 in the unit area along the drilling direction, Pa; V d,i+2 is the feed speed recorded at the data point No. i+2 in the unit area along the drilling direction, m / s; P t,i+3 is the push pressure recorded at the i+3 data point in the unit area along the drilling direction, Pa; P r,i+3 is the rotation pressure recorded at the i+3 data point in the unit area along the drilling direction, Pa; P c,i+3 is the impact pressure recorded at the i+3 data point in the unit area along the drilling direction, Pa; V d,i+3 is the feed speed recorded at the i+3 data point in the unit area along the drilling direction, m / s; P t,i+4 is the push pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; P r,i+4 is the rotation pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; P c,i+4 is the impact pressure recorded at the i+4 data point in the unit area along the drilling direction, Pa; V d,i+4 is the feed speed recorded at the i+4 data point in the unit area along the drilling direction, m / s; σ 1,j,i is the first principal stress of the surrounding rock in the i-hole area in the drilling direction of the j-hole, Pa; σ 2,j,i is the second principal stress of the surrounding rock in the i-hole area in the drilling direction of the j-hole, Pa; σ 3,j,i is the third principal stress of the surrounding rock in the i-hole area in the drilling direction of the j-hole, Pa; E j,i is the elastic modulus of the surrounding rock in the i-hole area in the drilling direction of the j-hole, Pa; v j,i is the Poisson's ratio of the surrounding rock in the i-hole area in the drilling direction of the j-hole; S23, calculating the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the target working face according to the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock in each unit area of the target working face; S3, constructing an extended sample library based on the mapping relationship between the drilling parameters and the mechanical parameters of the surrounding rock and the basic sample library; S4, constructing an intelligent rockburst grade prediction model based on deep learning, and bringing the extended sample library into the intelligent rockburst grade prediction model for training; S5, finely predicting the rockburst grade of the surrounding rock in front of the working face according to the constructed intelligent rockburst grade prediction model.

2. The method according to claim 1, wherein the method is characterized by: The step S1 constructs the basic sample library, the average values of each drilling parameter are used as the representative values of the drilling parameters of the target working face according to the raw data of the feed speed, impact pressure, pushing pressure and rotation pressure of each blasthole collected by the full-computer three-arm drill rig; the rock hardness and the rockburst grade of the working face are determined through the geological sketch of the working face, the rock hardness comprises five types of extremely hard rock, hard rock, relatively soft rock, soft rock and extremely soft rock, and the rockburst grade comprises five types of no rockburst, slight rockburst, medium rockburst, strong rockburst and extremely strong rockburst.

3. The method according to claim 1, wherein the method is characterized by: The step S21 specifically comprises: a unit volume of rock mass unit produces deformation under the action of a machine, it is assumed that there is no heat exchange with the outside world in this process, the energy output by the drill rig in the impact-pushing-rotation rock breaking process is all converted into internal energy in the rock mass, mainly including the dissipation energy of the surrounding rock and the elastic deformation energy of the rock mass, then In the formula: K is the conversion coefficient of the energy output by the rock drilling rig into the elastic deformation energy of the surrounding rock, which is determined based on numerical simulation or laboratory test results; P c The pressure sensor monitors the oil pressure at the inlet of the impact cylinder, in Pa; P t It is the oil pressure at the inlet of the impact cylinder monitored by the pressure sensor, in Pa; V d It is the drill bit feed rate monitored by the flow meter, in m / s; D cA It is the diameter of the piston in the impact cylinder, in meters (m); D cB It is the diameter of the front end of the impact cylinder piston, in meters (m); S c It refers to the piston design stroke, in meters (m). c D is the mass of the impact piston (kg); D is the borehole diameter (m); P t D is the pressure in the rear chamber of the piston propulsion cylinder, which is also the pressure at the inlet of the propulsion cylinder, i.e., the propulsion pressure, a monitored drilling parameter, in Pa. t It refers to the diameter of the hydraulic cylinder piston, in meters (m); P r The inlet pressure is the pressure at the oil inlet when the motor rotates, i.e., the rotary pressure of the monitored drilling parameters, in Pa; V r q is the drill bit rotation speed, measured in r / s; r The displacement of the hydraulic motor is the flow rate discharged per revolution of the motor, measured in ml / r or cc; r σ is the reduction ratio of the rotary motor; σ1 is the first principal stress, Pa; σ2 is the second principal stress, Pa; σ3 is the third principal stress, Pa; E is the elastic modulus of the rock mass, Pa; υ is the Poisson's ratio of the rock mass.

4. The method according to claim 1, wherein the method is characterized by: The step S23 specifically comprises: on the basis of the analysis result of the surrounding rock mechanical parameters of the borehole unit area, performing weighted average calculation on the initial ground stress analysis result of each unit area surrounding rock of the single borehole, and taking the weighted average calculation result as the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio of the surrounding rock of the tunnel face, and the specific calculation formula is: In the formula, σ1 is the first principal stress of the target tunnel face, Pa; σ2 is the second principal stress of the target tunnel face, Pa; σ3 is the third principal stress of the target tunnel face, Pa; E is the elastic modulus of the target tunnel face surrounding rock, Pa; v is the Poisson's ratio of the target tunnel face surrounding rock; m is the number of boreholes of the target tunnel face; and n is the number of regional division of the borehole.

5. The method according to claim 1, wherein the method is characterized by: The number of layers of the rock burst grade intelligent prediction model network constructed in the step S4 is 6 layers, namely 1 input layer, 4 hidden layers and 1 output layer. The step S4 constructs the rock burst grade intelligent prediction model network based on deep learning, and the number of layers of the rock burst grade intelligent prediction model network is 6 layers, namely 1 input layer, 4 hidden layers and 1 output layer. 6.The method according to claim 1, characterized in that: The step S4 constructs the rock burst grade intelligent prediction model network based on deep learning, and the number of layers of the rock burst grade intelligent prediction model network is 6 layers, namely 1 input layer, 4 hidden layers and 1 output layer. The step S5 specifically comprises: according to the cross section form and the borehole arrangement, taking 2 m as a unit, the tunnel face is divided into 5 layers and 4 columns, and a total of 18 blocks; and on the basis of the two-dimensional block, the surrounding rock in front of the tunnel face is further segmented in the longitudinal direction with a granularity of 20 cm according to the interval of the hole body support structure. The system specifically comprises:

7. The method according to claim 6, characterized in that: The basic sample library construction module (110) constructs a basic sample library, and the basic sample library comprises drilling parameters, rock hardness and rock burst grade information corresponding to the tunnel face; 8. A rock burst grade refined prediction system for a drill-and-blast tunnel construction stage according to the rock burst grade refined prediction method of any one of claims 1-7, characterized in that: The mapping relationship establishment module (120) establishes a mapping relationship between the drilling parameters and the surrounding rock mechanical parameters of the first principal stress, the second principal stress, the third principal stress, the elastic modulus and the Poisson's ratio based on the energy method; The extended sample library construction module (130) constructs an extended sample library based on the mapping relationship between the drilling parameters and the surrounding rock mechanical parameters on the basis of the basic sample library; The rock burst grade intelligent prediction model construction module (140) constructs a rock burst grade intelligent prediction model based on deep learning, and the extended sample library is brought into the rock burst grade intelligent prediction model for training; The prediction module (150) finely predicts the rock burst grade of the surrounding rock in front of the tunnel face according to the constructed rock burst grade intelligent prediction model. ​ ​

Images