Overlying rock type rock burst risk assessment method based on while-drilling multi-source logging data fusion

Abstract

A while-drilling multi-source logging data fusion overlying strata type rock burst risk assessment method comprises the steps that drilling is conducted in a roadway at the interval of 5-50 m, and drilling parameters such as torque, thrust, drilling speed and rotating speed and natural gamma, resistivity and sound wave velocity curves are obtained while drilling; the impact danger levels near different drilling positions are judged in combination with monitoring data such as microseism, stress and support resistance, and the corresponding relation between multi-source logging characteristic parameters and the impact danger levels is established through continuously accumulated drilling multi-source logging information and data of the impact danger levels near the corresponding positions. According to the method, the rock burst risk level near the drilling position in the next construction step is predicted, so that while-drilling real-time identification of the overlying strata type rock burst risk is realized, the hysteresis quality and low efficiency of risk assessment after traditional coring testing are avoided, and the disaster prevention and control timeliness is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of rock burst risk assessment, and specifically to a method for assessing the risk of overburden rock burst by fusing multi-source logging data while drilling. Background Technique

[0002] The rock burst disasters induced by hard roof are a very dangerous and complex dynamic disaster. This is because the overburden structure in coal mines is complex and variable, and is still regarded as a black box structure. The complexity and variability of the overburden structure make the strong mine tremors and rock burst disasters in coal mines extremely sudden, seriously threatening the safe production of coal mines.

[0003] The current risk assessment of overburden rock burst is as follows: based on the data of the rock layer structure of one or several limited geological exploration boreholes near the coal mining face, using theoretical analysis such as roof characteristic parameters and key layer calculation, or adopting numerical simulation methods to evaluate. However, the overburden structures in most areas of the actual working face vary greatly, and many parameters need to be simplified and empirically valued in the theoretical and numerical simulation methods. Therefore, the current risk assessment of overburden rock burst has poor reliability, large suddenness of overburden-induced rock burst disasters, and is difficult to predict and prevent. At the same time, in order to improve the exploration accuracy, some mines drill a large number of core holes in the gateway of the working face, and then bring the rock samples back to the laboratory for processing and testing to obtain the mechanical parameters of each borehole and each rock layer, and conduct empirical evaluation according to the strength and thickness of the thick and hard rock layers. This method has a large amount of work and high cost. Therefore, there is currently a lack of a method that can evaluate the risk of overburden rock burst in real time, economically and accurately underground. Summary of the Invention

[0004] The purpose of the present invention is to solve the problems of poor reliability, difficult to determine the impact risk in most areas, and large suddenness of disasters caused by the current overburden rock burst risk assessment relying on numerical simulation, theoretical calculation and limited borehole rock layer structure data, and to invent a method for assessing the risk of overburden rock burst by fusing multi-source logging data while drilling. It drills holes at intervals of 5m to 50m in the roadway, obtains drilling parameters such as torque, thrust, drilling speed, and rotational speed, as well as natural gamma, resistivity, and acoustic velocity curves while drilling, combines monitoring data such as microseismic, stress, and support resistance to judge the impact hazard levels near different borehole positions, and establishes the corresponding relationship between multi-source logging characteristic parameters and impact hazard levels through the continuously accumulated multi-source logging information of boreholes and the data of impact hazard levels near the corresponding positions, so as to predict the impact rock burst hazard level near the next construction borehole position, thereby realizing real-time, economic and accurate risk assessment of overburden rock burst.

[0005] The technical solution of the present invention is as follows:

[0006] A method for risk assessment of overburden rockburst during drilling, characterized by comprising the following steps:

[0007] Step 1: Conduct borehole logging at intervals of 5m to 50m in the coal mine roadway to obtain data such as drilling vibration velocity, drilling speed, rotational speed, torque, thrust, natural gamma, resistivity, and acoustic velocity curves, and record the borehole position coordinates, drilling time, and displacement;

[0008] Step 2: During the working face mining process, combine one or more of microseismic monitoring, stress on-line monitoring, support resistance monitoring, ground sound, and drill cuttings monitoring, and combine with the spatial coordinates of the borehole position to establish a discrimination criterion to determine the actual rockburst risk level near each borehole position of the working face. The actual risk level R s is divided into three levels: no danger, yellow warning, and red warning;

[0009] Step 3: Obtain the rock stratum strength through borehole sampling and laboratory mechanical parameter determination, and use mathematical statistics or machine learning methods to establish the corresponding relationship E i = f(T, n, F, v, S v , Y1, Y2, Y3), where: E is the rock stratum strength, σ2 is the confining pressure, T is the torque, n is the rotational speed, F is the thrust, v is the drilling speed, S v is the vibration velocity, Y1 is the natural gamma, Y2 is the resistivity, and Y3 is the acoustic velocity. According to the change of E i at the interface of two rock strata, obtain the rock stratum thickness M i , and according to the cumulative thickness of the drilled rock strata, obtain the distance L i from the current rock stratum to the coal seam, and establish the corresponding relationship between the strength, thickness, and distance from the coal seam of the hard rock stratum and the rockburst risk level as follows:

[0010]

[0011] Note: m i ‘ is the effective thickness of the i-th hard rock stratum, M i ‘ = a×M i , S1, S2 are critical values; k1, k2 are constants;

[0012] A total of at least 100m of rock strata above the coal seam are taken, and each rock stratum greater than 60Mpa is calculated according to the above table. According to the comparison of the results with S1 and S2, select the risk level. When there are multiple hard rock strata, add up the results of the risk levels, and the highest level is the red warning.

[0013] Correspond the actual rockburst risk level R s obtained from the actual monitoring in Step 2 to R p, modify the values of a, S1, S2, k1, and k2 so that the judgment formula can better predict the risk level of overburden rock burst.

[0014] Or use machine learning methods to directly establish the corresponding relationship R between the multi-source logging curve characteristics and the rock burst risk level monitored in Step 2 p = f(T, n, F, v, S v , Y1, Y2, Y3); The machine learning methods include but are not limited to convolutional neural network (CNN), long short-term memory network (LSTM), and support vector machine (SVM);

[0015] Step 4: According to the multi-source logging data of the newly drilled borehole, predict the rock burst risk level of the new borehole area through the corresponding relationship between the multi-source logging characteristic parameters and the rock burst risk level in Step 3;

[0016] Step 5: When the working face is mined to the position of the new borehole, judge the current rock burst risk level according to the actually monitored microseismic, in-situ stress, support resistance, rock noise, and drill cuttings data, and compare it with the prediction result. If there is an error, then combine the new data to correct the corresponding relationship between the multi-source logging characteristic parameters and the rock burst risk level.

[0017] The discrimination criteria in Step 2 are as follows:

[0018] (1) Microseismic monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is judged as a yellow warning; when the energy and frequency change rates exceed the set red threshold, it is judged as a red warning;

[0019] (2) In-situ stress online monitoring data: When the daily stress value and stress change rate exceed the set yellow threshold, it is judged as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is judged as a red warning;

[0020] (3) Rock noise monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is judged as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is judged as a red warning;

[0021] (4) Support resistance data: When the frequency of the daily resistance value exceeding 90% of the rated working resistance is greater than 10%, it is judged as a yellow warning; when the frequency of the resistance value exceeding 90% of the rated working resistance is greater than 30%, it is judged as a red warning;

[0022] (5) Except when rock bursts occur and are directly judged as red warnings, when the daily energy of microseismic monitoring exceeds the set yellow threshold and is directly judged as a yellow warning, and when the daily energy of microseismic monitoring exceeds the set red threshold and is directly judged as a red warning, it is necessary to combine microseismic, in-situ stress, rock noise, and support resistance monitoring to comprehensively judge the rock burst risk level. Only when two or more monitoring methods both judge a warning can the warning be confirmed, and the comprehensive warning level is the maximum level determined by the above warning methods.

[0023] (6) The initial values of the above yellow and red warning thresholds are determined based on the rock burst monitoring data and occurrence conditions of working faces under similar conditions; during the actual mining and excavation of the working face, the warning thresholds should be revised according to the maximum microseismic energy level and the occurrence of rock bursts.

[0024] During the judgment process of the discrimination criteria in step 2, it is necessary to comprehensively judge whether the rock burst risk is caused by overlying strata according to the microseismic positioning height and support resistance. If so, then this warning is judged as a rock burst warning of the overlying strata type.

[0025] In step 3, a takes an initial value of 0.1 - 3.0, k1 takes an initial value of 0.1 - 1.0, k2 takes an initial value of 0.2 - 3.0, S1 takes an initial value of 10 - 30, and S2 takes an initial value of 20 - 50.

[0026] The drilling speed, rotation speed, drilling time, displacement, torque, and thrust in step 1 are obtained from the drill equipped with sensors or the near-bit measurement sub; the drilling vibration velocity, natural gamma, resistivity, and acoustic velocity are obtained during drilling from the near-bit measurement sub, or obtained by an existing borehole logging analyzer after drilling.

[0027] The beneficial effects of the present invention are:

[0028] By inventing a risk assessment method for overlying strata type rock bursts with real-time fusion of multi-source logging data while drilling, the present invention realizes real-time identification of the risk of overlying strata type rock bursts while drilling, solves the problems of difficult exploration of overlying strata structures and difficult prevention and control of disasters in traditional methods. It avoids the lag and inefficiency of traditional core sampling tests followed by risk assessment, and can significantly improve the timeliness of disaster prevention and control. Based on this invention, a closed-loop workflow of "data acquisition while drilling - real-time risk assessment - optimization of roof cutting measures" can be established in the mine, improving the intelligent level and efficiency of disaster prevention and control of overlying strata type rock bursts. The present invention has high innovation and practicality, and can significantly improve the safety of coal mining. Brief Description of the Drawings

[0029] Figure 1 is a schematic diagram of borehole logging carried out at intervals of 5m - 50m in a coal mine underground roadway.

[0030] Figure 2 is a schematic diagram of multi-source logging data.

[0031] Figure 3 It is a schematic diagram of the AI model structure, showing the combination method of the convolutional neural network (CNN), long short-term memory network (LSTM) and support vector machine (SVM).

[0032] Figure 4 It is a schematic diagram of the impact hazard level prediction at different drilling positions, showing the corresponding relationship between multi-source logging characteristic parameters and the impact hazard level, and predicting the impact hazard level at different drilling positions. Specific implementation manners

[0033] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0034] Embodiment 1:

[0035] A method for evaluating the risk of overburden rockburst during drilling with multi-source logging data fusion, the specific steps are as follows:

[0036] Step 1: As shown in Figure 1 (in Figure 1 1 is the roadway, 2 is the drill rig, 3 is the borehole), borehole logging is carried out at intervals of 5m to 50m in the coal mine roadway, and the borehole depth is greater than 30m; drilling vibration velocity, drilling speed, rotational speed, torque, thrust, and natural gamma, resistivity, and acoustic velocity curve data are obtained (as shown in Figure 2 ), and the borehole position coordinates, drilling time, and displacement are recorded.

[0037] Step 2: During the working face mining process, through the combination of one or more of microseismic monitoring, stress on-line monitoring, support resistance monitoring, ground noise, and drill cuttings, combined with the spatial coordinates of the borehole position, a discrimination criterion is established to determine the actual rockburst hazard level near each borehole position of the working face. The actual hazard level R s is divided into three levels: no danger, yellow warning, and red warning. The discrimination criteria are as follows:

[0038] (1) Microseismic monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is determined as a yellow warning; when the energy and frequency change rates exceed the set red threshold, it is determined as a red warning;

[0039] (2) Stress on-line monitoring data: When the daily stress value and stress change rate exceed the set yellow threshold, it is determined as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is determined as a red warning;

[0040] (3) Ground noise monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is determined as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is determined as a red warning;

[0041] (4) Support resistance data: When the frequency of the daily resistance value exceeding 90% of the rated working resistance is greater than 10%, it is determined as a yellow warning; when the frequency of the resistance value exceeding 90% of the rated working resistance is greater than 30%, it is determined as a red warning.

[0042] (5) Except that when rock burst occurs, it is directly determined as a red warning, when the daily energy of microseismic monitoring exceeds the set yellow threshold, it is directly determined as a yellow warning, and when the daily energy of microseismic monitoring exceeds the set red threshold, it is directly determined as a red warning. It is necessary to combine microseismic, in-situ stress online, rock noise, and support resistance monitoring to comprehensively judge the impact risk level. Only when two or more monitoring methods both judge a warning can the warning be confirmed, and the comprehensive warning level is the maximum level determined by the above warning methods.

[0043] (6) The above yellow and red warning thresholds are determined based on the initial values according to the impact ground pressure monitoring data and occurrence conditions of working faces under similar conditions; during the actual mining and excavation of the working face, the warning thresholds should be revised according to the maximum energy level of microseismic and the occurrence of impacts.

[0044] The above warning thresholds are determined based on the initial values according to the impact ground pressure monitoring data and occurrence conditions of working faces under similar conditions; during the actual mining and excavation of the working face, the warning thresholds should be revised according to the maximum energy level of microseismic and the occurrence of impacts to more accurately determine the impact risk level.

[0045] For example, when using one or more of microseismic, in-situ stress online, rock noise, and support resistance monitoring, the following judgment criteria are implemented: 1) When the maximum energy of a certain microseismic monitoring working face is greater than the yellow threshold of 1E4 J, it is determined as a yellow warning; when the maximum energy of a certain working face is greater than the red threshold of 1E5 J, it is determined as a red warning; when the total frequency or total energy within 24 h increases by more than 2 times the average total frequency or total energy of the previous 5 days, it is a yellow warning, and when it exceeds 3 times the average total frequency or total energy of the previous 5 days, it is a red warning; 2) For in-situ stress online monitoring, when the stress value of the shallow measuring point is greater than 10 Mpa, it is a yellow warning, and when it is greater than 12 MPa, it is a red warning; when the stress value of the deep measuring point is greater than 12 Mpa, it is a yellow warning, and when it is greater than 14 MPa, it is a red warning; when the stress value increment within 24 h reaches 2 MPa, it is a yellow warning, and when the increment reaches 4 MPa, it is a red warning; 3) When the change rate of rock noise energy and frequency exceeds 2 times, it is a yellow warning, and when it exceeds 3 times, it is a red warning; 4) When the frequency of the support resistance value exceeding 90% of the rated working resistance is greater than 10%, it is determined as a yellow warning; when the frequency of the resistance value exceeding 90% of the rated working resistance is greater than 30%, it is determined as a red warning; 5) When rock burst occurs, it is directly determined as a red warning.

[0046] During the judgment process, it is necessary to comprehensively judge whether the impact ground pressure risk is caused by overlying strata according to support resistance, microseismic positioning height, and the structural conditions near the borehole. If so, record the multi-source logging data of the borehole and the impact ground pressure risk level data.

[0047] Step 3: Obtain the rock stratum strength through drilling sampling and laboratory mechanical parameter determination, and establish the corresponding relationship E between the two using mathematical statistics or machine learning methods i = f(T, n, F, v, S v , Y1, Y2, Y3), where: E is the rock stratum strength, σ2 is the confining pressure, T is the torque, n is the rotational speed, F is the thrust, v is the drilling speed, S v is the vibration velocity, Y1 is the natural gamma ray, Y2 is the resistivity, and Y3 is the acoustic wave velocity. Based on the change of E i at the interface of two rock strata, obtain the rock stratum thickness M i , and based on the cumulative thickness of the drilled rock strata, obtain the distance L from the current rock stratum to the coal seam i , and establish the corresponding relationship between the strength, thickness, and distance from the coal seam of the hard rock stratum and the impact hazard level as follows:

[0048]

[0049] Note: M‘ i is the effective thickness of the i-th layer of hard rock stratum, M‘ i = a × M i , S1 and S2 are critical values; k1 and k2 are constants;

[0050] A total of at least 100 m of rock strata above the coal seam are taken, and each rock stratum with a strength greater than 60 Mpa is calculated according to the above table. Based on the comparison of the results with S1 and S2, select the hazard level. When there are multiple layers of hard rock strata, accumulate the results of the hazard levels, and the highest level is a red warning.

[0051] Correspond the impact hazard level R s actually monitored in Step 2 to R p , and modify the values of a, S1, S2, k1, and k2 so that the judgment formula can better predict the impact hazard level of overlying rock strata

[0052] Characterized in that, the initial value of k1 is taken as 0.1 - 1.0, the initial value of k2 is taken as 0.2 - 3.0, the initial value of S1 is 10 - 30, and the initial value of S2 is taken as 20 - 50.

[0053] For example: It is judged that the strength of the first layer of hard rock stratum M1 is greater than 60 Mpa, the thickness is 13 m, and the distance from the coal seam is 10 m according to the vibration velocity and the peak value of the natural gamma ray curve. The initial values of a, k1, and S1 are taken as 1.0, 0.3, and 10 respectively, and calculate M i - k1L i = 13 - 0.3 × 10 = 10 ≤ S1, therefore, the value is 0. The strength of the second layer of hard rock stratum M2 is greater than 60 Mpa, the thickness is 30 m, and the distance from the coal seam is 25 m. The initial values of k1, k2, S1, and S2 are taken as 0.3, 0.5, 10, and 20 respectively, and calculate:

[0054]

[0055] Therefore, it is determined that the corresponding danger value of the hard rock layer M2 is 1, and a yellow warning is issued.

[0056] At this time, if the strength of the third hard rock layer M3 is greater than 60 Mpa, the thickness is 35 m, and the distance from the coal seam is 60 m, and the initial values of k1, k2, S1, and S2 are taken as 0.3, 0.5, 10, and 20 respectively, calculate:

[0057]

[0058] If there are only three hard rock layers M1, M2, and M3 in the rock layer 100 m above the coal seam, then take M1 + M2 + M3 = 2, and the final danger value is 2, and the danger level is a red warning.

[0059] Continuously correct the values of a, S1, S2, k1, and k2 according to the impact danger level monitored in Step 2, and establish a corresponding relationship formula between the strength, thickness, and distance from the coal seam of the hard rock layer and the impact danger level monitored in Step 2.

[0060] Step 4: According to the multi-source logging data of the newly collected boreholes, through the corresponding relationship between the multi-source logging characteristic parameters and the impact danger level in Step 3, predict the impact danger level of the new borehole area;

[0061] Step 5: When the working face is mined to the position of the new borehole, judge the current impact danger level according to the actually monitored microseismic, stress online, support resistance, ground noise, and drill cuttings data, and compare it with the prediction result. If there is an error, then combine the new data to correct the corresponding relationship between the multi-source logging characteristic parameters and the impact danger level.

[0062] Example 2:

[0063] As Figures 3 - 4 shown.

[0064] The difference between this embodiment and Embodiment 1 is that in Step 3, a machine learning method is used to directly establish the corresponding relationship R between the multi-source logging curve characteristics and the impact danger level monitored in Step 2 p = f(T, n, F, v, S v , Y1, Y2, Y3); The machine learning method includes but is not limited to convolutional neural network (CNN), long short-term memory network (LSTM), and support vector machine (SVM).

[0065] As Figure 3As shown, a machine learning model is established, and the multi-source logging data and rock burst danger level data obtained in Step 1 and Step 2 are used to train the AI model, which includes but is not limited to convolutional neural network (CNN), long short-term memory network (LSTM), support vector machine (SVM), etc.;

[0066] For example, the training process of the AI model includes:

[0067] (1) Data preprocessing: Denoise and normalize the multi-source logging data to eliminate the dimensional difference;

[0068] (2) Feature extraction: Use the convolutional neural network (CNN) to extract the local features of the logging curves;

[0069] (3) Temporal modeling: Use the long short-term memory network (LSTM) to capture the temporal features of the logging curves;

[0070] (4) Classification prediction: Use the support vector machine (SVM) or fully connected neural network (FCN) to perform classification prediction of the rock burst danger level.

[0071] Step 4: As Figure 4 shown, according to the multi-source logging data of the newly collected borehole, predict the rock burst danger level of the new borehole area through the corresponding relationship between the multi-source logging characteristic parameters and the rock burst danger level in Step 3;

[0072] Step 5: When the working face is mined to the position of the new borehole, judge the current rock burst danger level according to the actually monitored microseismic, stress online, support resistance, ground sound, and drill cuttings data, and compare it with the prediction result. If there is an error, then combine the new data to correct the corresponding relationship between the multi-source logging characteristic parameters and the rock burst danger level.

[0073] The parts not involved in the present invention are the same as or can be implemented by the prior art.

Claims

1. A risk assessment method for overburden rockburst with multi-source logging data fusion while drilling, characterized in that It includes the following steps: Step 1: Conduct borehole logging at intervals of 5m to 50m in the underground roadway of the coal mine to obtain data on drilling vibration velocity, drilling speed, rotational speed, torque, thrust, and natural gamma, resistivity, and acoustic velocity curve data, and record the borehole position coordinates, drilling time, and displacement; Step 2: During the working face mining process, combine one or more of microseismic monitoring, on-line stress monitoring, support resistance monitoring, rock noise, and drill cuttings, and combine with the spatial coordinates of the borehole positions to establish a discrimination criterion to determine the actual rock burst danger level near each borehole position on the working face. The actual danger level R s is divided into three levels: no danger, yellow warning, and red warning; Step 3: Obtain the rock stratum strength through drilling sampling and laboratory mechanical parameter determination, and establish the corresponding relationship E between the two by using mathematical statistics or machine learning methods i = f(T, n, F, v, S v , Y1, Y2, Y3), where: E is the rock stratum strength, σ2 is the confining pressure, T is the torque, n is the rotational speed, F is the thrust, v is the drilling speed, S v is the vibration velocity, Y1 is the natural gamma ray, Y2 is the resistivity, and Y3 is the acoustic velocity. Based on the change of E i at the interface of the two rock strata, obtain the rock stratum thickness M i , and based on the cumulative thickness of the drilled rock strata, obtain the distance L from the current rock stratum to the coal seam i , and establish the corresponding relationship between the strength, thickness, and distance from the coal seam of the hard rock stratum and the impact hazard level as follows: Note: M i ‘ is the effective thickness of the i-th hard rock layer, M i ‘ = a × M i , S1, S2 are critical values; k1, k2 are constants; A total of at least 100m of rock strata above the coal seam is taken. Each rock stratum with a strength greater than 60Mpa is calculated with reference to the above table. According to the comparison of the results with S1 and S2, the risk level is selected. When there are multiple hard rock strata, the results of the risk levels are accumulated, and the highest level is a red warning. The impact risk level R actually monitored in step 2 s corresponding to R p , modify the values of a, S1, S2, k1, and k2 so that the judgment formula can better predict the risk level of overburden rock burst. Or use machine learning methods to directly establish the corresponding relationship R between the characteristics of multi-source logging curves and the impact hazard levels monitored in step 2 p = f(T, n, F, v, S v , Y1, Y2, Y3); The machine learning methods include but are not limited to convolutional neural network (CNN), long short-term memory network (LSTM), and support vector machine (SVM); Step 4: Based on the multi-source logging data of the newly collected boreholes, through the corresponding relationship between the multi-source logging characteristic parameters and the rock burst risk level in Step 3, predict the rock burst risk level of the new borehole area; Step 5: When the working face is mined to the position of the new borehole, judge the current rock burst risk level according to the actually monitored microseismic, in-situ stress, support resistance, rock noise, and drill cuttings data, and compare it with the predicted result. If there is an error, then combine the new data to correct the corresponding relationship between the multi-source logging characteristic parameters and the rock burst risk level.

2. The method according to claim 1, characterized in that, The discrimination criteria in Step 2 are as follows: (1) Microseismic monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is judged as a yellow warning; when the energy and frequency change rates exceed the set red threshold, it is judged as a red warning; (2) In-situ stress online monitoring data: When the daily stress value and stress change rate exceed the set yellow threshold, it is judged as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is judged as a red warning; (3) Rock noise monitoring data: When the daily energy, frequency, and their change rates exceed the set yellow threshold, it is judged as a yellow warning; when the stress value and stress change rate exceed the set red threshold, it is judged as a red warning; (4) Support resistance data: When the frequency of the daily resistance value exceeding 90% of the rated working resistance is greater than 10%, it is judged as a yellow warning; When the frequency of the resistance value exceeding 90% of the rated working resistance is greater than 30%, it is judged as a red warning; (5) Except for directly judging as a red warning when a rock burst occurs, directly judging as a yellow warning when the daily energy of microseismic monitoring exceeds the set yellow threshold, and directly judging as a red warning when the daily energy of microseismic monitoring exceeds the set red threshold, it is necessary to comprehensively judge the rock burst risk level by combining microseismic, in-situ stress online, rock noise, and support resistance monitoring. Only when two or more monitoring methods judge a warning can the warning be confirmed, and the comprehensive warning level is the maximum level determined by the above warning methods. (6) The above yellow and red warning thresholds are determined based on the initial values of the rock burst monitoring data and occurrence conditions of working faces under similar conditions; during the actual mining and excavation of the working face, the warning thresholds should be revised according to the maximum energy level of microseismic and the occurrence of rock bursts.

3. The method according to claim 1, characterized in that, During the judgment process of the discrimination criteria in Step 2, it is necessary to judge whether the rock burst risk is caused by overlying strata according to the microseismic positioning height and support resistance combination. If so, then judge this warning as an overlying strata type rock burst warning.

4. The method according to claim 1, wherein The drilling rate, rotation speed, drilling time, displacement, torque, and thrust in step 1 are obtained by a drill rig equipped with sensors or a near-bit measurement sub; the drilling vibration velocity, natural gamma, resistivity, and acoustic velocity are obtained by the near-bit measurement sub during drilling or by an existing borehole logging analyzer after drilling.

5. The method according to claim 1, wherein In step 3, the initial value of a is 0.1 - 3.0, the initial value of k1 is 0.1 - 1.0, the initial value of k2 is 0.2 - 3.0, the initial value of S1 is 10 - 30, and the initial value of S2 is 20 - 50.

Images