A method for predicting rockburst disaster level in deep tunnels

Abstract

The present invention provides a method for predicting the level of rockburst disasters in deep-buried tunnels, and relates to the technical field of tunnel rockburst disaster prediction. The present invention divides the input data into sub-intervals, counts and calculates the distribution probabilities of different sub-intervals, and then uses the interval combination and its probability to further calculate the rockburst probability to generate a risk coefficient. This method not only comprehensively considers the joint effects of various environmental factors, but also quantifies the environmental differences between the monitoring area and the rockburst occurrence area by dynamically adjusting the reference, thereby significantly improving the accuracy and practicality of risk assessment. It is convenient for managers to take corresponding prevention and control measures in a timely manner and optimize the decision-making process. In addition, the flexible adaptability and scientific dynamic adjustment mechanism enable the method to effectively respond to environmental changes in different monitoring areas, greatly enhance the reliability and efficiency of rockburst risk prediction, and improve the safety and risk management level of the overall construction process.

Description

Technical Field

[0001] The invention relates to the technical field of tunnel rock burst disaster prediction, and in particular to a method for predicting the level of rock burst disaster in a deep buried tunnel. Background Art

[0002] In deep tunnel projects, rockburst is an extremely dangerous geological disaster that often has a serious impact on construction safety and project progress. Most of the existing rockburst risk prediction methods are based on static analysis of a single factor, ignoring the dynamic changes and interactions of multiple environmental factors such as stress, temperature, and humidity, resulting in low prediction accuracy and unsatisfactory risk warning effects. Traditional risk assessment methods often rely on empirical judgments, lack scientific basis and systematicity, and are difficult to accurately assess and classify rockburst risks in different regions. Therefore, there is an urgent need for a rockburst prediction and risk assessment method that can integrate multiple factors, dynamically adjust, and have high-precision risk assessment capabilities to improve the safety and efficiency of deep tunnel construction.

[0003] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0004] The purpose of the present invention is to provide a method for predicting the rockburst disaster level in a deep tunnel to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for predicting rock burst disaster level in a deep tunnel, comprising the following specific steps:

[0007] S1: Collect several groups of historical environmental data of rock burst occurrence areas and real-time environmental data of monitoring areas, build a data analysis model, and first use the historical environmental data as input data of the data analysis model;

[0008] S2: The data analysis model preprocesses the input data, divides the preprocessed input data into intervals, generates a probability distribution diagram of the input data, generates several groups of interval combinations according to the interval distribution of each data in the input data, calculates the combination probability of each group of interval combinations, generates a combination probability distribution diagram, and counts the number of rock bursts under each interval combination, and calculates the rock burst probability under different interval combinations;

[0009] S3: Use the real-time environmental data of the monitoring area as the input data of the data analysis model, repeat step S2, generate the corresponding interval combination and probability distribution map, and then generate the risk coefficient based on the combined probability of the interval combination and the rock burst probability, and perform risk prediction based on the risk coefficient.

[0010] Preferably, the historical environmental data includes historical stress data, historical temperature data, and historical humidity data; the real-time environmental data includes real-time stress data, real-time temperature data, and real-time humidity data;

[0011] The historical stress data, historical temperature data, and historical humidity data are calibrated as The superscript i represents the regional number of the rock burst region, i = 1, 2, 3, ..., M, and the subscript j represents the sampling number of the data at the time of sampling, j = 1, 2, 3, ..., N. The real-time stress data, real-time temperature data, and real-time humidity data are calibrated as Fs respectively. j , Ts j , Hs j .

[0012] Preferably, the method for preprocessing the input data includes: normalizing the cleaned input data, removing outliers from the normalized input data, and finally supplementing missing values.

[0013] Preferably, after the input data preprocessing is completed, the overall interval A of the input data is expanded and evenly divided into K sub-intervals A k , the length of each subinterval is calibrated as d, and the calculation method is:

[0014]

[0015] After expanding the overall interval A, A=[0,1], A k =[(k-1)d,kd], where the subscript k represents the interval number of the subinterval, k = 1, 2, 3, ..., K, where Respectively represent the maximum and minimum values ​​of the input data after preprocessing,

[0016] Count the number of input data contained in each subinterval a k , and according to the number of input data contained in the subinterval and the total number of input data, calculate the distribution probability P of the input data in different subintervals k , the calculation method is:

[0017]

[0018] Then, a probability distribution graph is generated based on the distribution probability of the input data in different sub-intervals.

[0019] Preferably, the historical stress data, historical temperature data, and historical humidity data are used as input data to generate corresponding sub-intervals. And calculate the distribution probability Then, the three seed intervals are arranged and combined to generate interval combination B q , where the subscript q represents the number of the interval combination, q = 1, 2, 3, ..., K 3 ;

[0020] The combined probability Pz of each interval combination q The calculation method is:

[0021]

[0022] Finally, calculate the rockburst probability Pb q , calculated as:

[0023]

[0024] Rockburst probability Pb q The value method is:

[0025]

[0026] Where b q Indicates the number of rock bursts under each interval combination.

[0027] Preferably, the real-time stress data, real-time temperature data, and real-time humidity data of the monitoring area are used as input data to generate the corresponding sub-interval A k (Fs j ), A k (Ts j ), A k (Hs j ) and the corresponding distribution probability P k (Fs j ), P k (Ts j ), P k (Hs j );

[0028] Then generate the corresponding interval combination B according to the sub-interval of the monitoring area q ′, corresponding combination probability Pz q ' is calculated as:

[0029] Pz q ′=P k (Fs j )·P k (Ts j )·P k (Hs j )

[0030] The risk coefficient is generated based on the combined probability of each interval combination in the monitoring area and the combined probability and rock burst probability. The calculation method is:

[0031]

[0032] Finally, the risk factor Compared with the preset risk threshold Compare and get the corresponding risk level, where

[0033] Preferably, when When the monitoring area is at low risk,

[0034] when When the risk is higher than 1%, the monitored area is considered to be in a medium-risk state;

[0035] when The monitored area is considered to be in a high-risk state.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The present invention divides the input data into sub-intervals, counts and calculates the distribution probability of different sub-intervals, and then uses the interval combination and its probability to further calculate the rockburst probability, and finally generates a risk coefficient. Compared with the prior art, this method not only comprehensively considers the joint effects of various environmental factors such as stress, temperature, and humidity, but also quantifies the environmental differences between the monitoring area and the rockburst occurrence area by dynamically adjusting the reference, significantly improving the accuracy and practicality of risk assessment. Through the clear division of risk levels, it is convenient for management personnel to take corresponding prevention and control measures in a timely manner and optimize the decision-making process. In addition, the flexible adaptability and scientific dynamic adjustment mechanism enable this method to effectively respond to environmental changes in different monitoring areas, greatly enhancing the reliability and efficiency of rockburst risk prediction. In summary, this scheme provides a scientific, systematic and efficient solution for rockburst risk management in deep tunnel engineering, improving the safety and risk management level of the overall construction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0039] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0040] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0041] Example:

[0042] See also Figure 1 , the present invention provides a technical solution:

[0043] A method for predicting rock burst disaster level in a deep tunnel, comprising the following specific steps:

[0044] S1: Collect several groups of historical environmental data of rock burst occurrence areas and real-time environmental data of monitoring areas, build a data analysis model, and first use the historical environmental data as input data of the data analysis model.

[0045] The historical environmental data includes historical stress data, historical temperature data, and historical humidity data; the real-time environmental data includes real-time stress data, real-time temperature data, and real-time humidity data. The historical stress data, historical temperature data, and historical humidity data are calibrated as The superscript i represents the regional number of the rock burst region, i = 1, 2, 3, ..., M, and the subscript j represents the sampling number of the data at the time of sampling, j = 1, 2, 3, ..., N. The real-time stress data, real-time temperature data, and real-time humidity data are calibrated as Fs respectively. j , Ts j , Hs j .

[0046] During data collection, the lengths of the observation windows for historical environmental data and real-time environmental data are different, and the lengths of the observation windows in different rockburst occurrence areas are the same. The longer the observation window for historical environmental data, the more data samples provided, and the more accurate the final prediction results. The observation window for real-time environmental data is shorter to improve the real-time performance and calculation speed of the prediction results. It can be understood that the maximum value N in the subscript j represents the maximum value of the data samples in an observation window, and is not a fixed value.

[0047] In this step, by collecting historical environmental data from multiple rockburst occurrence areas and real-time environmental data from the current monitoring area, the comprehensiveness and diversity of the data are ensured, so that the prediction model can better capture the rockburst occurrence laws under different conditions from the three most influential directions: stress, temperature, and humidity.

[0048] S2: The data analysis model preprocesses the input data, divides the preprocessed input data into intervals, generates a probability distribution diagram of the input data, generates several groups of interval combinations according to the interval distribution of each data in the input data, calculates the combination probability of each group of interval combinations, generates a combination probability distribution diagram, and counts the number of rock bursts under each interval combination, and calculates the rock burst probability under different interval combinations.

[0049] The method for preprocessing the input data is:

[0050] After cleaning, the input data is normalized and the calculation method is:

[0051]

[0052] In the formula represents the normalized input data, X j max represents the maximum value of the input data, X j min represents the minimum value of the input data;

[0053] Then calculate the input data after normalization The mean And standard deviation σ, clean the input data according to the mean and standard deviation. When there is a p-th group of data in the input data that satisfies Considering this group of data as abnormal data, we traverse the input data to find all abnormal data and replace them uniformly with the mean, which is expressed as:

[0054] Since the calculation methods of the mean and standard deviation are common calculation methods, they will not be repeated here. It can be understood that the input data is not a single type of data, but a general term for multiple types of data, including historical stress data, historical temperature data, historical humidity data, real-time stress data, real-time temperature data, and real-time humidity data. When historical environmental data is used as input data for calculation, the corresponding data range is the historical environmental data of all rock burst occurrence areas. When the implementation environment data is used as input data for calculation, the corresponding data range is only the real-time environmental data of the monitoring area.

[0055] In this step, the input data is normalized to eliminate the differences between different data dimensions, so that all data are in the same scale range (0 to 1), which helps to improve the convergence speed of the model and the accuracy of prediction. At the same time, the 3σ principle is used to clean the input data for outliers. By detecting and replacing abnormal data, the uniformity and consistency of the data set are ensured. The method of replacing outliers with the mean avoids the negative impact of extreme values ​​on model training and prediction.

[0056] After the input data preprocessing is completed, the overall interval A of the input data is expanded and evenly divided into K sub-intervals A k , the length of each subinterval is calibrated as d, and the calculation method is:

[0057]

[0058] After expanding the overall interval A, A=[0,1], A k =[(k-1)d,kd], where the subscript k represents the interval number of the subinterval, k = 1, 2, 3, ..., K, where Respectively represent the maximum and minimum values ​​of the input data after preprocessing, Here, when averaging the subintervals, the overall interval of the input data is changed from The expansion to [0,1] is to align the input data from different regions.

[0059] Count the number of input data contained in each subinterval a k , and according to the number of input data contained in the subinterval and the total number of input data, calculate the distribution probability P of the input data in different subintervals k , the calculation method is:

[0060]

[0061] Then, a probability distribution graph is generated based on the distribution probability of the input data in different sub-intervals.

[0062] In this embodiment, assume that the input data has 100 data samples, the overall interval of the input data is A∈[0.2,0.6], and it is desired to be equally divided into 5 subintervals A1 to A5, then K=5, d=0.2, and there are 21 data falling in the first subinterval A1, that is, in the range of [0,0.2], then the corresponding a k =21,N=100,P k =0.21, the distribution probability reflects the probability of falling into each different sub-interval when the input data is unknown.

[0063] Use historical stress data, historical temperature data, and historical humidity data as input data to generate corresponding sub-intervals And calculate the distribution probability Then, the three seed intervals are arranged and combined to generate interval combination B q , interval combination B q It is expressed as:

[0064]

[0065] Where the subscript q represents the number of the interval combination, q = 1, 2, 3, ..., K 3 , it can be understood that the interval combination B q It is composed of three different types of sub-intervals.

[0066] The combined probability Pz of each interval combination q The calculation method is:

[0067]

[0068] Finally, calculate the rockburst probability Pb q , calculated as:

[0069]

[0070] Rockburst probability Pb q The value method is:

[0071]

[0072] Where b q represents the number of rock bursts under each interval combination. Here, M·N·Pz q This term refers to the total number of M·N data samples that theoretically fall within the corresponding interval combination, so Pb q It reflects the ratio of the number of rockbursts under each interval combination to the total number of samples in the corresponding interval combination. It is a theoretical estimate, and its specific size changes with the size of the data sample M·N. The longer the observation window and the more data samples, the higher the rockburst probability Pb. q The value is more accurate.

[0073] Continuing to use the above assumptions to explain, in data sampling, whether it is historical environmental data or real-time environmental data, each sampling includes three groups: stress, temperature, and humidity. That is to say, the sampling frequency and number of the three are the same, and they correspond to each other according to the order of sampling number j. Take 100 groups of historical stress data, historical temperature data, and historical humidity data as input data, and generate the corresponding 5 sub-intervals respectively: The corresponding distribution probabilities are: Perform permutations and combinations to form interval combination B q , for example:

[0074]

[0075] Taking interval combination B1 as an example, suppose that 20 groups of historical stress data fall within the sub-interval There are 25 groups of historical temperature data that fall within the sub-interval There are 30 groups of historical humidity data that fall within the sub-interval Within, then the corresponding interval probability Pz q =(20*25*30) / 100 3 = 0.015. That is, the interval probability reflects the probability of forming the corresponding combined interval when three sets of unknown input data are input simultaneously.

[0076] When the rock burst occurs, the corresponding historical stress data, historical temperature data, and historical humidity data are counted, and the corresponding sub-intervals and interval combinations of the three are found, and the number of rock bursts under each interval combination can be obtained. q Assuming that there are 2 rockbursts in 100 samplings, and the historical stress data, historical temperature data, and historical humidity data at the time of rockburst satisfy the interval combination B1, then the rockburst probability corresponding to this interval combination is Pb1=2 / 1.5, which is obviously greater than 1. Therefore, the rockburst probability is taken as Pb1=1. If the total number of samples is expanded, it is found that there are still only 2 rockbursts in 1000 samplings, then the corresponding rockburst probability is Pb1=2 / 15=0.133, and the rockburst probabilities corresponding to other interval combinations are Pb2~Pb 125 are all 0, indicating the probability of rock burst under this interval combination. This is why the value of rock burst probability changes with the size of the data sample M·N. The longer the observation window and the more data samples, the higher the rock burst probability Pb q The value is more accurate.

[0077] In this step, by dividing the overall interval into several sub-intervals, the distribution of the input data can be refined, making the statistics of each sub-interval more specific and accurate, which helps to capture subtle changes and distribution characteristics of the data. Moreover, by calculating the distribution probability of each sub-interval and the combined probability of interval combinations, the probability distribution of the input data in each interval can be intuitively reflected, providing a solid foundation for subsequent rock burst risk assessment. At the same time, by arranging and combining historical stress data, historical temperature data, and historical humidity data, and comprehensively considering the joint effects of various environmental factors, the rock burst risk is analyzed and evaluated in multiple dimensions, which improves the comprehensiveness and accuracy of the prediction.

[0078] S3: Use the real-time environmental data of the monitoring area as the input data of the data analysis model, repeat step S2, generate the corresponding interval combination and probability distribution map, and then generate the risk coefficient based on the combined probability of the interval combination and the rock burst probability, and perform risk prediction based on the risk coefficient.

[0079] The real-time stress data, real-time temperature data, and real-time humidity data of the monitoring area are used as input data to generate the corresponding sub-interval A k (Fs j ), A k (Ts j ), A k (Hs j ) and the corresponding distribution probability P k (Fs j ), P k (Ts j ), P k (Hs j );

[0080] Then generate the corresponding interval combination B according to the sub-interval of the monitoring area q ′, interval combination B q ′ is expressed as:

[0081] B q ′←"A k (Fs j )"+"A k (Ts j )"+"A k (Hs j )"

[0082] Corresponding combination probability Pz q ' is calculated as:

[0083] Pz q ′=P k (Fs j )·P k (Ts j )·P k (Hs j )

[0084] In this step, the steps and logic for calculating the real-time stress data, real-time temperature data, and real-time humidity data of the monitored area as input data are the same as the steps and logic for calculating the historical stress data, historical temperature data, and historical humidity data as input data in the above example, so they are not elaborated here.

[0085] The risk coefficient is generated based on the combined probability of each interval combination in the monitoring area and the combined probability and rock burst probability. The calculation method is:

[0086]

[0087] Finally, the risk factor Compared with the preset risk threshold Compare and get the corresponding risk level, where

[0088] when When the monitoring area is at low risk,

[0089] when When the risk is higher than 1%, the monitored area is considered to be in a medium-risk state;

[0090] when The monitored area is considered to be in a high-risk state.

[0091] From the definition of risk factor, we can see that Pz q ′·Pb q It represents the probability of rock burst when the input data falls into a certain interval combination. Since rock burst may correspond to a variety of different interval combinations of pressure, temperature, and humidity, all interval combinations are traversed. It means that the input data falls within all possible rock burst interval combinations and the probability of rock burst. The lower part of the risk factor, |Pz q ′-Pz q | represents the absolute value of the difference between a certain combination probability in the monitoring area and the corresponding combination probability in the rock burst occurrence area, reflecting the difference between the two. Since the combination probability of a certain interval combination is determined by the distribution of input data in different sub-intervals, and this distribution is mainly affected by environmental factors such as geological features in different regions, |Pz q ′-Pz q |This absolute value of the difference is equivalent to taking the rockburst occurrence area as a reference, which is used to quantify the environmental difference between the monitoring area and the rockburst occurrence area. The smaller the environmental difference, the stronger the reference of the rockburst probability calculated by taking the rockburst occurrence area as a reference to the monitoring area. On the contrary, the larger the environmental difference, the weaker the reference of the rockburst probability calculated by taking the rockburst occurrence area as a reference to the monitoring area. Therefore, the risk coefficient is It can well quantify and reflect the risks of the monitored area.

[0092] In summary, by dividing the input data into sub-intervals, the distribution probability of different sub-intervals is counted and calculated, and then the interval combination and its probability are used to further calculate the rockburst probability, and finally generate the risk coefficient. Compared with the existing technology, this method not only comprehensively considers the joint effects of various environmental factors such as stress, temperature, and humidity, but also quantifies the environmental differences between the monitoring area and the rockburst occurrence area by dynamically adjusting the reference, which significantly improves the accuracy and practicality of risk assessment. Through the clear risk level division, it is convenient for managers to take corresponding prevention and control measures in time and optimize the decision-making process. In addition, the flexible adaptability and scientific dynamic adjustment mechanism enable this method to effectively respond to environmental changes in different monitoring areas, greatly enhancing the reliability and efficiency of rockburst risk prediction. In summary, this scheme provides a scientific, systematic and efficient solution for rockburst risk management in deep tunnel engineering, which improves the safety and risk management level of the overall construction process.

[0093] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0094] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0095] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0096] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for predicting rockburst disaster level in a deep tunnel, characterized in that: The specific steps include: S1: Collect several groups of historical environmental data of rock burst occurrence areas and real-time environmental data of monitoring areas, build a data analysis model, and first use the historical environmental data as input data of the data analysis model; S2: The data analysis model preprocesses the input data, divides the preprocessed input data into intervals, generates a probability distribution diagram of the input data, generates several groups of interval combinations according to the interval distribution of each data in the input data, calculates the combination probability of each group of interval combinations, generates a combination probability distribution diagram, and counts the number of rock bursts under each interval combination, and calculates the rock burst probability under different interval combinations; The historical environmental data includes historical stress data, historical temperature data, and historical humidity data. The historical stress data, historical temperature data, and historical humidity data are calibrated as The superscript i represents the regional number of the rock burst occurrence area, i = 1, 2, 3, ..., M, M is the total number of rock burst occurrence areas, and the subscript j represents the sampling number of the data at the time of sampling, j = 1, 2, 3, ..., N, N is the total number of sampled data; Use historical stress data, historical temperature data, and historical humidity data as input data to generate corresponding sub-intervals And calculate the distribution probability Then, the three seed intervals are arranged and combined to generate interval combination B q , where the subscript q represents the number of the interval combination, q = 1, 2, 3, ..., K 3 ; S3: Using the real-time environmental data of the monitoring area as the input data of the data analysis model, repeating step S2, generating the corresponding interval combination and probability distribution map, and then generating the risk coefficient based on the combined probability of the interval combination and the rock burst probability, and performing risk prediction based on the risk coefficient; Real-time environmental data includes real-time stress data, real-time temperature data, and real-time humidity data. The real-time stress data, real-time temperature data, and real-time humidity data are calibrated as Fs respectively. j , Ts j , Hs j ; The real-time stress data, real-time temperature data, and real-time humidity data of the monitoring area are used as input data to generate the corresponding sub-interval A k (Fs j ), A k (Ts j ), A k (Hs j ) and the corresponding distribution probability P k (Fs j ), P k (Ts j ), P k (Hs j ), and then generate the corresponding interval combination B according to the sub-interval of the monitoring area q ′; The risk factor is calculated as: In the formula represents the risk factor, Pb q Represents the probability of rock burst.

2. A method for predicting rockburst disaster level in a deep tunnel according to claim 1, characterized in that: The method for preprocessing the input data includes: normalizing the cleaned input data, removing outliers from the normalized input data, and finally supplementing missing values.

3. A method for predicting rockburst disaster level in a deep tunnel according to claim 2, characterized in that: After the input data preprocessing is completed, the overall interval A of the input data is expanded and evenly divided into K sub-intervals A k , the length of each subinterval is calibrated as d, and the calculation method is: After expanding the overall interval A, A=[0,1], A k =[(k-1)d,kd], where the subscript k represents the interval number of the subinterval, k = 1, 2, 3, ..., K, where Respectively represent the maximum and minimum values ​​of the input data after preprocessing, Count the number of input data contained in each subinterval a k , and according to the number of input data contained in the subinterval and the total number of input data, calculate the distribution probability P of the input data in different subintervals k , the calculation method is: Then, a probability distribution graph is generated based on the distribution probability of the input data in different sub-intervals.

4. A method for predicting rockburst disaster level in a deep tunnel according to claim 3, characterized in that: When historical stress data, historical temperature data, and historical humidity data are used as input data, the combined probability Pz of each interval combination q The calculation method is: Finally, calculate the rockburst probability Pb q , calculated as: Rockburst probability Pb q The value method is: Where b q Indicates the number of rock bursts under each interval combination.

5. A method for predicting rockburst disaster level in a deep tunnel according to claim 4, characterized in that: When the real-time stress data, real-time temperature data, and real-time humidity data of the monitoring area are used as input data, the corresponding combination probability Pz q ' is calculated as: Pz q ′=P k (Fs j )·P k (Ts j )·P k (Hs j ) The risk coefficient is generated based on the combined probability of each interval combination in the monitoring area and the combined probability and rock burst probability. Finally, the risk factor Compared with the preset risk threshold Compare and get the corresponding risk level, where 6. A method for predicting rockburst disaster level in a deep tunnel according to claim 5, characterized in that: when When the monitoring area is at low risk, when When the risk is higher than 1%, the monitored area is considered to be in a medium-risk state; when The monitored area is considered to be in a high-risk state.

Images