Coal and gas outburst intelligent early warning method and system based on T-B cell immune algorithm

Abstract

The invention discloses an intelligent early warning method and system for coal and gas outburst based on a T-B cellular immune algorithm. The method comprises the following steps: acquiring various monitoring data in a coal mine; preprocessing the monitoring data to obtain processed data; carrying out dimensionality reduction and local feature value extraction on the processed data by utilizing a deep learning combination model; setting a characteristic value threshold value, and screening the processed data according to the characteristic value threshold value and the local characteristic value to obtain abnormal data and normal data; constructing a T-B cell immune recognition model, and inputting the abnormal data into the T-B cell immune recognition model; and setting an early warning threshold value of the danger level, and outputting the danger level according to the detector type generated by the T cells and the set threshold value so as to realize intelligent early warning of coal and gas outburst. According to the invention, the T-B cellular immune algorithm is creatively combined with a mature underground Internet of Things system, so that the reliability and accuracy of accident early warning are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of coal-rock power disaster monitoring and early warning, and in particular relates to a coal and gas outburst intelligent early warning method and system based on TB cell immune algorithm. Background Art

[0002] Coal and gas outburst is an extremely complex dynamic phenomenon in coal mines, posing a serious threat to coal mine safety production and stable coal supply. With the increase of mining depth and mining intensity, outburst disasters are becoming more and more serious. At present, my country's coal mine gas extraction adheres to the principle of "extracting as much as possible, taking multiple measures, and balancing extraction and excavation". Mining operations can only be arranged after the extraction meets the standards. At the technical level, my country has formed a gas extraction technology system including pre-extraction of coal seam gas in ground wells, pre-extraction of coal seam gas in sections by underground through-layer drilling or along-layer drilling. In addition, daily prediction of outbursts uses a large number of drilling static prediction methods such as drill cuttings S, initial velocity of gas outburst q, gas desorption index K1 or Δh2, gas emission index ΔP, coal body Proctor coefficient f, and gas pressure p. These methods have played an important role in curbing the occurrence of disasters.

[0003] Although the existing gas extraction technology and prediction methods have played a positive role in preventing and controlling coal and gas outbursts to a certain extent, there are still some obvious shortcomings. First, the gas extraction effect is affected by many complex factors such as permeability, and with the increase of mining depth, the permeability of coal seams decreases, and the difficulty of gas extraction increases. Secondly, the traditional prediction method requires the construction of a certain number of drilling holes in the prediction process, which takes up a certain amount of working time, and due to the discontinuity of time and space, the relatively single index and the low integration, there are certain limitations in the prediction of outburst danger. In addition, domestic and foreign outburst accident cases show that outbursts often occur in geological structural belts or places where coal seams change sharply, but not all structural coal areas have the conditions for outbursts; there are usually signs before the outburst occurs, but there are also quite a few places on the scene where the signs appear, but the outburst does not occur. This shows that the existing prediction technology and methods need to be improved in terms of accuracy and real-time performance. Comprehensive index prediction usually selects multiple indexes for analysis, but due to the complex nonlinear relationship between coal and gas outbursts and various factors, the existing methods have limitations in dealing with this complexity. In short, the existing technology cannot improve the prediction accuracy and real-time performance of coal and gas outbursts. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a coal and gas outburst intelligent early warning method and system based on TB cell immune algorithm to solve the problems existing in the above prior art.

[0005] To achieve the above objectives, in a first aspect, the present invention provides an intelligent early warning method for coal and gas outburst based on TB cell immune algorithm, comprising:

[0006] Acquire various monitoring data in the coal mine; pre-process the monitoring data to obtain processed data;

[0007] Using a deep learning combined model to reduce the dimension of the processed data and extract local eigenvalues; setting an eigenvalue threshold, and screening the processed data according to the eigenvalue threshold and the local eigenvalue to obtain abnormal data and normal data; wherein the normal data produces immune tolerance;

[0008] Constructing a TB cell immune recognition model, inputting the abnormal data into the TB cell immune recognition model, and outputting a detector type; wherein the TB cell immune recognition model includes T cells and B cells;

[0009] A warning threshold of the danger level is set, and the danger level is outputted together according to the detector type and the warning threshold, so as to realize intelligent warning of coal and gas outburst.

[0010] Preferably, obtaining a variety of monitoring data in a coal mine includes:

[0011] By using gas concentration sensors, gas pressure sensors and stress gauges to monitor changes in gas concentration, gas pressure and ground stress respectively, a variety of monitoring data in the coal mine are obtained.

[0012] Preferably, preprocessing the monitoring data includes:

[0013] Using an oversampling method to supplement edge data in the monitoring data to obtain new sample data;

[0014] Using SVM to perform data denoising on the monitoring data to obtain a decision boundary, and deleting invalid noise points through the decision boundary to obtain valid data points;

[0015] The monitoring data is normalized to obtain processed data.

[0016] Preferably, inputting the abnormal data into the TB cell immune recognition model comprises:

[0017] Generate a detector through T cells, perform matching calculation on the detector and the abnormal data, and identify the detector type;

[0018] A learning vector is generated by B cells, and a learning code is generated by the learning vector. After the learning code is transmitted to T cells, the detector of the T cells is controlled to move or proliferate, so as to optimize the T cells.

[0019] Preferably, the method further comprises the step of memorizing:

[0020] The output danger level is stored in the database and memorized. If there is abnormal data of the same category, an immune response is directly triggered.

[0021] In a second aspect, the present invention also provides a coal and gas outburst intelligent early warning system based on TB cell immune algorithm, comprising:

[0022] A data preprocessing module is used to obtain various monitoring data in the coal mine; preprocess the monitoring data to obtain processed data;

[0023] A data screening module is used to reduce the dimension of the processed data and extract local eigenvalues ​​using a deep learning combined model; set an eigenvalue threshold, and screen the processed data according to the eigenvalue threshold and the local eigenvalue to obtain abnormal data and normal data; wherein the normal data produces immune tolerance;

[0024] A risk identification module, used to construct a TB cell immune recognition model, input the abnormal data into the TB cell immune recognition model, and output the T cell detector type; wherein the TB cell immune recognition model includes T cells and B cells;

[0025] The early warning module is used to set the early warning threshold of the danger level. According to the detector type generated in the T module and the set early warning threshold, the danger level is jointly output to realize intelligent early warning of coal and gas outburst.

[0026] In a third aspect, the present invention further discloses a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.

[0027] In a fourth aspect, the present invention further discloses a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0028] In a fifth aspect, the present invention further discloses a computer program product, including a computer program, which implements the steps of the method described in the first aspect when executed by a processor.

[0029] Compared with the prior art, the present invention has the following advantages and technical effects:

[0030] The present invention discloses an intelligent early warning method for coal and gas outburst based on TB cell immune algorithm. First, a variety of monitoring data in a coal mine are obtained; the monitoring data are preprocessed to obtain processed data; secondly, a deep learning combination model is used to reduce the dimension of the processed data and extract local eigenvalues; an eigenvalue threshold is set, and the processed data is screened according to the eigenvalue threshold and the local eigenvalue to obtain abnormal data and normal data; wherein the normal data produces immune tolerance; then, a TB cell immune recognition model is constructed, the abnormal data is input into the TB cell immune recognition model, and a T cell detector type is output; wherein the TB cell immune recognition model includes T cells and B cells; finally, an early warning threshold of a danger level is set, and according to the detector type generated in the T module, the danger level is jointly output in combination with the set threshold to realize intelligent early warning of coal and gas outburst.

[0031] The present invention combines the data processing method of machine learning with the underground Internet of Things data monitoring system to jointly construct an underground outburst hazard monitoring and early warning system, which solves many problems of the current underground monitoring and early warning system, such as inaccuracy, long warning calculation time, and inability to judge the accident category in real time. At the same time, it has a memory function, can trigger the warning faster, extend the evacuation time, and is safer. The present invention makes full use of the small amount of outburst-related data that can be detected underground, expands the data sample, makes full use of each piece of data, and inputs more feature vectors into the recognition system. The present invention innovatively combines the TB cell immune algorithm with a mature underground Internet of Things system to improve the reliability and accuracy of accident warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0033] Figure 1 Schematic diagram of the method steps of an embodiment of the present invention. DETAILED DESCRIPTION

[0034] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0035] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] With the continuous increase in the depth and intensity of coal mining, the phenomenon of "three highs and one disturbance" (high ground stress, high osmotic pressure, high geothermal temperature, and strong mining disturbance) has become more obvious, and the threat of the intensity and frequency of coal and gas outburst disasters has become more serious, which can easily lead to major accidents.

[0037] As one of the hot topics in artificial intelligence research, the artificial immune system draws on many excellent characteristics of the biological immune system and has been widely used in many fields such as fault diagnosis, information processing, machine learning, and intelligent optimization. It can get rid of the dependence on experience to a certain extent and is more autonomous. In view of the uncertainty and spatiotemporal coupling of coal and gas outbursts in the coal mine production process, it is unknown when and where the influx into the system is unknown, which is very similar to the immune system. It is also unknown when and where the antigen invades the immune system. The immune system can accurately identify known or unknown antigens, and produce corresponding antibodies to eliminate the antigens, and can maintain the stability of the body in a constantly changing environment. This is in line with the requirements for effective identification and prevention and control of the risks of coal and gas outbursts in the dynamically changing underground environment.

[0038] Accurate prediction is the key to coal and gas outburst prevention and control. It can detect the potential outburst hazards in the mine as early as possible and take effective prevention and control measures to ensure the safety of life and property of coal mine workers and the normal production of the mine. Coal and gas outburst prediction is to evaluate the potential risks and possibilities of coal and gas outburst by analyzing coal mine gas geological parameters, gas outburst volume and other information.

[0039] Embodiment 1

[0040] like Figure 1 As shown, this embodiment provides a coal and gas outburst intelligent early warning method based on TB cell immune algorithm, including:

[0041] S1. Obtain various monitoring data in the coal mine;

[0042] Specifically, data monitoring and collection: monitor changes in gas concentration, gas pressure, and ground stress through gas concentration sensors, gas pressure sensors, and stress gauges, and transmit the collected data to the first data processing center;

[0043] S2. Preprocessing the monitoring data to obtain processed data;

[0044] Specifically, the first data processing is to pre-process the data collected by the sensor, and the methods include: data supplementation, data denoising, and data normalization;

[0045] The data supplement adopts the oversampling method (BSMOTE) to synthesize new samples for the minority class sample data at the boundary to improve the data marginalization problem.

[0046] S21. Data supplementation calculation is as follows:

[0047] a. Calculate the probability that the noise is true:

[0048]

[0049] in, is the judgment value of the true noise of the jth noise point in cluster i; α is the balance coefficient; is the Euclidean distance from the jth noise point in cluster i to the cluster center; δ is a very small positive number to ensure is positive; is the near-neighborhood density of the jth noise point in cluster i.

[0050] b. Remove j noise points in probability order, and the remaining data points are cluster i'.

[0051] c. Calculate the amount of information of positive samples in the cluster:

[0052]

[0053]

[0054] in are the information amounts of wrong and correct classified samples in cluster i respectively; ω * is the hyperplane normal vector; b * is the hyperplane threshold.

[0055] d. Synthesize new samples, the calculation formula is:

[0056]

[0057] Where g is a new synthesized sample; is the positive sample containing the maximum amount of information; For sample points in the near-field set; ε' is a parameter for adjusting the sample position.

[0058] (2) Data denoising calculation is as follows:

[0059] Use SVM to determine a reasonable decision boundary, separate valid data points from invalid noise points to the greatest extent, and delete invalid noise points.

[0060] First, calculate the distance from each data point to the decision surface:

[0061]

[0062] Among them, ω T is the decision surface normal vector.

[0063] The optimized objective function is:

[0064]

[0065] When the result of this formula is the smallest, it is the optimal decision surface.

[0066] (3) Data normalization calculation is as follows:

[0067]

[0068] Among them, X is the data sample, X min is the minimum value of the data, X max is the maximum value of the data, X n is the normalized target matrix.

[0069] S3, using a deep learning combined model to reduce the dimension of the processed data and extract local eigenvalues;

[0070] Specifically, the preprocessed data is input into the deep learning combined model, and a convolutional neural network is used to reduce the dimension of the data and extract local feature parameter values.

[0071] The method of extracting features from the preprocessed input data in the process of convolutional neural network reducing the dimension of data and extracting local feature parameter values ​​is as follows:

[0072]

[0073] Among them, Y n is the output feature, n is the nth convolutional layer, f is the activation function, i is the number of filters, W is the filter weight, and Y n-1 is the output feature of the previous layer, and b is the bias.

[0074] The method for selecting the extracted features in the process of reducing the dimension of data and extracting local feature parameter values ​​using convolutional neural networks is:

[0075] X=f[αS(x)+b] (9)

[0076] Where X is the output, f is the activation function, α is the multiplicative bias, S(x) is the downsampling function, and b is the bias.

[0077] S4, setting a characteristic value threshold, and screening the processed data according to the characteristic value threshold and the local characteristic value to obtain abnormal data and normal data; wherein the normal data generates immunity;

[0078] As an additional implementation method, after extracting the features, the data is screened according to the set thresholds of each feature value to screen out normal data and abnormal data, and all the data are input into the TB immune recognition algorithm.

[0079] For normal data after screening, immune tolerance will be generated and no immune system response will be triggered. At this time, the control monitoring system continues to monitor and the system is stable;

[0080] Abnormal data after screening, which is equivalent to antigens in the biological body, will trigger a response from the immune system.

[0081] After adopting this type of data processing method, the output real-valued data can be input into the recognition system, which can also output real-valued data instead of binary signals, which is more conducive to identification and the early warning is more intuitive.

[0082] S5, constructing a TB cell immune recognition model, inputting the abnormal data into the TB cell immune recognition model, and outputting a risk level; wherein the TB cell immune recognition model includes T cells and B cells;

[0083] As an additional implementation method, the immune recognition model is mainly divided into a T module and a B module. The T module mainly generates a detector after receiving abnormal data, and then matches and calculates the detector with the received data to identify the danger level and then output the danger level; the B module is mainly responsible for optimizing the T module. Since the detectors generated by the T module are randomly distributed and lack learning ability, and since the distribution has no emphasis, there will be a problem of insufficient detection sensitivity. The B module generates learning vectors and clone vectors, and then generates learning codes. After the learning codes are transmitted to the T module, the detectors of the T module can be controlled to move or proliferate, and optimization is performed in these two ways.

[0084] S51, T module:

[0085] a. Detector generation;

[0086] The detectors in the module are randomly generated. The generated detectors are divided into self-friendly and non-self-friendly. The T module itself will make the self-friendly detectors die, and the non-self-friendly detectors enter the mature set to wait for maturity. This process is repeated until the number of detectors in the mature detector set meets the requirements, and then the input data is judged. According to the distribution of detectors in different areas, the abnormality level can be divided into: general prominence, large prominence, serious prominence, and particularly serious prominence.

[0087] b. Matching calculation;

[0088] After the mature detector set meets the requirements, it is necessary to perform matching calculations on the input abnormal data, determine the affinity between the abnormal data and the detector, and calculate the distance between the feature vector and the detector. When the distance is less than the threshold, it means that the two have high affinity and the match is successful. The feature vector whose distance exceeds the threshold cannot be recognized, and the detector needs to be optimized at this time.

[0089] S52, B module:

[0090] a. Generate learning vectors and clone vectors;

[0091] After receiving abnormal data, module B first randomly generates a preset number of learning vectors, and then performs clone learning to generate clone vectors. When the total vector reaches the preset value, the cloning process is stopped to avoid too many clone vectors affecting the calculation speed.

[0092] b. Generate learning code;

[0093] Learning codes are generated in the process of generating vectors. After the complete learning codes are generated, they are transmitted to the T module to optimize the T module, generate more detectors or move the detectors to make recognition more accurate and faster.

[0094] S53. The immune recognition model outputs the risk level.

[0095] S6. Set a warning threshold of the danger level, and compare the warning threshold with the output danger level to achieve intelligent warning of coal and gas outburst.

[0096] Specifically, the threshold method and the trend method are used to determine the alarm threshold.

[0097] (1) Threshold method;

[0098] First, the initial threshold is determined based on experience and relevant standards, with reference to coal mine safety regulations and previous coal and gas outburst accident case experience. At the same time, statistical analysis is used to assist, using the "mean + k times standard deviation" method (the k value is determined according to the risk acceptance level, such as k = 2). If the mean of the abnormal data is μ and the standard deviation is σ, then the warning threshold can be set to μ+2σ. At the same time, considering the coal body structure factors, if a large amount of broken coal or mylonitic coal appears in a certain area, even if the gas pressure, concentration and stress do not reach the threshold, an early warning should be issued.

[0099] (2) Trend method;

[0100] The moving average method is used to analyze the trend of data such as gas outburst volume, which can clearly show the long-term trend. If the moving average value of gas outburst volume continues to rise over a period of time, it may indicate an increase in the risk of coal and gas outburst. At the same time, linear regression or nonlinear regression is used to analyze the relationship between data such as gas pressure change rate and time. Calculate the slope of the trend line of data such as gas concentration. When the slope changes significantly, such as from close to 0 to a large positive value, this may be a key node in the change of coal and gas outburst trend. For data showing nonlinear trends, such as the change curve of gas pressure change rate, analyze its second-order derivative. The zero point or sign change point of the second-order derivative may be the turning point of the trend, that is, the key node. The key node is also the warning threshold point.

[0101] S7. Data memory.

[0102] As an additional implementation method, for newly emerging data types and data values, the function of memory cells is used to memorize them and store them in the database. The next time there is abnormal data of the same category, it can directly trigger an immune response, and the emergency response speed is faster. The collected abnormal data is shared to the cloud and referenced by multiple coal mining units.

[0103] Beneficial effects of this embodiment:

[0104] First, the recognition function is an indispensable function of the biological immune system to defend against antigens, and is also a prerequisite for generating an effective immune response. The coal and gas outburst hazard recognition system established in this embodiment based on the recognition ability of the immune system has a better recognition ability;

[0105] Secondly, the biological immune system realizes the memory of antigens through the memory function. When the same antigen invades again, the immune system can recognize the antigen at a faster speed and produce antibodies to eliminate it. The memory function of the TB cell immune model in this embodiment improves the recognition ability of the immune system to a certain extent, which is also an indispensable function in the recognition process.

[0106] Next, the environment in which the immune system is located will face many complex unknown antigens, just like the underground mine will also face different types of danger categories and data. This embodiment innovatively combines the adaptability of the immune system to improve the adaptability of the early warning system, and has the ability to process and remember unknown data, so as to respond to unknown data more quickly;

[0107] In addition, the various components of the biological immune system are distributed throughout the body, so as to achieve the purpose of comprehensive protection of the body. Moreover, the biological immune system is not controlled by a certain control center. This feature ensures that the various functions of the biological immune system will not fail as a whole due to damage to a single component. Inspired by this, due to the complexity of the underground environment, there is uncertainty about when and where coal and gas outbursts occur, and continuous monitoring of each monitoring point on the working face is required. Therefore, the coal and gas outburst risk identification system based on the immune mechanism in this embodiment should also have distributed capabilities.

[0108] Finally, the recognition function of the biological immune system is achieved through the synergistic action of multiple immune cells. For the identification of underground coal and gas outburst risks, the characteristics of factor changes shown at different monitoring locations may be different, so this embodiment needs to make accurate identification and effective decisions through collaborative analysis.

[0109] Embodiment 2

[0110] Based on the same inventive concept, this embodiment provides a coal and gas outburst intelligent early warning system based on TB cell immune algorithm, including:

[0111] A data preprocessing module is used to obtain various monitoring data in the coal mine; preprocess the monitoring data to obtain processed data;

[0112] A data screening module is used to reduce the dimension of the processed data and extract local eigenvalues ​​by using a deep learning combined model; set an eigenvalue threshold, and screen the processed data according to the eigenvalue threshold and the local eigenvalue to obtain abnormal data and normal data; wherein the normal data is immune;

[0113] A risk identification module, used to construct a TB cell immune recognition model, input the abnormal data into the TB cell immune recognition model, and output the risk level; wherein the TB cell immune recognition model includes T cells and B cells;

[0114] The early warning module is used to set the early warning threshold of the danger level, and compare the early warning threshold with the output danger level to achieve intelligent early warning of coal and gas outburst.

[0115] The intelligent early warning system for coal and gas outburst based on TB cell immune algorithm provided in this embodiment has all the advantages of the intelligent early warning method for coal and gas outburst based on TB cell immune algorithm provided in Example 1.

[0116] Embodiment 3

[0117] This embodiment further discloses a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first embodiment.

[0118] Embodiment 4

[0119] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first embodiment are implemented.

[0120] Embodiment 5

[0121] This embodiment also discloses a computer program product, including a computer program, which implements the steps of the method described in the first embodiment when executed by a processor.

[0122] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent early warning method for coal and gas outburst based on TB cell immune algorithm, characterized in that: The following steps are involved: Obtain various monitoring data in coal mines; Preprocessing the monitoring data to obtain processed data; Using a deep learning combined model to reduce the dimension of the processed data and extract local eigenvalues; Setting a characteristic value threshold, and screening the processed data according to the characteristic value threshold and the local characteristic value to obtain abnormal data and normal data; wherein the normal data generates immune tolerance; Constructing a TB cell immune recognition model, inputting the abnormal data into the TB cell immune recognition model, and outputting a detector type; wherein the TB cell immune recognition model includes T cells and B cells; A warning threshold of the danger level is set, and the danger level is outputted together according to the detector type and the set warning threshold, so as to realize intelligent warning of coal and gas outburst.

2. The method according to claim 1, characterized in that Obtain various monitoring data in coal mines including: By using gas concentration sensors, gas pressure sensors and stress gauges to monitor changes in gas concentration, gas pressure and ground stress respectively, a variety of monitoring data in the coal mine are obtained.

3. The method according to claim 1, characterized in that Preprocessing the monitoring data includes: Using an oversampling method to supplement edge data in the monitoring data to obtain new sample data; Using SVM to perform data denoising on the monitoring data to obtain a decision boundary, and deleting invalid noise points through the decision boundary to obtain valid data points; The monitoring data is normalized to obtain processed data.

4. The method according to claim 1, characterized in that Inputting the abnormal data into the TB cell immune recognition model comprises: The detector is generated by T cells, the detector and the abnormal data are matched and calculated, and the danger level is identified; A learning vector is generated by B cells, and a learning code is generated by the learning vector. After the learning code is transmitted to T cells, the detector of the T cells is controlled to move or proliferate, so as to optimize the T cells.

5. The method according to claim 1, characterized in that: Also includes memorization steps: The output danger level is stored in the database and memorized. If there is abnormal data of the same category, an immune response is directly triggered.

6. An intelligent early warning system for coal and gas outburst based on TB cell immune algorithm, characterized in that: include: Data preprocessing module, used to obtain various monitoring data in coal mines; Preprocessing the monitoring data to obtain processed data; A data screening module, used for reducing the dimension of the processed data and extracting local eigenvalues ​​by using a deep learning combined model; Setting a characteristic value threshold, and screening the processed data according to the characteristic value threshold and the local characteristic value to obtain abnormal data and normal data; wherein the normal data generates immune tolerance; A risk identification module, used to construct a TB cell immune recognition model, input the abnormal data into the TB cell immune recognition model, and output the detector type; wherein the TB cell immune recognition model includes T cells and B cells; The early warning module is used to set the early warning threshold of the danger level. According to the detector type and the set early warning threshold, the danger level is jointly output to realize intelligent early warning of coal and gas outburst.

7. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claims 1 to 5 are implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to claims 1 to 5 are implemented.

Images