Microseismic event identification method and device

Abstract

The invention provides a micro-seismic event identification method and device, and the method comprises the steps: monitoring a coal mine region, so as to obtain a to-be-identified micro-seismic signal; performing dynamic frequency slice wavelet transform (DFSWT) on the micro-seismic signal to be identified to obtain a corresponding target time-frequency diagram; performing feature extraction on the to-be-identified micro-seismic signal to obtain a corresponding target feature; based on the target time-frequency graph and the target features, adopting a target recognition model to obtain prediction probabilities of a plurality of microseismic events output by the target recognition model; wherein the prediction probability is used for indicating the probability of occurrence of the corresponding micro-seismic event; and determining a target micro-seismic event occurring in the coal mine area from the plurality of micro-seismic events according to the prediction probabilities of the plurality of micro-seismic events. According to the target time-frequency graph obtained by carrying out dynamic frequency slice wavelet transformation on the to-be-identified micro-seismic signals and the target features obtained by extracting the features of the to-be-identified micro-seismic signals, automatic identification of the micro-seismic events occurring in the coal mine area is realized, and the accuracy and effectiveness of micro-seismic event identification are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a method and device for identifying microseismic events. Background Art

[0002] During the hydraulic fracturing of the coal seam roof, the identification of microseismic events generated by hydraulic fracturing is the basis and prerequisite for ensuring the accurate positioning of the fracture network and accurate evaluation of the fracturing effect. However, during the microseismic monitoring process, the microseismic signals caused by hydraulic fracturing events are easily mixed with the signals caused by a large number of non-hydraulic fracturing events, such as roof breakage, coal mining machine cutting, drilling, etc. The signals are similar in time domain and partially overlap in frequency domain, which increases the difficulty of identification. Accurately identifying and classifying microseismic events generated by multiple sources is of great significance for accurate positioning of the fracture network and early warning of dynamic disasters in the coal seam roof. Summary of the invention

[0003] The present disclosure provides a method and device for identifying microseismic events, which can at least solve one of the technical problems in the related art to a certain extent. The technical solution of the present disclosure is as follows:

[0004] According to one aspect of the present disclosure, a method for identifying microseismic events is provided, comprising:

[0005] Monitor the coal mine area to obtain microseismic signals to be identified;

[0006] Performing dynamic frequency slicing wavelet transform (DFSWT) on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram;

[0007] Extracting features of the microseismic signal to be identified to obtain corresponding target features;

[0008] Based on the target time-frequency diagram and the target characteristics, a target recognition model is used to obtain the predicted probabilities of multiple microseismic events output by the target recognition model; wherein the predicted probability is used to indicate the probability of the occurrence of the corresponding microseismic event;

[0009] According to the predicted probabilities of the multiple microseismic events, a target microseismic event occurring in the coal mine area is determined from the multiple microseismic events.

[0010] According to another aspect of the present disclosure, another microseismic event identification device is provided, comprising:

[0011] A monitoring module is used to monitor the coal mine area to obtain microseismic signals to be identified;

[0012] A transformation module, used for performing dynamic frequency slicing wavelet transform DFSWT on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram;

[0013] An extraction module, used for extracting features from the microseismic signal to be identified to obtain corresponding target features;

[0014] A processing module, configured to obtain, based on the target time-frequency diagram and the target feature, a predicted probability of a plurality of microseismic events output by the target recognition model using a target recognition model; wherein the predicted probability is used to indicate the probability of occurrence of a corresponding microseismic event;

[0015] A determination module is used to determine a target microseismic event occurring in the coal mine area from among the multiple microseismic events according to the predicted probabilities of the multiple microseismic events.

[0016] According to another aspect of the present disclosure, an electronic device is provided, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the microseismic event identification method proposed in the above aspect of the present disclosure is implemented.

[0017] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium of computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the microseismic event identification method proposed in the above aspect of the present disclosure.

[0018] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method for identifying microseismic events proposed in the above aspect of the present disclosure is implemented.

[0019] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0020] The coal mine area is monitored to obtain the microseismic signal to be identified; the dynamic frequency slicing wavelet transform DFSWT is performed on the microseismic signal to be identified to obtain the corresponding target time-frequency diagram; the feature extraction of the microseismic signal to be identified is performed to obtain the corresponding target feature; based on the target time-frequency diagram and the target feature, the target recognition model is used to obtain the predicted probability of multiple microseismic events output by the target recognition model; wherein the predicted probability is used to indicate the probability of the corresponding microseismic event; according to the predicted probability of multiple microseismic events, the target microseismic event occurring in the coal mine area is determined from the multiple microseismic events. Thus, it is possible to automatically identify the microseismic events occurring in the coal mine area based on the target time-frequency diagram obtained by the dynamic frequency slicing wavelet transform of the microseismic signal to be identified and the target feature extracted from the feature of the microseismic signal to be identified, thereby improving the accuracy and effectiveness of microseismic event identification.

[0021] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will be obvious from the following description or learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and / or additional aspects and advantages of the present disclosure will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0023] Figure 1 A schematic diagram of the flow of a microseismic event identification method provided in the first embodiment of the present disclosure;

[0024] Figure 2 A schematic diagram of the flow of a microseismic event identification method provided in the second embodiment of the present disclosure;

[0025] Figure 3 The time domain waveform and time-frequency diagram of the microseismic signal corresponding to the hydraulic fracturing event provided by the present disclosure;

[0026] Figure 4 The time domain waveform and time-frequency diagram of the microseismic signal corresponding to the roof breaking event provided by the present disclosure;

[0027] Figure 5 The sum of the time-frequency diagram of the microseismic signal corresponding to the coal cutting event of the coal mining machine provided in the present disclosure;

[0028] Figure 6 A schematic diagram of the structure of a microseismic event identification device provided in Embodiment 3 of the present disclosure;

[0029] Figure 7 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown. DETAILED DESCRIPTION

[0030] Embodiments of the present disclosure are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0031] The microseismic event identification method and device according to the embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0032] Figure 1 This is a flow chart of the microseismic event identification method provided in the first embodiment of the present disclosure.

[0033] The embodiment of the present disclosure takes the microseismic event recognition method configured in a microseismic event recognition device as an example. The microseismic event recognition device can be applied to any electronic device so that the electronic device can perform a microseismic event recognition function.

[0034] Among them, the electronic device can be any device with computing capabilities, such as a personal computer (PC), an industrial computer, a host computer, a mobile terminal, a server, etc. The mobile terminal can be, for example, a mobile phone, a tablet computer, a personal digital assistant, a wearable device, and other hardware devices with various operating systems, touch screens and / or display screens.

[0035] like Figure 1 As shown, the microseismic event identification method may include the following steps:

[0036] Step 101: monitor the coal mine area to obtain microseismic signals to be identified.

[0037] As a possible implementation method, an acoustic sensor can be used to monitor the coal mine area to be monitored in real time, so as to collect the microseismic signals to be identified in the coal mine area through the acoustic sensor. The acoustic sensor can be, for example, a three-component geophone, etc., and the present disclosure does not limit this.

[0038] Step 102: Perform dynamic frequency slice wavelet transform (DFSWT) on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram.

[0039] As a possible implementation method, the following steps can be used to obtain the target video image corresponding to the microseismic signal to be identified:

[0040] Step 1021 , performing fast Fourier transform on the microseismic signal to be identified to obtain the main frequency value of the microseismic signal to be identified.

[0041] Specifically, a fast Fourier transform (FFT) is performed on the microseismic signal to be identified to convert the microseismic signal to be identified from the time domain to the frequency domain, that is, to obtain a frequency spectrum of the microseismic signal to be identified, and then based on the frequency spectrum of the microseismic signal to be identified, the main frequency value of the microseismic signal to be identified, that is, the frequency with the strongest signal energy, is determined.

[0042] Step 1022: filter the microseismic signal to be identified to obtain a target microseismic signal.

[0043] As a possible implementation method, a target filtering range can be set based on the main frequency value, and the filter with the set target filtering range is used to filter the microseismic signal to be identified, thereby obtaining the target microseismic signal. As an example, assuming that the main frequency value is f m , set the target filter range to 0~4f mThe low-pass filter of the target filtering range is used to filter the microseismic signal to be identified, and the frequencies greater than 4f in the microseismic signal to be identified can be removed. m noise signal to obtain the target microseismic signal.

[0044] As another possible implementation method, first, empirical mode decomposition (EMD) can be performed on the microseismic signal to be identified to obtain multiple IMF (Intrinsic Mode Function) components and residual signals; secondly, multiple target IMF components are determined from the multiple IMF components; thirdly, the multiple target intrinsic mode function IMF components are first fused to obtain a first fused signal, and the intrinsic mode function IMF components other than the target intrinsic mode function IMF components among the multiple intrinsic mode function IMF components are second fused with the residual signal to obtain a second fused signal; finally, based on the second fused signal, signal extraction is performed on the first fused signal to obtain the target microseismic signal.

[0045] In the embodiment of the present application, multiple target IMF components can be determined from multiple IMF components. It should be noted that the present application does not limit the number of the determined target IMF components.

[0046] In order to determine the target IMF component from multiple IMF components, as a possible implementation method, the Pearson correlation coefficient algorithm can be used to determine the correlation coefficient between each IMF component in the multiple IMF components and the microseismic signal to be identified; the multiple IMF components are sorted in descending order according to the corresponding correlation coefficients to obtain a sorted sequence; and the IMF component in the sorted sequence whose sorting number is less than or equal to the set threshold is determined as the target IMF component.

[0047] As an example, assuming that the IMF components include IMF1, IMF2, IMF3, IMF4, and IMF5, and the threshold is set to 2, the Pearson correlation coefficient algorithm is used to determine the correlation coefficients between the above IMF components and the microseismic signals to be identified, which are 0.7, 0.2, 0.5, 0.6, and 0.9 respectively. Then, the above IMF components are arranged in descending order according to the correlation coefficients, and the resulting sorting sequence is:

[0048] IMF5, IMF1, IMF4, IMF3, IMF2;

[0049] The IMF components with sorting numbers less than or equal to 2 in the sorting sequence, namely, IMF5 and IMF1, are determined as target components.

[0050] In an embodiment of the present application, a plurality of target intrinsic mode function IMF components may be first fused to obtain a first fused signal, and the intrinsic mode function IMF components other than the target intrinsic mode function IMF components among the plurality of intrinsic mode function IMF components may be second fused with the residual signal to obtain a second fused signal. As an example, a plurality of target intrinsic mode function IMF components are superimposed to obtain a first fused signal, and the intrinsic mode function IMF components other than the target intrinsic mode function IMF components among the plurality of intrinsic mode function IMF components are superimposed with the residual signal to obtain a second fused signal.

[0051] In the embodiment of the present application, the first fusion signal can be extracted based on the second fusion signal to obtain the target microseismic signal. As an example, a subtraction operation is performed on the first fusion signal and the second fusion signal, for example, the first fusion signal is subtracted from the second fusion signal to obtain the target microseismic signal.

[0052] Step 1023, determining the target frequency based on the main frequency value.

[0053] The target frequency may indicate the highest frequency allowed to be observed.

[0054] In the embodiment of the present disclosure, the target frequency can be determined based on the main frequency value. As an example, the main frequency value of the set multiple is determined as the target frequency. For example, assuming that the set multiple is 2 and the main frequency value is f m , then the target frequency can be 2f m .

[0055] Step 1024 , based on the target frequency, and the set frequency resolution and time-frequency resolution coefficient, a dynamic frequency slicing wavelet transform DFSWT is performed on the target microseismic signal to obtain a target time-frequency diagram of the microseismic signal to be identified.

[0056] Among them, the frequency resolution and the video resolution coefficient are both pre-set, and the present disclosure does not limit their values.

[0057] Step 103: extract features from the microseismic signal to be identified to obtain corresponding target features.

[0058] The target characteristics may include, for example, amplitude, duration, and energy, etc., which are not limited in the present disclosure.

[0059] In the embodiment of the present disclosure, the microseismic signal to be identified may be analyzed to extract the target feature from the microseismic signal to be identified.

[0060] Step 104 , based on the target time-frequency diagram and the target features, a target recognition model is used to obtain the predicted probabilities of multiple microseismic events output by the target recognition model.

[0061] The predicted probability is used to indicate the probability of the corresponding microseismic event occurring.

[0062] In the embodiments of the present disclosure, microseismic events may include, for example, hydraulic fracturing events, roof break events, coal mining machine cutting events, underground drilling events, ground truck noise events, distant mine earthquake events, etc., and the present disclosure does not impose any limitations on this.

[0063] In the embodiments of the present disclosure, a target recognition model can be used based on the target time-frequency diagram and target features to obtain the predicted probabilities of multiple microseismic events output by the target recognition model. For example, the target time-frequency diagram and target features can be input into the target recognition model to obtain the predicted probabilities of multiple microseismic events output by the target recognition model.

[0064] Step 105: determining a target microseismic event occurring in the coal mine area from among the multiple microseismic events according to the predicted probabilities of the multiple microseismic events.

[0065] In the embodiment of the present disclosure, the target microseismic event may indicate a microseismic event occurring in a coal mine area.

[0066] It should be noted that the target microseismic event may be one or more, and the present disclosure does not impose any limitation on this.

[0067] In the disclosed embodiment, the target fault type occurring in the coal mine area can be determined from the multiple microseismic events according to the predicted probabilities of the multiple microseismic events.

[0068] As a possible implementation method, a target probability may be determined from the predicted probabilities of multiple microseismic events; and a microseismic event corresponding to the target probability may be determined as a target microseismic event corresponding to the microseismic signal to be identified.

[0069] Among them, in order to determine the target probability among the predicted probabilities of multiple microseismic events, as an example, the microseismic event with the largest predicted probability can be used as the target fault type occurring in the coal mine area; as another example, the microseismic event with a predicted probability greater than a set probability threshold can be used as the target fault type occurring in the coal mine area. Thus, the determination of the target probability can be achieved, so that the target microseismic event can be determined according to the target probability in the future.

[0070] The microseismic event identification method of the disclosed embodiment monitors the coal mine area to obtain the microseismic signal to be identified; performs dynamic frequency slicing wavelet transform DFSWT on the microseismic signal to be identified to obtain the corresponding target time-frequency diagram; performs feature extraction on the microseismic signal to be identified to obtain the corresponding target feature; based on the target time-frequency diagram and the target feature, a target recognition model is used to obtain the predicted probability of multiple microseismic events output by the target recognition model; wherein the predicted probability is used to indicate the probability of the occurrence of the corresponding microseismic event; and according to the predicted probability of multiple microseismic events, the target microseismic event occurring in the coal mine area is determined from the multiple microseismic events. Thus, it is possible to automatically identify the microseismic events occurring in the coal mine area according to the target time-frequency diagram obtained by the dynamic frequency slicing wavelet transform of the microseismic signal to be identified and the target feature extracted from the feature of the microseismic signal to be identified, thereby improving the accuracy and effectiveness of microseismic event identification.

[0071] In order to clearly explain how to obtain the target recognition model in the above embodiments of the present disclosure, the present disclosure also proposes a microseismic event recognition method.

[0072] Figure 2 This is a flow chart of the microseismic event identification method provided in the second embodiment of the present disclosure.

[0073] like Figure 2 As shown, based on the above embodiments of the present disclosure, the microseismic event identification method may further include the following steps:

[0074] Step 201 : for any microseismic event among a plurality of microseismic events, a plurality of sample microseismic signals of the microseismic event are obtained.

[0075] As an example, during the time window period, a three-component detector array set up on the ground in the coal mine area can be used to collect sample microseismic signals corresponding to any microseismic event among multiple microseismic events, wherein the number of sample microseismic signals corresponding to any microseismic event is greater than 1,000.

[0076] Step 202: Obtain a training time-frequency diagram and training features of each sample microseismic signal, and construct a feature data set based on the training time-frequency diagram and training features of each sample microseismic signal.

[0077] In the disclosed embodiments, for any sample microseismic signal, DFSWT can be performed on the sample microseismic signal to obtain a training time-frequency diagram of the sample microseismic signal, and feature extraction can be performed on the sample microseismic signal to obtain training features of the sample microseismic signal; thus, a feature data set can be constructed based on the training time-frequency diagram and training features of each sample microseismic signal.

[0078] Step 203 , dividing the feature data set into a first training set, a first validation set, and a first test set according to the order of sampling time of each sample microseismic signal.

[0079] As an example, the feature data set may be divided into a first training set, a first validation set, and a first test set according to the order of sampling time of each sample microseismic signal and a set division ratio, wherein the set division ratio is, for example, 6:2:2.

[0080] Step 204, training, verifying and testing the initial recognition model based on the first training set, the first verification set and the first test set, respectively, to obtain a target recognition model.

[0081] In a possible implementation of the embodiment of the present disclosure, for any microseismic event among multiple microseismic events, multiple sample microseismic signals of the microseismic event are obtained; a sub-feature data set corresponding to the microseismic event is constructed based on the training time-frequency diagram and training features of each sample microseismic signal under the microseismic event; according to the sequence of sampling times of each sample microseismic signal in the sub-feature data set corresponding to the microseismic event, the corresponding sub-feature data set is divided into corresponding sub-training sets, sub-validation sets and sub-test sets; the sub-training sets, sub-validation sets and sub-test sets corresponding to each microseismic event are fused in turn to obtain a second training set, a second validation set and a second test set; finally, based on the second training set, the second validation set and the second test set, the initial recognition model is trained, validated and tested to obtain a target recognition model.

[0082] For example, assuming that the microseismic events include event A, event B, and event C, for event A, multiple sample microseismic signals of event A are obtained, and a sub-feature data set a corresponding to event A is constructed based on the training time-frequency diagram and training features of each sample microseismic signal under event A; then, according to the order of sampling time of each sample microseismic signal in the sub-feature data set a corresponding to event A, the sub-feature data set a is divided into sub-training set a1, sub-validation set a2, and sub-test set a3; and so on, for event B, sub-training set b1, sub-validation set b2, and Sub-test set b3, for event C, obtain sub-training set c1, sub-validation set c2 and sub-test set c3; fuse sub-training set a1, sub-training set b1 and sub-training set c1 to obtain the training set, and fuse sub-training set a2, sub-training set b2 and sub-training set c2 to obtain the validation set, and fuse sub-training set a3, sub-training set b3 and sub-training set c3 again to obtain the test set; finally, the initial recognition model can be trained, verified and tested based on the training set, validation set and test set respectively to obtain the target recognition model.

[0083] In this way, the data in the acquired training set, validation set, and test set can be evenly distributed, which helps to improve the accuracy and adaptability of the trained target recognition model.

[0084] Wherein, based on the training set, training the initial recognition model may include the following steps:

[0085] Step 2041, for any statistical feature in the training set, input the training time-frequency graph in the statistical feature into the first network model in the initial recognition model to obtain the image feature output by the first network model; wherein the statistical feature includes the training time-frequency graph and training feature corresponding to the sample microseismic signal.

[0086] The first network model may be, for example, a sliding window Transformer network.

[0087] In an embodiment of the present disclosure, when the statistical features include training time-frequency graphs and training features corresponding to sample microseismic signals, for any statistical feature in the training set, the training time-frequency graph in the statistical features can be input into the first network model in the initial recognition model, thereby obtaining the image features of the training time-frequency graph output by the first network model.

[0088] It should be noted that the image feature may be a one-dimensional feature, or a two-dimensional feature, etc., and the present disclosure does not impose any limitation on this.

[0089] Step 2042, concatenate the training features and the image features in the statistical features to obtain concatenated features.

[0090] As an example, the training features and image features in the statistical features may be spliced ​​to obtain spliced ​​features of a left-right structure or a top-bottom structure.

[0091] Step 2043: input the splicing features into the second network model to obtain output probabilities of multiple microseismic events output by the second network model.

[0092] Among them, the output probability is used to represent the probability of the corresponding microseismic event occurring.

[0093] As an example, the second network model may include a target network and a Softmax layer. The target network may be, for example, a DNN (Deep Neural Network), and the target network may include an input layer, two hidden layers, and an output layer, and the number of neurons in each layer is 4, 16, 16, and 6, respectively, and the output layer is connected to the Softmax layer.

[0094] In the disclosed embodiment, the splicing feature may be input into the second network model, thereby obtaining the output probabilities of multiple microseismic events output by the second network model.

[0095] Step 2044: Determine a predicted microseismic event corresponding to the statistical feature from the multiple microseismic events according to the output probabilities of the multiple microseismic events.

[0096] As an example, the microseismic event with the highest output probability may be used as the predicted microseismic event corresponding to the statistical feature.

[0097] As another example, a microseismic event with an output probability greater than a set probability threshold may be regarded as a predicted microseismic event corresponding to the statistical feature.

[0098] Step 2045 , training the initial recognition model based on the annotated microseismic events of each statistical feature and the corresponding predicted microseismic events.

[0099] In the embodiment of the present disclosure, the labeled microseismic event may be used to indicate the microseismic event to which the corresponding sample microseismic signal belongs.

[0100] It is understandable that the microseismic events of each sample microseismic signal can be obtained, so that the microseismic events to which the statistical features corresponding to each sample microseismic signal belong can be determined, and based on the microseismic events to which the statistical features belong, the statistical features can be labeled to obtain the labeled microseismic events of the statistical features.

[0101] As an example, the statistical features of the sample microseismic signals may be annotated manually to obtain annotated microseismic events of the statistical features.

[0102] As another example, a machine annotation method may be used to annotate the statistical features of the sample microseismic signals to obtain annotated microseismic events with statistical features.

[0103] In the disclosed embodiment, the initial recognition model may be trained based on the annotated microseismic events of various statistical features and the corresponding predicted microseismic events.

[0104] In a possible implementation of the embodiment of the present disclosure, a first loss function is used to determine a loss value based on the annotated microseismic events of each statistical feature and the corresponding predicted microseismic events; and an initial recognition model is trained based on the loss value.

[0105] Optionally, the first loss function may be a cross entropy function.

[0106] As an example, a target loss function can be used to determine a loss value based on the difference between the labeled microseismic events of each statistical feature and the corresponding predicted microseismic events, and the model parameters in the initial identification model can be adjusted based on the loss value to minimize the loss value.

[0107] It should be noted that the above only uses the termination condition of model training as the minimization of the loss value as an example. In actual application, other termination conditions can also be set, such as the number of training times reaching a set number of times, the training time reaching a set time, the loss value convergence, etc. The present disclosure does not impose any restrictions on this.

[0108] In a possible implementation of the embodiment of the present disclosure, the parameters of the initial recognition model are adjusted using a stochastic gradient descent method based on the differences between the labeled microseismic events of each statistical feature and the corresponding predicted microseismic events, as well as a second loss function.

[0109] The microseismic event identification method of the disclosed embodiment obtains multiple sample microseismic signals of a microseismic event for any microseismic event among multiple microseismic events; obtains the training time-frequency diagram and training features of each sample microseismic signal, and constructs a feature data set based on the training time-frequency diagram and training features of each sample microseismic signal; divides the feature data set into a first training set, a first verification set, and a first test set according to the order of sampling time of each sample microseismic signal; and trains, verifies, and tests the initial recognition model based on the first training set, the first verification set, and the first test set, respectively, to obtain a target recognition model. Thus, by constructing a training set, a verification set, and a test set, the initial recognition model is trained, verified, and tested in turn, thereby achieving the acquisition of a target recognition model.

[0110] In order to clearly illustrate the microseismic event identification method disclosed in the present invention, the above process is described in detail below with reference to examples.

[0111] As an example, the microseismic event identification method disclosed in the present invention may include the following steps:

[0112] 1. Model training

[0113] 1.1 During the time window, the three-component geophone array on the ground in the mining area is used to collect microseismic signals caused by six types of events, namely hydraulic fracturing events, roof breakage events, coal cutting events by coal mining machines, underground drilling events, ground truck noise events and distant mine earthquake events. Invalid channel signals caused by poor coupling or instrument failure are removed. There are no less than 1,000 microseismic signals of each type, and each of the above microseismic signals is used as a sample microseismic signal;

[0114] 1.2 For any sample microseismic signal, perform DFSWT processing on the sample microseismic signal; specifically:

[0115] 1.2.1 Perform fast Fourier transform on the sample microseismic signal to obtain its main frequency value f m , and round up the main frequency value to get the target main frequency value f main ;

[0116] 1.2.2 Filter the sample microseismic signal to obtain a first microseismic signal;

[0117] As a possible implementation method, empirical mode decomposition is performed on the sample microseismic signal to obtain multiple first IMF components and a first residual signal; multiple second IMF components are determined from the multiple first IMF components; the multiple second IMF components are subjected to a third fusion to obtain a third fusion signal, and the first IMF components except the second IMF components in the multiple first IMF components are subjected to a fourth fusion with the first residual signal to obtain a fourth fusion signal; based on the fourth fusion signal, signal extraction is performed on the third fusion signal to obtain the first microseismic signal.

[0118] 1.2.3 Based on the target main frequency value, determine the frequency observation range as 0~2f main Based on the frequency observation range, the set frequency resolution ratio (i.e. 0.025), the time-frequency resolution coefficient (i.e. ), perform DFSWT transformation on the first microseismic signal to obtain the time-frequency diagram of the sample microseismic signal.

[0119] As an example, the time domain waveform of the microseismic signal corresponding to the hydraulic fracturing event is as follows: Figure 3 As shown in (a), the corresponding time-frequency diagram is as follows Figure 3 As shown in (b); the time domain waveform of the microseismic signal corresponding to the roof breaking event is as follows Figure 4 As shown in (a), the corresponding time-frequency diagram is as follows Figure 4 As shown in (b); the time domain waveform of the microseismic signal corresponding to the coal cutting event of the coal mining machine is as follows Figure 5 As shown in (a), the corresponding time-frequency diagram is as follows Figure 5 As shown in (b).

[0120] 1.3 Extract features of the sample microseismic signals to obtain features of the sample microseismic signals, where the features include amplitude, duration and energy.

[0121] 1.4 The time-frequency diagram and corresponding features of each sample microseismic signal are used to construct a cross-modal feature data set (referred to as a feature data set in this disclosure);

[0122] 1.5 According to the order of sampling time of each sample microseismic signal, the cross-modal feature data set is divided into a training set, a validation set, and a test set in a ratio of 6:2:2;

[0123] 1.6 Train the cross-modal recognition model, specifically:

[0124] 1.6.1 For any time-frequency graph in the training set, input the time-frequency graph into the sliding window Transformer network model to obtain its one-dimensional feature quantity (referred to as image feature in this disclosure);

[0125] 1.6.2 The characteristics of the sample microseismic signal corresponding to the time-frequency diagram and its one-dimensional characteristic quantity are input into a deep neural network (DNN). The DNN consists of an input layer, two hidden layers and an output layer. The number of neurons in each network layer is 4, 16, 16 and 6 respectively.

[0126] 1.6.3 The Softmax layer connected to the output layer of the DNN is used to output the output probabilities of the above six types of microseismic events, and the microseismic event corresponding to the maximum value of the output probabilities of the six types of microseismic events is determined as the predicted microseismic event.

[0127] 1.6.4 Based on the difference between the predicted microseismic events and the annotated microseismic events corresponding to each sample microseismic signal, the initial recognition model is trained. As an example, the hyperparameters of the cross-modal recognition model can be continuously adjusted by the decrease of the loss function on the training set and the validation set, and finally the batch sample number batch_size is determined to be 32, the number of iterations epoch is 100, and the learning rate is set to 0.003, where the loss function can use the cross entropy function and the optimizer is the stochastic gradient descent (SGD) method.

[0128] 1.6.5 After the cross-modal recognition model is trained, the network parameters are solidified, and the test set is used to predict the output microseismic signal category. The accuracy and precision of multi-source microseismic signal recognition are determined by combining the annotated microseismic event statistical analysis model of each sample microseismic signal in the test set.

[0129] 2. Model Application

[0130] The microseismic signal to be identified is obtained, and the dynamic frequency slicing wavelet transform (DFSWT) is performed on the microseismic signal to be identified to obtain the corresponding target time-frequency diagram; the feature of the microseismic signal to be identified is extracted to obtain the corresponding target feature; based on the target time-frequency diagram and the target feature, a target recognition model is used to obtain the predicted probabilities of multiple microseismic events output by the target recognition model; wherein the predicted probability is used to indicate the probability of the corresponding microseismic event occurring; according to the predicted probabilities of multiple microseismic events, the target microseismic event corresponding to the microseismic signal to be identified is determined from the multiple microseismic events.

[0131] The microseismic event identification method disclosed in the present invention performs time-frequency analysis on microseismic signals through dynamic frequency slicing wavelet transform, dynamically adjusts the observation frequency range according to the main frequency of the signal, and obtains a two-dimensional time-frequency diagram that can accurately reflect the characteristics of multi-source microseismic signals in the process of hydraulic fracturing of the coal seam roof; based on a data set constructed from multi-source microseismic signals measured on-site in the coal mine, the network parameters of a reasonable cross-modal multi-source microseismic signal identification model are determined, and the one-dimensional characteristics and two-dimensional image characteristics of the microseismic signals can be learned simultaneously, ensuring the actual generalization ability of microseismic signal identification; the accuracy of multi-source microseismic signal identification in the process of hydraulic fracturing of the coal seam roof is better than other methods, reaching more than 95%.

[0132] With the above Figure 1 to Figure 2 Corresponding to the microseismic event identification method provided in the embodiment, the present disclosure also provides a microseismic event identification device. Figure 1 to Figure 2 The microseismic event identification method provided in the embodiment corresponds to the microseismic event identification method, so the implementation of the microseismic event identification method is also applicable to the microseismic event identification device provided in the embodiment of the present disclosure, and will not be described in detail in the embodiment of the present disclosure.

[0133] Figure 6 This is a schematic diagram of the structure of a microseismic event identification device provided in Embodiment 3 of the present disclosure.

[0134] like Figure 6 As shown, the microseismic event identification device 600 may include: a monitoring module 601 , a transformation module 602 , an extraction module 603 , a processing module 604 and a determination module 605 .

[0135] The monitoring module 601 is used to monitor the coal mine area to obtain microseismic signals to be identified.

[0136] The transformation module 602 is used to perform dynamic frequency slicing wavelet transform (DFSWT) on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram.

[0137] The extraction module 603 is used to extract features from the microseismic signal to be identified to obtain corresponding target features.

[0138] The processing module 604 is used to obtain the predicted probabilities of multiple microseismic events output by the target recognition model based on the target time-frequency diagram and the target characteristics, using the target recognition model; wherein the predicted probability is used to indicate the probability of the corresponding microseismic event occurring.

[0139] The determination module 605 is used to determine a target microseismic event occurring in the coal mine area from among the multiple microseismic events according to the predicted probabilities of the multiple microseismic events.

[0140] In a possible implementation of the embodiment of the present disclosure, the microseismic event identification device 600 may also include a first acquisition module, wherein the first acquisition module is used to: for any microseismic event, obtain multiple sample microseismic signals of the microseismic event; obtain the training time-frequency diagram and training features of each sample microseismic signal, and construct a feature data set based on the training time-frequency diagram and training features of each sample microseismic signal; divide the feature data set into a first training set, a first verification set and a first test set according to the sequence of sampling times of each sample microseismic signal; and train, verify and test the initial recognition model based on the first training set, the first verification set and the first test set, respectively, to obtain a target recognition model.

[0141] In a possible implementation of the embodiment of the present disclosure, the first acquisition module is used to: for any statistical feature in the training set, input the training time-frequency diagram in the statistical feature into the first network model in the initial recognition model to obtain the image feature output by the first network model; wherein the statistical feature includes the training time-frequency diagram and training features corresponding to the sample microseismic signal; splice the training features and image features in the statistical feature to obtain the spliced ​​features; input the spliced ​​features into the second network model to obtain the output probabilities of multiple microseismic events output by the second network model; determine the predicted microseismic event corresponding to the statistical feature from the multiple microseismic events according to the output probabilities of the multiple microseismic events; and train the initial recognition model based on the labeled microseismic events of each statistical feature and the corresponding predicted microseismic events.

[0142] In a possible implementation of the embodiment of the present disclosure, the first acquisition module is used to: determine the loss value by using the first loss function based on the annotated microseismic events of each statistical feature and the corresponding predicted microseismic events; and train the initial recognition model based on the loss value.

[0143] In a possible implementation of the embodiment of the present disclosure, the first acquisition module is used to adjust the parameters of the initial recognition model using the stochastic gradient descent method based on the differences between the labeled microseismic events of each statistical feature and the corresponding predicted microseismic events, and the second loss function.

[0144] In a possible implementation of the embodiment of the present disclosure, the microseismic event identification device 600 may also include a second acquisition module, wherein the second acquisition module is used to: for any microseismic event, obtain multiple sample microseismic signals of the microseismic event; construct a sub-feature data set corresponding to the microseismic event based on the training time-frequency diagram and training features of each sample microseismic signal under the microseismic event; divide the corresponding sub-feature data set into corresponding sub-training sets, sub-validation sets and sub-test sets according to the sequence of sampling times of each sample microseismic signal in the sub-feature data set corresponding to the microseismic event; fuse each sub-training set, each sub-validation set and each sub-test set in turn to obtain a second training set, a second validation set and a second test set; and train, validate and test the initial recognition model based on the second training set, the second validation set and the second test set, respectively, to obtain a target recognition model.

[0145] In a possible implementation of the embodiment of the present disclosure, the transformation module 602 is used to: perform fast Fourier transform on the microseismic signal to be identified to obtain the main frequency value of the microseismic signal to be identified; perform filtering processing on the microseismic signal to be identified to obtain the target microseismic signal; determine the target frequency based on the main frequency value; perform dynamic frequency slicing wavelet transform DFSWT on the target microseismic signal based on the target frequency and the set frequency resolution and time-frequency resolution coefficient to obtain the target time-frequency diagram of the microseismic signal to be identified.

[0146] In a possible implementation of the embodiment of the present disclosure, the transformation module 602 is used to: perform empirical mode decomposition on the microseismic signal to be identified to obtain multiple intrinsic mode function IMF components and a residual signal; determine multiple target intrinsic mode function IMF components from the multiple intrinsic mode function IMF components; perform a first fusion on the multiple target intrinsic mode function IMF components to obtain a first fusion signal; perform a second fusion on the intrinsic mode function IMF components other than the target intrinsic mode function IMF components among the multiple intrinsic mode function IMF components with the residual signal to obtain a second fusion signal; and perform signal extraction on the first fusion signal based on the second fusion signal to obtain a target microseismic signal.

[0147] In a possible implementation of the embodiment of the present disclosure, the determination module 605 is used to: determine a target probability from the predicted probabilities of multiple microseismic events; and determine the microseismic event corresponding to the target probability as the target microseismic event corresponding to the microseismic signal to be identified.

[0148] The microseismic event identification device of the disclosed embodiment monitors the coal mine area to obtain the microseismic signal to be identified; performs dynamic frequency slicing wavelet transform DFSWT on the microseismic signal to be identified to obtain the corresponding target time-frequency diagram; performs feature extraction on the microseismic signal to be identified to obtain the corresponding target feature; based on the target time-frequency diagram and the target feature, a target recognition model is used to obtain the predicted probability of multiple microseismic events output by the target recognition model; wherein the predicted probability is used to indicate the probability of the occurrence of the corresponding microseismic event; and according to the predicted probability of multiple microseismic events, the target microseismic event occurring in the coal mine area is determined from the multiple microseismic events. Thus, it is possible to automatically identify the microseismic events occurring in the coal mine area according to the target time-frequency diagram obtained by the dynamic frequency slicing wavelet transform of the microseismic signal to be identified and the target feature extracted from the feature of the microseismic signal to be identified, thereby improving the accuracy and effectiveness of microseismic event identification.

[0149] In order to implement the above embodiments, the present disclosure also proposes an electronic device, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the microseismic event identification method proposed in any of the aforementioned embodiments of the present disclosure is implemented.

[0150] In order to implement the above embodiments, the present disclosure further proposes a non-temporary computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the microseismic event identification method proposed in any of the above embodiments of the present disclosure is implemented.

[0151] In order to implement the above embodiments, the present disclosure further provides a computer program product. When instructions in the computer program product are executed by a processor, the microseismic event identification method proposed in any of the above embodiments of the present disclosure is executed.

[0152] like Figure 7 As shown, the electronic device 12 is in the form of a general purpose computing device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that connects various system components (including the system memory 28 and the processing unit 16).

[0153] The bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor or a local bus using any of a variety of bus structures. For example, these architectures include but are not limited to Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus and Peripheral Component Interconnection (PCI) bus.

[0154] The electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.

[0155] The memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 7 not shown, usually called a "hard drive"). Although Figure 7Not shown in the figure, a disk drive for reading and writing a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing a removable non-volatile optical disk (e.g., a compact disc read only memory (CD-ROM), a digital versatile disc read only memory (DVD-ROM), or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the various embodiments of the present disclosure.

[0156] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in the memory 28, such program modules 42 including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 42 generally perform the functions and / or methods of the embodiments described in the present disclosure.

[0157] The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 12, and / or may communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 12 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via a bus 18. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0158] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the methods mentioned in the above embodiments.

[0159] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0160] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0161] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present disclosure includes additional implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present disclosure belong.

[0162] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0163] It should be understood that the various parts of the present disclosure can be implemented in hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0164] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0165] In addition, each functional unit in each embodiment of the present disclosure may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0166] The storage medium mentioned above may be a read-only memory, a disk or an optical disk, etc. Although the embodiments of the present disclosure have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present disclosure. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present disclosure.

Claims

1. A microseismic event identification method, characterized in that: The method comprises: Monitor the coal mine area to obtain microseismic signals to be identified; Performing dynamic frequency slicing wavelet transform (DFSWT) on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram; Extracting features of the microseismic signal to be identified to obtain corresponding target features; Based on the target time-frequency diagram and the target features, a target recognition model is used to obtain the predicted probabilities of multiple microseismic events output by the target recognition model; wherein the predicted probabilities are used to indicate the probability of occurrence of corresponding microseismic events; According to the predicted probabilities of the multiple microseismic events, a target microseismic event occurring in the coal mine area is determined from the multiple microseismic events.

2. The method according to claim 1, characterized in that The acquisition of the target recognition model includes: For any of the microseismic events, obtaining a plurality of sample microseismic signals of the microseismic event; Acquire a training time-frequency diagram and training features of each of the sample microseismic signals, and construct a feature data set based on the training time-frequency diagram and training features of each of the sample microseismic signals; According to the order of sampling time of each of the sample microseismic signals, the feature data set is divided into a first training set, a first validation set and a first test set; The initial recognition model is trained, verified and tested based on the first training set, the first verification set and the first test set respectively to obtain the target recognition model.

3. The method according to claim 2, characterized in that The step of training the initial recognition model based on the training set includes: For any statistical feature in the training set, the training time-frequency graph in the statistical feature is input into the first network model in the initial recognition model to obtain the image feature output by the first network model; wherein the statistical feature includes the training time-frequency graph and the training feature corresponding to the sample microseismic signal; Splicing the training features in the statistical features and the image features to obtain spliced ​​features; Inputting the splicing feature into a second network model to obtain output probabilities of a plurality of microseismic events output by the second network model; Determining a predicted microseismic event corresponding to the statistical feature from the multiple microseismic events according to the output probabilities of the multiple microseismic events; The initial recognition model is trained based on the annotated microseismic events of each of the statistical features and the corresponding predicted microseismic events.

4. The method according to claim 3, characterized in that The training of the initial recognition model based on the annotated microseismic events of each of the statistical features and the corresponding predicted microseismic events includes: Based on the annotated microseismic events of the statistical characteristics and the corresponding predicted microseismic events, a first loss function is used to determine a loss value; Based on the loss value, the initial recognition model is trained.

5. The method according to claim 3, characterized in that: The training of the initial recognition model based on the annotated microseismic events of each of the statistical features and the corresponding predicted microseismic events includes: Based on the difference between the annotated microseismic events of each of the statistical features and the corresponding predicted microseismic events, and the second loss function, the parameters of the initial recognition model are adjusted using a stochastic gradient descent method.

6. The method according to claim 1, characterized in that The acquisition of the target recognition model includes: For any of the microseismic events, obtaining a plurality of sample microseismic signals of the microseismic event; Based on the training time-frequency diagram and training features of each of the sample microseismic signals under the microseismic event, construct a sub-feature data set corresponding to the microseismic event; According to the order of sampling time of each of the sample microseismic signals in the sub-feature data set corresponding to the microseismic event, the corresponding sub-feature data set is divided into a corresponding sub-training set, a sub-validation set and a sub-test set; The sub-training sets, the sub-validation sets and the sub-test sets are sequentially merged to obtain a second training set, a second validation set and a second test set; Based on the second training set, the second validation set and the second test set, the initial recognition model is trained, validated and tested respectively to obtain the target recognition model.

7. The method according to claim 1, characterized in that The step of performing a dynamic frequency slicing wavelet transform (DFSWT) on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram includes: Performing a fast Fourier transform on the microseismic signal to be identified to obtain a main frequency value of the microseismic signal to be identified; Filtering the microseismic signal to be identified to obtain a target microseismic signal; Based on the main frequency value, determining a target frequency; Based on the target frequency, and the set frequency resolution and time-frequency resolution coefficient, a dynamic frequency slicing wavelet transform (DFSWT) is performed on the target microseismic signal to obtain a target time-frequency diagram of the microseismic signal to be identified.

8. The method according to claim 7, characterized in that The filtering process of the microseismic signal to be identified to obtain a target microseismic signal includes: Performing empirical mode decomposition on the microseismic signal to be identified to obtain a plurality of intrinsic mode function IMF components and residual signals; Determine a plurality of target intrinsic mode function IMF components from the plurality of intrinsic mode function IMF components; Performing a first fusion on the multiple target intrinsic mode function IMF components to obtain a first fusion signal; Performing a second fusion on the intrinsic mode function IMF components other than the target intrinsic mode function IMF component among the multiple intrinsic mode function IMF components and the residual signal to obtain a second fused signal; Based on the second fusion signal, signal extraction is performed on the first fusion signal to obtain the target microseismic signal.

9. The method according to any one of claims 1 to 8, characterized in that: Determining a target microseismic event occurring in the coal mine area from the multiple microseismic events according to the predicted probabilities of the multiple microseismic events includes: determining a target probability from the predicted probabilities of the plurality of microseismic events; The microseismic event corresponding to the target probability is determined as the target microseismic event corresponding to the microseismic signal to be identified.

10. A microseismic event identification device, characterized in that: The method comprises: A monitoring module is used to monitor the coal mine area to obtain microseismic signals to be identified; A transformation module, used for performing dynamic frequency slicing wavelet transform DFSWT on the microseismic signal to be identified to obtain a corresponding target time-frequency diagram; An extraction module, used for extracting features from the microseismic signal to be identified to obtain corresponding target features; A processing module, configured to obtain, based on the target time-frequency diagram and the target feature, a predicted probability of a plurality of microseismic events output by the target recognition model using a target recognition model; wherein the predicted probability is used to indicate the probability of occurrence of a corresponding microseismic event; A determination module is used to determine a target microseismic event occurring in the coal mine area from among the multiple microseismic events according to the predicted probabilities of the multiple microseismic events.

Images