Microearthquake effective event intelligent identification method, device, equipment, medium and program

Abstract

The invention relates to the technical field of microseism event identification and estimation, and discloses a microseism effective event intelligent identification method, device, equipment, medium and program, and the method comprises the steps: carrying out the data enhancement of each piece of simulated seismic data in a simulated seismic data set, and obtaining an original seismic database; performing event marking on the seismic data in the original seismic database to obtain a marked seismic database; performing model training on a pre-constructed neural network model according to the marked seismic database to obtain a seismic event recognition model; performing feature extraction on the actual micro-seismic monitoring signal by using the seismic recognition model to obtain a seismic feature curve; and identifying a corresponding earthquake effective event according to the earthquake characteristic curve. According to the invention, the accuracy of effective earthquake event identification of the actual micro-earthquake monitoring signal can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of microseismic event identification and estimation, and in particular to a method, device, equipment, medium and program for intelligently identifying effective microseismic events. Background Art

[0002] This section is intended to provide a background or context to the embodiments recited in the claims. No admission is made that the description herein is prior art by inclusion in this section.

[0003] In recent years, in order to keep up with the pace of development of the world's oil and gas industry, my country has also begun to vigorously develop the exploration and development of unconventional oil and gas. Unlike conventional oil and gas reservoirs, unconventional oil and gas reservoirs have small porosity and low permeability, and their exploitation depends on horizontal wells and fracturing technology. Among them, fracturing technology requires the use of microseismic monitoring for construction guidance, and effective event identification is a key link in microseismic monitoring. The accuracy of automatic identification of effective events determines the correctness of subsequent source positioning, which in turn affects the interpretation of the fracture network. However, conventional microseismic effective event automatic identification methods (such as STA / LTA, AR-AIC, etc.) mainly construct characteristic functions based on the different statistical characteristics of effective events and noise in the time domain or frequency domain, and realize automatic identification of effective events by optimizing the characteristic functions. However, in the actual collected microseismic data, the effective signal and the noise signal are intertwined, and it is difficult to use simple time-frequency domain information to effectively identify the complex signal characteristics of microseismic events, making it difficult for conventional effective event automatic identification methods to accurately and automatically identify microseismic effective events. Secondly, conventional microseismic effective event automatic identification methods strongly rely on the selection of key parameters such as window length and threshold function. The identification parameters of microseismic monitoring data from different regions, different acquisition schemes and different source mechanisms vary greatly. In actual processing, it is generally necessary to rely on the processing personnel to test a large number of different parameter combinations based on experience to obtain the optimal identification results, which seriously affects the speed and reliability of this type of method in identifying effective events.

[0004] With the rise of artificial intelligence technology, artificial neural networks (ANN) have also been applied to the automatic identification of effective events in microseismic monitoring. Artificial neural networks are an automatic identification algorithm with strong "fault tolerance". They use monitoring data and expected output as input and feedback of the network model, do not rely on parameter selection, have strong nonlinear fitting ability and adaptability, and can effectively identify different types of microseismic signals in strong background noise without human intervention after training. Among them, recurrent neural networks (RNN) are widely used in automatic signal recognition due to their "memory ability", but simple RNNs will produce gradient vanishing and gradient explosion during multi-layer training, so that the entire network can only remember some of the last input signals, resulting in low accuracy in the identification of effective microseismic events. Summary of the invention

[0005] In view of the above problems, embodiments of the present invention provide a method, device, equipment, medium and program for intelligently identifying effective microseismic events.

[0006] In a first aspect, an embodiment of the present invention provides a method for intelligently identifying effective microseismic events, comprising:

[0007] Acquire a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database;

[0008] Marking each earthquake data in the original earthquake database to obtain a marked earthquake database;

[0009] Constructing a neural network model using preset model parameters, and training the neural network model according to the marked earthquake database to obtain an earthquake event recognition model;

[0010] Acquire an actual microseismic monitoring signal, and use the earthquake recognition model to perform feature extraction on the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal;

[0011] The effective earthquake events corresponding to the actual microseismic monitoring signal are identified according to the earthquake characteristic curve.

[0012] According to an embodiment of the present invention, performing data enhancement of different strengths on each simulated seismic data in the simulated seismic data set to obtain an original seismic database includes:

[0013] Calculating the average energy of each simulated seismic data in the simulated seismic data set;

[0014] Pre-constructed Gaussian white noise is added to each of the simulated seismic data according to the average energy to obtain an original seismic database corresponding to the simulated seismic data set.

[0015] According to an embodiment of the present invention, the step of marking each earthquake data in the original earthquake database to obtain a marked earthquake database includes:

[0016] Normalizing each seismic data in the original seismic database to obtain standard seismic data;

[0017] The event waveform in the standard seismic data is identified, and a Hanning window is added to the event waveform to obtain a marked earthquake database.

[0018] According to an embodiment of the present invention, constructing a neural network model using preset model parameters includes:

[0019] The neural network model is shown in the following formula:

[0020] h t =H(W xh X t +W hh h t-1 +b n )

[0021] o t =W ho h t +b o

[0022]

[0023] Among them, h t represents the hidden vector at time step t, H represents the hidden layer function in the neural network model, and W Xh , W Xi , W Xf , W Xo , W Xc Respectively represent the weight matrices between the input vector X and the hidden vector h, the input gate i, the forget gate f, the output gate o and the unit activation function c of the neural network model, X t , X t-1 Respectively represent the input vector at time step t and time t-1, W hh represents the weight matrix of the hidden vector h, i t 、f t , o t 、c t Respectively represent the output gate, forget gate, output gate and output result of the unit activation vector in the neural network at time step t, σ represents the logic function, W hi , W hf , W ho , W hc Represents the weight matrix between the hidden vector h and the input gate i, forget gate f, output gate o and unit activation function c, respectively. t-1 represents the output result of the unit activation vector at time step t-1, W ci , W cf , W co Respectively represent the weight matrices between the activation function c and the input gate i, the forget gate f, and the output gate o, b i , b f , b o , b c They represent the bias vectors of input gate i, forget gate f, output gate o and unit activation function c respectively, and tanh() represents the matrix tangent function.

[0024] According to an embodiment of the present invention, the step of training the neural network model according to the labeled earthquake database to obtain an earthquake event recognition model includes:

[0025] Dividing the labeled earthquake data in the labeled earthquake database into a training set and a validation set;

[0026] Using the neural network model to make an initial prediction on the training set to obtain a predicted seismic signal, and using a preset loss function to calculate a loss value between the predicted seismic signal and the training set;

[0027] Gradient adjustment is performed on the model parameters in the neural network model according to the loss value until the number of training times for the training set reaches a preset iteration cycle, thereby obtaining an initial earthquake event recognition model;

[0028] The initial earthquake event recognition model is used to calculate the verification loss value of the verification set, and when the verification loss value is less than a preset loss threshold, an earthquake event recognition model is obtained.

[0029] According to an embodiment of the present invention, the step of extracting features from the actual microseismic monitoring signal using the earthquake recognition model to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal includes:

[0030] Calculating candidate memory unit values ​​of the actual microseismic signal using the earthquake recognition model;

[0031] Calculating the output value of the candidate memory unit value using the hidden layer in the earthquake recognition model;

[0032] An activation calculation is performed on the output value to obtain a characteristic signal corresponding to the actual microseismic monitoring signal.

[0033] In a second aspect, an embodiment of the present invention provides an intelligent identification device for effective microseismic events, characterized in that it includes:

[0034] A data enhancement module is used to obtain a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database;

[0035] An event marking module is used to mark each earthquake data in the original earthquake database to obtain a marked earthquake database;

[0036] A model training module, used to construct a neural network model using preset model parameters, and to perform model training on the neural network model according to the marked earthquake database to obtain an earthquake event recognition model;

[0037] A characteristic curve extraction module is used to obtain an actual microseismic monitoring signal, and use the earthquake recognition model to extract characteristics of the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal;

[0038] The effective event identification module is used to identify the effective earthquake events corresponding to the actual microseismic monitoring signal according to the earthquake characteristic curve.

[0039] In a third aspect, an embodiment of the present invention provides an electronic device, comprising:

[0040] processor;

[0041] a memory for storing instructions executable by the processor;

[0042] Wherein, the processor is configured to execute the instructions to implement a method for intelligently identifying effective microseismic events as described in the first aspect.

[0043] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for intelligently identifying effective microseismic events as described in the first aspect.

[0044] In a fifth aspect, an embodiment of the present invention provides a computer program, characterized in that when the program is executed by a processor, it implements a method for intelligently identifying effective microseismic events as described in the first aspect.

[0045] Compared with the prior art, the above technical solution of the present invention has the following beneficial effects:

[0046] The embodiment of the present invention performs data enhancement of different intensities on each simulated earthquake data in a simulated earthquake data set to obtain an original earthquake database, which can make the original earthquake database closer to reality and improve the applicability of the original earthquake database, thereby improving the accuracy of subsequent model training; a marked earthquake database that marks events in the original earthquake database is used to perform model training on a pre-constructed RNN+LSTM neural network model to obtain an earthquake event recognition model with more accurate feature extraction; the earthquake event recognition model extracts features from actual microseismic monitoring signals that are actually monitored, and has the ability to save, read, reset and update long-distance historical information, thereby making full use of historical time series information in neurons, effectively improving the accuracy of feature extraction, and thereby improving the accuracy of effective microseismic event recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 The flowchart of the method for intelligently identifying effective microseismic events according to the first embodiment of the present invention is shown;

[0049] Figure 2 A schematic diagram of a process of training a neural network model according to the first embodiment of the present invention is shown;

[0050] Figure 3 A schematic diagram of a process of extracting features from actual microseismic monitoring signals according to the first embodiment of the present invention is shown;

[0051] Figure 4 The functional module diagram of the device for intelligently identifying effective microseismic events according to the third embodiment of the present invention is shown;

[0052] Figure 5 A schematic diagram of the structure of an electronic device for implementing the method for intelligently identifying effective microseismic events according to the fourth embodiment of the present invention is shown. DETAILED DESCRIPTION

[0053] The present disclosure is further described below in conjunction with the embodiments shown in the accompanying drawings.

[0054] 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 invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0055] The present invention proposes an intelligent identification method for effective microseismic events. The method performs data enhancement of different intensities on each simulated seismic data in a simulated seismic data set to obtain an original seismic database, which can make the original seismic database closer to reality and improve the applicability of the original seismic database, thereby improving the accuracy of subsequent model training; a marked seismic database for event marking of the original seismic database is used to perform model training on a pre-constructed RNN+LSTM neural network model to obtain an earthquake event identification model with more accurate feature extraction; the earthquake event identification model extracts features from actual microseismic monitoring signals of actual monitoring, and has the ability to save, read, reset and update long-distance historical information, thereby making full use of historical time series information in neurons, effectively improving the accuracy of feature extraction, and thereby improving the accuracy of effective microseismic event identification.

[0056] Embodiment 1

[0057] like Figure 1 As shown, the present invention proposes a method for intelligently identifying effective microseismic events, comprising the following steps:

[0058] S1. Acquire a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database.

[0059] In an embodiment of the present invention, the simulated seismic data set is microseismic simulation data of different source mechanisms and different acquisition schemes simulated by the finite element difference method, wherein the microseismic simulation data is seismic signal data, and background noise of different intensities is added to the simulated seismic data set to obtain simulated seismic data of different signal-to-noise ratios, so as to perform data enhancement on each simulated seismic data.

[0060] In the embodiment of the present invention, performing data enhancement of different strengths on each simulated seismic data in the simulated seismic data set to obtain an original seismic database includes:

[0061] Calculating the average energy of each simulated seismic data in the simulated seismic data set;

[0062] Pre-constructed Gaussian white noise is added to each of the simulated seismic data according to the average energy to obtain an original seismic database corresponding to the simulated seismic data set.

[0063] In an embodiment of the present invention, the average energy of each simulated seismic data is the summed mean of the squared signal amplitude values ​​at each frequency point of the simulated seismic data. The average energy can also be calculated in the frequency domain. The noise of each simulated seismic data can be calculated through the average energy and noise. Then, pre-constructed Gaussian white noise is added to each simulated seismic data according to the average energy to obtain enhanced simulated seismic data with different signal-to-noise ratios, so that the original seismic database is closer to reality and the applicability of the original seismic database is improved.

[0064] In an embodiment of the present invention, actual field microseismic data collected can be added to the original seismic database and integrated with the data-enhanced simulated seismic data set to obtain an original seismic database that is closer to actual seismic data, which can improve the accuracy of subsequent model training.

[0065] S2. Mark each earthquake data in the original earthquake database to obtain a marked earthquake database.

[0066] In an embodiment of the present invention, event marking is to mark the waveform signal representing the earthquake event in the original earthquake database, and mark the remaining waveform signals as 0, as the expected signal for subsequent model training. For example, the waveform signal of earthquake data 1 contains changes in the velocity of the earthquake wave, changes in the signal amplitude, etc., which are identifications of effective earthquake events. The embodiment of the present invention can set recognition rules based on historical experience to identify the effective events of each earthquake data in the original earthquake database.

[0067] In the embodiment of the present invention, the step of marking each earthquake data in the original earthquake database to obtain a marked earthquake database includes:

[0068] Normalizing each seismic data in the original seismic database to obtain standard seismic data;

[0069] The event waveform in the standard seismic data is identified, and a Hanning window is added to the event waveform to obtain a marked earthquake database.

[0070] In an embodiment of the present invention, normalization processing is to map the waveform signal of each seismic data to a limited range to eliminate the influence of different dimensions of seismic data such as velocity, frequency, and phase on the seismic waveform signal, thereby improving the accuracy of subsequent model training.

[0071] In an embodiment of the present invention, the waveform rules corresponding to when a valid earthquake event occurs in the seismic data can be set based on historical experience. For example, a change in the signal amplitude exceeding a threshold indicates a corresponding valid earthquake event, and a change in the velocity of the seismic wave exceeding a threshold and an amplitude of the waveform signal exceeding a threshold correspond to a valid earthquake event, thereby being able to identify the valid events of each seismic data.

[0072] In the embodiment of the present invention, the Hanning window (hanning) is a window function, which is also called a raised cosine window. The Hanning window can be regarded as the sum of the spectra of three rectangular time windows, or the sum of three sinc (t) type functions, and the two items in the brackets are moved to the left and right by π / T respectively relative to the first spectrum window, so that the side lobes cancel each other out, eliminating high-frequency interference and leakage energy. The waveform signal corresponding to the effective event of the seismic data can be truncated through the Hanning window, so that the seismic data is event marked to obtain a marked seismic database.

[0073] S3. Construct a neural network model using preset model parameters, and perform model training on the neural network model according to the labeled earthquake database to obtain an earthquake event recognition model.

[0074] In an embodiment of the present invention, the neural network model is obtained by combining RNN (Recurrent Neural Networks) + LSTM (Long Short Term Memory) networks, wherein LSTM is used as the hidden layer of the RNN neural network, and the number of neurons in each hidden layer is set by preset model parameters.

[0075] In the embodiment of the present invention, the input data in the RNN is input into the network in sequence according to its time sequence, and the time sequence represents the concept of time step. In the RNN, the hidden state is extremely important. The hidden state is the intermediate value connecting each neuron in each hidden layer. In the first layer, at time step t, the RNN hidden layer neuron obtains the hidden state h t At time step t+1, the hidden vector h of the previous hidden layer is received t As the input vector, we get the new hidden vector h t+1 .

[0076] In the embodiment of the present invention, LSTM is a connection of three units: input gate, forget gate, and output gate. Each unit corresponds to a network layer of different network functions. The input gate is used to check whether there is any information input, and then the forget gate is used to determine whether to choose to forget the information memorized in the LSTM. Finally, the output gate is used to determine whether to output the information at this moment, so that the input information can be screened, and the features of the earthquake information can be accurately extracted.

[0077] In an embodiment of the present invention, the use of preset model parameters to construct a neural network model includes:

[0078] The neural network model is shown in the following formula:

[0079] h t =H(W xh X t +W hh h t-1 +b h )

[0080] o t =W ho h t +b o

[0081]

[0082] Among them, h t represents the hidden vector at time step t, H represents the hidden layer function in the neural network model, and W Xh , W Xi , W Xf , WXo , W Xc Respectively represent the weight matrices between the input vector X and the hidden vector h, the input gate i, the forget gate f, the output gate o and the unit activation function c of the neural network model, X t , X t-1 Respectively represent the input vector at time step t and time t-1, W hh represents the weight matrix of the hidden vector h, i t 、f t , o t 、c t Respectively represent the output gate, forget gate, output gate and output result of the unit activation vector in the neural network at time step t, σ represents the logic function, W hi , W hf , W ho , W hc Represents the weight matrix between the hidden vector h and the input gate i, forget gate f, output gate o and unit activation function c, respectively. t-1 represents the output result of the unit activation vector at time step t-1, W ci , W cf , W co Respectively represent the weight matrices between the activation function c and the input gate i, the forget gate f, and the output gate o, b i , b f , b o , b c They represent the bias vectors of input gate i, forget gate f, output gate o and unit activation function c respectively, and tanh() represents the matrix tangent function.

[0083] In the embodiment of the present invention, the weight matrix, bias vector and other model parameters in the neural network model need to be trained to ensure the accuracy of subsequent feature extraction.

[0084] In the embodiment of the present invention, refer to Figure 2 As shown, the neural network model is trained according to the marked earthquake database to obtain an earthquake event recognition model, including:

[0085] S21, dividing the labeled earthquake data in the labeled earthquake database into a training set and a verification set;

[0086] S22, using the neural network model to make an initial prediction on the training set to obtain a predicted seismic signal, and using a preset loss function to calculate a loss value between the predicted seismic signal and the training set;

[0087] S23, performing gradient adjustment on the model parameters in the neural network model according to the loss value, until the number of training times for the training set reaches a preset iteration cycle, thereby obtaining an initial earthquake event recognition model;

[0088] S24. Calculate the verification loss value of the verification set using the initial earthquake event recognition model, and obtain the earthquake event recognition model when the verification loss value is less than a preset loss threshold.

[0089] In an embodiment of the present invention, the marked seismic data in the marked seismic database are divided into a training set and a validation set of preset proportions, for example, divided into 80% of the training set and 20% of the validation set, to avoid overfitting in model training, and then the loss value between the predicted seismic signal and the marked seismic data is calculated by the mean_squared_error (mse) loss function to determine whether the predicted seismic signal is consistent with the earthquake event waveform marked by the Hanning window, and then the model parameters are adjusted by a gradient learning algorithm with a learning rate of 0.00001 to improve the accuracy of model training.

[0090] In an embodiment of the present invention, since there are too many labeled earthquake data in the training set, it is impossible to train the labeled earthquake data at one time. The complete process of predicting all the labeled earthquake data in the training set is called an iteration cycle, until the training number of the training set reaches a preset iteration cycle. For example, the training set is trained for 10 or 15 iteration cycles to obtain an initial earthquake event recognition model, and then the initial earthquake event recognition model is verified using the verification set to avoid over-fitting of the model and obtain an earthquake event recognition model with more accurate feature extraction.

[0091] S4. Acquire an actual microseismic monitoring signal, and use the earthquake recognition model to extract features of the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal.

[0092] In an embodiment of the present invention, the actual microseismic monitoring signal is a seismic waveform signal actually monitored in the field that requires effective seismic event identification. The earthquake identification model can be used to extract waveform features of the actual microseismic detection signal to obtain a characteristic signal corresponding to the actual microseismic monitoring signal, thereby enabling the characteristic signal to be used to identify effective seismic events.

[0093] In the embodiment of the present invention, refer to Figure 3 As shown, the feature extraction of the actual microseismic monitoring signal by using the earthquake recognition model to obtain the earthquake characteristic curve corresponding to the actual microseismic monitoring signal includes:

[0094] S31, calculating the candidate memory unit value of the actual microseismic signal by using the earthquake recognition model;

[0095] S32, using the hidden layer in the earthquake recognition model to calculate the output value of the candidate memory unit value;

[0096] S33, performing activation calculation on the output value to obtain a characteristic signal corresponding to the actual microseismic monitoring signal.

[0097] In the embodiment of the present invention, the earthquake characteristic curve can be determined by the process of constructing the neural network model. First, the candidate memory unit value at the current moment is calculated. At the same time, the input gate of the hidden layer is used to control the influence of the current data input on the memory unit state value. Then, the forget gate is used to control the influence of historical information on the current memory unit state value. Finally, the output gate is used to control the output of the memory unit state value. Then, the activation function is used to perform activation calculation to obtain the final earthquake characteristic curve output h. t , where the activation function can be the logistic sigmoid function.

[0098] In the embodiment of the present invention, by introducing a gated design, the LSTM unit in the earthquake recognition model has the ability to save, read, reset and update long-distance historical information, thereby being able to fully utilize the historical timing information of neurons in the earthquake event recognition model to further improve the accuracy of feature extraction.

[0099] S5. Identify effective earthquake events corresponding to the actual microseismic monitoring signal according to the earthquake characteristic curve.

[0100] In an embodiment of the present invention, the earthquake characteristic curve is a characteristic waveform signal corresponding to the actual microseismic monitoring signal, so that effective earthquake events are identified based on the waveform information contained in the earthquake characteristic curve, for example, effective events are identified based on characteristic waveform information such as the maximum value and average value of the curve.

[0101] In an embodiment of the present invention, identifying effective earthquake events corresponding to the actual microseismic monitoring signal according to the earthquake characteristic curve includes:

[0102] Extracting curve characteristic values ​​from the earthquake characteristic curve;

[0103] When the curve characteristic value is greater than a preset characteristic threshold, it is determined that the actual microseismic monitoring signal corresponds to a valid earthquake event.

[0104] When the curve characteristic value is not greater than a preset characteristic threshold, it is determined that the actual microseismic monitoring signal does not correspond to a valid earthquake event.

[0105] In an embodiment of the present invention, a characteristic threshold can be used to determine whether there is an effective earthquake event in the actual microseismic monitoring signal. For example, if the maximum value in the curve characteristic value is greater than a preset threshold, then the actual microseismic monitoring signal corresponds to the existence of an effective earthquake event, thereby quickly and accurately identifying the effective microseismic event.

[0106] Embodiment 2

[0107] In order to more clearly understand the present invention, a second embodiment is used below to further explain the present invention in which a neural network model is constructed using preset model parameters, and the neural network model is trained according to the labeled earthquake database to obtain an earthquake event recognition model.

[0108] In an embodiment of the present invention, the neural network model is a multi-layer CNN (Convolutional Neural Networks, CNN) convolutional neural network model, wherein the convolutional neural network model includes multiple layers of convolutional layers and fully connected layers, and different convolutional layers include convolution kernels, pooling layers and activation functions of different scales, thereby extracting feature maps of different scales, so that the extracted feature information is more comprehensive.

[0109] In the embodiment of the present invention, the neural network model is trained according to the marked earthquake database to obtain an earthquake event recognition model, including:

[0110] Using the convolution layer of the neural network model to perform data convolution and pooling processing on each labeled seismic data in the labeled seismic data, to obtain convolution feature maps of different convolution layers;

[0111] Performing a first full connection on the convolution feature map to obtain a fully connected feature map, and performing a second full connection on the fully connected feature map to obtain a predicted label for each labeled earthquake data in the labeled earthquake database;

[0112] The loss value between the predicted label and the event label corresponding to the marked earthquake database is calculated, and the parameter gradient of the neural network model is adjusted according to the loss value until the number of training times for the marked earthquake database reaches a preset iteration cycle, thereby obtaining an earthquake event recognition model.

[0113] In an embodiment of the present invention, data convolution and pooling processing are performed on the labeled earthquake data through each convolution layer to obtain convolution feature maps of different convolution layers, and the convolution feature maps of different convolution layers are fully connected to the eigenvalues ​​through the first full connection, and then the eigenvalues ​​are pre-set to the label vector through the second full connection to obtain the predicted label of each labeled earthquake data, and then the parameter gradient of the neural network model can be adjusted until the training times for each labeled earthquake data in the labeled earthquake database reaches the preset period, and the earthquake event recognition model is obtained.

[0114] In the embodiment of the present invention, feature extraction is performed through a multi-layer convolutional neural network, which can calculate the features of the labeled seismic data at different scales, and then comprehensively and accurately extract the features of each labeled seismic data, thereby improving the feature extraction accuracy of the earthquake event recognition model.

[0115] Embodiment 3

[0116] like Figure 4 As shown, this embodiment also provides a functional module diagram of a device for intelligently identifying effective microseismic events.

[0117] The microseismic effective event intelligent identification device 400 described in this embodiment can be installed in an electronic device. According to the functions to be implemented, the microseismic effective event intelligent identification device 400 may include a data enhancement module 401, an event marking module 402, a model training module 403, a characteristic curve extraction module 404 and an effective event identification module 405. The module described in the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0118] In this embodiment, the functions of each module / unit are as follows:

[0119] The data enhancement module 401 is used to obtain a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database;

[0120] The event marking module 402 is used to mark each earthquake data in the original earthquake database to obtain a marked earthquake database;

[0121] The model training module 403 is used to construct a neural network model using preset model parameters, and to perform model training on the neural network model according to the marked earthquake database to obtain an earthquake event recognition model;

[0122] The characteristic curve extraction module 404 is used to obtain the actual microseismic monitoring signal, and use the earthquake recognition model to extract the characteristics of the actual microseismic monitoring signal to obtain the earthquake characteristic curve corresponding to the actual microseismic monitoring signal;

[0123] The effective event identification module 405 is used to identify the effective earthquake events corresponding to the actual microseismic monitoring signal according to the earthquake characteristic curve.

[0124] In detail, each module described in the device 400 for intelligently identifying effective microseismic events in the embodiment of the present invention adopts the same technical means as the method for intelligently identifying effective microseismic events described in Embodiment 1 and Embodiment 2 when used, and can produce the same technical effects, which will not be repeated here.

[0125] Embodiment 4

[0126] like Figure 5 As shown, this embodiment also provides a computer electronic device, which may include a processor 501, a memory 502, a communication bus 503 and a communication interface 504, and may also include a computer program stored in the memory 502 and executable on the processor 501, such as an intelligent identification program for effective microseismic events.

[0127] In some embodiments, the processor 501 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 501 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 502 (for example, executing a microseismic effective event intelligent identification program, etc.), and calls data stored in the memory 502 to execute various functions of the electronic device and process data.

[0128] The memory 502 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 502 may be an internal storage unit of an electronic device in some embodiments, such as a mobile hard disk of the electronic device. The memory 502 may also be an external storage device of an electronic device in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 502 may also include both an internal storage unit of the electronic device and an external storage device. The memory 502 may not only be used to store application software and various types of data installed in the electronic device, such as the code of the microseismic effective event intelligent identification program, but may also be used to temporarily store data that has been output or is to be output.

[0129] The communication bus 503 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 502 and at least one processor 501, etc.

[0130] The communication interface 504 is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0131] The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0132] For example, although not shown, the electronic device may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 501 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, etc. The electronic device may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0133] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0134] The microseismic effective event intelligent identification program stored in the memory 502 in the electronic device is a combination of multiple instructions. When running in the processor 501, it can achieve:

[0135] Acquire a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database;

[0136] Marking each earthquake data in the original earthquake database to obtain a marked earthquake database;

[0137] Constructing a neural network model using preset model parameters, and training the neural network model according to the marked earthquake database to obtain an earthquake event recognition model;

[0138] Acquire an actual microseismic monitoring signal, and use the earthquake recognition model to perform feature extraction on the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal;

[0139] The effective earthquake events corresponding to the actual microseismic monitoring signal are identified according to the earthquake characteristic curve.

[0140] Specifically, the specific implementation method of the processor 501 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0141] Furthermore, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0142] Embodiment 5

[0143] This embodiment provides a storage medium storing a computer program. When the computer program is executed by a processor, the steps of the method for intelligently identifying effective microseismic events as described above are implemented.

[0144] These program codes can also be loaded onto a computer or other programmable data processing device so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions executed on the computer or other programmable device for implementing the process. Figure 1 The steps of a specified function in a process or multiple processes.

[0145] Storage media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, modules of programs or other data. Examples of storage media can include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0146] Embodiment 6

[0147] A computer program, characterized in that when the computer program is executed by a processor of an electronic device, it can achieve:

[0148] Acquire a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database;

[0149] Marking each earthquake data in the original earthquake database to obtain a marked earthquake database;

[0150] Constructing a neural network model using preset model parameters, and training the neural network model according to the marked earthquake database to obtain an earthquake event recognition model;

[0151] Acquire an actual microseismic monitoring signal, and use the earthquake recognition model to perform feature extraction on the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal;

[0152] The effective earthquake events corresponding to the actual microseismic monitoring signal are identified according to the earthquake characteristic curve.

[0153] A computer program is a set of instructions that a computer can recognize and execute, and runs on an electronic computer. It is written in a programming language and runs on a target architecture. In order for a computer program to run, the computer needs to load code and data. From the bottom layer of the computer, this is a high-level language (such as Java, C / C++, C#, etc.) code translated into machine language and understood by the CPU for loading. If on a computer that meets most standards, an operating system such as Windows, Linux, etc., loads and executes many programs, in this case, each program is a separate mapping, not all executable programs on the computer. A computer program refers to a coded instruction sequence that can be executed by a computer or other device with information processing capabilities in order to obtain a certain result, or a symbolic instruction sequence or symbolic statement sequence that can be automatically converted into a coded instruction sequence.

[0154] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. When the terms "include" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0155] It is to be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the application described herein are, for example, capable of operation in sequences other than those illustrated or described herein.

[0156] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0157] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0159] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0160] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0161] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0162] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for intelligently identifying effective microseismic events, characterized in that: The method comprises: Acquire a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database; Marking each earthquake data in the original earthquake database to obtain a marked earthquake database; Constructing a neural network model using preset model parameters, and training the neural network model according to the marked earthquake database to obtain an earthquake event recognition model; Acquire an actual microseismic monitoring signal, and use the earthquake recognition model to perform feature extraction on the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal; The effective earthquake events corresponding to the actual microseismic monitoring signal are identified according to the earthquake characteristic curve.

2. The method for intelligently identifying effective microseismic events according to claim 1, characterized in that: The step of performing data enhancement of different strengths on each simulated seismic data in the simulated seismic data set to obtain an original seismic database includes: Calculating the average energy of each simulated seismic data in the simulated seismic data set; Pre-constructed Gaussian white noise is added to each of the simulated seismic data according to the average energy to obtain an original seismic database corresponding to the simulated seismic data set.

3. The method for intelligently identifying effective microseismic events according to claim 1, characterized in that: The step of marking each earthquake data in the original earthquake database to obtain a marked earthquake database comprises: Normalizing each seismic data in the original seismic database to obtain standard seismic data; The event waveform in the standard seismic data is identified, and a Hanning window is added to the event waveform to obtain a marked earthquake database.

4. The method for intelligently identifying effective microseismic events according to claim 1, characterized in that: The method of constructing a neural network model using preset model parameters includes: The neural network model is shown in the following formula: h t =H(W Xh X t +W hh h t-1 +b h ) o t =W ho h t +b o Among them, h t represents the hidden vector at time step t, H represents the hidden layer function in the neural network model, and W Xh , W Xi , W Xf , W Xo , W Xc Respectively represent the weight matrices between the input vector X and the hidden vector h, the input gate i, the forget gate g, the output gate o and the unit activation function c of the neural network model, X t , X t-1 Respectively represent the input vector at time step t and time t-1, W hh represents the weight matrix of the hidden vector h, i t 、f t , o t 、c t Respectively represent the output gate, forget gate, output gate and output result of the unit activation vector in the neural network at time step t, σ represents the logic function, W hi , W hf , W ho , W hc Represents the weight matrix between the hidden vector h and the input gate i, forget gate f, output gate o and unit activation function c, respectively. t-1 represents the output result of the unit activation vector at time step t-1, W ci , W cf , W co Respectively represent the weight matrices between the activation function c and the input gate i, the forget gate f, and the output gate o, b i , b f , b o , b c They represent the bias vectors of input gate i, forget gate f, output gate o and unit activation function c respectively, and tanh() represents the matrix tangent function.

5. The method for intelligently identifying effective microseismic events according to claim 1, characterized in that: The step of training the neural network model according to the marked earthquake database to obtain an earthquake event recognition model includes: Dividing the labeled earthquake data in the labeled earthquake database into a training set and a validation set; Using the neural network model to make an initial prediction on the training set to obtain a predicted seismic signal, and using a preset loss function to calculate a loss value between the predicted seismic signal and the training set; Gradient adjustment is performed on the model parameters in the neural network model according to the loss value until the number of training times for the training set reaches a preset iteration cycle, thereby obtaining an initial earthquake event recognition model; The initial earthquake event recognition model is used to calculate the verification loss value of the verification set, and when the verification loss value is less than a preset loss threshold, an earthquake event recognition model is obtained.

6. The method for intelligently identifying effective microseismic events according to claim 1, characterized in that: The step of extracting features from the actual microseismic monitoring signal using the earthquake recognition model to obtain an earthquake characteristic curve corresponding to the actual microseismic monitoring signal includes: Calculating candidate memory unit values ​​of the actual microseismic signal using the earthquake recognition model; Calculating the output value of the candidate memory unit value using the hidden layer in the earthquake recognition model; An activation calculation is performed on the output value to obtain a characteristic signal corresponding to the actual microseismic monitoring signal.

7. An intelligent identification device for effective microseismic events, characterized in that: The device comprises: A data enhancement module is used to obtain a simulated earthquake data set, and perform data enhancement of different strengths on each simulated earthquake data in the simulated earthquake data set to obtain an original earthquake database; An event marking module is used to mark each earthquake data in the original earthquake database to obtain a marked earthquake database; A model training module, used to construct a neural network model using preset model parameters, and to perform model training on the neural network model according to the marked earthquake database to obtain an earthquake event recognition model; A characteristic curve extraction module is used to obtain an actual microseismic monitoring signal, and use the earthquake recognition model to extract characteristics of the actual microseismic monitoring signal to obtain a seismic characteristic curve corresponding to the actual microseismic monitoring signal; The effective event identification module is used to identify the effective earthquake events corresponding to the actual microseismic monitoring signal according to the earthquake characteristic curve.

8. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for intelligently identifying effective microseismic events as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for intelligently identifying effective microseismic events as described in any one of claims 1 to 7 is implemented.

10. A computer program, characterized in that When the program is executed by a processor, the method for intelligently identifying effective microseismic events as described in any one of claims 1 to 7 is implemented.

Images