A coal rock hydraulic fracturing signal recognition method based on an MLP neural network model

Abstract

The application discloses a coal rock hydraulic fracturing signal identification method based on an MLP neural network model, microseismic characteristic parameters capable of representing different coal rock lithologies are screened out first, then microseismic characteristic parameters corresponding to different types of coal rock lithologies are obtained, and the above data is divided into a training set and a verification set; then an MLP neural network model with a specific structure is constructed, the model is trained and learned by using the training set, the weights and bias terms of the MLP neural network model are optimized and updated by using a BP algorithm, after the optimization training is finally completed, all microseismic data in the hydraulic fracturing process are input into the model, the model can accurately output the coal rock lithology labels corresponding to each group of data after identification, and the coal rock hydraulic fracturing signal distinguishing process is completed; thereby, the lithology and horizon information of the hydraulic fracturing coal rock rupture are identified, the search range of microseismic positioning inversion is reduced, and finally the efficiency and accuracy of the microseismic positioning inversion are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of microseismic monitoring technology, specifically relating to a method for identifying hydraulic fracturing signals in coal and rock based on an MLP neural network model. Background Technology

[0002] Hydraulic fracturing is a key technology for increasing coal seam permeability and improving gas extraction efficiency. Microseismic monitoring is an important means of detecting the range of hydraulic fracturing. Microseismic monitoring technology allows for the dynamic observation of the fracture propagation state in underground coal seams, evaluating the permeability-enhancing effect of hydraulic fracturing, and is of great significance for the precise implementation of underground coal seam hydraulic fracturing and efficient gas extraction. Microseismic source location is one of the core technologies of microseismic monitoring; accurately detecting the range of coal seam hydraulic fracturing is crucial for scientifically evaluating the permeability-enhancing effect of hydraulic fracturing and improving the efficiency of coal seam gas extraction.

[0003] Currently, most coal seams used for hydraulic fracturing and permeability enhancement in underground coal mines are generally less than 5 meters thick. Performing hydraulic fracturing in such thin coal seams easily leads to fracturing of the roof and floor strata, causing further fracturing. This results in microseismic monitoring signals containing not only coal seam fracturing microseismic signals but also microseismic signals generated by the fracturing of surrounding rock strata. Consequently, when using microseismic monitoring technology to evaluate the permeability enhancement range of coal and rock hydraulic fracturing, the mixing of multiple microseismic signals causes fracturing signals from other rocks to interfere with the coal seam hydraulic fracturing signal during data analysis, leading to lower microseismic positioning accuracy.

[0004] Therefore, the research direction required by this invention is to provide a new method that can identify the collected microseismic signals, thereby accurately distinguishing the hydraulic fracturing signals corresponding to different types of coal and rock lithology, and at a relatively fast speed. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method for identifying hydraulic fracturing signals in coal and rock based on an MLP neural network model. This method can identify the collected microseismic signals, thereby accurately distinguishing the hydraulic fracturing signals corresponding to different types of coal and rock lithology, and it is relatively fast.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for identifying coal and rock hydraulic fracturing signals based on an MLP neural network model, comprising the following steps:

[0007] Step 1: First, screen out the microseismic characteristic parameters that can represent different coal and rock lithologies. Then, collect coal samples from the coal seam to be hydraulically fracturing and the corresponding rock samples from the roof and floor strata. Each type of sample contains multiple samples. Simulate the hydraulic fracturing process for all samples of different types and obtain the microseismic characteristic parameters of each sample. Then, assign different labels to different types of samples.

[0008] Step 2: Select a portion of the multiple sets of microseismic characteristic parameters for each sample in Step 1 as the training set, and the remainder as the validation set;

[0009] Step 3: Construct an MLP (Multilayer Perceptron) neural network model, which consists of an input layer, hidden layers, and an output layer;

[0010] Step 4: Set the maximum number of iterations, initialize the weights and biases of the MLP neural network model, and use the BP algorithm to learn the weight parameters of the MLP neural network model;

[0011] Step 5: Input the training set from Step 2 into the MLP neural network model with the parameters set in Step 4 for iterative training. After reaching the maximum number of iterations, input all the data from the validation set in Step 2 into the trained MLP neural network model for recognition and output the label corresponding to each group of data. If the error value is less than the set threshold, proceed to Step 7; otherwise, proceed to Step 6.

[0012] Step 6: Optimize the weights and biases of the MLP neural network model from Step 5 using the BP algorithm and update the parameters. Repeat Step 5 after completion.

[0013] Step 7: Conduct hydraulic fracturing experiments on coal and rock samples and collect all microseismic data during the hydraulic fracturing process. Input each set of microseismic data into the MLP neural network model for identification, thereby outputting the label of the corresponding type for each set of data, and completing the process of distinguishing coal and rock hydraulic fracturing signals.

[0014] Furthermore, the microseismic characteristic parameters representing different coal and rock lithologies in step one include amplitude, rise time, ring count, energy, centroid frequency, and peak frequency. These parameters were derived by the inventors through research, and by combining these parameters with subsequent models, different types of coal and rock lithologies can be accurately identified.

[0015] Furthermore, the simulated hydraulic fracturing in step one is conducted in a laboratory, and the same hydraulic fracturing parameters are used for each sample. This ensures that the microseismic data obtained for each sample are derived using the same hydraulic fracturing method, guaranteeing the accuracy of data acquisition.

[0016] Furthermore, in step two, the MLP neural network model has four layers, including one input layer with six neurons; two hidden layers, each with 64 neurons, using ReLU as the activation function; and one output layer with four neurons, using Softmax as the activation function. Using these model parameters effectively ensures the accuracy of the model's recognition.

[0017] Furthermore, the maximum number of iterations in step four is 3000.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] 1. This invention constructs a specific structured MLP neural network model and combines it with various microseismic characteristic parameters that represent different coal and rock lithologies obtained from research. The specific structured MLP neural network model has a strong adaptive learning capability, enabling it to be trained and learned on the microseismic characteristic parameters of different coal and rock lithologies. After completion, it extracts and intelligently identifies the features of microseismic signals from hydraulic fracturing of coal and rock, outputs the coal and rock lithology corresponding to different hydraulic fracturing signals, thereby identifying the stratigraphic information of hydraulic fracturing of coal and rock, narrowing the search range of microseismic location inversion, and ultimately effectively improving the efficiency and accuracy of microseismic location inversion.

[0020] 2. The MLP neural network model with a specific structure in this invention does not require adjustment of the gradient descent algorithm hyperparameters, which brings the greatest possible convenience to the user. Therefore, this method has high ease of use. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0022] The present invention will be further described below.

[0023] like Figure 1 As shown, this embodiment includes the following steps:

[0024] Step 1: First, select microseismic characteristic parameters that represent different coal and rock lithologies, including amplitude, rise time, ring count, energy, centroid frequency, and peak frequency. These parameters were derived by the inventors through research. Using these parameters in combination with subsequent models can accurately identify different types of coal and rock lithologies. Next, collect coal samples from the required hydraulically fracturing coal seams and corresponding rock samples from the roof and floor strata, namely coal, sandstone, shale, and mudstone. Each type of sample contains multiple samples. All samples of the four types undergo simulated hydraulic fracturing processes in the laboratory, and each sample uses the same hydraulic fracturing parameters. This ensures that the microseismic data obtained for each sample are obtained using the same hydraulic fracturing, guaranteeing the accuracy of data acquisition. Then, obtain the microseismic characteristic parameters for all samples of each type. One sample in each type corresponds to one set of microseismic characteristic parameters, so samples of the same type have multiple sets of microseismic characteristic parameters. Different labels are assigned to different types of samples; the labels for coal, sandstone, shale, and mudstone are "0", "1", "2", and "3", respectively.

[0025] Step 2: The 500 sets of microseismic characteristic parameters for each sample in Step 1 are used as the training set as shown in Table 1, and the remaining 50 sets for each sample are used as the validation set as shown in Table 2.

[0026] Table 1 Training Set of Microseismic Signals for Hydraulic Fracturing of Coal and Rock

[0027]

[0028]

[0029] Table 2 Verification Set of Microseismic Signals for Hydraulic Fracturing of Coal and Rock

[0030]

[0031] Step 3: Construct an MLP (Multilayer Perceptron) neural network model. This model has four layers: one input layer with 6 neurons; two hidden layers, each with 64 neurons and ReLU activation function; and one output layer with 4 neurons and Softmax activation function. Using these model parameters effectively ensures the accuracy of the model's recognition.

[0032] Step 4: Set the maximum number of iterations for the MLP neural network model to 3000, initialize the weights and biases, and use the BP algorithm to learn the weight parameters of the MLP neural network model;

[0033] Step 5: Input the training set from Step 2 into the MLP neural network model with the parameters set in Step 4 for iterative training. After reaching the maximum number of iterations, input all the data from the validation set in Step 2 into the trained MLP neural network model for recognition and output the label corresponding to each group of data. If the error value is less than the set threshold, proceed to Step 7; otherwise, proceed to Step 6.

[0034] Step 6: Optimize the weights and biases of the MLP neural network model from Step 5 using the BP algorithm and update the parameters. Repeat Step 5 after completion.

[0035] Step 7: The optimized MLP neural network model is used to identify the validation sets of various types of samples. Through actual testing, it is found that the model's recognition accuracy reaches 96%, indicating that the model of the present invention can accurately identify and distinguish hydraulic fracturing signals of different types of coal and rock lithology.

[0036] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying hydraulic fracturing signals in coal and rock based on an MLP neural network model, characterized in that, Includes the following steps: Step 1: First, screen out the microseismic characteristic parameters that can represent different coal and rock lithologies, including amplitude, rise time, ring count, energy, centroid frequency, and peak frequency; then collect coal samples from the coal seam to be hydraulically fracturing and the corresponding rock samples from the roof and floor strata; each type of sample contains multiple samples, and simulate the hydraulic fracturing process for all samples of different types, obtain the microseismic characteristic parameters of each type of sample, and assign different labels to different types of samples; Step 2: Select a portion of the multiple sets of microseismic characteristic parameters for each sample in Step 1 as the training set, and the remainder as the validation set; Step 3: Construct an MLP neural network model, which has 4 layers, including 1 input layer with 6 neurons; 2 hidden layers with 64 neurons each, using ReLU as the activation function; and 1 output layer with 4 neurons, using Softmax as the activation function. Step 4: Set the maximum number of iterations, initialize the weights and biases of the MLP neural network model, and use the BP algorithm to learn the weight parameters of the MLP neural network model; Step 5: Input the training set from Step 2 into the MLP neural network model with the parameters set in Step 4 for iterative training. After reaching the maximum number of iterations, input all the data from the validation set in Step 2 into the trained MLP neural network model for recognition and output the label corresponding to each group of data. If the error value is less than the set threshold, proceed to Step 7. Otherwise, proceed to step six; Step 6: Optimize the weights and biases of the MLP neural network model from Step 5 using the BP algorithm and update the parameters. Repeat Step 5 after completion. Step 7: Conduct hydraulic fracturing experiments on coal and rock samples and collect all microseismic data during the hydraulic fracturing process. Input each set of microseismic data into the MLP neural network model for identification, thereby outputting the label of the corresponding type for each set of data, and completing the process of distinguishing coal and rock hydraulic fracturing signals.

2. The method for identifying coal and rock hydraulic fracturing signals based on an MLP neural network model according to claim 1, characterized in that, The simulated hydraulic fracturing in step one was carried out in the laboratory, and the same hydraulic fracturing parameters were used for each sample.

3. The method for identifying coal and rock hydraulic fracturing signals based on an MLP neural network model according to claim 1, characterized in that, The maximum number of iterations in step four is 3000.

Images