Coal-rock interface recognition method and system based on drilling parameters and imaging hyperspectral

Abstract

The present invention discloses a method and system for identifying coal-rock interface based on drilling parameters and imaging hyperspectral, and relates to the technical field of coal-rock interface identification, including: real-time acquisition of drilling parameter data during formation drilling construction; based on the acquired data, filtering out the influence of drilling mechanical parameters on drilling speed, acquiring drilling process index data, and then acquiring drilling signal time series data of the current effective drilling period after preprocessing; extracting the time-frequency domain characteristics of the drilling signal, and outputting the preliminary identification result of the coal-rock interface through a first classification model; if the coal-rock interface is preliminarily identified, collecting drilling cuttings within a set range before and after drilling the coal-rock interface, and scanning to obtain imaging hyperspectral data, so as to perform secondary identification of the coal-rock interface and output it; if the coal-rock interface is not identified, repeating the acquisition and identification of drilling parameter data for the next period. The present invention can achieve coal-rock interface identification with lower cost and higher accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of coal-rock interface recognition, and in particular to a method and system for coal-rock interface recognition based on drilling parameters and imaging hyperspectrum. Background Art

[0002] Coal-rock interface recognition technology is a technology that distinguishes the relative position of the coal-rock interface through the difference in coal-rock properties and corresponding detection methods. Intelligent recognition of coal-rock interface is a bottleneck technical obstacle for automatic cutting of coal seams. At present, based on the different principles of coal-rock interface recognition, the existing coal-rock recognition technologies include: X-ray detection method, electromagnetic wave detection method using the reflection characteristics of electromagnetic waves at the coal-rock interface, infrared imaging method, vibration detection method, cutting force method based on the temperature rise and vibration caused by the different hardness of coal and rock, image recognition method based on the difference in color and texture of the coal wall surface, and hyperspectral method based on the difference in absorption and reflection spectra of coal and rock, etc. However, the above existing methods often have problems such as high development difficulty, high cost or low recognition accuracy.

[0003] The drilling process is a process of interaction between the drill and the rock mass. The drill response information (i.e., drilling parameters) contains a large amount of rock mass information. There is a certain response relationship between the physical properties of some rocks and the drilling parameters. Under ideal conditions, the difference in drilling parameters can clearly distinguish different simulated strata, and the interface of the strata can be identified more accurately. Therefore, coal-rock interface identification while drilling is one of the commonly used identification technologies. However, although drilling parameters are easy to use, in practical applications, due to the complexity of geological conditions and the randomness of the drilling process, the working parameters of the drill are affected by many factors and change dynamically. It is difficult to use an accurate threshold range to distinguish different strata types, and the classification accuracy is poor. In addition, using only a single data type (such as using only a single parameter data such as drilling speed or drill bit drilling speed) for identification also has the problems of low classification accuracy and poor model robustness, making it difficult to establish an accurate coal-rock interface prediction model.

[0004] The use of hyperspectral technology to identify coal-rock interfaces has the advantages of high accuracy and rich information, and is one of the currently commonly used identification methods. However, if data collection and hyperspectral imaging processing are performed on the entire rock formation, sample collection is difficult and costly, and processing and analysis require a large amount of computing resources. Summary of the invention

[0005] In order to solve the deficiencies of the above-mentioned prior art, the present invention provides a method and system for identifying coal-rock interface based on drilling parameters and imaging hyperspectral, which collects real-time data during the drilling process of drilling, takes drilling speed data as a benchmark and filters out the influence of drilling mechanical parameters on drilling speed, obtains drilling process index data, avoids the adverse effects of single collected data and data fluctuations on the identification results, and then performs preliminary identification of the coal-rock interface by analyzing data features and combining classification models to determine the approximate range of the coal-rock interface. After that, the preliminary identification results are subjected to secondary precise classification and identification based on drilling cuttings imaging hyperspectral data to ensure the accuracy of the final coal-rock interface identification. This method avoids the problems of large data collection volume and high cost of traditional hyperspectral identification technology.

[0006] In a first aspect, the present invention provides a method for identifying coal-rock interface based on drilling parameters and imaging hyperspectrum.

[0007] A coal-rock interface recognition method based on drilling parameters and imaging hyperspectral includes:

[0008] Real-time collection of drilling parameter data during formation drilling construction; the drilling parameters include drilling speed, drilling pressure and drill bit drilling speed;

[0009] Based on the drilling parameter data, the influence of the drilling mechanical parameters on the drilling speed is filtered out to obtain the drilling process index data, and then after preprocessing, the drilling signal time series data of the current effective drilling period is obtained;

[0010] Based on the acquired drilling signal time series data, the time-frequency domain features of the drilling signal are extracted, and based on the time-frequency domain features, a preliminary recognition result of the coal-rock interface is output through a first classification model;

[0011] If the coal-rock interface is preliminarily identified, the drill cuttings within the set range before and after drilling the coal-rock interface are collected, and the imaging hyperspectral data are scanned to obtain the coal-rock interface. The coal-rock interface is then secondary identified and output based on the hyperspectral data. If the coal-rock interface is not identified, the drilling parameter data collection and identification for the next time period are repeated.

[0012] In a second aspect, the present invention provides a coal-rock interface recognition system based on drilling parameters and imaging hyperspectral.

[0013] A coal-rock interface recognition system based on drilling parameters and imaging hyperspectral, comprising:

[0014] The data acquisition module is used to collect drilling parameter data in real time during the formation drilling construction process; the drilling parameters include drilling speed, drilling pressure and drill bit drilling speed;

[0015] The data processing module is used to filter out the influence of drilling mechanical parameters on drilling speed based on drilling parameter data, obtain drilling process index data, and then obtain drilling signal timing data of the current effective drilling period after preprocessing;

[0016] A preliminary identification module is used to extract the time-frequency domain characteristics of the drilling signal based on the acquired drilling signal time series data, and output the preliminary identification result of the coal-rock interface through the first classification model based on the time-frequency domain characteristics;

[0017] The secondary identification module is used to collect drill cuttings within a set range before and after drilling the coal-rock interface if the coal-rock interface is preliminarily identified, and scan to obtain imaging hyperspectral data, and then perform secondary identification of the coal-rock interface based on the hyperspectral data and output it; if the coal-rock interface is not identified, repeat the drilling parameter data collection and identification for the next time period.

[0018] In a third aspect, the present invention further provides an electronic device comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the method described in the first aspect are completed.

[0019] In a fourth aspect, the present invention further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the method described in the first aspect.

[0020] One or more of the above technical solutions have the following beneficial effects:

[0021] 1. The present invention provides a method and system for identifying coal-rock interface based on drilling parameters and imaging hyperspectral, collects real-time drilling parameter data during the drilling process, takes the drilling speed in the drilling parameters as a benchmark and filters out the influence of other drilling mechanical parameters on the drilling speed, obtains drilling process index data, and pre-processes the obtained data to avoid the adverse effects of single collected data and data fluctuations on the identification results, and then performs preliminary identification of the coal-rock interface by analyzing data features and combining with a classification model. Through this improved method for identifying coal-rock interface based on drilling parameters, higher identification accuracy is achieved at low cost; on the basis of the above-mentioned determination of the approximate range of the coal-rock interface, the preliminary identification results are subjected to secondary precise classification and identification based on the imaging hyperspectral data of the drilling cuttings to ensure the accuracy of the final coal-rock interface identification, and this method only obtains the imaging hyperspectral data of part of the drilling cuttings, avoiding the problems of large data collection volume and high cost of traditional hyperspectral identification technology.

[0022] 2. In the coal-rock interface identification method proposed in the present invention, the real-time drilling speed, drilling pressure and drill bit drilling speed data during the drilling process are collected by sensors. Based on the collected data, the drilling speed after filtering out the influence of drilling pressure and drill bit drilling speed is normalized and expressed as a drilling process index (DPI) parameter, and the parameter data is used for coal-rock interface identification. In this way, the problem that the data collected in the existing method is single and the data is affected by many aspects and fluctuates, which affects the final identification accuracy, is solved, and the identification accuracy is effectively improved.

[0023] 3. In the present invention, a joint identification method of coal-rock interface identification based on drilling parameters and hyperspectral identification technology is adopted, wherein the coal-rock interface identification method based on drilling parameters has the characteristics of low cost and high economic benefit, and the hyperspectral identification technology has the advantages of high precision and rich information in identifying the coal-rock interface. The combination of the two takes into account efficiency, accuracy and economy, and is suitable for large-scale promotion and application.

[0024] 4. In the present invention, the above-mentioned coal-rock interface recognition system is deployed in electronic equipment to achieve accurate coal cutting and coal placement, breaking through the difficult problem of unmanned real-time online recognition of the coal-rock interface, which is in line with the future technical development trend of safe, efficient and green coal mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0026] Figure 1 This is a flow chart of a method for identifying coal-rock interface based on drilling parameters and imaging hyperspectral according to an embodiment of the present invention;

[0027] Figure 2 This is a flowchart of a method for identifying a coal-rock interface based on drilling parameters and imaging hyperspectral according to an embodiment of the present invention;

[0028] Figure 3 It is a schematic diagram of the structure of the electronic device for deploying the coal-rock interface identification system in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] It should be noted that the following detailed descriptions are exemplary only, are intended to describe specific embodiments, are intended to provide further explanation of the present invention, and are not intended to limit exemplary embodiments according to the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those of ordinary skill in the art to which the present invention belongs. In addition, it should also be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0030] Embodiment 1

[0031] This embodiment provides a method for identifying coal-rock interface based on drilling parameters and imaging hyperspectral. Figure 1 As shown, the specific steps include:

[0032] Step S1, real-time acquisition of drilling parameter data during the formation drilling construction process; the drilling parameters include drilling speed, drilling pressure and drill bit drilling speed;

[0033] Step S2: based on the drilling parameter data, filter out the influence of the drilling mechanical parameters on the drilling speed, obtain the drilling process index data, and then obtain the drilling signal time series data of the current effective drilling period after preprocessing;

[0034] Step S3, based on the acquired drilling signal time series data, extracting the time-frequency domain characteristics of the drilling signal, and based on the time-frequency domain characteristics, outputting the preliminary recognition result of the coal-rock interface through the first classification model;

[0035] Step S4: If the coal-rock interface is preliminarily identified, the drill cuttings within the set range before and after the drilling of the coal-rock interface are collected, and the imaging hyperspectral data are scanned to obtain the secondary identification of the coal-rock interface based on the hyperspectral data and output; if the coal-rock interface is not identified, the drilling parameter data collection and identification of the next time period are repeated.

[0036] like Figure 2 The coal-rock interface identification method based on drilling parameters and imaging hyperspectral proposed in this embodiment is introduced in more detail through the following content.

[0037] Step S1: sample data collection.

[0038] The drilling rig used in underground drilling construction in coal mines is a hydraulic drilling rig. The main method of hole formation is rotary cutting. The drilling rig power head drives the drill rod and the drill bit to rotate, and applies a certain feed force to the drill bit, thereby completing the cutting and crushing of the rock or coal seam. The whole process includes two main work parts: rotation and propulsion. Therefore, in this embodiment, the main drilling parameter data in the formation drilling construction process are collected through multiple sensors. The drilling parameters include drilling speed (referred to as drilling speed), drilling pressure (referred to as drilling pressure) and drill bit drilling speed (i.e., the rotary drilling speed of the drill bit), etc. According to the drilling parameter data obtained above, after pre-processing, an appropriate model is used to further calculate the work energy and identify the formation changes according to the data changes in the drilling process, so as to realize the recognition of the coal-rock interface.

[0039] Step S2: data preprocessing.

[0040] Step S2.1, based on the drilling parameter data, filter out the influence of the drilling mechanical parameters on the drilling speed, and obtain the drilling process index data. Specifically, according to research, even in homogeneous materials, the influence of drilling mechanical parameters (including drilling pressure and drill bit drilling speed, etc.) on the drilling speed is significant. When the drilling pressure and drill bit speed change, the discreteness of the drilling speed is very large. Therefore, it is unreasonable to use only the drilling speed as the only criterion for evaluating rock mass parameters. For this reason, this embodiment proposes a parameter called the drilling process index (DPI), and uses the parameter data as the input layer data of the coal-rock interface identification method for identification.

[0041] The drilling process index DPI is obtained by taking the drilling speed as the benchmark index for evaluating rock mass parameters, filtering out the influence of drilling mechanical parameters including drilling pressure and drill bit drilling speed on the drilling speed benchmark index, and obtaining the normalized parameter of drilling speed, namely the drilling process index DPI. The DPI index is unique in homogeneous materials and reflects the difficulty of rock crushing. It can be defined as:

[0042] ;

[0043] In the above formula, α is a constant and a parameter related to rock strength; V is the drilling speed; F is the drilling pressure; N Drilling speed of the drill bit.

[0044] It can be seen from the above formula (1) that the drilling process index is mainly determined by three basic parameters and is a derived parameter of the drilling parameters. Therefore, it is only necessary to collect the drilling parameters during the drilling process to obtain the drilling process index of the roof rock layer and the coal seam, which approximately reflects the strength of the current drilling formation, thereby identifying the drilling formation. Under the conditions of fixed construction equipment and technical level, the drilling process index is closely correlated with the physical and mechanical properties of the rock, and can more accurately characterize the response relationship between the physical parameters and the formation information during the drilling process.

[0045] Step S2.2: pre-process the drilling process index data sequence acquired during the current period.

[0046] First, the drilling process index data (hereinafter referred to as drilling data or drilling signal) in the inactive period is excluded, such as the initial drilling stage and the stage of adding a new drill pipe joint. The drilling data collected in this stage does not have interpretation value. Through the above steps, the effective drilling period is screened out and the drilling signal time series data of the current effective drilling period is obtained.

[0047] Then, considering that the drilling data collection process is subject to mechanical instability vibration and external interference, the drilling data collected by the sensor has certain fluctuations, which in turn causes certain fluctuations and noise in the DPI time series data. Therefore, in order to further improve the accuracy of the drilling data, this embodiment uses the wavelet threshold method to reduce the noise of the real-time drilling signal, and the formula is as follows:

[0048] ;

[0049] In the above formula, is the approximate coefficient after threshold processing, is the detail coefficient after threshold processing, and They are time series functions constructed based on wavelet function and scale function, respectively, and are used to represent the local characteristics of drilling signals at different scales and times; j represents the scale of wavelet transform and scaling transform, k Represents the translation parameters of wavelet coefficients and scale transformation. In practical applications, by adjusting the parameters j and k The value of can be used to analyze and process different parts of the signal (such as high-frequency noise, low-frequency trends, etc.).

[0050] Step S3: Preliminary identification of the coal-rock interface.

[0051] Step S3.1, based on the acquired drilling signal time series data, extract the time-frequency domain features of the drilling signal. Specifically, comprehensively consider the influence of the time-domain features and time-frequency domain features of the drilling signal on the identification of the coal-rock interface, extract the fluctuation features closely related to the coal-rock interface based on the pre-processed drilling signal time series data, that is, the time-domain features and the frequency-domain features, and select the time-domain features of the drilling signal and the frequency-domain features of the drilling signal that are closely related to the coal-rock interface.

[0052] The above time domain characteristics reflect the changing rules of the drilling signal in the time domain, including the intensity characteristics, waveform and change characteristics of the drilling signal. The intensity characteristics of the drilling signal include the maximum value, minimum value, peak-to-peak value, average value, root mean square value, standard deviation, total signal value, etc.; the kurtosis, skewness, peak factor, pulse factor, waveform index, margin factor, etc. of the drilling signal waveform and change characteristics.

[0053] On the basis of the above, the time-frequency analysis method is used to convert the time domain drilling signal into the frequency domain, and the spectrum of the drilling signal is obtained. Then, the frequency domain features are extracted based on the spectrum of the drilling signal to obtain the frequency domain characteristics of the drilling signal. The frequency domain characteristics of the drilling signal include the frequency domain amplitude average, centroid frequency, mean square frequency, and frequency variance.

[0054] Step S3.2: Based on the time-frequency domain characteristics, the first classification model is used to output the preliminary recognition result of the coal-rock interface.

[0055] In this embodiment, an existing machine learning model based on a supervised learning algorithm (such as a random forest, a BP neural network, a long short-term memory network, etc.) is used to construct a first classification model. The first classification model outputs whether the current drilling formation interface is a coal-rock interface, and pre-calibrates the drilling parameters to complete the model training.

[0056] Considering that the identification of the coal-rock interface during drilling is a complex and constantly changing nonlinear process, the result of the preliminary identification of the coal-rock interface based on the drilling parameters is an approximate range, and the exact position of the coal-rock interface cannot be determined. In order to guide the intelligent coal mining machine to cut and release coal, in this embodiment, the imaging hyperspectral technology is combined on the above basis to further identify a more precise coal-rock interface position. That is, the drilling signal time series data characteristics of the current effective drilling period are obtained and input into the first classification model that has been trained, and the preliminary identification result of the coal-rock interface is output. If the coal-rock interface is preliminarily identified, the hyperspectral identification step of the next step is performed. If the coal-rock interface is not identified, the drilling parameter data collection and identification steps of the next period of time of steps S1 to S3 above are repeated.

[0057] Step S4: secondary identification of the coal-rock interface.

[0058] When the coal-rock interface is preliminarily identified in the above step S3, this step is performed, i.e., the drill cuttings within the set range before and after drilling of the preliminarily identified coal-rock interface are collected, and imaging hyperspectral data are obtained by scanning, and then the coal-rock interface is secondary identified and output based on the hyperspectral data.

[0059] Step S4.1, obtain the drill cuttings within the set range before and after the coal-rock interface is drilled, and scan to obtain imaging hyperspectral data. Specifically, compared with indirectly obtaining coal-rock information through drilling parameters, the drill cuttings generated during drilling can directly obtain information on the coal-rock interface, and the drill cuttings and data are easier to obtain. When the coal-rock interface is identified in the previous step, the drill cuttings (or drill residues) within 10m before and after the interface are collected. After the drill cuttings are dried, a hyperspectral scan is performed at fixed lengths to obtain imaging hyperspectral data.

[0060] Step S4.2: preprocess the acquired hyperspectral data, which includes spectral correction processing, denoising processing based on principal component analysis, band selection processing, etc.

[0061] Specifically, first, the acquired hyperspectral data is spectrally corrected to eliminate the influence of light source, detector and sensor, so as to ensure the accuracy of spectral reflectance; secondly, the principal component analysis (PCA) method is used to remove the noise in the hyperspectral data; finally, according to the specific requirements of coal-rock interface identification, the most discriminative spectral band is selected to reduce redundant data; among them, the selected spectral bands usually include 600~700nm, 700~900nm and 1700~2300nm.

[0062] The specific steps of using the PCA method to remove noise from hyperspectral data are as follows:

[0063] (1) Assume that the spectral data is a m×n The sample matrix X , which is a matrix representing the number of samples, where Represents a pixel point. As a sample, , represents the total sample size; n Indicates the band. n As a feature, , Represents the total dimension of the feature;

[0064] (2) Data centering. For the sample matrix X Centering is done by subtracting the mean value calculated from the original data from each feature μ , to ensure that the mean of the data is zero and the origin of the new coordinate system is at the center of the data;

[0065] (3) Covariance matrix calculation. Based on the sample matrix after data centering, the covariance matrix between data features is calculated. COV , the covariance matrix describes the correlation and variance between different features;

[0066] (4) Eigenvalue decomposition. For the covariance matrix COV Perform eigenvalue decomposition to obtain eigenvalues ​​and corresponding eigenvectors. The eigenvectors represent the directions of the new coordinate system, while the eigenvalues ​​represent the variance of the data in each direction.

[0067] (5) Select the principal component. According to the size of the eigenvalue, save the previous one with the largest eigenvalue. n ′ eigenvectors, i.e., the number of principal components. Usually, the number of principal components can be determined based on the cumulative variance contribution rate curve of the eigenvalues ​​to retain sufficient information;

[0068] (6) Projection data. X Project to the selected n The matrix of ′ eigenvectors YIn the new space with , we get the sample matrix after dimensionality reduction X ′= X×Y The sample matrix obtained at this time is X ′ is the hyperspectral data after noise removal.

[0069] Step S4.3: Based on the preprocessed hyperspectral data, the hyperspectral features of the joint spatial features and spectral features are extracted through the second classification model based on the convolutional neural network, and classification and recognition are performed based on the hyperspectral features to output the secondary recognition result of the coal-rock interface.

[0070] Specifically, the hyperspectral image (i.e., hyperspectral data) can be regarded as a three-dimensional data cube based on pixel point × band × reflectance, i.e., spatial dimension plus spectral band. Convolutional neural network (CNN) can be used to extract the joint features of space and spectrum, thereby completing the effective extraction of hyperspectral features. ,in, H and W is the space size, D is the number of spectral bands, then the feature Y output by the convolution operation can be expressed as:

[0071] ;

[0072] in, K h 、K w 、K d are the sizes of the convolution kernel in height, width, and spectral dimensions, respectively. W is the weight matrix.

[0073] Furthermore, like the above-mentioned first classification model, this embodiment adopts one of the existing neural network models based on the supervised learning algorithm to construct a second classification model. The second classification model will judge whether the current drilling formation interface is a coal-rock interface as an output, and has completed the model training in advance using calibrated hyperspectral data.

[0074] The trained second classification model is used to perform secondary classification and recognition on the extracted hyperspectral features, and the accurate coal-rock interface position is further determined based on the differences in the hyperspectral features of different coal and rock strata. If the secondary recognition result is that the coal-rock interface is identified, that is, the current drilling stratum interface is judged to be the coal-rock interface, then the accurate coal-rock interface position is output, and the classification result can guide the intelligent coal mining machine to cut and release coal; on the contrary, if the secondary recognition result is that the coal-rock interface is not identified, that is, the current drilling stratum interface is judged to be non-coal-rock interface, then the drilling parameter data collection and recognition of the next period are repeated.

[0075] Preferably, the classification result output by the secondary recognition can guide the intelligent coal mining machine to cut and release coal. At the same time, it can also serve as the basis for optimizing the parameters of the first classification model in step S3 to further optimize the performance and recognition accuracy of the first classification model.

[0076] Embodiment 2

[0077] This embodiment provides a coal-rock interface recognition system based on drilling parameters and imaging hyperspectral, which specifically includes:

[0078] The data acquisition module is used to collect drilling parameter data in real time during the formation drilling construction process; the drilling parameters include drilling speed, drilling pressure and drill bit drilling speed;

[0079] The data processing module is used to filter out the influence of drilling mechanical parameters on drilling speed based on drilling parameter data, obtain drilling process index data, and then obtain drilling signal timing data of the current effective drilling period after preprocessing;

[0080] A preliminary identification module is used to extract the time-frequency domain characteristics of the drilling signal based on the acquired drilling signal time series data, and output the preliminary identification result of the coal-rock interface through the first classification model based on the time-frequency domain characteristics;

[0081] The secondary identification module is used to collect drill cuttings within a set range before and after drilling the coal-rock interface if the coal-rock interface is preliminarily identified, and scan to obtain imaging hyperspectral data, and then perform secondary identification of the coal-rock interface based on the hyperspectral data and output it; if the coal-rock interface is not identified, repeat the drilling parameter data collection and identification for the next time period.

[0082] Embodiment 3

[0083] This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps in the coal-rock interface identification method based on drilling parameters and imaging hyperspectral are completed as described above.

[0084] like Figure 3 As shown, the electronic device proposed in this embodiment includes a processor, a memory, a peripheral controller, a communication bus and a communication interface. Among them, the memory stores a computer program, that is, the framework of the trained coal-rock interface recognition system; the communication interface is connected to the data acquisition module of the drilling data; the processor connects the memory and the communication interface through the communication bus, and after receiving the execution instruction, it executes the computer program stored in the memory, determines the coal-rock interface, and exchanges data with the memory and the peripheral controller through the communication bus, and outputs the recognition result to the peripheral controller of the intelligent coal mining machine, instructing the intelligent coal mining machine to adaptively adjust the drum height, or pre-plan the cutting height, so as to achieve accurate coal cutting and coal placing.

[0085] Embodiment 4

[0086] This embodiment also provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps in the coal-rock interface identification method based on drilling parameters and imaging hyperspectral are completed as described above.

[0087] The steps involved in the above embodiments 2 to 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0088] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0089] The above description is only a preferred embodiment of the present invention. Although the specific implementation mode of the present invention is described in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present invention.

Claims

1. A coal-rock interface recognition method based on drilling parameters and imaging hyperspectral, characterized in that: include: Real-time collection of drilling parameter data during formation drilling construction; The drilling parameters include drilling speed, drilling pressure and drill bit drilling speed; Based on the drilling parameter data, the influence of the drilling mechanical parameters on the drilling speed is filtered out to obtain the drilling process index data, and then after preprocessing, the drilling signal time series data of the current effective drilling period is obtained; Based on the acquired drilling signal time series data, the time-frequency domain features of the drilling signal are extracted, and based on the time-frequency domain features, a preliminary recognition result of the coal-rock interface is output through a first classification model; the time-frequency domain features of the drilling signal are used as fluctuation features related to the coal-rock interface, including time-domain features and frequency-domain features; The time domain characteristics are used to reflect the variation law of the drilling signal in the time domain, including intensity characteristics, waveform and variation characteristics; the intensity characteristics include the maximum value, minimum value, peak-to-peak value, average value, root mean square value, standard deviation, and total signal value of the drilling signal; the waveform and variation characteristics include the kurtosis, skewness, peak factor, pulse factor, waveform index, and margin factor of the drilling signal; The extraction of the frequency domain features includes: The time-domain drilling signal is converted into the frequency domain by using the time-frequency analysis method to obtain the spectrum of the drilling signal; the frequency domain feature is extracted based on the spectrum of the drilling signal to obtain the frequency domain feature of the drilling signal; the frequency domain feature of the drilling signal includes the frequency domain amplitude average value, the center of gravity frequency, the mean square frequency and the frequency variance; If the coal-rock interface is initially identified, the drill cuttings within the set range before and after the drilling of the coal-rock interface are collected, and the imaging hyperspectral data are scanned to obtain the secondary identification of the coal-rock interface based on the hyperspectral data and output; if the coal-rock interface is not identified, the drilling parameter data collection and identification of the next period are repeated; Among them, based on the hyperspectral data of drilling cuttings, the secondary recognition and output of the coal-rock interface are carried out, including: Preprocessing the acquired hyperspectral data; the preprocessing includes spectral correction processing, denoising processing based on principal component analysis, and band selection processing; Based on the preprocessed hyperspectral data, the hyperspectral features of the joint spatial features and spectral features are extracted through the second classification model based on the convolutional neural network, and classification and identification are performed based on the hyperspectral features to output the secondary identification results of the coal-rock interface.

2. The method for identifying coal-rock interface based on drilling parameters and imaging hyperspectral according to claim 1, characterized in that: The acquisition of the drilling process index DPI includes: Taking the drilling speed as the benchmark indicator for evaluating rock mass parameters, the influence of drilling mechanical parameters including drilling pressure and drill bit drilling speed on the drilling speed benchmark indicator is filtered out to obtain the normalized parameter of drilling speed, namely the drilling process index DPI, which is defined as: ； In the above formula, α is a constant and a parameter related to rock strength; V is the drilling speed; F is the drilling pressure; N Drilling speed of the drill bit.

3. The method for identifying coal-rock interface based on drilling parameters and imaging hyperspectral according to claim 1, characterized in that: The pre-processing comprises: For the drilling process index data sequence obtained in the current period, the drilling process index data in the inactive period is excluded, and the wavelet threshold method is used to perform noise reduction on the drilling process index time series data after exclusion processing to obtain the drilling signal time series data of the current effective drilling period; The inactive state period includes the initial drilling stage and the stage of adding a new drill pipe joint.

4. The method for identifying coal-rock interface based on drilling parameters and imaging hyperspectral according to claim 1, characterized in that: The recognition result of the coal-rock interface is to determine whether the current drilling formation interface is a coal-rock interface. If the secondary recognition result is that the coal-rock interface is identified, that is, the current drilling formation interface is judged to be a coal-rock interface, then the accurate coal-rock interface position is output; conversely, if the secondary recognition result is that the coal-rock interface is not identified, that is, the current drilling formation interface is judged to be a non-coal-rock interface, then the drilling parameter data collection and identification of the next time period are repeated.

5. A coal-rock interface recognition system based on drilling parameters and imaging hyperspectral, characterized in that: include: The data acquisition module is used to collect drilling parameter data in real time during the formation drilling construction process; the drilling parameters include drilling speed, drilling pressure and drill bit drilling speed; The data processing module is used to filter out the influence of drilling mechanical parameters on drilling speed based on drilling parameter data, obtain drilling process index data, and then obtain drilling signal timing data of the current effective drilling period after preprocessing; A preliminary identification module is used to extract the time-frequency domain characteristics of the drilling signal based on the acquired drilling signal time series data, and output the preliminary identification result of the coal-rock interface through the first classification model based on the time-frequency domain characteristics; the time-frequency domain characteristics of the drilling signal are used as the fluctuation characteristics related to the coal-rock interface, including time-domain characteristics and frequency-domain characteristics; The time domain characteristics are used to reflect the variation law of the drilling signal in the time domain, including intensity characteristics, waveform and variation characteristics; the intensity characteristics include the maximum value, minimum value, peak-to-peak value, average value, root mean square value, standard deviation, and total signal value of the drilling signal; the waveform and variation characteristics include the kurtosis, skewness, peak factor, pulse factor, waveform index, and margin factor of the drilling signal; The extraction of the frequency domain features includes: The time-domain drilling signal is converted into the frequency domain by using the time-frequency analysis method to obtain the spectrum of the drilling signal; the frequency domain feature is extracted based on the spectrum of the drilling signal to obtain the frequency domain feature of the drilling signal; the frequency domain feature of the drilling signal includes the frequency domain amplitude average value, the center of gravity frequency, the mean square frequency and the frequency variance; The secondary identification module is used to collect drill cuttings within a set range before and after drilling the coal-rock interface if the coal-rock interface is initially identified, and scan to obtain imaging hyperspectral data, and then perform secondary identification of the coal-rock interface based on the hyperspectral data and output it; if the coal-rock interface is not identified, repeat the drilling parameter data collection and identification for the next period; Among them, based on the hyperspectral data of drilling cuttings, the secondary recognition and output of the coal-rock interface are carried out, including: Preprocessing the acquired hyperspectral data; the preprocessing includes spectral correction processing, denoising processing based on principal component analysis, and band selection processing; Based on the preprocessed hyperspectral data, the hyperspectral features of the joint spatial features and spectral features are extracted through the second classification model based on the convolutional neural network, and classification and identification are performed based on the hyperspectral features to output the secondary identification results of the coal-rock interface.

6. An electronic device, characterized in that: The method comprises a memory and a processor and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of a method for identifying coal-rock interface based on drilling parameters and imaging hyperspectrum as described in any one of claims 1 to 4 are completed.

7. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of a coal-rock interface identification method based on drilling parameters and imaging hyperspectrum as described in any one of claims 1-4.

Images