CN-122016877-A - Coal rock identification method

CN122016877ACN 122016877 ACN122016877 ACN 122016877ACN-122016877-A

Abstract

The invention discloses a coal and rock identification method, which is based on a Compton scattering imaging technology, a motion bracket, a data processing and identification module and the like of an X-ray source and a photon counting detector, and by combining a multi-channel energy spectrum fusion and boundary regression fitting algorithm and a third-order B spline interpolation and Savitzky-Golay smooth filtering, a confidence correction mechanism is introduced to generate a high-precision and anti-noise continuous coal and rock interface track, so that the identification accuracy and robustness are improved. The invention can realize multi-energy spectrum resolution on scattered rays of the measured object, and can accurately distinguish coal beds from thin-layer gangue by combining a multi-channel energy spectrum fusion algorithm.

Inventors

LIU ZAIBIN
DANG XINXIN
LIU QIANG
FAN TAO
MA LIANG
LI GUIHONG
Ju Chaohui
CHEN CHANGYUAN
LI PENG
LI XU

Assignees

西安煤科透明地质科技有限公司

Dates

Publication Date: 20260512
Application Date: 20250904

Claims (10)

1. A coal rock identification method is characterized in that the method realizes the identification of a coal rock interface based on a coal rock identification device, wherein the coal rock identification device comprises an X-ray source, a photon counting detector, a moving bracket and a data processing and identification module, wherein the X-ray source and the photon counting detector are arranged on the moving bracket and are arranged on the same side of a coal rock to be detected; the method comprises the following steps: step 1, moving an X-ray source and a photon counting detector to the front of a coal-rock area to be detected through a moving bracket; step 2, the X-ray source emits high-frequency pulse X-rays, namely primary rays I0, according to the set parameters, and irradiates the area to be detected; step 3, the photon counting detector receives Compton back scattering rays emitted by the area to be detected, namely Compton back scattering photons I bs , collects multi-energy spectrum scattering data to extract multi-channel energy spectrum characteristics, and transmits the multi-channel energy spectrum characteristics to the data processing and identifying module; Step 4, the data processing and identifying module analyzes substances in a detected area according to the original ray I0, compton back-scattered photons I bs and the multi-channel energy spectrum characteristics, and carries out coal and rock identification based on a coal and rock interface generating algorithm of multi-channel energy spectrum fusion and boundary regression fitting by combining third-order B spline interpolation and Savitzky-Golay smooth filtering and introducing confidence correction; And 5, moving the X-ray source and the photon counting detector to different areas through the moving support, repeatedly executing the steps on a plurality of parallel sections, integrating coal-rock interface tracks on the sections, generating a three-dimensional space distribution diagram of the coal-rock interface, and providing more comprehensive information for mining planning and geological analysis.
2. The coal rock identification method of claim 1, wherein the X-ray source is configured to emit X-rays, and the energy of the X-rays is adjusted by adjusting an operating voltage of the X-ray source; The photon counting detector is used for receiving Compton back scattering rays and measuring the radiation dose and energy distribution of the Compton back scattering rays, is provided with a collimator module, and receives and measures the rays in a specific direction by adjusting the direction of the collimator so as to realize the measurement of substances in a specific area and a specific depth; The motion support is used for supporting the X-ray source and the photon counting detector, and detection of different areas is achieved through movement.
3. The method for identifying coal and rock according to claim 2, wherein the radiation emitting end face of the X-ray source and the signal collecting end face of the photon counting detector face towards the coal and rock detection area, compton backscattering occurs after the X-rays act on the coal and rock substances, and the photon counting detector collects and detects the reflected X-ray backscattering signals; The included angle theta between the central axis of the X-ray source and the central axis of the photon counting detector is selected within the range of 150-180 degrees.
4. The coal rock identification method of claim 1, wherein the step 3 includes: Step 3.1, the photon counting detector receives Compton backscattered photons I bs ; step 3.2, distributing the detected Compton back scattering photons into n preset energy channels according to the energy size; step 3.3, dividing the surface to be detected into sampling grids of i columns and j rows; Step 3.4, recording the number of photons detected by each energy channel at each sampling grid point (x i , z j ) to form a multi-channel energy spectrum count vector: (3)。
5. The coal rock identification method of claim 4, wherein step 4 includes: Step 4.1, the classifier identifies the coal/rock and confidence: Using a Classifier to judge coal/rock, inputting the coal/rock into the multi-channel energy spectrum counting vector f ij extracted in the step 3, and outputting the coal/rock as a classification label l ij and a confidence coefficient p ij ; step 4.2, boundary jump detection: Defining the mutation position of the coal and rock classification labels of adjacent scanning points with different depths in the same column as a jump point according to the classification label l ij output in the step 4.1, defining the reliable jump point coordinates screened by the confidence degree as boundary points according to the confidence degree p ij output in the step 4.1, and detecting to obtain an ith column boundary point set B i ; and 4.3, performing regression fitting and smooth filtering on boundary points: Taking the boundary point set of all columns detected in the step 4.2 as an input node, and generating a continuous function f (x) representing the change of a coal-rock interface along with the horizontal position x by using a third-order B spline interpolation fitting; Inputting the continuous function f (x) into a Savitzky-Golay filter for local smoothing to obtain a smoothed coal-rock interface track function f smoothed (x); Step 4.4, confidence correction: Aiming at boundary points on the track function f smoothed (x), if the classification confidence coefficient is lower than the boundary restoration confidence coefficient threshold and the confidence coefficient of the adjacent boundary points is not lower than the boundary restoration confidence coefficient threshold, correcting the point value by using weighted interpolation, otherwise, keeping f smoothed (x), and obtaining the track function f corrected (x) corrected by the confidence coefficient weighted interpolation; Step 4.5, outputting tracks: The corrected trajectory function z=f corrected (x) is used as a final coal-rock interface trajectory, and the high-precision smooth continuous coal-rock interface trajectory realizes the identification of a coal-rock interface and can guide the cutting path of the coal cutter.
6. The method of coal and rock identification according to claim 5, wherein in step 4.1, the Classifier is used to perform coal/rock judgment, the feature vector f koukou is input, and the trained Classifier is used to perform prediction: (4) Wherein, classifying the label 0 Is the coal bed, 1 is the rock stratum, confidence The classification reliability is quantified, and a larger value indicates a more reliable classification result.
7. The coal rock identification method according to claim 5, wherein in the step 4.2, the jump point is a position (x i , z j ) where the classification tag is mutated, so as to satisfy scanning points l ij ≠l i(j+1) with different depths in the same column; The boundary points meet max (p ij , p i(j+1) ) > T high , wherein max (p ij , p i(j+1) ) is the confidence maximum value of two adjacent points (x i , z j ) and (x i , z j+1 ), and T high is a boundary confidence screening threshold; The ith column boundary point set B i is: (6) Wherein x i is the abscissa of the sampling point; an ordinate of the boundary point detected in the i-th column; the ordinate of the corresponding sampling point is the row index j.
8. The coal rock identification method of claim 5, wherein in step 4.3, the set of boundary points for all columns is: (7) Wherein M is the column number of the sampling grid, and x i is the abscissa of the sampling points; an ordinate of the boundary point detected in the i-th column; The continuous function f (x): (8) K is the index of the B spline basis function; Representing the control coefficient of the B spline curve; Representing a basis function of a B spline curve; the smoothed trajectory function f smoothed (x): (10) Wherein w is the half width of a local window, a m is the SG core coefficient, and m is the relative index in the filtering window.
9. The method of claim 5, wherein in step 4.4, if the confidence level p i ＜T low is classified by the boundary point x i , and the confidence levels p i-1 ≥T low and p i+1 ≥T low ,T low at the adjacent boundary points x i-1 and x i+1 are the boundary restoration confidence level threshold, the trajectory value at the point x i is considered unreliable, and the point value is corrected by using weighted interpolation; confidence weighted interpolation corrected trajectory function f corrected (x): (11) wherein, alpha and beta are weight coefficients.
10. The method for recognizing coal and rock according to claim 1, wherein in the step 5, the movement of the moving support in the vertical direction is controlled to move along the Y axis, the detection is repeated for a plurality of parallel sections with different Y values, and the trajectories z=f corrected (x, Y) of the coal and rock interfaces on the plurality of Y sections are integrated to generate the three-dimensional spatial distribution map of the coal and rock interfaces.

Description

Coal rock identification method Technical Field The invention belongs to the technical fields of mining, mine design construction and mine mining safety, and relates to a coal-rock interface identification method and a coal cutting track determination method. Background The coal-rock interface identification is a key for realizing intelligent mining of the coal mine, and directly influences the quality of the coal, the resource utilization rate and the service life of the coal mining machine. Due to complex geological conditions of a coal seam, a coal cutter is often used for cutting rocks or gangue due to unclear coal-rock interfaces, so that pick wear is increased, efficiency is reduced, and even equipment is damaged. The existing identification technology comprises gamma ray, infrared, image, radar, vibration, acoustic signal, multi-sensor fusion AI and other methods, but all the methods have the limitations that the gamma ray signal is weak, the infrared is easily influenced by motion parameters, the image identification is restricted by environment, the radar is severely attenuated in a thick coal seam, the vibration and acoustic method is easily interfered by noise, and the multi-sensor fusion method has high hardware requirement, large processing delay and information conflict. In addition, most methods are difficult to adapt to the severe environments with high noise, high dust and strong electromagnetic interference in the pit, and have the problems of calculation errors, low reliability, response lag and the like, so that the wide application of the method in actual production is restricted. Disclosure of Invention Aiming at the defects in the prior art, the invention aims to provide the coal and rock identification method, so as to solve the defects in the prior art, improve the mining efficiency and safety of the coal mine, stably work in a severe environment, have high accuracy and good cost benefit, and have important practical significance and wide application prospect for promoting the intelligent and unmanned mining of the coal mine. In order to solve the technical problems, the invention adopts the following technical scheme: The coal rock identification method is based on a coal rock identification device, and the coal rock identification device comprises an X-ray source, a photon counting detector, a moving bracket and a data processing and identification module, wherein the X-ray source and the photon counting detector are arranged on the moving bracket and are arranged on the same side of a coal rock to be detected; the method comprises the following steps: step 1, moving an X-ray source and a photon counting detector to the front of a coal-rock area to be detected through a moving bracket; step 2, the X-ray source emits high-frequency pulse X-rays, namely primary rays I0, according to the set parameters, and irradiates the area to be detected; step 3, the photon counting detector receives Compton back scattering rays emitted by the area to be detected, namely Compton back scattering photons I bs, collects multi-energy spectrum scattering data to extract multi-channel energy spectrum characteristics, and transmits the multi-channel energy spectrum characteristics to the data processing and identifying module; Step 4, the data processing and identifying module analyzes substances in a detected area according to the original ray I0, compton back-scattered photons I bs and the multi-channel energy spectrum characteristics, and carries out coal and rock identification based on a coal and rock interface generating algorithm of multi-channel energy spectrum fusion and boundary regression fitting by combining third-order B spline interpolation and Savitzky-Golay smooth filtering and introducing confidence correction; And 5, moving the X-ray source and the photon counting detector to different areas through the moving support, repeatedly executing the steps on a plurality of parallel sections, integrating coal-rock interface tracks on the sections, generating a three-dimensional space distribution diagram of the coal-rock interface, and providing more comprehensive information for mining planning and geological analysis. The invention also comprises the following technical characteristics: specifically, the X-ray source is used for emitting X-rays, and the energy of the X-rays is adjusted by adjusting the working voltage of the X-ray source; The photon counting detector is used for receiving Compton back scattering rays and measuring the radiation dose and energy distribution of the Compton back scattering rays, is provided with a collimator module, and receives and measures the rays in a specific direction by adjusting the direction of the collimator so as to realize the measurement of substances in a specific area and a specific depth; The motion support is used for supporting the X-ray source and the photon counting detector, and detection of different areas is achieved through movement. Sp