Coal and coal gangue recognition method based on electrical capacitance tomography

Abstract

This invention discloses a method for identifying coal and coal gangue based on capacitance tomography. It utilizes a differential capacitance tomography electrode sensor to collect capacitance measurement data of coal, coal gangue, or mixtures thereof within the measured area. Direct coupling capacitance between electrodes is canceled by applying an inverse excitation signal. The raw data is preprocessed, and stable and valid capacitance measurements are selected to form a data vector. A regularized iterative algorithm is used to solve for the dielectric constant distribution, combining a preset sensitivity distribution function. An initial solution is obtained through regularization, and then iteratively updated until convergence. Image reconstruction is performed based on the dielectric constant distribution. After normalization and edge detection, a capacitance tomography image representing the location, size, and dielectric differences of the materials is generated. The image grayscale values ​​are analyzed, and coal and coal gangue regions are automatically identified and marked according to a preset dielectric constant threshold range. Simultaneously, their centroid coordinates and equivalent areas are calculated, achieving non-destructive and accurate identification of the mixture components.

Description

Technical Field

[0001] This invention relates to the field of geological exploration, and in particular to a method for identifying coal and coal gangue based on capacitance tomography. Background Technology

[0002] In coal mining and sorting, real-time identification of coal gangue directly impacts production efficiency and resource utilization. Existing technologies primarily rely on manual sorting, density sorting, or traditional imaging methods. However, manual sorting is inefficient and susceptible to subjective interference; density sorting equipment is complex and sensitive to particle morphology; and traditional capacitance tomography (CT) technology suffers from significant direct coupling capacitance between sensor electrodes, resulting in noise in the measurement signal, insufficient reconstructed image resolution, and difficulty in accurately distinguishing coal from coal gangue with similar dielectric properties. Furthermore, conventional iterative algorithms exhibit poor stability when solving ill-conditioned equations, further limiting identification accuracy. Therefore, there is an urgent need to develop an automated identification method with strong anti-interference capabilities and high reconstruction accuracy. Summary of the Invention

[0003] This invention proposes a method for identifying coal and coal gangue based on capacitance tomography, comprising: S1. Acquire capacitance measurement data of coal or coal gangue or a mixture of coal and coal gangue in the measured area using a differential capacitance tomography electrode sensor. S2. Preprocess the collected capacitance measurement data to remove invalid measurement data; S3. Using the preset sensitivity distribution function and preprocessed capacitance measurement data, combined with the Landweber iterative algorithm based on Tikhonov regularization, calculate the dielectric constant distribution in the measured area. S4. Based on the calculated dielectric constant distribution, perform image reconstruction to generate a capacitance tomography image that characterizes the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue within the measured area. S5. Analyze the capacitance tomography image to identify the location areas of coal and coal gangue.

[0004] Furthermore, the acquisition of capacitance measurement data of coal or coal gangue or a mixture of coal and coal gangue within the measured area using a differential capacitance tomography electrode sensor specifically includes: S11. Set the excitation source of the differential capacitance tomography electrode sensor to a sinusoidal signal with a preset frequency and a preset voltage value. S12. The sinusoidal signal is input to the upper and lower corresponding electrodes of the differential capacitive tomography electrode sensor, wherein the phase of the signal input to the upper electrode is opposite to the phase of the signal input to the lower electrode. S13. By using the differential capacitance measurement circuit built into the differential capacitance tomography electrode sensor, the direct coupling capacitance between the upper and lower electrode pairs is canceled out, and the output capacitance measurement data only reflects the difference in dielectric constant of the measured coal or coal gangue or a mixture of coal and coal gangue. S14. Use a data acquisition card to record the capacitance measurement data output by the differential capacitance measurement circuit.

[0005] Furthermore, the preprocessing of the collected capacitance measurement data to remove invalid measurement data specifically includes: S21. Receive the raw capacitance measurement data sequence from the data acquisition card. The raw capacitance measurement data sequence contains the capacitance measurement values ​​of all effective electrode pairs of the differential capacitance tomography electrode sensor under a preset number of measurements. S22. Analyze the stability of capacitance measurement values ​​for each electrode pair under multiple measurements, and identify abnormal measurement values ​​with significant fluctuations. S23. According to the preset measurement data stability judgment criteria, discard the measurement data of each electrode pair for the first preset number of times, and only retain the last stable capacitance measurement value as the effective capacitance measurement value. S24. Combine all valid capacitance measurements retained after discarding abnormal data to form a preprocessed capacitance measurement data vector, which will be used for subsequent calculation of dielectric constant distribution.

[0006] Furthermore, the step of using a preset sensitivity distribution function and preprocessed capacitance measurement data, combined with a Landweber iterative algorithm based on Tikhonov regularization, to calculate the dielectric constant distribution within the measured region specifically includes: S31. Obtain the sensitivity distribution function matrix of the differential capacitance tomography electrode sensor, which has been calculated in advance through finite element simulation. S32. Associate the preprocessed capacitance measurement data vector with the sensitivity distribution function matrix to establish a linear equation model describing the relationship between the dielectric constant distribution and the measured capacitance; S33. Solve the linear equation model using the Landweber iterative algorithm based on Tikhonov regularization; S34. The Landweber iterative algorithm based on Tikhonov regularization includes: setting regularization parameters and the maximum number of iterations, and using the initial gray value of the dielectric constant distribution obtained by the Tikhonov regularization algorithm as the starting point of the iteration. S35. In each iteration, estimate the theoretical capacitance value based on the current dielectric constant distribution gray value, and calculate the residual between the theoretical capacitance value and the actual measured capacitance value. S36. Using the product of the transpose of the sensitivity distribution function matrix and the residual vector, combined with the regularization parameter and the step size factor, update the gray value of the dielectric constant distribution. S37. Repeat the iterative update process until the preset maximum number of iterations is reached or the residual meets the convergence condition, and finally output the calculated dielectric constant distribution grayscale image.

[0007] Furthermore, the step of reconstructing the image based on the calculated dielectric constant distribution to generate a capacitance tomography image characterizing the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue within the measured area specifically includes: S41. Receive the grayscale image of dielectric constant distribution calculated from the output of the Landweber iterative algorithm based on Tikhonov regularization; S42. Normalize the dielectric constant distribution grayscale image and map the grayscale values ​​to a preset display range; S43. Apply an image edge detection algorithm to identify the boundaries of high grayscale regions in the grayscale image of dielectric constant distribution, wherein the high grayscale regions correspond to materials with high dielectric constants; S44. Based on the shape and area of ​​the identified high grayscale region boundary, and combined with the preset typical range of dielectric constant values ​​for coal and coal gangue, generate a capacitance tomography image that clearly shows the location of coal and coal gangue, their relative size and spatial relationship.

[0008] Furthermore, the analysis of the capacitance tomography image to identify the location regions of coal and coal gangue specifically includes: S51. Obtain the capacitance tomography image generated in the image reconstruction step; S52. Extract the pixel grayscale values ​​of different regions in the capacitance tomography image; S53. Compare the pixel grayscale value with the preset grayscale threshold range corresponding to the typical dielectric constant of coal and the grayscale threshold range corresponding to the typical dielectric constant of coal gangue. S54. Mark the region whose gray value falls within the gray threshold range corresponding to the typical dielectric constant of coal as the coal location region. S55. Mark the area where the gray value falls within the gray threshold range corresponding to the typical dielectric constant of coal gangue as the coal gangue location area. S56. Calculate the centroid coordinates and equivalent area of ​​each marked region, and output the location information of coal, the location information of coal gangue, and the mixed state information of coal and coal gangue.

[0009] Furthermore, the differential capacitance tomography electrode sensor used in the method specifically includes: two parallel electrode arrays with identical structures; an excitation source interface for applying excitation sinusoidal signals with the same amplitude and opposite phase to the two electrode arrays; a signal output interface for connecting the differential capacitance measurement circuit; and a shielding layer surrounding the electrode arrays. The differential capacitance measurement circuit is used to differentially process the capacitance signal output from the upper electrode array and the capacitance signal output from the lower electrode array, eliminating the direct coupling capacitance component between the electrodes and amplifying the effective capacitance signal component reflecting the change in the dielectric constant of the measured coal or gangue.

[0010] Furthermore, the method also includes constructing a capacitance tomography system, the system comprising: a differential capacitance tomography electrode sensor; an excitation source that generates sinusoidal signals of a preset frequency and a preset voltage value; a data acquisition card that acquires the capacitance signals output by the differential capacitance measurement circuit; a computer that stores a preset sensitivity distribution function matrix, preset algorithm parameters, and preset recognition thresholds; and computer software that runs image reconstruction algorithms and coal gangue recognition algorithms; wherein the differential capacitance tomography electrode sensor, excitation source, data acquisition card, and computer are connected and work together through a preset interface.

[0011] Furthermore, the present invention also proposes a computer-readable storage medium storing a computer program thereon, characterized in that the computer program, when executed by a processor, implements the steps of the coal and coal gangue identification method based on capacitance tomography.

[0012] This invention applies an inverse excitation signal using a differential capacitance tomography electrode sensor, effectively canceling the direct coupling capacitance between electrodes and significantly improving the detection sensitivity of dielectric constant differences. Combined with a Landweber iterative algorithm based on Tikhonov regularization, the regularization parameter suppresses model ill-conditioned behavior, and iterative updates gradually approach a stable solution, greatly improving the calculation accuracy of dielectric constant distribution. High-contrast capacitance tomography images are generated using normalization processing and edge detection techniques, and coal and gangue regions are automatically marked based on a preset grayscale threshold. Simultaneously, the centroid coordinates and equivalent area parameters are output, thereby achieving rapid and non-destructive identification of mixture components and providing a reliable basis for intelligent sorting systems. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating a method for identifying coal and coal gangue based on capacitance tomography proposed in this invention. Figure 2 This is a schematic diagram illustrating the principle of S1, which utilizes a differential capacitance tomography electrode sensor to collect capacitance measurement data of coal or coal gangue or a mixture of coal and coal gangue within the measured area, in a method for identifying coal and coal gangue based on capacitance tomography proposed in this invention. Figure 3 This is the original image showing the state of coal gangue in six comparative experiments in S1 of the coal and coal gangue identification method based on capacitance tomography proposed in this invention. Figure 4 To and Figure 3 The original status display of Group 6 Figure 1 One corresponds to six sets of capacitance tomography images generated after image reconstruction based on the calculated dielectric constant distribution in S4, representing the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue in the tested area. Figure 5 This is a schematic diagram of the sensor used in this invention. Detailed Implementation

[0014] refer to Figure 1-4 This invention proposes a method for identifying coal and coal gangue based on capacitance tomography, comprising: First, in step S1, a differential capacitance tomography electrode sensor is used to acquire capacitance measurement data of coal, coal gangue, or a mixture of coal and coal gangue within the measured area. Then, in step S2, the acquired capacitance measurement data is preprocessed to remove invalid data. Next, in step S3, a preset sensitivity distribution function and the preprocessed capacitance measurement data are used, combined with a Landweber iterative algorithm based on Tikhonov regularization, to calculate the dielectric constant distribution within the measured area. Then, in step S4, image reconstruction is performed based on the calculated dielectric constant distribution to generate a capacitance tomography image characterizing the differences in location, size, and dielectric characteristics of coal, coal gangue, or a mixture of coal and coal gangue within the measured area. Finally, in step S5, the capacitance tomography image is analyzed to identify the location regions of coal and coal gangue.

[0015] The differential capacitance tomography electrode sensor is specifically a sensor with two parallel electrode arrays of identical structure, used to apply excitation signals of the same amplitude but opposite phase to cancel out direct coupling capacitance. The capacitance measurement data is specifically a sequence of capacitance values ​​obtained by the sensor, reflecting the dielectric constant distribution within the measured region. Preprocessing involves filtering the raw capacitance measurement data to remove invalid or unstable measurements. The sensitivity distribution function is specifically a matrix describing the influence of different locations in the sensor space on the capacitance measurement values, obtained in advance through finite element simulation. The Landweber iterative algorithm based on Tikhonov regularization is specifically an image reconstruction algorithm combining Tikhonov regularization and Landweber iteration; the former handles ill-conditioned problems, and the latter approximates the solution stepwise. The dielectric constant distribution is specifically a numerical distribution map characterizing the dielectric properties of materials at various points within the measured region, such as the difference between coal and coal gangue. Image reconstruction specifically generates a visualized image using the calculated dielectric constant distribution. A capacitance tomography image is a reconstructed image reflecting the differences in location, size, and dielectric characteristics of materials within a measured area. These materials include coal, coal gangue, and mixtures thereof. Identification specifically involves distinguishing the location areas of coal and coal gangue based on the grayscale or regional characteristics of the capacitance tomography image.

[0016] In specific implementation, this invention utilizes a differential capacitance tomography electrode sensor to collect capacitance change data caused by coal, coal gangue, or mixtures. The collected raw data first undergoes a preprocessing step, such as analyzing the stability of multiple measurements and discarding potentially fluctuating data from the earlier measurements, retaining only the last stable measurement value to form an effective capacitance measurement vector. Next, using a sensitivity matrix pre-calculated through finite element simulation, the capacitance measurement vector is correlated with the dielectric constant distribution. Then, the Landweber iterative algorithm based on Tikhonov regularization is applied to solve this correlation equation: the first step uses Tikhonov regularization to obtain an initial grayscale estimate of the dielectric constant distribution; starting from this initial value, iteration is performed, each iteration including three steps: calculating the theoretical capacitance value corresponding to the current grayscale estimate and obtaining the residual with the actual measured capacitance value; using the transpose of the sensitivity matrix and the product of the residual vector, combined with the regularization parameter and step size factor, the grayscale value of the dielectric constant distribution is updated; this iterative process is repeated until the preset maximum number of iterations is reached or the residual satisfies the convergence condition, finally outputting a grayscale image of the dielectric constant distribution. Image reconstruction is performed based on this grayscale image, including normalization and edge detection, to generate a capacitance tomography image. Finally, by analyzing the grayscale values ​​of each region in the image and comparing them with the grayscale threshold range corresponding to the preset typical values ​​of the dielectric constants of coal and coal gangue, the location regions of coal and coal gangue can be identified and marked, and their centroid coordinates and equivalent areas can be calculated.

[0017] Further details of step S1: The acquisition of capacitance measurement data for coal, coal gangue, or a mixture of coal and coal gangue within the measured area using a differential capacitance tomography electrode sensor specifically includes: In stage S11, setting the excitation source of the differential capacitance tomography electrode sensor to a sinusoidal signal with a preset frequency and preset voltage value; In stage S12, inputting the sinusoidal signal to the upper and lower corresponding electrodes of the differential capacitance tomography electrode sensor, wherein the signal phase input to the upper electrode is opposite to the signal phase input to the lower electrode; In stage S13, using the differential capacitance measurement circuit built into the differential capacitance tomography electrode sensor, canceling the direct coupling capacitance between the upper and lower electrode pairs, and outputting capacitance measurement data that only reflects the difference in dielectric constant of the measured coal, coal gangue, or mixture; In stage S14, using a data acquisition card to record the capacitance measurement data output by the differential capacitance measurement circuit.

[0018] Here, the preset frequency and preset voltage values ​​of the sinusoidal signal specifically refer to the excitation signal parameters set according to system requirements, such as a 350kHz frequency and a 6V voltage. The upper and lower corresponding electrodes specifically refer to paired electrodes in a differential sensor located in the same spatial position but belonging to upper and lower layers. Opposite signal phase specifically means that the sinusoidal signals applied to the upper and lower corresponding electrodes differ in phase by 180 degrees. The differential capacitance measurement circuit is specifically a circuit design used to receive signals from the upper and lower electrodes and perform differential processing. Direct coupling capacitance specifically refers to the inherent capacitive coupling component between the sensor electrodes, independent of the measured object. The data acquisition card is specifically a hardware device used to acquire and digitize capacitance measurement signals, such as the PXI5112.

[0019] In practical operation, when acquiring capacitance measurement data, the excitation source is first set to a specific sinusoidal signal (e.g., 350kHz, 6V). This signal is then input to the corresponding upper and lower electrodes of the differential sensor, with the key point being that the signal input to the upper electrode is strictly out of phase with the signal input to the lower electrode. The sensor's built-in differential capacitance measurement circuit simultaneously receives signals from both electrodes. Due to the symmetrical structure of the upper and lower electrodes and the application of opposite-phase excitation signals, the direct coupling capacitance components between them are canceled out in the differential circuit. At this point, the signal output by the differential circuit primarily reflects the capacitance change component caused by the different dielectric properties of the measured object, i.e., the effective capacitance measurement data. This data is ultimately recorded and digitized by a data acquisition card, providing input for subsequent processing.

[0020] Further details of step S2: Preprocessing the acquired capacitance measurement data to remove invalid measurement data specifically includes: In stage S21, receiving the raw capacitance measurement data sequence from the data acquisition card, the raw capacitance measurement data sequence containing the capacitance measurement values ​​of all valid electrode pairs of the differential capacitance tomography electrode sensor under a preset number of measurements; In stage S22, analyzing the stability of the capacitance measurement values ​​of each electrode pair under multiple measurements, identifying abnormal measurement values ​​with significant fluctuations; In stage S23, according to the preset measurement data stability judgment criteria, discarding the measurement data of each electrode pair for the first preset number of measurements, retaining only the last stable capacitance measurement value as the valid capacitance measurement value; In stage S24, combining all the valid capacitance measurement values ​​retained after discarding abnormal data to form a preprocessed capacitance measurement data vector, which is used for subsequent calculation of the dielectric constant distribution.

[0021] Here, the raw capacitance measurement data sequence specifically refers to the raw dataset recorded by the data acquisition card, containing the capacitance values ​​of all valid electrode pairs under multiple measurements, with a default measurement count of 3. An electrode pair specifically refers to a pair of electrodes used for excitation and detection in the sensor. The preset measurement count specifically refers to the number of times each electrode pair is measured repeatedly during data acquisition. Significantly fluctuating abnormal measurement values ​​specifically refer to individual data points that deviate significantly from stable values ​​in multiple measurements. The measurement data stability criterion specifically refers to the rules for judging whether the data is stable, such as a strategy of discarding the first two measurements and retaining only the last data. Valid capacitance measurement values ​​specifically refer to the single capacitance measurement values ​​that are considered reliable after being filtered according to the stability criterion. The preprocessed capacitance measurement data vector specifically refers to a vector composed of the valid capacitance measurement values ​​of all electrode pairs in sequence.

[0022] In practice, when preprocessing the acquired raw capacitance measurement data, the system first receives a raw data sequence from the data acquisition card. This sequence contains the capacitance measurements of all valid electrode pairs of the sensor under a preset number of measurements. Next, the stability of the data for each electrode pair under multiple measurements is analyzed, for example, by calculating the variance of the three measurements or observing their trends, identifying outliers with significant deviations or fluctuations. Then, the data is processed according to a preset data stability criterion. The method used is to discard the first two measurements for each electrode pair, retaining only the last stable measurement as the valid capacitance measurement value for that electrode pair. Finally, the retained valid capacitance measurements for all electrode pairs are combined and arranged in a specific order to form a preprocessed capacitance measurement data vector. This vector will be used for subsequent dielectric constant distribution calculations.

[0023] Further details of step S3: Using a preset sensitivity distribution function and preprocessed capacitance measurement data, combined with a Landweber iterative algorithm based on Tikhonov regularization, the calculation of the dielectric constant distribution within the measured region specifically includes: In stage S31, obtaining the differential capacitance tomography electrode sensor sensitivity distribution function matrix obtained in advance through finite element simulation; In stage S32, associating the preprocessed capacitance measurement data vector with the sensitivity distribution function matrix to establish a linear equation model describing the relationship between the dielectric constant distribution and the measured capacitance; In stage S33, applying the Landweber iterative algorithm based on Tikhonov regularization to the linear equation model. The equation model is solved; the algorithm includes the following sub-steps: S34: Set the regularization parameter and the maximum number of iterations, and use the initial gray value of the dielectric constant distribution obtained by the preliminary solution of the Tikhonov regularization algorithm as the starting point of the iteration; S35: In each iteration, estimate the theoretical capacitance value based on the current gray value of the dielectric constant distribution, and calculate the residual between the theoretical capacitance value and the actual measured capacitance value; S36: Update the gray value of the dielectric constant distribution by using the product of the transpose of the sensitivity distribution function matrix and the residual vector, combined with the regularization parameter and the step size factor; S37: Repeat the iterative update process until the preset maximum number of iterations is reached or the residual meets the convergence condition, and finally output the calculated dielectric constant distribution gray map.

[0024] Here, finite element simulation specifically refers to using numerical calculation methods to simulate the sensor's physical field on a computer to obtain the sensitivity distribution. The sensitivity distribution function matrix is ​​specifically a matrix pre-calculated through finite element simulation, describing the influence of spatial grid points on the capacitance of each electrode pair. The linear equation model is specifically an equation describing the relationship between the capacitance measurement vector, the sensitivity matrix, and the dielectric constant distribution vector. The regularization parameter is specifically a coefficient introduced in Tikhonov regularization to control the smoothness and stability of the solution. The maximum number of iterations is specifically the maximum number of steps allowed by the Landweber iterative algorithm. The initial grayscale value of the dielectric constant distribution is specifically the estimated value of the dielectric constant distribution obtained through the preliminary solution of the Tikhonov regularization algorithm. The iteration start point is specifically the initial value at which the Landweber iterative algorithm begins calculation. The theoretical capacitance value is specifically the capacitance value calculated based on the currently estimated dielectric constant distribution grayscale value and the sensitivity matrix. The residual is specifically the difference vector between the theoretical capacitance value and the actual pre-processed measured capacitance value. The step size factor is specifically a coefficient controlling the update amplitude in the Landweber iteration. The dielectric constant distribution grayscale image is specifically a numerical matrix or image that is finally calculated and represents the distribution of dielectric constant within the measured area.

[0025] In practice, when calculating the dielectric constant distribution, the sensitivity distribution function matrix of the differential sensor, pre-calculated through finite element simulation, is first obtained. The pre-processed capacitance measurement data vector is then associated with this matrix to establish a linear equation model. Next, the Landweber iterative algorithm based on Tikhonov regularization is applied to solve this model: Regularization parameters and a maximum number of iterations are set. Initial grayscale values ​​of the dielectric constant distribution are calculated using the Tikhonov regularization formula as the starting point for iteration. During each iteration, the theoretical capacitance value is calculated based on the current grayscale value and the sensitivity matrix, and the residual between this value and the actual measured capacitance value is obtained. Then, the grayscale values ​​of the dielectric constant distribution are updated according to the Landweber iterative formula using the product of the transpose of the sensitivity matrix and the residual vector, combined with the regularization parameters and step size factor. This iterative update process is repeated until the maximum number of iterations is reached or the residual satisfies the preset convergence condition. Finally, the calculated dielectric constant distribution grayscale image is output.

[0026] Further details of step S4: Image reconstruction based on the calculated dielectric constant distribution to generate a capacitance tomography image representing the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue within the measured area specifically includes: In stage S41, receiving a grayscale image of the dielectric constant distribution calculated by the Landweber iterative algorithm based on Tikhonov regularization; In stage S42, normalizing the grayscale image of the dielectric constant distribution to map the grayscale values ​​to a preset display range; In stage S43, applying an image edge detection algorithm to identify the boundaries of high grayscale regions in the dielectric constant distribution grayscale image, where the high grayscale regions correspond to substances with higher dielectric constants; In stage S44, based on the shape and area of ​​the identified high grayscale region boundaries, combined with the preset typical range of dielectric constant values ​​for coal and coal gangue, generating a capacitance tomography image that clearly displays the location areas of coal and coal gangue, their relative sizes, and spatial relationships.

[0027] Here, the dielectric constant distribution grayscale image is specifically the numerical matrix calculated in the preceding steps, where each element represents the relative magnitude of the dielectric constant at each spatial location. Normalization specifically maps grayscale values ​​linearly or non-linearly to a standard range for display and analysis, such as 0-255. The preset display range is the target range of normalized grayscale values. The image edge detection algorithm is an algorithm used to identify region boundaries in the image, such as the Canny or Sobel algorithm. High grayscale regions are regions with higher normalized grayscale values, typically corresponding to materials with higher dielectric constants, such as coal gangue. Boundaries are the dividing lines between high and non-high grayscale regions. The preset typical dielectric constant range for coal and coal gangue is the dielectric constant interval for coal and coal gangue determined experimentally or empirically, used to guide image interpretation. Relative size is the area ratio of the coal region and the coal gangue region displayed in the image. Spatial relationships specifically refer to the relative positions of coal and gangue regions in the image, such as separation, adjacency, or overlap.

[0028] In practice, when reconstructing an image based on the calculated dielectric constant distribution grayscale image, the grayscale image is received first. Then, the grayscale image is normalized, for example, by linearly stretching or mapping the calculated grayscale value range to a display range of 0-255 to facilitate visualization. Next, an image edge detection algorithm is applied to identify the boundary contours of high grayscale regions in the normalized grayscale image. Based on the identified boundary shapes and calculated region areas, and referring to the preset typical range of dielectric constant values ​​for coal and coal gangue, a capacitance tomography image is finally generated. This image clearly shows the location of coal and coal gangue, as well as their relative sizes and spatial distribution relationships.

[0029] Further details of step S5: Analyzing the capacitance tomography image to identify the location regions of coal and gangue specifically includes: In stage S51, acquiring the capacitance tomography image generated in the image reconstruction step; in stage S52, extracting the pixel grayscale values ​​of different regions in the capacitance tomography image; in stage S53, comparing the pixel grayscale values ​​with preset grayscale threshold ranges corresponding to the typical dielectric constant of coal and the typical dielectric constant of gangue; in stage S54, marking regions whose grayscale values ​​fall within the grayscale threshold range corresponding to the typical dielectric constant of coal as coal location regions; in stage S55, marking regions whose grayscale values ​​fall within the grayscale threshold range corresponding to the typical dielectric constant of gangue as gangue location regions; in stage S56, calculating the centroid coordinates and equivalent area of ​​each marked region, and outputting the location region information of coal, the location region information of gangue, and the mixed state information of coal and gangue.

[0030] Here, the pixel grayscale value specifically refers to the brightness value of each pixel in the capacitance tomography image, representing the relative magnitude of the dielectric constant at that location. The preset grayscale threshold range corresponding to the typical dielectric constant of coal is a grayscale value range pre-defined based on the dielectric properties of coal. Values ​​below the lower limit of this range may represent background or air, while values ​​above the upper limit may represent coal gangue. The preset grayscale threshold range corresponding to the typical dielectric constant of coal gangue is a grayscale value range pre-defined based on the dielectric properties of coal gangue. Marking specifically refers to identifying specific areas in an image or data in a specific way, such as using color or labels. The centroid coordinates are the average of all pixel coordinates within the marked area, representing the center position of that area. The equivalent area is the total number of pixels occupied by the marked area or the converted actual area, representing the size of the area. Location area information specifically includes data describing location and size, such as centroid coordinates and equivalent area. The coal and coal gangue mixing state information specifically describes whether the two coexist and their spatial distribution relationship.

[0031] In practice, when analyzing capacitance tomography images to identify the location regions of coal and gangue, the reconstructed image is first acquired. The grayscale values ​​of pixels in different regions of the image are extracted. These grayscale values ​​are compared with two preset threshold ranges: one corresponding to the typical dielectric constant of coal, and the other to the typical dielectric constant of gangue. Pixel regions whose grayscale values ​​fall within the coal threshold range are marked as coal location regions. Pixel regions whose grayscale values ​​fall within the gangue threshold range are marked as gangue location regions. For each marked region, the average of all its pixel coordinates is calculated to obtain the centroid coordinates, and the number of pixels within the region is counted to obtain the equivalent area. The final output contains information on the location regions of both coal and gangue, and the mixing state of the two can be analyzed based on this information.

[0032] Further explanation of the sensor structure: The differential capacitance tomography electrode sensor used in the method specifically includes: two parallel electrode arrays with identical structures; an excitation source interface for applying excitation sinusoidal signals of the same amplitude but opposite phase to the two electrode arrays; a signal output interface for connecting the differential capacitance measurement circuit; and a shielding layer surrounding the electrode arrays. The differential capacitance measurement circuit is used to differentially process the capacitance signal output from the upper electrode array and the capacitance signal output from the lower electrode array, eliminating the direct coupling capacitance component between the electrodes and amplifying the effective capacitance signal component reflecting the change in the dielectric constant of the measured coal or gangue.

[0033] Here, the two parallel and identical electrode arrays constitute the core of the sensor, consisting of two geometrically identical electrode arrays placed parallel to each other. The excitation source interface is the electrical interface on the sensor used to connect to an external excitation signal source. The signal output interface is the electrical interface on the sensor used to connect to the differential capacitance measurement circuit. The shielding layer is a conductive layer surrounding the electrode array, used to shield against external electromagnetic interference. The differential capacitance measurement circuit is an electronic circuit that receives two capacitance signals from the upper and lower electrodes, cancels the common-mode signal (i.e., the directly coupled capacitance) through differential amplification, and amplifies the differential-mode signal, which is the effective signal reflecting the change in the dielectric constant of the measured object.

[0034] In practice, this differential capacitance tomography electrode sensor consists of two parallel electrode arrays with identical geometry. An external excitation source applies a sinusoidal signal of the same amplitude but opposite phase to these two electrode arrays through an excitation source interface. When the analyte is placed near the sensor, its dielectric properties change the capacitance between the electrodes. The capacitance signals sensed by the upper and lower electrode arrays are transmitted to the differential capacitance measurement circuit through a signal output interface. The core function of this circuit is to differentially process the signal from the upper electrode with the signal from the lower electrode. Due to the applied anti-phase excitation and symmetrical structure, the inherent direct coupling capacitance component between the upper and lower electrodes is significantly canceled out during the differential processing. Simultaneously, the effective capacitance change component caused by the analyte is extracted and amplified by the differential amplifier circuit. The entire sensor is surrounded by a shielding layer to reduce external electromagnetic interference. This design effectively improves the detection sensitivity for differences in dielectric constants between coal and coal gangue.

[0035] Further explanation of system construction: The method also requires the construction of a capacitance tomography system, which includes: a differential capacitance tomography electrode sensor; an excitation source that generates sinusoidal signals with preset frequency and preset voltage values; a data acquisition card that acquires the capacitance signals output by the differential capacitance measurement circuit; a computer that stores a preset sensitivity distribution function matrix, preset algorithm parameters, and preset recognition thresholds; and computer software that runs image reconstruction algorithms and coal gangue recognition algorithms. The differential capacitance tomography electrode sensor, excitation source, data acquisition card, and computer are connected and work together through a preset interface.

[0036] Here, the capacitance tomography system specifically refers to a complete hardware and software platform consisting of the aforementioned sensors, excitation source, data acquisition card, computer, and software. The excitation source is specifically a device that generates sinusoidal signals of a specific frequency and voltage. The data acquisition card is specifically hardware that acquires and digitizes the sensor output signals. The computer is specifically a device that runs the algorithm software. The preset sensitivity distribution function matrix is ​​specifically a pre-calculated sensitivity matrix stored in the computer. The preset algorithm parameters are specifically parameters stored in the computer for use by the reconstruction and recognition algorithms, including Tikhonov parameters, Landweber step size, maximum number of iterations, and grayscale threshold. The preset recognition threshold is specifically a grayscale threshold range stored in the computer used to distinguish between coal and coal gangue. The computer software is specifically an application program that implements the method steps described above. The preset interface is specifically a physical or electrical connection method for connecting the various hardware components, such as a USB, BNC, or PXI interface.

[0037] In specific deployment, the constructed capacitance tomography system includes the aforementioned differential capacitance tomography electrode sensor, an excitation source capable of generating sinusoidal signals of specific frequencies and voltages, a data acquisition card for acquiring and digitizing the differential capacitance signals output by the sensor, a computer storing a preset sensitivity distribution function matrix, preset algorithm parameters, and preset recognition thresholds, and software running on the computer to implement the aforementioned method steps. These components are connected and work together through preset interfaces: the excitation source drives the sensor, the sensor senses changes in the measured object and outputs signals, the data acquisition card acquires and converts signals, and the computer runs software to process data, calculate dielectric distribution, reconstruct images, identify coal gangue, and output results.

[0038] Finally, the present invention also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the described method for identifying coal and coal gangue based on capacitance tomography.

[0039] Computer-readable storage media specifically refers to physical or electronic media capable of storing computer programs, such as hard disks, solid-state drives, USB flash drives, optical discs, or memory. A computer program specifically refers to a set of code containing executable instructions. A processor specifically refers to a central processing unit or graphics processing unit that executes the computer program instructions. The implementation method specifically involves ensuring that when the processor executes the program on the storage medium, it can completely execute all the operational processes of the aforementioned coal and coal gangue identification method based on capacitance tomography.

[0040] The specific implementation is as follows: a specific computer program is stored on the computer-readable storage medium. When the program is loaded and executed by the processor, the processor will, according to the program instructions, progressively implement all the steps of the coal and gangue identification method based on capacitance tomography described in detail above. This includes controlling data acquisition, performing data preprocessing, calculating the dielectric constant distribution using preset parameters and algorithms, performing image reconstruction, analyzing the image and identifying coal and gangue regions, as well as possible sensor configuration or system control operations. The result of the program execution is the completion of the coal and gangue identification task.

[0041] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for identifying coal and coal gangue based on capacitance tomography, characterized in that, include: S1. Acquire capacitance measurement data of coal or coal gangue or a mixture of coal and coal gangue in the measured area using a differential capacitance tomography electrode sensor. S2. Preprocess the collected capacitance measurement data to remove invalid measurement data; S3. Using the preset sensitivity distribution function and preprocessed capacitance measurement data, combined with the Landweber iterative algorithm based on Tikhonov regularization, calculate the dielectric constant distribution in the measured area. S4. Based on the calculated dielectric constant distribution, perform image reconstruction to generate a capacitance tomography image that characterizes the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue within the measured area. S5. Analyze the capacitance tomography image to identify the location areas of coal and coal gangue.

2. The method as described in claim 1, characterized in that, The acquisition of capacitance measurement data of coal, coal gangue, or a mixture of coal and coal gangue within the measured area using a differential capacitance tomography electrode sensor specifically includes: S11. Set the excitation source of the differential capacitance tomography electrode sensor to a sinusoidal signal with a preset frequency and a preset voltage value. S12. The sinusoidal signal is input to the upper and lower corresponding electrodes of the differential capacitive tomography electrode sensor, wherein the phase of the signal input to the upper electrode is opposite to the phase of the signal input to the lower electrode. S13. By using the differential capacitance measurement circuit built into the differential capacitance tomography electrode sensor, the direct coupling capacitance between the upper and lower electrode pairs is canceled out, and the output capacitance measurement data only reflects the difference in dielectric constant of the measured coal or coal gangue or a mixture of coal and coal gangue. S14. Use a data acquisition card to record the capacitance measurement data output by the differential capacitance measurement circuit.

3. The method as described in claim 1, characterized in that, The preprocessing of the collected capacitance measurement data to remove invalid measurement data specifically includes: S21. Receive the raw capacitance measurement data sequence from the data acquisition card. The raw capacitance measurement data sequence contains the capacitance measurement values ​​of all effective electrode pairs of the differential capacitance tomography electrode sensor under a preset number of measurements. S22. Analyze the stability of capacitance measurement values ​​for each electrode pair under multiple measurements, and identify abnormal measurement values ​​with significant fluctuations. S23. According to the preset measurement data stability judgment criteria, discard the measurement data of each electrode pair for the first preset number of times, and only retain the last stable capacitance measurement value as the effective capacitance measurement value. S24. Combine all valid capacitance measurements retained after discarding abnormal data to form a preprocessed capacitance measurement data vector, which will be used for subsequent calculation of dielectric constant distribution.

4. The method as described in claim 1, characterized in that, The calculation of the dielectric constant distribution within the measured region using a preset sensitivity distribution function and preprocessed capacitance measurement data, combined with a Landweber iterative algorithm based on Tikhonov regularization, specifically includes: S31. Obtain the sensitivity distribution function matrix of the differential capacitance tomography electrode sensor, which has been calculated in advance through finite element simulation. S32. Associate the preprocessed capacitance measurement data vector with the sensitivity distribution function matrix to establish a linear equation model describing the relationship between the dielectric constant distribution and the measured capacitance; S33. Solve the linear equation model using the Landweber iterative algorithm based on Tikhonov regularization; S34. The Landweber iterative algorithm based on Tikhonov regularization includes: setting regularization parameters and the maximum number of iterations, and using the initial gray value of the dielectric constant distribution obtained by the Tikhonov regularization algorithm as the starting point of the iteration. S35. In each iteration, estimate the theoretical capacitance value based on the current dielectric constant distribution gray value, and calculate the residual between the theoretical capacitance value and the actual measured capacitance value. S36. Using the product of the transpose of the sensitivity distribution function matrix and the residual vector, combined with the regularization parameter and the step size factor, update the gray value of the dielectric constant distribution. S37. Repeat the iterative update process until the preset maximum number of iterations is reached or the residual meets the convergence condition, and finally output the calculated dielectric constant distribution grayscale image.

5. The method as described in claim 1, characterized in that, The step of reconstructing the image based on the calculated dielectric constant distribution to generate a capacitance tomography image characterizing the differences in location, size, and dielectric characteristics of coal or coal gangue or a mixture of coal and coal gangue within the measured area specifically includes: S41. Receive the grayscale image of dielectric constant distribution calculated from the output of the Landweber iterative algorithm based on Tikhonov regularization; S42. Normalize the dielectric constant distribution grayscale image and map the grayscale values ​​to a preset display range; S43. Apply an image edge detection algorithm to identify the boundaries of high grayscale regions in the grayscale image of dielectric constant distribution, wherein the high grayscale regions correspond to materials with high dielectric constants; S44. Based on the shape and area of ​​the identified high grayscale region boundary, and combined with the preset typical range of dielectric constant values ​​for coal and coal gangue, generate a capacitance tomography image that clearly shows the location of coal and coal gangue, their relative size and spatial relationship.

6. The method as described in claim 1, characterized in that, The analysis of the capacitance tomography image to identify the location regions of coal and coal gangue specifically includes: S51. Obtain the capacitance tomography image generated in the image reconstruction step; S52. Extract the pixel grayscale values ​​of different regions in the capacitance tomography image; S53. Compare the pixel grayscale value with the preset grayscale threshold range corresponding to the typical dielectric constant of coal and the grayscale threshold range corresponding to the typical dielectric constant of coal gangue. S54. Mark the region whose gray value falls within the gray threshold range corresponding to the typical dielectric constant of coal as the coal location region. S55. Mark the area where the gray value falls within the gray threshold range corresponding to the typical dielectric constant of coal gangue as the coal gangue location area. S56. Calculate the centroid coordinates and equivalent area of ​​each marked region, and output the location information of coal, the location information of coal gangue, and the mixed state information of coal and coal gangue.

7. The method as described in claim 1, characterized in that, The differential capacitance tomography electrode sensor used in the method specifically includes: two parallel electrode arrays with identical structures; an excitation source interface for applying excitation sinusoidal signals with the same amplitude and opposite phase to the two electrode arrays; a signal output interface for connecting to the differential capacitance measurement circuit; and a shielding layer surrounding the electrode arrays. The differential capacitance measurement circuit is used to differentially process the capacitance signal output from the upper electrode array and the capacitance signal output from the lower electrode array, eliminating the direct coupling capacitance component between the electrodes and amplifying the effective capacitance signal component reflecting the change in the dielectric constant of the measured coal or gangue.

8. The method as described in claim 1, characterized in that, The method further includes constructing a capacitance tomography system, the system comprising: a differential capacitance tomography electrode sensor; an excitation source that generates sinusoidal signals with preset frequency and preset voltage values; a data acquisition card that acquires capacitance signals output by a differential capacitance measurement circuit; a computer that stores a preset sensitivity distribution function matrix, preset algorithm parameters, and preset recognition thresholds; and computer software that runs image reconstruction algorithms and coal gangue recognition algorithms; wherein the differential capacitance tomography electrode sensor, excitation source, data acquisition card, and computer are connected and work together through a preset interface.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for identifying coal and coal gangue based on capacitance tomography as described in any one of claims 1 to 8.

Images