An online coal quality detection method based on adaptive focused laser induced breakdown spectroscopy

Abstract

The present invention relates to the technical field of coal quality detection, and in particular to an online coal quality detection method using adaptive focused laser induced breakdown spectroscopy. The present invention proposes the following scheme: firstly, the position of a laser focal point is obtained by image processing technology; secondly, a plasma is obtained by emitting laser light through a laser, and a spectral signal of the plasma is obtained; finally, the spectral signal of the plasma is processed to extract signal characteristics of internal components of the coal to be detected; and according to the signal characteristics of the internal components of the coal, the coal quality of the coal to be detected is analyzed to determine the coal quality of the coal to be detected.

Description

Technical Field

[0001] The invention relates to the technical field of coal quality detection, and in particular to an online coal quality detection method using adaptive focused laser induced breakdown spectroscopy. Background Art

[0002] Rapid coal quality detection is crucial to achieve low-carbon, flexible and intelligent operation of coal-fired power plants. Laser induced breakdown spectroscopy (LIBS) has broad application prospects in the field of rapid coal quality detection due to its advantages such as good safety, simple sample pretreatment, and simultaneous rapid analysis of multiple elements. However, the complex characteristics of coal exacerbate the interference of spectral data uncertainty on quantitative analysis, and it cannot be fully characterized by a single spectral information, which restricts the further improvement of coal quality detection accuracy.

[0003] For example, the Chinese patent with the publication number CN114965440A discloses a coal quality detection method based on the LIBS coal quality analysis system, which belongs to the field of measurement technology and includes a laser, a spectrometer, a powder supply device and a dichroic mirror. The powder supply device includes a powder drop port, the laser is arranged to align with the powder drop port, and the emission direction of the laser is perpendicular to the powder output direction. The dichroic mirror is arranged between the laser and the powder supply device, the spectrometer is aligned with the dichroic mirror, and also includes a collimating lens group, the collimating lens group includes a first collimating lens and a second collimating lens, the first collimating lens is arranged between the dichroic mirror and the spectrometer, the first collimating lens is close to the dichroic mirror, and the distance between the second collimating lens and the dichroic mirror is greater than the first collimating lens. The coal quality detection method includes powder output; laser focusing spectrum acquisition; information transmission; information processing, and can stably acquire the spectral signal generated by the ablated coal powder, overcoming the problem of drift in the acquisition of the spectral signal of the ablated coal powder.

[0004] The above patents have the problems raised by this background technology: when the focusing position is manually selected, the coal may not be penetrated or the spectral data may be inaccurate. The processing of spectral data depends on quantitative analysis models, and the estimation of the internal composition of the coal is inaccurate. To solve the above problems, this application designs an online coal quality detection method using adaptive focused laser-induced breakdown spectroscopy. Summary of the invention

[0005] The technical problem to be solved by the present invention is to address the deficiencies in the prior art and provide an online coal quality detection method using adaptive focused laser induced breakdown spectroscopy. First, the position of the laser focal point is obtained by image processing technology. Secondly, a laser is emitted to obtain plasma, and a spectral signal of the plasma is obtained. Finally, the spectral signal of the plasma is processed to extract the signal characteristics of the internal components of the coal to be detected. The coal to be detected is analyzed according to the signal characteristics of the internal components of the coal to determine the coal quality to be detected.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An online coal quality detection method using adaptive focused laser induced breakdown spectroscopy comprises the following steps:

[0008] S1: acquiring a surface image of the coal to be detected by an image sensor, processing the surface image of the coal to be detected, and acquiring the position of the laser focus point;

[0009] S2: emitting laser light through a laser, the laser light is reflected and transmitted through a reflector, and focused at the position of the laser focus point to ablate the coal to be inspected, thereby forming plasma;

[0010] S3: a light receiving probe receives light emitted by the plasma and transmits the light to a spectrometer, and the spectrometer obtains a spectrum signal of the plasma through photoelectric conversion;

[0011] S4: Processing the spectrum signal of the plasma to extract the signal characteristics of the internal components of the coal to be detected, performing coal quality analysis on the coal to be detected according to the signal characteristics of the internal components of the coal, and determining the coal quality of the coal to be detected.

[0012] The specific steps of processing the surface image of the coal to be detected are as follows:

[0013] S1.1: Preprocessing the surface image of the coal to be inspected to obtain a surface compressed reconstructed image, wherein the preprocessing includes denoising, image sparse representation, image cutting and image sampling compression;

[0014] S1.2: converting the surface compression reconstructed image into a surface row gradient map through a row gradient operator, checking the gradient response of each pixel in the surface row gradient map in the gradient direction, setting a response threshold, setting the gradient response in the surface row gradient map that is less than the response threshold to zero, and only retaining the gradient response in the surface row gradient map that is greater than or equal to the response threshold;

[0015] S1.3: Adaptively binarize the surface gradient map and perform edge detection, extract the continuous curve in the surface gradient map, perform straight line fitting on the curve, obtain the straight line and significant vanishing line in the surface gradient map, and use the intersection point of the significant vanishing line and the straight line as the key point of the surface gradient map;

[0016] S1.4: Calculate the color features and texture features of each key point, and calculate the spectral intensity of the key point based on the color features and texture features, sort the spectral intensities of the key points from small to large, and take the key point with the smallest spectral intensity as the position of the laser focus point.

[0017] The calculation formula of the spectral intensity of the key point is:

[0018]

[0019] Among them, S i represents the spectral intensity of the key point, It represents the channel characteristics of the key point h channel after the third-level wavelet transform. Represents the channel characteristics of the key point s channel after the third-level wavelet transform, It represents the channel characteristics of the key point v channel after the third-level wavelet transform, E represents the grayscale distribution entropy of the key point, A represents the color characteristics of the key point, B represents the texture characteristics of the key point, and h, s and v represent different channels of the key point.

[0020] The specific steps of S1.1 are as follows:

[0021] S1.1.1: De-noising the surface image of the coal to be detected, wherein the noise includes on-site light noise, fly ash noise inside the coal, and photoelectric tube noise of the image sensor;

[0022] S1.1.2: Sparsely represent the denoised surface image of the coal to be tested, and perform mask sampling through Fourier transform to obtain a sparse surface image;

[0023] S1.1.3: Cut the sparse surface image, cut the original image of size h*w into blocks according to the size of p*q, obtain the block image matrix, construct the observation matrix sequence, match different sampling rates according to the composition complexity of different areas of the block image, and perform sparse sampling compression in the Fourier domain in a distributed multi-point collaborative manner. The calculation formula for sparse sampling compression is:

[0024]

[0025] Where G represents the surface compression reconstructed image, a represents a single block image matrix, M represents the total number of block image matrices, and p a represents the sampling rate of the a-th block image matrix, θ represents the observation matrix sequence, represents the vector dot product operation, K represents the division rate, F[·] represents the Fourier transform function, I represents the external image of the ultrasonic motor, and γ represents the noise.

[0026] The specific steps of processing the spectrum signal of the plasma are as follows:

[0027] S4.1: preprocessing the spectrum signal of the plasma, wherein the preprocessing includes spectrum signal normalization, multivariate scattering correction and spectrum signal smoothing;

[0028] S4.2: converting the preprocessed spectral signal into a spectral frequency domain by Fourier transform, converting the spectral frequency domain into a spectral frequency distribution by spectrum analysis, and drawing a spectral frequency distribution histogram according to the spectral frequency distribution;

[0029] S4.3: extracting a characteristic spectrum curve according to the spectrum frequency distribution histogram, and determining the element corresponding to the characteristic spectrum curve through an atomic spectrum database;

[0030] S4.4: Establish a characteristic vector based on the characteristic spectral curve, perform Hilbert transform on the characteristic vector to calculate the envelope of the characteristic spectral curve, segment it according to the envelope, calculate the maximum value and mean value of each segment, and calculate the signal characteristics of the internal components of the coal to be detected based on the maximum value and mean value of each segment.

[0031] The calculation formula of the signal characteristics of the internal components of the coal to be detected is:

[0032]

[0033] Among them, I c represents the signal characteristics of coal internal components, i represents a single envelope, n represents the total number of envelopes of the characteristic spectrum curve, fi represents the frequency distribution interval of the i-th envelope, max i Indicates the maximum value of the element signal of the i-th envelope, max avg Represents the average value of the maximum value of all envelope element signals, avg i represents the average value of the element signal of the ith envelope, η represents a constant greater than zero, δ(·) represents the Dirac function, t i represents the horizontal axis coordinate of the ith envelope, j represents the imaginary unit, and ω represents the center frequency.

[0034] The coal quality analysis of the coal to be tested according to the coal internal component signal characteristics includes coal quality component classification and coal quality analysis;

[0035] The coal quality components include ash, volatile matter and moisture;

[0036] The coal quality includes high-quality coal, medium-quality coal, low-quality coal and inferior coal.

[0037] The coal quality component classification includes signal feature fitting and signal feature clustering, and the coal quality analysis includes coal quality calculation and coal quality classification;

[0038] The signal feature fitting is used to perform curve fitting on the signal features of the internal components of the coal through a high-dimensional space to extract a signal feature vector;

[0039] The signal feature clustering is used to perform regression analysis on the signal feature vector through the DBSCAN clustering algorithm to extract the internal element content of the coal;

[0040] The coal quality calculation is used to calculate the content of the coal quality components in the coal according to the content of the elements in the coal, and to calculate the coal quality mass fraction according to the content of the coal quality components in the coal;

[0041] The coal quality classification is used to determine the coal quality of the coal to be tested according to the coal quality fraction.

[0042] The DBSCAN clustering algorithm includes boundary division, parameter optimization, clustering division and error regression. The calculation formula of the coal quality score is:

[0043]

[0044] Among them, C q Indicates coal quality fraction, m c Indicates the moisture content inside coal, m s Indicates the standard content of moisture in coal, a c Indicates the ash content in coal, a s Indicates the standard content of ash in coal, v c Indicates the content of volatile matter in coal, v s Indicates the standard content of volatile matter in coal, ω m represents the influence factor of coal internal moisture content on coal quality, ω a represents the influence factor of coal internal ash content on coal quality, ω v It represents the influence factor of volatile matter content in coal on coal quality, and c1, c2 and c3 represent constants greater than 0.

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

[0046] 1. The present invention determines the laser focus position of coal through image processing technology, reduces human intervention, improves the automation level of the production line, and ensures that the laser is accurately focused on the target area of ​​each coal sample during detection, thereby improving the efficiency of coal quality detection;

[0047] 2. The present invention processes the spectral signal, calculates the components and content corresponding to the spectral signal through a cluster analysis algorithm, calculates the mass fraction of the coal according to the components and content, performs coal quality detection, and improves the accuracy of coal detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0049] Figure 1 This is a schematic diagram of the steps of an online coal quality detection method using adaptive focused laser induced breakdown spectroscopy according to Embodiment 1 of the present invention;

[0050] Figure 2 This is a flow chart of the steps of determining the laser focal point position in Example 1 of the present invention. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0052] Example 1

[0053] See also Figure 1 The present invention provides an embodiment: an online coal quality detection method using adaptive focused laser induced breakdown spectroscopy, comprising the following steps:

[0054] S1: Acquire a surface image of the coal to be detected by an image sensor, process the surface image of the coal to be detected, and acquire the position of the laser focus point.

[0055] S2: The laser is emitted by a laser, and the laser is reflected and transmitted by a reflector, and is focused and ablated at the position of the laser focal point to form plasma.

[0056] S3: The light receiving probe receives the light emitted by the plasma and transmits it to the spectrometer, and the spectrometer obtains the spectrum signal of the plasma through photoelectric conversion.

[0057] S4: Processing the spectrum signal of the plasma to extract the signal characteristics of the internal components of the coal to be detected, performing coal quality analysis on the coal to be detected according to the signal characteristics of the internal components of the coal, and determining the coal quality of the coal to be detected.

[0058] See also Figure 2 , a flowchart of the steps of determining the laser focus point position according to an embodiment of the present invention, wherein the specific steps of processing the surface image of the coal to be detected are as follows:

[0059] S1.1: Preprocess the surface image of the coal to be inspected to obtain a surface compressed and reconstructed image, wherein the preprocessing includes denoising, image sparse representation, image cutting and image sampling compression.

[0060] S1.2: The surface compression reconstructed image is converted into a surface row gradient map through a row gradient operator, the gradient response of each pixel in the surface row gradient map is checked in the gradient direction, a response threshold is set, the gradient responses in the surface row gradient map that are less than the response threshold are set to zero, and only the gradient responses in the surface row gradient map that are greater than or equal to the response threshold are retained.

[0061] S1.3: Adaptively binarize the surface line gradient map and perform edge detection, extract the continuous curve in the surface line gradient map, perform straight line fitting on the curve, obtain the straight line and significant vanishing line in the surface line gradient map, and use the intersection point of the significant vanishing line and the straight line as the key point of the surface line gradient map.

[0062] S1.4: Calculate the color features and texture features of each key point, and calculate the spectral intensity of the key point based on the color features and texture features, sort the spectral intensities of the key points from small to large, and take the key point with the smallest spectral intensity as the position of the laser focus point.

[0063] The calculation formula of the spectral intensity of the key point is:

[0064]

[0065] Among them, S i represents the spectral intensity of the key point, It represents the channel characteristics of the key point h channel after the third-level wavelet transform. Represents the channel characteristics of the key point s channel after the third-level wavelet transform, It represents the channel characteristics of the key point v channel after the third-level wavelet transform, E represents the grayscale distribution entropy of the key point, A represents the color characteristics of the key point, B represents the texture characteristics of the key point, and h, s and v represent different channels of the key point.

[0066] The calculation of the color features and texture features of each key point includes extracting color features and extracting texture features, the extracting color features includes color space quantization and color moment calculation, the extracting texture features includes calculating the grayscale co-occurrence matrix, and the calculating grayscale co-occurrence matrix includes energy features, contrast features, entropy features and correlation features.

[0067] The specific steps of S1.1 are as follows:

[0068] S1.1.1: De-noise the surface image of the coal to be inspected, wherein the noise includes on-site light noise, fly ash noise inside the coal, and image sensor photoelectric tube noise.

[0069] S1.1.2: The denoised surface image of the coal to be tested is sparsely represented, and mask sampling is performed through Fourier transform to obtain a sparse surface image.

[0070] S1.1.3: Cut the sparse surface image, cut the original image of size h*w into blocks according to the size of p*q, obtain the block image matrix, construct the observation matrix sequence, match different sampling rates according to the composition complexity of different areas of the block image, and perform sparse sampling compression in the Fourier domain in a distributed multi-point collaborative manner. The calculation formula for sparse sampling compression is:

[0071]

[0072] Where G represents the surface compression reconstructed image, a represents a single block image matrix, M represents the total number of block image matrices, and p a represents the sampling rate of the a-th block image matrix, θ represents the observation matrix sequence, represents the vector dot multiplication operation, K represents the division rate, F[·] represents the Fourier transform function, I represents the external image of the ultrasonic motor, and γ represents the noise. The sizes of p and q are determined by those skilled in the art according to actual conditions.

[0073] The specific steps of processing the spectrum signal of the plasma are as follows:

[0074] S4.1: Preprocessing the spectrum signal of the plasma, wherein the preprocessing includes spectrum signal normalization, multivariate scattering correction and spectrum signal smoothing.

[0075] S4.2: Convert the preprocessed spectral signal into a spectral frequency domain by Fourier transform, convert the spectral frequency domain into a spectral frequency distribution by spectrum analysis, and draw a spectral frequency distribution histogram according to the spectral frequency distribution.

[0076] S4.3: Extracting a characteristic spectrum curve according to the spectrum frequency distribution histogram, and determining the element corresponding to the characteristic spectrum curve through an atomic spectrum database.

[0077] S4.4: Establish a characteristic vector based on the characteristic spectral curve, perform Hilbert transform on the characteristic vector to calculate the envelope of the characteristic spectral curve, segment it according to the envelope, calculate the maximum value and mean value of each segment, and calculate the signal characteristics of the internal components of the coal to be detected based on the maximum value and mean value of each segment.

[0078] The calculation formula of the signal characteristics of the internal components of the coal to be detected is:

[0079]

[0080] Among them, I c represents the signal characteristics of coal internal components, i represents a single envelope, n represents the total number of envelopes of the characteristic spectrum curve, fi represents the frequency distribution interval of the i-th envelope, maxi Indicates the maximum value of the element signal of the i-th envelope, max avg Represents the average value of the maximum value of all envelope element signals, avg i represents the average value of the element signal of the ith envelope, η represents a constant greater than zero, δ(·) represents the Dirac function, t i represents the horizontal axis coordinate of the ith envelope, j represents the imaginary unit, and ω represents the center frequency.

[0081] The coal quality analysis of the coal to be tested according to the coal internal component signal characteristics includes coal quality component classification and coal quality analysis.

[0082] The coal components include ash, volatile matter and moisture.

[0083] The coal quality includes high-quality coal, medium-quality coal, low-quality coal and inferior coal.

[0084] The coal quality component classification includes signal feature fitting and signal feature clustering, and the coal quality analysis includes coal quality calculation and coal quality classification.

[0085] The signal feature fitting is used to perform curve fitting on the coal internal component signal features through high-dimensional space to extract signal feature vectors.

[0086] The signal feature clustering is used to perform regression analysis on the signal feature vector through the DBSCAN clustering algorithm to extract the internal element content of the coal.

[0087] The coal quality calculation is used to calculate the content of the coal quality components in the coal according to the content of the elements in the coal, and to calculate the coal quality mass fraction according to the content of the coal quality components in the coal.

[0088] The coal quality classification is used to determine the coal quality of the coal to be tested according to the coal quality fraction.

[0089] The DBSCAN clustering algorithm includes boundary division, parameter optimization, clustering division and error regression. The calculation formula of the coal quality score is:

[0090]

[0091] Among them, C q Indicates coal quality fraction, m c Indicates the moisture content inside coal, m s Indicates the standard content of moisture in coal, a c Indicates the ash content in coal, a s Indicates the standard content of ash in coal, v c Indicates the content of volatile matter in coal, vs Indicates the standard content of volatile matter in coal, ω m represents the influence factor of coal internal moisture content on coal quality, ω a represents the influence factor of coal internal ash content on coal quality, ω v It represents the influence factor of volatile matter content in coal on coal quality, and c1, c2 and c3 represent constants greater than 0.

[0092] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. An online coal quality detection method based on adaptive focused laser induced breakdown spectroscopy, characterized in that: The following steps are involved: S1: acquiring a surface image of the coal to be detected by an image sensor, processing the surface image of the coal to be detected, and acquiring the position of the laser focus point; S2: emitting laser light through a laser, the laser light is reflected and transmitted through a reflector, and focused at the position of the laser focus point to ablate the coal to be inspected, thereby forming plasma; S3: a light receiving probe receives light emitted by the plasma and transmits the light to a spectrometer, and the spectrometer obtains a spectrum signal of the plasma through photoelectric conversion; S4: Processing the spectrum signal of the plasma to extract the signal characteristics of the internal components of the coal to be detected, performing coal quality analysis on the coal to be detected according to the signal characteristics of the internal components of the coal, and determining the coal quality of the coal to be detected; The specific steps for processing the surface image of the coal to be inspected and obtaining the position of the laser focus point are as follows: S1.1: Preprocessing the surface image of the coal to be inspected to obtain a surface compressed reconstructed image, wherein the preprocessing includes denoising, image sparse representation, image cutting and image sampling compression; S1.2: converting the surface compression reconstructed image into a surface row gradient map through a row gradient operator, checking the gradient response of each pixel in the surface row gradient map in the gradient direction, setting a response threshold, setting the gradient response in the surface row gradient map that is less than the response threshold to zero, and only retaining the gradient response in the surface row gradient map that is greater than or equal to the response threshold; S1.3: Adaptively binarize the surface gradient map and perform edge detection, extract the continuous curve in the surface gradient map, perform straight line fitting on the curve, obtain the straight line and significant vanishing line in the surface gradient map, and use the intersection point of the significant vanishing line and the straight line as the key point of the surface gradient map; S1.4: Calculate the color features and texture features of each key point, and calculate the spectral intensity of the key point based on the color features and texture features, sort the spectral intensities of the key points from small to large, and take the key point with the smallest spectral intensity as the position of the laser focus point; The calculation formula of the spectral intensity of the key point is: Among them, S i represents the spectral intensity of the key point, It represents the channel characteristics of the key point h channel after the third-level wavelet transform. Represents the channel characteristics of the key point s channel after the third-level wavelet transform, It represents the channel characteristics of the key point v channel after the third-level wavelet transform, E represents the grayscale distribution entropy of the key point, A represents the color characteristics of the key point, B represents the texture characteristics of the key point, and h, s and v represent different channels of the key point.

2. According to claim 1, a method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy, characterized in that: The specific steps of S1.1 are as follows: S1.1.1: De-noising the surface image of the coal to be detected, wherein the noise includes on-site light noise, fly ash noise inside the coal, and photoelectric tube noise of the image sensor; S1.1.2: Sparsely represent the denoised surface image of the coal to be tested, and perform mask sampling through Fourier transform to obtain a sparse surface image; S1.1.3: Cut the sparse surface image, cut the original image of size h*w into blocks according to the size of p*q, obtain the block image matrix, construct the observation matrix sequence, match different sampling rates according to the composition complexity of different areas of the block image, and perform sparse sampling compression in the Fourier domain in a distributed multi-point collaborative manner. The calculation formula for sparse sampling compression is: Where G represents the surface compression reconstructed image, a represents a single block image matrix, M represents the total number of block image matrices, ρ a represents the sampling rate of the a-th block image matrix, θ represents the observation matrix sequence, represents the vector dot product operation, K represents the division rate, F[·] represents the Fourier transform function, I represents the external image of the ultrasonic motor, and γ represents the noise.

3. According to claim 1, a method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy, characterized in that: The specific steps of processing the spectral signal of the plasma and extracting the signal characteristics of the internal components of the coal to be detected are as follows: S4.1: preprocessing the spectrum signal of the plasma, wherein the preprocessing includes spectrum signal normalization, multivariate scattering correction and spectrum signal smoothing; S4.2: converting the preprocessed spectral signal into a spectral frequency domain by Fourier transform, converting the spectral frequency domain into a spectral frequency distribution by spectrum analysis, and drawing a spectral frequency distribution histogram according to the spectral frequency distribution; S4.3: extracting a characteristic spectrum curve according to the spectrum frequency distribution histogram, and determining the element corresponding to the characteristic spectrum curve through an atomic spectrum database; S4.4: Establish a characteristic vector based on the characteristic spectral curve, perform Hilbert transform on the characteristic vector to calculate the envelope of the characteristic spectral curve, segment it according to the envelope, calculate the maximum value and mean value of each segment, and calculate the signal characteristics of the internal components of the coal to be detected based on the maximum value and mean value of each segment.

4. The method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy according to claim 3 is characterized in that: The calculation formula of the signal characteristics of the internal components of the coal to be detected is: Among them, I c represents the signal characteristics of coal internal components, i represents a single envelope, n represents the total number of envelopes of the characteristic spectrum curve, and f i represents the frequency distribution interval of the i-th envelope, max i Indicates the maximum value of the element signal of the i-th envelope, max avg Represents the average value of the maximum value of all envelope element signals, avg i represents the average value of the element signal of the ith envelope, η represents a constant greater than zero, δ(·) represents the Dirac function, t i represents the horizontal axis coordinate of the ith envelope, j represents the imaginary unit, and ω represents the center frequency.

5. The method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy according to claim 4 is characterized in that: The coal quality analysis of the coal to be tested according to the coal internal component signal characteristics includes coal quality component classification and coal quality analysis; The coal quality components include ash, volatile matter and moisture; The coal quality includes high-quality coal, medium-quality coal, low-quality coal and inferior coal.

6. The method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy according to claim 5 is characterized in that: The coal quality component classification includes signal feature fitting and signal feature clustering, and the coal quality analysis includes coal quality calculation and coal quality classification; The signal feature fitting is used to perform curve fitting on the signal features of the internal components of the coal through a high-dimensional space to extract a signal feature vector; The signal feature clustering is used to perform regression analysis on the signal feature vector through the DBSCAN clustering algorithm to extract the internal element content of the coal; The coal quality calculation is used to calculate the content of the coal quality components in the coal according to the content of the elements in the coal, and to calculate the coal quality mass fraction according to the content of the coal quality components in the coal; The coal quality classification is used to determine the coal quality of the coal to be tested according to the coal quality fraction.

7. The method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy according to claim 6 is characterized in that: The DBSCAN clustering algorithm includes boundary partitioning, parameter optimization, clustering partitioning and error regression.

8. The method for online detection of coal quality by adaptive focused laser induced breakdown spectroscopy according to claim 7 is characterized in that: The calculation formula of the coal quality fraction is: Among them, C q Indicates coal quality fraction, m c Indicates the moisture content inside coal, m s Indicates the standard content of moisture in coal, a c Indicates the ash content in coal, a s Indicates the standard content of ash in coal, v c Indicates the content of volatile matter in coal, v s Indicates the standard content of volatile matter in coal, ω m represents the influence factor of coal internal moisture content on coal quality, ω a represents the influence factor of coal internal ash content on coal quality, ω v It represents the influence factor of volatile matter content in coal on coal quality, and c1, c2 and c3 represent constants greater than 0.

Images