A coal quality detection method, system, device and medium

Abstract

The application relates to a coal quality detection method, system, device and medium, wherein the method comprises the following steps: acquiring color texture feature parameters according to a standard coal image obtained in advance, and acquiring Raman spectrum feature parameters according to a standard coal Raman spectrum obtained in advance. A prediction model of ash content in coal is constructed by increasing the weight of the color texture feature parameters based on the Raman spectrum feature parameters. A prediction model of moisture content and volatile content in coal is constructed by increasing the weight of the Raman spectrum feature parameters based on the color texture feature parameters. The prediction model of ash content in coal and the prediction model of moisture content and volatile content in coal are coupled to obtain a target coal quality prediction model. Color texture feature parameters and Raman spectrum feature parameters of a coal sample to be detected are acquired, and the component content of the coal sample to be detected is acquired based on the target coal quality prediction model. The problem of low accuracy in coal quality detection in the related art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal quality detection, in particular to a coal quality detection method, system, device and medium. BACKGROUND

[0002] Coal, as an important industrial raw material, occupies an irreplaceable position in the current fossil energy system. Especially in the field of thermal power, the demand for coal is huge. However, the supply of coal sources is not stable, which means that power plants may use dozens or even hundreds of different coals at the same time. At the same time, real-time monitoring of coal composition (moisture, volatile matter, fixed carbon and ash content, etc.) is beneficial to the efficient thermal conversion (combustion, liquefaction, vaporization, etc.) of coal. Therefore, rapid detection of coal quality is of great significance.

[0003] Although the traditional industrial analysis method is accurate, its cumbersome process and long cycle are difficult to meet the demand of rapid detection. In recent years, coal quality detection methods based on spectral technology have made significant progress, providing new possibilities for rapid analysis. However, such methods mainly focus on the characterization of specific chemical structures of coal, and face bottlenecks in improving detection accuracy, and are easily affected by environmental factors on site, affecting the stability and reliability of detection. Therefore, the related art has the problem of low accuracy of coal quality detection. SUMMARY

[0004] The embodiments of the present application provide a coal quality detection method, system, device and medium to at least solve the problem of low accuracy of coal quality detection in the related art.

[0005] In a first aspect, the embodiments of the present application provide a coal quality detection method, which comprises:

[0006] obtaining color texture feature parameters according to a standard coal image obtained in advance, and obtaining Raman spectrum feature parameters according to a standard coal Raman spectrum obtained in advance;

[0007] based on the Raman spectrum feature parameters, constructing a prediction model of ash content in coal by increasing the weight of the color texture feature parameters;

[0008] based on the color texture feature parameters, constructing a prediction model of moisture and volatile matter content in coal by increasing the weight of the Raman spectrum feature parameters;

[0009] coupling the prediction model of ash content in coal and the prediction model of moisture and volatile matter content in coal to obtain a target coal quality prediction model;

[0010] obtaining color texture feature parameters and Raman spectrum feature parameters of a coal sample to be detected, and obtaining the component content of the coal sample to be detected based on the target coal quality prediction model.

[0011] In an embodiment, before the color texture feature parameters are acquired according to the pre-acquired standard coal images and the Raman spectrum feature parameters are acquired according to the pre-acquired standard coal Raman spectrum images, the method further comprises:

[0012] The standard coals of different coal components are subjected to image acquisition and Raman spectrum testing to acquire images and Raman spectrum images of the standard coals; wherein,

[0013] The coal components include moisture, volatile matter, fixed carbon and ash, and the same reference is adopted for the standard coal samples of the different coal components, and the reference is air-dry basis.

[0014] In an embodiment, before the standard coals of different coal components are subjected to image acquisition and Raman spectrum testing, the method further comprises:

[0015] The Raman spectrum of the standard coal is pre-processed, and the pre-processing includes segmenting the Raman spectrum into first-order Raman spectrum and second-order Raman spectrum;

[0016] The wavelength range of the first-order Raman spectrum is selected to be 400-2400 cm -1 , the wavelength range of the second-order Raman spectrum is selected to be 2400-3800 cm -1 , and the first-order Raman spectrum and the second-order Raman spectrum are subjected to baseline removal processing.

[0017] In an embodiment, the color texture feature parameters include first-order moment of gray, second-order moment of gray, third-order moment of gray, color standard deviation, dominant hue, energy, contrast, correlation, entropy and uniformity;

[0018] The Raman spectrum feature parameters include peak area, total peak area of the first-order Raman spectrum and the second-order Raman spectrum, and combinations of the peak area and the total peak area of the first-order Raman spectrum and the total peak area of the second-order Raman spectrum.

[0019] In an embodiment, before the prediction model of ash content in coal is constructed based on the Raman spectrum feature parameters by increasing the weight of the color texture feature parameters, the method further comprises:

[0020] The color texture feature parameters and the Raman spectrum feature parameters are scaled by Z-score standardization, so that the scales of the color texture feature parameters and the Raman spectrum feature parameters are within a preset range.

[0021] In an embodiment, after the color texture feature parameters and the Raman spectrum feature parameters are scaled by Z-score standardization, the method further comprises:

[0022] The number of the color texture feature parameters and the Raman spectrum feature parameters is selected using neighborhood component analysis, wherein a plurality of parameters with stronger correlation are selected as the color texture feature parameters, and a plurality of parameters with stronger correlation are selected as the Raman spectrum feature parameters.

[0023] In an embodiment, the component content of the coal sample to be detected includes fixed carbon content, ash content, moisture content, and volatile content, wherein the fixed carbon content is determined according to the ash content, the moisture content, and the volatile content.

[0024] In a second aspect, the embodiments of the present application provide a coal quality detection system, which includes an acquisition feature parameter module, a prediction model construction module of ash content in coal, a prediction model construction module of moisture and volatile content in coal, a target coal quality prediction model, and an acquisition coal quality detection result module, wherein:

[0025] The acquisition feature parameter module is configured to acquire color texture feature parameters according to a standard coal image acquired in advance, and acquire Raman spectrum feature parameters according to a standard coal Raman spectrum image acquired in advance.

[0026] The prediction model construction module of ash content in coal is configured to construct a prediction model of ash content in coal based on the Raman spectrum feature parameters by increasing the weight of the color texture feature parameters.

[0027] The prediction model construction module of moisture and volatile content in coal is configured to construct a prediction model of moisture and volatile content in coal based on the color texture feature parameters by increasing the weight of the Raman spectrum feature parameters.

[0028] The target coal quality prediction model is configured to couple the prediction model of ash content in coal and the prediction model of moisture and volatile content in coal to acquire a target coal quality prediction model.

[0029] The acquisition coal quality detection result module is configured to acquire color texture feature parameters and Raman spectrum feature parameters of a coal sample to be detected, and acquire component content of the coal sample to be detected based on the target coal quality prediction model.

[0030] In a third aspect, the embodiments of the present application provide a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements a coal quality detection method according to the first aspect.

[0031] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, which stores a computer program executable by a processor to implement a coal quality detection method according to the first aspect.

[0032] The coal quality detection method, system, device and medium provided by the embodiment of the application have at least the following technical effects.

[0033] The color texture feature parameters are obtained according to the standard coal image obtained in advance, and the Raman spectrum feature parameters are obtained according to the standard coal Raman spectrum obtained in advance. Based on the Raman spectrum feature parameters, the weight of the color texture feature parameters is increased to construct a prediction model of the ash content in the coal. Based on the color texture feature parameters, the weight of the Raman spectrum feature parameters is increased to construct a prediction model of the moisture and volatile content in the coal. The prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal are coupled to obtain a target coal quality prediction model. The color texture feature parameters and the Raman spectrum feature parameters of the coal sample to be detected are obtained, and the component content of the coal sample to be detected is obtained based on the target coal quality prediction model. The application realizes accurate prediction of the moisture, volatile content, ash content and fixed carbon in the coal by coupling image recognition and Raman spectrum detection and by dynamic adjustment of feature weights. The application has the characteristics of fast detection speed, high precision and good stability, and solves the problem of low accuracy in coal quality detection in the related art.

[0034] Details of one or more embodiments of the application are presented in the following drawings and description to make other features, objects and advantages of the application more apparent. BRIEF DESCRIPTION OF DRAWINGS

[0035] The drawings described herein are intended to provide further understanding of the application, form a part of the application, and the illustrative embodiments of the application and their descriptions are used to explain the application, and do not constitute improper limitations on the application. In the drawings:

[0036] Figure 1 is a flowchart of a coal quality detection method according to an embodiment of the application;

[0037] Figure 2 is a schematic diagram of standard coal according to an exemplary embodiment;

[0038] Figure 3 is a schematic diagram of a Raman spectrum according to an exemplary embodiment;

[0039] Figure 4 is a schematic diagram of baseline correction of a first-order Raman spectrum and a second-order Raman spectrum according to an exemplary embodiment;

[0040] Figure 5 is a block diagram of a coal quality detection system according to an exemplary embodiment;

[0041] Figure 6 is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0042] In order to make the purposes, technical solutions, and advantages of the present application clearer, the present application is described and explained below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. Based on the embodiments provided by the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort belong to the scope of the present application.

[0043] It is obvious that the accompanying drawings in the following description are only some examples or embodiments of the present application, and for those of ordinary skill in the art, the present application can be applied to other similar scenarios without creative effort based on the accompanying drawings. In addition, it can be understood that although the efforts made in the development process can be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacture or production changes based on the technical content disclosed in the present application are only routine technical means and should not be understood as insufficient disclosure of the content disclosed in the present application.

[0044] In the present application, "embodiments" means that the specific features, structures or properties described in conjunction with the embodiments can be included in at least one embodiment of the present application. The phrase appears at various places in the specification does not necessarily refer to the same embodiment, nor is it mutually exclusive or alternative embodiments to other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in the present application can be combined with other embodiments without conflict.

[0045] Unless otherwise defined, technical terms and scientific terms used in the present application shall have the meanings as commonly understood by one of ordinary skill in the art to which this application pertains. The terms "a", "an", "one", "this", and similar referents in the context of describing the application are to be construed to be open-ended, referring to one or more than one, unless otherwise noted. The terms "including", "containing", "having", and "comprising" and variations thereof in the context of this application are to be construed in an open-ended fashion, that is as "including, at least, the recited steps, or components, and those additional steps, or components, that can not be expressly listed." The term "coupled" or "connected" or "coupling" or "connecting" in the context of this application are to be construed in an open-ended fashion, that is as "connected, at least, or coupled, at least, or coupling, at least, or connecting, at least, or in some way." The term "plurality" refers to two or more. The term "and / or" describes association between or among multiple options, that is, A and / or B can mean A or B or both A and B. The term "and / or" is used in the context of the present application to mean "and" or "or", that is, "A and / or B" means "A and B" or "A or B". The term "first", "second", "third", etc. are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence.

[0046] In this document, the term "techniques" can be used to refer to a technique or a set of techniques, or other summary technical terms, for example, the term can include:

[0047] Raman spectra is a kind of scattering spectrum, which is used to characterize the structural characteristics of coal, and then to quickly identify the type of coal.

[0048] Z-score standardization normalizes data by subtracting the mean and dividing by the standard deviation, so that each feature has a zero mean and unit variance.

[0049] Neighborhood Components Analysis (NCA) is a supervised learning method, mainly used for feature selection.

[0050] In a first aspect, the embodiments of the present application provide a coal quality detection method, applied to the end of the stator winding of a steam turbine generator, Figure 1 is a flow chart of coal quality detection, as shown in Figure 1 A coal quality detection method includes:

[0051] In step S101, color texture feature parameters are obtained according to the standard coal image obtained in advance, and Raman spectrum feature parameters are obtained according to the standard coal Raman spectrum image obtained in advance.

[0052] Step S102, based on the Raman spectral characteristic parameters, a prediction model of the ash content in the coal is constructed by increasing the weight of the color texture characteristic parameters.

[0053] Step S103, based on the color texture characteristic parameters, a prediction model of the moisture and volatile content in the coal is constructed by increasing the weight of the Raman spectral characteristic parameters.

[0054] Step S104, the prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal are coupled to obtain a target coal quality prediction model.

[0055] Step S105, the color texture characteristic parameters and the Raman spectral characteristic parameters of the coal sample to be detected are obtained, and the component content of the coal sample to be detected is obtained based on the target coal quality prediction model.

[0056] In summary, the embodiment of the present application provides a coal quality detection method. The color texture characteristic parameters are obtained according to the pre-obtained standard coal image, and the Raman spectral characteristic parameters are obtained according to the pre-obtained standard coal Raman spectrum. Based on the Raman spectral characteristic parameters, a prediction model of the ash content in the coal is constructed by increasing the weight of the color texture characteristic parameters. Based on the color texture characteristic parameters, a prediction model of the moisture and volatile content in the coal is constructed by increasing the weight of the Raman spectral characteristic parameters. The prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal are coupled to obtain a target coal quality prediction model. The color texture characteristic parameters and the Raman spectral characteristic parameters of the coal sample to be detected are obtained, and the component content of the coal sample to be detected is obtained based on the target coal quality prediction model. The present application couples image recognition and Raman spectrum detection, dynamically adjusts the feature weight, and realizes accurate prediction of the moisture, volatile content, ash content and fixed carbon in the coal. The present application has the characteristics of fast detection speed, high precision and good stability, and solves the problem of low accuracy in related art.

[0057] In an embodiment, before step S101, the method further comprises:

[0058] image acquisition and Raman spectrum testing are performed on standard coals of multiple different coal quality components to obtain images and Raman spectrum graphs of the standard coals; wherein,

[0059] The coal quality components include moisture, volatile content, fixed carbon and ash content, and the same standard is used for standard coal samples of different coal quality components, and the standard is air-dry basis.

[0060] Optionally, standard coals of different coal quality components are selected, and the powders of the standard coals are subjected to image acquisition and Raman testing to obtain images and Raman spectra of the standard coals. The selected standard coals are all raw coal, and the coal quality components are obtained by industrial analysis according to GB / T 212-2008, including moisture, volatile matter, fixed carbon and ash, and the same reference is used for different standard coals, which is all air-dry basis.

[0061] Among them, the powders of the standard coals are subjected to image acquisition and Raman testing to obtain images and Raman spectra of the standard coals. It can be: using a high-resolution camera or other imaging equipment to take photos or videos of each standard coal sample, obtaining the image of the standard coal, considering factors such as light source, background and shooting angle when shooting to ensure image quality. Use a Raman spectrometer to measure each coal sample to obtain the Raman spectrum of the standard coal, and control the laser power, exposure time and other conditions during the experiment to obtain high-quality spectral data. The image-Raman database of the standard coal can be established, and the image and Raman spectrum of the standard coal are entered into the database for subsequent query and data analysis.

[0062] By obtaining the image of the standard coal, the color, texture and other appearance characteristics of the coal can be intuitively understood. By obtaining the Raman spectrum of the standard coal, information about the chemical composition and structure of the coal is provided. The combination of the two forms a comprehensive data set that contains both physical and chemical characteristics.

[0063] In an embodiment, before the image acquisition and Raman spectrum testing of the standard coals of different coal quality components, the method further comprises:

[0064] The Raman spectrum of the standard coal is preprocessed, and the preprocessing includes segmenting the Raman spectrum into first-order Raman spectrum and second-order Raman spectrum;

[0065] The wavelength range of the selected first-order Raman spectrum is 400-2400 cm -1 , and the wavelength range of the selected second-order Raman spectrum is 2400-3400 cm -1 , and the first-order Raman spectrum and the second-order Raman spectrum are subjected to baseline removal processing.

[0066] Optionally, the Raman spectrum of the standard coal is preprocessed, and the preprocessing includes segmenting the Raman spectrum into first-order Raman spectrum and second-order Raman spectrum, the wavelength range of the selected first-order Raman spectrum is 400-2400 cm -1 , and the wavelength range of the selected second-order Raman spectrum is 2400-3400 cm -1 , and the wavelength range of the first-order Raman spectrum in this example can be 800-2200 cm -1 , and the wavelength range of the selected second-order Raman spectrum can be 2200-3200 cm-1 and baseline processing of the first-order Raman spectrum and the second-order Raman spectrum.

[0067] By segmenting the Raman spectrum, different types of vibration modes can be better identified. The first-order Raman spectrum mainly contains the basic vibration modes of the molecule, while the second-order Raman spectrum may include combination frequency and overtone information. There are usually fluorescent background or other non-specific signals in the Raman spectrum, which will affect the identification of characteristic peaks. By baseline processing (such as using polynomial fitting), these background noises can be removed, thereby enhancing the true Raman signal, making the data clearer, and improving the quality of the sample data.

[0068] In step S101, color texture feature parameters are obtained according to the pre-acquired standard coal image, and Raman spectrum feature parameters are obtained according to the pre-acquired standard coal Raman spectrum image.

[0069] Optionally, the color texture feature parameters can be obtained from the standard coal image by color space conversion (such as RGB to HSV), such as gray image, average color value, color saturation, color standard deviation, etc. Texture information such as energy and contrast can be extracted using methods such as Local Binary Pattern (LBP). Specifically, the color texture feature parameters can be gray first moment, gray second moment, gray third moment, color standard deviation, dominant hue, energy, contrast, correlation, entropy, and uniformity. These features can reflect the fine structure and color distribution of the coal surface, and help to identify different types of coal.

[0070] The Raman spectrum feature parameters obtained from the standard coal Raman spectrum image can be peak area, total peak area of the first-order Raman spectrum, and total peak area of the second-order Raman spectrum. Specifically, the Raman spectrum feature parameters include peak area I D , I V , I G , I 2D , I 2G , I (D+G) , I 2S , total peak area of the first-order Raman spectrum S1 and total peak area of the second-order Raman spectrum S2, and their corresponding combinations I D / S1, I V / S1, I G / S1, I 2D / S2, I 2G / S2, I (D+G) / S2, S1 / S2, I V / I 2D , I 2D / I 2S , I (D+G) / I 2S , I (D+G) / I 2S etc.

[0071] By obtaining color texture feature parameters, the physical appearance of coal such as glossiness, particle size, and impurity content can be directly reflected, which helps to improve the detection accuracy. By obtaining Raman spectrum feature parameters, detailed information about the chemical composition of coal can be provided, which can help to determine the types and structures of organic matter in coal.

[0072] In an embodiment, before step S102, based on the Raman spectrum feature parameters, a prediction model of the ash content in coal is constructed by increasing the weight of the color texture feature parameters, the method further comprises:

[0073] The color texture feature parameters and the Raman spectrum feature parameters are scaled using Z-score standardization, so that the scales of the color texture feature parameters and the Raman spectrum feature parameters are in a preset range.

[0074] The number of color texture feature parameters and Raman spectrum feature parameters is selected using neighborhood component analysis, wherein the first several parameters with stronger correlation are selected for color texture feature parameters, and the first several parameters with stronger correlation are selected for Raman feature parameters.

[0075] Optionally, the color texture feature parameters and the Raman spectrum feature parameters are scaled using Z-score standardization to ensure that the scales of all feature parameters are in a preset range. In this example, the preset range can be a range of 0-1. Different features may have different dimensions and scales, and direct use may cause some features to have too large or too small influence on the model. Z-score standardization eliminates such influence, so that all features are compared on the same scale, and the standardized data is more stable, which helps to improve the robustness and generalization ability of the model. Neighborhood component analysis (NCA) is used for feature selection to reduce the number of features and improve the performance of the model. In this embodiment, the first 8 parameters with stronger correlation can be selected for color texture feature parameters, and the first 12 parameters with stronger correlation can be selected for Raman feature parameters.

[0076] Standardizing the feature parameters and selecting the features can improve the prediction accuracy of the model, and reducing the number of features can significantly reduce the time and computational resource requirements for model training. Standardization ensures the consistency of the data and reduces the bias caused by the difference in feature scales.

[0077] In an embodiment, in step S102, based on the Raman spectrum feature parameters, a prediction model of the ash content in coal is constructed by increasing the weight of the color texture feature parameters. Specifically, it includes:

[0078] After the neighborhood component analysis (NCA) is used for feature selection, a weight factor is set for the color texture feature parameter, and the weight factor is multiplied to increase the weight of the color texture feature parameter. A support vector machine (SVM) is used as a prediction model. The SVM is a powerful classification and regression algorithm and is suitable for high-dimensional data. A prediction model of the ash content in the coal is constructed. The prediction model of the ash content in the coal takes the color texture feature parameter and the Raman spectrum feature parameter as input and outputs the ash content.

[0079] By increasing the weight of the color texture feature parameter, the image information can be better utilized, and thus the prediction accuracy of the model is improved.

[0080] In an embodiment, step S103, based on the color texture feature parameter, a prediction model of the moisture and volatile content in the coal is constructed by increasing the weight of the Raman spectrum feature parameter. Specifically, the step includes:

[0081] After the neighborhood component analysis (NCA) is used for feature selection, a weight factor is set for the Raman spectrum feature parameter, and the weight factor is multiplied to increase the weight of the Raman spectrum feature parameter. The prediction model of the moisture and volatile content in the coal is constructed. The prediction model of the moisture and volatile content in the coal takes the Raman spectrum feature parameter and the color texture feature parameter as input and outputs the moisture content and the volatile content.

[0082] By increasing the weight of the Raman spectrum feature parameter, the chemical composition information can be better utilized, and thus the prediction accuracy of the model is improved.

[0083] It is worth noting that in step S102 and step S103, different weights are used for the prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal, respectively. This is because there is a difference in the relevance of different weight parameters to the prediction object. For example, in the prediction model of the ash content in the coal, the color texture parameter has a strong direct relevance (linear correlation) to the ash content, and the Raman parameter is relatively weak. However, only the color texture parameter cannot be used to directly establish an accurate prediction model. That is, the ash content is associated with both the color texture parameter and the Raman parameter. When the prediction model is constructed, the color texture parameter is used as the main parameter, and the Raman parameter is used as the auxiliary parameter. Therefore, the weight of the color texture parameter is increased. For the prediction model of the moisture and volatile content in the coal, the weight of the Raman parameter is increased. In an embodiment, step S104, the prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal are coupled to obtain a target coal quality prediction model. Specifically, the step includes:

[0084] The prediction model of the ash content in the coal and the prediction model of the moisture content and the volatile content in the coal are combined to obtain a target coal quality prediction model. By coupling the prediction model of the ash content in the coal and the prediction model of the moisture content and the volatile content in the coal, a comprehensive target coal quality prediction model is constructed, and rapid detection of the coal quality is realized, and the overall prediction accuracy is improved.

[0085] In an embodiment, step S105, the color texture feature parameters and the Raman spectrum feature parameters of the coal sample to be detected are obtained, and the ingredient content of the coal sample to be detected is obtained based on the target coal quality prediction model. Specifically, it includes:

[0086] The same image acquisition and Raman test as step S101 are performed on the coal sample to be detected, and the corresponding color texture and Raman spectrum feature parameters are extracted, and then the Z-score standardization and neighborhood component analysis are used to process the color texture and Raman spectrum feature parameters to obtain the color texture feature parameters and the Raman spectrum feature parameters of the target coal sample to be detected. At this time, the color texture feature parameters and the Raman spectrum feature parameters of the target coal sample to be detected correspond one-to-one with the parameters used by the target coal quality prediction model, and no additional parameter screening is required using the NCA algorithm. The color texture feature parameters and the Raman spectrum feature parameters of the target coal sample to be detected are input into the target coal quality prediction model to obtain the ingredient content of the coal sample to be detected, which includes the fixed carbon content, the ash content, the moisture content and the volatile content. The fixed carbon content is determined according to the ash content, the moisture content and the volatile content. Specifically, the calculation method of the fixed carbon content of the coal sample to be detected is: FC ad = 100 - A ad - M ad - V ad , FC ad represents fixed carbon, A ad represents ash, M ad represents moisture, and V ad represents volatile matter.

[0087] Step S105 is based on image recognition and Raman spectrum, does not need to prepare a sample, has wide applicability, and is non-destructive, non-toxic and non-contact to the sample itself, and can be applied to complex production environments. The color texture feature parameters and the Raman spectrum feature parameters based on the image are coupled to make up for the shortcomings of the two in detecting the coal quality ingredients alone, improve the detection accuracy and the detection range. In addition, the detection process is simple and fast, and real-time online detection of the coal quality can be realized to meet the needs of industrial production for quickly obtaining coal quality information.

[0088] The following is an application example based on the first aspect.

[0089] First, 26 different coal rank and ash content of standard coal were taken, respectively ground to 74-105 μm, and numbered 1-30, respectively, and the proximate analysis was carried out, and the results were adopted on an air-dried basis. Here, the composition parameters of 6 typical standard coals are listed as shown in Table 1.

[0090] Table 1

[0091]

[0092] Then, the 30 standard coals were respectively subjected to image acquisition and Raman detection, and the standard coal images and Raman spectra obtained, Figure 2 is a schematic diagram of standard coal according to an example embodiment, Figure 3 is a schematic diagram of Raman spectrum according to an example embodiment, as Figure 2 and Figure 3 shown, it can be found that the Raman spectrum has baseline drift. In order to accurately obtain the Raman spectrum characteristic parameters of the coal sample, the first-order Raman peak and the second-order Raman peak should be respectively subjected to baseline correction processing. The method of baseline correction is two-point method, Figure 4 is a schematic diagram of baseline correction of first-order Raman spectrum and second-order Raman spectrum according to an example embodiment, as Figure 4 shown, the left is a first-order Raman spectrum baseline correction schematic diagram, wherein the upper left curve diagram is the original data and baseline, and the lower left curve diagram is the corrected data, and the right is a second-order Raman spectrum baseline correction schematic diagram, wherein the upper right curve diagram is the original data and baseline, and the lower right curve diagram is the corrected data.

[0093] In order to extract as many and accurate color texture feature parameters of the image as possible to participate in the Raman spectrum characteristic parameters, the color texture feature parameters extracted herein include gray first moment, gray second moment, gray third moment, color standard deviation, energy, contrast, correlation, entropy, and uniformity; the Raman spectrum characteristic parameters include peak area I D , I V , I G , I 2D , I 2G , I (D+G) , I 2S , first-order Raman spectrum total peak area and second-order Raman spectrum total peak area S1, S2, and corresponding combinations I D / S1, I V / S1, I G / S1, I 2D / S2, I 2G / S2, I (D+G) / S2, S1 / S2, I V / I 2D , I 2D / I 2S , I (D+G) / I 2S 、I (D+G) / I 2S and so on, the Raman peak area parameters are obtained by peak fitting, and the characteristic parameters of one of the standard coals are taken as an example, Table 2 is the color texture characteristic parameters of the standard coal, and Table 3 is the color texture characteristic parameters of the standard coal, as shown in Table 2 and Table 3, thereby constructing the image-Raman database.

[0094] Table 2

[0095]

[0096] Table 3

[0097]

[0098] According to the Z-score standardization and neighborhood component analysis, the characteristic parameters are further processed to predict the ash content in the coal. After normalizing all the characteristic parameters, a weight factor is set, which can be 0.5-3, and in the example, 2 can be used. The weight of the color texture parameter is increased by multiplying the weight factor, and the Raman spectrum characteristic parameter is coupled without weight. Then, the NCA algorithm is used to screen the parameters in the color texture and Raman spectrum that are more strongly related to the ash content. Among them, the first 8 color texture characteristic parameters with stronger correlation are selected, and the first 12 Raman characteristic parameters with stronger correlation are selected. The processed and screened characteristic parameters and the corresponding ash content are input into the SVM model to construct the prediction model of the ash content of the coal.

[0099] The same steps are used to construct the prediction models of the moisture and volatile matter in the coal. Since it is found that the Raman parameters have stronger correlation with the moisture and volatile matter, the weight of the Raman spectrum characteristic parameter is increased for modeling. Finally, all the prediction models are combined.

[0100] The image acquisition and Raman test are performed on the coal sample to be detected in the same way as step S101, and the corresponding color texture and Raman spectrum characteristic parameters are extracted. Then, the Z-score standardization and neighborhood component analysis are used to process the color texture and Raman spectrum characteristic parameters to obtain the color texture characteristic parameters and Raman spectrum characteristic parameters of the target coal sample to be detected.

[0101] The processed color texture characteristic parameters and Raman spectrum characteristic parameters of the target coal sample to be detected are input into the prediction model to obtain the contents of the ash, moisture and volatile matter in the coal. The content of the fixed carbon in the coal sample to be detected is calculated as follows: FC ad = 100 - A ad - M ad - V ad , FC ad represents the fixed carbon, A ad represents the ash content, and Mad moisture, V ad moisture, V

[0102] In summary, the embodiment of the present application provides a coal quality detection method. Color texture feature parameters are obtained according to pre-acquired standard coal images, and Raman spectrum feature parameters are obtained according to pre-acquired standard coal Raman spectrum images. Based on the Raman spectrum feature parameters, a prediction model of the ash content in coal is constructed by increasing the weight of the color texture feature parameters. Based on the color texture feature parameters, a prediction model of the moisture and volatile content in coal is constructed by increasing the weight of the Raman spectrum feature parameters. The prediction model of the ash content in coal and the prediction model of the moisture and volatile content in coal are coupled to obtain a target coal quality prediction model. The color texture feature parameters and the Raman spectrum feature parameters of a coal sample to be detected are obtained, and the component content of the coal sample to be detected is obtained based on the target coal quality prediction model. The present application realizes accurate prediction of the moisture, volatile, ash and fixed carbon in coal by coupling image recognition and Raman spectrum detection and by dynamic adjustment of feature weights. The present application has the characteristics of fast detection speed, high precision and good stability, and solves the problem of low precision in coal quality detection in related technologies.

[0103] In a second aspect, the embodiment of the present application provides a coal quality detection system. Figure 5 is a block diagram of a system for grading relay protection device defects according to an exemplary embodiment. As shown in Figure 5 the system includes a feature parameter acquisition module 510, a prediction model of the ash content in coal construction module 520, a prediction model of the moisture and volatile content in coal construction module 530, a target coal quality prediction model 540 and a coal quality detection result acquisition module 550, wherein:

[0104] The feature parameter acquisition module 510 is configured to obtain color texture feature parameters according to pre-acquired standard coal images, and obtain Raman spectrum feature parameters according to pre-acquired standard coal Raman spectrum images;

[0105] The prediction model of the ash content in coal construction module 520 is configured to construct a prediction model of the ash content in coal based on the Raman spectrum feature parameters by increasing the weight of the color texture feature parameters;

[0106] The prediction model of the moisture and volatile content in coal construction module 530 is configured to construct a prediction model of the moisture and volatile content in coal based on the color texture feature parameters by increasing the weight of the Raman spectrum feature parameters;

[0107] The target coal quality prediction model 540 is configured to couple the prediction model of the ash content in coal and the prediction model of the moisture and volatile content in coal to obtain a target coal quality prediction model;

[0108] The coal quality detection result acquisition module 550 is configured to acquire the color texture feature parameter and the Raman spectrum feature parameter of the coal sample to be detected, and acquire the component content of the coal sample to be detected based on the target coal quality prediction model.

[0109] To sum up, the coal quality detection system provided in the embodiments of the present application solves the problem of low accuracy in coal quality detection in the related art by the feature parameter acquisition module 510, the prediction model of the ash content in coal construction module 520, the prediction model of the moisture and volatile content in coal construction module 530, the target coal quality prediction model, and the coal quality detection result acquisition module 540. Specifically, the color texture feature parameter is acquired according to the pre-acquired standard coal image, and the Raman spectrum feature parameter is acquired according to the pre-acquired standard coal Raman spectrum image. The prediction model of the ash content in coal is constructed by increasing the weight of the color texture feature parameter based on the Raman spectrum feature parameter. The prediction model of the moisture and volatile content in coal is constructed by increasing the weight of the Raman spectrum feature parameter based on the color texture feature parameter. The prediction model of the ash content in coal and the prediction model of the moisture and volatile content in coal are coupled to obtain the target coal quality prediction model. The color texture feature parameter and the Raman spectrum feature parameter of the coal sample to be detected are acquired, and the component content of the coal sample to be detected is acquired based on the target coal quality prediction model. The present application realizes accurate prediction of the moisture, volatile content, ash content, and fixed carbon in coal by coupling image recognition and Raman spectrum detection and by dynamic adjustment of the feature weight. The present application has the characteristics of fast detection speed, high accuracy, and good stability, and solves the problem of low accuracy in coal quality detection in the related art.

[0110] It should be noted that the system for coal quality detection provided in the embodiments is used to implement the above-described embodiments, and details are not repeated. As used above, the terms "module", "unit", "sub-unit", and the like can be a combination of software and / or hardware that realizes a predetermined function. Although the devices described in the above embodiments are preferably implemented in software, hardware or a combination of software and hardware can also be implemented.

[0111] In a third aspect, the embodiments of the present application provide an electronic device, Figure 6 is a block diagram of an electronic device according to an exemplary embodiment. As Figure 6 indicated, the electronic device can include a processor 61 and a memory 62 storing computer program instructions.

[0112] Specifically, the processor 61 described above can include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or can be configured as one or more integrated circuits that implement the embodiments of the present application.

[0113] The memory 62 can include a mass storage for data or instructions. By way of example, and without limitation, the memory 62 can include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), a flash drive, a compact disc (CD) or DVD, a tape, a magnetic or optical or semiconductor memory, or any combination of two or more of these. The memory 62 can be removable and / or non-removable (or fixed) as appropriate. The memory 62 can be internal or external as appropriate. In certain embodiments, the memory 62 is a non-volatile memory. In certain embodiments, the memory 62 includes read-only memory (ROM) and random access memory (RAM). The ROM can be mask programmed ROM, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), electrically alterable ROM (EAROM), or FLASH, or a combination of two or more of these, as appropriate. The RAM can be static random access memory (SRAM) or dynamic random access memory (DRAM), which can be Fast Page Mode Dynamic Random Access Memory (FPMDRAM), Extended Data Output Dynamic Random Access Memory (EDODRAM), Synchronous Dynamic Random Access Memory (SDRAM), or the like, as appropriate.

[0114] The memory 62 can be used to store or buffer various data files required for processing and / or communication, and possible computer program instructions executed by the processor 61.

[0115] The processor 61 reads and executes the computer program instructions stored in the memory 62 to implement any of the coal quality detection methods in the above embodiments.

[0116] In an embodiment, the coal quality detection device can further include a communication interface 63 and a bus 60. As shown in the figure, the processor 61, the memory 62, and the communication interface 63 are connected through the bus 60 and complete communication with each other. Figure 6

[0117] The communication interface 63 is used to realize communication between various modules, devices, units and / or equipment in the embodiments of the present application. The communication interface 63 can also realize data communication with other components, such as external devices, image / data acquisition devices, databases, external storage, image / data processing workstations, etc.

[0118] ​Bus 60 includes hardware, software, or both, to couple various components of a coal quality detection device to each other. Bus 60 includes, but is not limited to, at least one of the following: a data bus, an address bus, a control bus, an expansion bus, a local bus. By way of example and not limitation, bus 60 can include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable bus or combination of two or more of these. Where appropriate, bus 60 can include one or more buses. Although this application describes and illustrates a particular bus, this application contemplates any suitable bus or interconnect.

[0119] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, having stored thereon a program, wherein the program is executed by a processor to implement the coal quality detection method according to the first aspect.

[0120] More specifically, the computer readable storage medium can include, but is not limited to, a portable disc, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0121] In possible implementation manners, the present application can also be implemented in the form of a program product, which comprises program codes for causing a terminal device to execute steps of a coal quality detection method provided by the first aspect when the program product is run on the terminal device.

[0122] The program codes for executing the present application can be written in any combination of one or more programming languages, and can be executed completely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or completely on a remote device.

[0123] The technical features of the above embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist in contradiction, they shall be considered as falling within the scope of the present application.

[0124] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it shall not be understood as a limitation on the patent scope of the present application. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present application, and these shall fall within the protection scope of the present application. Therefore, the patent protection scope of the present application shall be subject to the appended claims.

Claims

1. A method of coal quality detection, characterized by, The method comprises: obtaining color texture feature parameters according to a pre-acquired standard coal image, and obtaining Raman spectrum feature parameters according to a pre-acquired standard coal Raman spectrum image; based on the Raman spectrum feature parameters, a prediction model of the ash content in the coal is constructed by increasing the weight of the color texture feature parameters; specifically, after feature selection by neighborhood component analysis, a weight factor is set for the color texture feature parameters, and the weight factor is multiplied to increase the weight of the color texture feature parameters, thereby constructing the prediction model of the ash content in the coal; based on the color texture feature parameters, a prediction model of the moisture and volatile content in the coal is constructed by increasing the weight of the Raman spectrum feature parameters; specifically, after feature selection by neighborhood component analysis, a weight factor is set for the Raman spectrum feature parameters, and the weight factor is multiplied to increase the weight of the Raman spectrum feature parameters, thereby constructing the prediction model of the moisture and volatile content in the coal; coupling the prediction model of the ash content in the coal and the prediction model of the moisture and volatile content in the coal to obtain a target coal quality prediction model; obtaining color texture feature parameters and Raman spectrum feature parameters of a coal sample to be detected, and obtaining the component content of the coal sample to be detected based on the target coal quality prediction model.

2. The method of claim 1, wherein, Before the color texture feature parameters are obtained according to the pre-acquired standard coal image, and the Raman spectrum feature parameters are obtained according to the pre-acquired standard coal Raman spectrum image, the method further comprises: image acquisition and Raman spectrum testing are performed on standard coals of multiple different coal components to obtain images and Raman spectrum images of the standard coals; wherein, the coal components include moisture, volatile matter, fixed carbon and ash, and the same standard is used for the standard coal samples of the different coal components, and the standard is an air-dry basis.

3. The method of claim 2, wherein, Before the image acquisition and Raman spectrum testing are performed on the standard coals of multiple different coal components, the method further comprises: preprocessing the Raman spectrum of the standard coal, and the preprocessing comprises segmenting the Raman spectrum into first-order Raman spectrum and second-order Raman spectrum; The wavelength range of the first-order Raman spectrum is selected as 400-2400 cm -1 , the wavelength range of the second-order Raman spectrum is selected as 2400-3800 cm -1 , and the first-order Raman spectrum and the second-order Raman spectrum are subjected to baseline removal.

4. The method of claim 1, wherein the color texture feature parameters include gray first moment, gray second moment, gray third moment, color standard deviation, dominant hue, energy, contrast, correlation, entropy and uniformity; the Raman spectrum feature parameters include peak area, total peak area of first-order Raman spectrum and total peak area of second-order Raman spectrum, and a combination of the peak area and the total peak area of the first-order Raman spectrum and the total peak area of the second-order Raman spectrum.

5. The method of claim 1, wherein, Before the prediction model of the ash content in the coal is constructed based on the Raman spectrum feature parameters by increasing the weight of the color texture feature parameters, the method further comprises: scaling the color texture feature parameters and the Raman spectrum feature parameters by Z-score standardization, so that the scales of the color texture feature parameters and the Raman spectrum feature parameters are within a preset range.

6. The method of claim 5, wherein, After the color texture feature parameters and the Raman spectrum feature parameters are scaled by Z-score standardization, the method further comprises: The number of the color texture feature parameters and the Raman spectrum feature parameters is selected by using neighborhood component analysis, wherein, several parameters with stronger correlation are selected as the color texture feature parameters, and several parameters with stronger correlation are selected as the Raman spectrum feature parameters.

7. The method of claim 1, wherein, The component content of the coal sample to be detected includes fixed carbon content, ash content, moisture content and volatile content, wherein the fixed carbon content is determined according to the ash content, the moisture content and the volatile content.

8. A coal quality detection system characterized by, The system includes an acquisition feature parameter module, a prediction model construction module of ash content in coal, a prediction model construction module of moisture and volatile content in coal, a target coal quality prediction model module and an acquisition coal quality detection result module, wherein: The acquisition feature parameter module is configured to acquire color texture feature parameters from a standard coal image acquired in advance and acquire Raman spectrum feature parameters from a standard coal Raman spectrum image acquired in advance; The prediction model construction module of ash content in coal is configured to construct a prediction model of ash content in coal based on the Raman spectrum feature parameters by increasing the weight of the color texture feature parameters; specifically, after feature selection by using neighborhood component analysis, a weight factor is set for the color texture feature parameters, and the color texture feature parameters are multiplied by the weight factor to increase the weight of the color texture feature parameters, so as to construct the prediction model of ash content in coal; the prediction model construction module of moisture and volatile content in coal is configured to construct a prediction model of moisture and volatile content in coal based on the color texture feature parameters by increasing the weight of the Raman spectrum feature parameters; specifically, after feature selection by using neighborhood component analysis, a weight factor is set for the Raman spectrum feature parameters, and the Raman spectrum feature parameters are multiplied by the weight factor to increase the weight of the Raman spectrum feature parameters, so as to construct the prediction model of moisture and volatile content in coal; The target coal quality prediction model module is configured to couple the prediction model of ash content in coal and the prediction model of moisture and volatile content in coal to obtain a target coal quality prediction model; The acquisition coal quality detection result module is configured to acquire color texture feature parameters and Raman spectrum feature parameters of a coal sample to be detected, and acquire component content of the coal sample to be detected based on the target coal quality prediction model.

9. An electronic device, comprising: A computer program is stored in a memory and executable on a processor, and the processor implements a coal quality detection method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executable on a processor to implement a coal quality detection method according to any one of claims 1 to 7.

Images