Coal mining in-situ identification method and system based on element and spectral characteristics

Abstract

The invention relates to the technical field of coal seam mining, and particularly discloses a coal seam mining in-situ identification method and system based on element and spectral characteristics, and the method comprises the steps: obtaining spectral data and element content data of a mining interface in real time, and obtaining a coal-rock proportion identification result of the current mining interface based on a pre-constructed coal-rock identification model; if the identified coal-rock proportion is higher than a set threshold value, mining work is carried out; acquiring spectral data and element content data of the mined material in real time, and based on a pre-constructed coal gangue identification model, obtaining an identification result that the material belongs to coal, coal gangue or a coal and coal gangue mixed material; the sorted coal and coal gangue are conveyed respectively, spectral data and element content data of coal materials in the conveying process are obtained, and coal types are recognized based on a pre-constructed coal type recognition model; according to the method, in-situ identification and optimization of coal types in the mining process are achieved, and the purity and efficiency of coal mining are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of coal seam mining, and in particular to an in-situ identification method and system for coal seam mining based on element and spectral characteristics. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] During the mining process of coal mine shafts and tunnels, the sudden change of the boundary between coal and rock increases the difficulty of mining and easily causes damage to construction machinery; the mined coal is often mixed with gangue, which greatly affects the quality of the output coal; and the similarity in appearance between different types of coal makes it difficult to effectively distinguish between multiple types of coal, seriously affecting the economic benefits of coal.

[0004] Therefore, the rapid and effective identification of coal-rock boundaries, coal gangue and coal types is of great value in the entire coal mine mining process and can improve the mining efficiency of coal mine projects.

[0005] Existing technologies often only screen and identify a single process in the mining process, such as single coal rock identification or single coal gangue identification, etc., and cannot achieve online in-situ identification of different processes in the entire coal mine shaft and tunnel mining process, and cannot guarantee full process identification of the entire mining process.

[0006] In addition, for the identification of coal-rock boundaries, gangue or coal types, the existing technologies mainly use process signal monitoring, infrared thermal imaging, images or reflection spectra and other methods to achieve them. However, the methods of process signal monitoring or infrared thermal imaging are too dependent on single indirect characteristics of coal, gangue or rock, such as heat, hardness, etc., and cannot distinguish them from the essence of the composition; these characteristics may be similar between different substances and cannot be distinguished from the fundamental composition, which can easily lead to confusion between coal and gangue or rock, affecting the accuracy of identification and the efficiency of mining.

[0007] The image recognition method mainly relies on appearance features. Some types of coal and coal gangue are very similar in appearance and are difficult to distinguish with the naked eye or simple image recognition methods. In coal mines, coal, coal gangue and rock are often mixed, further increasing the difficulty of identification; therefore, the image recognition method is prone to confusion among coal rock, coal gangue and coal types, and falls into the disadvantage of inaccuracy.

[0008] Reflection spectroscopy and element identification are fast, non-destructive, and efficient, and can reflect the corresponding characteristics from the essence of the material. However, mixtures of multiple substances are prone to cause overlapping, annihilation and other variation characteristics of the mixed spectrum, which are difficult to interpret with high precision using traditional spectroscopy knowledge, resulting in reduced accuracy in coal-rock boundary identification, coal gangue identification, or coal type identification, making it difficult to provide technical guidance for actual coal mine engineering construction. Summary of the invention

[0009] In order to solve the above problems, the present invention proposes an in-situ identification method and system for coal seam mining based on elemental and spectral characteristics, which can accurately identify the coal-rock interface during the cutting stage, and can accurately distinguish coal gangue from coal after mining is completed to improve the quality of coal and further identify the type of coal.

[0010] In some embodiments, the following technical solutions are adopted: A coal seam mining in-situ identification method based on element and spectral characteristics, comprising: Acquire the spectrum data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model; If the identified coal-rock ratio is higher than the set threshold, mining is carried out; the spectral data and element content data of the mined material are obtained in real time, and based on the pre-built coal gangue identification model, the identification result of the material belonging to coal, coal gangue, or a mixture of coal and coal gangue is obtained; the coal and coal gangue mixture are separated to obtain coal; The sorted coal and gangue are transported separately, the spectral data and element content data of the coal materials during the transportation process are obtained, and the coal types are identified based on the pre-built coal type identification model; Among them, coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, the mixed spectrum and element content data of each mixed sample are obtained respectively to form a first data set, and the coal-rock recognition model is trained using the first data set.

[0011] As a further solution, it also includes: mixing coal and gangue samples in different proportions to prepare a variety of mixed samples of coal and gangue, respectively obtaining end member spectra and element contents of the coal and gangue samples, and mixed spectra and element content data of each mixed sample to form a second data set, and using the second data set to train the gangue recognition model.

[0012] As a further solution, it also includes: mixing bituminous coal and anthracite samples in different proportions to prepare a variety of mixed coal samples, respectively obtaining end member spectra and element contents of the bituminous coal samples and the anthracite samples, and mixed spectra and element content data of each mixed sample to form a third data set, and using the third data set to train the coal type recognition model.

[0013] As a further solution, the spectral data includes reflectivity data corresponding to a plurality of different wavelengths, and a corresponding spectral curve is obtained by fitting.

[0014] As a further solution, during the training of the coal type identification model, a quantitative relationship between the spectral features and the contents of different minerals in the mixed sample is established through linear regression, specifically: I 1 C-H =I a1 C-H ·C 无烟煤 +I b1 C-H ·C 有烟煤 ; C 无烟煤 +C 有烟煤 =1; Among them, I 1 C-H is the intensity of the CH characteristic peak in the mixed spectrum, C 无烟煤 is the relative content of anthracite in the mixed sample, I a1 C-H is the intensity of the CH characteristic peak in the anthracite end member spectrum; C 有烟煤 is the relative content of bituminous coal in the mixed sample; I b1 C-H is the intensity of the CH characteristic peak in the end-member spectrum corresponding to bituminous coal.

[0015] As a further solution, during training, the coal rock identification model, coal gangue identification model or coal type identification model extracts the change characteristics of the signature element content of the mixed samples at different mixing ratios compared with the signature element content in anthracite based on the element sequence and content of the mixed samples at different mixing ratios, and establishes a corresponding relationship between the mixed samples at different mixing ratios and the change characteristics of the signature element content; during actual prediction, based on the acquired signature element content in the sample to be tested, the mineral type and proportion in the sample to be tested are matched; wherein the signature elements include at least: carbon, hydrogen, sulfur and silicon.

[0016] As a further solution, the coal rock identification model, coal gangue identification model or coal type identification model is implemented using a convolutional neural network structure.

[0017] In other embodiments, the following technical solutions are adopted: An in-situ identification system for coal seam mining based on element and spectral characteristics, comprising: The coal-rock identification module is used to obtain the spectral data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model; if the identified coal-rock ratio is higher than the set threshold, mining work is carried out; The gangue identification module is used to obtain the spectral data and element content data of the mined materials in real time after the mining is completed. Based on the pre-built gangue identification model, the identification result of whether the material belongs to coal, gangue or a mixture of coal and gangue is obtained; the coal and gangue mixture are separated to obtain coal; The coal type identification module is used to transport the sorted coal and coal gangue separately, obtain the spectral data and element content data of the coal materials during the transportation process, and identify the coal type based on the pre-built coal type identification model; Among them, coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, the mixed spectrum and element content data of each mixed sample are obtained respectively to form a first data set, and the coal-rock recognition model is trained using the first data set.

[0018] In other embodiments, the following technical solutions are adopted: A terminal device includes a processor and a memory, wherein the processor is used to implement instructions; the memory is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executing the above-mentioned in-situ identification method for coal seam mining based on elemental and spectral characteristics.

[0019] In other embodiments, the following technical solutions are adopted: A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor of a terminal device and executing the above-mentioned in-situ identification method for coal seam mining based on elemental and spectral features.

[0020] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention can identify the coal-rock mixed zone in real time during the mining process and perform quantitative and precise cutting, which can significantly reduce the mixing of non-coal materials; at the same time, it can accurately and effectively sort coal and coal gangue through coal gangue identification; for the screened coal samples, bituminous coal and anthracite are distinguished through coal type identification, realizing in-situ identification and optimization of coal types during the mining process, and improving the purity and efficiency of coal mining.

[0021] (2) The present invention pre-constructs mixed samples of coal rock, coal gangue and coal types with different mixing ratios based on actual engineering experience, obtains spectral data and elemental content data of each mixed sample, and forms a data set of mixed samples with different mixing ratios and their corresponding spectral and elemental content data. The data set is used to train the corresponding coal rock identification model, coal gangue identification model or coal type identification model, and accurate mineral type and content identification results can be obtained, which solves the problem that the mixed sample spectrum cannot be directly interpreted in the prior art; this precise identification can maximize the use of useful resources in the ore deposit, reduce resource waste, and improve overall resource utilization and sorting efficiency.

[0022] (3) This patent combines spectral and elemental features with machine learning, which enables the system to more accurately identify complex mixed spectral features. Machine learning can provide higher accuracy and reliability through large amounts of data training. The present invention establishes an identification system based on elemental and spectral variation features, which can accumulate a large amount of spectral data and elemental variation feature data; these data can not only be used for real-time decision-making, but also for optimization training of recognition models, providing important data support for the subsequent dynamic optimization of recognition models.

[0023] Other features and advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flow chart of an in-situ identification method for coal seam mining based on element and spectral characteristics in an embodiment of the present invention; Figure 2 Schematic diagram of the coal-rock identification process in an embodiment of the present invention; Figure 3 Schematic diagram of end member spectra of anthracite and carbon-containing coal gangue and mixed spectrum data of mixed samples in an embodiment of the present invention; Figure 4 This is a schematic diagram of the content of characteristic elements mixed in mineral samples at different proportions in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0026] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0027] Embodiment 1 In one or more embodiments, a method for in-situ identification of coal seam mining based on elemental and spectral features is disclosed, combined with Figure 1 , specifically including the following process: S101: Acquire the spectrum data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model.

[0028] In this embodiment, firstly, ore deposit exploration and geological exploration are carried out on the mining area to be tested, and the geological characteristics of the mining area are fully investigated to provide geological information constraints for the qualitative identification of minerals; wherein, the geological characteristics include geological structure, lithology distribution, stability, etc.

[0029] Methods for conducting mineral deposit exploration and geological exploration include but are not limited to: geological radar, drone aerial photography, drilling exploration, geological mapping and other methods; the main purpose of conducting mineral deposit exploration and geological exploration in this embodiment is to obtain detailed information such as the geological structure, lithological characteristics, stability, etc. of the mining area, evaluate the geological conditions and potential risks of the mining area, and provide geological information constraints for qualitative identification of minerals.

[0030] Specifically, surveys were conducted along the mining area, and geological radar, geological mapping and drone aerial photography were used to identify and record the geological structure of the mining area; core samples were obtained through drilling exploration, and petrological analysis was performed, and the spatial distribution of lithology distribution was determined by combining geological radar and geological mapping; the stability of the rock mass was evaluated through field observations and geotechnical engineering tests; and it was finally found that the overall structure of the mining area was mainly monocline, with small folds and faults developed locally, the main rock types in the mining area included sandstone, and no major geological tectonic movements had occurred in history. The model trained with data from this region in this embodiment is best suited in this region or in an environment with similar geological characteristics to this region.

[0031] In this embodiment, before mining, the coal-rock ratio of the current mining interface is first identified. If the coal / rock ratio is lower than the set threshold (for example, the threshold can be set to 0.75), it means that the coal content is low, which may cause the mined coal to be unclean, causing trouble for subsequent processing and affecting economic benefits; if the coal / rock ratio is higher than the set threshold, it means that the coal content is high, and subsequent mining work will continue.

[0032] This embodiment can ensure the accuracy of the excavation process and avoid unnecessary waste of resources by identifying the coal-rock ratio at the mining interface.

[0033] S102: If the identified coal-rock ratio is higher than the set threshold, mining is carried out; the spectral data and element content data of the mined material are obtained in real time, and based on the pre-built gangue identification model, the identification result of the material belonging to coal, gangue or a mixture of coal and gangue is obtained; the coal and gangue mixture are separated to obtain coal.

[0034] In this embodiment, after the coal seam mining is completed, the raw coal obtained by mining is first preliminarily screened and impurities are removed, and then the raw coal is crushed into smaller block materials. The material is evenly spread on the conveyor belt, and the element content data and spectral data of the material are collected in real time using a spectral sensor and a LIBS analyzer. The collected element content and spectral data are input into a pre-built coal gangue identification model for real-time analysis and identification, identifying the mineral type of a certain piece of material, and obtaining the relative content of coal and coal gangue in a certain piece of material. Pretreatment is performed, and the separation device is controlled to select pure coal, pure coal gangue, and coal-coal gangue mixed materials respectively; the selected coal-coal gangue mixed materials are further separated using existing technology to obtain pure coal.

[0035] S103: The sorted coal and coal gangue are transported separately, the spectral data and element content data of the coal materials during the transportation process are obtained, and the coal types are identified based on the pre-built coal type identification model.

[0036] In this embodiment, the obtained pure coal is evenly spread on a conveyor belt, and the element content data and spectrum data of the pure coal are collected in real time using a spectral sensor and a LIBS analyzer. The collected element content and spectrum data are input into a pre-built coal type identification model for real-time analysis and identification, and the coal type in a piece of pure coal is identified, and the relative content of bituminous coal and anthracite is obtained. Pretreatment is performed, and a separation device is controlled to select bituminous coal, anthracite and mixed coal respectively. The selected mixed coal is further separated using existing technology to obtain pure anthracite and pure bituminous coal.

[0037] The method of this embodiment realizes accurate in-situ identification of bituminous coal and anthracite during coal mining based on element content and spectral data, and will not be affected by the superposition of spectral data.

[0038] In this embodiment, the inputs of the coal rock identification model, the coal gangue identification model and the coal type identification model are all element content data and spectral data; wherein the spectral data includes corresponding reflectivity data at multiple different wavelengths, and the corresponding spectral curve is obtained by fitting.

[0039] Coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, as well as the mixed spectrum and element content data of each mixed sample are obtained to form a first data set, which is used to train the coal-rock recognition model.

[0040] Coal and gangue samples are mixed in different proportions to prepare a variety of mixed samples of coal and gangue, and the end member spectra and element contents of the coal and gangue samples, as well as the mixed spectrum and element content data of each mixed sample are obtained to form a second data set, which is used to train the gangue recognition model.

[0041] Bituminous coal and anthracite samples are mixed in different proportions to prepare a variety of mixed coal samples. The end member spectra and element contents of the bituminous coal samples and the anthracite samples, as well as the mixed spectra and element content data of each mixed sample are obtained to form a third data set, which is used to train the coal type recognition model.

[0042] Among them, the characteristics of endmember spectra refer to the spectral data obtained from pure samples (such as anthracite, carbon-containing coal gangue, and sandstone); including the characteristic peak position, intensity, and shape of each pure mineral, representing the absorption and reflection characteristics of these samples; the characteristics of mixed spectra refer to the spectral data of mixed samples, representing the combination of spectral characteristics of mixed minerals, manifested as changes in characteristic peaks, shifts in spectral peaks, and changes in the intensity of absorption bands.

[0043] When two minerals of different proportions are mixed together, their mixed spectral characteristics will change, showing a phenomenon inconsistent with the spectral characteristics of a single pure mineral, such as: shift of spectral peak, annihilation of characteristic absorption bands, etc. The variation characteristics of the mixed spectrum, such as the band shape, absorption peak position and intensity, contain the spectral information of each component, and become an important basis for identifying the mixed sample pattern. Figure 3 The spectral data of 20% anthracite mixed with 80% carbon gangue are given, as well as the end member spectra of anthracite and carbon gangue. It can be seen that the peak intensity and position of the spectrum after mixing have shifted, so the spectral data can reflect the change in the proportion of the mixed sample.

[0044] In addition, this embodiment pre-tested the main elements in anthracite, carbon-containing coal gangue and sandstone, and found that the carbon content was the highest in anthracite, followed by carbon-containing coal gangue, and the lowest in sandstone. The carbon content decreased significantly after mixing, and the higher the ratio of coal gangue / rock after mixing, the greater the decrease. The hydrogen content was higher in anthracite, and the hydrogen content decreased slightly after mixing. The sulfur content was higher in anthracite, and the sulfur content decreased slightly after mixing. The silicon content was lower in anthracite, and the silicon content was higher in carbon-containing coal gangue and sandstone. The silicon content increased significantly after mixing, and the higher the mixing ratio, the greater the increase. The changing trends and characteristic patterns of these elements can be used to identify the presence and relative content of each component, reflecting changes in the proportion of mixed samples; Figure 4 A schematic diagram of the changes in the contents of the characteristic elements carbon (C), hydrogen (H), sulfur (S), and silicon (Si) in the mixed samples of 10% coal + 90% rock, 15% coal + 85% rock, 20% coal + 80% rock, and 25% coal + 75% rock is given.

[0045] This embodiment combines the two characteristics of spectrum and elements to more accurately analyze and identify the composition of the mixed sample.

[0046] The recognition processes of the coal rock recognition model, the coal gangue recognition model and the coal type recognition model in this embodiment are basically the same, except that the training data sets and input data of different models are different.

[0047] As a specific example, take coal rock identification as an example to illustrate. Figure 2 , the specific method is as follows: S1011: Select samples of common rock and coal types in the coal mining process, and use a high-resolution spectrometer to obtain endmember spectral information of the above samples. The spectral information covers the spectral range from visible light to near-infrared and even mid-infrared, that is, it includes corresponding reflectance data at multiple different wavelengths. The corresponding spectral curve can be obtained by fitting multiple (wavelength, reflectance) data; the obtained endmember spectral data is stored in a dedicated database to establish an endmember spectral library.

[0048] When selecting samples, representative samples of coal and rock should be collected in the coal seam and rock stratum enrichment areas according to the mining area exploration results in (1). Ensure that the samples cover coal qualities (such as anthracite, bituminous coal, lignite), different lithologies (such as sandstone, shale, mudstone), and the collected samples are numbered, cleaned, dried, and properly stored in a light-proof and moisture-proof environment for subsequent experimental use.

[0049] The high-resolution spectrometers used include Fourier transform infrared spectrometer and Raman spectrometer. In this example, Fourier transform infrared spectrometer is used to perform spectral testing on the collected samples, and the spectral data of the spectral library is denoised and smoothed to improve the accuracy and readability of the data, and the data is standardized and normalized to eliminate the differences between different data sources.

[0050] S1012: According to the actual engineering ratio, different types of coal and rock are mixed to prepare a variety of mixed samples of coal and rock to ensure that the mixed samples are representative and diverse. Collect mixed samples for spectral analysis to obtain spectral data at different mixing ratios.

[0051] Specifically, the coal and rock ratio selected in this embodiment covers multiple gradients of mixing in actual mining projects, including: 10% coal + 90% rock, 15% coal + 85% rock, 20% coal + 80% rock, 25% coal + 75% rock; spectral data of these mixed samples are obtained.

[0052] In order to improve the quality of data, highlight the spectral features, and improve the accuracy of spectral analysis, this embodiment preprocesses the acquired spectral data, including: removing the baseline drift in the spectral data to make the spectral curve smoother for subsequent analysis; using digital filtering technology, such as smoothing filtering, detrending filtering, etc., to reduce random noise in the spectral data and improve the signal-to-noise ratio; normalizing the spectral data to eliminate the intensity differences caused by factors such as sample quantity and instrument response, so that the spectral data of different samples are comparable.

[0053] In this embodiment, in order to reduce the impact of environmental factors on spectral data, the temperature and humidity in the spectrometer laboratory are controlled to be maintained within a constant range, and the temperature can be controlled at 20-25°C and the humidity can be controlled at 40%-60%RH. At the same time, direct exposure to external light sources is reduced to avoid vibration interference, ensuring that the spectrometer is in a stable working environment; during the acquisition process, attention should be paid to controlling environmental conditions to ensure the accuracy and consistency of the data.

[0054] S1013: Conduct detailed elemental testing on the mixed samples in S1012 and analyze the elemental composition of the mixed samples. Establish elemental sequences of samples under different mixing ratios, extract characteristic elemental variation characteristics under different mixing modes, form a data set, and provide basic data for subsequent analysis.

[0055] The element testing method of this embodiment adopts laser induced breakdown spectroscopy testing; the characteristic element change feature refers to the change in the content of the main elements after mixing mineral samples in different proportions, etc., specifically refers to the change in the content of elements such as carbon, hydrogen, sulfur, and silicon at different mixing proportions (extracting the element content of the mixed sample at different mixing proportions).

[0056] S1014: Construct a coal-rock identification model; train the coal-rock identification model using the mixed samples of coal and rock mixed in different proportions obtained in S1011-S1013, and their corresponding spectral data and element content data.

[0057] In this embodiment, the coal-rock identification model performs feature analysis on the obtained spectral data, and uses spectral analysis technology to extract end-member spectra and mixed spectral features. Based on the end-member spectra, the variation law of the mixed spectra with the change of end-member content and type is analyzed. A quantitative relationship between spectral features and mineral content is established.

[0058] Specifically, the spectral analysis techniques used include derivative spectroscopy, peak-valley identification, multivariate curve resolution, etc., and the position, intensity, shape and other parameters of characteristic peaks are focused on when extracting endmember spectroscopy and mixed spectral features. In this embodiment, the endmember spectral data obtained from pure samples (anthracite, carbonaceous coal gangue, sandstone) are analyzed to obtain the main characteristic peaks of the anthracite endmember spectrum at 3000-2800 cm -1 (CH stretching vibration) and 1600 cm -1 (C=C aromatic ring vibration); the main characteristic peaks of the end member spectrum of carbon-containing coal gangue are at 3600-3200 cm -1 (OH stretching vibration), 1000-1200 cm -1 (Si-O stretching vibration); the main characteristic peak of the sandstone end member spectrum is at 3600-3200 cm -1 (OH stretching vibration), 1000-1200 cm -1 (Si-O stretching vibration); In this example, the variation law of the mixed spectrum with the change of endmember content and type is analyzed based on the endmember spectrum as the endmember content and type change. That is, the spectral data obtained from samples with different mixing ratios are analyzed, and the characteristic peak position of the mixed spectrum is basically consistent with the characteristic peak position of the endmember spectrum, but the peak intensity and shape change with the change of endmember content. The intensity of the characteristic peaks of CH, C=C, OH and Si-O in the mixed spectrum changes with the endmember content. The intensity of the CH characteristic peak increases with the increase of anthracite content, and the intensity of the OH and Si-O characteristic peaks increases with the increase of carbon-containing coal gangue and sandstone content; the shape of the characteristic peaks in the mixed spectrum (such as peak width, symmetry) changes with the type and content of the endmember. The CH characteristic peak in the mixed sample is narrower and symmetrical when the anthracite content is high, and wider and asymmetric when the carbon-containing coal gangue and sandstone content is high.

[0059] The quantitative relationship between spectral characteristics and mineral content is established by linear regression. The specific formula is as follows: ; ; in, is the intensity of the kth characteristic peak in the mixed sample, is the intensity of the kth characteristic peak of the ith mineral, is the relative content of the i-th mineral, n is the number of mineral species involved in the mixed sample, n≥2, k=n-1.

[0060] The above formula is universal for coal rock identification model, coal gangue identification model and coal type identification model; for example: When identifying coal types, there are two types of minerals: bituminous coal and anthracite.

[0061] The quantitative relationship is: I 1 C-H =I a1 C-H *C 无烟煤 +I b1 C-H *C 有烟煤 ; C 无烟煤 +C 有烟煤 =1; Among them, I 1 C-H is the intensity of the CH characteristic peak in the mixed spectrum, C 无烟煤 is the relative content of anthracite in the mixed sample, I a1 C-H is the intensity of the CH characteristic peak in the anthracite end member spectrum; C 有烟煤 is the relative content of bituminous coal in the mixed sample; I b1 C-H is the intensity of the CH characteristic peak in the end-member spectrum corresponding to bituminous coal.

[0062] When identifying coal gangue, there are usually four types of minerals: bituminous coal, anthracite, carbon-containing gangue and carbon-free gangue. The quantitative relationship is: I 2 C-H =I a2 C-H ·C 无烟煤 +I b2 C-H ·C 有烟煤 +I c2 C-H ·C 含碳煤矸石 +I d2 C-H ·C 无碳煤矸石 ; I 2 O-H =I a2 O-H ·C 无烟煤+I b2 O-H ·C 有烟煤 +I c2 O-H ·C 含碳煤矸石 +I d2 O-H ·C 无碳煤矸石 ; I 2 SI-O =I a2 SI-O ·C 无烟煤 +I b2 SI-O ·C 有烟煤 +I c2 SI-O ·C 含碳煤矸石 +I d2 SI-O ·C 无碳煤矸石 ; C 无烟煤 +C 有烟煤 +C 含碳煤矸石 +C 无碳煤矸石 =1; Among them, I 2 C-H is the intensity of the CH characteristic peak in the mixed spectrum, C 无烟煤 is the relative content of anthracite in the mixed sample, I a2 C-H is the intensity of the CH characteristic peak in the anthracite end member spectrum; C 有烟煤 is the relative content of bituminous coal in the mixed sample; I b2 C-H is the intensity of the CH characteristic peak in the end member spectrum corresponding to bituminous coal; C 含碳煤矸石 is the relative content of carbon-containing coal gangue in the mixed sample, I c2 C-H is the intensity of the CH characteristic peak in the end member spectrum corresponding to the carbon-containing coal gangue; C 无碳煤矸石 is the relative content of carbon-free coal gangue in the mixed sample, I d2 C-H is the intensity of the CH characteristic peak in the end member spectrum of carbon-free coal gangue; I 2 O-H ,I 2 SI-O are the intensities of the OH characteristic peak and SI-O characteristic peak in the mixed spectrum, I a2 O-H ,I a2 SI-O are the intensities of the OH characteristic peak and SI-O characteristic peak in the anthracite end member spectrum; I b2 O-H ,I b2SI-O are the intensities of the OH characteristic peak and SI-O characteristic peak in the end-member spectrum corresponding to bituminous coal; I c2 O-H ,I c2 SI-O are the intensities of the OH characteristic peak and SI-O characteristic peak in the end-member spectrum corresponding to carbon-containing coal gangue; I d2 O-H ,I d2 SI-O They are respectively the intensities of the OH characteristic peak and the SI-O characteristic peak in the end-member spectrum corresponding to carbon-free coal gangue.

[0063] In coal rock identification, minerals usually include: bituminous coal, anthracite, carbon-containing coal gangue, carbon-free coal gangue, sandstone and shale. Based on the same principle, the intensities of the CH characteristic peak, OH characteristic peak, SI-O characteristic peak, C=C characteristic peak and SO characteristic peak in the mixed spectrum can be used to construct a relationship.

[0064] In the above relationship, the intensity of each characteristic peak in the mixed spectrum and the intensity of each characteristic peak in the end-member spectra of different minerals can be obtained from the end-member spectrum library or through collection, so the proportion of different mineral contents can be determined.

[0065] In this embodiment, the coal-rock identification model performs characteristic analysis on the input element content data, and based on the element sequence and content of the mixed samples at different mixing ratios, extracts the change characteristics of the contents of signature elements such as carbon, hydrogen, sulfur and silicon in the mixed samples at different mixing ratios compared with the signature element contents in anthracite, and establishes a correspondence between the mixed samples at different mixing ratios and the change characteristics of the signature element contents. During actual prediction, based on the acquired signature element contents in the sample to be tested, the mineral types and proportions in the sample to be tested can be matched.

[0066] S1015: The spectral data and element content data of the mining interface are acquired in real time. Based on the coal-rock identification model constructed and trained in S1014, the coal-rock ratio identification result of the current mining interface can be output.

[0067] The identification process of gangue identification and coal type identification is basically the same as the above-mentioned coal rock identification process, and the identification principles of the constructed gangue identification model and coal type identification model are also exactly the same as the identification principles of the coal rock identification model. The difference is: When the gangue identification model is trained, the endmember spectra, mixed spectra and element content data of different types of pure coal (such as anthracite, bituminous coal, lignite), pure gangue (such as carbonaceous gangue, carbon-free gangue) and mixed samples of pure coal and pure gangue in different proportions are obtained, such as: 10% coal + 90% gangue, 15% coal + 85% gangue, 20% coal + 80% gangue, 25% coal + 75% gangue; during actual prediction, the spectral data and element content data of the mined material are obtained, and based on the pre-built and trained gangue identification model, the identification results of whether the material belongs to coal, gangue or a mixture of coal and gangue and their respective proportions are obtained.

[0068] When the coal type identification model is trained, the endmember spectra, mixed spectra and element content data of different types of pure coal (such as anthracite, bituminous coal, lignite) and mixed samples of different types of pure coal in different proportions are obtained; during actual prediction, the spectral data and element content data of the coal material during the transportation process are obtained, and the identification results of anthracite and bituminous coal and their proportions are obtained based on the pre-built and trained coal type identification model.

[0069] Embodiment 2 In one or more embodiments, a coal seam mining in-situ identification system based on elemental and spectral features is disclosed, comprising: The coal-rock identification module is used to obtain the spectral data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model; if the identified coal-rock ratio is higher than the set threshold, mining work is carried out; The gangue identification module is used to obtain the spectral data and element content data of the mined materials in real time after the mining is completed. Based on the pre-built gangue identification model, the identification result of whether the material belongs to coal, gangue or a mixture of coal and gangue is obtained; the coal and gangue mixture are separated to obtain coal; The coal type identification module is used to transport the sorted coal and coal gangue separately, obtain the spectral data and element content data of the coal materials during the transportation process, and identify the coal type based on the pre-built coal type identification model; Among them, coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, the mixed spectrum and element content data of each mixed sample are obtained respectively to form a first data set, and the coal-rock recognition model is trained using the first data set.

[0070] It should be noted that the specific implementation of the above modules is the same as that in Example 1 and will not be described in detail.

[0071] Embodiment 3 In one or more embodiments, a terminal device is disclosed, which includes a processor and a memory, wherein the processor is used to implement instructions; the memory is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executing the in-situ identification method for coal seam mining based on elemental and spectral characteristics described in Example 1.

[0072] It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general-purpose processors, digital signal processors DSP, application-specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0073] The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0074] In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in a processor or an instruction in the form of software.

[0075] Embodiment 4 In one or more embodiments, a computer-readable storage medium is disclosed, in which a plurality of instructions are stored, wherein the instructions are suitable for being loaded by a processor of a terminal device and executing the in-situ identification method for coal seam mining based on elemental and spectral features described in Example 1.

[0076] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A method for in-situ identification of coal seam mining based on element and spectral characteristics, characterized in that: include: Acquire the spectrum data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model; If the identified coal-rock ratio is higher than the set threshold, mining is carried out; the spectral data and element content data of the mined material are obtained in real time, and based on the pre-built coal gangue identification model, the identification result of the material belonging to coal, coal gangue, or a mixture of coal and coal gangue is obtained; the coal and coal gangue mixture are separated to obtain coal; The sorted coal and gangue are transported separately, the spectral data and element content data of the coal materials during the transportation process are obtained, and the coal types are identified based on the pre-built coal type identification model; Among them, coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, the mixed spectrum and element content data of each mixed sample are obtained respectively to form a first data set, and the coal-rock recognition model is trained using the first data set.

2. The method for in-situ identification of coal seam mining based on element and spectral characteristics as claimed in claim 1, characterized in that: Also includes: Coal and gangue samples are mixed in different proportions to prepare a variety of mixed samples of coal and gangue, and the end member spectra and element contents of the coal and gangue samples, as well as the mixed spectrum and element content data of each mixed sample are obtained to form a second data set, and the second data set is used to train the gangue recognition model.

3. The method for in-situ identification of coal seam mining based on element and spectral characteristics as claimed in claim 1, characterized in that: Also includes: Bituminous coal and anthracite samples are mixed in different proportions to prepare a variety of mixed coal samples, and the end member spectra and element contents of the bituminous coal samples and the anthracite samples, as well as the mixed spectra and element content data of each mixed sample are obtained to form a third data set, and the third data set is used to train the coal type recognition model.

4. A method for in-situ identification of coal seam mining based on element and spectral characteristics as claimed in any one of claims 1 to 3, characterized in that: The spectral data includes reflectivity data corresponding to a plurality of different wavelengths, and the corresponding spectral curve is obtained by fitting.

5. The method for in-situ identification of coal seam mining based on element and spectral characteristics as claimed in claim 1, characterized in that: When training the coal type identification model, a quantitative relationship between spectral features and the contents of different minerals in the mixed sample is established through linear regression, specifically: I 1 C-H =I a1 C-H ·C 无烟煤 +I b1 C-H ·C 有烟煤 ; C 无烟煤 +C 有烟煤 =1; Among them, I 1 C-H is the intensity of the CH characteristic peak in the mixed spectrum, C 无烟煤 is the relative content of anthracite in the mixed sample, I a1 C-H is the intensity of the CH characteristic peak in the anthracite end member spectrum; C 有烟煤 is the relative content of bituminous coal in the mixed sample; I b1 C-H is the intensity of the CH characteristic peak in the end-member spectrum corresponding to bituminous coal.

6. The method for in-situ identification of coal seam mining based on element and spectral characteristics according to claim 1, characterized in that: When training, the coal rock identification model, coal gangue identification model or coal type identification model extracts the change characteristics of the signature element content of the mixed samples at different mixing ratios compared with the signature element content in anthracite based on the element sequence and content of the mixed samples at different mixing ratios, and establishes a corresponding relationship between the mixed samples at different mixing ratios and the change characteristics of the signature element content; during actual prediction, based on the acquired signature element content in the sample to be tested, the mineral type and proportion in the sample to be tested are matched; wherein the signature elements include at least: carbon, hydrogen, sulfur and silicon.

7. The method for in-situ identification of coal seam mining based on element and spectral characteristics according to claim 1, characterized in that: The coal rock identification model, coal gangue identification model or coal type identification model is implemented using a convolutional neural network structure.

8. An in-situ identification system for coal seam mining based on element and spectral characteristics, characterized in that: include: The coal-rock identification module is used to obtain the spectral data and element content data of the mining interface in real time, and obtain the coal-rock ratio identification result of the current mining interface based on the pre-built coal-rock identification model; if the identified coal-rock ratio is higher than the set threshold, mining work is carried out; The gangue identification module is used to obtain the spectral data and element content data of the mined materials in real time after the mining is completed. Based on the pre-built gangue identification model, the identification result of whether the material belongs to coal, gangue or a mixture of coal and gangue is obtained; the coal and gangue mixture are separated to obtain coal; The coal type identification module is used to transport the sorted coal and coal gangue separately, obtain the spectral data and element content data of the coal materials during the transportation process, and identify the coal type based on the pre-built coal type identification model; Among them, coal and rock samples are mixed in different proportions to prepare a variety of coal-rock mixed samples, and the end member spectra and element contents of the coal and rock samples, the mixed spectrum and element content data of each mixed sample are obtained respectively to form a first data set, and the coal-rock recognition model is trained using the first data set.

9. A terminal device, comprising a processor and a memory, wherein the processor is used to implement instructions; and the memory is used to store multiple instructions, characterized in that: The instructions are suitable for being loaded by a processor and executing the in-situ identification method for coal seam mining based on elemental and spectral features as described in any one of claims 1-7.

10. A computer-readable storage medium storing a plurality of instructions, characterized in that: The instructions are suitable for being loaded by a processor of a terminal device and executing the in-situ identification method for coal seam mining based on elemental and spectral features as described in any one of claims 1-7.

Images