Multi-dimensional fusion coal quality detection method and system
Through the multi-dimensional fusion coal quality detection method, coal sample feature information is collected and input into the pre-trained model for detection, which solves the accuracy and stability of online detection of coal quality coal in the existing technology, and achieves high intelligence and self-learning Accuracy detection meets the comprehensive analysis needs of the coal industry.
Patent Information
- Application Number
- CN202411161874.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-29
AI Technical Summary
The prior art has problems in the online detection of coal-quality coal quantity, low detection accuracy, environmental factors affect equipment stability and lack of comprehensive data analysis and processing, resulting in a lack of applicability and flexibility.
Using multi-dimensional fusion coal quality detection method, coal sample images, physical and chemical information are collected, and input them into the pre-trained coal quality detection model for detection. Based on historical data, the gold standard data set is constructed to improve detection accuracy. and robustness.
It has realized the online accurate detection of intelligent coal quality coal quality, built a gold standard data set, and realized self-learning high-precision multi-dimensional fusion coal quality inspection to meet the coal industry's needs for comprehensive analysis.
Smart Images

Figure CN119046684A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of coal quality analysis, and in particular to a multi-dimensional fusion coal quality detection method, a multi-dimensional fusion coal quality detection system, a machine-readable storage medium and an electronic device. Background Art
[0002] At present, the digital transformation of the coal industry urgently needs to grasp the comprehensive information of coal quality and quantity in real time. The existing technology usually detects coal quality and quantity based on different technologies (such as laser induced breakdown spectroscopy, X-ray fluorescence spectroscopy, near infrared spectroscopy and / or Raman spectroscopy). The existing technology has technical limitations in analyzing real-time coal quality and quantity online detection in industrial applications, including the following points: 1. The detection accuracy of different spectral methods is low; 2. Complex environmental factors will affect equipment stability; 3. Failure to conduct comprehensive and effective algorithmic analysis and processing of the large amount of data generated by different technologies; 4. Lack of technical applicability and flexibility in the entire coal industry chain.
[0003] Therefore, with regard to the comprehensive analysis and online detection of intelligent coal quality and quantity, how to realize accurate online detection of intelligent coal quality and quantity in order to grasp and predict comprehensive coal quality and quantity information is an urgent problem that needs to be solved. Summary of the invention
[0004] The purpose of the embodiments of the present invention is to provide a multi-dimensional fusion coal quality detection method and system to at least solve the above-mentioned problem of how to realize intelligent online accurate detection of coal quality and coal quantity to grasp and predict comprehensive coal quality and coal quantity information.
[0005] In order to achieve the above object, the first aspect of the present invention provides a multi-dimensional fusion coal quality detection method, comprising: Collect various coal quality characteristic information of the coal sample to be tested; Input various coal quality characteristic information into the pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; The coal quality detection model is obtained by training the initial model based on the coal quality detection and analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection and analysis atlas database. The coal quality detection and analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal samples and the manual test results corresponding to the historical coal samples. The initial model is used to predict the coal quality information of the historical coal samples.
[0006] Optionally, before collecting a plurality of coal quality characteristic information of the coal sample to be tested, the method further includes: Control the sampling device to transmit the coal sample to be tested, and pre-process the coal sample to be tested transmitted by the sampling device; wherein, The preprocessing process includes: The agitator is controlled to mix the coal sample to be tested, and the pressing roller is controlled to shape the mixed coal sample to be tested.
[0007] Optionally, the above-mentioned coal quality characteristic information includes coal sample image information; The above-mentioned coal quality characteristic information collected from the coal sample to be tested includes: Control the 3D scanner to collect external features of the coal sample to be tested, and obtain the coal sample image information of the coal sample to be tested; wherein, The coal sample image information of the coal sample to be tested includes three-dimensional shape data, external contour data, volume data, surface area data, spatial distribution arrangement data of coal particles, pore characteristic data, crack characteristic data, particle size data of coal particles and shape factor parameters of coal particles.
[0008] Optionally, the above-mentioned coal quality characteristic information includes physical information of the coal sample; The above-mentioned coal quality characteristic information collected from the coal sample to be tested includes: Control the particle size analyzer to perform particle size analysis on the coal sample to be tested, and obtain the physical information of the coal sample to be tested; wherein, The physical information of the coal sample to be tested includes the particle size distribution curve data, average particle size data, particle size range data, specific surface area data and content data of the preset particle size of the coal sample to be tested.
[0009] Optionally, the above-mentioned coal quality characteristic information includes coal sample chemical information; The above-mentioned coal quality characteristic information collected from the coal sample to be tested includes: The laser induced breakdown spectrometer, near infrared spectrometer, X-ray fluorescence spectrometer and Raman spectrometer are sequentially controlled to collect chemical characteristics of the coal sample to be tested, so as to obtain the chemical information of the coal sample to be tested; wherein, The coal sample chemical information of the coal sample to be tested includes content data of preset trace elements, moisture content data in coal, volatile content data, fixed carbon content data, calorific value data, carbon structure information, and characteristic data and distribution data of aromatic ring structure.
[0010] Optionally, the training rules of the above coal quality detection model include: The coal quality data set in the coal quality detection and analysis atlas database is divided into a training set, a validation set and a test set according to a preset ratio; wherein each coal quality data set contains coal quality characteristic information of historical coal samples and corresponding coal quality detection results; The initial model is trained using the training set, and during the training process, the model parameters of the initial model are adjusted, the initial model with different model parameters is verified using the verification set, and the accuracy of the initial model with different model parameters is evaluated based on the verification result and a predetermined evaluation index; wherein the evaluation index includes accuracy and / or mean square error; The initial model with the highest training accuracy is determined as the coal quality detection model, and the coal quality detection model is tested using the test set; The initial model is built based on a neural network algorithm, a random forest algorithm, or a support vector machine.
[0011] Optionally, after the coal quality data set in the coal quality detection and analysis atlas database is divided into a training set, a validation set and a test set according to a preset ratio, the method further includes: Data enhancement processing is performed on the coal quality data groups in the training set, validation set and test set respectively.
[0012] A second aspect of the present invention provides a multi-dimensional fusion coal quality detection system, comprising: A coal quality characteristic information collection module is used to collect various coal quality characteristic information of the coal sample to be tested; The coal quality detection module is used to input various coal quality feature information into the pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; wherein, The coal quality detection model is obtained by training the initial model based on the coal quality detection and analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection and analysis atlas database. The coal quality detection and analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal samples and the manual test results corresponding to the historical coal samples. The initial model is used to predict the coal quality information of the historical coal samples.
[0013] In a third aspect of the present invention, a machine-readable storage medium is provided, on which instructions are stored, and when the instructions are executed by a processor, the processor is configured to execute the multi-dimensional fusion coal quality detection method.
[0014] In a fourth aspect of the present invention, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the multi-dimensional fusion coal quality detection method is implemented.
[0015] Through the above technical scheme, a multi-dimensional fusion coal quality detection method and system are provided to input the collected multiple coal quality feature information of the coal sample to be detected into the pre-trained coal quality detection model for coal quality detection, and obtain the coal quality detection result. The coal quality detection model is obtained by training the initial model based on the coal quality detection analysis atlas database, and the coal quality detection model can also iteratively correct the coal quality detection model according to the coal quality detection result and the coal quality detection analysis atlas database to improve the accuracy of coal quality detection model recognition and optimize the robustness of the coal quality detection model. The coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal sample and the manual test results corresponding to the historical coal sample, thereby providing a gold standard data set for the training of the coal quality detection model. The method and system not only realize the intelligent online accurate detection of coal quality and coal quantity to grasp and predict the purpose of comprehensive coal quality and coal quantity information, but also construct a gold standard data set (i.e., coal quality detection analysis atlas database) for the training of the coal quality detection model, and the coal quality detection model of the method and system realizes self-learning to further achieve the purpose of high-precision multi-dimensional fusion coal quality detection.
[0016] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings: Figure 1 It is a flow chart of a multi-dimensional fusion coal quality detection method provided by one embodiment of the present invention; Figure 2 It is a schematic diagram of an implementation of a multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention; Figure 3 It is a schematic diagram of another multi-dimensional fusion coal quality detection method provided by one embodiment of the present invention; Figure 4 It is a schematic diagram of another implementation of a multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention; Figure 5 It is a block diagram of a multi-dimensional fusion coal quality detection system provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0018] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.
[0019] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.
[0020] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.
[0021] Example 1 Figure 1 FIG. 1 is a flow chart of a multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a multi-dimensional fusion coal quality detection method, comprising: S110: Collecting various coal quality characteristic information of the coal sample to be tested; In some implementations of the present embodiment, the above-mentioned coal quality characteristic information includes coal sample image information; the above-mentioned collection of multiple coal quality characteristic information of the coal sample to be tested includes: controlling the 3D scanner to collect external characteristics of the coal sample to be tested, and obtaining coal sample image information of the coal sample to be tested; wherein the coal sample image information of the coal sample to be tested includes three-dimensional shape data, external contour data, volume data, surface area data, spatial distribution arrangement data of coal particles, pore characteristic data, crack characteristic data, particle size data of coal particles and shape factor parameters of coal particles.
[0022] Specifically, the method controls a 3D scanner to collect external features of the coal sample to be tested, and obtains the coal sample image information of the coal sample to be tested. Among them, the three-dimensional shape data and external contour data include the overall shape, concave-convex features, irregular parts, etc. The volume data and surface area data represent the accurately calculated volume size and surface area of the coal sample. The spatial distribution arrangement data of coal particles represent the distribution and mutual relationship of coal particles in space. The pore characteristic data and crack characteristic data represent the position, size and shape of the pores, cracks and other defects detected inside the coal. The particle size data of coal particles and the shape factor parameters of coal particles are used to represent the particle size and shape factor (such as roundness, aspect ratio, etc.) of coal particles obtained by analyzing the three-dimensional model (the three-dimensional model can be constructed based on the three-dimensional shape data, external contour data, volume data, surface area data, spatial distribution arrangement data of coal particles, pore characteristic data and crack characteristic data).
[0023] In some implementations of the present embodiment, the above-mentioned coal quality characteristic information includes physical information of the coal sample; the above-mentioned collection of multiple coal quality characteristic information of the coal sample to be tested includes: controlling the particle size analyzer to perform particle size analysis on the coal sample to be tested, and obtaining the coal sample physical information of the coal sample to be tested; wherein the coal sample physical information of the coal sample to be tested includes particle size distribution curve data, average particle size data, particle size range data, specific surface area data and preset particle size content data of the coal sample to be tested.
[0024] Among them, the particle size distribution curve data is used to intuitively display the proportion of coal particles in different particle size ranges. The average particle size data includes the arithmetic mean particle size, median diameter (D50), etc., which are used to reflect the overall size characteristics of coal particles. The particle size range data is used to indicate the maximum and minimum values of the coal particle size. The specific surface area data is used to reflect the proportional relationship between the surface area and volume of coal particles. The content data of the preset particle size (such as D10, D90, etc.) indicates that the cumulative proportion of particles smaller or larger than the particle size reaches 10% or 90%.
[0025] In some implementations of the present embodiment, the above-mentioned coal quality characteristic information includes coal sample chemical information; the above-mentioned collection of multiple coal quality characteristic information of the coal sample to be tested includes: sequentially controlling the laser induced breakdown spectrometer, the near-infrared spectrometer, the X-ray fluorescence spectrometer and the Raman spectrometer to collect chemical characteristics of the coal sample to be tested, and obtaining the coal sample chemical information of the coal sample to be tested; wherein the coal sample chemical information of the coal sample to be tested includes content data of preset trace elements, moisture content data in coal, volatile matter content data, fixed carbon content data, calorific value data, carbon structure information, and characteristic data and distribution data of the aromatic ring structure.
[0026] Specifically, the area for collecting the chemical information of the coal sample to be tested collects and analyzes the chemical characteristics of the coal sample to be tested in turn through LIBS (laser induced breakdown spectrometer), NIRS (near infrared spectrometer), XRF (X-ray fluorescence spectrometer) and Ramanm (Raman spectrometer) to obtain the chemical information of the coal sample to be tested. The chemical information of the coal sample to be tested includes the precise content of the main trace elements, qualitative and quantitative analysis of elemental composition, moisture content, volatile content, fixed carbon content, calorific value prediction information, preliminary inference results of the chemical form and valence state of the elements, microcrystalline structure information of carbon (such as the degree of graphitization and order), and the characteristics and distribution of aromatic ring structure.
[0027] S120: Input a variety of coal quality characteristic information into a pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; wherein, the coal quality detection model is obtained by training an initial model based on a coal quality detection analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection analysis atlas database, the coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal samples and the manual test results corresponding to the historical coal samples, and the initial model is used to predict the coal quality information of the historical coal samples.
[0028] Specifically, the coal quality detection model has two modes: learning mode and detection mode. When the coal quality detection model runs in learning mode, the collected information (including the prediction results of the initial model corresponding to the historical coal samples) is used to construct a data set and compared with the manual test results to form a gold standard data set (i.e., a coal quality detection and analysis atlas database) for training the coal quality detection model. When the coal quality detection model runs in detection mode, a variety of coal quality feature information is input into the pre-trained coal quality detection model for comprehensive coal quality analysis to obtain high-precision, multi-dimensional coal quality detection results. And when the coal quality detection model runs in learning mode, the coal quality detection model can also be iteratively corrected according to the coal quality detection results and the coal quality detection analysis atlas database to improve the accuracy of coal quality detection model recognition and optimize the robustness of the coal quality detection model.
[0029] In the above implementation process, the method inputs the collected coal quality characteristic information of the coal sample to be tested into the pre-trained coal quality detection model for coal quality detection to obtain the coal quality detection result. The coal quality detection model is obtained by training the initial model based on the coal quality detection analysis atlas database, and the coal quality detection model can also iteratively correct the coal quality detection model according to the coal quality detection result and the coal quality detection analysis atlas database to improve the accuracy of coal quality detection model recognition and optimize the robustness of the coal quality detection model. The coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal sample and the manual test results corresponding to the historical coal sample, thereby providing a gold standard data set for the training of the coal quality detection model. The method not only realizes the intelligent online accurate detection of coal quality and coal quantity to grasp and predict the purpose of comprehensive coal quality and coal quantity information, but also constructs a gold standard data set (i.e., coal quality detection analysis atlas database) for the training of the coal quality detection model, and the coal quality detection model of the method realizes self-learning to further achieve the purpose of high-precision multi-dimensional fusion coal quality detection.
[0030] In some implementations of this embodiment, before collecting multiple coal quality characteristic information of the coal sample to be tested, the method also includes: controlling the sampling device to transmit the coal sample to be tested, and pre-processing the coal sample to be tested transmitted by the sampling device; wherein the pre-processing process includes: controlling the agitator to mix the coal sample to be tested, and controlling the pressure roller to shape the mixed coal sample to be tested.
[0031] Specifically, after the coal sample to be tested arrives at the pretreatment area, the coal sample is mixed by a stirrer, and after the mixing is completed, the coal sample to be tested is shaped by a pressing roller.
[0032] Please refer to Figure 2 , Figure 2It is an implementation diagram of a multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention. First, the operator sends the coal sample to be tested into the sample setting area. The sample setting area has a flat and smooth container. The operator pours the coal sample to be tested into the container, and the container is transported to the pretreatment area by a belt. From the sample setting area, the coal sample to be tested is transported to the pretreatment area, 3D scanner, particle size analyzer, LIBS (laser induced breakdown spectrometer), NIRS (near infrared spectrometer), XRF (X-ray fluorescence spectrometer), Ramanm (Raman spectrometer) in sequence by belt transportation, and the collected coal sample image information, coal sample physical information and coal sample chemical information of the coal sample to be tested are sent to the data analysis platform for comprehensive processing, and then the remaining coal sample to be tested is sent to the discard port by belt transportation. The data analysis platform comprehensively processes the detection results of these devices (including the coal sample image information, coal sample physical information and coal sample chemical information of the coal sample to be tested), comprehensively analyzes the chemical composition, structural characteristics and chemical bonds of the coal, so as to more deeply evaluate the quality, reactivity and potential application value of the coal. The content of main elements can be obtained through LIBS (laser induced breakdown spectrometry), and combined with the detection of trace elements by XRF (X-ray fluorescence spectrometer), the purity and potential contaminants of coal can be more comprehensively evaluated; NIRS (near infrared spectrometer) predicts the combustion performance of coal through analysis of moisture and organic functional groups; the analysis results of carbon structure by Ramanm (Raman spectrometer) can reflect the degree of metamorphism and chemical stability of coal.
[0033] The sampling area includes transmission belts, sample carrying rotating devices and other equipment, which are mainly used to transmit coal samples to be tested. The sampling area is mainly used to collect coal samples to be tested, and send the coal samples to be tested to various testing devices (including 3D scanners, particle size analyzers, LIBS (laser induced breakdown spectrometers), NIRS (near infrared spectrometers), XRF (X-ray fluorescence spectrometers) and Ramanm (Raman spectrometers)) through belt transportation, and collect the data collected by various testing devices (including coal sample image information, coal sample physical information and coal sample chemical information of the coal samples to be tested).
[0034] Please refer to Figure 4 , Figure 4It is an implementation diagram of another multidimensional fusion coal quality detection method provided by an embodiment of the present invention. In some implementations of this embodiment, the training rules of the above-mentioned coal quality detection model include: dividing the coal quality data group in the coal quality detection analysis atlas database into a training set, a verification set and a test set according to a preset ratio; wherein each coal quality data group contains coal quality characteristic information of historical coal samples and corresponding coal quality detection results; using the training set to train the initial model, and in the training process, adjusting the model parameters of the initial model, using the verification set to verify the initial model with different model parameters, and evaluating the accuracy of the initial model with different model parameters based on the verification results and predetermined evaluation indicators; wherein the evaluation indicators include accuracy and / or mean square error; determining the highest accuracy of the trained initial model as the coal quality detection model, and using the test set to test the coal quality detection model; wherein the initial model is constructed based on a neural network algorithm, a random forest algorithm or a support vector machine.
[0035] In some implementations of this embodiment, after the coal quality data groups in the coal quality detection and analysis atlas database are divided into training set, verification set and test set according to a preset ratio, the method also includes: performing data enhancement processing on the coal quality data groups in the training set, verification set and test set respectively.
[0036] Specifically, when the coal quality detection model runs in learning mode, a suitable machine learning algorithm can be selected, such as neural network, random forest, support vector machine, etc. The preprocessed data is divided into training set, validation set and test set. The initial model is trained using the training set data, and the parameters of the initial model are adjusted to optimize the performance. The validation set data is used to evaluate the performance of the initial model during the training process and select the optimal model parameters (mainly including the rationality of feature selection: ① Select features that can effectively reflect the quality of coal and avoid the interference of irrelevant or redundant features on the model; ② Model complexity and computing resources: Select a model of appropriate complexity according to the actual application scenario and computing power). Data enhancement technology: If the amount of data is limited, data enhancement technology can be used to increase the diversity of data. Model evaluation indicators: Select appropriate evaluation indicators, such as accuracy, mean square error, etc., to evaluate the performance of the model. And compare the prediction results with the manual results, and finally form a gold standard set for mapping and comparing the coal sample image information, coal sample physical information, coal sample chemical information and coal quality information of the coal sample to be tested.
[0037] In some implementations of this embodiment, the coal quality detection model can obtain relevant detection data for a new coal sample and perform the same preprocessing steps as the training data. The preprocessed data is input into the trained coal quality detection model to obtain the predicted quality parameters. As new data continues to accumulate, the new data is regularly added to the training set and the coal quality detection model is retrained to improve the generalization ability and accuracy of the coal quality detection model.
[0038] Example 2 Figure 2 is a schematic diagram of a multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention, Figure 3 is a schematic diagram of another multi-dimensional fusion coal quality detection method provided by an embodiment of the present invention, such as Figure 2 and Figure 3 As shown, this embodiment provides a self-learning high-precision multi-dimensional fusion coal quality detection device, which corresponds to a multi-dimensional fusion coal quality detection method provided in Example 1. A self-learning high-precision multi-dimensional fusion coal quality detection device includes a sampling area, a detection area and a data analysis platform. The sampling area includes a transmission belt, a sample carrying rotating device and other equipment, which are mainly used to transmit the coal sample to be detected. The detection area collects coal sample image information, coal sample physical information and coal sample chemical information through a series of instruments, and then sends the coal sample image information, coal sample physical information and coal sample chemical information to the computer for analysis and obtains the final detection result. The instrument includes a near-infrared light source component, an X-ray fluorescence light source component, a Raman light source component, an optical energy and distribution detection device, a particle size analyzer, an optical path system (including beryllium window, lens, etc.), an optical vibration isolation platform and SSD and other detectors. The data analysis platform includes computer equipment and simulation analysis software for comprehensive analysis of data, training artificial intelligence models, and detecting new data. The self-learning high-precision multi-dimensional fusion coal quality detection device has two working modes: learning mode and detection mode. Use the following steps to use this device: S1: First, the operator sends the coal sample to be tested into the sample setting area. The sample setting area has a flat and smooth container. The operator pours the coal sample to be tested into the container, and the container is transported to the pretreatment area by a belt; S2: After the coal sample to be tested arrives at the pretreatment area, the coal sample is mixed by a stirrer, and after the mixing is completed, the coal sample to be tested is shaped by a roller; S3: The shaped coal sample is passed through a 3D scanner to obtain coal sample image information; wherein the coal sample image information includes the following data: ① Three-dimensional shape and external contour: including overall shape, concave and convex features, irregular parts, etc.; ② Volume and surface area data: can accurately calculate the volume and surface area of coal samples; ③ Spatial distribution and arrangement of particles: understand the distribution and relationship of coal particles in space; ④ Characteristics of pores and defects: The location, size and shape of pores, cracks and other defects inside the coal are detected; ⑤ Particle size and shape parameters: Obtain the particle size and shape factors (such as roundness, aspect ratio, etc.) of coal particles through analysis of the three-dimensional model; The coal sample image information is sent to the data analysis platform, and then the coal sample is sent to the particle size analyzer via belt transport.
[0039] S4: Physical information of the coal sample is obtained through a particle size analyzer, wherein the physical information of the coal sample includes the following data: ① Particle size distribution curve: intuitively displays the proportion of coal particles in different particle size ranges; ② Average particle size: including arithmetic mean particle size, median diameter (D50), etc., reflecting the overall size characteristics of coal particles; ③ Content of specific particle size: such as D10, D90, etc., indicating that the cumulative proportion of particles smaller or larger than this particle size reaches 10% or 90%; ④ Particle size range: gives the maximum and minimum values of coal particle size; ⑤ Specific surface area: reflects the ratio of the surface area to volume of coal particles; The physical information of the coal sample is transmitted to the data analysis platform, and then the coal sample is transported to the next area via a belt.
[0040] S5: This area is mainly analyzed by LIBS, NIRS, XRF, and Ramanm in sequence to obtain the chemical information of coal samples, among which the chemical information of coal samples includes: ① Accurate content of main trace elements, qualitative and quantitative analysis of element composition; ② Prediction information of moisture content, volatile matter content, fixed carbon content and calorific value in coal; ③ Preliminary inference of the chemical form and valence state of elements; ④ Carbon microstructure information: such as graphitization degree and order; ⑤ Characteristics and distribution of aromatic ring structures; The chemical information of the coal sample is transmitted to the data analysis platform for comprehensive processing, and then the remaining coal sample is sent to the waste outlet.
[0041] S6: The data analysis platform comprehensively processes the test results of these devices and comprehensively analyzes the chemical composition, structural characteristics and chemical bonds of coal, so as to more deeply evaluate the quality, reactivity and potential application value of coal. The main element content is known through LIBS (laser induced breakdown spectrometer), and combined with the detection of trace elements by XRF (X-ray fluorescence spectrometer), the purity and potential contaminants of coal can be more comprehensively evaluated; NIRS (near infrared spectrometer) predicts the combustion performance of coal through the analysis of moisture and organic functional groups; the analysis results of carbon structure by Ramanm (Raman spectrometer) can reflect the degree of coal deterioration and chemical stability.
[0042] S7: The data analysis platform combines coal sample image information, coal sample physical information and coal sample chemical information to form a coal spectrum database. The model prediction results are compared with the manual test results to form a gold standard data set, which can be used for the training of artificial intelligence models.
[0043] S8: In the "learning" mode, the data analysis platform uses the model prediction results to correct the model through the gold standard data set to improve the accuracy of the model and optimize the robustness of the model. In the "detection" mode, the data analysis platform uses computer software to comprehensively analyze the collected information and combine it with the trained artificial intelligence model prediction to obtain high-precision, multi-dimensional coal quality detection results.
[0044] In the above implementation process, a self-learning high-precision multi-dimensional fusion coal quality detection equipment can solve the needs of coal industry application scenarios for comprehensive analysis of coal quality and quantity technology, as well as the difficulties in technical application. It is of great significance to the development of comprehensive analysis of coal quality and quantity, and can effectively help coal enterprises to detect and control coal quality and quantity information in all links of coal production, transportation, processing and utilization in real time, and use effective information for industrial guidance.
[0045] In some implementations of this embodiment, the detection area is composed of a series of instruments for collecting coal sample image information, physical information and chemical information.
[0046] In some implementations of this embodiment, the optical seismic isolation platform is mainly used to prevent the detection device from being affected by vibrations from other devices during operation; In some implementations of this embodiment, the data analysis platform has two modes, one is a "learning" mode and the other is a "detection" mode, and has data collection and preprocessing capabilities, and can perform model updating and optimization.
[0047] Example 3 Figure 5 FIG. 1 is a block diagram of a multi-dimensional fusion coal quality detection system provided by an embodiment of the present invention. Figure 5 As shown, an embodiment of the present invention provides a multi-dimensional fusion coal quality detection system, comprising: A coal quality characteristic information collection module is used to collect various coal quality characteristic information of the coal sample to be tested; The coal quality detection module is used to input various coal quality feature information into a pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; wherein, the coal quality detection model is obtained by training the initial model based on the coal quality detection analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection analysis atlas database, the coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal samples and the manual test results corresponding to the historical coal samples, and the initial model is used to predict the coal quality information of the historical coal samples.
[0048] Specifically, the system inputs the collected coal quality characteristic information of the coal sample to be tested into the pre-trained coal quality detection model to perform coal quality detection and obtain the coal quality detection results. The coal quality detection model is obtained by training the initial model based on the coal quality detection analysis atlas database, and the coal quality detection model can also iteratively correct the coal quality detection model according to the coal quality detection results and the coal quality detection analysis atlas database to improve the accuracy of coal quality detection model recognition and optimize the robustness of the coal quality detection model. The coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to the historical coal sample and the manual test results corresponding to the historical coal sample, thereby providing a gold standard data set for the training of the coal quality detection model. The system not only realizes the intelligent online accurate detection of coal quality and coal quantity to grasp and predict the purpose of comprehensive coal quality and coal quantity information, but also constructs a gold standard data set (i.e., coal quality detection analysis atlas database) for the training of the coal quality detection model, and the coal quality detection model of the system realizes self-learning to further achieve the purpose of high-precision multi-dimensional fusion coal quality detection.
[0049] The optional implementation modes of the embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation modes. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical scheme of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.
[0050] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.
[0051] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions to enable a single-chip microcomputer, a chip or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0052] In addition, various implementation modes of the embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed by the embodiments of the present invention.
Claims
1. A multi-dimensional fusion coal quality detection method, It is characterized in that include: Collect various coal quality characteristic information of the coal sample to be tested; Input various coal quality characteristic information into the pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; The coal quality detection model is obtained by training an initial model based on a coal quality detection analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection analysis atlas database. The coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to historical coal samples and the manual test results corresponding to the historical coal samples. The initial model is used to predict the coal quality information of historical coal samples.
2. The multi-dimensional fusion coal quality detection method according to claim 1, It is characterized in that Before collecting a plurality of coal quality characteristic information of the coal sample to be tested, the method further comprises: Control the sampling device to transmit the coal sample to be tested, and pre-process the coal sample to be tested transmitted by the sampling device; wherein, The pre-treatment process includes: The agitator is controlled to mix the coal sample to be tested, and the pressing roller is controlled to shape the mixed coal sample to be tested.
3. The multi-dimensional fusion coal quality detection method according to claim 1, It is characterized in that The coal quality characteristic information includes coal sample image information; The multiple coal quality characteristic information of the coal sample to be tested is collected, including: Control the 3D scanner to collect external features of the coal sample to be tested, and obtain the coal sample image information of the coal sample to be tested; wherein, The coal sample image information of the coal sample to be tested includes three-dimensional shape data, external contour data, volume data, surface area data, spatial distribution arrangement data of coal particles, pore characteristic data, crack characteristic data, particle size data of coal particles and shape factor parameters of coal particles.
4. The multi-dimensional fusion coal quality detection method according to claim 1, It is characterized in that The coal quality characteristic information includes physical information of the coal sample; The multiple coal quality characteristic information of the coal sample to be tested is collected, including: Control the particle size analyzer to perform particle size analysis on the coal sample to be tested, and obtain the physical information of the coal sample to be tested; wherein, The coal sample physical information of the coal sample to be tested includes particle size distribution curve data, average particle size data, particle size range data, specific surface area data and content data of a preset particle size of the coal sample to be tested.
5. The multi-dimensional fusion coal quality detection method according to claim 1, It is characterized in that The coal quality characteristic information includes coal sample chemical information; The multiple coal quality characteristic information of the coal sample to be tested is collected, including: The laser induced breakdown spectrometer, near infrared spectrometer, X-ray fluorescence spectrometer and Raman spectrometer are sequentially controlled to collect chemical characteristics of the coal sample to be tested, so as to obtain the chemical information of the coal sample to be tested; wherein, The coal sample chemical information of the coal sample to be tested includes content data of preset trace elements, moisture content data in coal, volatile content data, fixed carbon content data, calorific value data, carbon structure information, and characteristic data and distribution data of aromatic ring structure.
6. The multi-dimensional fusion coal quality detection method according to claim 1, It is characterized in that The training rules of the coal quality detection model include: The coal quality data set in the coal quality detection and analysis atlas database is divided into a training set, a validation set and a test set according to a preset ratio; wherein each coal quality data set contains coal quality characteristic information of historical coal samples and corresponding coal quality detection results; The initial model is trained using the training set, and during the training process, the model parameters of the initial model are adjusted, the initial model with different model parameters is verified using the verification set, and the accuracy of the initial model with different model parameters is evaluated based on the verification result and a predetermined evaluation index; wherein the evaluation index includes accuracy and / or mean square error; The initial model with the highest training accuracy is determined as the coal quality detection model, and the coal quality detection model is tested using the test set; The initial model is constructed based on a neural network algorithm, a random forest algorithm or a support vector machine.
7. The multi-dimensional fusion coal quality detection method according to claim 6, It is characterized in that After the coal quality data set in the coal quality detection and analysis atlas database is divided into a training set, a validation set and a test set according to a preset ratio, the method further includes: Data enhancement processing is performed on the coal quality data groups in the training set, validation set and test set respectively.
8. A multi-dimensional coal quality detection system, It is characterized in that include: A coal quality characteristic information collection module is used to collect various coal quality characteristic information of the coal sample to be tested; The coal quality detection module is used to input various coal quality feature information into the pre-trained coal quality detection model to perform coal quality detection and obtain coal quality detection results; wherein, The coal quality detection model is obtained by training an initial model based on a coal quality detection analysis atlas database, and the model iteration is maintained in real time based on the latest coal quality detection results and the coal quality detection analysis atlas database. The coal quality detection analysis atlas database is obtained by comparing and verifying the prediction results of the initial model corresponding to historical coal samples and the manual test results corresponding to the historical coal samples. The initial model is used to predict the coal quality information of historical coal samples.
9. A machine-readable storage medium having instructions stored thereon, It is characterized in that When the instruction is executed by a processor, the processor is configured to execute the multi-dimensional fusion coal quality detection method as described in any one of claims 1 to 7.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the multi-dimensional fusion coal quality detection method described in any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Coal quality on-line analysis system and method based on incoming factory
CN116297273A
Coal quality evaluation method, system, equipment and medium
CN117074643A
Coal quality analysis model training method and device, electronic equipment and storage medium
CN117216558A
Coal quality detection method and device, electronic equipment and storage medium
CN118225752A
Cited By
Spectrum pretreatment method for online detection of coal quality through multi-spectrum combination and related equipment
CN120427566A
Fingerprint tracing method based on coal consumption process
CN120853744A
A fingerprinting method based on coal consumption process
CN120853744B
Coal quality multi-source data anomaly detection method and device, electronic equipment and storage medium
CN121030591A
Interlayer gold purity evaluation method and device based on hybrid model, equipment and medium
CN121502310A
