A method, apparatus, equipment and medium for coal quality detection based on infrared detection

Abstract

This application discloses a method, apparatus, equipment, and medium for coal quality testing based on infrared detection. The method includes: identifying key characteristic absorption peaks in coal based on its infrared spectral information, performing cluster analysis to generate proportional information of different components in the coal; obtaining a coal quality grade assessment result using the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm based on the proportional information; performing anomaly detection on low-quality coal in the quality grade assessment result, locating the key characteristic absorption peaks of the anomalies from the infrared spectral information, and assessing the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference, generating a preliminary report; predicting abnormal substance information using a pre-set backpropagation neural network based on the preliminary report, and determining the source direction and distance of the abnormal substances by combining a geographic information system, ultimately generating a coal quality testing report. This application can improve the efficiency and accuracy of coal quality testing.

Description

Technical Field

[0001] This application relates to the field of coal quality testing technology, and in particular to a coal quality testing method, apparatus, equipment and medium based on infrared detection. Background Technology

[0002] As an important energy resource, coal quality testing is crucial during its mining, processing, and use. Existing coal quality testing methods primarily rely on chemical analysis and physical tests, such as elemental analysis, ash content determination, and volatile matter determination. By measuring the content of elements such as sulfur, nitrogen, and oxygen, as well as indicators like ash and volatile matter, detailed information about the coal's composition can be provided. Automated equipment allows for the automatic sampling, preparation, and testing of coal samples on a production line, improving testing efficiency and data consistency.

[0003] However, existing coal quality testing methods have the following shortcomings: traditional methods require a long time to complete a full testing cycle, which is unacceptable in large-scale production environments; due to a lack of in-depth understanding of the internal structure of coal, traditional testing methods may give inaccurate or misleading results in some cases; traditional methods have difficulty effectively identifying and quantifying abnormal substances in coal, especially when the content of these substances is extremely low or their properties are very close to normal components; existing methods are difficult to provide information on the source of abnormal substances, which makes pollution control and accountability difficult. Summary of the Invention

[0004] To address the above technical issues, this application provides a method, apparatus, equipment, and medium for coal quality detection based on infrared detection, which can improve the efficiency and accuracy of coal quality detection.

[0005] This application provides a coal quality detection method based on infrared detection, including: Obtain infrared spectral information of coal; Based on the infrared spectral information, key characteristic absorption peaks of organic components and minerals in coal are identified, and cluster analysis is performed on the key characteristic absorption peaks to generate information on the proportion of different components in coal. Based on the aforementioned proportion information, the coal quality is assessed using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm to obtain the coal quality grade assessment results. Anomaly detection is performed on low-quality coal in the quality grade assessment results. The key characteristic absorption peaks of the anomaly are located from the infrared spectral information. Based on Bayesian inference, the probability and uncertainty of the presence of abnormal substances in the coal are assessed to generate a preliminary report. Based on the preliminary report, a pre-set backpropagation neural network is used to predict abnormal material information, and combined with a geographic information system, the source direction and distance of the abnormal material are determined, and finally a coal quality inspection report is generated.

[0006] As an improvement to the above scheme, the step of identifying key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, and performing cluster analysis on the key characteristic absorption peaks to generate proportional information of different components in coal includes: Based on the infrared spectral information, an adaptive wavelength selection algorithm is used to dynamically adjust the wavelength selection window of the infrared spectral signal to obtain candidate key characteristic absorption peaks with typical infrared absorption characteristics representing different components in the coal. Based on the candidate key feature absorption peaks, key feature absorption peaks with similar absorption characteristics are grouped into one category by machine learning clustering analysis to obtain the classified key feature absorption peaks. Based on the key characteristic absorption peaks after classification, Fourier transform and inverse transform are used to calculate and analyze the area ratio of the key characteristic absorption peaks to generate the proportion information of different components in coal.

[0007] As an improvement to the above scheme, the step of using the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm to assess coal quality based on the proportional information, and obtaining the coal quality grade assessment result, includes: Based on the aforementioned proportion information, the importance of different components to coal quality assessment is evaluated to generate importance assessment results for different components; Based on the importance assessment results, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the judgment matrix is ​​filled with historical data analysis to generate a judgment matrix that reflects the relative importance among different components. Based on the judgment matrix, the weights of different components are applied to the fuzzy comprehensive evaluation algorithm to construct a standard coal quality assessment model. The coal quality is assessed based on the standard coal quality assessment model to obtain the coal quality grade assessment result.

[0008] As an improvement to the above scheme, the method involves performing anomaly detection on low-quality coal in the quality grade assessment results, locating key characteristic absorption peaks of the anomalies from the infrared spectral information, and assessing the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report, including: For coal with low quality grades in the quality grade assessment results, anomaly detection is performed on its infrared spectral information to identify and locate abnormal key characteristic absorption peaks that deviate from the normal component ratio. The key characteristic absorption peaks of the anomaly are analyzed to determine the differences between the key characteristic absorption peaks of the anomaly and the known coal composition characteristics, and the analysis results of the key characteristic anomaly are generated. Based on the analysis results of the aforementioned key anomalies, the probability of the existence of the anomalous substance and the uncertainty assessment results are determined by Bayesian inference. Based on the probability and uncertainty assessment results of the presence of the anomalous substance, a preliminary report on the probability of the presence of the anomalous substance is generated. The preliminary report includes the probability of the presence of the anomalous substance, the uncertainty assessment, the location information of the key characteristic absorption peak of the anomalous substance, and the comparison with the composition of standard coal.

[0009] As an improvement to the above scheme, the step of determining the probability of the existence of anomalous substances and the uncertainty assessment results based on the analysis results of the anomalous key features through Bayesian inference includes: Based on the analysis results of the aforementioned key anomalies, a description of the differences between the key anomalies and the characteristics of standard coal composition is extracted; Based on the description of the differences, a prior probability distribution of the existence of the anomalous substance is constructed. Based on the position and intensity information of the absorption peaks of the key anomalous features in the infrared spectrum, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate a posterior probability distribution of the existence of the anomalous substance. The probability of the presence of the anomalous substance is determined based on the posterior probability distribution, and the uncertainty of the probability of the presence of the anomalous substance is analyzed based on the width of the posterior probability distribution to generate an uncertainty assessment result.

[0010] As an improvement to the above scheme, the step of constructing a prior probability distribution of the existence of anomalous substances based on the difference description, and using Bayesian inference to analyze the probability of the existence of anomalous substances based on the position and intensity information of the absorption peaks of key anomalous features in infrared spectral information, to generate a posterior probability distribution of the existence of anomalous substances, includes: Based on the description of the differences, the differences between the compositional characteristics of the anomalous substance and the compositional characteristics of standard coal are extracted to obtain the key characteristics reflecting the existence of the anomalous substance. Based on the key features and the historical situation of the existence of the anomalous substance, the probability of the existence of the anomalous substance is estimated in order to construct the prior probability distribution of the existence of the anomalous substance. Based on the prior probability distribution, key evidence reflecting the characteristics of the anomalous substance is obtained according to the position and intensity information of the abnormal key feature absorption peaks in the infrared spectral information. Based on the prior probability distribution and the key evidence, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate the posterior probability distribution of the existence of the anomalous substance.

[0011] As an improvement to the above solution, based on the preliminary report, a preset backpropagation neural network is used to predict anomalous substance information, and combined with a geographic information system, the source direction and distance of the anomalous substance are determined, ultimately generating a coal quality inspection report, including: The data in the preliminary report were configured and processed using a backpropagation neural network to obtain a backpropagation neural network model; The preliminary report is analyzed and predicted based on the backpropagation neural network model, and specific information about the anomalous substance is generated, including elemental composition and molecular structure. Based on the specific information of the anomalous substance, and in conjunction with a geographic information system, the distribution of the pollution sources of the anomalous substance is analyzed to determine the source direction and distance of the anomalous substance. Based on the source direction and distance of the abnormal substance, the quality grade information of the coal is integrated to finally generate a coal quality inspection report that includes the quality grade of the coal and specific information about the abnormal substance.

[0012] This application also provides a coal quality detection device based on infrared detection, comprising: The data acquisition module is used to acquire the infrared spectral information of coal. The information identification module is used to identify key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, perform cluster analysis on the key characteristic absorption peaks, and generate proportional information of different components in coal. The quality assessment module is used to assess coal quality based on the ratio information using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm, and to obtain the coal quality grade assessment result. An anomaly detection module is used to detect anomalies in low-quality coal in the quality grade assessment results, locate the key characteristic absorption peaks of the anomalies from the infrared spectral information, and assess the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report. The quality inspection module is used to predict abnormal material information based on the preliminary report using a preset backpropagation neural network, and in conjunction with a geographic information system, to determine the source direction and distance of the abnormal material, and finally generate a coal quality inspection report.

[0013] This application also provides a computer device, including a processor and a memory, wherein the memory stores a computer program and the computer program is configured to be executed by the processor, wherein the processor executes the computer program to implement the coal quality detection method based on infrared detection described in any of the above claims.

[0014] This application also provides a computer-readable storage medium storing a computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the coal quality detection method based on infrared detection described above.

[0015] Compared to existing technologies, the beneficial effects of the infrared detection-based coal quality detection method, apparatus, equipment, and medium provided in this application are as follows: By acquiring the infrared spectral signal of coal, an adaptive wavelength selection algorithm is used to identify key characteristic absorption peaks, and machine learning cluster analysis is combined to calculate the area ratio, generating an information table reflecting the proportion of coal components. A quality assessment model constructed based on the analytic hierarchy process and fuzzy comprehensive evaluation algorithm, optimized with historical data, ensures the accuracy and reliability of the quality grade assessment results. For coal judged to be of low quality grade, anomaly detection algorithms are used to locate the key abnormal characteristic absorption peaks in the infrared spectral signal, and Bayesian inference is used to assess the probability and uncertainty of the presence of abnormal substances. The generated preliminary report records in detail the probability of the presence of abnormal substances, uncertainty assessment, location information of characteristic absorption peaks, and comparison with standard coal components, providing a scientific basis for a deeper understanding of the root causes of low-quality coal problems. By analyzing the key feature set of anomalies, the differences between the anomaly features and the composition features of standard coal are extracted. A prior probability distribution of the existence of anomalous substances is constructed, and a posterior probability distribution is generated using Bayesian inference. The possibility and uncertainty of the existence of anomalous substances are quantified and evaluated. This not only accurately quantifies the probability of the existence of anomalous substances, but also assesses its reliability by analyzing the width of the posterior probability distribution, thereby enhancing the credibility of coal quality detection results and decision support capabilities. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of a coal quality detection method based on infrared detection provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a coal quality detection device based on infrared detection provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0018] Please see Figure 1 , Figure 1 This is a schematic flowchart of a coal quality detection method based on infrared detection provided in an embodiment of this application. The coal quality detection method based on infrared detection includes: S1: Obtain infrared spectral information of coal; S2: Based on the infrared spectral information, identify the key characteristic absorption peaks of organic components and minerals in coal, perform cluster analysis on the key characteristic absorption peaks, and generate the proportion information of different components in coal; S3: Based on the aforementioned proportion information, the coal quality is assessed using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm to obtain the coal quality grade assessment results; S4: Perform anomaly detection on the low-quality coal in the quality grade assessment results, locate the key characteristic absorption peaks of the anomaly from the infrared spectral information, and assess the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report. S5: Based on the preliminary report, a pre-set backpropagation neural network is used to predict abnormal material information, and combined with a geographic information system, the source direction and distance of the abnormal material are determined, and finally a coal quality inspection report is generated.

[0019] Specifically, in step S1, a high-resolution Fourier transform infrared spectrometer is used to perform a non-contact scan of the coal, enabling the acquisition of infrared spectral information over a wide frequency range. The infrared spectrometer emits infrared light and measures the changes in light intensity reflected or transmitted by the coal to obtain absorption at different wavelengths. These data reflect changes in the vibrational modes of chemical bonds in the coal and are used to identify the material composition. Consistency processing ensures the comparability of infrared spectral signals from different batches or sources, generating standardized infrared spectral signals and providing a reliable data foundation for subsequent analysis.

[0020] In step S2, an adaptive wavelength selection algorithm is used to identify key characteristic absorption peaks of organic components and minerals in coal, including aromatic hydrocarbons, aliphatic hydrocarbons, and minerals such as silicates. These absorption peaks correspond to the vibrational frequencies of specific chemical bonds and are important indicators for identifying the composition of substances. Machine learning cluster analysis is used to classify similar absorption peaks and calculate the proportional relationships of each component. The area ratio of key characteristic absorption peaks can reflect the relative content of different components in coal, forming an information table that provides a quantitative basis for quality assessment.

[0021] In step S3, a standard coal quality assessment model constructed using the Analytic Hierarchy Process (AHP) and fuzzy comprehensive evaluation algorithm is used to evaluate the coal's quality grade by comparing historical data with standard reference values. The AHP is used to determine the importance weights of each component proportion, while fuzzy comprehensive evaluation handles uncertainty and subjective factors, ensuring the comprehensiveness and objectivity of the assessment results. In one example, the generated coal quality grade assessment results show that the coal contains a high proportion of aromatic hydrocarbons, but its mineral content is slightly higher than the standard for high-quality coal, ultimately being rated as a medium-quality grade. This assessment result provides important decision-making reference for the procurement department.

[0022] In step S4, for coal classified as low-quality in the quality grade assessment results, an anomaly detection algorithm is used to locate the key characteristic absorption peaks of the anomaly in the infrared spectral signal. Bayesian inference is used to assess the probability and uncertainty of the presence of the anomalous substance, generating a preliminary report on the probability of the anomalous substance's presence. This step helps to further understand the causes of low-quality coal and provides guidance for subsequent processing.

[0023] In step S5, a backpropagation neural network is used to predict specific information about the anomalous substance, including its type and concentration. Combined with a geographic information system (GIS), the source direction and distance of the anomalous substance are determined, generating a complete coal quality inspection report. This process not only improves the accuracy of the inspection but also traces the source of the problem, providing support for improving procurement strategies. In one example, to further investigate the cause of low-quality coal, technicians used a backpropagation neural network to predict the specific information about the anomalous substance, confirming that the main pollutant was a rare mineral impurity. Using the GIS, technicians discovered an abandoned mine near the coal mining site, which may be one of the pollution sources. The final coal quality inspection report not only includes detailed component analysis and quality grade assessment but also clarifies the source direction and distance of the anomalous substance, providing strong evidence for the procurement department to adjust the supply chain.

[0024] This application's embodiments achieve a comprehensive and accurate assessment of coal quality. By acquiring the infrared spectral information of coal using a high-resolution infrared spectrometer and performing consistency processing to ensure data reliability, a quality assessment model constructed based on the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm, combined with historical data correction and optimization, accurately determines the coal's quality grade. A backpropagation neural network is used to predict the specific information of anomalous substances, and a geographic information system is combined to determine their source, generating a detailed coal quality inspection report. This series of steps not only improves detection accuracy but also provides a scientific basis for tracing the source of problems, effectively ensuring coal quality and supply chain management.

[0025] As one optional embodiment, the step of identifying key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, and performing cluster analysis on the key characteristic absorption peaks to generate proportional information of different components in the coal includes: Based on the infrared spectral information, an adaptive wavelength selection algorithm is used to dynamically adjust the wavelength selection window of the infrared spectral signal to obtain candidate key characteristic absorption peaks with typical infrared absorption characteristics representing different components in the coal. Based on the candidate key feature absorption peaks, key feature absorption peaks with similar absorption characteristics are grouped into one category by machine learning clustering analysis to obtain the classified key feature absorption peaks. Based on the key characteristic absorption peaks after classification, Fourier transform and inverse transform are used to calculate and analyze the area ratio of the key characteristic absorption peaks to generate the proportion information of different components in coal.

[0026] Specifically, an adaptive wavelength selection algorithm is used to dynamically adjust the wavelength selection window of the infrared spectral signal to optimize the identification accuracy of key characteristic absorption peaks of organic components and minerals in coal, thereby obtaining candidate key characteristic absorption peaks with typical infrared absorption characteristics representing different components in coal. These candidate key characteristic absorption peaks are then analyzed to remove noise interference and non-specific key characteristic absorption peaks, generating purified key characteristic absorption peaks to ensure the effectiveness and accuracy of coal component analysis. Based on the purified key characteristic absorption peaks, machine learning clustering analysis is used to classify them, grouping key characteristic absorption peaks with similar absorption characteristics into one category to distinguish different chemical components or substance types in coal, resulting in a classified key characteristic absorption peak group. Based on the classified key characteristic absorption peak group, Fourier transform and inverse transform are used to calculate and analyze the area ratio of the key characteristic absorption peaks, generating an information table reflecting the proportional relationship of different components in coal.

[0027] The adaptive wavelength selection algorithm dynamically adjusts the wavelength selection window of the infrared spectral signal, optimizing the identification accuracy of key characteristic absorption peaks of organic components and minerals in coal. The purification process removes noise interference and non-specific absorption peaks, ensuring the effectiveness and accuracy of component analysis. Machine learning clustering analysis groups similar absorption characteristics together, distinguishing different chemical components. Fourier transform and inverse transform are used to calculate the area ratio of key characteristic absorption peaks, generating an information table reflecting the proportional relationships of different components in coal, providing data support for subsequent quality assessment.

[0028] This application first utilizes an adaptive wavelength selection algorithm to dynamically adjust the wavelength selection window and identify candidate key characteristic absorption peaks with typical infrared absorption properties. Next, these candidate absorption peaks are analyzed to remove noise and non-specific absorption peaks, generating purified key characteristic absorption peaks. Then, based on the purified absorption peaks, machine learning cluster analysis is used for classification, grouping absorption peaks with similar absorption characteristics into one category. Finally, through Fourier transform and inverse transform, the area ratio of each key characteristic absorption peak is calculated, generating an information table reflecting the proportional relationships of different components in the coal, providing basic data for subsequent steps such as quality assessment, anomaly detection, and source determination.

[0029] For example, to further improve the efficiency of coal quality testing, technicians in the coal testing laboratory used an adaptive wavelength selection algorithm to optimize the wavelength selection window of the infrared spectral signal, identifying candidate key characteristic absorption peaks in a batch of coal. After analysis and processing, noise interference and non-specific absorption peaks were removed, generating purified key characteristic absorption peaks. Through machine learning cluster analysis, absorption peaks with similar absorption characteristics were grouped into one category, forming a classified key characteristic absorption peak group. Based on this, technicians calculated the area ratio of the key characteristic absorption peaks and generated a detailed information table. This information table not only provides an important basis for the quality grade assessment in step S3, but also helps identify abnormal substances in low-quality coal in step S4, and in step S5, combined with a geographic information system, determines the specific information of the abnormal substances and their source direction and distance, ensuring the comprehensiveness and accuracy of coal quality testing.

[0030] As one optional embodiment, the step of using the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm to assess coal quality based on the proportional information, and obtaining the coal quality grade assessment result, includes: Based on the aforementioned proportion information, the importance of different components to coal quality assessment is evaluated to generate importance assessment results for different components; Based on the importance assessment results, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the judgment matrix is ​​filled with historical data analysis to generate a judgment matrix that reflects the relative importance among different components. Based on the judgment matrix, the weights of different components are applied to the fuzzy comprehensive evaluation algorithm to construct a standard coal quality assessment model. The coal quality is assessed based on the standard coal quality assessment model to obtain the coal quality grade assessment result.

[0031] Specifically, based on the information table showing the proportions of different components in coal, the importance of each component to coal quality assessment is evaluated to generate importance assessment results for each component. Based on these assessment results, the Analytic Hierarchy Process (AHP) is applied to construct a judgment matrix, which is then populated using historical data analysis to generate a judgment matrix reflecting the relative importance of different components. Based on this judgment matrix, the weights of different components are applied to a fuzzy comprehensive evaluation algorithm to generate an evaluation system that includes the weights of each component. Based on this evaluation system, the weight allocation ratio in the fuzzy comprehensive evaluation algorithm is intelligently adjusted to generate a standard coal quality assessment model. Finally, based on this standard coal quality assessment model, historical data is used for correction and optimization, and the results are compared with standard reference values ​​to generate a coal quality grade assessment result.

[0032] The Analytic Hierarchy Process (AHP) is used to assess the importance of different components in coal to its quality. By constructing a judgment matrix and filling it with historical data analysis results, the relative importance of each component is quantified. The fuzzy comprehensive evaluation algorithm combines these weights to handle uncertainties and subjective factors, constructing a comprehensive quality assessment model. This model intelligently adjusts the weight configuration ratio to ensure the accuracy and reliability of the assessment results, ultimately generating a coal quality grade assessment result, providing a scientific basis for decision-making.

[0033] This application's embodiments assess the importance of each component to coal quality based on the component ratios in the information table, generating an importance assessment result. Then, an analytic hierarchy process (AHP) is applied to construct a judgment matrix, which is then filled using historical data analysis to reflect the relative importance of each component. Finally, these weights are applied to a fuzzy comprehensive evaluation algorithm, forming an evaluation system containing different component weights. Based on this evaluation system, the weight configuration ratios in the fuzzy comprehensive evaluation algorithm are intelligently adjusted to optimize the standard coal quality assessment model. Combined with historical data correction and optimization, and compared with standard reference values, a coal quality grade assessment result is generated.

[0034] For example, in a coal testing laboratory, after obtaining the component ratios from an information table, technicians assess the importance of each component to coal quality and construct a judgment matrix reflecting their relative importance. Based on this matrix, weights are applied to a fuzzy comprehensive evaluation algorithm, forming a detailed evaluation system. By intelligently adjusting the weight configuration ratios, the standard coal quality assessment model is optimized. In subsequent steps, this model helps identify a batch of low-quality coal and, combined with anomaly detection algorithms and a geographic information system, further analyzes the source direction and distance of anomalous substances. The final generated coal quality testing report not only includes detailed component analysis and quality grade assessment but also clarifies the specific information and source of anomalous substances, ensuring that supply chain adjustments are based on evidence and guaranteeing stable coal quality. This series of steps improves testing accuracy and provides reliable decision support.

[0035] As one optional embodiment, the process involves performing anomaly detection on low-quality coal in the quality grade assessment results, locating key characteristic absorption peaks of the anomalies from the infrared spectral information, and assessing the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report, including: For coal with low quality grades in the quality grade assessment results, anomaly detection is performed on its infrared spectral information to identify and locate abnormal key characteristic absorption peaks that deviate from the normal component ratio. The key characteristic absorption peaks of the anomaly are analyzed to determine the differences between the key characteristic absorption peaks of the anomaly and the known coal composition characteristics, and the analysis results of the key characteristic anomaly are generated. Based on the analysis results of the aforementioned key anomalies, the probability of the existence of the anomalous substance and the uncertainty assessment results are determined by Bayesian inference. Based on the probability and uncertainty assessment results of the presence of the anomalous substance, a preliminary report on the probability of the presence of the anomalous substance is generated. The preliminary report includes the probability of the presence of the anomalous substance, the uncertainty assessment, the location information of the key characteristic absorption peak of the anomalous substance, and the comparison with the composition of standard coal.

[0036] Specifically, based on the quality grade assessment results, coal classified as low-quality coal is analyzed using its infrared spectral signals to obtain infrared spectral signal analysis results. Based on these results, anomaly detection algorithms are used to analyze the absorption peaks in the infrared spectral signals, identifying and locating anomalous key characteristic absorption peaks that deviate from the normal component ratio, thus obtaining a set of anomalous key features. This set of anomalous key features is then analyzed to determine the differences between the anomalous key features and known coal component characteristics, generating a result for analyzing anomalous key features. Based on this result, Bayesian inference is used to analyze the probability of the presence of anomalous substances, quantifying and assessing the likelihood of their existence, resulting in a probability and uncertainty assessment of the anomalous substances. Based on the probability and uncertainty assessment of the anomalous substances, a preliminary report on their probability of existence is generated, including the probability of the anomalous substances, uncertainty assessment, location information of the anomalous key characteristic absorption peaks, and a comparison with standard coal components.

[0037] As one optional embodiment, the step of determining the probability of the presence of the anomalous substance and the uncertainty assessment result based on the analysis results of the anomalous key features through Bayesian inference includes: Based on the analysis results of the aforementioned key anomalies, a description of the differences between the key anomalies and the characteristics of standard coal composition is extracted; Based on the description of the differences, a prior probability distribution of the existence of the anomalous substance is constructed. Based on the position and intensity information of the absorption peaks of the key anomalous features in the infrared spectrum, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate a posterior probability distribution of the existence of the anomalous substance. The probability of the presence of the anomalous substance is determined based on the posterior probability distribution, and the uncertainty of the probability of the presence of the anomalous substance is analyzed based on the width of the posterior probability distribution to generate an uncertainty assessment result.

[0038] Specifically, the analysis results of key anomaly features are used to extract the differences between the key features of the anomaly and the compositional characteristics of standard coal to obtain a description of the differences. Based on the description of the differences, a prior probability distribution of the existence of the anomalous substance is constructed. Based on the position and intensity information of the absorption peaks of the key features of the anomaly in the infrared spectral signal, Bayesian inference is used to analyze the probability of the existence of the anomalous substance to generate a posterior probability distribution of the existence of the anomalous substance. The possibility of the existence of the anomalous substance is quantified and evaluated using the posterior probability distribution to obtain the probability of the existence of the anomalous substance. Based on the probability of the existence of the anomalous substance, the width of the posterior probability distribution is analyzed to analyze the uncertainty of the probability of the existence of the anomalous substance, which is used to determine the reliability of the probability of the existence of the anomalous substance to generate an uncertainty assessment result.

[0039] As one optional embodiment, the step of constructing a prior probability distribution of the presence of the anomalous substance based on the difference description, and using Bayesian inference to analyze the probability of the presence of the anomalous substance based on the position and intensity information of the anomalous key feature absorption peaks in the infrared spectrum, to generate a posterior probability distribution of the presence of the anomalous substance, includes: Based on the description of the differences, the differences between the compositional characteristics of the anomalous substance and the compositional characteristics of standard coal are extracted to obtain the key characteristics reflecting the existence of the anomalous substance. Based on the key features and the historical situation of the existence of the anomalous substance, the probability of the existence of the anomalous substance is estimated in order to construct the prior probability distribution of the existence of the anomalous substance. Based on the prior probability distribution, key evidence reflecting the characteristics of the anomalous substance is obtained according to the position and intensity information of the abnormal key feature absorption peaks in the infrared spectral information. Based on the prior probability distribution and the key evidence, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate the posterior probability distribution of the existence of the anomalous substance.

[0040] Specifically, by utilizing differential description, the differences between the compositional characteristics of anomalous substances and those of standard coal are extracted to obtain key features reflecting the existence of anomalous substances. Based on these key features and the historical data of anomalous substance existence, the probability of anomalous substance existence is estimated to construct a prior probability distribution for anomalous substance existence. Based on the prior probability distribution of anomalous substance existence, the position and intensity information of the absorption peaks of anomalous key features in infrared spectral signals are used to obtain key evidence reflecting the characteristics of anomalous substances. Combining the prior probability distribution of anomalous substance existence and key evidence, Bayesian inference is used to analyze the probability of anomalous substance existence to generate a posterior probability distribution for anomalous substance existence.

[0041] In coal quality inspection, anomaly detection algorithms are used to identify anomalous key characteristic absorption peaks in infrared spectral signals that deviate from the normal component ratios. These absorption peaks reflect potential impurities or anomalous substances in the coal. Bayesian inference, based on known data and new evidence, quantifies and assesses the probability and uncertainty of the presence of anomalous substances. A difference description, including the differences between the anomalous features and standard coal component characteristics, is used to construct a prior probability distribution. The posterior probability distribution combines prior information and new evidence to provide a more accurate probability assessment, ensuring that the initial report includes not only the probability of presence but also uncertainty and reliability analysis.

[0042] This application first analyzes the infrared spectral signal of low-quality coal to identify anomalous key absorption peaks and form a set of anomalous key features. Next, it analyzes the differences between these features and known coal compositions, generating anomalous key feature analysis results. Then, using Bayesian inference, a prior probability distribution is constructed based on the difference description, and a posterior probability distribution is calculated by combining the position and intensity information of anomalous absorption peaks in the infrared spectral signal. Finally, the probability and uncertainty of the presence of anomalous substances are quantified and evaluated, generating a detailed preliminary report to provide a scientific basis for subsequent processing.

[0043] For example, in a coal testing laboratory, technicians discovered a batch of coal that had been assessed as low-quality. They conducted an in-depth analysis of the infrared spectral signals of this coal, identifying several anomalous absorption peaks that deviated from the normal component ratios. Using Bayesian inference, the technicians constructed a prior probability distribution for the presence of the anomalous substance and, combined with the actual absorption peak positions and intensities, calculated a posterior probability distribution. The results indicated the possible presence of a rare mineral impurity in this batch of coal, with a relatively high probability of presence but some uncertainty. The resulting preliminary report not only provided the probability of the anomalous substance's presence but also indicated the specific absorption peak positions and a comparison with standard coal components, helping the procurement department adjust its supply chain strategy to ensure the stability of subsequent coal quality.

[0044] Furthermore, embodiments of this application employ Bayesian inference with adaptive weights, a time-dependent exponential decay factor, and spatial correlation to assess the probability and uncertainty of the presence of anomalous substances in coal, in order to generate a preliminary report on the probability of the presence of anomalous substances.

[0045] The Bayesian inference evaluation calculation formula is as follows:

[0046] in, This represents the probability of the presence of an anomalous substance given the observational data. Indicates that in the assumption of the first The probability of observing the current data under these circumstances; Represents the prior probability, indicating the estimate of the first probability based on historical data before observational data is obtained. The probability of such a situation occurring refers to the prediction of the likelihood of the presence of the anomalous substance before observation; An exponential decay factor, representing time dependence and taking into account spatial correlation, is used to adjust the influence of the prior probability on changes in time and spatial location. It is the first The timestamp or relative position of the observation; It is the first Spatial distance between the secondary observation point and the known pollution source; Represents adaptive weights, reflecting the first The importance of the first observation in probability updates; and It is the summation symbol; Indicates that in the assumption of the first The probability of observing the current data assuming the following condition holds true; Represents the prior probability, indicating the prediction of the first [number]th ... The probability of this situation occurring; The exponential decay factor, representing time dependence and spatial correlation, is used to adjust for the influence of changes in prior probability with time and spatial location. It is the first The timestamp or relative position of the observation; It is the first Spatial distance between the secondary observation point and the known pollution source.

[0047] This embodiment employs a Bayesian inference method, combined with adaptive weights, a time-dependent exponential decay factor, and spatial correlation, to provide a more scientific, flexible, and reliable means of coal quality detection. It not only improves the accuracy of assessing the probability of anomalous substances but also addresses complex spatiotemporal variations and uncertainties, providing strong support for coal quality management and supply chain optimization.

[0048] In a specific example, when conducting quality testing on a batch of coal, technicians collected data from multiple observation points, including the timestamp of each observation. Spatial distance between observation point and known pollution source Based on historical data, the prior probability of the presence of anomalous substances in each case was estimated. Simultaneously, the probability of observing the current data under each assumed condition is calculated. .

[0049] Assume there are three observation points The specific data is as follows: Observation point 1: ; Observation point 2: ; Observation point 3: ; Time-dependent exponential decay factor that also considers spatial correlation Represented as:

[0050] Assumption and Then we have:

[0051] Substitute these values ​​into the Bayesian inference evaluation calculation formula:

[0052] Calculated .

[0053] The calculation results show that the probability of the presence of anomaly in this batch of coal is approximately 82.2%, indicating a high likelihood. This result confirms the presence of anomaly in the coal, necessitating further investigation into its source and properties. Furthermore, by introducing adaptive weights, a time-dependent exponential decay factor, and spatial correlation, the Bayesian inference method not only considers the immediacy and spatial distribution of the observed data but also incorporates prior information from historical data, making the assessment results more reliable. The final preliminary report not only provides the probability of the presence of the anomaly but also identifies the specific observation points and times, providing strong support for subsequent steps such as predicting specific information about the anomaly and determining its source.

[0054] As one optional embodiment, the step of using a preset backpropagation neural network to predict anomalous substance information based on the preliminary report, and combining it with a geographic information system to determine the source direction and distance of the anomalous substance, ultimately generating a coal quality inspection report, includes: The data in the preliminary report were configured and processed using a backpropagation neural network to obtain a backpropagation neural network model; The preliminary report is analyzed and predicted based on the backpropagation neural network model, and specific information about the anomalous substance is generated, including elemental composition and molecular structure. Based on the specific information of the anomalous substance, and in conjunction with a geographic information system, the distribution of the pollution sources of the anomalous substance is analyzed to determine the source direction and distance of the anomalous substance. Based on the source direction and distance of the abnormal substance, the quality grade information of the coal is integrated to finally generate a coal quality inspection report that includes the quality grade of the coal and specific information about the abnormal substance.

[0055] Specifically, a backpropagation neural network is used to assess the probability and uncertainty of the presence of anomalous substances in the preliminary report, as well as the location information of the absorption peaks of key anomalous features, and to process the data to obtain a backpropagation neural network model. Based on the backpropagation neural network model, the preliminary report is analyzed, predicted, and specific information on anomalous substances, including elemental composition and molecular structure, is generated. Based on the specific information on anomalous substances, combined with a geographic information system, the distribution of pollution sources of anomalous substances is analyzed to obtain the source direction and distance of anomalous substances. According to the source direction and distance of anomalous substances, coal quality grade information is integrated to generate a coal quality inspection report that includes the coal quality grade and specific information on anomalous substances.

[0056] Among these technologies, the backpropagation neural network is a deep learning model that can predict the specific composition of anomalous substances, including elemental composition and molecular structure, by configuring and processing information on the probability and uncertainty of the presence of anomalous substances in the preliminary report and the location of absorption peaks of key anomalous features. The geographic information system (GIS) is used to analyze the distribution of pollution sources and determine the direction and distance of the anomalous substances' origin. Combining these two technologies can generate detailed coal quality inspection reports, including not only quality grades but also specific information on anomalous substances, providing a scientific basis for supply chain management and environmental monitoring.

[0057] This application first utilizes a backpropagation neural network to configure and process the data in the preliminary report, constructing a neural network model. This model learns the relationship between anomalous substance characteristics and infrared spectral signals through training, thereby accurately predicting the specific information of the anomalous substance. Next, based on the prediction results, it combines a geographic information system to analyze the pollution source distribution of the anomalous substance, determining its source direction and distance. Finally, it integrates coal quality grade information to generate a comprehensive coal quality inspection report, ensuring that the report is detailed and instructive.

[0058] For example, to further improve the efficiency of coal quality testing, technicians in a coal testing laboratory used a backpropagation neural network to make detailed predictions about anomalous substances in low-quality coal based on preliminary reports. After training, the neural network model successfully predicted the specific elemental composition and molecular structure of the anomalous substances, confirming the presence of a rare mineral impurity. Combined with a geographic information system, technicians discovered an abandoned mine near the coal mining site, which may be one of the sources of pollution. The final coal quality testing report not only included detailed component analysis and quality grade assessment but also clearly identified the source direction and distance of the anomalous substances, providing strong support for the procurement department to adjust the supply chain and ensuring the stability and safety of subsequent coal quality.

[0059] Furthermore, based on the preliminary report, this application utilizes a backpropagation neural network to calculate the difference between the predicted and actual values ​​of anomalous substances using weighted mean square error, and introduces weights to adjust the error contribution of anomalous substances to predict specific information about them. The formula for using weighted mean square error is as follows:

[0060] in, It is the weighted mean square error, used to analyze the difference between the predicted and actual values ​​of abnormal substances; It refers to the number of samples containing abnormal substances, which is the total number of samples involved in the error calculation. Representing the The weight of each sample of an anomalous substance reflects the importance of the sample in the overall error assessment. It is the exponential decay coefficient, used to adjust the effect of time or data volume on the weighted mean square error; Representing the The timestamp of a sample of an anomalous substance indicates the time information of the sample; Representing the The actual value of a sample of an anomalous substance refers to the true label of the anomalous substance; It is the first The predicted value of a sample of anomalous substance, and the characteristics of the anomalous substance predicted by a backpropagation neural network model.

[0061] In a specific example, suppose the technician has three samples of anomalous substances. The specific data is as follows: Sample 1: ; Sample 2: ; Sample 3: ; Substituting into the weighted mean square error formula, we can calculate the result. The result indicates that the difference between the predicted and actual values ​​is small, suggesting that the backpropagation neural network model is relatively accurate in predicting the properties of anomalous substances.

[0062] In coal quality inspection, accurately predicting the specific information of anomalous substances and determining their source direction and distance is crucial to ensuring coal quality and safety. Traditional detection methods often rely on fixed statistical models or simple threshold judgments, which are difficult to handle the influence of complex time, data volume, and spatial correlation. Therefore, this embodiment combines backpropagation neural network with weighted mean square error assessment and spatial analysis of geographic information system to more accurately predict the characteristics of anomalous substances and determine their source location. This method not only improves the accuracy of detection but also copes with the influence of changes in time and space.

[0063] Furthermore, in determining the source direction and distance of the anomalous substance based on specific information about the substance and in conjunction with a geographic information system, the calculation formulas for the source direction and distance are as follows:

[0064] in, It represents the great circle distance between two points and is used in geographic information systems to calculate the shortest distance between two points on the Earth's surface. Represents the Earth's radius; Latitude, representing the starting and ending points, in radians; Longitude representing the starting and ending points, in radians; The altitude of the starting and ending points, in meters; It is the Earth's mean radius of curvature, used to convert elevation differences into distance correction terms, and... same; It is an exponential decay coefficient that controls the influence of altitude, used to adjust for the impact of terrain changes on distance calculations. The parameters are derived from the coal sample collection location information and the pollution source location information.

[0065] In a specific example, suppose the sampling location information and pollution source location information of a coal sample are known as follows: Collection location: Latitude ,longitude altitude rice; Location of pollution source: latitude ,longitude altitude rice; Earth radius kilometers, exponential decay coefficient ; After converting the angle to radians, calculate d to obtain... Calculations show that the anomalous substance originated approximately from the northeast. The direction and distance are approximately 146.64 kilometers. This suggests that the anomalous substance may have originated from a relatively distant but clearly defined direction, which will help in further investigation of the specific source of pollution.

[0066] This embodiment provides a more scientific, flexible, and reliable method for coal quality detection by combining backpropagation neural networks, weighted mean square error assessment, and geographic information system spatial analysis. This method not only improves the accuracy of predicting abnormal material properties but also copes with complex spatiotemporal variations and uncertainties, providing strong support for coal quality management and supply chain optimization. By integrating multi-source data and an adaptive adjustment mechanism, the comprehensiveness and robustness of the assessment results are ensured, improving detection accuracy and the effectiveness of decision support.

[0067] This application has the following beneficial effects: This application employs a high-resolution Fourier transform infrared spectrometer for non-contact scanning of coal, rapidly acquiring infrared spectral information over a wide frequency range, and generating infrared spectral signals through consistency processing. An adaptive wavelength selection algorithm is used to identify key characteristic absorption peaks, and machine learning cluster analysis is combined to calculate area ratios, generating an information table reflecting the proportions of coal components. A quality assessment model based on the analytic hierarchy process and fuzzy comprehensive evaluation algorithm, optimized using historical data, ensures the accuracy and reliability of the quality grade assessment results.

[0068] Furthermore, for coal classified as low-quality, anomaly detection algorithms are used to locate key characteristic absorption peaks in the infrared spectral signal, and Bayesian inference is used to assess the probability and uncertainty of the presence of anomalous substances. The generated preliminary report details the probability of the presence of anomalous substances, the uncertainty assessment, the location information of characteristic absorption peaks, and a comparison with the composition of standard coal, providing a scientific basis for a deeper understanding of the root causes of low-quality coal.

[0069] Furthermore, by analyzing the key feature set of anomalies, the differences between the anomaly features and the compositional characteristics of standard coal are extracted, and a prior probability distribution of the presence of anomalous substances is constructed. A posterior probability distribution is generated using Bayesian inference, quantifying and assessing the likelihood and uncertainty of the presence of anomalous substances. This process not only accurately quantifies the probability of anomalous substances but also assesses its reliability by analyzing the width of the posterior probability distribution, enhancing the credibility of coal quality testing results and their decision support capabilities.

[0070] Accordingly, this application also provides a coal quality detection device based on infrared detection, which can realize all the processes of the coal quality detection method based on infrared detection in the above embodiments.

[0071] Please see Figure 2 , Figure 2 This is a schematic diagram of a coal quality detection device based on infrared detection provided in an embodiment of this application. The infrared detection-based coal quality detection device includes: The data acquisition module 201 is used to acquire the infrared spectral information of coal; Information identification module 202 is used to identify key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, perform cluster analysis on the key characteristic absorption peaks, and generate proportional information of different components in coal. The quality assessment module 203 is used to assess the quality of coal based on the ratio information, using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm, to obtain the quality grade assessment result of the coal. Anomaly detection module 204 is used to detect anomalies in low-quality coal in the quality grade assessment results, locate the key characteristic absorption peaks of the anomaly from the infrared spectral information, and assess the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report. The quality inspection module 205 is used to predict abnormal material information based on the preliminary report using a preset backpropagation neural network, and in conjunction with a geographic information system, to determine the source direction and distance of the abnormal material, and finally generate a coal quality inspection report.

[0072] Preferably, the step of identifying key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, and performing cluster analysis on the key characteristic absorption peaks to generate proportional information of different components in the coal includes: Based on the infrared spectral information, an adaptive wavelength selection algorithm is used to dynamically adjust the wavelength selection window of the infrared spectral signal to obtain candidate key characteristic absorption peaks with typical infrared absorption characteristics representing different components in the coal. Based on the candidate key feature absorption peaks, key feature absorption peaks with similar absorption characteristics are grouped into one category by machine learning clustering analysis to obtain the classified key feature absorption peaks. Based on the key characteristic absorption peaks after classification, Fourier transform and inverse transform are used to calculate and analyze the area ratio of the key characteristic absorption peaks to generate the proportion information of different components in coal.

[0073] Preferably, the step of using the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm to assess coal quality based on the proportional information, and obtaining the coal quality grade assessment result, includes: Based on the aforementioned proportion information, the importance of different components to coal quality assessment is evaluated to generate importance assessment results for different components; Based on the importance assessment results, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the judgment matrix is ​​filled with historical data analysis to generate a judgment matrix that reflects the relative importance among different components. Based on the judgment matrix, the weights of different components are applied to the fuzzy comprehensive evaluation algorithm to construct a standard coal quality assessment model. The coal quality is assessed based on the standard coal quality assessment model to obtain the coal quality grade assessment result.

[0074] Preferably, the process involves performing anomaly detection on low-quality coal in the quality grade assessment results, locating key characteristic absorption peaks of the anomalies from the infrared spectral information, and assessing the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report, including: For coal with low quality grades in the quality grade assessment results, anomaly detection is performed on its infrared spectral information to identify and locate abnormal key characteristic absorption peaks that deviate from the normal component ratio. The key characteristic absorption peaks of the anomaly are analyzed to determine the differences between the key characteristic absorption peaks of the anomaly and the known coal composition characteristics, and the analysis results of the key characteristic anomaly are generated. Based on the analysis results of the aforementioned key anomalies, the probability of the existence of the anomalous substance and the uncertainty assessment results are determined by Bayesian inference. Based on the probability and uncertainty assessment results of the presence of the anomalous substance, a preliminary report on the probability of the presence of the anomalous substance is generated. The preliminary report includes the probability of the presence of the anomalous substance, the uncertainty assessment, the location information of the key characteristic absorption peak of the anomalous substance, and the comparison with the composition of standard coal.

[0075] Preferably, the step of determining the probability of the presence of the anomalous substance and the uncertainty assessment result based on the analysis results of the anomalous key features using Bayesian inference includes: Based on the analysis results of the aforementioned key anomalies, a description of the differences between the key anomalies and the characteristics of standard coal composition is extracted; Based on the description of the differences, a prior probability distribution of the existence of the anomalous substance is constructed. Based on the position and intensity information of the absorption peaks of the key anomalous features in the infrared spectrum, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate a posterior probability distribution of the existence of the anomalous substance. The probability of the presence of the anomalous substance is determined based on the posterior probability distribution, and the uncertainty of the probability of the presence of the anomalous substance is analyzed based on the width of the posterior probability distribution to generate an uncertainty assessment result.

[0076] Preferably, the step of constructing a prior probability distribution of the existence of the anomalous substance based on the difference description, and analyzing the probability of the existence of the anomalous substance using Bayesian inference based on the position and intensity information of the anomalous key feature absorption peaks in the infrared spectrum, to generate a posterior probability distribution of the existence of the anomalous substance, includes: Based on the description of the differences, the differences between the compositional characteristics of the anomalous substance and the compositional characteristics of standard coal are extracted to obtain the key characteristics reflecting the existence of the anomalous substance. Based on the key features and the historical situation of the existence of the anomalous substance, the probability of the existence of the anomalous substance is estimated in order to construct the prior probability distribution of the existence of the anomalous substance. Based on the prior probability distribution, key evidence reflecting the characteristics of the anomalous substance is obtained according to the position and intensity information of the abnormal key feature absorption peaks in the infrared spectral information. Based on the prior probability distribution and the key evidence, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate the posterior probability distribution of the existence of the anomalous substance.

[0077] Preferably, the step of predicting anomalous substance information using a preset backpropagation neural network based on the preliminary report, and combining this with a geographic information system to determine the source direction and distance of the anomalous substance, ultimately generating a coal quality inspection report, includes: The data in the preliminary report were configured and processed using a backpropagation neural network to obtain a backpropagation neural network model; The preliminary report is analyzed and predicted based on the backpropagation neural network model, and specific information about the anomalous substance is generated, including elemental composition and molecular structure. Based on the specific information of the anomalous substance, and in conjunction with a geographic information system, the distribution of the pollution sources of the anomalous substance is analyzed to determine the source direction and distance of the anomalous substance. Based on the source direction and distance of the abnormal substance, the quality grade information of the coal is integrated to finally generate a coal quality inspection report that includes the quality grade of the coal and specific information about the abnormal substance.

[0078] In specific implementation, the working principle, control process and technical effects of the coal quality detection device based on infrared detection provided in this application are the same as those of the coal quality detection method based on infrared detection in the above embodiments, and will not be repeated here.

[0079] See Figure 3 , Figure 3This is a structural block diagram of a computer device provided in an embodiment of this application. The computer device includes a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, it implements the steps in the above-described embodiments of the coal quality detection method based on infrared detection. Alternatively, when the processor 301 executes the computer program, it implements the functions of each module / unit in the above-described device embodiments.

[0080] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 302 and executed by the processor 301 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the computer device.

[0081] The computer device may include, but is not limited to, processor 301 and memory 302. Those skilled in the art will understand that the schematic diagram is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0082] The processor 301 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 301 is the control center of the computer device, connecting various parts of the entire computer device through various interfaces and lines.

[0083] The memory 302 can be used to store the computer programs and / or modules. The processor 301 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 302 and calling the data stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0084] Wherein, if the modules / units integrated into the computer device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by the processor 301, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0085] This application also provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the infrared detection-based coal quality detection method described in any of the above embodiments.

[0086] Compared to existing technologies, this application provides a method, apparatus, equipment, and medium for coal quality detection based on infrared detection. Its advantages lie in: acquiring infrared spectral signals of coal, identifying key characteristic absorption peaks using an adaptive wavelength selection algorithm, and calculating area ratios using machine learning cluster analysis to generate an information table reflecting the proportions of coal components. A quality assessment model constructed based on the analytic hierarchy process and fuzzy comprehensive evaluation algorithm, optimized with historical data, ensures the accuracy and reliability of the quality grade assessment results. For coal judged to be of low quality, anomaly detection algorithms locate abnormal key characteristic absorption peaks in the infrared spectral signal, and Bayesian inference assesses the probability and uncertainty of the presence of abnormal substances. The generated preliminary report details the probability of the presence of abnormal substances, uncertainty assessment, location information of characteristic absorption peaks, and comparison with standard coal components, providing a scientific basis for a deeper understanding of the root causes of low-quality coal problems. By analyzing the key feature set of anomalies, the differences between the anomaly features and the composition features of standard coal are extracted. A prior probability distribution of the existence of anomalous substances is constructed, and a posterior probability distribution is generated using Bayesian inference. The possibility and uncertainty of the existence of anomalous substances are quantified and evaluated. This not only accurately quantifies the probability of the existence of anomalous substances, but also assesses its reliability by analyzing the width of the posterior probability distribution, thereby enhancing the credibility of coal quality detection results and decision support capabilities.

[0087] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A coal quality detection method based on infrared detection, characterized in that, include: Obtain infrared spectral information of coal; Based on the infrared spectral information, key characteristic absorption peaks of organic components and minerals in coal are identified, and cluster analysis is performed on the key characteristic absorption peaks to generate information on the proportion of different components in coal. Based on the aforementioned proportion information, the coal quality is assessed using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm to obtain the coal quality grade assessment results. Anomaly detection is performed on low-quality coal in the quality grade assessment results. The key characteristic absorption peaks of the anomaly are located from the infrared spectral information. Based on Bayesian inference, the probability and uncertainty of the presence of abnormal substances in the coal are assessed to generate a preliminary report. Based on the preliminary report, a pre-set backpropagation neural network is used to predict abnormal material information, and combined with a geographic information system, the source direction and distance of the abnormal material are determined, and finally a coal quality inspection report is generated.

2. The coal quality detection method based on infrared detection as described in claim 1, characterized in that, The process of identifying key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, performing cluster analysis on the key characteristic absorption peaks, and generating proportional information of different components in coal includes: Based on the infrared spectral information, an adaptive wavelength selection algorithm is used to dynamically adjust the wavelength selection window of the infrared spectral signal to obtain candidate key characteristic absorption peaks with typical infrared absorption characteristics representing different components in the coal. Based on the candidate key feature absorption peaks, key feature absorption peaks with similar absorption characteristics are grouped into one category by machine learning clustering analysis to obtain the classified key feature absorption peaks. Based on the key characteristic absorption peaks after classification, Fourier transform and inverse transform are used to calculate and analyze the area ratio of the key characteristic absorption peaks to generate the proportion information of different components in coal.

3. The coal quality detection method based on infrared detection as described in claim 1, characterized in that, The coal quality assessment is performed based on the aforementioned proportion information using the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation algorithm to obtain the coal quality grade assessment result, including: Based on the aforementioned proportion information, the importance of different components to coal quality assessment is evaluated to generate importance assessment results for different components; Based on the importance assessment results, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP), and the judgment matrix is ​​filled with historical data analysis to generate a judgment matrix that reflects the relative importance among different components. Based on the judgment matrix, the weights of different components are applied to the fuzzy comprehensive evaluation algorithm to construct a standard coal quality assessment model. The coal quality is assessed based on the standard coal quality assessment model to obtain the coal quality grade assessment result.

4. The coal quality detection method based on infrared detection as described in claim 1, characterized in that, The process involves performing anomaly detection on low-quality coal in the quality grade assessment results, locating key characteristic absorption peaks of the anomalies from the infrared spectral information, and assessing the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report, including: For coal with low quality grades in the quality grade assessment results, anomaly detection is performed on its infrared spectral information to identify and locate abnormal key characteristic absorption peaks that deviate from the normal component ratio. The key characteristic absorption peaks of the anomaly are analyzed to determine the differences between the key characteristic absorption peaks of the anomaly and the known coal composition characteristics, and the analysis results of the key characteristic anomaly are generated. Based on the analysis results of the aforementioned key anomalies, the probability of the existence of the anomalous substance and the uncertainty assessment results are determined by Bayesian inference. Based on the probability and uncertainty assessment results of the presence of the anomalous substance, a preliminary report on the probability of the presence of the anomalous substance is generated. The preliminary report includes the probability of the presence of the anomalous substance, the uncertainty assessment, the location information of the key characteristic absorption peak of the anomalous substance, and the comparison with the composition of standard coal.

5. The coal quality detection method based on infrared detection as described in claim 4, characterized in that, The determination of the probability of the presence of the anomalous substance and the uncertainty assessment results based on the analysis results of the key abnormal features through Bayesian inference include: Based on the analysis results of the aforementioned key anomalies, a description of the differences between the key anomalies and the characteristics of standard coal composition is extracted; Based on the description of the differences, a prior probability distribution of the existence of the anomalous substance is constructed. Based on the position and intensity information of the absorption peaks of the key anomalous features in the infrared spectrum, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate a posterior probability distribution of the existence of the anomalous substance. The probability of the presence of the anomalous substance is determined based on the posterior probability distribution, and the uncertainty of the probability of the presence of the anomalous substance is analyzed based on the width of the posterior probability distribution to generate an uncertainty assessment result.

6. The coal quality detection method based on infrared detection as described in claim 5, characterized in that, The step of constructing a prior probability distribution for the existence of anomalous substances based on the difference description, and using Bayesian inference to analyze the probability of the existence of anomalous substances based on the position and intensity information of the key anomalous feature absorption peaks in the infrared spectrum, to generate a posterior probability distribution for the existence of anomalous substances, includes: Based on the description of the differences, the differences between the compositional characteristics of the anomalous substance and the compositional characteristics of standard coal are extracted to obtain the key characteristics reflecting the existence of the anomalous substance. Based on the key features and the historical situation of the existence of the anomalous substance, the probability of the existence of the anomalous substance is estimated in order to construct the prior probability distribution of the existence of the anomalous substance. Based on the prior probability distribution, key evidence reflecting the characteristics of the anomalous substance is obtained according to the position and intensity information of the abnormal key feature absorption peaks in the infrared spectral information. Based on the prior probability distribution and the key evidence, Bayesian inference is used to analyze the probability of the existence of the anomalous substance in order to generate the posterior probability distribution of the existence of the anomalous substance.

7. The coal quality detection method based on infrared detection as described in claim 1, characterized in that, Based on the preliminary report, a pre-set backpropagation neural network is used to predict anomalous substance information, and combined with a geographic information system, the source direction and distance of the anomalous substance are determined, ultimately generating a coal quality inspection report, including: The data in the preliminary report were configured and processed using a backpropagation neural network to obtain a backpropagation neural network model; The preliminary report is analyzed and predicted based on the backpropagation neural network model, and specific information about the anomalous substance is generated, including elemental composition and molecular structure. Based on the specific information of the anomalous substance, and in conjunction with a geographic information system, the distribution of the pollution sources of the anomalous substance is analyzed to determine the source direction and distance of the anomalous substance. Based on the source direction and distance of the abnormal substance, the quality grade information of the coal is integrated to finally generate a coal quality inspection report that includes the quality grade of the coal and specific information about the abnormal substance.

8. A coal quality detection device based on infrared detection, characterized in that, include: The data acquisition module is used to acquire the infrared spectral information of coal. The information identification module is used to identify key characteristic absorption peaks of organic components and minerals in coal based on the infrared spectral information, perform cluster analysis on the key characteristic absorption peaks, and generate proportional information of different components in coal. The quality assessment module is used to assess coal quality based on the ratio information using the analytic hierarchy process and the fuzzy comprehensive evaluation algorithm, and to obtain the coal quality grade assessment result. An anomaly detection module is used to detect anomalies in low-quality coal in the quality grade assessment results, locate the key characteristic absorption peaks of the anomalies from the infrared spectral information, and assess the probability and uncertainty of the presence of abnormal substances in the coal based on Bayesian inference to generate a preliminary report. The quality inspection module is used to predict abnormal material information based on the preliminary report using a preset backpropagation neural network, and in conjunction with a geographic information system, to determine the source direction and distance of the abnormal material, and finally generate a coal quality inspection report.

9. A computer device, characterized in that, The device includes a processor and a memory, wherein the memory stores a computer program and the computer program is configured to be executed by the processor, wherein the processor executes the computer program to implement the coal quality detection method based on infrared detection as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the coal quality detection method based on infrared detection as described in any one of claims 1 to 7.

Images