Gray clean coal component analysis and detection system and method

Abstract

The present invention relates to the field of coal detection technology, and in particular to a gray clean coal component analysis and detection system and method, comprising a gray clean coal extraction module, a sample collection module, a component analysis module, a combustion performance analysis module, and a management database. The system extracts gray clean coal from medium coal and other types of coal, and collects gray clean coal samples to be analyzed based on the transportation mode and storage status of the gray clean coal, performs preprocessing, obtains component data and particle size interval distribution information of the gray clean coal samples through multiple analysis methods, calculates the correlation coefficient matrix between variables, screens and obtains each principal component, calculates the correlation between the principal component and the combustion performance index, evaluates the gray clean coal combustion performance coefficient, and dynamically optimizes the flotation parameters according to the evaluation results. The system can clearly understand the gray clean coal composition and clarify the quantitative relationship between the composition, particle size distribution and other characteristics of the gray clean coal and the combustion performance.

Description

Technical Field

[0001] The present invention relates to the technical field of coal detection, in particular to a gray clean coal component analysis detection system and method. Background Art

[0002] As an important basic energy source in my country, coal occupies a key position in the development of the national economy. With the continuous development and utilization of coal resources, the rational disposal and utilization of by-products such as middlings and other types of coal produced during coal washing have received increasing attention. Gray clean coal is a product further extracted from middlings and other types of coal in coal washing, and it is of great significance to conduct component analysis and testing on it.

[0003] Accurately analyzing the composition of clean coal helps improve the comprehensive utilization rate of coal resources. By gaining a deeper understanding of the content and characteristics of minerals and organic components in clean coal, we can tap into its potential value, maximize resource utilization, reduce resource waste, and meet the strategic requirements of sustainable development. At the same time, the results of clean coal composition analysis are crucial for the subsequent processing and utilization of coal. For example, in combustion applications, understanding the composition of clean coal can optimize the design and operating parameters of combustion equipment, improve combustion efficiency, reduce pollutant emissions, and achieve clean and efficient energy utilization.

[0004] However, the existing technology still has some shortcomings. For example, the existing Chinese patent with publication number 201810447623.1 discloses a coal composition analysis method based on coal spectral data. This scheme uses spectral data and coal industry analysis and measurement results to establish a coal composition analysis model. The model uses a convolutional neural network to extract spectral feature data, and outputs the spectral feature data to an extreme learning machine to obtain the coal composition corresponding to the coal spectral data obtained by coal industry analysis and measurement. During the prediction process, an artificial bee colony algorithm is used to optimize the weights and deviations of the extreme learning machine, thereby obtaining an optimized coal composition analysis model.

[0005] For example, the existing Chinese patent publication number 201610811928.7 discloses an intelligent monitoring method for coal quality composition of coal mills in thermal power plants. This solution continuously collects the operating data, coal quantity of each coal mill, and the power generation load of the unit to calculate the raw coal moisture MNj and equivalent electrical load QNj of each coal mill, establishes a coal quality characteristic analysis database, and uses the raw coal moisture Mi and equivalent electrical load Qi monitored in real time during the operation of the coal mill as indexes to query and output the raw coal quality characteristics. If no results are found, the raw coal of the coal mill is a new type of coal or mixed coal. Manual sampling is then performed for coal quality composition and characteristic analysis, and the coal type is added to the coal quality characteristic database. The coal quality characteristics of each coal mill can be obtained online with high accuracy and complete analysis data, guiding boiler combustion optimization.

[0006] However, all of these patents suffer from the following issues: First, they focus solely on the detection technology itself, lacking attention and design to the coal extraction process, and ignoring the prerequisite of obtaining a representative and pure coal sample. Coal, in its natural state or after mining, is often mixed with various impurities, and its internal composition distribution is not uniform. Without a scientific and reasonable extraction process, direct testing is likely to result in inaccurate composition data due to impurity interference and sample bias.

[0007] 2. The above scheme adopts a single sampling method, and the samples taken do not take into account the differences in characteristics over time and space, resulting in the samples being unable to accurately reflect the overall characteristics and dynamic changes of coal characteristics over time, thereby generating sample bias and affecting the accuracy of subsequent analysis and processing.

[0008] 3. The above scheme only analyzes the coal composition and does not involve further in-depth analysis, such as providing a quantitative assessment of the combustion performance of gray clean coal. It cannot directly provide specific guidance on combustion performance for the combustion optimization and operation adjustment of the coal mill. Summary of the Invention

[0009] In order to overcome the shortcomings of the background technology, the embodiments of the present invention provide a gray clean coal composition analysis and detection system and method, which can effectively solve the problems involved in the above background technology.

[0010] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a gray clean coal component analysis and detection system, including: a gray clean coal extraction module, which is used to extract high-purity gray clean coal from medium coal and other coal types through the synergistic effect of compressed air and hydrophobic flocculation.

[0011] The sample collection module is used to combine time interval sampling and spatial stratification sampling to obtain coal samples of various ash clean coals.

[0012] The component analysis module is used to obtain the component data and particle size distribution information of the gray clean coal sample through multiple analysis methods.

[0013] The combustion performance analysis module is used to screen the main components based on the composition data of the ash clean coal sample and the distribution information of each particle size interval, and analyze the direction and degree of contribution of each main component to the combustion performance index.

[0014] The evaluation feedback module is used to evaluate the combustion performance coefficient of the gray clean coal and dynamically optimize the flotation parameters based on the evaluation results.

[0015] The management database is used to store the composition data of the ash clean coal samples and the distribution information of each particle size interval of the ash clean coal samples.

[0016] Preferably, the specific operation method of extracting the gray clean coal is: pre-treating the collected middling coal and other coal types and adding them into a flotation tank together with a flotation agent to form flocs through hydrophobic flocculation.

[0017] Air is introduced into the flotation tank through a porous distributor, so that the air forms a large number of uniform microbubbles in the slurry. The microbubbles fully collide and adhere with the flocs of medium coal and other coal types after hydrophobic flocculation. The buoyancy of the microbubbles drives the flocs to float to the surface of the slurry, thereby forming a foam layer.

[0018] The foam layer is scraped off with a scraper device at a set speed and frequency to collect the foam product containing gray clean coal. The foam product is sequentially dehydrated to remove excess water, and then subjected to secondary flotation purification to further separate impurities, ultimately obtaining high-purity gray clean coal.

[0019] Preferably, the specific analysis method of the sample collection module is: for the gray clean coal conveyed by the belt, a sampling time interval is set, and the gray clean coal samples at different times during the belt conveying process are obtained using a belt middle sampler according to the time interval.

[0020] For the stockpiled gray clean coal, the stockpiled gray clean coal is divided into different heights, and multiple sampling points are set at each height. Layered sampling is carried out from each sampling point at each height to obtain samples of the stockpiled gray clean coal at different spatial positions.

[0021] The coal samples obtained by the two methods of time interval sampling during belt conveying and spatial stratification sampling of stockpiled clean ash coal are comprehensively mixed to obtain a total sample, and the total sample is evenly divided into several clean ash coal samples.

[0022] Preferably, the specific method for obtaining the composition data of the ash clean coal sample is: weigh a portion of the ash clean coal sample, place it in a crucible after pretreatment, record the initial total mass of the ash clean coal sample and the crucible, place it in the sample cell of the thermogravimetric analyzer, start the heating program, and raise the temperature from room temperature to the set moisture detection temperature. After a fixed period of time, record the total mass of the ash clean coal sample and the crucible at this time, and calculate the moisture content of the ash clean coal sample by calculating the reduction in the initial total mass of the ash clean coal sample and the crucible.

[0023] The temperature is continued to be raised to the set volatilization detection temperature in the same way, and the volatile matter content of the ash clean coal sample is calculated.

[0024] The ash content of the coal sample is calculated based on the total mass of the final ash clean coal sample residue and the crucible, and then the fixed carbon content of the ash clean coal sample is obtained by subtracting the moisture content, volatile matter content and ash content from the initial total mass of the coal sample components.

[0025] Preferably, the specific method for obtaining the composition data of the gray clean coal sample also includes: obtaining the content data of each component of a large number of standard coal samples with different contents of each component, collecting the coal sample spectrum of the standard coal sample through a near-infrared spectrometer, and using the partial least squares method to establish a quantitative model of the near-infrared spectrum and the coal sample composition.

[0026] A sample of clean ash coal is taken, and its near-infrared spectrum is collected by a near-infrared spectrometer. The collected spectral data is input into a quantitative model of near-infrared spectrum and coal sample composition to predict the component content of the clean ash coal sample.

[0027] Preferably, the specific method for obtaining the distribution information of each particle size interval is: taking a gray clean coal sample, adding it to the sample pool of the laser particle size analyzer, allowing the sample to be fully dispersed in the dispersion medium, starting the laser particle size analyzer, and irradiating the laser beam of the instrument onto the gray clean coal sample in the sample pool.

[0028] The angle and intensity distribution of scattered light generated by particles of different particle sizes to the laser are measured, and the collected scattered light intensity data at different angles are substituted into the Mie scattering theory formula. The particle size distribution model parameters are adjusted through iterative calculation to achieve the best fit between the calculated and measured scattered light intensity distributions. The particle size distribution model that best fits the measured data is determined to obtain the distribution information of each particle size interval.

[0029] Preferably, the specific analysis method of the principal component analysis module is: summarizing the component data of the gray coal sample and the distribution information of each particle size interval, judging whether there are missing values ​​therein; if there are missing component data, making reasonable estimates and filling them based on the statistical data of similar samples of the gray coal sample; if there is a missing proportion of a certain particle size interval in the distribution information of each particle size interval, performing interpolation calculation based on the adjacent particle size interval data to fill the missing values.

[0030] The filled data are standardized, the correlation coefficient matrix between the variables is calculated, and then the linear equations are solved for each eigenvalue to obtain the corresponding eigenvector, and the proportion of each eigenvalue to the total eigenvalue is calculated.

[0031] The proportion of each eigenvalue to the total eigenvalue is sorted in descending order, the proportion of each eigenvalue to the total eigenvalue is accumulated in sequence, and the corresponding eigenvectors when the accumulation reaches the set proportion are recorded as principal components.

[0032] Preferably, the specific analysis method of the combustion performance analysis module is: using the Pearson correlation coefficient to calculate the correlation coefficient between each principal component and the combustion performance index, setting the significance level, and using the t-test method to perform a significance test on the calculated correlation coefficient. When the calculated t-statistic is greater than the critical value of the t-distribution, it is considered that the correlation between the principal component and the combustion performance index is significant.

[0033] The principal components with significant correlation with combustion performance indicators were screened out, a linear regression model was constructed, the regression model was fitted using the least squares method, and the regression coefficient was estimated. Based on the size and sign of the regression coefficient, the direction and degree of contribution of each principal component to the combustion performance indicators were analyzed.

[0034] The correlation between the principal components and the combustion performance is determined based on the contribution direction of each principal component to the combustion performance index. The contribution direction includes positive and negative directions. A positive direction indicates that the principal component is positively correlated with the combustion performance, and a negative direction indicates that the principal component is negatively correlated with the combustion performance.

[0035] The specific analysis method of the evaluation feedback module is: obtaining the combustion performance coefficient by multiplying each principal component by its corresponding regression coefficient and then adding them, setting the expected standard of the combustion performance coefficient, and adjusting the flotation parameters if the combustion performance coefficient is less than the expected standard of the combustion performance coefficient.

[0036] The principal components that are positively correlated and negatively correlated with the combustion performance are screened out respectively, and the lower limit adjustment threshold and the upper limit adjustment threshold are set for them.

[0037] If the content of a main component that is positively correlated with combustion performance is less than the lower adjustment threshold, the dosage of the flotation agent is adjusted; if the content of a main component that is negatively correlated with combustion performance is greater than the upper adjustment threshold, the flotation time is adjusted.

[0038] Within a fixed period of time after adjusting the flotation parameters, the ash clean coal sample is re-extracted for component analysis and combustion performance testing to obtain the adjusted combustion performance coefficient.

[0039] The adjusted combustion performance coefficient is compared with the expected combustion performance coefficient standard. If the adjusted combustion performance coefficient is less than the expected combustion performance coefficient standard, the flotation parameters are adjusted again until the adjusted combustion performance coefficient reaches or exceeds the expected combustion performance coefficient standard.

[0040] Preferably, the present invention provides a method for analyzing and detecting the components of gray clean coal, and the specific steps of the detection method are as follows: S1. Gray clean coal extraction: high-purity gray clean coal is extracted from medium coal and other types of coal through the synergistic effect of compressed air and hydrophobic flocculation.

[0041] S2. Sample collection: Combine time interval sampling and spatial stratification sampling to obtain coal samples of each ash clean coal.

[0042] S3. Composition analysis: The composition data and particle size distribution information of the ash clean coal sample are obtained through multiple analysis methods.

[0043] S4. Combustion performance analysis: The main components are selected based on the composition data of the ash clean coal sample and the distribution information of each particle size interval, and the direction and degree of contribution of each main component to the combustion performance index are analyzed.

[0044] S5. Evaluation feedback: Evaluate the combustion performance coefficient of the clean coal and dynamically optimize the flotation parameters based on the evaluation results.

[0045] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: 1. The present invention extracts gray clean coal from medium coal and other types of coal through the synergistic effect of compressed air and hydrophobic flocculation. The large number of microbubbles generated by compressed air provide a powerful buoyancy carrier for the flocs of medium coal and other types of coal after hydrophobic flocculation. After the microbubbles fully collide and adhere with the flocs, they can more effectively carry the gray clean coal particles originally dispersed in the slurry to the surface of the slurry, significantly improving the recovery rate of the gray clean coal.

[0046] 2. The present invention effectively eliminates the sample bias caused by the traditional single sampling method by combining time interval sampling and spatial stratified sampling, so that the obtained gray clean coal sample can more accurately represent the characteristics of the overall gray clean coal.

[0047] 3. The present invention obtains the composition data and particle size distribution information of the gray clean coal sample through multiple analysis methods, screens out the main components, evaluates the direction and degree of contribution of each main component to the combustion performance index, increases the content of the main component in a targeted manner, improves the combustion calorific value of the gray clean coal, and meets the specific requirements of different industrial applications for combustion performance.

[0048] 4. The present invention calculates the correlation between the main components and the combustion performance indicators, thereby constructing a linear regression equation to evaluate the combustion performance coefficient of the gray clean coal. It can clarify the quantitative relationship between the composition, particle size distribution and other characteristics of the gray clean coal and the combustion performance, and reveal the key factors affecting the combustion performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 This is a module connection diagram of an ash clean coal composition analysis and detection system.

[0051] Figure 2 for Figure 1 Flowchart of pretreatment in the medium ash clean coal extraction module.

[0052] Figure 3The figure is a flow chart of a method for analyzing and detecting the components of ash clean coal. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0054] See also Figure 1 As shown, a gray clean coal component analysis and detection system includes a gray clean coal extraction module, a sample collection module, a component analysis module, a combustion performance analysis module, and a management database.

[0055] The management database is connected to the ash clean coal extraction module, the sample collection module, the component analysis module, and the combustion performance analysis module; the sample collection module is connected to the component analysis module and the ash clean coal extraction module.

[0056] The gray clean coal extraction module is used to extract high-purity gray clean coal from medium coal and other types of coal through the synergistic effect of compressed air and hydrophobic flocculation.

[0057] See also Figure 2 As shown, the specific operation method for extracting gray clean coal is as follows: the collected medium coal and other types of coal are pretreated and then added to the flotation tank together with the flotation agent to form flocs through hydrophobic flocculation; the hydrophobic flocculation is used to promote particle agglomeration, so that the effective capture range of the target substance in the flotation process is expanded, and the clean coal components in the medium coal and other types of coal can be more efficiently separated from the complex mixture, which significantly improves the selectivity and efficiency of flotation.

[0058] Air is introduced into the flotation cell through a porous distributor, causing the air to form a large number of uniform microbubbles in the slurry. The microbubbles fully collide and adhere to the flocs of medium coal and other coal types after hydrophobic flocculation. The buoyancy of the microbubbles drives the flocs to float to the surface of the slurry, thereby forming a foam layer. The uniform and stable microbubble system ensures the consistency and reliability of the flotation process, reduces the fluctuation of the flotation effect caused by uneven bubble size or distribution, and makes the entire flotation process easier to control and optimize.

[0059] A scraper device is used to scrape the foam layer at a set speed and frequency to collect a foam product containing gray clean coal. The foam product is then dehydrated to remove excess water and then subjected to secondary flotation purification to further separate impurities, ultimately obtaining high-purity gray clean coal. Secondary flotation purification can further separate impurities and significantly improve the purity of the gray clean coal. The resulting high-purity gray clean coal has higher value in industrial applications.

[0060] It should be noted that the medium coal and other types of coal are intermediate products produced in the coal washing process. During the coal washing process, the raw coal will be separated into products of different qualities after a series of processes such as crushing, screening, jigging, and flotation. Among them are medium coal and other types of coal. Through a series of steps such as pretreatment, flotation, scraping, dehydration, and secondary flotation purification of the medium coal and other types of coal, gray clean coal is finally obtained.

[0061] It should be noted that the flotation agent includes a collector and a foaming agent. The collector is hydrocarbon oil, and its dosage is 150-250g per ton of medium coal and other coal types. The foaming agent is pine oil, and its addition dosage is 80-120g per ton of medium coal and other coal types, so that the medium coal and other coal types undergo hydrophobic flocculation in the slurry to form flocs.

[0062] The sample collection module is used to combine time interval sampling and spatial stratification sampling to obtain coal samples of various ash clean coals.

[0063] The specific analysis method of the sample collection module is: for the gray clean coal conveyed by the belt, a sampling time interval is set, and the gray clean coal samples at different times during the belt conveying process are obtained using the belt middle sampler according to the time interval.

[0064] For the stockpiled gray clean coal, the stockpiled gray clean coal is divided into different heights, and multiple sampling points are set at each height. Layered sampling is carried out from each sampling point at each height to obtain samples of the stockpiled gray clean coal at different spatial positions.

[0065] The coal samples obtained by the two methods of time interval sampling through belt conveyor and spatial stratification sampling of stockpiled ash clean coal are comprehensively mixed to obtain a total sample, and the total sample is divided into several ash clean coal samples; by combining sampling in the time dimension and the space dimension, the sample deviation caused by the traditional single sampling method is effectively eliminated, so that the obtained ash clean coal samples can more accurately represent the characteristics of the overall ash clean coal.

[0066] The component analysis module is used to obtain the component data and particle size distribution information of the gray clean coal sample through multiple analysis methods.

[0067] The specific method for obtaining the composition data of the ash clean coal sample is as follows: weigh a portion of the ash clean coal sample, place it in a crucible after pretreatment, record the initial total mass of the ash clean coal sample and the crucible, place it in the sample cell of the thermogravimetric analyzer, start the heating program, and raise the temperature from room temperature to the set moisture detection temperature. After a fixed period of time, record the total mass of the ash clean coal sample and the crucible at this time, and calculate the moisture content of the ash clean coal sample by calculating the reduction in the initial total mass of the ash clean coal sample and the crucible; during the combustion process, the moisture content will affect the combustion efficiency and heat output of the coal. Understanding the moisture content can better optimize the combustion process and improve energy utilization efficiency.

[0068] It should be noted that the specific analysis method of the moisture content of the ash clean coal sample is: assuming the initial total mass of the ash clean coal sample and the crucible is After the moisture detection temperature is kept at a fixed time, the total mass of the gray clean coal sample and the crucible is , through the formula The mass reduction is calculated to be the mass of water in the ash clean coal sample.

[0069] The pretreatment steps include: A1. adjusting the crusher discharge port size according to the set screening requirements, and pouring the collected gray clean coal sample into the crusher feed port for crushing according to the set rotation speed; by accurately setting the discharge port size and controlling the crusher rotation speed, the particle size of the crushed coal sample can be made as close to the expected range as possible, providing raw materials with uniform particle size and meeting the requirements for subsequent processes.

[0070] It should be noted that the set rotation speed is determined according to the hardness of the ash clean coal sample to be processed.

[0071] For example, a jaw crusher with a processing capacity of 80-240 tons / hour is suitable for coarsely crushing gray fine coal samples. For the jaw crusher, if the coal sample is relatively hard, such as the Mohs hardness reaches 4-5, the speed can be adjusted to 250 revolutions / minute, because coal samples with high hardness require a slower crushing process to avoid excessive wear and overload of the equipment.

[0072] A2. According to the crushing particle size range of the crusher, select a sieve with a corresponding aperture to screen the crushed gray clean coal sample, and collect the gray clean coal sample with a particle size that meets the requirements under the sieve; through screening, the coal sample with unqualified particle size is separated and can be crushed again or otherwise processed, avoiding the waste of unqualified coal samples in subsequent processes.

[0073] A3. Place the screened ash-clean coal sample into a drying container and place it in the drying equipment. Set the drying temperature range and drying time. Repeat the drying, cooling, and weighing operations until the coal sample reaches a constant weight. Removing excess moisture from the coal sample can improve the performance of the coal sample in subsequent processing. The dried coal sample is easier to achieve the target particle size during grinding, reduces agglomeration caused by moisture, and improves grinding efficiency.

[0074] For example, for a gray clean coal sample with a moisture content of about 15%, the oven temperature is set to 105° C. and the drying time is set to 3 hours.

[0075] A4. Place the dried ash clean coal sample into a grinding container for grinding. After grinding, check whether the coal sample particle size meets the standard. If it does not meet the standard, continue grinding to obtain a ground ash clean coal sample.

[0076] It should be noted that the coal sample particle size is detected by making the ground coal sample into a suspension, placing it into a laser particle size analyzer for measurement, and obtaining the average particle size of the gray clean coal sample, which is then compared with the upper and lower limits of the standard particle size. If the average particle size is within the upper and lower limits of the standard particle size, it is determined that the coal sample particle size meets the standard.

[0077] Continue to raise the temperature to the set volatilization detection temperature in the same way, and calculate the volatile matter content of the gray clean coal sample; volatile matter is a gaseous substance produced by the thermal decomposition of coal under specific conditions. The accurate determination of its content is of great significance for evaluating the combustion characteristics and processing and utilization performance of coal.

[0078] The ash content of the coal sample is calculated based on the total mass of the final ash-cleaned coal sample residue and the crucible, and then the fixed carbon content of the ash-cleaned coal sample is obtained by subtracting the moisture content, volatile matter content and ash content from the initial total mass of the coal sample components. Ash content is one of the important quality indicators of coal, which will affect the combustion performance, thermal efficiency and environmental impact of coal. Accurate determination of ash content helps to rationally utilize coal resources and reduce the cost and environmental burden of ash treatment. Fixed carbon is an important component of the combustible components in coal, and its content directly affects the calorific value and combustion performance of coal. Accurate determination of fixed carbon content can provide a more comprehensive understanding of the quality and combustion characteristics of coal, and provide a more accurate basis for the rational utilization of coal.

[0079] It should be noted that the initial total mass of the weighed ash clean coal sample and the crucible is , the mass of the empty crucible is The total mass of the final ash clean coal sample residue and crucible is , the ash content of the coal sample The calculation formula is .

[0080] The specific method for obtaining the composition data of the gray clean coal sample also includes: obtaining the content data of each component of a large number of standard coal samples, collecting the coal sample spectrum of the standard coal sample through a near-infrared spectrometer, and using the partial least squares method to establish a quantitative model of near-infrared spectrum and coal sample composition.

[0081] A sample of ash-clean coal is taken, and its near-infrared spectrum is collected using a near-infrared spectrometer. The collected spectral data is input into a quantitative model of near-infrared spectrum and coal sample composition to predict the component content of the ash-clean coal sample. The model has been trained and optimized with a large amount of standard coal sample data and can more accurately predict the component content of the ash-clean coal sample. Accurate prediction results help to evaluate and classify the quality of coal samples, providing a scientific basis for the rational utilization and sale of coal.

[0082] It should be noted that the data on the content of each component of the standard coal samples are divided into a training set and a validation set.

[0083] A near-infrared spectrometer was used to perform spectral scanning on the standard coal samples of each component content in the training set to obtain a series of spectral data. The spectral data of the standard coal samples of each component content were arranged in wavelength order to form a one-dimensional data vector. The data vectors of the standard coal samples of each component content were arranged in rows to form a spectral data matrix.

[0084] Perform singular value decomposition on the spectral data matrix, extract the principal components, calculate the cumulative contribution rate to determine the number of principal components, and select the The left singular vectors corresponding to the singular values ​​are taken as the principal component matrix, denoted as .

[0085] It should be noted that a threshold of cumulative contribution rate is set in advance. When the cumulative contribution rate reaches or exceeds the threshold for the first time, the corresponding is the number of principal components selected.

[0086] Assume that the component content data matrix of the standard coal sample is , the dimension is ,in is the number of principal components, The number of standard coal samples for component content is used to establish the principal component matrix and ingredient content data matrix Building a regression model ,in is the regression coefficient matrix, is the error matrix, and the regression coefficient matrix obtained by solving , which is the quantitative model of near-infrared spectrum and coal sample composition, reflecting the quantitative relationship between the main component and the component content.

[0087] The content data of each component of the standard coal samples in the verification set were input into the quantitative model of near-infrared spectroscopy and coal sample composition for output verification.

[0088] The specific method for obtaining the distribution information of each particle size interval is as follows: taking a gray clean coal sample, adding it to the sample pool of the laser particle size analyzer, allowing the sample to be fully dispersed in the dispersion medium, starting the laser particle size analyzer, and irradiating the laser beam of the instrument onto the gray clean coal sample in the sample pool; the laser particle size analyzer has the characteristics of rapid measurement and can obtain a large amount of particle size distribution data in a short time.

[0089] The angle and intensity distribution of scattered light generated by particles of different particle sizes in response to laser light are measured, and the collected scattered light intensity data at different angles are substituted into the Mie scattering theory formula. The particle size distribution model parameters are adjusted through iterative calculation to achieve the best fit between the calculated and measured scattered light intensity distributions. The particle size distribution model that best fits the measured data is determined, thereby obtaining distribution information for each particle size interval. Mie scattering theory is a mature light scattering theory that is suitable for the calculation of scattered light from spherical particles. It can accurately describe the scattering characteristics of particles of different particle sizes in response to laser light, thus providing a theoretical basis for the accurate measurement of particle size distribution.

[0090] It should be noted that the specific steps for measuring the angle and intensity distribution of scattered light generated by particles of different particle sizes to the laser are: preparing a suspension of the gray clean coal sample, slowly injecting it into the sample cell, starting the stirring device to keep the sample in a uniform suspension state in the sample cell, and when the laser particle size analyzer emits a laser beam through the suspension in the sample cell, particles of different particle sizes will scatter the laser, and the scattered light will propagate in all directions, and then the detector array will collect scattered light intensity data at different angles in real time.

[0091] It should be noted that the specific method for constructing the particle size distribution model is: setting the particle size range, setting the mean at the middle of the range, and taking a preliminary value of the standard deviation based on the expected width of the particle size distribution, thereby establishing an initial particle size distribution model.

[0092] According to the set initial particle size distribution model and the Mie scattering theory formula, the scattered light intensity distribution at each scattering angle is calculated. The calculated scattered light intensity is compared with the actual measured intensity, and the root mean square error is used as the error evaluation function.

[0093] The parameters of the particle size distribution model are adjusted using a genetic algorithm. The parameters of the particle size distribution model are encoded to form population individuals. The inverse of the error evaluation function is used as the fitness function to calculate the fitness of each individual. The higher the fitness, the better the fit between the model and the measured data.

[0094] Through genetic operations such as selection, crossover and mutation, new populations are generated, and continuous iterative optimization is carried out so that the individuals in the population gradually approach the optimal solution. That is, the model parameters are continuously adjusted so that the error between the calculated scattered light intensity distribution and the measured data gradually decreases.

[0095] A convergence threshold is set. When the error between the scattered light intensity distribution and the measured data is less than the threshold in multiple consecutive iterations, the iterative process is considered to have converged. The model parameters at this time are the parameters that make the calculated and measured scattered light intensity distributions achieve a good fit.

[0096] When the best fit is achieved, the distribution information of each particle size interval is determined based on the fitted particle size distribution model, and the particle size range is divided into several small intervals. For each small interval, the probability density function of the particle size distribution model is integrated over the interval to calculate the relative content of the particles in the interval, thereby obtaining the distribution information of each particle size interval.

[0097] The formula for the probability density function of the particle size distribution model is: ,in Indicates particle size, represents the average particle size, is the standard deviation, which reflects the degree of dispersion of particle size distribution. is pi.

[0098] The combustion performance analysis module is used to screen the main components based on the composition data of the ash clean coal sample and the distribution information of each particle size interval, and analyze the direction and degree of contribution of each main component to the combustion performance index.

[0099] The specific analysis method of the main component screening module is: summarizing the component data and particle size distribution information of the gray clean coal sample, judging whether there are missing values ​​therein, if there are missing component data, making reasonable estimation and filling according to the statistical data of similar samples of the gray clean coal sample, if there is a missing proportion of a certain particle size interval in the distribution information of each particle size interval, performing interpolation calculation according to the adjacent particle size interval data to fill the missing value; when there is missing component data, making reasonable estimation and filling according to the statistical data of similar samples of the gray clean coal sample, the statistical data of similar samples reflect the distribution law of the component characteristics of similar coal samples, and using these data for estimation and filling can make the filling result of the missing value more in line with the actual situation, reduce the error caused by missing values, and improve the accuracy and reliability of the component data.

[0100] It should be noted that the missing values ​​are divided into missing component data and missing particle size interval ratio. The missing component data means that the content of a certain chemical component of the gray coal sample has not been measured or recorded, resulting in a null value for the data bit. The missing particle size interval ratio means that in the distribution information of each particle size interval, the relative content of particles in a specific particle size interval has not been counted or calculated, resulting in a missing situation.

[0101] Please refer to Table 1 for details, which lists some representative data.

[0102] Table 1. Composition data of some collected ash clean coal samples

[0103]

[0104] It should be noted that when missing values ​​are found in the component data, the range of samples belonging to the same type as the gray clean coal sample is determined, and for the determined similar samples, the mean and median of the component content of the item with the missing data value are calculated respectively.

[0105] The median arranges the content of the component in samples of the same type in ascending order, and takes the value in the middle position. If the number of samples is an even number, the average of the two middle numbers is taken.

[0106] Select appropriate statistics based on the characteristics of the data and actual situation. If the data distribution is relatively uniform, select the mean. If there are some outliers in the data, select the median. Fill the position where there are missing values ​​with the selected statistics to complete the filling of missing values ​​of the component data.

[0107] The filled data is standardized, the correlation coefficient matrix between the variables is calculated, and then the linear equations are solved for each eigenvalue to obtain the corresponding eigenvector, and the proportion of each eigenvalue to the total eigenvalue is calculated; by solving the eigenvector and calculating the proportion of the eigenvalue to the total eigenvalue, the original multiple correlated variables can be converted into a few independent principal components.

[0108] The proportion of each eigenvalue to the total eigenvalues ​​is sorted in descending order, and the proportion of each eigenvalue to the total eigenvalue is accumulated in sequence. The corresponding eigenvectors when the accumulation reaches the set proportion are recorded as principal components. The principal component can retain the information of the original data to the greatest extent, while reducing the data dimension and the complexity of the data, which facilitates subsequent data analysis and model construction and improves analysis efficiency.

[0109] It should be noted that for the data set after filling in the missing values, the mean and standard deviation of each variable are calculated respectively. The data of all variables are converted to a standard scale with a mean of 0 and a standard deviation of 1 through normalization. The correlation coefficient between the variables is obtained using the Pearson correlation coefficient. The variables are arranged in order to construct a correlation coefficient matrix.

[0110] For the correlation coefficient matrix, solve the characteristic equation to obtain each eigenvalue and its corresponding eigenvector, calculate the proportion of each eigenvalue to the total eigenvalues, and arrange the proportion of each eigenvalue to the total eigenvalues ​​in order from large to small.

[0111] A ratio threshold is set, and from the accumulated ratio values, the first corresponding eigenvector when the accumulated ratio is greater than or equal to the set ratio threshold is found, and the eigenvector is determined as each principal component.

[0112] The specific analysis method of the combustion performance analysis module is as follows: using the Pearson correlation coefficient to calculate the correlation coefficient between each principal component and the combustion performance index, setting the significance level, and using the t-test method to perform a significance test on the calculated correlation coefficient. When the calculated t-statistic is greater than the critical value of the t-distribution, it is considered that the correlation between the principal component and the combustion performance index is significant. After calculating the correlation coefficient, setting the significance level and using the t-test method to perform a significance test on the correlation coefficient helps to screen out the principal components that truly have a significant impact on the combustion performance.

[0113] The Pearson correlation coefficient formula is ,in, represents the principal component data column, A data column representing a combustion performance indicator, Indicates the The number of the data point, , The first Sample data, It is the average value of the data of the main component data column and the combustion performance index data column.

[0114] It should be noted that the specific method of the significance test is: determine a significance level, and use the formula for each calculated Pearson correlation coefficient ,in Indicates the number of samples, that is, the number of data points involved in the analysis, and calculates the corresponding Statistics used to measure whether the correlation between variables is statistically significant, based on the set significance level and degrees of freedom , check Distribution table, obtain the corresponding two-sided critical value, and calculate the Statistics and The critical value of the distribution is compared. The absolute value of the statistic is greater than If the critical value of the distribution is less than 0.05, the null hypothesis is rejected, and it is considered that the correlation between the principal component and the combustion performance index is significant; otherwise, the null hypothesis cannot be rejected, that is, it is considered that the correlation between the principal component and the combustion performance index is not significant.

[0115] It should be noted that middle, It is an adjustment term related to sample size used when calculating the t statistic. In the t distribution, the degrees of freedom , It is a treatment of the correlation coefficient R, which comprehensively reflects the impact of the degree of correlation on the calculation of the t statistic. It plays a scaling role on R, so that the t statistic can reasonably reflect the significance of the correlation.

[0116] The principal components with significant correlation with combustion performance indicators were screened out, a linear regression model was constructed, the regression model was fitted using the least squares method, and the regression coefficient was estimated. Based on the size and sign of the regression coefficient, the direction and degree of contribution of each principal component to the combustion performance indicators were analyzed.

[0117] It should be noted that the specific method for analyzing the direction and degree of contribution of each principal component to the combustion performance index is as follows: based on the results of the significance test, the principal components with significant correlation with the combustion performance index are selected, a linear regression model is constructed, and a set of regression coefficients is found using the least squares method to minimize the sum of squared errors between the observed values ​​and the predicted values. By taking partial derivatives of the sum of squared errors and setting these partial derivatives equal to 0, a system of equations is obtained, and solving this system of equations can obtain the estimated values ​​of the regression coefficients.

[0118] The correlation between the main components and the combustion performance is judged according to the contribution direction of each main component to the combustion performance index. The contribution direction includes positive and reverse directions. The positive direction indicates that the main component is positively correlated with the combustion performance, and the reverse direction indicates that the main component is negatively correlated with the combustion performance. For the main components that are positively correlated with the combustion performance, measures can be taken to increase their content in the coal sample or strengthen their effect. For the main components that are negatively correlated, efforts can be made to reduce their influence, thereby achieving effective regulation of the combustion performance of the gray clean coal and improving the combustion efficiency and quality of the coal.

[0119] It should be noted that the direction of the contribution of the principal component to the combustion performance index is judged based on the positive or negative value of the regression coefficient. If the regression coefficient is greater than 0, it means that the principal component is positively correlated with the combustion performance index, that is, an increase in the value of the principal component will lead to an increase in the value of the combustion performance index, and has a positive contribution to the combustion performance. If the regression coefficient is less than 0, it means that the corresponding principal component is negatively correlated with the combustion performance index, that is, an increase in the value of the principal component will lead to a decrease in the value of the combustion performance index, and has an inverse contribution to the combustion performance.

[0120] The absolute value of the regression coefficient reflects the contribution of the main component to the combustion performance index. The larger the absolute value, the greater the influence of the main component on the combustion performance index and the higher the contribution.

[0121] The evaluation feedback module is used to evaluate the combustion performance coefficient of the gray clean coal and dynamically optimize the flotation parameters based on the evaluation results.

[0122] The specific analysis method of the evaluation feedback module is: obtaining the combustion performance coefficient by multiplying each principal component by its corresponding regression coefficient and then adding them, setting the expected standard of the combustion performance coefficient, and adjusting the flotation parameters if the combustion performance coefficient is less than the expected standard of the combustion performance coefficient.

[0123] The principal components that are positively correlated and negatively correlated with the combustion performance are screened out respectively, and the lower limit adjustment threshold and the upper limit adjustment threshold are set for them.

[0124] If the content of a main component that is positively correlated with combustion performance is less than the lower adjustment threshold, the dosage of the flotation agent is adjusted; if the content of a main component that is negatively correlated with combustion performance is greater than the upper adjustment threshold, the flotation time is adjusted.

[0125] Within a fixed period of time after adjusting the flotation parameters, the ash clean coal sample is re-extracted for component analysis and combustion performance testing to obtain the adjusted combustion performance coefficient.

[0126] It should be noted that the adjustment amount of the flotation parameters was determined by conducting small-scale experiments using a trial-and-error method, with different adjustment gradients pre-set, wherein the flotation agent dosage was set to increase by 2%, 4%, and 6%, respectively, and the flotation time was set to extend by 5 minutes, 10 minutes, and 15 minutes, respectively. Subsequently, the changes in the content of each main component and the impact on the combustion performance indicators under the above-mentioned different adjustment conditions were observed, and the adjustment amount that could achieve the expected main component content and was most beneficial to the combustion performance was selected, thereby determining the optimal adjustment amount of the flotation agent dosage and flotation time.

[0127] The adjusted combustion performance coefficient is compared with the expected combustion performance coefficient standard. If the adjusted combustion performance coefficient is less than the expected combustion performance coefficient standard, the flotation parameters are adjusted again until the adjusted combustion performance coefficient reaches or exceeds the expected combustion performance coefficient standard.

[0128] The management database is used to store the composition data of the ash clean coal samples and the distribution information of each particle size interval of the ash clean coal samples.

[0129] See also Figure 3 As shown, in addition, the present invention provides a method for analyzing and detecting the components of gray clean coal, and the specific steps of the detection method are as follows: S1. Gray clean coal extraction: through the synergistic effect of compressed air and hydrophobic flocculation, high-purity gray clean coal is extracted from medium coal and other types of coal.

[0130] S2. Sample collection: Combine time interval sampling and spatial stratification sampling to obtain coal samples of each ash clean coal.

[0131] S3. Composition analysis: The composition data and particle size distribution information of the ash clean coal sample are obtained through multiple analysis methods.

[0132] S4. Combustion performance analysis: Based on the composition data of the ash clean coal sample and the distribution information of each particle size interval, the main components are screened and the direction and degree of contribution of each main component to the combustion performance index are analyzed.

[0133] S5. Evaluation feedback: Evaluate the combustion performance coefficient of the clean coal and dynamically optimize the flotation parameters based on the evaluation results.

[0134] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered by the scope of protection of the present invention.

Claims

1. A gray clean coal component analysis and detection system, characterized in that: The system specifically includes the following modules: Gray clean coal extraction module, used to extract high-purity gray clean coal from medium coal through the synergistic effect of compressed air and hydrophobic flocculation; The sample collection module is used to combine time interval sampling and spatial stratification sampling to obtain coal samples of various ash clean coals; The component analysis module is used to obtain the component data and particle size distribution information of the gray clean coal sample through multiple analysis methods; The combustion performance analysis module is used to screen the main components based on the composition data of the ash clean coal sample and the distribution information of each particle size interval, and analyze the direction and degree of contribution of each main component to the combustion performance index; Evaluation feedback module, used to evaluate the combustion performance coefficient of clean coal and dynamically optimize flotation parameters based on the evaluation results; A management database for storing the composition data of clean coal samples and the distribution information of each particle size interval of clean coal samples; The specific analysis method of the evaluation feedback module is: The combustion performance coefficient is obtained by multiplying each principal component by its corresponding regression coefficient and then adding them together, and an expected standard of the combustion performance coefficient is set. If the combustion performance coefficient is less than the expected standard of the combustion performance coefficient, the flotation parameters are adjusted; The principal components that are positively correlated and negatively correlated with the combustion performance are screened out respectively, and the lower limit adjustment threshold and the upper limit adjustment threshold are set for them; If the content of a main component that is positively correlated with combustion performance is less than the lower adjustment threshold, the amount of flotation agent is adjusted; if the content of a main component that is negatively correlated with combustion performance is greater than the upper adjustment threshold, the flotation time is adjusted; Within a fixed period of time after adjusting the flotation parameters, the gray clean coal sample is re-extracted, and component analysis and combustion performance test are performed to obtain the adjusted combustion performance coefficient; The adjusted combustion performance coefficient is compared with the expected combustion performance coefficient standard. If the adjusted combustion performance coefficient is less than the expected combustion performance coefficient standard, the flotation parameters are adjusted again until the adjusted combustion performance coefficient reaches or exceeds the expected combustion performance coefficient standard.

2. The gray clean coal component analysis and detection system according to claim 1, characterized in that: The specific operation method of extracting gray clean coal is: After pre-treatment, the collected middling coal is added to the flotation tank together with the flotation agent to form flocs through hydrophobic flocculation; Air is introduced into the flotation tank through a porous distributor, so that the air forms a large number of uniform microbubbles in the slurry. The microbubbles fully collide and adhere with the hydrophobically flocculated medium coal flocs, and the buoyancy of the microbubbles drives the flocs to float to the surface of the slurry, thereby forming a foam layer. The foam layer is scraped off with a scraper device at a set speed and frequency to collect the foam product containing gray clean coal. The foam product is sequentially dehydrated to remove excess water, and then subjected to secondary flotation purification to further separate impurities, ultimately obtaining high-purity gray clean coal.

3. The gray clean coal component analysis and detection system according to claim 2, characterized in that: The specific analysis method of the sample collection module is: For the gray clean coal conveyed by the belt, a sampling time interval is set, and the gray clean coal samples at different times during the belt conveying process are obtained using a belt middle sampler according to the time interval; For the stockpiled clean coal, the stockpiled clean coal is divided into different heights, and multiple sampling points are set at each height. Layered sampling is carried out from each sampling point at each height to obtain samples of the stockpiled clean coal at different spatial positions; The coal samples obtained by the two methods of time interval sampling during belt conveying and spatial stratification sampling of stockpiled clean ash coal are comprehensively mixed to obtain a total sample, and the total sample is evenly divided into several clean ash coal samples.

4. The gray clean coal component analysis and detection system according to claim 1, characterized in that: The specific method for obtaining the composition data of the gray clean coal sample is: Weigh a portion of the clean coal sample, place it in a crucible after pretreatment, record the initial total mass of the clean coal sample and the crucible, place it in the sample cell of the thermogravimetric analyzer, start the heating program, and raise the temperature from room temperature to the set moisture detection temperature. After a fixed period of time, record the total mass of the clean coal sample and the crucible. Calculate the moisture content of the clean coal sample by calculating the decrease in the initial total mass of the clean coal sample and the crucible; Continue to raise the temperature to the set volatilization detection temperature in the same way, and calculate the volatile matter content of the gray clean coal sample; The ash content of the coal sample is calculated based on the total mass of the final ash clean coal sample residue and the crucible, and then the fixed carbon content of the ash clean coal sample is obtained by subtracting the moisture content, volatile matter content and ash content from the initial total mass of the coal sample components.

5. The gray clean coal component analysis and detection system according to claim 4, characterized in that: The specific method for obtaining the composition data of the gray clean coal sample also includes: Obtaining data on the content of various components of a large number of standard coal samples, collecting coal sample spectra of the standard coal samples through a near-infrared spectrometer, and using the partial least squares method to establish a quantitative model of near-infrared spectra and coal sample components; A sample of clean ash coal is taken, and its near-infrared spectrum is collected by a near-infrared spectrometer. The collected spectral data is input into a quantitative model of near-infrared spectrum and coal sample composition to predict the component content of the clean ash coal sample.

6. The gray clean coal component analysis and detection system according to claim 5, characterized in that: The specific method for obtaining the distribution information of each particle size interval is: Taking a gray clean coal sample, adding it to the sample cell of the laser particle size analyzer, allowing the sample to be fully dispersed in the dispersion medium, starting the laser particle size analyzer, and irradiating the laser beam of the instrument onto the gray clean coal sample in the sample cell; The angle and intensity distribution of scattered light generated by particles of different particle sizes to the laser are measured, and the collected scattered light intensity data at different angles are substituted into the Mie scattering theory formula. The particle size distribution model parameters are adjusted through iterative calculation to achieve the best fit between the calculated and measured scattered light intensity distributions. The particle size distribution model that best fits the measured data is determined to obtain the distribution information of each particle size interval.

7. The gray clean coal component analysis and detection system according to claim 1, characterized in that: The specific analysis method for screening the main components is: Summarize the composition data and particle size distribution information of the ash clean coal sample to determine whether there are missing values. If there are missing composition data, make reasonable estimates based on the statistical data of similar samples of the ash clean coal sample. If there is a missing proportion of a certain particle size interval in the particle size distribution information, interpolate the data of adjacent particle size intervals to fill in the missing value. The filled data are standardized, the correlation coefficient matrix between the variables is calculated, and then the linear equation system is solved for each eigenvalue to obtain the corresponding eigenvector, and the proportion of each eigenvalue to the total eigenvalue is calculated; The proportion of each eigenvalue to the total eigenvalue is sorted in descending order, the proportion of each eigenvalue to the total eigenvalue is accumulated in sequence, and the corresponding eigenvectors when the accumulation reaches the set proportion are recorded as principal components.

8. The gray clean coal component analysis and detection system according to claim 7, characterized in that: The specific analysis method of the combustion performance analysis module is: The Pearson correlation coefficient was used to calculate the correlation coefficient between each principal component and the combustion performance index. The significance level was set and the calculated correlation coefficient was tested for significance using the t-test method. When the calculated t-statistic was greater than the critical value of the t-distribution, the correlation between the principal component and the combustion performance index was considered significant. The principal components with significant correlation with combustion performance indicators were screened out, a linear regression model was constructed, and the regression model was fitted using the least squares method to estimate the regression coefficient. Based on the size and sign of the regression coefficient, the contribution direction and degree of each principal component to the combustion performance indicator were analyzed; The correlation between the principal components and the combustion performance is determined based on the contribution direction of each principal component to the combustion performance index. The contribution direction includes positive and negative directions. A positive direction indicates that the principal component is positively correlated with the combustion performance, and a negative direction indicates that the principal component is negatively correlated with the combustion performance.

9. A method for analyzing and detecting the composition of clean coal, performed by the system for analyzing and detecting the composition of clean coal according to any one of claims 1 to 8, characterized in that: The specific steps of this detection method are as follows: S1. Ash Clean Coal Extraction: Through the synergistic effect of compressed air and hydrophobic flocculation, high-purity ash clean coal is extracted from medium coal; S2. Sample collection: Combine time interval sampling and spatial stratification sampling to obtain coal samples of each ash clean coal; S3. Composition Analysis: Analyze the composition data and particle size distribution of the clean coal sample using a multi-analysis method. S4. Combustion Performance Analysis: Based on the compositional data and particle size distribution information of the clean coal sample, key principal components are selected. The correlation between the principal components and combustion performance indicators is calculated. The principal components are further selected and the direction and degree of contribution of each principal component to the combustion performance indicators are analyzed. S5. Evaluation feedback: Evaluate the combustion performance coefficient of the clean coal, classify the combustion performance level, and dynamically optimize the flotation parameters based on the evaluation results.

Images