Grinding and waste gas treatment linkage control method and system based on fineness online detection

Abstract

This invention discloses a method and system for the coordinated control of a vertical mill and exhaust gas treatment based on online fineness detection, belonging to the field of automation control technology. The method acquires time-series detection values ​​of the finished product fineness, generates fineness evolution trend parameters and fineness response state parameters; identifies the causes of fineness deviations by combining system airflow, mill internal pressure difference, classifier speed, and return material volume; constructs an exhaust gas transport capacity index based on the cause type, and calculates the estimated load of ultrafine particles entering the dust removal system in the future; generates and executes airflow, classifier, and dust removal adjustment commands based on the cause type, exhaust gas transport capacity index, and ultrafine particle load estimated value; and switches the operating mode according to the changing trend of the exhaust gas transport capacity index and ultrafine particle load estimated value within a preset period. This invention achieves adaptive coordinated control of multiple systems, including vertical mill fineness, exhaust gas transport, and dust removal.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, and in particular to a method and system for the linkage control of vertical mill and exhaust gas treatment based on online fineness detection. Background Technology

[0002] Vertical roller mills, widely used in modern cement, slag, steel slag, and non-metallic mineral powder preparation processes, integrate grinding, grading, and drying. During operation, materials are compressed and ground between the grinding disc and rollers, forming particles of varying sizes. These particles then enter the classifying zone under airflow for particle grading. Finished particles meeting size requirements are transported with exhaust gas to the subsequent dust collection system for gas-solid separation, while coarse particles are returned to the mill for further grinding. Therefore, product fineness, airflow conditions, and dust collection operation status collectively constitute crucial components of the continuous operation of a vertical roller mill system. With the development of online monitoring, process control, and industrial automation technologies, real-time monitoring of the vertical roller mill system's operating status using online particle size detection, airflow monitoring, differential pressure monitoring, and dust collection parameter monitoring, along with adjustments to classifying, airflow, and dust collection parameters based on the monitoring results, has become a common control method in grinding production processes.

[0003] Existing vertical mill fineness control and exhaust gas treatment control mostly adopt independent adjustment methods. Due to the lack of cross-system state correlation analysis and collaborative decision-making mechanism, system mismatch problems occur, such as fineness control meeting the standard but dust removal load deteriorating, or dust removal operation being normal but fineness being out of control. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides a method and system for the linkage control of vertical mill and exhaust gas treatment based on online fineness detection. It aims to improve the existing vertical mill fineness control and exhaust gas treatment control, which mostly adopt independent adjustment methods, easily leading to system mismatch problems such as fineness control meeting the standard but dust removal load deteriorating, or dust removal operation being normal but fineness control being out of control.

[0005] This invention provides the following technical solution: a method for coordinated control of a vertical mill and exhaust gas treatment based on online fineness detection, comprising the following steps: S1. Obtain the time-series detection values ​​of the fineness of the finished product, generate fineness evolution trend parameters and fineness response state parameters, and obtain the system air volume, mill internal pressure difference, classifier speed and return material amount; S2. Based on the correspondence between fineness evolution trend parameters, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount, the cause type of fineness deviation is identified by comparing the distance between the deviation feature vector composed of the above parameters and the reference feature vector corresponding to each cause type. S3. Based on the cause type, system air volume, mill internal pressure difference, return material amount and fineness evolution trend parameters, and combined with the cause type weight, construct the exhaust gas transport capacity index, and calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume and fineness evolution trend parameters. The estimated load of ultrafine particles is used to characterize the amount of ultrafine particles that will enter the dust removal system after a preset time under the current working conditions. S4. Using the type of formation, the exhaust gas transport capacity index, and the estimated load of ultrafine particles as input conditions, determine the air volume adjustment command, the powder selection adjustment command, and the dust removal adjustment command, and execute the air volume adjustment command, the powder selection adjustment command, and the dust removal adjustment command. S5. Obtain the changing trend of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period, and switch between multiple preset operating modes according to the changing trend.

[0006] By adopting the above technical solution, fineness evolution trend parameters and fineness response state parameters are generated by acquiring the time-series detection values ​​of finished product fineness. Combined with system air volume, mill internal pressure difference, classifier speed and return material amount, the cause type of fineness deviation is identified. Furthermore, the exhaust gas transport capacity index and ultrafine particle load prediction value are constructed. Based on the cause type, exhaust gas transport capacity index and ultrafine particle load prediction value, air volume adjustment command, classifier adjustment command and dust removal adjustment command are determined. In this way, the linkage control relationship between fineness state, transport state and dust removal state is established, thereby improving the traditional vertical mill fineness control and exhaust gas treatment control, which mostly adopt independent adjustment methods. Due to the lack of cross-system state correlation analysis and collaborative decision-making mechanism, the system mismatch problem is that fineness control meets the standard but the dust removal load deteriorates or the dust removal operation is normal but the fineness is out of control.

[0007] Further, in S1, the step of generating the fineness evolution trend parameters and the fineness response state parameters includes: A fineness sequence within a continuous time window is constructed based on the time-series detection values ​​of the finished product fineness, and the fineness change rate, fineness change acceleration, and fineness fluctuation characteristics are calculated based on the fineness sequence. Fineness evolution trend parameters are generated based on the fineness change rate, fineness change acceleration, and fineness fluctuation characteristics, wherein the fineness evolution trend parameters according to Calculation generates, where For the rate of change of fineness, For the acceleration of fineness change, It is characterized by fineness fluctuation. For the preset weighting coefficients, satisfy ; Obtain at least one of the following within a preset time range: system air volume adjustment record, air classifier speed adjustment record, and circulating fan frequency adjustment record; The fineness response state parameters are generated based on the change in fineness evolution trend parameters before and after the occurrence of the control action.

[0008] Further, in S2, the step of identifying the cause type of fineness deviation includes: A combination of deviation characteristics is constructed based on parameters of fineness evolution trend, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount; The parameters in the deviation feature combination are normalized, and a deviation feature vector is generated based on the normalized parameters. Calculate the Euclidean distance between the deviation feature vector and the reference feature vector corresponding to each cause type; The cause type corresponding to the minimum Euclidean distance is identified as the cause type of the fineness deviation.

[0009] Furthermore, in S3, the step of constructing the exhaust gas transport capacity index includes: The airflow transport characteristics are generated based on the system air volume and the pressure difference inside the mill. Particle recycling characteristics are generated based on parameters related to the amount of recycled material and the evolution trend of fineness. Construct a set of related features based on airflow transport characteristics, particle circulation characteristics, and genetic types; The waste gas transport capacity index is generated based on the set of associated features.

[0010] Furthermore, in S3, the step of calculating the estimated load of ultrafine particles entering the dust removal system in the future includes: Based on the fineness evolution trend parameters, an ultrafine particle generation trend is generated to characterize the ability to generate ultrafine particles; Based on the exhaust gas transport capacity index and system air volume, an ultrafine particle transport trend is generated to characterize the ability of ultrafine particles to be transported to the dust removal system. A set of features for predicting particle load is constructed based on the generation and transport trends of ultrafine particles. The estimated value of ultrafine particle load entering the dust removal system in the future is calculated based on the particle load prediction feature set.

[0011] Furthermore, in S4, the step of using the causal type, exhaust gas transport capacity index, and estimated ultrafine particle load as input conditions includes: The causal types are mapped to quantified values ​​to generate causal demand features; The exhaust gas transport capacity index is normalized to generate transport status characteristics; The estimated load of the ultrafine particles is normalized to generate load state characteristics; A set of control decision features is constructed based on the characteristics of cause and demand, transportation status, and load status.

[0012] Furthermore, in S4, the steps of determining the airflow adjustment command, the powder selection adjustment command, and the dust removal adjustment command include: Establish the correspondence between the set of control decision features and the adjustment actions, and generate a set of control rules; The set of control decision features is matched with the set of control rules to determine the target adjustment combination; Generate air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands based on the target adjustment combination; Execute air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands.

[0013] Furthermore, in S5, the step of switching between multiple preset operating modes based on the changing trend includes: The operating trend combination is constructed based on the changing trend of the exhaust gas transport capacity index and the changing trend of the ultrafine particle load estimate within the preset period. Operational status features are generated based on a combination of operational trends; Match the running status characteristics with the mode switching conditions corresponding to multiple preset running modes; Among them, the multiple preset operating modes include quality priority mode, energy consumption priority mode and dust removal safety priority mode; The execution mode is switched according to the target operating mode.

[0014] This invention also provides a vertical mill and exhaust gas treatment linkage control system based on online fineness detection, including: The data acquisition and processing module is connected to the online particle size analyzer of the finished product and the recording unit of the DCS system. It is configured to acquire the time-series detection values ​​of the fineness of the finished product and generate fineness evolution trend parameters and fineness response state parameters. The cause identification module is configured to identify the cause type of fineness deviation by comparing the distance between the deviation feature vector composed of the above parameters and the reference feature vector corresponding to each cause type, based on the correspondence between the fineness evolution trend parameters, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount. The load prediction module is configured to construct an exhaust gas transport capacity index based on the cause type, system air volume, mill internal pressure difference, return material volume and fineness evolution trend parameters, combined with the cause type weight, and calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume and fineness evolution trend parameters. The linkage control module is electrically connected to the frequency converter of the circulating fan, the drive unit of the classifier and the pulse jet control unit of the dust collector. It is configured to use the cause type, the exhaust gas transport capacity index and the estimated load of ultrafine particles as input conditions to determine the air volume adjustment command, the classifier adjustment command and the dust removal adjustment command, and execute the air volume adjustment command, the classifier adjustment command and the dust removal adjustment command. The mode management module is configured to acquire the changing trends of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period, and to switch between multiple preset operating modes based on the changing trends.

[0015] The present invention has the following beneficial effects: 1. In this invention, fineness evolution trend parameters and fineness response state parameters are generated by acquiring the time-series detection values ​​of finished product fineness. Combined with system air volume, mill internal pressure difference, classifier speed and return material amount, the cause type of fineness deviation is identified. Furthermore, the exhaust gas transport capacity index and ultrafine particle load prediction value are constructed. Based on the cause type, exhaust gas transport capacity index and ultrafine particle load prediction value, air volume adjustment command, classifier adjustment command and dust removal adjustment command are determined. In this way, the linkage control relationship between fineness state, transport state and dust removal state is established, thereby improving the traditional vertical mill fineness control and exhaust gas treatment control, which mostly adopt independent adjustment methods. Due to the lack of cross-system state correlation analysis and collaborative decision-making mechanism, the system mismatch problem is that fineness control meets the standard but the dust removal load deteriorates or the dust removal operation is normal but the fineness is out of control.

[0016] 2. In this invention, an exhaust gas transport capacity index is constructed based on parameters such as system air volume, mill internal pressure difference, return material quantity, and fineness evolution trend. Based on the exhaust gas transport capacity index, system air volume, and fineness evolution trend parameters, the estimated load of ultrafine particles entering the dust removal system in the future is calculated. Thus, information on dust removal load changes is obtained before ultrafine particles enter the dust removal system, thereby improving the problem that traditional exhaust gas treatment control mostly adopts a feedback adjustment method based on the current dust removal operating status. Due to the lack of advance analysis of particle load changes entering the dust removal system, the dust removal system is in a passive response state for a long time.

[0017] 3. In this invention, by obtaining the changing trend of the waste gas transport capacity index and the estimated load of ultrafine particles within a preset period, and performing mode switching between multiple preset operating modes according to the changing trend, the corresponding operating mode is adjusted according to the changes in the system operating status. This improves the problem that traditional vertical mills and waste gas treatment systems mostly adopt fixed operating modes, and the operating mode cannot be adjusted with changes in fineness transport status and particle load status, resulting in high system energy consumption. Attached Figure Description

[0018] Figure 1This is a schematic flowchart of the vertical mill and exhaust gas treatment linkage control method based on online fineness detection proposed in this invention. Figure 2 This is a schematic diagram of the fineness deviation cause identification process in the vertical mill and exhaust gas treatment linkage control method based on online fineness detection proposed in this invention. Figure 3 This is a schematic diagram of the process of constructing the waste gas transport capacity index and calculating the estimated load of ultrafine particles in the vertical mill and waste gas treatment linkage control method based on online fineness detection proposed in this invention. Figure 4 This is a schematic diagram of the linkage control decision generation process in the linkage control method for vertical mill and exhaust gas treatment based on online fineness detection proposed in this invention. Figure 5 This is a schematic diagram of the operation mode switching process in the vertical mill and exhaust gas treatment linkage control method based on online fineness detection proposed in this invention. Detailed Implementation

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] In embodiments of the present invention, a method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection is provided, such as... Figure 1 As shown, the process includes the following steps: S1, acquiring time-series detection values ​​of finished product fineness, generating fineness evolution trend parameters and fineness response state parameters, and acquiring system airflow, mill internal pressure difference, classifier speed, and return material quantity; the system airflow is collected in real time by an airflow detection unit installed on the circulating air duct; the mill internal pressure difference is collected in real time by a pressure difference transmitter installed on the airflow channel inside the mill; the classifier speed is collected in real time by a classifier frequency converter or speed sensor; the return material quantity is collected in real time by a return material conveying and metering device, which includes at least one of a belt scale, a bucket elevator current metering unit, or a return material flow meter.

[0021] Furthermore, in S1, the steps for generating the fineness evolution trend parameters and the fineness response state parameters include: A fineness sequence within a continuous time window is constructed based on the time-series detection values ​​of the finished product fineness, and the fineness change rate, fineness change acceleration, and fineness fluctuation characteristics are calculated based on the fineness sequence. Fineness evolution trend parameters are generated based on the fineness change rate, fineness change acceleration, and fineness fluctuation characteristics. Obtain at least one of the following within a preset time range: system air volume adjustment record, air classifier speed adjustment record, and circulating fan frequency adjustment record; The fineness response state parameters are generated based on the change in fineness evolution trend parameters before and after the occurrence of the control action.

[0022] Specifically, firstly, the time-series detection values ​​of the fineness of the finished product are continuously acquired, and a fineness sequence is formed according to the time order. The fineness detection values ​​are obtained from online particle size analysis, preferably using the sieve residue value or characteristic particle size value of the finished product particles as the fineness characterization quantity. In this embodiment, this is denoted as... , indicating the first The fineness detection value corresponding to each sampling time, the sampling time is denoted as . Constructing a fine-grained sequence within a continuous time window using a sliding time window method. In this scheme, the time window is preferably set to continuous sampling data within the most recent 5 to 15 minutes; the fineness change rate is calculated based on the fineness detection values ​​at adjacent time points, and the fineness change rate is denoted as... The calculation method is as follows: ; in, This represents the detail detection value at the current moment. This indicates the detail detection value at the previous moment. This represents the time interval between adjacent sampling points; further, the fineness change acceleration is calculated based on the fineness change rate at two consecutive moments, and the fineness change acceleration is denoted as... The calculation method is as follows: ; in, This represents the rate of change of fineness at the current moment. , This represents the rate of change of fineness at the previous moment. Then, the fluctuation characteristics of the fineness detection values ​​within the current time window are calculated. The fluctuation characteristics are represented by the coefficient of variation, denoted as . The calculation method is as follows: ; in, This represents the standard deviation of the fineness measurement values ​​within the time window. This represents the average value of the fineness detection within the time window; it also represents the rate of change of fineness. Acceleration of fineness change and fluctuation characteristics Construct the fineness evolution trend parameter, denoted as . The calculation method is as follows: ; in, For the preset weighting coefficients, satisfy In one implementation, Take 0.5, Take 0.3, Take 0.2; It is understandable that the specific values ​​of the aforementioned weighting coefficients can be determined based on the actual operating conditions of the vertical mill and the degree of attention paid to the characteristics of different fineness variations, while satisfying the following conditions. Adaptive adjustments are made under constraints. Fineness evolution trend parameters are used to characterize the development direction and degree of change of the current fineness state, and serve as the data basis for subsequent identification of the causes of fineness deviations; when As the size increases, the characterization becomes coarser. As the value continues to decrease, the characterization becomes finer. When the fluctuation exceeds a preset threshold, it indicates a fine oscillation state. and When all parameters are close to zero, the fineness is in a stable state. After obtaining the fineness evolution trend parameters, further operational adjustment records within a preset time range are acquired. These records include at least one of the following: system airflow adjustment records, classifier speed adjustment records, and circulating fan frequency adjustment records. For each control action record, the time of adjustment is recorded. And extract the fineness evolution trend parameters corresponding to the time window before the regulation occurs. Fineness evolution trend parameters corresponding to the time window after the adjustment occurs A trend response quantity is constructed based on the difference in the fineness evolution trend parameters before and after adjustment, and the trend response quantity is denoted as... The calculation method is as follows: ; in, This represents the parameter indicating the trend of fineness evolution after adjustment. This represents the fineness evolution trend parameters obtained before adjustment; further, the trend response quantities corresponding to multiple adjustment events are statistically analyzed, and fineness response state parameters are generated based on the magnitude of change in the trend response quantities. In one embodiment, the fineness response state parameters... The calculation method is as follows: ; in, To control the number of times the action is recorded within a preset time range, For the first The trend response quantity corresponding to the next control action; when A response level above the first threshold corresponds to a high response state. A value below the second threshold corresponds to a low response state, while other values ​​correspond to a medium response state. The fineness response state parameter describes the degree of response of the fineness state to operational adjustment behavior and serves as input data in the subsequent fineness deviation cause identification process. A higher response state corresponds to a sustained large change in the trend response value; a lower response state corresponds to a small change in the trend response value; and a lower response state is indicated when the trend response value remains at a low level after multiple consecutive adjustments. This indicates that the current fineness state is weakly responsive to operational adjustment behavior. This fineness response state parameter quantifies the dynamic sensitivity of fineness to control actions in a physical sense. When the FS value is below the preset second threshold, it indicates that the system has high inertia or that the response is sluggish due to abnormal operating conditions. In this case, conventional feedback adjustment may not be effective, and a more predictive feedforward strategy needs to be introduced in the subsequent cause identification and control decision-making. The generated fineness evolution trend parameter and fineness response state parameter are then input into the fineness deviation cause identification process, forming the data basis for fineness deviation analysis together with system airflow, mill internal pressure difference, classifier speed, and return material amount.

[0023] like Figure 2 As shown in Figure S2, based on the correspondence between fineness evolution trend parameters, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount, identify the types of causes of fineness deviation; Furthermore, in S2, the step of identifying the cause type of fineness deviation includes: A combination of deviation characteristics is constructed based on parameters of fineness evolution trend, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount; The parameters in the deviation feature combination are normalized, and a deviation feature vector is generated based on the normalized parameters. Calculate the Euclidean distance between the deviation feature vector and the reference feature vector corresponding to each cause type; The cause type corresponding to the minimum Euclidean distance is identified as the cause type of the fineness deviation.

[0024] Specifically, after obtaining the fineness evolution trend parameters, the main causes of fineness deviations are further identified by combining system airflow, mill pressure difference, classifier speed, and return material volume. The fineness evolution trend parameters characterize the development direction and change characteristics of the current fineness state, while the fineness response state parameters characterize the degree of response of the fineness state to operational adjustments. The system airflow is denoted as... This represents the gas flow rate through the vertical mill system per unit time, with data derived from airflow measurement results; the internal pressure difference is recorded as... This represents the pressure difference generated during the airflow process inside the mill; the data comes from pressure difference detection results. The classifier speed is denoted as... This indicates the operating speed of the classifier impeller, data sourced from the frequency converter operation record; the return material quantity is recorded as... This represents the material flow rate returning to the mill for further grinding per unit time. The data comes from the statistical results of the return material conveying volume. First, the fineness evolution trend parameters, fineness response state parameters, and system air volume are included. Differential pressure inside the mill Air classifier speed and the amount of recycled materials Synchronous correlation is performed at the same time to form a combination of deviation characteristics; in this scheme, to avoid direct comparison of parameters with different dimensions, each parameter is normalized, and the normalization result is denoted as... The calculation method is as follows: ; in, This indicates the current value of the parameter to be processed. This indicates the minimum value of the corresponding parameter within the statistical period. This represents the maximum value of the corresponding parameter within the statistical period; after normalization, it forms the deviation eigenvector: ; in, This represents the normalized result corresponding to the fineness evolution trend parameter. This represents the normalized result corresponding to the fineness response state parameters. This represents the normalized result of the system air volume. This represents the normalized result of the internal pressure difference in the mill. This indicates the normalized result of the air classifier speed. This represents the normalized result of the returned material quantity. Subsequently, a fineness deviation causal feature library is established, containing reference feature vectors corresponding to insufficient transport, insufficient powder selection, insufficient grinding, and combined coupling types. In this scheme, the reference feature vectors are obtained by: during system debugging or self-learning phases, data is collected by artificially simulating typical abnormal operating conditions. For example, under the premise of ensuring safety, the "insufficient transport" condition is simulated by gradually reducing the frequency of the circulating fan. The complete combination of deviation features, including fineness evolution trend parameters, system airflow, and mill internal pressure difference, is collected and recorded in real time. After multiple experiments, the normalized mean vector is taken and stored in the database as the initial reference feature vector for this causal type. During long-term operation, the system can continuously optimize and update the reference feature vectors corresponding to each causal type by combining manual annotation or automatic identification, thereby continuously improving the accuracy of diagnosis. Among them, the reference feature vector corresponding to the insufficient transport type has a lower normalized value for system air volume and a higher normalized value for return material volume; the reference feature vector corresponding to the insufficient classifier type has a lower normalized value for classifier speed; the reference feature vector corresponding to the insufficient grinding type has a lower normalized value for mill internal pressure difference; the reference feature vector corresponding to the comprehensive coupling type is composed of abnormal feature combinations of parameters corresponding to the aforementioned types; the feature distance between the deviation feature vector and each reference feature vector is calculated, and the feature distance is denoted as... Using Euclidean distance, the calculation method is as follows: ; in, The first eigenvector represents the deviation eigenvector. 1 eigenvalue, Indicates the first The first in the class reference feature vector 1 eigenvalue, This indicates the categories of causes for different levels of detail deviation; after calculating multiple feature distances, the category corresponding to the smallest feature distance is selected as the matching result: ; in, The final matching category is indicated; the cause type of fineness deviation is determined based on the matching result; when the matching result corresponds to insufficient transport type, it indicates that the fineness deviation is mainly related to changes in airflow transport capacity; when the matching result corresponds to insufficient powder selection type, it indicates that the fineness deviation is mainly related to changes in powder selection and classification capacity; when the matching result corresponds to insufficient grinding type, it indicates that the fineness deviation is mainly related to changes in grinding state within the mill; when the matching result corresponds to comprehensive coupling type, it indicates that multiple operating factors participate in the formation of fineness deviation simultaneously; the obtained cause type of fineness deviation serves as important input data for the subsequent construction process of the exhaust gas transport capacity index, and together with system airflow, mill pressure difference, return material amount, and fineness evolution trend parameters, participates in the calculation of the exhaust gas transport capacity index, thereby establishing the correlation between fineness state and exhaust gas transport state.

[0025] like Figure 3 As shown, S3, construct the exhaust gas transport capacity index based on the cause type, system air volume, mill internal pressure difference, return material amount and fineness evolution trend parameters, and calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume and fineness evolution trend parameters. The estimated load of ultrafine particles is used to characterize the amount of ultrafine particles that will enter the dust removal system after a preset time under the current working conditions. Furthermore, in S3, the steps for constructing the exhaust gas transport capacity index include: The airflow transport characteristics are generated based on the system air volume and the pressure difference inside the mill. Particle recycling characteristics are generated based on parameters related to the amount of recycled material and the evolution trend of fineness. Construct a set of related features based on airflow transport characteristics, particle circulation characteristics, and genetic types; The waste gas transport capacity index is generated based on the set of associated features.

[0026] Specifically, after identifying the causes of fineness deviation, a waste gas transport capacity index is further constructed to characterize the current waste gas flow's ability to carry and transport fine particles, and serves as input data for subsequent calculations of ultrafine particle load estimates; whereby the system airflow is denoted as... This represents the gas flow rate through the vertical mill system per unit time, with data sourced from airflow monitoring results; the internal pressure difference is recorded as... This represents the pressure difference generated during the airflow process inside the mill; the data comes from pressure difference detection results. The return material quantity is recorded as... This represents the particle flow rate returning to the mill for further grinding per unit time; the data is derived from statistics on return material conveying volume. The fineness evolution trend parameter is denoted as... The results are derived from the analysis of the fineness time series detection values; firstly, based on the system air volume... With the pressure difference inside the mill The airflow transport characteristics are generated; since the system air volume reflects the gas transport capacity and the internal pressure difference reflects the flow resistance of the airflow through the internal space of the mill, the ratio of the two is used to characterize the airflow transport capacity per unit flow resistance. The airflow transport characteristics are denoted as... The calculation method is as follows: ; in, The larger the value, the stronger the airflow transport capacity per unit pressure difference; subsequently, based on the return material volume... With fineness evolution trend parameters The particle recycling feature is generated; the amount of recycled material reflects the scale of particles that are not carried out in time, and the fineness evolution trend parameter reflects the direction of change in particle size state. Therefore, these two parameters are used to construct the particle recycling feature, which is denoted as [missing information]. The calculation method is as follows: ; in, The baseline return material quantity is the average value of the return material quantity during the stable operation phase of the system. The stable operation phase can be determined based on operating parameters such as system airflow, mill pressure differential, and finished product fineness. For example, a stable operation phase is defined as a period of two consecutive hours where the finished product fineness meets the standard and the fluctuation range is below a preset threshold, while the system airflow and mill pressure differential remain stable. The arithmetic mean of the return material quantity during this period is then taken as the baseline return material quantity. . This represents the absolute value of the fineness evolution trend parameter, used to eliminate the influence of the trend direction on the calculation results; when the amount of recycled material increases and the fineness change trend weakens, the particle circulation characteristic increases accordingly; subsequently, based on the airflow transport characteristics... Particle cycle characteristics and the types of causes Construct a set of associated features; in this scheme, different causal types correspond to different influence weights, and the causal weight corresponding to the insufficient transport type is denoted as... The higher the weight of the cause corresponding to insufficient powder selection, the more significant the cause. Secondly, the causal weight corresponding to insufficient grinding is denoted as... The causal weights for the comprehensive coupling type are relatively low, and are determined based on historical operational statistics. In one implementation, the causal weights preferably use normalized weights in the calculation. The waste gas transport capacity index is generated based on the associated feature set, and is denoted as... The calculation method is as follows: ; in, It is a dimensionless evaluation index used to comprehensively reflect the actual transport capacity of the current exhaust gas system for fine particles; when the airflow transport characteristics improve, Increase; when the particle cycle characteristics improve, Reduce; after obtaining the exhaust gas transport capacity index, combine the exhaust gas transport capacity index with the system air volume. and fineness evolution trend parameters The process of predicting ultrafine particle load is input together; among them, the exhaust gas transport capacity index is used to characterize the ability of exhaust gas to carry fine powder particles, the system air volume is used to characterize the particle transport scale, and the fineness evolution trend parameter is used to characterize the fine powder generation trend. The three together serve as the basis for calculating the estimated value of ultrafine particle load entering the dust removal system in the future, thereby establishing a quantitative correlation between fineness state and exhaust gas transport state.

[0027] Furthermore, in S3, the steps for calculating the estimated load of ultrafine particles entering the dust removal system in the future include: Based on the fineness evolution trend parameters, an ultrafine particle generation trend is generated to characterize the ability to generate ultrafine particles; Based on the exhaust gas transport capacity index and system air volume, an ultrafine particle transport trend is generated to characterize the ability of ultrafine particles to be transported to the dust removal system. A set of features for predicting particle load is constructed based on the generation and transport trends of ultrafine particles. The estimated value of ultrafine particle load entering the dust removal system in the future is calculated based on the particle load prediction feature set.

[0028] Specifically, in obtaining the exhaust gas transport capacity index Then, combined with the fineness evolution trend parameters With system air volume Calculate the estimated load of ultrafine particles entering the dust removal system in the future; among which, the fineness evolution trend parameter Derived from the analysis results of time-series fineness detection values, used to characterize the current generation state of ultrafine particles; exhaust gas transport capacity index Derived from the evaluation results of exhaust gas transport capacity, it is used to characterize the ability of exhaust gas to carry fine particles; system air volume Derived from airflow measurement results, it is used to characterize the gas flow rate involved in particle transport per unit time; firstly, based on the fineness evolution trend parameters... The generation trend of ultrafine particles is constructed by combining its absolute value with the trend direction, since the fineness evolution trend parameter reflects the direction and intensity of fineness change. This ultrafine particle generation trend is denoted as […]. The calculation method is as follows: ; in, This represents the maximum value of the detail evolution trend parameter within the statistical period. This represents the minimum value of the detail evolution trend parameter within the statistical period. This is a normalized indicator of the formation trend of ultrafine particles; in one embodiment, the statistical period preferably uses historical fineness data from the most recent continuous operating cycle; subsequently, it is based on the exhaust gas transport capacity index. With system air volume The trend of ultrafine particle conveying is generated; the exhaust gas transport capacity index reflects the ability of fine powder to be carried out by the airflow, and the system air volume reflects the conveying scale. Therefore, the trend of ultrafine particle conveying is constructed. : ; in, The baseline airflow can be the average airflow during the system's stable operation phase. The criteria for determining the stable operation phase can refer to the aforementioned method, and the arithmetic mean of the system airflow during that period can be used as the baseline airflow. . This serves as an indicator of the transport trend of dimensionless ultrafine particles. In this scheme, the preferred reference air volume is the average air volume over a continuous operating cycle under historical stable conditions; subsequently, the ultrafine particle generation trend is considered. With the trend of ultrafine particle conveying A particle load prediction feature set is constructed; this feature set is used to simultaneously characterize the generation and transport capacity of ultrafine particles, and can be expressed as: ; After obtaining the feature set for predicting particulate load, the estimated load of ultrafine particles entering the dust removal system in the future is calculated; the estimated load of ultrafine particles is denoted as... The unit is mass flow rate, representing the mass of ultrafine particles entering the dust removal system within a predicted time window; the calculation method can be expressed as: ; in, The baseline ultrafine particle mass flow rate is derived from the average conveyed mass flow rate of ultrafine particles obtained from historical stable operation phases; in one embodiment, The method for obtaining the mass flow rate is as follows: within a continuous time period determined to be stable operation (e.g., 2 consecutive hours), the total mass of particles carried by the airflow through the finished product pipeline per unit time is calculated based on the finished product fineness parameters and the system airflow. This total mass is multiplied by a preset ultrafine particle proportion coefficient to obtain the ultrafine particle mass flow rate. The arithmetic mean is then taken for multiple similar time periods, and this average value is used as the benchmark ultrafine particle mass flow rate. The calculated estimated load of ultrafine particles This refers to the predicted load of ultrafine particles entering the dust removal system within a future time window. This result is used to characterize the particle load level that the dust removal system may bear in the future, and serves as the data input for the generation of subsequent dust removal adjustment commands and the judgment of operation mode switching. This enables the linkage control process to not only adjust according to the current operating status, but also to make control decisions based on the future load change trend.

[0029] like Figure 4 As shown, S4, using the cause type, exhaust gas transport capacity index and ultrafine particle load estimate as input conditions, determines the air volume adjustment command, powder selection adjustment command and dust removal adjustment command, and executes the air volume adjustment command, powder selection adjustment command and dust removal adjustment command. Furthermore, in S4, the steps using the genetic type, exhaust gas transport capacity index, and estimated ultrafine particulate load as input conditions include: The causal types are mapped to quantified values ​​to generate causal demand features; The exhaust gas transport capacity index is normalized to generate transport status characteristics; The estimated load of the ultrafine particles is normalized to generate load state characteristics; A set of control decision features is constructed based on the characteristics of cause and demand, transportation status, and load status.

[0030] Specifically, after obtaining the causes of fineness deviation, the exhaust gas transport capacity index, and the estimated load of ultrafine particles, a set of control decision features is further constructed to provide a unified decision input for the generation of subsequent airflow adjustment commands, powder selection adjustment commands, and dust removal adjustment commands; among which, the causes of fineness deviation are... Waste gas transport capacity index and estimated load of ultrafine particles The above steps have already been completed; firstly, causal demand characteristics are generated based on the causal type; since the causal type is a discrete state quantity, a state coding method is used for quantization representation, and the causal demand characteristics are denoted as... In one implementation, the insufficient transport type is assigned a value of 1, the insufficient powder selection type is assigned a value of 2, the insufficient grinding type is assigned a value of 3, and the comprehensive coupling type is assigned a value of 4, thus forming a unified expression of the causal demand characteristics. Subsequently, transport state characteristics are generated based on the waste gas transport capacity index. To eliminate the influence of differences in the numerical range of the waste gas transport capacity index under different operating conditions, the waste gas transport capacity index is normalized, and the transport state characteristics are denoted as... The calculation method is as follows: ; in, This represents the maximum value of the waste gas transport capacity index within the statistical period. To determine the minimum value of the waste gas transport capacity index within the statistical period, a continuous and stable operating period can be used. In one implementation, the statistical period preferably uses the operating data of the most recent 24 hours. Subsequently, load state characteristics are generated based on the estimated ultrafine particle load. Since the estimated ultrafine particle load reflects the future dust removal load level, a normalization method is used to construct the load state characteristics, which are denoted as […]. The calculation method is as follows: ; in, This represents the maximum estimated value of ultrafine particle load within the statistical period. This represents the minimum estimated load of ultrafine particles within the statistical period; after processing, , and The current system operating status is described from three dimensions: causal state, transport state, and load state; then, based on the causal demand characteristics... Transportation status characteristics and load condition characteristics Construct a set of control decision features, denoted as . : ; Among them, the control decision feature set serves as the unified input data structure for subsequent control decision stages. After the control decision feature set is formed, it is input into the regulation rule matching process. By analyzing the regulation demand corresponding to the causal demand characteristics, the transportation capacity level corresponding to the transportation state characteristics, and the dust removal load level corresponding to the load state characteristics, the direction of air volume regulation, the direction of powder selection regulation, and the direction of dust removal regulation are further determined. These directions serve as the basis for generating subsequent air volume regulation commands, powder selection regulation commands, and dust removal regulation commands, thereby achieving linkage decision-making between fineness state, transportation state, and dust removal state.

[0031] Furthermore, in S4, the steps for determining the airflow adjustment command, powder selection adjustment command, and dust removal adjustment command include: Establish the correspondence between the set of control decision features and the adjustment actions, and generate a set of control rules; The set of control decision features is matched with the set of control rules to determine the target adjustment combination; Generate air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands based on the target adjustment combination; Execute air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands.

[0032] Specifically, after obtaining the control decision feature set, airflow adjustment commands, powder selection adjustment commands, and dust removal adjustment commands are further generated, thereby achieving coordinated control of the vertical mill system and the exhaust gas treatment system; whereby the control decision feature set is denoted as... Derived from the construction results of causal demand characteristics, transport state characteristics, and load state characteristics, it can be expressed as: ;in, The characteristics of the causal demand are derived from the coding results of the cause type of fineness deviation. The characteristics of the transport status are derived from the normalized results of the waste gas transport capacity index. The load state characteristics are derived from the normalized results of ultrafine particle load prediction. First, a correspondence is established between the control decision characteristic set and the adjustment actions, generating a control rule set. This control rule set is denoted as... This is used to establish the mapping relationship between different operating states and adjustment actions; in one implementation, the control rule set is constructed in the form of a rule table, where each rule contains a combination of causal demand characteristics, transport state characteristics, and load state characteristics, as well as the corresponding adjustment action, i.e.: ;in, Indicates the first Control rules; This indicates the causal requirements and characteristic conditions of the corresponding rule; The transport state characteristic conditions corresponding to the rule; This represents the load state characteristic conditions of the corresponding rule; This represents the target adjustment action under the corresponding rule; in one implementation, when the cause-demand characteristic corresponds to an insufficient transport state and the transport state characteristic is lower than a preset level, the target adjustment action preferably includes an air volume increase action; when the cause-demand characteristic corresponds to an insufficient powder selection state, the target adjustment action preferably includes a powder selection enhancement action; when the load state characteristic is higher than a preset level, the target adjustment action preferably includes a dust removal enhancement action; subsequently, the control decision characteristic set is matched with the control rule set to determine the target adjustment combination; to quantify the degree of matching between the current operating state and each rule, a rule matching degree is defined. : ; in, This indicates the matching results of causal demand features; This indicates the matching result of transport status features; This indicates the result of load state feature matching; This represents the corresponding weight coefficient, which satisfies: In one implementation, The preferred value is 0.4. The preferred value is 0.3. The preferred value is 0.3; the maximum matching value is determined based on the rule matching degree corresponding to all control rules. ; The control rule corresponding to the maximum matching value is determined as the target adjustment combination, denoted as . After obtaining the target adjustment combination, generate airflow adjustment commands, powder selection adjustment commands, and dust removal adjustment commands based on the target adjustment combination; among them, the airflow adjustment command is denoted as... Used to adjust the frequency of the circulating fan or the operating parameters of the induced draft fan; the powder selection adjustment command is denoted as... Used to adjust the operating parameters of the air classifier; dust removal adjustment commands are denoted as... This is used to adjust the dust removal cycle or the blowing frequency; the target adjustment combination can be expressed as: ; Then execute the air volume adjustment command. Powder selection adjustment instructions and dust removal adjustment instructions This process updates the airflow, powder selection, and dust removal status according to the target adjustment combination. After execution, new system airflow, mill internal pressure difference, powder classifier speed, and dust removal operation status are generated, and the next control cycle is entered to continue acquiring fineness detection data, thus forming a closed-loop control process of fineness status identification, transport capacity evaluation, load prediction, and linkage adjustment.

[0033] like Figure 5 As shown, S5 obtains the changing trend of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period, and performs mode switching between multiple preset operating modes based on the changing trend. Furthermore, in S5, the steps for switching between multiple preset operating modes based on changing trends include: The operating trend combination is constructed based on the changing trend of the exhaust gas transport capacity index and the changing trend of the ultrafine particle load estimate within the preset period. Operational status features are generated based on a combination of operational trends; Match the running status characteristics with the mode switching conditions corresponding to multiple preset running modes; Among them, the multiple preset operating modes include quality priority mode, energy consumption priority mode and dust removal safety priority mode; The execution mode is switched according to the target operating mode.

[0034] Specifically, after completing airflow adjustment, powder selection adjustment, and dust removal adjustment, the operating mode is switched based on changes in operating status over a longer time scale. This allows the control process to not only adjust the current operating conditions but also adjust the overall control objectives according to the long-term operating trend of the system. The waste gas transport capacity index is denoted as... The value is derived from the evaluation results of waste gas transport capacity; the estimated load of ultrafine particles is denoted as... The data is derived from particle load prediction results; the preset period is denoted as... This is used to characterize the operational mode evaluation cycle; in one implementation, The preferred method is a statistical time window consisting of multiple consecutive control cycles; firstly, the sequence of exhaust gas transport capacity indexes corresponding to consecutive moments within the preset cycle is obtained: ; and the estimated sequence of ultrafine particle load: ;in, Indicates the first The exhaust gas transport capacity index corresponding to each moment. Indicates the first The estimated ultrafine particle load at each time point; subsequently, the trend of waste gas transport capacity change is calculated based on the waste gas transport capacity index sequence, denoted as . : ; The load change trend was calculated based on the estimated load sequence of ultrafine particles, denoted as : ; in, Used to characterize the long-term trend of changes in waste gas transport capacity. Used to characterize the development trend of dust removal load; subsequently, based on the changing trend of waste gas transport capacity... and load change trends Constructing a portfolio of operating trends: ; After obtaining the combination of operational trends, operational status features are generated. To unify the expression of data with different dimensions, the change trends are first normalized to obtain normalized transport trend features. With normalized load trend characteristics : ; ; in, and These represent the maximum and minimum values ​​of the waste gas transport capacity trend within the historical statistical period, respectively. and These represent the maximum and minimum values ​​of the load change trend within the historical statistical period, respectively; subsequently, the operating status characteristics are constructed: ; The operational status characteristics are used to characterize the long-term operational status of the current system in terms of both transport capacity and dust removal load. Subsequently, the operational status characteristics are matched with the mode switching conditions corresponding to multiple preset operational modes. In this scheme, the preset operational modes include quality-priority mode, energy-priority mode, and dust removal safety-priority mode. The set of mode characteristic conditions corresponding to each operational mode is denoted as follows: , , ;in, Corresponding to the quality-first mode, Corresponding to the energy-priority mode, Corresponding to the dust removal safety priority mode; to quantify the degree of matching, a mode matching degree is defined. : ; in, This indicates the matching result between operational status characteristics and corresponding mode transport conditions. This represents the matching result between operating status characteristics and corresponding mode load conditions, where the quality-first mode... The corresponding transport conditions are: In the medium range, the load conditions are At a lower range; energy-first mode The corresponding transport conditions are: At a higher range, the load conditions are At a low level; dust removal safety priority mode. The corresponding load conditions are It is in a relatively high range. In one implementation, ,in For the first Reference values ​​for transport conditions corresponding to each mode; ,in For the first Reference values ​​for load conditions corresponding to each mode; mode matching degree The smaller the value, the higher the degree of matching; the minimum value corresponds to the target operating mode. value, and Represents the weighting coefficients, satisfying: In one implementation, The preferred value is 0.5. The preferred value is 0.5; then the minimum pattern matching degree is determined: ; The operating mode corresponding to the minimum pattern matching degree is the target operating mode, denoted as . After obtaining the target operating mode, a mode switch is executed, causing the subsequent control rule generation process to operate according to the optimization target corresponding to the target operating mode. When the target operating mode is a quality-first mode, the subsequent control rules are tilted towards finer stability; when the target operating mode is an energy-first mode, the subsequent control rules are tilted towards wind turbine energy consumption optimization. For example, in a certain preset evaluation cycle... Within, if the normalized load trend characteristics are calculated... A value greater than 0.7 indicates that the estimated load of ultrafine particles continues to rise rapidly, regardless of transport trends. At what level, according to the rules for calculating pattern matching degree, the current operating state is closest to the condition reference value vector corresponding to the "dust removal safety priority mode," and the pattern matching degree is considered to be at that level. Once the minimum value is reached, the control system will automatically select and switch to the "dust removal safety priority mode". When the target operating mode is the dust removal safety priority mode, subsequent control rules will be tilted towards dust removal load control; the target operating mode will participate in the control decision-making process as a mode parameter in the control strategy generation stage of the next operating cycle, thus forming a two-layer closed-loop control mechanism with the synergistic effect of parameter adjustment and operating mode.

[0035] Accordingly, one embodiment of the present invention also provides a vertical mill and exhaust gas treatment linkage control system based on online fineness detection. This system is used to implement the control method described in the above embodiments. The system includes: The data acquisition and processing module communicates with the online particle size analyzer installed on the finished product conveying pipeline and the various sensor recording units in the DCS system. It is used to acquire the time-series detection values ​​of the fineness of the finished product, the system air volume adjustment records, the classifier speed adjustment records, etc., and calculates and generates fineness evolution trend parameters and fineness response state parameters according to the preset algorithm.

[0036] The cause identification module receives parameters output by the data acquisition and processing module, as well as real-time data on system air volume, mill internal pressure difference, classifier speed, and return material volume. By constructing a deviation feature vector and performing distance matching with the internally stored reference feature vector library, it identifies the cause type of the current fineness deviation.

[0037] The load prediction module is configured to construct an exhaust gas transport capacity index based on the cause type, system air volume, mill internal pressure difference, return material volume, and fineness evolution trend parameters, combined with the cause type weight, and to calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume, and fineness evolution trend parameters.

[0038] The linkage control module is electrically connected to the frequency converter of the circulating fan, the drive unit of the classifier, and the pulse jet control unit of the dust collector. Based on the cause type, the exhaust gas transport capacity index, and the estimated load of ultrafine particles, it generates air volume adjustment commands, classifier adjustment commands, and dust removal adjustment commands for the circulating fan, classifier, and dust collector through matching of built-in control rule sets, and sends the commands to the corresponding actuators.

[0039] The mode management module continuously monitors the changing trends of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period. When the trend characteristics meet the preset mode switching conditions, it sends a mode switching signal to the linkage control module, enabling it to switch between operating modes such as quality priority, energy consumption priority, or dust removal safety priority, thereby changing the matching logic of subsequent control rules.

[0040] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for the closed-loop control of a vertical mill and exhaust gas treatment based on online detection of fineness, characterized in that, Includes the following steps: S1. Obtain the time-series detection values ​​of the fineness of the finished product, generate fineness evolution trend parameters and fineness response state parameters, and obtain the system air volume, mill internal pressure difference, classifier speed and return material amount; S2. Based on the correspondence between fineness evolution trend parameters, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount, the cause type of fineness deviation is identified by comparing the distance between the deviation feature vector composed of the above parameters and the reference feature vector corresponding to each cause type. S3. Based on the cause type, system air volume, mill internal pressure difference, return material amount and fineness evolution trend parameters, and combined with the cause type weight, construct the exhaust gas transport capacity index, and calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume and fineness evolution trend parameters. The estimated load of ultrafine particles is used to characterize the amount of ultrafine particles that will enter the dust removal system after a preset time under the current working conditions. S4. Using the type of formation, the exhaust gas transport capacity index, and the estimated load of ultrafine particles as input conditions, determine the air volume adjustment command, the powder selection adjustment command, and the dust removal adjustment command, and execute the air volume adjustment command, the powder selection adjustment command, and the dust removal adjustment command. S5. Obtain the changing trend of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period, and switch between multiple preset operating modes according to the changing trend.

2. The method according to claim 1, wherein, In S1, the steps of generating the fineness evolution trend parameters and the fineness response state parameters include: A fineness sequence within a continuous time window is constructed based on the time-series detection values ​​of the finished product fineness, and the fineness change rate, fineness change acceleration, and fineness fluctuation characteristics are calculated based on the fineness sequence. The fineness evolution trend parameters are generated according to the fineness change rate, the fineness change acceleration and the fineness fluctuation characteristics, wherein the fineness evolution trend parameters According to The calculation generates, wherein The fineness change rate is The fineness change acceleration is The fineness fluctuation characteristics are The preset weight coefficient satisfies ; Obtain at least one of the following within a preset time range: system air volume adjustment record, air classifier speed adjustment record, and circulating fan frequency adjustment record; The fineness response state parameters are generated based on the change in fineness evolution trend parameters before and after the occurrence of the control action.

3. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S2, the step of identifying the cause type of fineness deviation includes: A combination of deviation characteristics is constructed based on parameters of fineness evolution trend, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount; The parameters in the deviation feature combination are normalized, and a deviation feature vector is generated based on the normalized parameters. Calculate the Euclidean distance between the deviation feature vector and the reference feature vector corresponding to each cause type; The cause type corresponding to the minimum Euclidean distance is identified as the cause type of the fineness deviation.

4. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S3, the step of constructing the exhaust gas transport capacity index includes: The airflow transport characteristics are generated based on the system air volume and the pressure difference inside the mill. Particle recycling characteristics are generated based on parameters related to the amount of recycled material and the evolution trend of fineness. Construct a set of related features based on airflow transport characteristics, particle circulation characteristics, and genetic types; The waste gas transport capacity index is generated based on the set of associated features.

5. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S3, the step of calculating the estimated load of ultrafine particles entering the dust removal system in the future includes: Based on the fineness evolution trend parameters, an ultrafine particle generation trend is generated to characterize the ability to generate ultrafine particles; Based on the exhaust gas transport capacity index and system air volume, an ultrafine particle transport trend is generated to characterize the ability of ultrafine particles to be transported to the dust removal system. A set of features for predicting particle load is constructed based on the generation and transport trends of ultrafine particles. The estimated value of ultrafine particle load entering the dust removal system in the future is calculated based on the particle load prediction feature set.

6. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S4, the step of using the causal type, exhaust gas transport capacity index, and estimated ultrafine particle load as input conditions includes: The causal types are mapped to quantified values ​​to generate causal demand features; The exhaust gas transport capacity index is normalized to generate transport status characteristics; The estimated load of the ultrafine particles is normalized to generate load state characteristics; A set of control decision features is constructed based on the characteristics of cause and demand, transportation status, and load status.

7. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S4, the steps of determining the air volume adjustment command, the powder selection adjustment command, and the dust removal adjustment command include: Establish the correspondence between the set of control decision features and the adjustment actions, and generate a set of control rules; The set of control decision features is matched with the set of control rules to determine the target adjustment combination; Generate air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands based on the target adjustment combination; Execute air volume adjustment commands, powder selection adjustment commands, and dust removal adjustment commands.

8. The method for coordinated control of vertical mill and exhaust gas treatment based on online fineness detection according to claim 1, characterized in that, In S5, the step of switching between multiple preset operating modes based on the changing trend includes: The operating trend combination is constructed based on the changing trend of the exhaust gas transport capacity index and the changing trend of the ultrafine particle load estimate within the preset period. Operational status features are generated based on a combination of operational trends; Match the running status characteristics with the mode switching conditions corresponding to multiple preset running modes; Among them, the multiple preset operating modes include quality priority mode, energy consumption priority mode and dust removal safety priority mode; The execution mode is switched according to the target operating mode.

9. A vertical mill and exhaust gas treatment linkage control system based on online fineness detection, characterized in that, For performing the method according to any one of claims 1 to 8, comprising: The data acquisition and processing module is connected to the online particle size analyzer of the finished product and the recording unit of the DCS system. It is configured to acquire the time-series detection values ​​of the fineness of the finished product and generate fineness evolution trend parameters and fineness response state parameters. The cause identification module is configured to identify the cause type of fineness deviation by comparing the distance between the deviation feature vector composed of the above parameters and the reference feature vector corresponding to each cause type, based on the correspondence between the fineness evolution trend parameters, fineness response state parameters, system air volume, mill internal pressure difference, classifier speed and return material amount. The load prediction module is configured to construct an exhaust gas transport capacity index based on the cause type, system air volume, mill internal pressure difference, return material volume and fineness evolution trend parameters, combined with the cause type weight, and calculate the estimated load of ultrafine particles entering the dust removal system in the future based on the exhaust gas transport capacity index, system air volume and fineness evolution trend parameters. The linkage control module is electrically connected to the frequency converter of the circulating fan, the drive unit of the classifier and the pulse jet control unit of the dust collector. It is configured to use the cause type, the exhaust gas transport capacity index and the estimated load of ultrafine particles as input conditions to determine the air volume adjustment command, the classifier adjustment command and the dust removal adjustment command, and execute the air volume adjustment command, the classifier adjustment command and the dust removal adjustment command. The mode management module is configured to acquire the changing trends of the exhaust gas transport capacity index and the estimated load of ultrafine particles within a preset period, and to switch between multiple preset operating modes based on the changing trends.

Images