A data analysis method for intelligent production scheduling of a platform paper production line

Abstract

The application provides a kind of platform paper production line intelligent production scheduling data analysis method, comprising: obtaining fiber length distribution and current screen aperture size from slurry pool, simultaneously obtaining real-time capacity level and production speed of each section from storage bin monitoring device, data fusion is carried out to obtain initial dynamic relationship matrix;If the agglomeration density is higher than the preset risk threshold, then extract the relevant subset about fiber flow speed and screen resistance from the dynamic relationship matrix, and adjust the screen aperture size by the fiber agglomeration density, while integrating the production speed, to obtain the adjusted material passing rate;Material passing rate is compared with current storage bin capacity level to obtain capacity matching degree, whether to expand buffer storage space is judged according to capacity matching degree, if material passing rate exceeds capacity level preset threshold, then increase the buffer space allocation ratio.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for intelligent scheduling data analysis of a board paper production line. Background Technology

[0002] In the pulp and paper industry, the stability of paper quality is crucial to production efficiency and product competitiveness, especially in benchtop paper production lines, where the handling of fiber raw materials directly determines the thickness uniformity and production continuity of the finished paper. However, existing methods have significant shortcomings in dealing with fiber agglomeration and production speed fluctuations. Many solutions rely too heavily on adjusting or fixing parameters of a single piece of equipment, ignoring the dynamic relationship between fiber characteristics, equipment parameters, and production rhythm. This results in an inability to flexibly adapt to changes in raw materials from different batches or the material flow requirements between multiple stages of the production line. This limitation makes it difficult for the production line to maintain a stable material supply and uniform paper quality when facing complex operating conditions. Specifically, fiber agglomeration is a core technical challenge affecting production stability. Agglomerates formed by fiber length differences in the pulp tank can be redistributed by adjusting the screen aperture of the screening equipment, thus affecting the production speed of subsequent stages. While adjusting the screen aperture can change the distribution of agglomerates, it introduces a new problem: the difficulty of matching the storage bin capacity with the production speed. For example, when operators increase the screen aperture to reduce agglomerate blockage, too many short fibers may pass through, causing speed fluctuations in subsequent processes due to changes in material properties, and even leading to overflowing or empty storage bins. Such fluctuations not only disrupt the production rhythm but can also cause uneven paper thickness. Therefore, optimizing the buffer storage space configuration between processes by analyzing the dynamic relationship between storage bin capacity, fiber length distribution, screen aperture adjustment values, and production speeds in real time, to ensure continuous and stable material flow and maintain uniform finished paper thickness, becomes a critical issue. Summary of the Invention

[0003] This invention provides an intelligent scheduling data analysis method for a board paper production line, mainly including:

[0004] Fiber length distribution and screen aperture data are obtained from the slurry tank, and capacity level and production speed of each section are obtained from the storage silo monitoring equipment. These data are then fused to construct an initial dynamic relationship matrix. Based on this matrix, fiber length is identified, and fibers are classified as long, medium, and short fibers. The proportion of long fibers is calculated, and a weighted average is used to obtain the ratio of fiber length to screen aperture, determining the agglomeration density. Based on this agglomeration density, a subset of fiber flow velocity and screen resistance is extracted from the initial dynamic relationship matrix. The screen aperture is adjusted, and the material throughput is calculated in conjunction with the production speed. Finally, the material throughput is compared with the storage silo capacity level. The process involves calculating the capacity matching degree and adjusting the buffer space allocation ratio accordingly. After adjustment, the fiber passage speed fluctuations in each section are extracted, and the fiber length distribution is integrated. A moving average filtering algorithm is then used to obtain a smoothed production speed. The smoothed production speed is compared with the initial dynamic relationship matrix to calculate flow stability. Based on the flow stability, the storage silo capacity requirement is adjusted to determine the target material flow parameter set. The paper thickness uniformity of the target material flow parameter set is predicted, and the adjusted screen aperture and buffer space allocation ratio are integrated to calculate the thickness standard deviation and coefficient of variation, thus determining the production scheduling parameters.

[0005] Furthermore, the process of acquiring fiber length distribution and screen aperture data from the slurry tank, acquiring capacity levels and production speeds of each section from the storage silo monitoring equipment, and fusing the data to construct an initial dynamic relationship matrix includes:

[0006] Samples were collected from different locations in the pulp tank, fiber lengths were scanned, the proportions of long, medium, and short fibers were statistically analyzed, and screen aperture values ​​were read. Capacity level data was obtained from the storage silo, timestamped, and a capacity change sequence was generated. Production speeds were collected from the pulping, screening, and forming stages, and the material transfer rate between adjacent stages was calculated. Based on the fiber length distribution and screen aperture values, a fiber throughput vector was calculated, and the corresponding values ​​for long fibers in the fiber throughput vector were adjusted. An initial dynamic relationship matrix was constructed by arranging the fiber throughput vector, material transfer rate, and capacity change sequence in chronological order.

[0007] Furthermore, the step of identifying fiber length based on the initial dynamic relationship matrix, classifying fibers into long, medium, and short fibers, calculating the proportion of long fibers, and using a weighted average to obtain the ratio of fiber length to screen aperture to determine the aggregation density includes:

[0008] Fiber throughput vectors are extracted from the initial dynamic relationship matrix to identify long, medium, and short fibers; the proportion of long fibers is statistically analyzed, and the concentration of long fibers at different locations in the pulp tank is collected and weighted by depth to obtain the comprehensive long fiber proportion; the ratio is calculated by dividing the comprehensive long fiber proportion by the screen aperture value; the fiber throughput at continuous time points is extracted, and the absolute value of the throughput difference is calculated to determine the agglomeration fluctuation index; the agglomeration density is calculated based on the agglomeration fluctuation index, the ratio, and the fiber suspension concentration.

[0009] Furthermore, the step of extracting a subset of fiber flow velocity and screen resistance from the initial dynamic relationship matrix based on agglomeration density, adjusting the screen aperture, and calculating the material throughput rate in conjunction with the production speed includes:

[0010] Extract the material transfer rate corresponding to the low throughput row from the initial dynamic relationship matrix as the fiber flow velocity; calculate the reciprocal of the throughput change to determine the screen resistance coefficient and construct a subset; calculate the screen aperture adjustment amount based on the mean of the screen resistance coefficient and the difference in agglomeration density to determine the target screen aperture; adjust the production speed of each section according to the target screen aperture; calculate the fiber mass flow rate by multiplying the screening section speed by the target screen aperture and dividing by the fiber suspension concentration to determine the material throughput.

[0011] Furthermore, the step of comparing the material throughput with the storage silo capacity level, calculating the capacity matching degree, and adjusting the buffer space allocation ratio according to the capacity matching degree includes:

[0012] The material inflow rate is calculated by multiplying the material throughput rate by the screening section speed; the storage silo capacity occupancy rate is obtained, and the ratio of the material inflow rate to the remaining capacity is calculated as the capacity matching degree; the capacity matching degree is compared with a threshold to calculate the difference and determine the buffer space expansion demand coefficient; the additional buffer space demand is calculated using the buffer space expansion demand coefficient, the buffer space allocation ratio increment is determined, and the allocation ratio is adjusted.

[0013] Furthermore, the extracted and adjusted fiber speed fluctuations of each section, combined with the fiber length distribution, are processed using a moving average filtering algorithm to obtain a smoothed production speed, including:

[0014] Extract the fiber speed difference from each process section to form a fluctuation sequence; apply a moving average filter to the fluctuation sequence, calculate the average value within the window, and obtain a smoothed fluctuation sequence; calculate a comprehensive fluctuation index by weighted fusion of the smoothed fluctuation sequence and the fiber length distribution; determine the material flow trend through the difference of the comprehensive fluctuation index; adjust the filter window size to process the original production speed and obtain a smoothed production speed.

[0015] Furthermore, the process of calculating flow stability by comparing the smoothed production speed with the initial dynamic relationship matrix, adjusting the storage silo capacity requirement based on the flow stability, and determining the target material flow parameter set includes:

[0016] By comparing the smoothed production speed with the material transfer rate point by point, the standard deviation of the speed difference is calculated. Combined with the capacity change sequence, the flow stability index is determined. The stability deviation is calculated by comparing the flow stability index with the benchmark value. The capacity adjustment coefficient is calculated based on the stability deviation and the fiber throughput fluctuation. The capacity of the storage silo is adjusted by the capacity adjustment coefficient to determine the target production speed and screen aperture, and a target material flow parameter set is constructed.

[0017] Furthermore, the process of predicting paper thickness uniformity based on the target material flow parameter set, integrating the adjusted screen aperture and buffer space allocation ratio, calculating the thickness standard deviation and coefficient of variation, and determining production scheduling parameters includes:

[0018] A thickness mapping relationship is established using the target material flow parameter set, and the paper thickness distribution is calculated. Based on the paper thickness distribution, the thickness standard deviation and coefficient of variation are calculated. The production coordination coefficient is determined by combining the thickness standard deviation with the adjusted screen aperture ratio and the buffer space ratio. The speed of each section is adjusted using the production coordination coefficient, the material transfer time interval is calculated, and the optimal production speed combination and section coordination sequence are determined.

[0019] Furthermore, the process of predicting paper thickness uniformity based on the target material flow parameter set, integrating the adjusted screen aperture and buffer space allocation ratio, calculating the thickness standard deviation and coefficient of variation, and determining production scheduling parameters includes:

[0020] By combining the screen aperture and storage bin capacity from the target material flow parameter set with the thickness variation coefficient, the speed matching coefficient between the pulping section and the forming section is calculated; the material transfer time interval is adjusted according to the speed matching coefficient to determine the coordination sequence of the sections; the buffer time is calculated by dividing the storage bin capacity by the material consumption rate; by comparing the buffer time with the screen adjustment time, the screen adjustment frequency and speed change cycle are determined, the section start sequence and speed switching timing are set, and a production scheduling plan is constructed.

[0021] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0022] This invention discloses an intelligent scheduling data analysis method for a sheet metal production line. Addressing the problems of fiber agglomeration, unstable material flow, and low production efficiency during pulp preparation and paper forming, it constructs an initial dynamic relationship matrix through data fusion, identifies and classifies fiber lengths, calculates the ratio of long fibers to screen aperture, and determines agglomeration density. When the agglomeration density exceeds the standard, the screen aperture is dynamically adjusted and the production speed is optimized to improve material throughput. Further, through capacity matching analysis, the buffer storage space is expanded, and moving average filtering is used to smooth production speed fluctuations, predict material flow trends, and assess flow stability. If the stability deviation exceeds a threshold, the storage silo capacity requirement is recalculated, a target material flow parameter set is generated, paper thickness uniformity is predicted, and the coordination sequence and speed matching of work sections are optimized. Finally, an intelligent scheduling scheme including start-stop sequence and speed switching timing is formed. This invention achieves full-process optimization from fiber separation to paper forming by dynamically controlling screen aperture and production parameters, improving paper thickness uniformity and production efficiency, and reducing material waste and equipment wear. Attached Figure Description

[0023] Figure 1 This is a flowchart of an intelligent scheduling data analysis method for a board paper production line according to the present invention.

[0024] Figure 2 This is a schematic diagram of an intelligent scheduling data analysis method for a board paper production line according to the present invention. Detailed Implementation

[0025] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0026] like Figure 1-2 This embodiment of an intelligent scheduling data analysis method for a board paper production line may specifically include:

[0027] S101. Obtain the fiber length distribution and current screen aperture size from the slurry tank, and simultaneously obtain the real-time capacity level and production speed of each section from the storage silo monitoring equipment. Perform data fusion to obtain the initial dynamic relationship matrix.

[0028] Slurry samples were collected from sampling ports at the bottom and middle of the slurry tank. The length of each fiber in the sample was scanned using the image recognition function of a fiber analyzer, and the distribution data of long, medium, and short fiber proportions were statistically obtained. Simultaneously, the current screen aperture value was read from the control panel of the screening equipment. Real-time capacity level data was acquired through an ultrasonic level sensor built into the storage silo. The current production speed values ​​of each section were collected from the PLC controllers of the pulping, screening, and forming sections. The capacity levels were timestamped according to the collection time to obtain a capacity change sequence. Based on the ratio of fiber length distribution to screen aperture value, the probability value of fibers of different lengths passing through the screen was calculated, constructing a fiber throughput vector. If the proportion of long fibers exceeds a preset threshold, the value at the corresponding position in the throughput vector is adjusted according to the difference between the proportion of long fibers and the threshold. The material transfer rate between adjacent sections was calculated using the difference in production speed between each section. Using the fiber throughput vector as the first column of the matrix, the material transfer rate as the second column, and the capacity change sequence as the third column, the matrix is ​​arranged in chronological order to form a row vector, thus constructing an initial dynamic relationship matrix that includes fiber characteristics, equipment parameters, and production rhythm.

[0029] In one embodiment, the fiber analyzer employs a linear CCD camera combined with an LED backlight to form an image acquisition system. The pulp sample is fixed to the detection area using a glass slide. The image recognition function identifies the outline of individual fibers using an edge detection algorithm, and measures the straight-line distance between the two ends of the fiber as the fiber length value. Fibers longer than 5 mm are classified as long fibers, those between 3 and 5 mm as medium fibers, and those shorter than 3 mm as short fibers. The distribution data is obtained by calculating the percentage of each type of fiber out of the total number of fibers.

[0030] It should be noted that the ultrasonic level sensor is installed on the top of the storage silo. It calculates the liquid level height based on the time difference by emitting 40kHz ultrasonic pulses and receiving the reflected signals. The capacity change sequence refers to a sequence formed by arranging data pairs consisting of the capacity level value at each acquisition moment and the corresponding timestamp in chronological order.

[0031] Specifically, the fiber throughput vector is constructed by calculating the ratio of fiber length to screen aperture. When the fiber length is less than 0.8 times the screen aperture, the throughput probability is set to 0.95; when the fiber length is between 0.8 and 1.2 times the screen aperture, the throughput probability decreases linearly from 0.95 to 0.1; and when the fiber length is greater than 1.2 times the screen aperture, the throughput probability is set to 0.1. The preset threshold for the proportion of long fibers is usually set to 35%. When the actual proportion exceeds this value, the throughput value at the corresponding position decreases by 0.02 for every 1 percentage point exceeding this value.

[0032] In one possible implementation, the material transfer rate is calculated by multiplying the difference in production speed between adjacent sections by the cross-sectional area of ​​the pipeline. The transfer rate between the pulping section and the screening section reflects the pulp conveying capacity, while the transfer rate between the screening section and the molding section reflects the supply rate of qualified pulp.

[0033] Preferably, the initial dynamic relationship matrix is ​​constructed using a time-aligned approach, with each row representing a snapshot of the system state at the same moment. The columns of the matrix are arranged in the order of fiber throughput vector, material transfer rate, and capacity change sequence, forming a multi-dimensional data structure that reflects the system's operating state. This matrix structure can intuitively demonstrate the impact of changes in fiber characteristics on the production process.

[0034] S102. Identify fiber length based on the initial dynamic relationship matrix, classify fibers in the slurry pool according to fiber length, including long fibers, medium fibers, and short fibers, identify the proportion of long fibers, perform weighted average processing on the proportion of long fibers, obtain the ratio of fiber length to screen aperture, and determine the agglomeration density.

[0035] The numerical distribution of fiber throughput vectors is extracted from the initial dynamic relationship matrix. K-means clustering is used to group the throughput values, with three clusters. Based on the center value of each group's throughput value, a low throughput group is identified for long fibers, a medium throughput group for medium fibers, and a high throughput group for short fibers. The number of data points in each group is counted to obtain the quantity distribution of each type of fiber. The percentage of long fibers in the total fiber quantity is calculated based on this distribution. Long fiber concentration data are collected from the bottom, middle, and top of the pulp tank. The proportion of long fibers at each location is weighted and summed according to a preset depth weighting coefficient, with the bottom having the largest weight and the top having the smallest weight, to obtain a comprehensive long fiber proportion value. This comprehensive long fiber proportion value is divided by the current screen aperture value to obtain the fiber length to screen aperture ratio. If the ratio exceeds a preset agglomeration risk threshold, the fiber throughput values ​​at five consecutive time points in the matrix are extracted, and the sum of the absolute values ​​of the differences in throughput at adjacent time points is calculated as an agglomeration fluctuation index. The degree of fiber aggregation per unit volume is obtained by multiplying the agglomeration fluctuation index by the ratio of fiber length to screen aperture, and then by the fiber suspension concentration in the slurry tank. This aggregation degree value is compared with a preset critical density threshold. If the aggregation degree is higher than the critical density threshold, the aggregation degree value is defined as the agglomeration density, and the agglomeration density value is determined.

[0036] In one embodiment, the K-means clustering method begins with the fiber throughput vector of the initial dynamic relation matrix. The throughput value at each time point in the matrix reflects the probability of fibers of different lengths passing through the sieve, with throughput values ​​distributed between 0.1 and 0.95. The clustering algorithm first randomly selects three initial cluster centers, corresponding to throughput values ​​of 0.15, 0.5, and 0.85, respectively. During iteration, the algorithm calculates the Euclidean distance from each throughput value to the three cluster centers and assigns the value to the nearest cluster group. After each iteration, the average value of each group is recalculated as the new cluster center, until the change in cluster centers is less than 0.01 or the number of iterations reaches 100. In the resulting three cluster groups, the group with the smallest center value corresponds to long fibers, as long fibers are difficult to pass through the sieve; the group with the middle center value corresponds to medium fibers; and the group with the largest center value corresponds to short fibers, as short fibers easily pass through the sieve. The number of data points in each cluster group is counted and divided by the total number of data points to obtain the distribution of the proportion of each type of fiber.

[0037] It should be noted that multi-location sampling in the slurry tank reflects the vertical distribution characteristics of the fibers. Due to the higher density and easier settling of long fibers, the concentration of long fibers at the bottom of the tank is usually higher than that at the top. Sampling locations were set at three heights: 0.5 meters, 2 meters, and 3.5 meters from the bottom of the tank, representing the bottom, middle, and top, respectively.

[0038] Specifically, the depth weighting coefficients are set based on fiber settling theory and practical production experience. The bottom weight is set to 0.5 because the fiber concentration in this area best reflects the actual fiber composition entering the screening equipment; the middle weight of 0.3 reflects the fiber state in the transition zone; and the upper weight of 0.2 represents the distribution of suspended fibers. The weighted summation calculation process involves multiplying the proportion of long fibers at each of the three locations by their corresponding weights, and then summing them to obtain the comprehensive value.

[0039] In one possible implementation, calculating the fiber length to screen aperture ratio involves determining the average fiber length. Based on the clustering results, the average length of long fibers is taken as 6 mm, medium fibers as 4 mm, and short fibers as 2 mm. The weighted average fiber length is obtained by multiplying the proportion of long fibers by 6 mm, the proportion of medium fibers by 4 mm, and the proportion of short fibers by 2 mm. This length is then divided by the screen aperture value to obtain the ratio.

[0040] Preferably, the agglomeration risk threshold is set to 1.5. When the ratio exceeds this value, it indicates that the fiber length is too large relative to the screen aperture, making agglomeration more likely. The agglomeration fluctuation index is calculated by selecting five consecutive time points with a time interval of 30 seconds. The difference in the throughput between adjacent time points reflects the instability of fiber passage; the larger the absolute value of the difference, the more severe the agglomeration phenomenon. Four differences are generated from the five time points, and the agglomeration fluctuation index is obtained by adding the absolute values ​​of these four differences.

[0041] For example, the calculation of the aggregation degree value integrates three key parameters. The agglomeration fluctuation index reflects the instability in the time dimension, the fiber length to screen aperture ratio reflects the blockage tendency in the spatial dimension, and the fiber suspension concentration reflects the density of the material. The value obtained by multiplying these three values ​​characterizes the degree of fiber entanglement and aggregation per unit volume. When the fiber suspension concentration in the slurry is 15 grams per liter, the agglomeration fluctuation index is 0.8, and the ratio is 1.8, the aggregation degree value is 21.6.

[0042] Understandably, setting the critical density threshold requires consideration of the production line's processing capacity and product quality requirements. A threshold that is too low will lead to frequent alarms and unnecessary adjustments, while a threshold that is too high may miss early warnings of agglomeration risks. Statistical analysis of historical production data shows that when the agglomeration level exceeds 20, the probability of blockages in subsequent processes increases significantly.

[0043] In one embodiment, the final determination of agglomeration density also considers the influence of temperature. Increased slurry temperature reduces the coefficient of friction between fibers, decreasing the tendency to agglomerate. When the slurry temperature is above 40 degrees Celsius, the agglomeration degree is multiplied by a temperature correction factor of 0.9; below 30 degrees Celsius, it is multiplied by a correction factor of 1.1; and in between, linear interpolation is used. The corrected value serves as the final agglomeration density, providing a quantitative basis for subsequent screen adjustments and production speed control. Furthermore, the application of the agglomeration density value can guide the optimization and adjustment of the production process. When the agglomeration density exceeds a critical value, the screening equipment automatically increases the screen vibration frequency or adjusts the screen tilt angle to break up the formed fiber agglomerates. Simultaneously, a dispersant is injected into the slurry tank to reduce the adhesion between fibers, fundamentally reducing the probability of agglomeration formation.

[0044] S103. If the agglomeration density is higher than the preset risk threshold, extract the relevant subsets of fiber flow velocity and screen resistance from the dynamic relationship matrix, adjust the screen aperture size by adjusting the fiber agglomeration density, and integrate the production speed to obtain the adjusted material throughput.

[0045] If the agglomeration density is higher than a preset risk threshold, the row index of the fiber throughput vector that is less than the preset throughput threshold is located from the dynamic relationship matrix. The material transfer rate data corresponding to the row index is extracted as the fiber flow velocity. The reciprocal of the throughput change at adjacent time points is calculated as the screen resistance coefficient, and a relevant subset containing the flow velocity and resistance coefficient is constructed. Based on the mean of the screen resistance coefficient in the relevant subset, the difference between the agglomeration density and the preset risk threshold is calculated. The difference is multiplied by a preset aperture adjustment coefficient to obtain the screen aperture adjustment amount. The original screen aperture plus the adjustment amount yields the target screen aperture value. Using the ratio of the target screen aperture value to the original aperture value, the production speed of each section is adjusted. The pulping section speed is divided by the ratio to obtain the new pulping speed, the screening section speed is multiplied by the ratio to obtain the new screening speed, and the forming section speed is adjusted accordingly based on the material balance principle to obtain the adjusted production speed combination. The fiber mass flow rate passing through the screen per unit time is calculated by multiplying the screening section speed and the target screen aperture value in the adjusted production speed combination and dividing by the fiber suspension concentration in the slurry tank. The ratio of the fiber mass flow rate to the total mass flow rate of the input slurry is determined as the adjusted material throughput. Based on the adjusted material throughput, it is determined whether it is lower than a preset minimum throughput threshold. If it is lower, the screen vibration frequency is increased. The increase in vibration frequency is proportional to the deviation of the material throughput. The material throughput is recalculated under the new vibration conditions.

[0046] In one embodiment, the construction of the relevant subset begins with data filtering of the dynamic relationship matrix. When the aggregation density exceeds a risk threshold, the fiber throughput values ​​at all time points in the matrix are automatically scanned. The preset throughput threshold is typically set to 0.3; values ​​below this indicate severe obstruction of fiber flow. All rows with throughput below 0.3 are identified, and the corresponding material transfer rate data are extracted. The material transfer rate reflects the actual flow velocity of the fiber in the pipe, measured in meters per second. The screen resistance coefficient is calculated based on the characteristics of throughput changes between adjacent time points. When the throughput decreases from 0.4 to 0.2, the change is 0.2, and its reciprocal of 5 is used as the resistance coefficient. The larger the resistance coefficient, the stronger the obstruction of fiber flow by the screen.

[0047] It should be noted that the average screen resistance coefficient is calculated using a sliding window method, with a window size of 10 consecutive time points. This method can smooth out short-term fluctuations and reflect the overall trend of screen resistance. The difference between agglomeration density and the risk threshold directly reflects the urgency of the current production status; a larger difference indicates a more severe agglomeration problem, requiring more significant adjustments.

[0048] Specifically, the aperture adjustment coefficient is an empirical value derived from historical production data. When the agglomeration density exceeds the risk threshold by 1 unit, the screen aperture needs to be increased by 0.1 mm; when it exceeds by 2 units, it needs to be increased by 0.25 mm; and when it exceeds 3 units or more, the increase increases logarithmically to avoid excessive adjustment that could lead to significant loss of short fibers. Determining the target screen aperture value also needs to consider the physical limitations of the equipment; the screen aperture cannot exceed the maximum allowable value of the equipment, typically 8 mm.

[0049] Preferably, the adjustment of production speed combinations follows the principle of material conservation. As the source section, the speed adjustment of the pulping section directly affects the subsequent material supply. When the screen aperture increases, the resistance to fiber passage decreases, increasing the processing capacity of the screening section; therefore, the screening speed is directly proportional to the aperture ratio. The pulping section speed needs to be reduced accordingly to avoid excessive material accumulation before the screening section. Speed ​​adjustment in the forming section is more complex, requiring dynamic adjustment based on the actual amount of qualified pulp received to maintain the storage silo level within a safe range.

[0050] In one possible implementation, the material balance principle is specifically applied through the mass flow rate conservation equation. The total mass flow rate of the input slurry equals the pulping section velocity multiplied by the slurry density and the pipe cross-sectional area. The fiber mass flow rate through the screen is determined by the screening section velocity, the target screen aperture, and the fiber suspension concentration. The fiber suspension concentration typically fluctuates between 10 and 20 grams per liter; the higher the concentration, the more fiber is present per unit volume of slurry, and the greater the fiber mass flow rate through a screen with the same aperture. The adjusted material throughput is defined as the ratio of the actual fiber mass flow rate to the total input mass flow rate; this ratio directly reflects the screening efficiency.

[0051] For example, when the material throughput is lower than the preset minimum throughput threshold of 0.4, it indicates that even with adjustments to the screen aperture, fiber passage remains difficult. At this point, the vibration frequency adjustment mechanism is activated. Vibration breaks up fiber entanglement, improving throughput efficiency. The increase in vibration frequency is calculated by multiplying the deviation of the material throughput from the minimum threshold by a frequency adjustment coefficient, typically 100 Hz per unit deviation.

[0052] Understandably, vibration frequency cannot be increased indefinitely, as excessively high frequencies can cause fatigue damage to the screen. The upper limit of vibration frequency for industrial screening equipment is typically 50 Hz. Once this limit is reached, the system further enhances the vibration effect by increasing the amplitude. Increasing the amplitude from the standard 2 mm to 4 mm can increase material throughput by 15% to 20%. Furthermore, the recalculation of material throughput under the new vibration conditions takes into account the promoting effect of vibration on fiber dispersion. Vibration loosens fiber aggregates, reducing the actual fiber aggregate size, which is equivalent to increasing the effective passing area of ​​the screen. A vibration effect coefficient is introduced into the calculation formula, which is proportional to the product of vibration frequency and amplitude, with a typical value between 1.2 and 1.5.

[0053] For example, in actual production, when the detected agglomeration density increases from 20 to 25, exceeding the risk threshold of 22, the system automatically executes an adjustment procedure. First, the average resistance coefficient is extracted to be 4.5, calculating that the screen aperture needs to be increased by 0.3 mm. The original aperture of 5 mm is adjusted to 5.3 mm, resulting in an aperture ratio of 1.06. The pulping speed is reduced from 100 liters per minute to 94 liters per minute, and the screening speed is increased from 80 liters per minute to 85 liters per minute. After the adjustment, the material throughput increases from 0.35% to 0.45%, meeting normal production requirements.

[0054] S104. Compare the material throughput rate with the current storage silo capacity level to obtain the capacity matching degree. Based on the capacity matching degree, determine whether it is necessary to expand the buffer storage space. If the material throughput rate exceeds the preset threshold of the capacity level, increase the buffer space allocation ratio.

[0055] The adjusted material throughput rate and the current capacity percentage of the storage silo are obtained. The material throughput rate is multiplied by the adjusted screening section production speed to obtain the material inflow per unit time. The real-time capacity occupancy rate is read from the storage silo level sensor, and the ratio of the material inflow to the remaining capacity of the storage silo is calculated as the capacity matching degree. The capacity matching degree is compared with a preset capacity matching degree threshold, which is set as the critical value for the safe operation of the storage silo. If the capacity matching degree is greater than the threshold, it is determined that the current buffer space is insufficient. The difference between the capacity matching degree and the threshold is used to determine the buffer space expansion demand coefficient. The buffer space expansion demand coefficient is multiplied by the current storage silo capacity to obtain the additional buffer space demand. If the material throughput rate exceeds the preset capacity threshold, the additional buffer space demand is increased by a preset percentage as an emergency buffer amount. The sum of the additional buffer space demand and the emergency buffer amount is divided by the current storage silo capacity to obtain the buffer space allocation ratio increment. The original buffer space allocation ratio is added to the increment to determine the increased buffer space allocation ratio.

[0056] In one embodiment, the capacity matching degree is calculated based on real-time monitoring data and dynamic flow analysis. The material inflow rate is measured by a flow meter at the screening section outlet, which uses electromagnetic induction to accurately measure the volumetric flow rate of fibrous slurry. When the material throughput is 0.5 and the screening section production speed is 85 liters per minute, the material inflow rate is 42.5 liters per minute. The remaining capacity of the storage silo is obtained by subtracting the currently occupied capacity from the total capacity. If the total capacity of the storage silo is 5000 liters and the current capacity occupancy rate is 70%, then the remaining capacity is 1500 liters. The capacity matching degree is 42.5 liters per minute divided by 1500 liters, which is approximately 0.028 per minute.

[0057] It should be noted that the capacity matching threshold is set considering both production continuity and safety margin. The threshold is typically set at 0.02 per minute, meaning the remaining capacity can support at least 50 minutes of continuous production. When the capacity matching exceeds this threshold, it indicates that the storage silo will reach full capacity in a short period of time.

[0058] Specifically, the buffer space expansion demand coefficient reflects the degree of deviation between actual demand and safety standards. When the capacity matching degree is 0.028 and the threshold is 0.02, a difference of 0.008 represents a capacity pressure increase of 0.8% per minute. Multiplying this coefficient by the current storage silo capacity of 5000 liters yields an additional buffer space requirement of 400 liters. This means that an additional 400 liters of buffer capacity is needed to maintain safe operation.

[0059] Preferably, the emergency buffer capacity is set using a tiered response mechanism. When the material throughput exceeds a preset threshold of 0.6 times the capacity level, the system determines it to be a high-flow state, requiring additional emergency buffer capacity. The preset percentage is typically 20%, so the additional demand of 400 liters becomes 480 liters after a 20% increase.

[0060] In one possible implementation, the buffer space allocation ratio is adjusted through a gradual increment. Assuming the original buffer space allocation ratio is 30%, the additional 480 liters divided by 5000 liters yields a 9.6% increment, ultimately adjusting the buffer space allocation ratio to 39.6%. This gradual adjustment avoids drastic system fluctuations and ensures a smooth production transition. In practical applications, the data is relayed to the storage silo control system, automatically adjusting the feed valve opening and discharge speed to achieve dynamic management of the buffer space.

[0061] S105. Extract the fiber speed fluctuations of each section after increasing the buffer space allocation ratio, filter and smooth the fiber speed fluctuations of each section, and integrate the fiber length distribution to predict the material flow trend. Obtain the smoothed production speed through the moving average filtering algorithm.

[0062] After increasing the buffer space allocation ratio, fiber throughput speed data from the pulping, screening, and forming sections are extracted. The absolute value of the speed difference between adjacent time points in each section is calculated as the speed fluctuation value. These speed fluctuation values ​​are arranged in chronological order to form a fluctuation sequence, resulting in a fiber throughput speed fluctuation dataset for each section. For each fluctuation sequence in the fiber throughput speed fluctuation dataset, a moving average filtering algorithm with a preset time window is applied. The arithmetic mean of the fluctuation values ​​within the window is calculated to replace the center point value. The filtering process is completed point by point by sliding the window to obtain a smoothed fluctuation sequence. The smoothed fluctuation sequence is then weighted and fused with the fiber length distribution. The proportion of long fibers is used as a weighting coefficient multiplied by the fluctuation value of the pulping section, the proportion of medium fibers is multiplied by the fluctuation value of the screening section, and the proportion of short fibers is multiplied by the fluctuation value of the forming section. The weighted fluctuation values ​​are summed to obtain a comprehensive fluctuation index. By calculating the difference of the comprehensive fluctuation index at consecutive time points, the time series change rate is obtained. If the change rate is positive, the material flow shows an accelerating trend; if it is negative, it shows a decelerating trend. The material flow trend value for the next time period is predicted based on the magnitude of the absolute value of the change rate. The size of the moving average filter window is adjusted using the material flow trend value. When the trend is accelerating, the window is shrunk to the original window multiplied by a preset shrinkage coefficient. When the trend is decelerating, the window is expanded to the original window multiplied by a preset expansion coefficient. The adjusted window is used to perform moving average filtering on the original production speed data of each section to obtain the smoothed production speed of each section.

[0063] In one embodiment, the fiber extraction process begins from the system state after buffer space adjustment, through a velocity fluctuation extraction process. As the buffer space allocation ratio increases from 30% to 39.6%, the operating characteristics of each section change. The fiber throughput velocity in the pulping section fluctuates from 94 liters per minute to 96 liters per minute, in the screening section from 85 liters to 88 liters per minute, and in the forming section from 78 liters to 82 liters per minute. The velocity fluctuation value is calculated using the difference method between adjacent sampling points, with a sampling interval set to 10 seconds. If the velocity in the pulping section is 94 liters per minute at the first time point and 95.5 liters per minute at the second time point, the velocity fluctuation value is 1.5 liters per minute. These fluctuation values ​​are arranged chronologically to form a fluctuation sequence reflecting the dynamic characteristics of each section.

[0064] It's important to note that the window size in the moving average filtering algorithm directly affects the smoothing effect. The initial window size is typically set to 5 data points, corresponding to a 50-second time span. During filtering, the five fluctuation values ​​within the window are summed and divided by 5 to obtain the average value, which replaces the original value at the window's center point. The window slides forward point by point, moving one data point at a time, until the entire sequence has been traversed.

[0065] Specifically, the weighted fusion of fiber length distribution and fluctuation sequence reflects the differences in the impact of different fiber types on each stage. Long fibers mainly affect the pulping stage because they require more mechanical action to disperse during pulping; medium fibers mainly affect the screening stage, as the matching relationship between their length and screen aperture determines the screening efficiency; short fibers mainly affect the forming stage, as the uniform distribution of short fibers directly affects the consistency of paper thickness. Assuming a long fiber ratio of 40%, a medium fiber ratio of 35%, and a short fiber ratio of 25%, with a smoothed fluctuation value of 1.2 for the pulping stage, 0.8 for the screening stage, and 0.5 for the forming stage, the comprehensive fluctuation index is 0.4×1.2+0.35×0.8+0.25×0.5=0.885.

[0066] Preferably, the prediction of material flow trends is based on the time evolution characteristics of a comprehensive volatility index. The comprehensive volatility index is calculated continuously for 10 time points. If the index value gradually increases from 0.885 to 1.05, it indicates that the system volatility is increasing. The time series rate of change is calculated using the difference between the index values ​​at two adjacent time points. A positive rate of change indicates increased volatility and an accelerating trend in material flow; a negative rate of change indicates decreased volatility and a tendency for material flow to stabilize. The absolute value of the rate of change reflects the severity of the trend change; the larger the absolute value, the faster the system state changes.

[0067] In one possible implementation, the moving average filter window's dynamic adjustment mechanism automatically responds based on the predicted flow trend. When an accelerating trend is detected, the system determines that a faster response is needed and therefore shrinks the filter window. The preset shrinkage factor is typically 0.6, reducing the original window of 5 data points to 3. A smaller window can track speed changes faster, but the smoothing effect is reduced. Conversely, when a decelerating trend is detected, the system determines that a greater delay is acceptable for better smoothing, with a preset expansion factor of 1.5, expanding the window to 7 or 8 data points.

[0068] For example, the adjusted filtering is applied to the raw production speed data for each section. Raw production speed refers to real-time acquired data without any processing, containing all noise and disturbances. Filtering this raw data using an adjusted window effectively suppresses noise while maintaining system responsiveness. The pulping section uses a 3-point window to obtain a smooth speed with a fast response, the screening section uses a standard 5-point window, and the forming section uses a large 7-point window to obtain a highly smooth speed curve.

[0069] Understandably, the smoothed production speeds of each section constitute a set of coordinated control parameters. These parameters reflect the stable operating speed that each section should maintain under the current fiber distribution and agglomeration state. The smoothed speed of the pulping section stabilizes at 95 liters per minute, the screening section at 86 liters per minute, and the molding section at 80 liters per minute. Furthermore, the smoothed production speeds also need to consider the material balance between sections. The output of the pulping section must match the processing capacity of the screening section, and the qualified pulp output of the screening section must meet the needs of the molding section. The smoothed speeds obtained through moving average filtering eliminate the interference of short-term fluctuations, enabling each section to operate in coordination at a stable speed and avoiding material accumulation or flow interruptions caused by speed fluctuations.

[0070] S106. By comparing the smoothed production speed with the dynamic relationship matrix, the flow stability is evaluated. If the material flow stability deviation is not less than the preset deviation threshold, the storage silo capacity requirement is recalculated based on the stability deviation value to obtain the target material flow parameter set.

[0071] By comparing the smoothed production speed of each section with the material transfer rate at the corresponding time point in the dynamic relationship matrix, the speed difference at each time point is calculated. The standard deviation of the speed difference sequence is calculated, and the ratio of the standard deviation to the difference between the maximum and minimum values ​​of the capacity change sequence in the matrix is ​​multiplied to obtain the flow stability index. Based on the difference between the flow stability index and the preset stability benchmark value, the stability deviation value is obtained. If the stability deviation value is not less than the preset deviation threshold, the material flow is determined to be unstable. The difference between the maximum and minimum fiber throughput values ​​within the time period corresponding to the deviation value is extracted from the dynamic relationship matrix as the fluctuation amplitude. The product of the fluctuation amplitude and the stability deviation value is divided by the current storage silo capacity occupancy rate to obtain the capacity adjustment coefficient. The original storage silo capacity is multiplied by the capacity adjustment coefficient and then by the preset expansion coefficient to recalculate the target storage silo capacity requirement. The target production speed is obtained by dividing the target storage silo capacity requirement by the material consumption per unit time. The calculation formula is as follows:

[0072]

[0073] V t C represents the target production speed. s R represents the target storage silo capacity requirement. c This represents the material consumption per unit time; the adjusted screen aperture is determined by dividing the target production speed by the weighted sum of fiber lengths corresponding to the proportions of long, medium, and short fibers, respectively.

[0074] L w =P l ·L l +Pm ·L m +P s ·L s

[0075] L w P represents the weighted sum of fiber lengths. l L represents the proportion of long fibers. l P represents the fiber length corresponding to long fibers. m L represents the proportion of medium fiber. m P represents the fiber length corresponding to the medium fiber. s L represents the proportion of short fibers. s This represents the fiber length corresponding to short fibers; the adjusted screen aperture Da = Vt / Lw. By integrating the adjusted screen aperture, target production speed, and target storage silo capacity requirement, a target material flow parameter set is constructed. The adjusted screen aperture serves as the screening control parameter, the target production speed as the coordination parameter for each section, and the target storage silo capacity requirement as the buffer configuration parameter.

[0076] In one embodiment, the assessment of flow stability is based on a precise comparison between the smoothed production speed and the dynamic relationship matrix. The smoothed production speed represents the stable operating state after filtering, while the material transfer rate in the dynamic relationship matrix reflects the actual material flow. The point-by-point comparison method compares two sets of data at the same time stamp, calculating the speed difference at each time point. For example, assuming the smoothed pulping speed at a certain time point is 95 liters per minute, and the corresponding material transfer rate in the matrix is ​​92 liters per minute, the difference is 3 liters per minute. Difference data from 100 consecutive time points are collected to form a difference sequence. The standard deviation is calculated using statistical methods: first, the average of the difference sequence is calculated; then, the sum of squares of the deviations of each difference from the average is calculated, divided by the number of data points, and the square root is taken to obtain the standard deviation. The difference between the maximum and minimum values ​​of the capacity change sequence reflects the fluctuation range of the storage silo capacity; when the storage silo capacity fluctuates from 3500 liters to 4200 liters, the difference is 700 liters. Multiplying the standard deviation by the ratio of 700 liters yields a flow stability index that comprehensively reflects the coupling effect of velocity deviation and capacity fluctuation.

[0077] It should be noted that the stability benchmark value is set based on statistical analysis of historical production data. By analyzing the distribution of flow stability indicators under normal production conditions, the upper limit of the 95% confidence interval is taken as the benchmark value. When the actual indicators exceed the benchmark value, it indicates that the system has deviated from its normal operating state.

[0078] Specifically, extracting the fluctuation range of fiber throughput requires precisely locating the time period corresponding to the deviation. When the stability deviation value is 0.15, the system automatically searches the dynamic relationship matrix for data segments 30 seconds before and after the occurrence of this deviation value. Within this time period, the fiber throughput may fluctuate from 0.45 to 0.65, with a fluctuation range of 0.2. This fluctuation range directly reflects the degree of influence of fiber agglomeration on throughput efficiency.

[0079] Preferably, the calculation of the capacity adjustment coefficient comprehensively considers the interaction of multiple factors. The product of the fluctuation amplitude of 0.2 and the stability deviation value of 0.15 is 0.03, which characterizes the degree of instability of the system. If the current storage silo capacity occupancy rate is 70%, then the capacity adjustment coefficient is 0.03 divided by 0.7, approximately 0.043. The original storage silo capacity of 5000 liters multiplied by 1.043 equals 5215 liters, which, multiplied by the preset expansion coefficient of 1.1, results in a final target storage silo capacity requirement of 5736 liters. The preset expansion coefficient is introduced to provide additional safety margin and avoid insufficient capacity under extreme operating conditions.

[0080] In one possible implementation, the reverse calculation of the target production speed involves material balance calculations. Material consumption per unit time is determined by the paper output rate of the forming section; assuming a production rate of 80 kg of sheet paper per minute, the corresponding pulp consumption is 100 liters per minute. The target storage silo capacity of 5736 liters divided by 100 liters per minute yields a buffer time of 57.36 minutes. To maintain this buffer time, the target production speed needs to be adjusted to a function of capacity demand and buffer time.

[0081] For example, the adjusted screen aperture is determined using a weighted average method. A 40% long fiber percentage corresponds to a fiber length of 6 mm, a 35% medium fiber percentage corresponds to 4 mm, and a 25% short fiber percentage corresponds to 2 mm. The weighted sum yields an average fiber length of 4.3 mm. Dividing the target production speed of 100 liters per minute by 4.3 mm gives a speed-to-length ratio of approximately 23.3. Based on an empirical formula, the screen aperture should be set to 0.22 times the speed-to-length ratio, i.e., 5.1 mm. This aperture value ensures efficient throughput while minimizing the loss of excessive short fibers.

[0082] Understandably, the construction of the target material flow parameter set achieves synergistic optimization of multiple parameters. The adjusted screen aperture of 5.1 mm serves as a screening control parameter, directly affecting fiber screening efficiency; the target production speed of 100 liters per minute serves as a coordination parameter for each stage, ensuring balanced material flow between stages; and the target storage silo capacity of 5736 liters serves as a buffer configuration parameter, providing sufficient buffer space to cope with production fluctuations. Furthermore, there are interdependent relationships among the parameters in the parameter set. Increasing the screen aperture improves material throughput but may reduce paper quality; increasing production speed increases output but increases pressure on the storage silo; expanding the storage silo capacity provides greater buffer space but increases equipment investment. By integrating these three parameters into a unified parameter set, the system can find a balance between quality, efficiency, and cost.

[0083] For example, in practical applications, when the flow stability index is detected to rise from 0.8 to 1.2, the system automatically initiates the parameter adjustment program. Through the above calculation process, the adjusted parameter combination is obtained: a screen aperture of 5.1 mm, a target production speed of 100 liters per minute, and a target storage bin capacity of 5736 liters. This set of parameters can effectively cope with production fluctuations caused by fiber agglomeration and maintain the stable operation of the board paper production line.

[0084] S107. Perform paper forming quality prediction processing on the target material flow parameter set, integrate the adjusted screen aperture and buffer storage space to obtain a uniform paper thickness distribution index, and obtain production scheduling parameters related to intelligent scheduling of the paper production line by combining the material flow parameter set.

[0085] Linear regression is performed on the adjusted screen aperture, target production speed, and target storage bin capacity of the target material flow parameters to establish a mapping relationship between the parameters and paper thickness. The predicted paper thickness at different locations is calculated using this mapping relationship, yielding paper thickness distribution data. Based on this paper thickness distribution data, the sum of squares of the differences between the thickness values ​​at each sampling point and the average thickness is calculated. This sum is divided by the number of sampling points and then squared to obtain the standard deviation of thickness. The standard deviation is then divided by the average thickness to obtain the coefficient of variation of thickness. The standard deviation and coefficient of variation are used as indicators of uniform paper thickness distribution. The ratio of the standard deviation of thickness to the adjusted screen aperture is used, along with the proportion of buffer storage space to the total capacity, to obtain a production coordination coefficient. This production coordination coefficient is used to determine the speed adjustment ratio for each process section. The adjusted speed values ​​for each process section are obtained by multiplying the speed adjustment ratio by the target speeds of the pulping, screening, and forming sections. The adjusted speed values ​​of the three sections are combined to form the optimal production speed combination. The material transfer time interval is calculated based on the speed difference between adjacent sections. Based on the material transfer time interval, the delayed start time of the screening section after the pulping section starts, and the delayed start time of the forming section after the screening section starts, are determined. The optimal production speed combination is integrated with the start delay time of each section to construct a production scheduling parameter set.

[0086] In one embodiment, the linear regression process establishes a correlation model between parameters and thickness based on historical production data. The target material flow parameter set includes three key variables: adjusted screen aperture of 5.1 mm, target production speed of 100 liters per minute, and target storage bin capacity of 5736 liters. These parameters are input into the regression model as independent variables, with paper thickness as the dependent variable. The regression coefficients are determined using the least squares method. The regression coefficient for screen aperture reflects the influence of aperture on fiber distribution, the coefficient for production speed reflects the effect of speed on fiber deposition, and the coefficient for storage bin capacity characterizes the contribution of buffer time to fiber uniformity. The established mapping relationship expression is: predicted thickness equals base thickness plus screen aperture coefficient multiplied by aperture value, plus speed coefficient multiplied by speed value, plus capacity coefficient multiplied by capacity value. Through this mapping relationship, the thickness values ​​at different positions in the transverse and longitudinal directions of the paper can be predicted, forming a complete thickness distribution data matrix.

[0087] It should be noted that the standard deviation of thickness and the coefficient of variation are core indicators for evaluating paper quality. The standard deviation reflects the absolute degree of variation in thickness, while the coefficient of variation eliminates the influence of the average thickness and reflects relative variation. Sampling points are typically set at nine locations on the paper: the four corners, the midpoints of the four sides, and the center point, ensuring coverage of the entire paper surface.

[0088] Specifically, the production coordination coefficient calculation integrates two dimensions: quality control and capacity balancing. The ratio of the thickness standard deviation of 0.05 mm to the screen aperture of 5.1 mm is approximately 0.0098, reflecting the impact of screening accuracy on thickness control. The buffer storage space of 2000 liters accounts for 0.35 of the total capacity of 5736 liters, representing the system's buffering capacity. The two are integrated by multiplying the ratio by the proportion and then taking the square root, yielding a production coordination coefficient of approximately 0.059. This coefficient is used to determine the speed adjustment ratio for each section; a larger coefficient indicates a greater need for speed adjustments.

[0089] Preferably, a differentiated speed adjustment strategy is adopted for each section. The pulping section, as the source section, has its speed adjustment ratio set at 1.2 times the production coordination coefficient, i.e., 0.071; the screening section, as an intermediate section, has an adjustment ratio equal to the production coordination coefficient, 0.059; and the molding section, as the final section, has an adjustment ratio of 0.8 times the coefficient, i.e., 0.047. Multiplying the original target speed of 100 liters per minute by their respective adjustment ratios, the adjusted speeds are 107.1 liters per minute for the pulping section, 105.9 liters per minute for the screening section, and 104.7 liters per minute for the molding section. This decreasing speed configuration avoids material accumulation between sections.

[0090] In one possible implementation, the material transfer time interval is calculated based on the speed difference between adjacent sections and the pipe volume. The pipe volume from the pulping section to the screening section is 500 liters, the speed difference is 1.2 liters per minute, and the transfer time interval is 500 divided by 1.2, approximately 417 seconds. The pipe volume from the screening section to the forming section is 400 liters, the speed difference is 1.2 liters per minute, and the transfer time interval is 333 seconds. These time intervals determine the timing of section startup.

[0091] For example, the design of the process coordination sequence follows the principle of continuous material flow. When the pulping section starts at time zero, it needs to wait for the first batch of pulp to fill the pipeline and reach the inlet of the screening section; therefore, the delayed start time of the screening section is set to 420 seconds. Similarly, the molding section starts 335 seconds after the screening section. This tiered start-up method ensures that each section starts working only after receiving a stable material flow, avoiding idling or flow interruption.

[0092] Understandably, the construction of the production scheduling parameter set achieves a balance between quality control and production efficiency. The optimal production speed combination includes adjusted speed values ​​for the three work sections. This set of speed values, after coordination and optimization, ensures both uniform paper thickness and high production efficiency. The work section coordination sequence includes the start-up delay time and shutdown sequence of each work section, ensuring continuous material flow and coordinated equipment operation. Furthermore, this set of production scheduling parameters can interface with the intelligent scheduling system of the board paper production line. The scheduling system, based on order requirements and equipment status, calls the corresponding parameter set and automatically sets the operating parameters for each work section. When producing board paper of different specifications, the system recalculates and adjusts the parameter set to achieve flexible production.

[0093] For example, in actual production, when receiving an order for board paper with a thickness requirement of 2.5 mm and a uniformity requirement of less than 5% coefficient of variation, the system first runs a quality prediction model to confirm that the current parameter set meets the requirements. Then, it starts each section sequentially according to the coordinated timing of the work sections: the pulping section runs at a speed of 107.1 liters per minute, the screening section starts at 105.9 liters per minute after 7 minutes, and the forming section starts at 104.7 liters per minute after another 5.5 minutes. This precise timing control and speed matching ensures that the uniformity of the board paper thickness meets the quality standards while maintaining the efficient operation of the production line.

[0094] Based on the screen aperture and storage bin capacity configuration of the target material flow parameters, and combined with the predicted paper thickness uniformity, the speed matching relationship between the pulp preparation section and the forming section is determined. The material transfer rhythm between each section is adjusted according to the thickness distribution uniformity requirements, and the coordinated sequence of sections from fiber separation to paper forming is obtained. The buffer time of each section is determined by the storage bin capacity configuration. The frequency of screen adjustment and the production speed change cycle are integrated to form a production scheduling scheme that includes the start and stop sequence of sections and the timing of speed switching.

[0095] Based on the screen aperture values ​​and storage bin capacity configuration data from the target material flow parameter set, and combined with the predicted paper thickness uniformity, the ratio of the output flow rate of the pulping section to the received flow rate of the forming section is calculated. When the thickness variation coefficient is less than a preset threshold, this ratio is determined as the speed matching coefficient, thus obtaining the speed matching relationship between the pulping and forming sections. According to the speed matching relationship and the thickness distribution uniformity requirements, the time interval for material transfer between each section is calculated. The transfer time from the pulping section to the screening section and the transfer time from the screening section to the forming section are proportionally adjusted to obtain the material transfer cycle sequence. The time difference between the start times of the pulping section, the screening section, and the forming section is determined using this material transfer cycle sequence, forming a section coordination sequence. The buffer time of each section is calculated by dividing the storage bin capacity configuration by the material consumption rate of each section. By comparing the buffer time with the time required to adjust the screen aperture, when the buffer time is greater than a preset multiple of the adjustment time, the screen adjustment frequency is determined to be the integer part of the buffer time divided by the adjustment time. The larger value between the time interval corresponding to the screen adjustment frequency and the production speed change interval is taken to obtain the speed change cycle. Based on the coordination sequence of the work sections and the speed change cycle, the work section start sequence is set as pulping, screening, and molding starting sequentially. A speed switching opportunity is set at the beginning of each speed change cycle. The start / stop sequence and speed switching opportunity are integrated to form a production scheduling plan.

[0096] In one embodiment, the determination of the speed matching coefficient is based on the dual constraints of flow balance principle and quality control requirements. The output flow rate of the pulping section is measured by a flow meter on the outlet pipe, typically 107 liters per minute. The receiving flow rate of the forming section needs to consider fiber loss and moisture evaporation, with an actual receiving volume of approximately 102 liters per minute. The ratio of the two, 1.049, reflects the material transfer efficiency of the system. When the coefficient of variation of paper thickness is below a preset threshold of 5%, it indicates that the current flow ratio can guarantee thickness uniformity, and 1.049 is determined as the speed matching coefficient. The physical meaning of this coefficient is that for every 1.049 units of pulp output from the pulping section, the forming section can process exactly 1 unit, with the excess 0.049 units being discharged through the screening section's slag and absorbed by the buffer in the storage silo. Establishing the speed matching relationship allows each section to maintain material balance at different speeds, preventing any one section from becoming a production bottleneck.

[0097] It should be noted that the material transfer cycle sequence needs to take into account pipeline length, slurry flow rate, and equipment response time. The pipeline length from the pulping section to the screening section is 50 meters, the average slurry flow rate is 2 meters per second, and the theoretical transfer time is 25 seconds. However, in reality, a 15-second start-up preparation time for the screening equipment needs to be added, so the actual transfer time is 40 seconds.

[0098] Specifically, the core of adjusting the cycle time ratio lies in maintaining continuous operation of each section. When the requirement for uniform thickness distribution increases, the processing time of the screening section needs to be extended to ensure sufficient fiber separation. The ratio of the 40-second transfer time from pulping to screening to the 30-second transfer time from screening to forming is adjusted. If the uniformity requirement increases from a coefficient of variation of 5% to 3%, the ratio needs to be adjusted from 4:3 to 5:3, correspondingly extending the processing cycle of the screening section. The adjusted material transfer cycle time sequence is 50 seconds and 30 seconds, and this sequence determines the operating rhythm of each section.

[0099] Preferably, the design of the coordinated timing sequence of the work sections adopts a feedforward control approach. After the pulping section starts at time 0, it takes 50 seconds for the first batch of pulp to reach the inlet of the screening section. The screening section starts at time 50, and it takes 20 seconds to process the first batch of pulp. Adding the 30-second transfer time to the molding section, the molding section should start at time 100. This tiered start method forms a coordinated timing sequence of work sections at 0 seconds, 50 seconds, and 100 seconds. The buffer time of each work section is calculated based on the storage silo capacity and material consumption rate. The storage silo capacity is 5736 liters, the pulping section consumption rate is 107 liters per minute, and the buffer time is 53.6 minutes; the screening section storage silo is 4000 liters, the consumption rate is 105 liters per minute, and the buffer time is 38.1 minutes; the molding section storage silo is 3000 liters, the consumption rate is 102 liters per minute, and the buffer time is 29.4 minutes.

[0100] In one possible implementation, determining the frequency of screen adjustments requires balancing production efficiency and quality control. Adjusting the screen aperture from 5.0 mm to 5.2 mm takes 3 minutes, including a 1-minute downtime, 1.5 minutes of adjustment, and a 0.5-minute restart. When the pulping section's buffer time of 53.6 minutes is more than 15 times the 3-minute adjustment time, screen adjustments can be performed without affecting production continuity. The screen adjustment frequency is determined to be the integer part of 53.6 divided by 3, i.e., 17 times. This means that a maximum of 17 screen adjustments can be performed within one production cycle.

[0101] For example, the determination of the speed change cycle combines two factors: screen adjustment and production rhythm. A screen adjustment frequency of 17 times corresponds to a time interval of 3.15 minutes, while the production speed may change every 5 minutes depending on order requirements. The larger of these two values, 5 minutes, is taken as the speed change cycle to ensure the system has sufficient time to complete adjustments and reach a stable state.

[0102] Understandably, the core of the production scheduling scheme is coordinating the start-up, shutdown, and speed switching of each section. The start-up sequence of the sections strictly follows the order of pulping, screening, and molding, starting at 0 seconds, 50 seconds, and 100 seconds respectively. At the beginning of each 5-minute speed change cycle, the system first adjusts the speed of the pulping section, then adjusts the speed of the screening section after a 50-second delay, and finally adjusts the speed of the molding section after another 50 seconds. This wave-like speed adjustment avoids system oscillation. Furthermore, the timing of speed switching needs to consider the buffering effect of the storage silo. At the 0th minute of each cycle, the pulping section switches from 107 liters per minute to 110 liters per minute; at the 0.83rd minute, the screening section switches from 105 liters per minute to 108 liters per minute; and at the 1.67th minute, the molding section switches from 102 liters per minute to 105 liters per minute. This time difference setting utilizes the buffering capacity of the storage silo to smooth the impact of speed switching on the system.

[0103] For example, in actual production scheduling, the early shift starts production at 8:00 AM, with the pulping section starting first, followed by the screening section at 8:00:50 AM, and the forming section at 8:01:40 AM. The first speed adjustment is made at 8:05 AM, increasing capacity by 5% according to order requirements. The pulping section speed is increased again at 8:05 AM, followed by the screening section at 8:05:50 AM, and the forming section completes its adjustment at 8:06:40 AM. Throughout the adjustment process, the liquid levels in each storage bin are maintained within a safe range of 30% to 70%, achieving a smooth speed transition. This production scheduling scheme, through precise timing control and speed coordination, achieves efficient and stable operation of the board paper production line.

[0104] The above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the invention. Those skilled in the art will understand that implementing all or part of the above-described embodiments and making equivalent changes in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for intelligent scheduling data analysis of a board paper production line, characterized in that, include: Data on fiber length distribution and screen aperture were obtained from the slurry tank, and data on capacity level and production speed of each section were obtained from the storage silo monitoring equipment. The data were then integrated to construct an initial dynamic relationship matrix. Fiber length is identified based on the initial dynamic relationship matrix. Fibers longer than 5 mm are classified as long fibers, fibers between 3 and 5 mm as medium fibers, and fibers shorter than 3 mm as short fibers. The proportion of long fibers is calculated. The concentration of long fibers at the bottom, middle, and top of the slurry tank is collected. The proportion of long fibers is weighted and summed using depth weighting coefficients of 0.5, 0.3, and 0.2 to obtain the comprehensive proportion of long fibers. The weighted average is then used to obtain the ratio of fiber length to screen aperture. The absolute value of the difference in fiber throughput at continuous time points is extracted to determine the agglomeration fluctuation index. The agglomeration density is calculated based on the agglomeration fluctuation index, the ratio of fiber length to screen aperture, and the fiber suspension concentration. Based on the agglomeration density, if the agglomeration density is higher than a preset risk threshold, locate the row index of the fiber throughput vector less than 0.3 preset threshold from the initial dynamic relationship matrix, extract the material transfer rate of the corresponding row as the fiber flow velocity, calculate the reciprocal of the throughput change at adjacent time points as the screen resistance coefficient, construct a subset of fiber flow velocity and screen resistance, calculate the screen aperture adjustment amount based on the difference between the average screen resistance coefficient and the agglomeration density and the preset risk threshold, adjust the screen aperture to determine the target screen aperture, adjust the production speed of each section based on the target screen aperture, combine the adjusted production speed, calculate the fiber mass flow rate passing through the screen per unit time by dividing the product of the screening section speed and the target screen aperture by the fiber suspension concentration, and calculate the material throughput by the ratio of fiber mass flow rate to the total mass flow rate of the input slurry. The material throughput rate is compared with the storage silo capacity level to calculate the capacity matching degree, and the buffer space allocation ratio is adjusted according to the capacity matching degree. After extracting and adjusting the fiber speed fluctuations of each section, the fiber length distribution is integrated and processed using a moving average filtering algorithm to obtain a smoothed production speed. By comparing the smoothed production speed with the material transfer rate at the same moment in the initial dynamic relationship matrix point by point, the standard deviation of the speed difference sequence is calculated. The flow stability is obtained by combining the range of the capacity change sequence. The stability deviation between the flow stability and the preset stability benchmark value is calculated based on the flow stability. The capacity adjustment coefficient is calculated by combining the fiber throughput fluctuation amplitude. The storage silo capacity requirement is adjusted. The adjusted screen aperture, target production speed and target storage silo capacity are integrated to determine the target material flow parameter set. The paper thickness uniformity is predicted based on the target material flow parameter set. The adjusted screen aperture and buffer space allocation ratio are integrated, the thickness standard deviation and coefficient of variation are calculated, and the production scheduling parameters are determined.

2. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process involves acquiring fiber length distribution and screen aperture data from the slurry tank, obtaining capacity levels and production speeds for each section from the storage silo monitoring equipment, and fusing the data to construct an initial dynamic relationship matrix, including: Samples were collected from different locations in the pulp tank, fiber lengths were scanned, the proportions of long, medium, and short fibers were statistically analyzed, and screen aperture values ​​were read. Capacity level data was obtained from the storage silo, timestamped, and a capacity change sequence was generated. Production speeds were collected from the pulping, screening, and forming stages, and the material transfer rate between adjacent stages was calculated. Based on the fiber length distribution and screen aperture values, a fiber throughput vector was calculated, and the corresponding values ​​for long fibers in the fiber throughput vector were adjusted. An initial dynamic relationship matrix was constructed by arranging the fiber throughput vector, material transfer rate, and capacity change sequence in chronological order.

3. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process of identifying fiber length based on an initial dynamic relationship matrix, classifying fibers into long, medium, and short fibers, calculating the proportion of long fibers, and using a weighted average to obtain the ratio of fiber length to screen aperture to determine agglomeration density includes: Fiber throughput vectors are extracted from the initial dynamic relationship matrix to identify long, medium, and short fibers; the proportion of long fibers is statistically analyzed, and the concentration of long fibers at different locations in the pulp tank is collected and weighted by depth to obtain the comprehensive long fiber proportion; the ratio is calculated by dividing the comprehensive long fiber proportion by the screen aperture value; the fiber throughput at continuous time points is extracted, and the absolute value of the throughput difference is calculated to determine the agglomeration fluctuation index; the agglomeration density is calculated based on the agglomeration fluctuation index, the ratio, and the fiber suspension concentration.

4. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process of extracting a subset of fiber flow velocity and screen resistance from the initial dynamic relationship matrix based on agglomeration density, adjusting the screen aperture, and calculating the material throughput in conjunction with the production speed includes: Extract the material transfer rate corresponding to the low throughput row from the initial dynamic relationship matrix as the fiber flow velocity; calculate the reciprocal of the throughput change to determine the screen resistance coefficient and construct a subset; calculate the screen aperture adjustment amount based on the mean of the screen resistance coefficient and the difference in agglomeration density to determine the target screen aperture; adjust the production speed of each section according to the target screen aperture; calculate the fiber mass flow rate by multiplying the screening section speed by the target screen aperture and dividing by the fiber suspension concentration to determine the material throughput.

5. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The step of comparing the material throughput with the storage silo capacity level, calculating the capacity matching degree, and adjusting the buffer space allocation ratio according to the capacity matching degree includes: The material inflow rate is calculated by multiplying the material throughput rate by the screening section speed; the storage silo capacity occupancy rate is obtained, and the ratio of the material inflow rate to the remaining capacity is calculated as the capacity matching degree; the capacity matching degree is compared with a threshold to calculate the difference and determine the buffer space expansion demand coefficient; the additional buffer space demand is calculated using the buffer space expansion demand coefficient, the buffer space allocation ratio increment is determined, and the allocation ratio is adjusted.

6. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The extracted and adjusted fiber speed fluctuations of each section are analyzed, and the fiber length distribution is integrated. A moving average filtering algorithm is then used to process the data to obtain a smoothed production speed, including: Extract the fiber speed difference from each process section to form a fluctuation sequence; apply a moving average filter to the fluctuation sequence, calculate the average value within the window, and obtain a smoothed fluctuation sequence; calculate a comprehensive fluctuation index by weighted fusion of the smoothed fluctuation sequence and the fiber length distribution; determine the material flow trend through the difference of the comprehensive fluctuation index; adjust the filter window size to process the original production speed and obtain a smoothed production speed.

7. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process involves comparing the smoothed production speed with the initial dynamic relationship matrix to calculate flow stability, adjusting the storage silo capacity requirement based on the flow stability, and determining the target material flow parameter set, including: By comparing the smoothed production speed with the material transfer rate point by point, the standard deviation of the speed difference is calculated. Combined with the capacity change sequence, the flow stability index is determined. The stability deviation is calculated by comparing the flow stability index with the benchmark value. The capacity adjustment coefficient is calculated based on the stability deviation and the fiber throughput fluctuation. The capacity of the storage silo is adjusted by the capacity adjustment coefficient to determine the target production speed and screen aperture, and a target material flow parameter set is constructed.

8. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process of predicting paper thickness uniformity based on the target material flow parameter set, integrating the adjusted screen aperture and buffer space allocation ratio, calculating the thickness standard deviation and coefficient of variation, and determining production scheduling parameters includes: A thickness mapping relationship is established using the target material flow parameter set, and the paper thickness distribution is calculated. Based on the paper thickness distribution, the thickness standard deviation and coefficient of variation are calculated. The production coordination coefficient is determined by combining the thickness standard deviation with the adjusted screen aperture ratio and the buffer space ratio. The speed of each section is adjusted using the production coordination coefficient, the material transfer time interval is calculated, and the optimal production speed combination and section coordination sequence are determined.

9. The intelligent scheduling data analysis method for a board paper production line according to claim 1, characterized in that, The process of predicting paper thickness uniformity based on the target material flow parameter set, integrating the adjusted screen aperture and buffer space allocation ratio, calculating the thickness standard deviation and coefficient of variation, and determining production scheduling parameters includes: By combining the screen aperture and storage bin capacity from the target material flow parameter set with the thickness variation coefficient, the speed matching coefficient between the pulping section and the forming section is calculated; the material transfer time interval is adjusted according to the speed matching coefficient to determine the coordination sequence of the sections; the buffer time is calculated by dividing the storage bin capacity by the material consumption rate; by comparing the buffer time with the screen adjustment time, the screen adjustment frequency and speed change cycle are determined, the section start sequence and speed switching timing are set, and a production scheduling plan is constructed.

Images