Special paper production line process collaborative optimization method and system based on anomaly identification

Abstract

This invention provides a method and system for collaborative optimization of processes in a specialty paper production line based on anomaly identification. It relates to the field of process collaborative optimization technology, and includes: synchronously collecting operating parameters of each process node in the specialty paper production line; constructing a process influence matrix to characterize the influence relationships between process nodes; analyzing several deviation value sequences; identifying several abnormal process nodes based on these deviation value sequences; locating the first abnormal process node and marking it as the primary abnormal process node; and classifying the anomaly types of the specialty paper production line by identifying the anomalies of the primary abnormal process nodes, categorizing them into continuous anomalies and independent anomalies. This reduces the shortcomings of traditional control mechanisms that cannot consider the influence of preceding and following processes. Ultimately, this process significantly improves the production stability of the paper web, ensuring the consistency and quality stability of the multi-layered functional structure.

Description

Technical Field

[0001] This invention relates to the field of process collaborative optimization technology, and in particular to a method and system for collaborative optimization of special paper production lines based on anomaly identification. Background Technology

[0002] Multi-layered functional premium box packaging specialty paper is a type of composite paper material used in high-end packaging. It typically uses kraft paper or high-strength base paper as a base, and constructs functional layers such as decorative layers, waterproof layers, flame-retardant layers, and abrasion-resistant layers sequentially on one or both sides of the paper web to meet the comprehensive requirements of packaging materials in terms of appearance, moisture resistance, safety, and durability. Due to the large number of functional layers and complex structure of this type of specialty paper, and the high sensitivity of each functional layer to thickness uniformity, interlayer bonding, and initial material condition, its production process is usually completed in a continuous, multi-stage sequence.

[0003] In actual production, multi-layer functional premium box packaging specialty paper is typically manufactured using roll-to-roll continuous production lines. The production process generally includes base paper unwinding and pretreatment, decorative layer coating, functional layer coating, segmented drying or curing, flipping or double-sided treatment, subsequent functional layer construction, and winding or cutting. The corresponding production equipment mainly includes unwinding devices, various coating devices, drying devices, guiding and tension adjusting devices, and winding devices. The coating device forms the various functional layers on the surface of the paper web during operation. The drying device uses hot air, infrared radiation, or a combination of methods to evaporate the solvent and set the structure of the coated wet coating. All devices are physically connected sequentially via guide rollers and the continuous paper web, forming an uninterrupted production channel.

[0004] Existing specialty paper production lines typically employ a combination of main line speed setting and segmented tension control to achieve coordinated operation between various processes. Specifically, although each coating, drying, and post-processing unit uses independent drive and control, unified line speed setting, proportional following control, and closed-loop tension adjustment in the unwinding and rewinding sections ensure that the paper web can continuously pass through each process and maintain a relatively stable operating state. The main goal of this coordination mechanism is to ensure consistent operating rhythm between equipment, reduce obvious operational failures such as paper breaks, severe wrinkling, or equipment conflicts, thereby maintaining production line continuity and equipment safety.

[0005] The existing technology has the following technical problems:

[0006] Because the aforementioned production line consists of multiple independently controlled devices connected in series along the paper web's running direction, differences inevitably exist among these devices in terms of dynamic response characteristics, mechanical inertia, load variations, and control lag. During actual operation, changes in coating resistance, drying load, or material condition can lead to localized speed inconsistencies during transient adjustments, even under uniform linear speed settings. This can cause stretching, relaxation, or morphological changes in the paper web within certain processing stages. Furthermore, the paper web's moisture content, stiffness, and stress state continuously change during multiple coating and drying processes, making it difficult to maintain perfect tension consistency across different processing stages. All these factors combined result in tension deviations being objectively unavoidable during production.

[0007] Existing tension adjustment and compensation mechanisms typically use tension sensor readings or changes in the displacement of the gyratory rollers as trigger conditions. Upon detecting a deviation, the tension value is corrected by adjusting the driving torque of the unwinding, traction, or rewinding devices within the current control section. This type of compensation is essentially a localized adjustment of an already occurring deviation, primarily aimed at restoring the tension to the set range to ensure continuous paper web operation. However, this compensation mechanism cannot retrospectively correct for changes in the tensile state of the paper web, differences in actual processing time, or the initial formation state of functional layers. Simultaneously, subsequent processes continue to operate independently according to their respective speed and tension settings, failing to detect and coordinate adjustments to material anomalies already formed in preceding processes. This leads to these anomalies being carried over into subsequent coating, drying, and laminating processes as the paper web continues to run, gradually accumulating and amplifying under the combined effects of multiple processes, ultimately forming hidden technical problems that affect the consistency and quality stability of multilayer functional structures. Summary of the Invention

[0008] In view of this, embodiments of the present invention provide a method and system for collaborative optimization of special paper production line processes based on anomaly identification, which can realize a special paper production line with consistent multi-layer functional structure and stable quality.

[0009] The technical solution of this invention is implemented as follows:

[0010] This invention provides a method for collaborative optimization of special paper production line processes based on anomaly identification. The method includes: Step 1: Synchronously collecting the operating parameters of each process node in the special paper production line, constructing a process influence matrix to characterize the influence relationship between process nodes, and analyzing several deviation value sequences; Step 2: Based on the several deviation value sequences, analyzing several abnormal process nodes, locating the first abnormal process node, and marking it as the main abnormal process node. By identifying the abnormality of the main abnormal process node, the anomaly type of the special paper production line is divided into continuous anomalies and independent anomalies; Step 3: If the special paper production line has continuous anomalies, updating the process influence matrix and mapping the adjustment amount of each abnormal process node, generating specific execution content, and completing the allocation of adjustment amount and the coordinated configuration of roller speed; if the special paper production line has independent anomalies, after a deep judgment process, selecting an appropriate optimization strategy to implement the adjustment based on the judgment result.

[0011] This application also provides a collaborative optimization system for a specialty paper production line based on anomaly identification. The system includes: a deviation sequence analysis module, used to synchronously collect operating parameters of each process node in the specialty paper production line, construct a process influence matrix to characterize the influence relationship between process nodes, and analyze several deviation value sequences; an anomaly type classification module, used to analyze several abnormal process nodes based on several deviation value sequences, locate the first abnormal process node, and mark it as the main abnormal process node, classifying the anomaly types of the specialty paper production line into continuous anomalies and independent anomalies by identifying the anomalies of the main abnormal process nodes; and an anomaly strategy implementation module, used to update the process influence matrix and map the adjustment amount of each abnormal process node if the specialty paper production line has continuous anomalies, generate specific execution content, and complete the allocation of adjustment amount and coordinated configuration of roller speed; if the specialty paper production line has independent anomalies, after a deep judgment process, it selects an appropriate optimization strategy to implement adjustment based on the judgment result.

[0012] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0013] 1. This invention, by synchronously collecting operating parameters of each process node in a specialty paper production line and constructing a process influence matrix, effectively characterizes the influence relationships between process nodes and analyzes the deviation value sequence to locate abnormal process nodes. By identifying the main abnormal process nodes and classifying their abnormality types, comprehensive control over the abnormal state of the specialty paper production line can be achieved, solving the problems of speed inconsistency and tension deviation caused by differences in dynamic response between equipment in existing technologies. Specifically, the updating of the process influence matrix and the generation of adjustment values ​​enable each abnormal process node to adjust collaboratively, thereby solving problems such as local speed inconsistency and tension fluctuations, reducing the limitations of traditional technologies that can only make local adjustments and provide post-event compensation. Through this method, the occurrence of abnormalities can not only be identified at the source, but also, through the adjustment of optimization strategies, the consistency of tension and speed of the paper web across each process can be ensured, thereby effectively reducing the accumulation and amplification of stretching, slack, and morphological changes. In addition, by combining in-depth judgment processes and adaptive optimization strategies, targeted adjustments can be implemented according to different abnormality types, reducing the shortcomings of traditional control mechanisms that cannot consider the influence of preceding and following processes. Ultimately, this process significantly improves the production stability of the paper web, ensuring the consistency of the multi-layered functional structure and the stability of its quality.

[0014] 2. This invention effectively solves the problems of tension inconsistency and speed deviation caused by differences in dynamic response, mechanical inertia, load changes, and control lag between multiple independently controlled devices in the prior art by classifying the abnormality types of special paper production lines and combining them with adaptive updates of the process influence matrix. First, by updating the process influence matrix, the influence coefficients of each subsequent process node are extracted based on the current process node. When a process node with abnormal tension or speed is detected, increasing the influence coefficient of the relevant process helps improve the adjustment sensitivity of subsequent process nodes. Furthermore, by allocating adjustment amounts to the main abnormal process node and subsequent process nodes, it ensures that each abnormal process node can undertake adjustment tasks matching its process influence coefficient, thereby accurately adjusting the operating status of each subsequent process node. During this process, by coordinating the rotational speeds of the traction rollers, guide rollers, or take-up and unwind rollers of each subsequent process node, it ensures that the paper web maintains a consistent linear speed relationship throughout the production line, effectively correcting differences in paper stretching, morphological changes, and processing time, and reducing the accumulation of quality problems caused by abnormal transmission in subsequent processes. This method forms a precise closed-loop control system by adaptively adjusting the correlation device, which improves the stability of the entire production process and provides reliable quality assurance for complex multi-layer functional coating structures.

[0015] 3. This invention effectively solves the speed inconsistency and tension deviation problems caused by multiple independently controlled devices in the prior art by triggering a deep judgment process of the main abnormal process node and executing corresponding optimization strategies. First, by obtaining the total deviation of subsequent nodes, it can be determined whether the abnormality of the main abnormal process node will extend to subsequent processes. If the total deviation of subsequent nodes is higher than the defined total deviation of subsequent nodes, a collaborative optimization strategy based on the process influence matrix is ​​executed. By mapping the adjustment amount of the main abnormal process node and reasonably allocating it to subsequent process nodes, coordinated adjustment between multiple processes is achieved, ensuring that the speed adjustment of the traction roller, guide roller, or take-up and unwinding roller within each subsequent process node can be carried out synchronously, ultimately achieving a consistent linear speed and reducing the propagation of subsequent abnormalities caused by individual adjustments. If the total deviation of subsequent nodes is not higher than the defined threshold, a directional optimization strategy is executed. Based on the adjustment amount of the main abnormal process node, the preceding process nodes are adjusted. The influence strength of each preceding process node on the main abnormal process node is analyzed through the process influence matrix, and the adjustment amount is reasonably allocated and the speed is adjusted, thereby solving the problem of tension differences and speed inconsistency between local processes. Ultimately, this method enables coordinated optimization across the entire production line, ensuring the stability of the paper web across different processes, reducing hidden quality problems caused by incoordination between processes, and significantly improving the consistency and stability of specialty paper product quality. Attached Figure Description

[0016] Figure 1 This is a flowchart of a collaborative optimization method for special paper production line processes based on anomaly identification provided in an embodiment of the present invention;

[0017] Figure 2 This is a detailed flowchart illustrating the collaborative optimization process of the special paper production line provided in this embodiment of the invention.

[0018] Figure 3 This is a schematic diagram of the structure of the special paper production line process collaborative optimization system based on anomaly identification provided in an embodiment of the present invention;

[0019] Figure 4 This is a partial schematic diagram of a special paper double-sided multi-layer coating and flipping curing production line provided in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Specialty paper is a multi-layered functional packaging paper primarily used for high-end premium box packaging. This paper achieves excellent properties such as decorative appeal, water resistance, flame retardancy, and abrasion resistance by layering multiple functional layers onto a base paper. The main components of specialty paper include a base paper layer, a decorative layer, a waterproof layer, a flame retardant layer, and an abrasion-resistant layer. The base paper layer provides basic strength and structural stability, the decorative layer enhances the paper's appearance, the waterproof and flame-retardant layers provide protection, and the abrasion-resistant layer increases the packaging's durability and scratch resistance. To ensure the high quality of the final product, these functional layers must be precisely coated and dried during the production process.

[0022] In the production of this specialty paper, the base paper is first introduced into the production line via an unwinding device. After entering the coating unit, the base paper is coated with a decorative layer, a waterproof layer, a flame-retardant layer, and a wear-resistant layer in sequence. Each layer needs to be cured by a drying device after coating. After the front coating is completed, the paper web is flipped over by a flipping device to enter the reverse coating stage, where a waterproof layer, a flame-retardant layer, and a wear-resistant layer are coated. After that, it is cured by a drying device, and finally, the finished paper web is wound into a roll by a winding device.

[0023] Figure 4 This is a partial schematic diagram of a special paper double-sided multi-layer coating and flipping curing production line provided in an embodiment of the present invention. The base paper is introduced into the production line by an unwinding device (not shown in the figure) and enters the front coating section under the action of guide rollers and tension control. The front coating forms a front decorative layer, a front waterproof layer, a front flame retardant layer, and a front abrasion-resistant layer in sequence according to the process sequence. After each coating layer is completed, the paper web enters the corresponding drying device for heating, drying / curing and shaping (not shown in the figure) to ensure stable film formation and reliable interlayer bonding. After the front coating and curing are completed, the paper web is flipped by a flipping mechanism. The flipping guide roller in the figure illustrates its flipping guide path, so that the side to be coated is switched to the reverse side and enters the reverse coating section. In the reverse coating stage, the paper web is coated in sequence to form a reverse waterproof layer, a reverse flame retardant layer, and a reverse abrasion-resistant layer (the specific coating unit of the reverse abrasion-resistant layer is not shown in the figure), and is also cured by a drying device after each coating layer (not shown in the figure). Finally, the paper web, after double-sided multi-layer coating and curing, is wound into a roll by a winding device to obtain the finished special paper roll (not shown in the figure).

[0024] However, in the above production process, since the production line is composed of multiple independently controlled devices connected in series, there are differences in dynamic response characteristics, mechanical inertia, load changes, and control lags among the devices. Therefore, when the coating resistance changes, the drying load changes, or the material state changes, the paper web may have inconsistent local speeds in some process sections, resulting in stretching, relaxation, or morphological changes of the paper web. In addition, due to the continuous change of tension during multiple coating and drying processes, it is difficult for the paper web to maintain exactly the same tension between different process sections, thus causing deviations in the production process.

[0025] Therefore, a method for collaborative optimization of special paper production line processes based on anomaly recognition is proposed. Figure 1 As shown in the flowchart of the method for collaborative optimization of special paper production line processes based on anomaly recognition, the processing flow of this method can include the following steps:

[0026] First, on the special paper production line, for each process node, high-precision sensors are used to synchronously collect relevant operating parameters, mainly including key data such as speed and tension.

[0027] A process node is each independent and clearly functional technological link in the special paper production process. Each process node receives the output of the previous process node and provides input for the next process node. Process nodes are connected physically or functionally to form a production line. For example, a drying node is a drying device responsible for drying the coated paper web through heating or air flow.

[0028] Construct a process influence matrix to describe the influence relationship of the state change of the previous process node on the operating state of the subsequent process node. The process influence matrix M is a matrix containing multiple elements, where each row and each column represents a process node. The element a in the i-th row and j-th column of the matrix i,j represents the process influence coefficient of the i-th process node on the j-th process node. The process influence coefficient characterizes the influence intensity of the state change of one process node on the operating states of other subsequent process nodes. If the tension change in the previous process significantly affects the coating quality of the subsequent process, then the process influence coefficient between them will be relatively large. Here, both i and j represent the numbers of each process node, i = 1, 2, 3,..., y, where y represents the total number of process nodes. j has the same meaning as i. When i = j, it means that the element in the matrix corresponds to the same process node. In the process influence matrix, the influence of each process node on itself is usually set to the maximum value, that is, a ij = 1. When i < j, the element in the matrix represents the influence of the subsequent process node on the previous process node. If there is no direct influence between them, then a ij = 0.

[0029] It should be explained that the process influence matrix is ​​usually formulated by relevant technical personnel. The technical personnel first decompose the complete production process and systematically sort out each process node and its key process parameters. Then, combined with long-term production experience, historical quality data, defect statistics records and on-site test results, they analyze the influence intensity of the state change of one process node on the operating state of other subsequent process nodes and quantify the corresponding values.

[0030] Further, an effective speed deviation value sequence and a tension deviation value sequence are constructed. The effective speed deviation value is formed by synchronously collecting the actual running speed of each process node and subtracting the reference running speed set by the technicians. The effective speed deviation value sequence is formed by arranging the effective speed deviation values ​​of all process nodes in the order of paper web operation.

[0031] It should be noted that although each process node corresponds to different devices and functional units, the paper web follows a unified linear velocity system along the production line during continuous production of specialty paper. Therefore, the operating speed essentially does not refer to the local rotational speed of a single roller, but rather to the linear velocity of the paper web at the process node, that is, the speed at which the paper web passes through that process node in the direction of travel per unit time. Therefore, under normal operating conditions, the target linear velocity of each process node should remain consistent to ensure the continuous and smooth transfer of the paper web between different processes.

[0032] Tension deviation value refers to the difference between the actual tension at each process node and the tension reference value set by the technicians for each process node. The actual tension at each process node refers to the average value of several tension sensors at each process node. By arranging the tension deviation values ​​of all process nodes in the order of paper web operation, a tension deviation value sequence is formed.

[0033] The absolute value of the effective speed deviation value in the effective speed deviation value sequence is compared with the defined effective speed deviation value, where the defined effective speed deviation value represents the maximum allowed absolute value of the effective speed deviation value and is stored in the database.

[0034] If the absolute value of the effective speed deviation is higher than the defined effective speed deviation value, it indicates that the operating speed of the process node has deviated significantly from the reference operating speed. In this case, the process node corresponding to the effective speed deviation value is marked as a speed abnormal process node.

[0035] The absolute value of the tension deviation value in the tension deviation value sequence is compared with the defined tension deviation value, where the defined tension deviation value represents the maximum allowable absolute value of the tension deviation value and is stored in the database.

[0036] If the absolute value of the tension deviation is higher than the defined tension deviation value, it indicates that the tension of the process node has deviated significantly from the tension reference value. The process node corresponding to the tension deviation value is then marked as a tension abnormal process node.

[0037] The first occurrence of a speed or tension anomaly in a process node is identified and marked as the primary anomaly node. This allows for precise identification of the problem's source, ensuring that subsequent adjustment measures address the root cause and reduce the impact on later processes, thereby improving the optimization efficiency of the production process.

[0038] To ensure that the specialty paper production line can be effectively optimized and adjusted in the event of anomalies, we first mark several process nodes that are located after each process node as the subsequent process nodes corresponding to the process node, according to the operating sequence of the paper web on the production line. The subsequent process nodes refer to those process nodes on the specialty paper production line that are located after the current process node and are directly affected by the preceding process node. Their operating status may be affected by abnormal changes in the current process node.

[0039] Based on the anomalies observed in these subsequent process nodes, the type of anomaly in the production line is determined. If tension anomalies or speed anomalies exist in the subsequent process nodes of the main anomaly process node, it indicates that the anomaly has propagated along the production line and affected the stability of subsequent processes. The anomaly type of the specialty paper production line is then marked as a continuous anomaly, and a collaborative optimization strategy targeting each subsequent process node is implemented.

[0040] The collaborative optimization strategy, which takes each subsequent process node as the target, accurately adjusts the operating status of each process node by adaptively updating the process influence matrix, thereby optimizing the production process and solving the problem of inconsistent tension and speed in the production line.

[0041] The specialty paper production line includes multiple process nodes, each corresponding to a different process task (such as coating, drying, turning, etc.). Assuming the main abnormal process node is the coating process node (process 1), and among the subsequent process nodes, only the drying process node (process 2) is the tension abnormal process node, then 'a' can be obtained from the process influence matrix. 1,2 , which represents the process influence coefficient of the coating process node on the drying process node.

[0042] Assuming under normal circumstances, a 1,2 The value is 0.8, indicating that the coating process node has a certain impact on the drying process node. When an abnormal tension is detected in the coating process, it is necessary to extract the increment of the process influence coefficient from the database. Assuming that the increment is 0.2, the updated process influence coefficient is: a 1,2=0.8+0.2=1. The adaptive update process ensures that when an anomaly occurs at a certain process node, the influence coefficient of the main abnormal process node on that process node is enhanced, thereby improving the ability of the main abnormal process node to adjust subsequent processes, ensuring that the operation of subsequent processes can be coordinated more effectively and the production process optimized.

[0043] Among them, the process influence coefficient increment represents the value that the process influence coefficient needs to increase. The database stores mapping tables between different parameters and the process influence coefficient increment, such as the effective speed deviation value-process influence coefficient increment mapping table and the tension deviation value-process influence coefficient increment mapping table. If a node is a tension abnormal process node, the corresponding process influence coefficient increment is retrieved from the tension deviation value-process influence coefficient increment mapping table. If a node is a speed abnormal process node, the corresponding process influence coefficient increment is retrieved from the effective speed deviation value-process influence coefficient increment mapping table.

[0044] If a node is both a speed abnormality process node and a tension abnormality process node, the specific method for obtaining the process influence coefficient increment is as follows: Based on the effective speed deviation value of the node, the corresponding process influence coefficient increment is retrieved from the effective speed deviation value - process influence coefficient increment mapping table; based on the tension deviation value of the node, the corresponding process influence coefficient increment is retrieved from the tension deviation value - process influence coefficient increment mapping table.

[0045] The effective velocity deviation value of the node is normalized, and the tension deviation value of the node is also normalized. The ratio of the two is then processed, and the final result is marked as the proportional coefficient.

[0046] Multiply the increment of the process influence coefficient corresponding to the effective speed deviation value of the node by the proportional coefficient, then multiply the result of subtracting the proportional coefficient from 1 by the increment of the process influence coefficient corresponding to the tension deviation value of the node, and sum the two products to obtain the final increment of the process influence coefficient.

[0047] After the process influence matrix is ​​updated, the total process influence coefficient of each subsequent process node corresponding to the main abnormal process node is calculated. The total process influence coefficient of a certain process node is calculated by gradually accumulating the influence coefficients of the main abnormal process node and its subsequent process nodes on the target process node.

[0048] Suppose we need to calculate the total process influence coefficient of the third process node after the main abnormal process node. First, we calculate the direct influence of the main abnormal process node on the target process node (i.e., the third process node), which is i+3=k, representing element a in the process influence matrix. i,kLet represent the impact of the main abnormal process node on the target process node, where k represents the number of the target process node, k = 1, 2, 3, ..., y. Next, consider the impact of the first process node following the main abnormal process node on the target process node, i.e., a. (i+1),k And add the impact of the second process node after the main abnormal process node a. (i+2),k Finally, the influence coefficient 'a' of the target process node itself. k,k This is added to the total influence coefficient, representing the impact of changes in the state of this process node on its operational state. Total Process Influence Coefficient T total For a i,k +a (i+1),k +a (i+2),k +a k,k .

[0049] Subsequent process nodes whose total process influence coefficient exceeds the defined process influence coefficient are marked as abnormal process nodes. From the perspective of inter-process influence transmission and contribution to quality results, this further identifies nodes that do not show obvious parameter out-of-bounds behavior but have a key amplifying or superimposed effect on defect formation in the overall process chain. This step avoids missing hidden abnormal processes by relying solely on a single parameter threshold, thus achieving a more comprehensive and accurate location of abnormal processes and providing a more reliable basis for subsequent targeted optimization and process adjustment. The defined process influence coefficient represents the maximum allowable value of the total process influence coefficient and is stored in the database.

[0050] The main abnormal process node and the corresponding tension deviation value and effective speed deviation value of each abnormal process node are located. The tension deviation value is normalized and then multiplied by the weight corresponding to the tension deviation value. The effective speed deviation value is normalized and then multiplied by the weight corresponding to the effective speed deviation value. The two results are accumulated and marked as the total deviation of the main abnormal process node and the corresponding deviation of each abnormal process node. The weight is stored in the database and describes the degree of influence of different parameters on the total deviation.

[0051] The corresponding adjustment amount is retrieved from the total deviation-adjustment amount mapping table in the database. If the total deviation is less than 0, the adjustment direction is to increase; if the total deviation is greater than 0, the adjustment direction is to decrease.

[0052] The adjustment amounts of the main abnormal process node and each abnormal process node are allocated to the corresponding subsequent process nodes. Assume there are processes X, X+1, X+2 and X+3, where process X is the main abnormal process node and process X+2 is an abnormal process node.

[0053] Suppose that process X retrieves an adjustment amount of +10 units from the database, indicating an increasing adjustment direction. This adjustment amount needs to be allocated to subsequent process nodes (processes X+1, X+2, and X+3). Based on the data in the process influence matrix, assuming that the process influence coefficient of process X on X+1 is 0.7, the process influence coefficient of process X on X+2 is 0.2, and the process influence coefficient of X on X+3 is 0.1, then the allocation of the adjustment amount is as follows: the sub-adjustment amount allocated to X+1 is +10 units × 0.7 = +7 units, the sub-adjustment amount allocated to X+2 is +10 units × 0.2 = +2 units, and the sub-adjustment amount allocated to X+3 is +10 units × 0.1 = +1 unit.

[0054] Suppose that process X+2 retrieves an adjustment amount of -6 units from the database, indicating that the adjustment direction is decreasing. This adjustment amount needs to be allocated to the subsequent process node X+3. According to the data in the process influence matrix, assuming that the process influence coefficient of process X+2 on X+3 is 0.2, then the adjustment amount allocated to X+3 is -6 units × 0.2 = -1.2 units. Therefore, the total sub-adjustment amount of process X is +10 units, the total sub-adjustment amount of process X+1 is +7 units, the total sub-adjustment amount of process X+2 is +2 units + (-6 units) = -4 units, and the total sub-adjustment amount of X+3 is +1 unit + (-1.2 units) = -0.2 units.

[0055] Determine the speed adjustment values ​​of each traction roller, guide roller, or take-up / unwind roller within the subsequent process node, and coordinate the speed adjustment values ​​while maintaining continuous paper web operation, so that each traction roller, guide roller, or take-up / unwind roller within the subsequent process node maintains a consistent linear speed relationship after adjustment.

[0056] In this embodiment, the main abnormal process node of the specialty paper production line ensures the speed adjustment of the traction rollers, guide rollers, and take-up / unwind rollers of subsequent process nodes through precise adjustment amount allocation, thereby achieving the consistency of the linear speed of the entire production line. The adjustment amount allocation process is based on the process influence matrix and the corresponding adjustment ratio to ensure the coordination between the operating status of each process node and other nodes.

[0057] The adjustment amount of each process node is directly related to the change in its operating speed, which is specifically defined by the adjustment amount-operating speed change correlation table stored in the database.

[0058] Based on the calculated changes in operating speed, these speed changes are converted into specific roll speed adjustments by consulting the operating speed change-rotation adjustment mapping table in the database. Assume the rotational speed adjustment coefficients in the mapping table are as follows: Traction Roller: For every 1 m / s increase in speed change, the traction roller rotational speed is adjusted to 5 RPM; Guide Roller: For every 1 m / s increase in speed change, the guide roller rotational speed is adjusted to 3 RPM; Take-up and Untake-up Rollers: For every 1 m / s increase in speed change, the take-up and untake-up roller rotational speed is adjusted to 2 RPM.

[0059] Therefore, for the adjustment amount and running speed change of each process node, the corresponding roller speed adjustment value is obtained through the mapping table: the adjustment amount of process X+1 is +7 units, and the corresponding running speed change is found to be 3.5m / s from the section amount-running speed change association table. Therefore: traction roller speed adjustment: 3.5×5=+17.5 RPM; guide roller speed adjustment: 3.5×3=+10.5 RPM; take-up and unwinding roller speed adjustment: 3.5×2=+7 RPM; the same applies to processes X+2 and X+3.

[0060] In this implementation, both positive and negative signs indicate increase or decrease; that is, a positive sign indicates increase, and a negative sign indicates decrease.

[0061] By coordinating and configuring the rotation speed adjustment values ​​for each process node, the consistency of the paper web's linear speed on the specialty paper production line is ensured. While maintaining continuous paper web operation, the rotation speeds of the traction rollers, guide rollers, and take-up / unwind rollers in subsequent process nodes are adjusted to ensure consistent linear speed across all process nodes, eliminating tension and speed deviations caused by abnormal process nodes.

[0062] If there are no tension or speed abnormality process nodes in the subsequent process nodes of the main abnormal process node, it means that the abnormality is only concentrated in the main abnormal process node itself and has not spread to the subsequent process nodes. The abnormality type of the special paper production line is marked as an independent abnormality, triggering the in-depth judgment process of the main abnormal process node to determine the optimization strategy.

[0063] Obtaining the total deviation of subsequent nodes of the main abnormal process node specifically involves first analyzing the total deviation of each subsequent process node corresponding to the main abnormal process node, summing them up, and marking the summed result as the total deviation of subsequent nodes of the main abnormal process node. The total deviation of subsequent nodes serves as a benchmark to help determine whether the abnormality has spread further in the production line and affected the stability of subsequent processes.

[0064] If the total deviation of subsequent nodes is higher than the defined total deviation of subsequent nodes, it indicates that the abnormal impact has spread to multiple subsequent processes and may cause more extensive production problems. In this case, a collaborative optimization strategy for subsequent processes based on the process impact matrix is ​​executed. The defined total deviation of subsequent nodes represents the maximum allowable value of the total deviation of subsequent nodes and is used to determine whether the abnormality of the main abnormal process node continues to subsequent process nodes.

[0065] If the total deviation of subsequent nodes is not higher than the defined total deviation of subsequent nodes, it means that the impact of the anomaly is limited to the main abnormal process node itself and has not had a significant impact on subsequent processes. In this case, a targeted optimization strategy for the main abnormal process node will be executed.

[0066] The subsequent process collaborative optimization strategy based on the process influence matrix specifically refers to: based on the total deviation of the subsequent nodes of the main abnormal process node, querying the adjustment amount and direction of the main abnormal process node from the total deviation-adjustment amount mapping table; for example, if the main abnormal process node is process X+1, the query adjustment amount is +2 units, and the process influence coefficient of X+1 on X+3 is 0.2, according to the process influence matrix, assuming that the process influence coefficients of process X+1 on subsequent process nodes (such as processes X+2 and X+3) are 0.6 and 0.4 respectively. Based on these influence coefficients, the adjustment amount is allocated as follows: Sub-adjustment amount for process X+2: +2 × 0.6 = +1.2 units, indicating that the speed of process X+2 needs to be increased. Sub-adjustment amount for process X+3: +2 × 0.4 = +0.8 units, indicating that the speed of process X+3 also needs to be increased.

[0067] Based on the adjustment amount for each subsequent process node, the corresponding change in operating speed is calculated. Assuming that the speed adjustment values ​​for the traction roller, guide roller, and take-up / unwind roller per unit adjustment are 0.5 RPM for the traction roller, 0.3 RPM for the guide roller, and 0.2 RPM for the take-up / unwind roller, the operating speed change for each process node is calculated based on these ratios: The adjustment amount for process X+2 is +1.2 units, resulting in: Traction roller speed adjustment: +1.2 × 0.5 = +0.6 RPM; Guide roller speed adjustment: +1.2 × 0.3 = +0.36 RPM; Take-up / unwind roller speed adjustment: 1.2 × 0.2 = +0.24 RPM; The same applies to process X+3.

[0068] It's important to explain that optimization strategies for continuous anomalies focus on the coordinated adjustment of multiple processes. By ensuring the distribution of adjustment amounts among process nodes, they eliminate the continuous impact of anomalies on the production line. In contrast, optimization strategies for independent anomalies target the main anomalous process node, primarily addressing the anomaly at a single node and avoiding over-adjustment of the entire production line. Based on this difference, strategies for continuous anomalies can more effectively eliminate the chain reaction triggered by the main anomalous process node, ensuring coordination among process nodes on the production line. Independent anomalies, on the other hand, focus more on optimizing local anomalies, ensuring that they do not unnecessarily affect the normal operation of subsequent process nodes. Through this optimization, the most appropriate adjustment strategy can be selected under different anomaly scenarios, improving the overall stability of the production line and the consistency of product quality.

[0069] The targeted optimization strategy for the main abnormal process node specifically refers to: based on the total deviation of the subsequent nodes of the main abnormal process node, querying the adjustment amount and direction of the main abnormal process node from the total deviation-adjustment amount mapping table.

[0070] The adjustment amount of the main abnormal process node is compared with the defined adjustment amount. If the adjustment amount of the main abnormal process node is less than the defined adjustment amount, no targeted optimization is needed. The defined adjustment amount refers to the maximum allowable adjustment amount, stored in the database. The comparison here only focuses on the numerical part of the adjustment amount. The defined adjustment amount is set to ensure the rationality and efficiency of the optimization measures. When the adjustment amount of the main abnormal process node is less than the defined adjustment amount, it indicates that the abnormality of this process node is relatively minor, and its impact on the entire production line is relatively small. At this time, other process nodes on the production line can still maintain relatively stable operation, and there is no need to take overly aggressive adjustment measures. Continuing to execute the optimization strategy may lead to unnecessary adjustments and waste of resources. Therefore, in this case, the system chooses not to execute the targeted optimization strategy to avoid affecting production efficiency and stability.

[0071] If the adjustment amount of the main abnormal process node is not less than the defined adjustment amount, then the running speed of several preceding process nodes corresponding to the main abnormal process node is reduced. For example, the adjustment amount of the main abnormal process node X is +1 unit. Assume that the process influence coefficient of the preceding process node X-1 on process X is 0.6, and the process influence coefficient of the preceding process node X-2 on process X is 0.4. To ensure that the adjustment tasks of subsequent process nodes are consistent with their roles in production, the allocation of adjustment amounts is determined according to the influence coefficients: the adjustment amount allocated to process node X-1 is +1 × 0.6 = +0.6 units; the adjustment amount allocated to process node X-2 is +1 × 0.4 = +0.4 units. Then, based on the speed adjustment values ​​of the traction roller, guide roller, and take-up / unwind roller corresponding to each unit of adjustment amount, the speeds of process node X-1 and its traction roller, guide roller, and take-up / unwind roller are adjusted.

[0072] Figure 2 This is a detailed flowchart illustrating the collaborative optimization process of a specialty paper production line provided in this embodiment of the invention. During the continuous production of specialty paper, operating parameters are synchronously collected from each process node in the production line to construct a process influence matrix. This matrix characterizes the influence of changes in the state of preceding process nodes on the operating state of subsequent process nodes. The effective speed deviation and tension deviation values ​​of each process node are analyzed, thereby constructing effective speed deviation and tension deviation value sequences. The absolute values ​​of the effective speed deviation values ​​in the effective speed deviation value sequence are compared with the defined effective speed deviation value. If the absolute value of the effective speed deviation value is higher than the defined effective speed deviation value, the process node corresponding to that effective speed deviation value is marked as having a speed anomaly. Process nodes; compare the absolute value of the tension deviation value in the tension deviation value sequence with the defined tension deviation value; if the absolute value of the tension deviation value is higher than the defined tension deviation value, mark the process node corresponding to the tension deviation value as a tension abnormal process node; locate the first occurrence of a speed abnormal process node or a tension abnormal process node, mark it as a main abnormal process node, and classify the abnormality type of the special paper production line by identifying the abnormality of the main abnormal process node. If the abnormality type of the special paper production line is a continuous abnormality, execute a collaborative optimization strategy with each subsequent process node as the target; if the abnormality type of the special paper production line is an independent abnormality, trigger the deep judgment process of the main abnormal process node to determine the optimization strategy.

[0073] Figure 3 This is a schematic diagram of the structure of a collaborative optimization system for special paper production line processes based on anomaly identification provided in an embodiment of the present invention, including a deviation sequence analysis module, an anomaly type classification module, an anomaly strategy implementation module, and a database.

[0074] The database is used to store the parameters involved in the collaborative optimization system for special paper production line processes based on anomaly identification. It is built according to the parameter management, anomaly identification and strategy execution requirements of collaborative optimization of special paper production line processes. It classifies and stores various basic parameters, detection parameters, threshold parameters and coefficient parameters required for system operation, so as to realize the orderly management and rapid retrieval of parameters.

[0075] The deviation sequence analysis module is used to synchronously collect the operating parameters of each process node in the special paper production line, construct a process influence matrix to characterize the influence relationship between process nodes, and analyze several deviation value sequences.

[0076] The anomaly type classification module is used to analyze several abnormal process nodes based on several deviation value sequences, locate the first abnormal process node, and mark it as the main abnormal process node. By identifying the abnormal situation of the main abnormal process node, the anomaly type of the special paper production line is classified into continuous anomalies and independent anomalies.

[0077] The abnormal strategy implementation module is used to update the process influence matrix and map the adjustment amount of each abnormal process node if the special paper production line is a continuous abnormality, generate specific execution content and complete the allocation of adjustment amount and the coordinated configuration of roller speed. If the special paper production line is an independent abnormality, after the in-depth judgment process, the appropriate optimization strategy is selected to implement the adjustment based on the judgment result.

[0078] In Example 2, under the same conditions as in Example 1, if several abnormal process nodes follow the main abnormal process node, these subsequent abnormal process nodes may have a cumulative effect on the total process influence coefficient of certain process nodes, resulting in an abnormal amplification of the total process influence coefficient. To avoid overestimating the actual impact of certain nodes due to the cumulative effect, the abnormally amplified portion is usually quantified based on the increment of the total process influence coefficient, a corresponding correction value is matched, and the total process influence coefficient is corrected accordingly. In this way, the true contribution of each process node to the quality result can be more accurately reflected, avoiding misleading anomaly judgments or process optimization decisions due to the cumulative effect of subsequent abnormal processes.

[0079] Assuming the main abnormal process node is the Y-th process node, and there are two abnormal process nodes in its subsequent process nodes, namely the Y+1-th process node and the Y+2-th process node, then their increments will all affect the Y+3-th process node, causing the total process influence coefficient of the Y+3-th process node to increase.

[0080] Assuming under normal circumstances, a Y,Y+3 The value is 0.8, a Y+1,Y+3 The value is 0.5, a Y+2,Y+3 The value is 0.4, a Y+3,Y+3 The value of a is 1, before the update. Y+3,Y+3 The total process influence coefficient is 0.8 + 0.5 + 0.4 + 1 = 2.7. The process influence coefficient increment of the Y-th process node is extracted from the database as 0.25, the process influence coefficient increment of the (Y+1)-th process node is 0.15, the process influence coefficient increment of the (Y+2)-th process node is 0.1, and the total process influence coefficient increment of the (Y+3)-th process node is 0.25 + 0.15 + 0.1 = 0.5. The corresponding correction value is 0.8, which is found in the total process influence coefficient increment-correction value mapping table in the database. Here, the correction value represents the proportion of the total process influence coefficient increment that is corrected.

[0081] The total process influence coefficient of the updated Y+3 process node is: 2.7 + (0.25 + 0.15 + 0.1) × 0.8 = 3.1. The adaptive update process ensures that when a process node is abnormal, the influence coefficient of the main abnormal process node on that process node is enhanced, thereby improving the ability of the main abnormal process node to adjust subsequent processes, ensuring that the operation of subsequent processes can be more effectively coordinated and the production process optimized.

[0082] The above description is only an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A collaborative optimization method for special paper production line processes based on anomaly identification, characterized in that, The method includes: Step 1: Synchronously collect the operating parameters of each process node in the special paper production line, construct a process influence matrix to characterize the influence relationship between process nodes, and analyze several deviation value sequences. Step 2: Based on several deviation value sequences, analyze several abnormal process nodes, locate the first abnormal process node, and mark it as the main abnormal process node. By identifying the abnormal situation of the main abnormal process node, classify the abnormality type of the special paper production line into continuous abnormality and independent abnormality. Step 3: If the special paper production line is a continuous anomaly, update the process influence matrix and map the adjustment amount of each abnormal process node, generate specific execution content and complete the allocation of adjustment amount and the coordinated configuration of roller speed. If the special paper production line is an independent anomaly, after the in-depth judgment process, select the appropriate optimization strategy to implement the adjustment based on the judgment result.

2. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 1, characterized in that, The analysis yielded several deviation value sequences, specifically referring to: In the continuous production process of specialty paper, operating parameters are collected synchronously at each process node in the specialty paper production line; Construct a process influence matrix to characterize the relationship between the state changes of preceding process nodes and the operating state of subsequent process nodes; Analyze the effective speed deviation and tension deviation values ​​at each process node to construct effective speed deviation value sequences and tension deviation value sequences; The effective speed deviation value sequence refers to the numerical sequence formed by arranging the effective speed deviation values ​​of each process node in the special paper production line in the order of the process of paper web operation. The tension deviation value sequence refers to the numerical sequence formed by arranging the tension deviation values ​​of each process node in the special paper production line in the order of the process of paper web operation.

3. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 1, characterized in that, The analysis identified several abnormal process nodes. The specific analysis process is as follows: The absolute values ​​of the effective speed deviation values ​​in the effective speed deviation value sequence are compared with the defined effective speed deviation values; If the absolute value of the effective speed deviation is higher than the defined effective speed deviation value, then the process node corresponding to the effective speed deviation value is marked as a speed abnormal process node. The absolute values ​​of the tension deviation values ​​in the tension deviation value sequence are compared with the defined tension deviation values; If the absolute value of the tension deviation is higher than the defined tension deviation value, then the process node corresponding to the tension deviation value is marked as a tension abnormal process node. Locate the first abnormal speed or tension process node and mark it as the main abnormal process node. By identifying the abnormal situation of the main abnormal process node, classify the abnormality type of the special paper production line.

4. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 3, characterized in that, The classification of anomalies in the special paper production line is as follows: According to the sequence of processes in the paper web operation of the production line, several process nodes located after the process node are marked as the subsequent process nodes corresponding to the process node. Locate each subsequent process node corresponding to the main abnormal process node. If there are tension abnormal process nodes or speed abnormal process nodes among the subsequent process nodes of the main abnormal process node, mark the abnormality type of the special paper production line as a continuous abnormality and execute a collaborative optimization strategy with each subsequent process node as the target. If there are no tension abnormality process nodes or speed abnormality process nodes in the subsequent process nodes of the main abnormal process node, the abnormality type of the special paper production line will be marked as an independent abnormality, triggering the in-depth judgment process of the main abnormal process node to determine the optimization strategy.

5. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 4, characterized in that, The collaborative optimization strategy that targets each subsequent process node specifically refers to: Based on each subsequent process node, the process influence matrix is ​​adaptively updated sequentially; The adaptive update refers to extracting the process influence coefficient of the current process node on each corresponding subsequent process node from the process influence matrix, based on the current process node. If a subsequent process node is a tension abnormality process node or a speed abnormality process node, the process influence coefficient increment is matched from the database, thereby increasing the process influence coefficient of the current process node on the subsequent process node in the process influence matrix. After the process influence matrix is ​​updated, the total process influence coefficient of each subsequent process node corresponding to the main abnormal process node is calculated. Mark subsequent process nodes whose total process impact coefficient is greater than the defined process impact coefficient as abnormal process nodes; Map the main abnormal process node and the adjustment amount and direction of each abnormal process node; Based on the main abnormal process node and the adjustment amount of each abnormal process node, the specific execution content is generated.

6. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 5, characterized in that, The specific process for generating the execution content is as follows: Distribute the adjustment amounts of the main abnormal process node and each abnormal process node to the corresponding subsequent process nodes. During the allocation of adjustment amounts, the adjustment amount is divided into several sub-adjustment amounts based on the position of the main abnormal process node and the corresponding subsequent process nodes in the special paper production line and the corresponding process influence coefficient, so that each process node undertakes an adjustment task that matches the degree of influence. For any subsequent process node of the main abnormal process node, the total sub-adjustment amount of the subsequent process node is calculated, the speed adjustment value of each traction roller, guide roller or take-up and unwind roller inside the subsequent process node is determined, and the speed adjustment value is coordinated and configured under the premise of maintaining continuous paper web operation so that each traction roller, guide roller or take-up and unwind roller inside the subsequent process node maintains a consistent linear speed relationship after adjustment.

7. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 4, characterized in that, The depth determination process for the triggering major exception process node is used to determine the optimization strategy. The specific process is as follows: Obtain the total deviation of subsequent nodes of the main abnormal process node; If the total deviation of subsequent nodes is higher than the defined total deviation of subsequent nodes, then the collaborative optimization strategy of subsequent processes based on the process influence matrix is ​​executed. If the total deviation of subsequent nodes is not higher than the defined total deviation of subsequent nodes, then a targeted optimization strategy for the main abnormal process node is executed.

8. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 7, characterized in that, The subsequent process collaborative optimization strategy based on the process influence matrix specifically refers to: Based on the total deviation of subsequent nodes of the main abnormal process node, the adjustment amount and direction of the main abnormal process node are mapped. The adjustment amount of the main abnormal process node is allocated to several subsequent process nodes of the main abnormal process node. During the adjustment allocation process, the adjustment amount is divided into several sub-adjustment amounts of subsequent process nodes based on the position of the main abnormal process node in the special paper production line and the corresponding process influence coefficient. For any subsequent process node of the main abnormal process node, the speed adjustment value of each traction roller, guide roller or take-up and unwind roller inside the subsequent process node is determined according to the sub-adjustment amount of the subsequent process node. Under the premise of maintaining continuous paper web operation, the speed adjustment value is coordinated and configured so that each traction roller, guide roller or take-up and unwind roller inside the subsequent process node maintains a consistent linear speed relationship after adjustment.

9. The collaborative optimization method for special paper production line processes based on anomaly identification as described in claim 8, characterized in that, The targeted optimization strategy for the main abnormal process node specifically refers to: Based on the total deviation of subsequent nodes of the main abnormal process node, the adjustment amount of the main abnormal process node is mapped. Compare the adjustment amount of the main abnormal process node with the defined adjustment amount. If the adjustment amount of the main abnormal process node is less than the defined adjustment amount, then no targeted optimization is required. If the adjustment amount of the main abnormal process node is not less than the defined adjustment amount, then optimize the running speed of several preceding process nodes corresponding to the main abnormal process node.

10. A collaborative optimization system for special paper production line processes based on anomaly identification, characterized in that, The system includes: The deviation sequence analysis module is used to synchronously collect the operating parameters of each process node in the special paper production line, construct a process influence matrix to characterize the influence relationship between process nodes, and analyze several deviation value sequences. The anomaly type classification module is used to analyze several abnormal process nodes based on several deviation value sequences, locate the first abnormal process node, and mark it as the main abnormal process node. By identifying the abnormal situation of the main abnormal process node, the anomaly type of the special paper production line is classified into continuous anomalies and independent anomalies. The abnormal strategy implementation module is used to update the process influence matrix and map the adjustment amount of each abnormal process node if the special paper production line is a continuous abnormality, generate specific execution content and complete the allocation of adjustment amount and the coordinated configuration of roller speed. If the special paper production line is an independent abnormality, after the in-depth judgment process, the appropriate optimization strategy is selected to implement the adjustment based on the judgment result.

Images