A raw paper intelligent management system based on multi-modal data

Abstract

This invention relates to the field of intelligent manufacturing technology in logistics, and more particularly to an intelligent management system for raw paper based on multimodal data. The invention includes a feature acquisition module, a feature pre-analysis module, a feature recognition module, a region joint analysis module, and a raw paper storage optimization module. The feature pre-analysis module determines humidity trend parameters to assess the presence of abnormal risks in the raw paper. The feature recognition module performs optical response analysis on each monitored sub-region to mark the feature monitoring sub-regions. The region joint analysis module performs joint response analysis on each feature monitoring sub-region to determine the optimal adjustment method for the hot-pressing roller. The raw paper storage optimization module determines whether the storage of the raw paper needs optimization. This invention enables rapid identification of abnormal risks in raw paper and adaptive adjustment of storage parameters based on the characteristics of the raw paper, improving the effectiveness and reliability of intelligent raw paper management.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology in logistics, and in particular to an intelligent management system for raw paper based on multimodal data. Background Technology

[0002] As a key intermediate product in the papermaking industry, the uniformity of base paper directly determines the performance and yield of subsequent processing steps such as printing, coating, and die-cutting. In the field of base paper production, the uniformity and stability of product quality are core standards for measuring production levels, directly affecting the efficiency of downstream processing steps and the performance of the final product. The uniformity of key parameters such as thickness, moisture content, and fiber distribution of base paper are decisive factors affecting its printability, mechanical strength, and appearance quality. Although the industry has introduced automated equipment for monitoring and controlling the production process, challenges remain in terms of systematic, intelligent, and closed-loop management. Existing systems typically rely on independent sensors to monitor single parameters such as thickness, moisture content, or surface defects. However, the non-uniformity of base paper is often the result of multiple factors working together. Single test data can only reflect the superficial manifestation of final quality defects, and cannot effectively trace and distinguish whether the root cause is due to differences in drying efficiency or changes in mechanical stress. This can easily lead to misjudgment of quality and delayed control, making preventive control impossible. At the same time, the production and warehousing management processes are disconnected, forming information and management silos. Current production quality data is rarely systematically transmitted and used to guide subsequent warehousing and logistics decisions. This disconnect allows potential quality risks to worsen during the storage stage. For example, paper rolls with uneven thickness may undergo permanent deformation under improper stacking, or paper rolls with uneven humidity may experience increased deformation due to differences in moisture absorption during resting, ultimately resulting in value loss and affecting the effectiveness and reliability of intelligent management of raw paper. Therefore, improving the effectiveness and reliability of intelligent management of raw paper is an urgent technical problem to be solved.

[0003] For example, Chinese Patent Application Publication No. CN120013334A discloses a method and system for drying corrugated cardboard in corrugated paper production. Based on the parameters of the raw paper, factory environment, and manufacturer requirements before drying, the required drying characteristics are obtained. A corrugated cardboard production drying model is constructed based on these required drying characteristics and the corresponding corrugating machine drying control method. The real-time required drying characteristics obtained from processing the corrugated paper to be dried are input to obtain a real-time corrugating machine drying control method. This real-time control method assists corrugating machine managers in adjusting the operating parameters of the corrugating machine drying facilities, thereby managing the drying quality of corrugated cardboard production. By effectively processing end-to-end data parameters affecting the drying effect of corrugated paper and improving the drying control accuracy of the corrugating machine through an improved machine learning model, the system assists in managing the drying quality of corrugated cardboard production, reducing labor costs, improving the drying quality of corrugated paper, and increasing customer satisfaction with the high quality of corrugated paper products.

[0004] The following problems still exist in the existing technology:

[0005] Existing technologies do not take into account the product variations that may occur during the production of raw paper. Using the same storage method is difficult to adapt to the condition of all raw paper, which affects the utilization rate of raw paper. Existing technologies cannot quickly identify whether there are abnormal risks in raw paper, nor can they adaptively adjust storage parameters according to the characteristics of raw paper during the production stage, which affects the effectiveness and reliability of intelligent management of raw paper. Summary of the Invention

[0006] To address this, the present invention provides a raw paper intelligent management system based on multimodal data, which overcomes the problems of existing technologies that cannot quickly identify whether there are abnormal risks in raw paper, and cannot adaptively adjust storage parameters according to the characteristics of raw paper in the production stage, thus affecting the effectiveness and reliability of raw paper intelligent management.

[0007] To achieve the above objectives, the present invention provides a paper intelligent management system based on multimodal data, comprising:

[0008] The feature acquisition module is used to acquire the humidity parameters and grayscale parameters of several monitoring sub-regions on the dried paper.

[0009] The feature pre-analysis module, which is connected to the feature acquisition module, is used to determine the humidity trend parameter based on the comparison between the humidity parameters of each monitoring sub-region, so as to determine whether the raw paper has any abnormal risk.

[0010] A feature recognition module, which is connected to the feature acquisition module and the feature pre-analysis module respectively, is used to perform optical response analysis on each of the monitoring sub-regions to obtain morphological mapping characterization parameters, and to mark the feature monitoring sub-regions based on the morphological mapping characterization parameters.

[0011] A regional joint analysis module, which is connected to the feature recognition module, is used to perform joint response analysis on each feature monitoring sub-region to determine the abnormal risk trend category of the raw paper, and to determine the optimization adjustment method of the hot pressing roller based on the abnormal risk trend category of the raw paper. The optimization adjustment method includes determining the increase of the roller body temperature of the hot pressing roller, or determining the decrease of the roller gap of the hot pressing roller in the virtual monitoring area.

[0012] The raw paper storage optimization module, which is connected to the regional joint analysis module, is used to determine whether to optimize the storage of the raw paper based on the humidity trend parameters of the raw paper after optimization and adjustment on the hot pressing roller.

[0013] The optical response analysis involves illuminating each monitoring sub-region with different illumination angles to obtain grayscale parameters associated with each illumination angle. The virtual monitoring region is determined based on several feature monitoring sub-regions.

[0014] Furthermore, the feature pre-analysis module is used to determine whether the base paper has any abnormal risks, wherein,

[0015] The feature pre-analysis module determines that the raw paper has an abnormal risk based on the comparison of humidity parameters between the monitoring sub-regions and the determination result that the comparison meets the risk tendency conditions.

[0016] The risk tendency condition is that the humidity tendency parameter exceeds a preset humidity tendency parameter threshold, and the humidity tendency parameter is the variance of the humidity parameter of each monitoring sub-region.

[0017] Furthermore, the feature recognition module is used to perform optical response analysis on each of the monitored sub-regions to obtain morphological mapping characterization parameters, wherein,

[0018] The feature recognition module obtains the grayscale parameters of the monitored sub-region under different illumination angles, and determines the variance of the grayscale parameters as the morphological mapping representation parameter of the monitored sub-region.

[0019] Furthermore, the feature recognition module is used to label feature monitoring sub-regions based on the morphological mapping representation parameters, wherein,

[0020] The feature recognition module marks the monitored sub-region as a feature monitoring sub-region based on the determination result that the morphological mapping representation parameter of the monitored sub-region exceeds the preset morphological mapping representation parameter threshold.

[0021] Furthermore, the regional joint analysis module is used to perform joint response analysis on each of the feature monitoring sub-regions, wherein,

[0022] The regional joint analysis module is used to calculate the interval distance between any feature monitoring sub-region and the other feature monitoring sub-regions, and the mean of the interval distance is determined as the distribution tendency parameter.

[0023] Furthermore, the regional joint analysis module is used to determine the anomaly risk trend category of the raw paper, wherein,

[0024] Based on the determination result that the distribution tendency parameter exceeds the preset distribution tendency parameter threshold, the regional joint analysis module determines that the raw paper abnormal risk tendency category is the global abnormal risk tendency category.

[0025] Based on the determination result that the distribution tendency parameter does not exceed the preset distribution tendency parameter threshold, the abnormal risk tendency category of the raw paper is determined to be a non-global abnormal risk tendency category.

[0026] Furthermore, the regional joint analysis module is used to determine the optimal adjustment method for the hot-pressing roller, wherein,

[0027] If the abnormal risk trend category of the raw paper is the global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the roller temperature increase of the hot pressing roller according to the humidity trend parameter.

[0028] If the abnormal risk trend category of the raw paper is the non-global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the reduction of the roller gap of the hot pressing roller in the virtual monitoring area based on the morphological mapping characterization parameters of each feature monitoring sub-region.

[0029] Furthermore, the regional joint analysis module is used to determine the virtual monitoring area, wherein,

[0030] The region joint analysis module obtains the average interval distance between any two feature monitoring sub-regions in the direction perpendicular to the hot pressing roller. The width of the virtual monitoring region is the length of the original paper in the direction parallel to the hot pressing roller, and the length of the virtual monitoring region is the average interval distance.

[0031] Furthermore, the regional joint analysis module is used to determine the increase in roll body temperature of the hot-pressing roller and the decrease in roll gap in the virtual monitoring area, wherein,

[0032] The increase in roller temperature is positively correlated with the humidity trend parameter, and the decrease in roller gap is positively correlated with the mean value of the morphological mapping characterization parameter of each feature monitoring sub-region.

[0033] Furthermore, the raw paper storage optimization module is used to determine whether to optimize the storage of the raw paper, wherein,

[0034] The raw paper storage optimization module determines to optimize the storage of the raw paper based on the judgment result that the humidity trend parameter after the hot pressing roller optimization adjustment exceeds the preset optimization humidity trend threshold, and determines the reduction of the longest static storage period of the raw paper according to the optimization response parameter.

[0035] The shortening magnitude is negatively correlated with the optimized response parameter, which is the absolute value of the difference in humidity trend parameter before and after the optimization adjustment of the hot-pressing roller.

[0036] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention sets up a feature acquisition module, a feature pre-analysis module, a feature recognition module, a region joint analysis module, and a raw paper storage optimization module. The feature acquisition module acquires the humidity parameters and grayscale parameters of several monitoring sub-regions on the raw paper after drying. The feature pre-analysis module determines the humidity trend parameters based on the comparison between the humidity parameters of each monitoring sub-region to determine whether there is an abnormal risk in the raw paper. The feature recognition module performs optical response analysis on each monitoring sub-region to obtain morphological mapping characterization parameters. Based on the morphological mapping characterization parameters, the feature monitoring sub-regions are marked. The region joint analysis module performs joint response analysis on each feature monitoring sub-region to determine the abnormal risk trend category of the raw paper. Based on the abnormal risk trend category of the raw paper, the optimization adjustment method of the hot pressing roller is determined. The raw paper storage optimization module determines whether the storage of the raw paper should be optimized. Thus, the present invention achieves rapid identification of whether there is an abnormal risk in the raw paper, and adaptive adjustment of storage parameters according to the characteristics of the raw paper, thereby improving the effectiveness and reliability of intelligent management of raw paper.

[0037] In particular, this invention uses a feature pre-analysis module to determine whether there are any abnormal risks in the raw paper based on the comparison of humidity parameters between each monitoring sub-region. It is understood that the uniformity of humidity distribution after drying is a leading indicator of thickness issues. By calculating the humidity trend parameters of the entire area online in real time, an abnormal risk warning is issued in advance before the thickness unevenness fully manifests or causes irreversible effects, buying time for subsequent adjustments, changing post-event remediation to pre-event prevention, and reducing the scrap rate. At the same time, the traditional method of judging whether drying is uniform by relying on manual experience is highly subjective and has poor stability. By calculating the humidity trend parameters, the production quality control standards are unified and digitized, improving the standardization level of quality management and decision-making efficiency. It also avoids the system performing complex calculations on all data without differentiation. While ensuring detection accuracy, it optimizes the allocation of system computing resources, improves the overall response speed and operating economy, and thus achieves rapid identification of whether there are abnormal risks in the raw paper. Based on the characteristics of the raw paper, the storage parameters are adaptively adjusted, improving the effectiveness and reliability of intelligent management of raw paper.

[0038] In particular, this invention uses a feature recognition module to perform optical response analysis on each monitoring sub-region to mark the feature monitoring sub-region. It can be understood that by analyzing the scattering and reflection response of the paper surface to different incident light, the microstructural variations caused by thickness and uneven distribution can be captured. By determining the morphological mapping characterization parameters of each monitoring sub-region, the drawbacks of relying on manual visual inspection, such as strong subjectivity, low efficiency, and susceptibility to fatigue and errors, are avoided. This provides clear and reliable data support for subsequent joint analysis and precise control, thereby realizing the marking of feature monitoring sub-regions and improving the effectiveness and reliability of intelligent paper management.

[0039] In particular, this invention, through a regional joint analysis module, determines the temperature increase of the hot-pressing roller based on humidity trend parameters under the overall abnormal risk trend category. It can be understood that when the feature monitoring sub-regions are relatively widely distributed (i.e., under the overall abnormal risk trend category), it indicates that the problem is not a localized, occasional equipment failure, but more likely stems from a systemic upstream process deviation affecting the entire paper. A global roller temperature increase strategy is adopted to generate process optimization instructions that correct the deviation as a whole. A positive correlation is established between the temperature increase and the quantified humidity trend parameters. Based on intelligent risk control, slight unevenness triggers a small temperature increase adjustment. Severe unevenness triggers even greater temperature increases, effectively solving the problem while minimizing excessive energy consumption or insufficient adjustment. This achieves an optimal balance between quality control and production costs. Given different raw material batches, environmental conditions, or product specifications, the resulting systematic unevenness in process deviation is dynamically changing. Based on the specific degree of anomaly detected in each real-time step, the system dynamically provides the current optimal temperature rise compensation value, ensuring continuous stability of product quality under various production conditions. Furthermore, this allows for the determination of optimized adjustment methods for the hot-pressing rollers, improving the effectiveness and reliability of intelligent management of raw paper.

[0040] In particular, this invention, through a regional joint analysis module, determines the roll gap reduction of the hot-pressing roller in the virtual monitoring area based on the morphological mapping characterization parameters of each feature monitoring sub-region under the non-global anomaly risk trend category. It is understood that the distribution of the non-global anomaly risk trend category, i.e., the feature monitoring sub-regions, is relatively concentrated. If the optimization method of overall heating or pressure adjustment is used, not only will it fail to solve local problems, but it will also cause unnecessary thermal stress or overpressure risks to a large area of ​​normal areas, affecting overall quality and increasing energy consumption. Constructing a virtual monitoring area covering the feature monitoring sub-regions limits the control range to the local area where the problem occurs, efficiently solving local problems while protecting the overall paper quality. The performance and stability of the body, the mean value of the morphological mapping characterization parameter reflects the physical severity of local abnormal areas. By establishing a positive correlation between the roll gap reduction amplitude and the mean value of the morphological mapping characterization parameter, the optimal pressure compensation value matching the severity of defects is given. By optimizing the allocation of control resources, the dual goals of quality improvement and cost saving are achieved. The limited process adjustment energy is applied to the most needed position, avoiding the waste of energy and equipment load, which is conducive to maintaining the overall stability of the production process and provides a reliable guarantee for the continuous production of high-quality products. In addition, the optimal adjustment method of the hot pressing roller is determined, which improves the effectiveness and reliability of the intelligent management of raw paper.

[0041] In particular, this invention uses a raw paper storage optimization module to determine whether to optimize the storage of the raw paper based on the humidity trend parameters of the raw paper after optimization and adjustment on the hot-pressing rollers. It is understood that the optimization and adjustment of the hot-pressing rollers aims to correct uneven thickness of the raw paper, but process adjustments may not completely eliminate all defects. The optimized humidity trend parameters characterize the final uniformity of the internal moisture distribution of the paper after all online correction measures have been completed. Uneven moisture distribution means that there is a gradient in the bound water contained in the fibers of different areas of the paper. This gradient is the root cause of internal stress, curling, or deformation of the paper during storage due to different rates of moisture absorption or desiccation. The larger the optimized humidity trend parameters, the higher the risk of inherent deformation of the roll of raw paper, and the more necessary it is to suppress it by adjusting external storage conditions. For paper with inherent defects, quality deterioration is a cumulative process over time. The longer the resting time, the more opportunities there are for it to be affected by fluctuations in ambient temperature and humidity, and the greater the possibility of defect development and deterioration. Shortening the longest resting period can shorten the exposure time of this high-risk material in an uncontrollable storage environment. By accelerating its transfer to the next controlled production stage, such as coating and slitting, the probability of additional quality loss during storage can be minimized. The smaller the absolute value of the difference in humidity trend parameters before and after the optimization and adjustment of the hot pressing roller, that is, the weaker the effect of the front-end process adjustment on the abnormality, the more stubborn the quality problem is, and the higher its instability is expected during storage. Therefore, a larger reduction is needed to offset the risk. In this way, the storage parameters can be adaptively adjusted according to the characteristics of the raw paper, thereby improving the effectiveness and reliability of intelligent management of raw paper. Attached Figure Description

[0042] Figure 1 This is a functional block diagram of the intelligent paper management system based on multimodal data, according to an embodiment of the present invention.

[0043] Figure 2 This is a flowchart illustrating the logic of the feature pre-analysis module in this embodiment of the invention for determining whether there is any abnormal risk in the raw paper.

[0044] Figure 3 A flowchart illustrating the logic of the regional joint analysis module in this embodiment of the invention for determining the risk tendency category of raw paper anomalies;

[0045] Figure 4 This is a flowchart illustrating the logic of the paper storage optimization module in an embodiment of the present invention for determining whether to optimize the storage of paper. Detailed Implementation

[0046] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0047] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0048] It should be noted that in the description of this invention, the terms "upper," "lower," "inner," "outer," etc., which indicate the direction or positional relationship, are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0049] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation" and "connection" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0050] Please see Figure 1 The diagram shown is a functional block diagram of a multimodal data-based intelligent management system for raw paper according to an embodiment of the present invention. The present invention provides a multimodal data-based intelligent management system for raw paper, comprising:

[0051] The feature acquisition module is used to acquire the humidity parameters and grayscale parameters of several monitoring sub-regions on the dried paper.

[0052] Specifically, the embodiments of the present invention do not impose specific limitations on the structure of the feature acquisition module. Preferably, the humidity parameters of each monitoring sub-region can be obtained by an infrared moisture meter, and the grayscale parameters of each monitoring sub-region can be obtained by an industrial camera. The monitoring sub-region can be uniformly divided into grids. The humidity parameter can be the average humidity of the monitoring sub-region, and the grayscale parameter can be the average grayscale value of the monitoring sub-region. Of course, other forms can also be used, which will not be elaborated here.

[0053] The feature pre-analysis module, which is connected to the feature acquisition module, is used to determine the humidity trend parameter based on the comparison between the humidity parameters of each monitoring sub-region, so as to determine whether the raw paper has any abnormal risk.

[0054] Specifically, the embodiments of the present invention do not impose specific limitations on the structure of the feature pre-analysis module. Preferably, it can be a microprocessor to determine whether there is an abnormal risk in the raw paper. Of course, other forms can also be used, which will not be elaborated here.

[0055] A feature recognition module, which is connected to the feature acquisition module and the feature pre-analysis module respectively, is used to perform optical response analysis on each of the monitoring sub-regions to obtain morphological mapping characterization parameters, and to mark the feature monitoring sub-regions based on the morphological mapping characterization parameters.

[0056] Specifically, the embodiments of the present invention do not impose specific limitations on the structure of the feature recognition module. Preferably, it can be a microprocessor used to mark the feature monitoring sub-region. Of course, other forms can also be used, which will not be elaborated here.

[0057] A regional joint analysis module, which is connected to the feature recognition module, is used to perform joint response analysis on each feature monitoring sub-region to determine the abnormal risk trend category of the raw paper, and to determine the optimization adjustment method of the hot pressing roller based on the abnormal risk trend category of the raw paper. The optimization adjustment method includes determining the increase of the roller body temperature of the hot pressing roller, or determining the decrease of the roller gap of the hot pressing roller in the virtual monitoring area.

[0058] Specifically, the embodiments of the present invention do not specifically limit the structure of the regional joint analysis module. Preferably, it can be a processor used in a computer to determine the optimal adjustment method of the hot pressing roller. Of course, other forms can also be used, which will not be elaborated here.

[0059] The raw paper storage optimization module, which is connected to the regional joint analysis module, is used to determine whether to optimize the storage of the raw paper based on the humidity trend parameters of the raw paper after optimization and adjustment on the hot pressing roller.

[0060] Specifically, the embodiments of the present invention do not specifically limit the structure of the raw paper storage optimization module. Preferably, it can be a microprocessor to determine whether to optimize the storage of raw paper. Of course, other forms can also be used, which will not be elaborated here.

[0061] The optical response analysis involves illuminating each monitoring sub-region with different illumination angles to obtain grayscale parameters associated with each illumination angle. The virtual monitoring region is determined based on several feature monitoring sub-regions.

[0062] Specifically, the included angle between adjacent illumination angles can be set by those skilled in the art according to the accuracy requirements of the paper intelligent management system. The higher the accuracy requirement, the smaller the angle should be. The value range can be [5, 15], with the interval unit being °. Preferably, the included angle can be 8°, the initial illumination angle can be 45°, and the final illumination angle can be 135°. Five different illumination angles can be set.

[0063] Please see Figure 2The diagram shown is a flowchart illustrating the logic of the feature pre-analysis module in this embodiment of the invention for determining whether the raw paper has any abnormal risks. The feature pre-analysis module is used to determine whether the raw paper has any abnormal risks.

[0064] The feature pre-analysis module determines that the raw paper has an abnormal risk based on the comparison of humidity parameters between the monitoring sub-regions and the determination result that the comparison meets the risk tendency conditions.

[0065] Based on the judgment result that the comparison between the humidity parameters of each monitoring sub-region does not meet the risk tendency conditions, it is determined that the raw paper does not have any abnormal risk.

[0066] The risk tendency condition is that the humidity tendency parameter exceeds a preset humidity tendency parameter threshold, and the humidity tendency parameter is the variance of the humidity parameter of each monitoring sub-region.

[0067] Specifically, the preset humidity trend parameter threshold is the product of the humidity trend parameter reference value and the trend factor. The humidity trend parameter reference value is the average value of the humidity trend parameter under the same working conditions in historical data. The trend factor can be set by those skilled in the art according to the accuracy requirements of the paper intelligent management system. The higher the accuracy requirement, the smaller the value should be. The value range can be [1.15, 1.25], preferably 1.2.

[0068] Specifically, this invention uses a feature pre-analysis module to determine whether there are any abnormal risks in the raw paper based on the comparison of humidity parameters between each monitoring sub-region. It is understood that the uniformity of humidity distribution after drying is a leading indicator of thickness issues. By calculating the humidity trend parameters across the entire surface online in real time, an early warning of abnormal risks is issued before thickness unevenness fully manifests or causes irreversible effects, allowing time for subsequent adjustments and shifting from post-event remediation to pre-event prevention, thus reducing the scrap rate. Furthermore, the traditional method of judging drying uniformity based on manual experience is highly subjective and unstable. Calculating humidity trend parameters unifies and digitizes production quality control standards, improving the standardization level and decision-making efficiency of quality management. It also avoids the system performing complex, undifferentiated calculations on all data. While ensuring detection accuracy, it optimizes the allocation of system computing resources, improving overall response speed and operational economy. Therefore, it enables rapid identification of any abnormal risks in the raw paper and adaptive adjustment of storage parameters based on the characteristics of the raw paper, improving the effectiveness and reliability of intelligent management of the raw paper.

[0069] Specifically, it can be understood that during the drying process of raw paper, heat energy is transferred from the drying cylinder to the paper web, and moisture evaporates due to heat. For paper webs with uneven thickness distribution, there are differences in local thermal resistance and moisture resistance. Thicker areas have longer heat conduction paths and higher moisture content, so under the same drying conditions, their moisture evaporation rate will be slower than that of thinner areas, resulting in higher residual humidity in these areas after drying. After drying, the uneven distribution of humidity at various points in the transverse direction of the paper web essentially reflects the unevenness of thickness or basis weight formed in the previous process. The humidity tendency parameter is a mathematical characterization of the degree of dispersion of this transverse distribution. The larger the humidity tendency parameter, the worse the drying uniformity of the paper web, which in turn indicates a higher risk of underlying quality defects such as uneven thickness. Thus, it is possible to quickly identify whether there are abnormal risks in the raw paper, and to adaptively adjust storage parameters according to the characteristics of the raw paper, thereby improving the effectiveness and reliability of intelligent management of raw paper.

[0070] Specifically, the feature recognition module is used to perform optical response analysis on each of the monitored sub-regions to obtain morphological mapping characterization parameters, wherein,

[0071] The feature recognition module obtains the grayscale parameters of the monitored sub-region under different illumination angles, and determines the variance of the grayscale parameters as the morphological mapping representation parameter of the monitored sub-region.

[0072] Specifically, the feature recognition module is used to label feature monitoring sub-regions based on the morphological mapping representation parameters, wherein,

[0073] The feature recognition module marks the monitoring sub-region as a feature monitoring sub-region based on the determination result that the morphological mapping representation parameter of the monitoring sub-region exceeds the preset morphological mapping representation parameter threshold.

[0074] Based on the determination result that the morphological mapping representation parameters of the monitored sub-region do not exceed the preset morphological mapping representation parameter threshold, the monitored sub-region is not marked.

[0075] Specifically, the preset threshold for the morphology mapping representation parameter is the product of the reference value of the morphology mapping representation parameter and the morphology factor. The reference value of the morphology mapping representation parameter is the average value of the morphology mapping representation parameter under the same working conditions in historical data. The morphology factor can be set by those skilled in the art according to the accuracy requirements of the paper intelligent management system. The higher the accuracy requirement, the smaller the value should be. The value range can be [1.05, 1.2], preferably 1.1.

[0076] Specifically, in this embodiment of the invention, the optical response analysis of each monitoring sub-region is performed by a feature recognition module to mark the feature monitoring sub-region. It can be understood that by analyzing the scattering and reflection response of the paper surface to different incident light, the microstructural variations caused by thickness and uneven distribution can be captured. By determining the morphological mapping characterization parameters of each monitoring sub-region, the drawbacks of relying on manual visual inspection, such as strong subjectivity, low efficiency, and easy fatigue and error, are avoided. This provides clear and reliable data support for subsequent joint analysis and precise control, thereby realizing the marking of feature monitoring sub-regions and improving the effectiveness and reliability of intelligent paper management.

[0077] Specifically, it can be understood that as a multilayer porous medium composed of a fiber network, the optical properties of paper exhibit significant angle dependence. When a beam of light illuminates the paper surface at different angles, various optical phenomena such as specular reflection, volume scattering, and transmission occur. For regions with uniform structure, the surface smoothness and internal fiber arrangement are relatively consistent. Therefore, under different angles of illumination, the intensity of reflected or scattered light is reflected in a gradual change in grayscale value. For regions with abnormal thickness, their microstructure changes. For example, excessively thick regions are usually more porous and have stronger internal light scattering effects, and their surface micro-geometry is more irregular. This structural difference leads to a significantly different pattern of optical response variation with illumination angle compared to normal regions. By calculating the morphological mapping characterization parameters of the same sub-region at multiple different incident angles, the larger the morphological mapping characterization parameter, the stronger the optical response depends on the illumination angle, and the more significant the heterogeneity. This enables the marking of feature monitoring sub-regions, improving the effectiveness and reliability of intelligent paper management.

[0078] Specifically, the regional joint analysis module is used to perform joint response analysis on each of the feature monitoring sub-regions, wherein,

[0079] The regional joint analysis module is used to calculate the interval distance between any feature monitoring sub-region and the other feature monitoring sub-regions, and the mean of the interval distance is determined as the distribution tendency parameter.

[0080] Please see Figure 3 The diagram shown is a logical flowchart of the regional joint analysis module for determining the anomaly risk trend category of raw paper in an embodiment of the present invention. The regional joint analysis module is used to determine the anomaly risk trend category of raw paper.

[0081] Based on the determination result that the distribution tendency parameter exceeds the preset distribution tendency parameter threshold, the regional joint analysis module determines that the raw paper abnormal risk tendency category is the global abnormal risk tendency category.

[0082] Based on the determination result that the distribution tendency parameter does not exceed the preset distribution tendency parameter threshold, the abnormal risk tendency category of the raw paper is determined to be a non-global abnormal risk tendency category.

[0083] Specifically, the preset distribution tendency parameter threshold is the product of the distribution tendency parameter reference value and the distribution factor. The distribution tendency parameter reference value is the mean of the distribution tendency parameter under the same working conditions in historical data. The distribution factor can be set by those skilled in the art according to the accuracy requirements of the paper intelligent management system. The higher the accuracy requirement, the larger the value should be. The value range can be [1.1, 1.2], preferably 1.15.

[0084] Specifically, the regional joint analysis module is used to determine the optimal adjustment method for the hot-pressing roller, wherein,

[0085] If the abnormal risk trend category of the raw paper is the global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the roller temperature increase of the hot pressing roller according to the humidity trend parameter.

[0086] If the abnormal risk trend category of the raw paper is the non-global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the reduction of the roller gap of the hot pressing roller in the virtual monitoring area based on the morphological mapping characterization parameters of each feature monitoring sub-region.

[0087] Specifically, in this embodiment of the invention, the regional joint analysis module determines the temperature increase of the hot-pressing roller based on humidity trend parameters under the global anomaly risk trend category. It can be understood that when the feature monitoring sub-regions are relatively widely distributed (i.e., the global anomaly risk trend category), it indicates that the problem is not a localized, occasional equipment failure, but more likely stems from a systemic upstream process deviation affecting the entire paper. A global roller temperature increase strategy is adopted to generate process optimization instructions that correct the deviation as a whole. A positive correlation is established between the temperature increase and the quantified humidity trend parameters. Based on intelligent risk control, slight unevenness triggers small temperature increases. The temperature rise adjustment is adjusted accordingly, and severe unevenness triggers a larger temperature rise. This effectively solves the problem while minimizing excessive energy consumption or insufficient adjustment, achieving an optimal balance between quality control and production costs. Faced with different raw material batches, environmental conditions, or product specifications, the process deviation that causes systematic unevenness changes dynamically. Based on the specific degree of anomaly detected in real time each time, the current optimal temperature rise compensation value is dynamically given, ensuring the continuous stability of product quality under various production conditions. In turn, the optimized adjustment method of the hot pressing roller is determined, improving the effectiveness and reliability of intelligent management of raw paper.

[0088] Specifically, it can be understood that the overall abnormal risk trend category, i.e., thickness anomalies, has a relatively wide distribution range on the base paper. This usually indicates a general, trend-based deviation in upstream processes such as headbox sizing or hot air drying. Moisture trend parameters characterize the gradient of internal moisture distribution in the paper web before it enters the calender. Moisture acts as a plasticizer for paper fibers; areas with higher moisture content have softer, more malleable fibers, while areas with lower moisture content have stiffer fibers. Hot calendering uses the combined effects of temperature and pressure to induce controlled plastic deformation in the paper to adjust its thickness. When dealing with base paper with uneven moisture content, directly applying pressure will cause differential compression in areas with different plasticity, which is counterproductive. This may exacerbate uneven thickness. Increasing the temperature can provide additional heat energy to the entire paper, promoting the activation of bound water in the fibers and the softening of cellulose, thereby improving the overall plasticity of the paper fiber network. This reduces the deformation resistance of fibers in even drier areas, allowing subsequent pressure to act more evenly across the entire paper and compensating for the plasticity differences caused by initial uneven moisture content. The higher the humidity trend parameter, the greater the difference in plasticity within the paper web. The higher the overall thermoplasticity improvement required to equalize this difference, the greater the roll temperature rise is needed. This leads to the determination of the optimized adjustment method for the hot pressing roll, improving the effectiveness and reliability of intelligent management of the raw paper.

[0089] Specifically, the regional joint analysis module is used to determine the virtual monitoring area, wherein,

[0090] The region joint analysis module obtains the average interval distance between any two feature monitoring sub-regions in the direction perpendicular to the hot pressing roller. The width of the virtual monitoring region is the length of the original paper in the direction parallel to the hot pressing roller, and the length of the virtual monitoring region is the average interval distance.

[0091] Specifically, the vertical distance from the starting position of the virtual monitoring area along the length of the paper to the hot-pressing roller is the minimum vertical distance between each feature monitoring sub-area and the hot-pressing roller.

[0092] Specifically, the regional joint analysis module is used to determine the increase in roll body temperature of the hot-pressing roller and the decrease in roll gap in the virtual monitoring area, wherein,

[0093] The increase in roller temperature is positively correlated with the humidity trend parameter, and the decrease in roller gap is positively correlated with the mean value of the morphological mapping characterization parameter of each feature monitoring sub-region.

[0094] Specifically, the increase in roll temperature is calculated as the humidity trend parameter / reference value of humidity trend parameter × temperature factor, and the decrease in roll gap is calculated as the mean value of morphology mapping characterization parameter / reference value of mean value of morphology mapping characterization parameter × roll gap factor. The reference value of the mean value of morphology mapping characterization parameter is the average value of the mean values ​​of morphology mapping characterization parameter under the same working conditions in historical data. The temperature factor and roll gap factor can be calculated by those skilled in the art based on the average value of several historical experimental data. The temperature factor can be in the range of [0.1, 0.3] to avoid the roll temperature increase being too large or too small. Preferably, it can be 0.2. The roll gap factor can be in the range of [0.1, 0.2] to avoid the roll gap decrease being too large or too small. Preferably, it can be 0.15.

[0095] Specifically, in this embodiment of the invention, the regional joint analysis module determines the roll gap reduction of the hot-pressing roller in the virtual monitoring area based on the morphological mapping characterization parameters of each feature monitoring sub-region under the non-global anomaly risk trend category. It is understood that the distribution of the non-global anomaly risk trend category, i.e., the feature monitoring sub-regions, is relatively concentrated. If the optimization method of overall heating or pressure adjustment is used, it will not only fail to solve local problems but will also cause unnecessary thermal stress or overpressure risks to a large area of ​​normal areas, affecting overall quality and increasing energy consumption. Constructing a virtual monitoring area covering the feature monitoring sub-regions limits the control range to the local area where the problem occurs, efficiently solving local problems while protecting the paper. The overall performance and stability of the sheet are assessed, and the mean value of the morphological mapping characterization parameter reflects the physical severity of local abnormal areas. By establishing a positive correlation between the roll gap reduction and the mean value of the morphological mapping characterization parameter, an optimal pressure compensation value matching the severity of defects is provided. By optimizing the allocation of control resources, the dual goals of quality improvement and cost saving are achieved. This ensures that limited process adjustment energy is applied where it is most needed, avoiding waste of energy and equipment load, and helps maintain the overall stability of the production process. This provides a reliable guarantee for the continuous production of high-quality products. Furthermore, the optimal adjustment method of the hot pressing roller is determined, improving the effectiveness and reliability of intelligent management of the raw paper.

[0096] Specifically, it can be understood that the non-global anomaly risk tendency category feature monitoring sub-regions are relatively clustered in space. To correct this local thickening, greater pressure can be applied at the corresponding positions to achieve a greater compression rate than the surrounding normal areas, thereby catching up with the overall thickness and reducing the roll gap in the virtual monitoring area. Even if the hot-pressing roller applies enhanced pressure to the horizontal position covering the virtual monitoring area, the virtual monitoring area is defined by the full width of the paper web and the length by the average longitudinal span of the feature monitoring sub-region. The average value of the morphological mapping characterization parameter quantifies the average severity of the anomaly set within the virtual area. According to elasticity and material compression characteristics, for thicker or looser local areas, greater compression force is required to achieve the same compaction as the surrounding areas. The larger the average value of the morphological mapping characterization parameter, the more severe the average severity of the anomaly set within the virtual area, and the smaller the roll gap is required to bring greater compression force. Thus, the optimized adjustment method of the hot-pressing roller is determined, improving the effectiveness and reliability of intelligent management of the raw paper.

[0097] Please see Figure 4 The diagram shown is a flowchart illustrating the logic of the paper storage optimization module in an embodiment of the present invention for determining whether to optimize the storage of the paper. The paper storage optimization module is used to determine whether to optimize the storage of the paper.

[0098] The raw paper storage optimization module determines to optimize the storage of the raw paper based on the judgment result that the humidity trend parameter after the hot pressing roller optimization adjustment exceeds the preset optimization humidity trend threshold, and determines the reduction of the longest static storage period of the raw paper according to the optimization response parameter.

[0099] Based on the determination that the humidity trend parameters after the optimization and adjustment of the hot-pressing roller do not exceed the preset optimized humidity trend threshold, it is determined that the storage of the raw paper will not be optimized.

[0100] The shortening magnitude is negatively correlated with the optimized response parameter, which is the absolute value of the difference in humidity trend parameter before and after the optimization adjustment of the hot-pressing roller.

[0101] Specifically, the preset optimized humidity trend threshold is the product of the humidity trend parameter reference value and the optimization factor. The optimization factor can be set by those skilled in the art according to the accuracy requirements of the paper intelligent management system. The higher the accuracy requirement, the smaller the value should be. The value range can be [1.05, 1.15], and preferably, it can be 1.1.

[0102] Specifically, the reduction margin is calculated as the optimized response parameter reference value / optimized response parameter × duration factor. The optimized response parameter reference value is the average optimized response parameter under the same working conditions in historical data. The duration factor can be calculated by those skilled in the art based on several historical experimental data, and its value range can be [0.1, 0.3] to avoid the reduction margin being too large or too small. Preferably, it can be 0.2.

[0103] Specifically, in this embodiment of the invention, the raw paper storage optimization module determines whether to optimize the storage of the raw paper based on the humidity trend parameters of the raw paper after optimization and adjustment on the hot-pressing rollers. It is understood that the optimization and adjustment of the hot-pressing rollers aims to correct uneven thickness of the raw paper, but process adjustments may not completely eliminate all defects. The optimized humidity trend parameters characterize the final uniformity of the internal moisture distribution of the paper after all online correction measures have been completed. Uneven moisture distribution means that there is a gradient in the bound water contained in the fibers of different areas of the paper. This gradient is the root cause of internal stress, curling, or deformation of the paper during storage due to different rates of moisture absorption or desiccation. The larger the optimized humidity trend parameters, the higher the risk of inherent deformation of the roll of raw paper, and the more necessary it is to adjust external storage conditions to suppress it. For paper with inherent defects, its quality deterioration is a cumulative process over time. The longer the resting time, the more opportunities it has to be affected by fluctuations in ambient temperature and humidity, and the greater the possibility of defect development and deterioration. Shortening the longest resting period can shorten the exposure time of this high-risk material in an uncontrollable storage environment. By accelerating its flow to the next controlled production stage, such as coating and slitting, the probability of additional quality loss during storage can be minimized. The smaller the absolute value of the difference in humidity trend parameters before and after the optimization and adjustment of the hot pressing roller, that is, the weaker the effect of front-end process adjustment on the abnormality, the more stubborn the quality problem is, and the higher its instability is expected during storage. Therefore, a larger reduction is needed to offset the risk. Thus, storage parameters can be adaptively adjusted according to the characteristics of the raw paper, improving the effectiveness and reliability of intelligent management of raw paper.

[0104] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0105] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A paper intelligent management system based on multimodal data, characterized in that, include: The feature acquisition module is used to acquire the humidity parameters and grayscale parameters of several monitoring sub-regions on the dried paper. The feature pre-analysis module, which is connected to the feature acquisition module, is used to determine the humidity trend parameter based on the comparison between the humidity parameters of each monitoring sub-region, so as to determine whether the raw paper has any abnormal risk. A feature recognition module, which is connected to the feature acquisition module and the feature pre-analysis module respectively, is used to perform optical response analysis on each of the monitoring sub-regions to obtain morphological mapping characterization parameters, and to mark the feature monitoring sub-regions based on the morphological mapping characterization parameters. A regional joint analysis module, which is connected to the feature recognition module, is used to perform joint response analysis on each feature monitoring sub-region to determine the abnormal risk trend category of the raw paper, and to determine the optimization adjustment method of the hot pressing roller based on the abnormal risk trend category of the raw paper. The optimization adjustment method includes determining the increase of the roller body temperature of the hot pressing roller, or determining the decrease of the roller gap of the hot pressing roller in the virtual monitoring area. The raw paper storage optimization module, which is connected to the regional joint analysis module, is used to determine whether to optimize the storage of the raw paper based on the humidity trend parameters of the raw paper after optimization and adjustment on the hot pressing roller. The optical response analysis involves illuminating each monitoring sub-region with different illumination angles to obtain grayscale parameters associated with each illumination angle. The virtual monitoring region is determined based on several feature monitoring sub-regions.

2. The intelligent management system for raw paper based on multimodal data according to claim 1, characterized in that, The feature pre-analysis module is used to determine whether the base paper has any abnormal risks, wherein, The feature pre-analysis module determines that the raw paper has an abnormal risk based on the comparison of humidity parameters between the monitoring sub-regions and the determination result that the comparison meets the risk tendency conditions. The risk tendency condition is that the humidity tendency parameter exceeds a preset humidity tendency parameter threshold, and the humidity tendency parameter is the variance of the humidity parameter of each monitoring sub-region.

3. The intelligent management system for raw paper based on multimodal data according to claim 2, characterized in that, The feature recognition module is used to perform optical response analysis on each of the monitored sub-regions to obtain morphological mapping characterization parameters, wherein... The feature recognition module obtains the grayscale parameters of the monitored sub-region under different illumination angles, and determines the variance of the grayscale parameters as the morphological mapping representation parameter of the monitored sub-region.

4. The intelligent management system for raw paper based on multimodal data according to claim 3, characterized in that, The feature recognition module is used to mark feature monitoring sub-regions based on the morphological mapping representation parameters. in, The feature recognition module marks the monitored sub-region as a feature monitoring sub-region based on the determination result that the morphological mapping representation parameter of the monitored sub-region exceeds the preset morphological mapping representation parameter threshold.

5. The intelligent management system for raw paper based on multimodal data according to claim 4, characterized in that, The regional joint analysis module is used to perform joint response analysis on each of the feature monitoring sub-regions, wherein... The regional joint analysis module is used to calculate the interval distance between any feature monitoring sub-region and the other feature monitoring sub-regions, and the mean of the interval distance is determined as the distribution tendency parameter.

6. The intelligent management system for raw paper based on multimodal data according to claim 5, characterized in that, The regional joint analysis module is used to determine the anomaly risk trend category of the raw paper, wherein, Based on the determination result that the distribution tendency parameter exceeds the preset distribution tendency parameter threshold, the regional joint analysis module determines that the raw paper abnormal risk tendency category is the global abnormal risk tendency category. Based on the determination result that the distribution tendency parameter does not exceed the preset distribution tendency parameter threshold, the abnormal risk tendency category of the raw paper is determined to be a non-global abnormal risk tendency category.

7. The intelligent management system for raw paper based on multimodal data according to claim 6, characterized in that, The regional joint analysis module is used to determine the optimal adjustment method for the hot-pressing roller, wherein, If the abnormal risk trend category of the raw paper is the global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the roller temperature increase of the hot pressing roller according to the humidity trend parameter. If the abnormal risk trend category of the raw paper is the non-global abnormal risk trend category, the regional joint analysis module determines that the optimization adjustment method of the hot pressing roller is to determine the reduction of the roller gap of the hot pressing roller in the virtual monitoring area based on the morphological mapping characterization parameters of each feature monitoring sub-region.

8. The intelligent management system for raw paper based on multimodal data according to claim 7, characterized in that, The regional joint analysis module is used to determine the virtual monitoring area, wherein... The region joint analysis module obtains the average interval distance between any two feature monitoring sub-regions in the direction perpendicular to the hot pressing roller. The width of the virtual monitoring region is the length of the original paper in the direction parallel to the hot pressing roller, and the length of the virtual monitoring region is the average interval distance.

9. The intelligent management system for raw paper based on multimodal data according to claim 8, characterized in that, The regional joint analysis module is used to determine the increase in roll body temperature of the hot-pressing roller and the decrease in roll gap in the virtual monitoring area, wherein... The increase in roller temperature is positively correlated with the humidity trend parameter, and the decrease in roller gap is positively correlated with the mean value of the morphological mapping characterization parameter of each feature monitoring sub-region.

10. The intelligent management system for raw paper based on multimodal data according to claim 9, characterized in that, The raw paper storage optimization module is used to determine whether to optimize the storage of the raw paper, wherein... The raw paper storage optimization module determines to optimize the storage of the raw paper based on the judgment result that the humidity trend parameter after the hot pressing roller optimization adjustment exceeds the preset optimization humidity trend threshold, and determines the reduction of the longest static storage period of the raw paper according to the optimization response parameter. The shortening magnitude is negatively correlated with the optimized response parameter, which is the absolute value of the difference in humidity trend parameter before and after the optimization adjustment of the hot-pressing roller.

Images