Intelligent control system for ground pulping station

Abstract

The present invention belongs to the field of automatic control technology, and specifically relates to an intelligent control system for a ground pulping station. The system includes: a pulping sample feature analysis and evolution unit, a physical property analysis unit, and a feedback intelligent control unit; the pulping sample feature analysis and evolution unit is used to obtain the length, diameter, and orientation angle of each sample fiber from the pulping sample as sample fiber features; based on the fiber classification function, a fiber state evolution function is constructed; the physical property analysis unit is used to construct a rheological property model of the pulping sample according to the fiber state evolution function, and to construct a performance evolution model of the pulping sample as a global multi-objective optimization function of the pulping sample; the feedback intelligent control unit is used to solve the global multi-objective optimization function to obtain the temperature field, pressure field, velocity field, and concentration field corresponding to the maximum value of the global multi-objective optimization function. The present invention improves the efficiency, uniformity, and stability of the pulping process.

Description

Technical Field

[0001] The invention belongs to the technical field of automatic control, and in particular relates to an intelligent control system for a ground pulping station. Background Art

[0002] In the field of coal mining, pulping technology is mainly used in coal mine grouting reinforcement, water disaster prevention and control, and mining waste recovery. Its core task is to achieve reinforcement and penetration protection of coal body cracks or tunnel walls by preparing a specific slurry system. With the continuous expansion of coal mining depth and scale, higher requirements are placed on the efficiency of the pulping process, the stability of slurry performance, and the uniformity of grouting effects. However, the existing pulping technology still has many technical bottlenecks and difficulties in the slurry preparation process, especially how to achieve dynamic monitoring of the slurry fiber state, precise regulation of the slurry physical properties, and global optimization of process parameters. These problems seriously restrict the further improvement of the pulping process.

[0003] At present, coal mine pulping technology mainly relies on traditional slurry ratio experience and simple mechanical stirring process. These technical means can meet basic needs in specific scenarios, but with the increasing complexity of coal mining, their limitations are becoming increasingly significant. In the public related technologies, some studies have attempted to introduce fiber-reinforced slurry systems to improve the rheological properties and structural stability of slurries. However, the preparation process of fiber-reinforced slurries is still in the empirical control stage, lacking systematic research on fiber properties, slurry physical parameters, and multi-physical field dynamic coupling. In the prior art, the monitoring and analysis of slurry fiber properties mainly rely on offline measurement methods, such as sampling and analyzing fiber length, diameter, and orientation angle by optical microscopy or scanning electron microscopy (SEM). Although these methods can provide certain fiber geometric property information, they cannot realize real-time dynamic monitoring of fiber state during pulping, and cannot capture the evolution of fiber properties during pulping. In addition, these methods generally have the problems of low efficiency and measurement accuracy being greatly affected by the environment, making it difficult to adapt to the needs of industrial production. In the study of rheological properties of pulping samples, traditional technologies focus more on the macroscopic viscosity and fluidity parameters of the slurry, but lack in-depth exploration of the relationship between the microscopic behavior of the slurry and the fiber state and fluid mechanics properties. The regulation of slurry viscosity is usually achieved by adjusting the water-cement ratio or the concentration of chemical additives, but this method cannot accurately reflect the actual contribution of fiber distribution, orientation state and particle interaction in the slurry to the rheological properties. In existing studies, the slurry viscosity was only roughly measured under simple experimental conditions, ignoring the influence of the complex dynamic environment in the pulping process on the rheological properties. In addition, in the study of slurry flocculation behavior, existing technologies mainly optimize slurry performance by adjusting the aggregation of particles through chemical additives. However, the physical essence of flocculation behavior is the interaction between fiber particles in the slurry in the pressure field and fluid dynamic environment. Traditional technologies have failed to establish a systematic mathematical model to describe the dynamic characteristics of the flocculation process, making it difficult to accurately predict the dynamic changes of particle aggregation and its impact on slurry performance. Summary of the invention

[0004] The main purpose of the present invention is to provide an intelligent control system for a ground pulping station, which solves the problems of unstable performance, high energy consumption, lack of dynamic regulation, etc. in the traditional pulping process. The invention significantly improves the efficiency, uniformity and stability of the pulping process, realizes the intelligence and automation of the pulping process, and provides an efficient and intelligent technical solution for the coal mining field and other complex pulping scenarios.

[0005] In order to solve the above technical problems, the present invention provides an intelligent control system for a ground pulping station, the system comprising: a pulping sample feature analysis and evolution unit, a physical property analysis unit and a feedback intelligent control unit; the pulping sample feature analysis and evolution unit is used to obtain the length, diameter and orientation angle of each sample fiber from the pulping sample as the sample fiber feature; each sample fiber feature is regarded as a point in a three-dimensional feature space, so as to map each sample fiber to the three-dimensional feature space to obtain a fiber classification function; based on the fiber classification function, the pressure field of the pulping sample and the dynamic viscosity of the fluid during the pulping process are obtained, and a fiber state evolution function is constructed to describe the dynamic change of the state of the sample fiber during the pulping process; the The physical property analysis unit is used to construct a rheological property model of the pulping sample according to the fiber state evolution function to describe the stress change of the pulping sample during the pulping process; and to establish the flocculation kinetic equation of the pulping sample according to the rheological property model; on the basis of the flocculation kinetic equation, the rheological property model and the fiber state evolution function, a performance evolution model of the pulping sample is constructed as a global multi-objective optimization function of the pulping sample; the feedback intelligent control unit is used to solve the global multi-objective optimization function, obtain the temperature field, pressure field, velocity field and concentration field corresponding to the maximum value of the global multi-objective optimization function, as the operation control target of the ground pulping station, and control the operation of the ground pulping station according to the operation control target.

[0006] Furthermore, the pulping sample characteristic analysis and evolution unit uses an optical microscope or a scanning electron microscope to obtain a geometric image of the sample fibers in the pulping sample, and uses automatic image processing technology to extract the length, diameter and orientation angle of each sample fiber from the geometric image as the sample fiber characteristics; the resolution of the optical microscope or the scanning electron microscope is at least 1 micron / pixel.

[0007] Furthermore, the process of extracting the length, diameter and orientation angle of each sample fiber from the geometric image using automatic image processing technology specifically includes: converting the collected geometric image into a grayscale image; applying a filter to smooth the grayscale image and remove noise; using an edge detection algorithm to extract the boundaries of the sample fibers from the grayscale image and output an edge image; using a global threshold segmentation method to segment the edge image and using a connectivity analysis algorithm to mark the pixel area of ​​each sample fiber; in the pixel area of ​​each sample fiber, using a minimum circumscribed rectangle to calculate the length of the sample fiber; using the ratio of area to length to calculate the diameter; and obtaining the orientation angle of the sample fiber through a principal component analysis method.

[0008] Furthermore, the dimension classification function Use the following formula to express it:

[0009] ;

[0010] in, is the length variable of the fiber classification function; is an integer subscript index; is the number of sample fibers extracted from the geometric image; is the diameter variable of the fiber classification function; is the orientation angle variable of the fiber classification function; is the maximum length of the sample fiber; is the maximum diameter of the sample fiber; is the maximum orientation angle of the sample fiber; is the standard deviation of the length of the sample fibers; is the standard deviation of the diameter of the sample fibers; is the standard deviation of the orientation angle of the sample fibers; For the The length of the sample fiber; For the The diameter of the sample fiber; For the Orientation angle of the sample fibers; is the length integral differential; is the integral differential of the orientation angle; is the diameter integral differential; is the distance from the cluster center to the origin of the points corresponding to all sample fiber features in the feature space.

[0011] Furthermore, the fiber state evolution function Use the following formula to express it:

[0012] ;

[0013] in, For time; , is the diffusion coefficient matrix, is the preset diffusion coefficient value; is the Z-axis coordinate; is the X-axis coordinate; is the Y-axis coordinate; For time The pressure field at that time; is the dynamic viscosity of the fluid; ;in, is the mean of the orientation angles of all sample fibers; is the Laplace operator.

[0014] Furthermore, the physical property analysis unit constructs a rheological property model of the pulping sample according to the fiber state evolution function through the following formula:

[0015] ;

[0016] in, For time stress at the time; is the beating degree of the pulping sample; is the surface roughness of the pulped sample; For time The temperature field; is the time integral variable, ranging from 0 to ; For time The concentration field at .

[0017] Furthermore, the physical property analysis unit establishes a flocculation kinetic equation of the pulping sample according to the rheological property model through the following formula:

[0018] ;

[0019] in, For pulping samples at time The velocity field at time ; For pulping samples at time The number density of flocculated particles at is the flocculation diffusion coefficient, which is the set value; is the gradient operator.

[0020] Furthermore, the physical property analysis unit constructs a performance evolution model of the pulping sample as a global multi-objective optimization function based on the flocculation kinetic equation, the rheological property model and the fiber state evolution function through the following formula:

[0021] ;

[0022] in, Function value for a global multi-objective optimization function.

[0023] Furthermore, the constraints of the global multi-objective optimization function of the pulping sample are:

[0024] ;

[0025] .

[0026] The intelligent control system of the ground pulping station of the present invention has the following beneficial effects:

[0027] First, the present invention realizes the accurate extraction of sample fiber characteristics and real-time monitoring of dynamic evolution through the pulping sample feature analysis and evolution unit. Traditional pulping technology relies on fixed process ratios and empirical regulation, and it is difficult to reflect the subtle changes in the state of the fiber during the pulping process in real time. The present invention uses high-resolution microscopic imaging technology and automatic image processing technology to accurately extract key features such as fiber length, diameter and orientation angle from pulping samples, and comprehensively describes the physical properties and distribution patterns of the fibers through three-dimensional feature space mapping and classification function modeling. This real-time and refined feature extraction capability enables the system to accurately capture the dynamic changes in the fiber state during the pulping process, and provide high-quality data support for subsequent state evolution and optimization control.

[0028] Secondly, the present invention effectively realizes the accurate modeling and description of the rheological properties and flocculation dynamics of pulping samples through the construction of a physical property analysis unit. It is difficult for traditional pulping processes to fully analyze the complex rheological behavior of the slurry during the pulping process and the dynamic characteristics of the fiber flocculation process, which leads to lag and uncertainty in process control. By introducing the fiber state evolution function, the present invention systematically describes the dynamic evolution law of the fiber under the action of pressure field, fluid viscosity and temperature field; at the same time, through the rheological property model, a nonlinear relationship between stress and fiber state, concentration field and flocculation behavior is established, providing a scientific basis for optimizing slurry performance. In addition, through the flocculation kinetic equation, the evolution law of fiber flocculation particles in time and space is accurately revealed, so that the system can better control the distribution uniformity of the fiber and the fluidity of the slurry.

[0029] Third, the present invention further realizes the global optimization and dynamic control of pulping process parameters through the construction of a performance evolution model. The traditional pulping process is relatively limited in terms of multi-objective optimization, usually focusing only on a single performance indicator, and unable to find the optimal balance between process efficiency, slurry performance and energy consumption cost. The present invention organically combines the classification characteristics, dynamic state, flocculation behavior and rheological properties of the fiber by establishing a global multi-objective optimization function, and comprehensively describes the performance evolution law in the pulping process in the form of mathematical integrals. The multi-objective design of the optimization function enables the system to comprehensively weigh the distribution of the concentration field, the density of the flocculated particles and the stability of the stress, and achieve global optimization of the slurry performance under dynamically changing conditions, thereby greatly improving the overall efficiency and reliability of the pulping process.

[0030] Fourth, in addition, with the support of the feedback intelligent control unit, the present invention further deeply integrates mathematical modeling with actual process control to form an adaptive, closed-loop intelligent control framework. The control methods of traditional pulping processes often rely on manual judgment and experience-based decision-making, which is not only inefficient, but also prone to deviations due to changes in process conditions. The present invention uses the dynamic feedback mechanism of the intelligent control unit to convert the real-time monitored fiber state, physical properties and optimization results into specific process control parameters, such as adjusting the pressure field, temperature field and fluid viscosity, thereby realizing the automated control and rapid response of the pulping process. This closed-loop control capability significantly improves the adaptability and stability of the process, enabling the system to maintain efficient operation under complex and changeable production conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of the system structure of a power grid primary frequency regulation intelligent control system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The method of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments of the present invention.

[0033] Example 1, reference Figure 1 : An intelligent control system for a ground pulping station, the system comprising: a pulping sample feature analysis and evolution unit, a physical property analysis unit and a feedback intelligent control unit; the pulping sample feature analysis and evolution unit is used to obtain the length, diameter and orientation angle of each sample fiber from the pulping sample as the sample fiber feature; each sample fiber feature is regarded as a point in a three-dimensional feature space, so as to map each sample fiber to the three-dimensional feature space to obtain a fiber classification function; based on the fiber classification function, the pressure field of the pulping sample and the dynamic viscosity of the fluid during the pulping process are obtained, and a fiber state evolution function is constructed to describe the dynamic change of the state of the sample fiber during the pulping process; the physical property analysis unit The element is used to construct a rheological property model of the pulping sample according to the fiber state evolution function to describe the stress change of the pulping sample during the pulping process; and to establish the flocculation kinetic equation of the pulping sample according to the rheological property model; on the basis of the flocculation kinetic equation, the rheological property model and the fiber state evolution function, a performance evolution model of the pulping sample is constructed as a global multi-objective optimization function of the pulping sample; the feedback intelligent control unit is used to solve the global multi-objective optimization function, obtain the temperature field, pressure field, velocity field and concentration field corresponding to the maximum value of the global multi-objective optimization function, and use them as the operation control target of the ground pulping station, and control the operation of the ground pulping station according to the operation control target.

[0034] Specifically, in practice, the implementation of the intelligent control system of the ground pulping station first depends on the real-time sampling and feature extraction of pulping samples. The pulping samples contain a large number of fibers, and the physical properties of these fibers have a decisive influence on the pulping quality. In the pulping sample feature analysis and evolution unit, the length, diameter and orientation angle of the fibers are obtained from the samples through high-precision measurement equipment and calculation methods. Fiber length refers to the physical size of the fiber in the main axis direction, which determines the connection ability of the fiber in the network structure; the diameter reflects the thickness of the fiber, which is closely related to the mechanical properties and fluid dynamics of the fiber; the orientation angle describes the spatial direction distribution of the fiber in the flow, reflecting the randomness and orderliness of the fiber arrangement. These characteristic parameters are regarded as points in three-dimensional space, so that each fiber can be mapped to the three-dimensional feature space. By analyzing the distribution state of the fibers in this space, a fiber classification function can be established to group and classify the fibers. This classification is based on the physical properties of the fibers, which can not only distinguish the characteristics of the fibers in different states, but also provide a basis for subsequent dynamic analysis. When the fiber enters the pulping system, its state will change dynamically with the changes of external operating conditions (such as temperature, pressure, flow rate, etc.). This dynamic change is described by the fiber state evolution function. The fiber state evolution function combines the fiber classification function with the changes in the pressure field and fluid dynamic viscosity during the pulping process, and describes the forces and deformations of the fiber in the flow field. In practice, the realization of this process depends on the accurate measurement of the physical field inside the pulping equipment. Through the sensors installed in the equipment, the distribution of the pressure field and the dynamic viscosity of the fluid can be captured in real time. These data are used as input and combined with the fiber classification function to dynamically calculate the fiber state evolution function. The fiber state evolution function is a bridge between the microscopic fiber behavior and the macroscopic physical field in the pulping process. It reflects how the length, diameter and orientation angle of the fiber change over time, and formally expresses these changes through mathematical models.

[0035] The dynamic behavior of the fiber directly determines the macroscopic physical properties of the pulping sample. Therefore, in the physical property analysis unit, the system further constructs the rheological property model and flocculation kinetic equation of the pulping sample based on the fiber state evolution function. The core of the rheological property model is to describe the stress-shear rate relationship of the pulping sample, that is, how the internal stress of the sample responds when the sample is subjected to external shear force. As a complex fluid, the viscosity and flow characteristics of the pulping sample are not only related to the shear rate, but also significantly affected by the fiber concentration and orientation state. Through the rheological property model, the system can predict the flow behavior of the pulping sample under different operating conditions, thereby providing a basis for optimizing the process parameters. At the same time, in order to describe the interaction and aggregation behavior of the fibers in the flow field, the system also establishes the flocculation kinetic equation. Flocculation is a key microscopic phenomenon in the pulping process, which refers to the process in which the fibers aggregate into agglomerates under the action of interaction forces (such as van der Waals forces or electrostatic forces). Flocculation behavior directly affects the microstructure and macroscopic properties of the pulping sample. The flocculation kinetic equation describes the rearrangement and aggregation process of the fibers in the flow field by modeling the force conditions between the fibers. The parameters of this equation are determined by the flow field characteristics, fiber state and external operating conditions of the pulping sample, and can reflect the fiber aggregation behavior under the current process conditions in real time.

[0036] On the basis of rheological property model and flocculation kinetic equation, the system further integrates fiber state evolution function and constructs performance evolution model of pulping sample. The performance evolution model provides a unified mathematical framework for global optimization of process parameters by linking microscopic characteristics of pulping sample with macroscopic performance. Specifically, the model characterizes key performance indicators in pulping process, such as viscosity, flocculation degree, energy consumption and finished fiber quality, in the form of multi-objective optimization function. Through optimization algorithm, the performance evolution model can find the best trade-off between these objectives to ensure that the pulping process is both efficient and stable. The feedback intelligent control unit is the execution core of the whole system. Its role is to convert the optimization results of performance evolution model into specific control instructions to guide the operation of pulping equipment. In practice, this unit first collects the operation data of pulping station in real time through sensor network, including temperature field, pressure field, velocity field and concentration field. Then, these data are input into performance evolution model to calculate the value of multi-objective optimization function in the current state. Then, the system solves the function through optimization algorithm to find the operating conditions that optimize the overall performance indicators, such as the best temperature, pressure, flow rate and fiber concentration distribution. Finally, these optimization results are converted into equipment control parameters to adjust the operating status of the pulping station. This control process is dynamic and can respond to changes in the external environment and disturbances in process conditions in real time. For example, when the sensor detects an abnormal pressure field distribution, the system can quickly adjust the operating parameters of the equipment by recalculating the performance evolution model to ensure stable operation of the pulping process. The real-time regulation capability of the feedback intelligent control unit not only improves the operating efficiency of the pulping station, but also reduces energy consumption and resource waste. Compared with traditional pulping control technology, the intelligent control system of the ground pulping station is innovative in many aspects in principle.

[0037] First, by introducing three-dimensional feature space and fiber classification function, it quantifies the complex properties of fibers into mathematical expressions that can be directly calculated, greatly improving the accuracy and efficiency of fiber property analysis. Secondly, the establishment of the fiber state evolution function enables the system to dynamically describe how the physical properties of the fiber evolve as external conditions change, which is difficult to achieve in traditional technologies. In addition, the combination of rheological property models and flocculation kinetic equations provides new tools for analyzing the physical properties of pulping samples. These models can not only explain microscopic phenomena, but can also be used to guide macroscopic process optimization. Finally, the combination of performance evolution models and feedback intelligent control units transforms complex multi-objective optimization problems into practical and operational control strategies, achieving a seamless transition from theoretical models to engineering applications.

[0038] Embodiment 2: The pulping sample characteristic analysis and evolution unit uses an optical microscope or a scanning electron microscope to obtain a geometric image of the sample fibers in the pulping sample, and uses automatic image processing technology to extract the length, diameter and orientation angle of each sample fiber from the geometric image as the sample fiber characteristics; the resolution of the optical microscope or the scanning electron microscope is at least 1 micron / pixel.

[0039] Specifically, during the imaging process, optical microscopes rely on optical lens systems to magnify the image of sample fibers and record the image in digital format through digital cameras, while scanning electron microscopes use the interaction between electron beams and sample surfaces to generate high-resolution images. Their accuracy and ability to resolve surface details are particularly suitable for the observation of complex fiber samples. Regardless of the microscopic technique used, the generated fiber geometric image contains information about the microscopic state of the sample, which is converted into quantifiable data through automated image processing technology. In image processing, the system first pre-processes the original microscopic image, including noise reduction, contrast enhancement, and background correction, to improve the contrast between the fiber and the background, thereby ensuring clear visibility of the fiber boundary. Then, the fiber is separated from the background through advanced image segmentation algorithms. This process uses tools such as edge detection and deep learning segmentation technology to accurately identify the boundaries of each fiber, laying the foundation for subsequent geometric parameter extraction. After the segmentation is completed, the system performs a geometric analysis of the shape and position of each fiber, extracting its length, diameter, and orientation angle respectively. The length of the fiber is calibrated by counting the number of pixels in the direction of the main axis and combining it with the resolution of the microscope, while the fiber diameter is obtained by measuring the width in the direction perpendicular to the main axis and statistically averaging multiple measurement positions. The calculation of the orientation angle is based on the angle between the main axis of the fiber and a specific reference direction (such as the flow direction). This parameter reflects the arrangement of the fiber in the pulping fluid and thus characterizes the orientation characteristics of the pulping sample. Through this series of processing steps, the system converts the geometric properties of a large number of fibers in the sample into structured data, which is used as the input of the fiber classification function to analyze the distribution and classification of the fibers.

[0040] Example 3: The process of extracting the length, diameter and orientation angle of each sample fiber from a geometric image using automatic image processing technology specifically includes: converting the collected geometric image into a grayscale image; applying a filter to smooth the grayscale image and remove noise; using an edge detection algorithm to extract the boundaries of the sample fibers from the grayscale image and output an edge image; using a global threshold segmentation method to segment the edge image, and using a connectivity analysis algorithm to mark the pixel area of ​​each sample fiber; in the pixel area of ​​each sample fiber, using a minimum circumscribed rectangle to calculate the length of the sample fiber; using the ratio of area to length to calculate the diameter; and obtaining the orientation angle of the sample fiber through a principal component analysis method.

[0041] Specifically, the collected geometric images are usually stored in digital form, which may contain noise, background information, and complex geometric structures of fibers. In order to extract the key features of the fibers, these original geometric images must first be converted into grayscale images. Grayscale images reduce data complexity by compressing the color information of each pixel into a single intensity value, while retaining the structural information of the fiber. This step lays the foundation for subsequent image processing and improves the processing speed and efficiency of the algorithm. Next, the grayscale image is smoothed by a filter to remove noise. The fiber geometric images in the pulping sample are usually accompanied by random noise during the acquisition process, such as optical noise of the microscope or background texture interference. These noises may interfere with the edge detection and segmentation process. Using filtering techniques (such as Gaussian filters or median filters), the intensity changes in the image can be smoothed while maintaining the clarity of the fiber edges. This noise suppression method not only improves the accuracy of the processing results, but also reduces the false detection rate, providing a cleaner input image for subsequent edge extraction. After smoothing, the system uses an edge detection algorithm to extract the boundaries of the sample fibers from the grayscale image. The core of the edge detection algorithm is to identify areas in the image where the pixel intensity changes significantly, which usually correspond to the boundaries of the fiber. In practice, the classic Canny edge detection algorithm can be used, which identifies the high-contrast area of ​​the fiber boundary by calculating the gradient change of the image. The result of edge detection is a binary edge image in which the fiber boundary is clearly marked. This edge image is the key intermediate data for extracting fiber geometric information.

[0042] In order to segment the fiber area in the edge image, the system applies the global threshold segmentation method. Global threshold segmentation divides the image into foreground and background areas by setting a uniform grayscale value, thereby converting the boundary information of the fiber into a connected area. In order to ensure that each fiber area can be correctly identified, the pixel area of ​​each sample fiber is further marked using the connectivity analysis algorithm after segmentation. The role of connectivity analysis is to identify the relationship between adjacent pixels in the image and combine pixels belonging to the same fiber into independent marked areas. This process ensures that the system can accurately distinguish and separate multiple fibers in the sample. In each marked fiber area, the system uses the minimum enclosing rectangle method to calculate the fiber length. The minimum enclosing rectangle is the smallest rectangle that contains the fiber pixel area, and its length in the main axis direction is approximately the actual length of the fiber. With this method, the system can quickly and accurately estimate the fiber length, and even if the fiber presents a slightly curved shape, the robustness of the result can be guaranteed. For the calculation of fiber diameter, the system is based on the ratio of area to length. The area of ​​the fiber pixel area can be obtained by counting the number of pixels in the area. Combined with the fiber length calculated in the previous step, the diameter can be estimated by the ratio of area to length. This method can effectively handle irregular fiber cross-sections and provide an average value of the diameter as a characterization. The orientation angle is calculated using the principal component analysis (PCA) method. PCA is a linear dimensionality reduction technique that determines the main axis direction of the fiber by analyzing the distribution direction of the fiber pixel area and calculates its angle with the reference direction. The advantage of the PCA method is that it can handle fibers of complex shapes and provide stable orientation angle estimates even if the fiber boundaries are slightly irregular.

[0043] Example 4: Dimensional classification function Use the following formula to express it:

[0044] ;

[0045] in, is the length variable of the fiber classification function; is an integer subscript index; is the number of sample fibers extracted from the geometric image; is the diameter variable of the fiber classification function; is the orientation angle variable of the fiber classification function; is the maximum length of the sample fiber; is the maximum diameter of the sample fiber; is the maximum orientation angle of the sample fiber; is the standard deviation of the length of the sample fibers; is the standard deviation of the diameter of the sample fibers; is the standard deviation of the orientation angle of the sample fibers; For the The length of the sample fiber; For the The diameter of the sample fiber; For the Orientation angle of the sample fibers; is the length integral differential; is the integral differential of the orientation angle; is the diameter integral differential; is the distance from the cluster center to the origin of the points corresponding to all sample fiber features in the feature space.

[0046] Specifically, the basic structure of the fiber classification function can be understood as a type of weighted integral model, whose main goal is to accumulate the characteristic distribution of each fiber into the global classification result through triple integral calculation. , , They correspond to the length, diameter and orientation angle of the fiber respectively, which together constitute the three-dimensional characteristic space of the fiber. Through integral calculation, the function covers the continuous distribution range of the entire characteristic space, so that it can comprehensively characterize the state changes of the sample fiber in different characteristic dimensions. This process is not only a description of the individual characteristics of the fiber, but also an induction and summary of the overall fiber distribution law. In the specific implementation, the function first converts the actual measured characteristic value of each fiber into The formula is embedded into the feature space as discrete points. By accumulating the feature contributions of all sample fibers, the formula constructs a global classification framework to ensure that the results are statistically significant and robust. The mathematical core of the formula is reflected in its weighted and distance metric parts. Distance metric Describes the integration variable With sample fiber feature points This part adopts the form of standardized Euclidean distance, through the standard deviation parameter , , Normalize the contributions of different feature dimensions. Length, diameter, and orientation angle have different physical meanings and distribution characteristics in the pulping process. For example, the distribution of length may be more discrete, while the diameter usually has a more concentrated distribution pattern. Through the standard deviation parameter, the formula can dynamically adjust the weights of different feature dimensions in the calculation to ensure the balance and consistency of the classification results. This design is particularly suitable for complex industrial scenarios such as ground pulping stations, because in the actual pulping process, the different characteristics of the fiber may be affected by different physical conditions (such as pressure and temperature), and this weighting mechanism can effectively eliminate these interferences.

[0047] Based on the weighted distance metric, the formula is further improved by the Gaussian distribution function The distance is transformed nonlinearly, thereby enhancing the sensitivity to the feature center and suppressing the influence of edge data points. The physical significance of this design is that the distribution of fibers in the pulping process often shows a certain degree of aggregation. The central area of ​​the fiber characteristics can better reflect the main state of the sample, while the edge points may be noise or abnormal data. Through Gaussian weighting, the formula gives a higher weight to the feature center, thereby highlighting the contribution of the main features in the classification process. This nonlinear weighting mechanism makes the fiber classification function more stable and accurate when processing complex samples. The setting of the integral range is another key point of the formula. The upper and lower limits of the integral are respectively determined by the maximum length of the sample fiber. , Maximum diameter and the maximum orientation angle Determined. These maximum values ​​are determined by statistical analysis of sample data and have clear physical boundary significance. The setting of the integral range ensures that the calculation of the formula can cover all possible distributions of fiber characteristics, while also providing flexibility for processing diverse pulping samples. In industrial applications, different pulping conditions may lead to significant changes in the distribution of fiber characteristics. For example, higher pressure may lead to a more concentrated distribution of fiber orientation angles, while higher temperatures may increase the distribution range of fiber length. By dynamically adjusting the integral range, the formula can adapt to changes in characteristics under different process conditions, thereby enhancing the adaptability of the intelligent control system of the ground pulping station. In addition, the distance from the cluster center to the origin in the formula is a global constraint parameter, which mathematically defines the spatial distribution range of sample feature points. The introduction of this parameter not only strengthens the formula's ability to describe the global distribution, but also provides an additional constraint mechanism through the spatial aggregation of feature points. In the actual pulping process, the distribution of fiber characteristics is often affected by the combined effects of process conditions and equipment parameters. The existence of enables the formula to reflect the superposition effect of these influences at the global level. Through this parameter, the fiber classification function can more accurately capture the macroscopic characteristics of the pulping sample, providing more comprehensive data support for subsequent state evolution modeling and optimization control.

[0048] Example 5: Fiber state evolution function Use the following formula to express it:

[0049] ;

[0050] in, For time; , is the diffusion coefficient matrix, is the preset diffusion coefficient value; is the Z-axis coordinate; is the X-axis coordinate; is the Y-axis coordinate; For time The pressure field at that time; is the dynamic viscosity of the fluid; ;in, is the mean of the orientation angles of all sample fibers; is the Laplace operator.

[0051] Specifically, in the formula, the diffusion term is in the form of , whose core function is to describe the random diffusion behavior of fibers in three-dimensional space. The essence of diffusion behavior lies in the thermal motion of fibers in the fluid and the irregular motion caused by local flow field disturbances. In three-dimensional space, the diffusion coefficient matrix It is set as a diagonal matrix and assumes that the diffusion coefficient is equal in all directions, that is, isotropic. This design simplifies the computational complexity and can effectively capture the natural expansion trend of the fiber state in spatial distribution. It is a mathematical tool used in the formula to describe diffusion behavior. It is used to calculate the second-order derivative of the fiber state at each point in space, thereby reflecting the diffusion rate of the fiber state near that point. The existence of the diffusion term ensures the uniform distribution trend of the fiber state in the sample space, which is a necessary condition to ensure the stability of the internal structure of the pulping sample. In practical applications, the diffusion coefficient The size of is directly related to factors such as the dynamic viscosity, temperature and external disturbance of the fluid. The dynamic adjustment of can optimize the fiber diffusion rate to meet different process requirements. The migration term describes the migration behavior of the fiber under pressure drive through the coupling of the formalized pressure field and the fiber classification function. This term includes the pressure field gradient , fluid viscosity , fiber classification function Normalization, and along The directional derivative of the axis. Its physical significance is to reveal how the fiber migrates in space under the drive of external pressure changes. In the pulping process, the pressure field is one of the main driving forces affecting the fiber movement, especially when the fiber is in a non-uniform flow field, the pressure gradient will cause the fiber to migrate to the low-pressure area, thereby affecting the overall distribution state of the sample. The dynamic viscosity of the fluid It plays a role in regulating the migration rate, and its size reflects the degree of obstruction of the fiber in the fluid. When the viscosity is high, the movement of the fiber tends to be slow; conversely, when the viscosity is low, the fiber is more likely to respond to the drive of the pressure gradient. Fiber classification function in the migration term By weighting the fibers with different characteristics, the migration rate can reflect the differences in fiber characteristics. For example, longer or thicker fibers may migrate slower due to their characteristics, while thinner fibers may have greater migration flexibility. Combining these factors, the migration term accurately describes the dynamic behavior of the fiber under the combined action of pressure and viscosity, providing a theoretical basis for controlling the uniformity of fiber distribution. The source term is obtained by rotating the matrix The square root form of the fiber classification function further describes the fiber state changes caused by external disturbances or generated within the system. The average orientation angle of the sample fibers The decision directly reflects the overall rotation trend of the fiber under the action of external force or flow field. The orientation of the fiber is one of the important parameters affecting the pulping quality, and whether its distribution is uniform directly determines the structural properties of the fiber network. In actual processes, when the flow field or external pressure fluctuates violently, the orientation distribution of the fiber may be significantly affected, resulting in uneven flocculation or substandard performance. The source term simulates the dynamic adjustment behavior of the fiber under these disturbance conditions by adjusting the rotation matrix. In addition, The introduction of enables the source term to dynamically adjust the disturbance intensity according to the classification weight of the fiber and the overall characteristics of the sample, thereby achieving a flexible response to the changes in the fiber state during the pulping process. The physical significance of this mechanism is that when the internal or external environment of the pulping sample changes, the system can quickly restore the stable state of the fiber through the adjustment of the source term to avoid uneven distribution or performance degradation caused by disturbances. Overall, the fiber state evolution function comprehensively describes the dynamic behavior of the fiber in time and space through the coupling of the three processes of diffusion, migration and rotation. The core advantage of the formula lies in its unified description ability of multiple physical fields. By integrating the pressure field gradient, viscosity adjustment, fiber property weighting and dynamic adjustment of the rotation matrix, the formula can adapt to a variety of pulping process conditions. In the intelligent control system of the ground pulping station, this function is not only used as a theoretical tool to describe the fiber behavior, but also a key basis for dynamically adjusting the system operation parameters. By calculating the evolution trajectory of the fiber state in real time, the system can quickly optimize the process parameters, such as adjusting the pressure field distribution to improve the fiber uniformity, or adjusting the fluid viscosity to increase the migration rate.

[0052] Embodiment 6: The physical property analysis unit constructs a rheological property model of the pulping sample according to the fiber state evolution function using the following formula:

[0053] ;

[0054] in, For time stress at the time; is the beating degree of the pulping sample; is the surface roughness of the pulped sample; For time The temperature field; is the time integral variable, ranging from 0 to ; For time The concentration field at .

[0055] Specifically, in the rheological property model, stress is the fiber network at a particular moment The overall stress performance of the fiber directly reflects the interaction between the fiber and the fluid, and between the fibers. It plays a global regulatory role in the formula and is directly related to the distribution density of the fibers. The higher the fiber concentration, the more fibers there are in the sample and the greater the probability of interaction. Therefore, the contribution of the concentration field to the stress is positively correlated. During the pulping process, the distribution of the concentration field may be affected by fluid disturbances, pressure changes, or temperature fluctuations. By Incorporating the model, the system is able to capture in real time the effect of fiber concentration changes on stress behavior and adjust operating parameters accordingly to maintain uniform fiber distribution. It is an important parameter in stress calculation. Its physical meaning is to describe the resistance of fluid to fiber movement. The higher the viscosity, the more difficult it is for the fiber to move in the fluid, which leads to increased friction between the fibers and increased stress. In the pulping process, viscosity is usually affected by temperature, fiber surface roughness and pressure. The viscosity parameter in the formula is calculated by the fiber beating degree. and surface roughness Further modification fully reflects the influence of fiber properties on rheological behavior. It is an important indicator to measure the softness of fibers under mechanical shearing. The higher the value, the easier it is for the fiber to deform and form a network with other fibers. This network effect will enhance the viscous coupling between fibers, thereby significantly amplifying the stress. It reflects the subtle characteristics of the fiber surface structure. The higher the roughness, the larger the contact area between the fiber and the fluid, and the stronger the interaction between the fibers. The increase in this microscopic friction will also increase the stress value.

[0056] Fiber state evolution function The introduction of the formula enables the model to reflect the dynamic changes of fiber state and its influence on rheological properties. , the formula incorporates the current state of the fiber and its spatial distribution into the stress calculation. Describes the coupling effect between fiber state and beating degree, which indicates that the mechanical flexibility of the fiber will directly affect its interaction strength in the fluid. The diffusion effect of fiber state is introduced through the Laplace operator to describe the uniformity of fiber distribution in the pulping sample. Combining these two parts, the fiber state evolution function provides an accurate description of the dynamic changes of fibers in space and time, making the stress calculation closer to the actual process conditions. In the denominator of the formula, the integral term describes the accumulation and attenuation of stress over time by introducing the temperature field and pressure field gradients. Temperature field It is a time difference function, and its physical meaning is to reflect the influence of temperature change during the force application of the fiber. Higher temperature will reduce the rigidity of the fiber and the viscosity of the fluid, thereby weakening the interaction between the fibers and reducing the stress value; conversely, lower temperature will make the fiber harder and the stress will increase accordingly. Through time integration, the formula can capture the cumulative effect of temperature fluctuations at different time points on the current stress, making the model dynamic responsive. Equally important is the pressure field gradient. , which describes the force changes of the fiber under dynamic pressure environment. When the pressure gradient is large, the movement of the fiber in the fluid is significantly disturbed, which has a stronger impact on the stress behavior. By using the exponential decay function of the temperature field and pressure field gradient The modeling formula can effectively smooth out the drastic fluctuations of these external factors and ensure the stability and physical rationality of stress calculation.

[0057] Embodiment 7: The physical property analysis unit establishes the flocculation kinetic equation of the pulping sample according to the rheological property model through the following formula:

[0058] ;

[0059] in, For pulping samples at time The velocity field at time ; For pulping samples at time The number density of flocculated particles at is the flocculation diffusion coefficient, which is the set value; is the gradient operator.

[0060] Specifically, the basic structure of the flocculation kinetics equation consists of two core parts: the time derivative term and the spatial variation term. It describes the change of the number density of flocculation particles over time. Its physical significance lies in capturing the generation and dissipation rate of particles. In the pulping process, the generation of flocculation particles is affected by the interaction between fibers, pressure gradient and fiber surface characteristics, while the dissipation of particles is related to the turbulence intensity of the fluid and the deflocculation process of the fiber. Through this time derivative, the system can dynamically monitor the growth trend of the number of particles and then infer the overall progress of the flocculation process. The spatial variation term is calculated through the gradient operator Other terms related to diffusion describe the migration and distribution behavior of flocculation particles in space. The pulping sample is at time The physical meaning of the fluid velocity distribution during pulping is the motion characteristics of fiber particles driven by the fluid. During pulping, the non-uniformity of the velocity field can lead to significant differences in the distribution of fiber particles in different areas. For example, high-speed areas may cause particles to migrate faster, while low-speed areas may enhance the aggregation effect of particles. The formula can accurately describe the effect of velocity field changes on the spatial distribution of flocculation particles, thus providing data support for controlling the uniformity of fibers.

[0061] Terms in formulas It is one of the core of the kinetic equation, which combines fiber properties, pressure field gradient and diffusion effect to describe the movement law of fiber particles in the vertical direction. and fiber surface roughness The two important parameters that determine the strength of fiber-particle interaction are: higher beating degree makes the fiber softer and the particles more easily aggregated under pressure, while higher roughness increases the friction and adsorption between particles, thereby enhancing the flocculation effect. The pressure gradient has a negative impact on the flocculation behavior by inhibiting the rate of particle movement. Through the synergistic effect of these three parameters, the formula can dynamically adjust the particle migration rate to adapt to different pulping conditions. and fiber state diffusion term The movement law of flocculation particles in complex scenes is further refined. The physical meaning of pressure gradient is to describe the force distribution of fiber particles in the flow field. When the pressure gradient is large, the migration rate of fiber particles is significantly accelerated, while when the pressure gradient is small, particles are more likely to stay in a local area and aggregate. In addition, the fiber state diffusion term By describing the uniformity of the fiber state distribution in space, the influence of diffusion on flocculation behavior is revealed. The introduction of the diffusion term enables the formula to capture the spontaneous diffusion effect of fiber particles, which is an important mechanism for achieving the balance of the flocculation process. In the last term of the kinetic equation, , the formula further introduces the influence of fiber classification function and external disturbance. Rotation matrix Describes the average orientation angle of the fiber particles The physical meaning of the overall rotation effect under the action of external force is to reflect the rearrangement tendency of fibers under the action of external force. This mechanism is particularly important because the orientation distribution of fibers in the pulping process directly determines the geometry and distribution pattern of flocculated particles. In addition, the fiber classification function The influence of fibers with different characteristics is incorporated into the flocculation kinetics model through weighted processing. For example, fibers with longer length or larger diameter usually have higher adsorption capacity during flocculation, while fibers with larger orientation angles may cause more significant interference with the arrangement of particles. Used to describe the diffusion equilibrium behavior of flocculation particles, where the diffusion coefficient is a set value used to adjust the diffusion rate of particles in space. Through this correction term, the kinetic equation can ensure that at higher diffusion rates, the distribution of particle number density tends to be uniform; while at lower diffusion rates, particles gather more in local areas to form dense flocculation structures.

[0062] Embodiment 8: The physical property analysis unit constructs a performance evolution model of the pulping sample as a global multi-objective optimization function based on the flocculation kinetic equation, the rheological property model and the fiber state evolution function through the following formula:

[0063] ;

[0064] in, Function value for a global multi-objective optimization function.

[0065] Specifically, the core idea of ​​the performance evolution model is to calculate the fiber length ,diameter and orientation angle The global integration of these three characteristic parameters in the characteristic space comprehensively reflects the characteristics of all fibers in the sample and their contribution to the overall performance. The upper and lower limits of the integration are determined by the maximum length of the pulping sample. , Maximum diameter and the maximum orientation angle This ensures that the coverage of the integral completely includes the physical boundaries of the sample and provides comprehensive data support for performance optimization. The setting of the integral variable not only emphasizes the diversity of fiber microscopic properties, but also provides a high-precision description for global optimization through the cumulative calculation of the distribution of fiber properties within the sample. The first part of the formula is based on the concentration field , flocculation particle density ,stress Fiber status And fiber classification function As the core, a comprehensive performance index is constructed to describe the linkage relationship among fiber distribution, particle generation, mechanical properties and dynamic state. It is a key factor affecting the strength of fiber-to-fiber interaction. The higher the concentration, the more frequent the interaction between fibers in the sample, and the higher the flocculation efficiency and mechanical properties. However, too high a concentration may lead to excessive aggregation of fibers and affect the uniformity of the sample. Therefore, the regulation of the concentration field is crucial in the pulping process. Flocculation particle density Describes the number density of aggregates formed by fibers during the flocculation process, reflecting the generation and distribution of particles. A higher particle density usually means a stronger network structure, corresponding to better sample uniformity and mechanical properties; while a lower particle density may lead to reduced flocculation efficiency or even unevenness. Stress It is the main parameter that describes the mechanical behavior of pulping samples under pressure field and dynamic environment, and can characterize the sample's response ability to external forces and the stability of its internal structure. The introduction of enables the formula to dynamically capture the state changes of fibers during the pulping process. The changes in fiber state may be affected by multiple factors such as temperature field, pressure field and fluid viscosity. By incorporating the state evolution function into the model, the formula can reflect how the sample performance changes with the dynamic adjustment of process conditions. Fiber classification function By adjusting the weights, the contributions of fibers with different characteristics can be calculated separately. For example, fibers with longer length or thicker diameter may have a more significant enhancement effect on the mechanical properties of the sample, while fibers with more uniform orientation angle distribution may help improve the overall uniformity of the sample.

[0066] The formula also introduces an exponential decay term , which is used to describe the effect of the rate of change of fiber state on sample performance. When the fiber state changes rapidly, the value of the attenuation term is small, indicating that the stability of the sample is reduced; when the fiber state changes tend to be stable, the value of the attenuation term is large, indicating that the sample performance is more stable. Through this design, the performance evolution model can dynamically reflect the effect of process condition fluctuations on sample performance, and provide a quantitative basis for the optimization process. The second part of the formula is expressed by the gradient term The modulation of the trigonometric function further captures the spatial variation characteristics of the sample performance. The gradient term describes the spatial variation rate of the flocculation particle density, stress, fiber state and classification function. Its physical meaning is to reflect the local fluctuations of the sample performance in different spatial regions. For example, in high-density areas, the distribution of flocculation particles may be more drastic, and the corresponding stress changes are more significant, while in low-density areas, the changes in the fiber state may be relatively gentle. Through this gradient term, the performance evolution model can describe the complex behavior of the sample under non-uniform flow field conditions. The trigonometric function modulation term is obtained by , and Model the periodic changes of fiber length, diameter and orientation angle. The introduction of these trigonometric functions not only improves the flexibility of the model, but also reflects the contribution pattern of fibers with different characteristics in the sample performance. For example, the periodic changes in fiber length may be directly related to the network strength of the sample, while changes in diameter and orientation angle may affect the interaction and distribution uniformity between fibers. By embedding the performance evolution model into the intelligent control system of the ground pulping station, the system can calculate the global performance indicators of the sample in real time and dynamically adjust the process parameters according to the optimization objectives. Global multi-objective optimization function The core advantage of the system lies in its comprehensiveness and dynamics. It can not only describe the instantaneous state of sample performance during pulping, but also predict the long-term impact of changes in process conditions on sample performance. In practical applications, the system can quickly adjust the pressure field, temperature field and fluid viscosity through real-time monitoring and calculation of the optimization function value to achieve comprehensive optimization of flocculation efficiency, uniformity and mechanical properties.

[0067] Example 9: The constraints of the global multi-objective optimization function for pulping samples are:

[0068] ;

[0069] .

[0070] Specifically, the core of the first constraint is to control the number density of flocculation particles. Its mathematical form includes flocculation density , fiber state evolution , pressure gradient , diffusion term , and the fiber characteristic periodic variation term These terms act together on the dynamic behavior of the sample, reflecting the temporal and spatial dependence of the flocculation process. This constraint condition, in the form of an integral, limits the cumulative effect of the flocculation density and its dynamic influence to a certain threshold, i.e. This upper limit is determined by the fiber classification function and fluid viscosity The square root of indicates the constraints of fiber properties and fluid resistance on flocculation behavior. The physical significance of the constraint condition is to avoid the adverse effects caused by too high or too low number density of flocculation particles. Too high flocculation density may lead to excessive aggregation of fibers, thereby reducing the uniformity and fluidity of the sample; while too low flocculation density may lead to insufficient network structure and fail to form high-quality fiber products. By limiting the dynamic changes of flocculation density and its relationship with pressure gradient, diffusion and fiber orientation angle, the system can ensure the balance and stability of the flocculation process during dynamic operation.

[0071] The second constraint is mainly about the stress behavior of the pulped sample and its relationship with concentration field , flocculation particle diffusion This condition is expressed by the exponential term Describes the time decay effect of fiber state change on sample stress. The introduction of the exponential term shows that when the fiber state changes rapidly, the cumulative effect of stress is suppressed; and when the fiber state tends to be stable, the decay effect is weakened, so that the stress can better reflect the overall mechanical properties of the sample. The right side of this constraint is , indicating that the cumulative change of sample stress must be regulated by the dynamic behavior of temperature and pressure fields. The physical significance of this constraint is that by limiting the accumulation of stress and its dynamic influence, the sample can be prevented from being overloaded or unstable during the pulping process. The dynamic change of the pressure field gradient may cause drastic fluctuations in stress, while the coupling of concentration field and particle diffusion may lead to inhomogeneity in sample performance. By combining restrictions on these factors, the system can maintain the mechanical stability of the sample in actual operation while improving the efficiency of the pulping process and product quality. On the whole, the design of these constraints not only takes into account the core performance indicators of the pulping sample, but also fully combines the complex dynamic influences of fiber properties, fluid dynamics behavior and external physical fields. In practical applications, these constraints provide the intelligent control system with operational boundaries and optimization directions, which are calculated in real time and combined with the global multi-objective optimization function. The system can find the optimal performance solution while satisfying the constraints by combining the results of the optimization method. This optimization method is particularly suitable for dealing with complex situations with multiple variables, multiple objectives, and multiple constraints in the pulping process, such as optimizing energy consumption and time cost while ensuring flocculation efficiency and stress uniformity.

[0072] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. The intelligent control system of the ground pulping station is characterized by: The system comprises: a pulping sample feature analysis and evolution unit, a physical property analysis unit and a feedback intelligent control unit; the pulping sample feature analysis and evolution unit is used to obtain the length, diameter and orientation angle of each sample fiber from the pulping sample as the sample fiber feature; each sample fiber feature is regarded as a point in the three-dimensional feature space, so as to map each sample fiber to the three-dimensional feature space to obtain a fiber classification function; based on the fiber classification function, the pressure field of the pulping sample and the dynamic viscosity of the fluid during the pulping process are obtained, and a fiber state evolution function is constructed to describe the dynamic change of the state of the sample fiber during the pulping process; the physical property analysis unit is used to construct a rheological property model of the pulping sample according to the fiber state evolution function to describe the pulping sample during the pulping process. stress changes; and according to the rheological property model, the flocculation kinetic equation of the pulping sample is established; on the basis of the flocculation kinetic equation, the rheological property model and the fiber state evolution function, the performance evolution model of the pulping sample is constructed as the global multi-objective optimization function of the pulping sample; the feedback intelligent control unit is used to solve the global multi-objective optimization function, and obtain the temperature field, pressure field, velocity field and concentration field corresponding to the maximum value of the global multi-objective optimization function as the operation control target of the ground pulping station, and control the operation of the ground pulping station according to the operation control target; the physical property analysis unit, through the following formula, on the basis of the flocculation kinetic equation, the rheological property model and the fiber state evolution function, constructs the performance evolution model of the pulping sample as the global multi-objective optimization function: ； in, Function value for global multi-objective optimization function; For pulping samples at time The number density of flocculated particles at For time The concentration field at For time stress at the time; is the fiber classification function; is the length variable of the fiber classification function; is the maximum length of the sample fiber; is the maximum diameter of the sample fiber; is the maximum orientation angle of the sample fiber; is the diameter variable of the fiber classification function; is the orientation angle variable of the fiber classification function; is the gradient operator; is the time integral variable, ranging from 0 to ; is the mean of the orientation angles of all sample fibers; is the fiber state evolution function.

2. The intelligent control system for a ground pulping station according to claim 1, characterized in that: The pulping sample characteristic analysis and evolution unit uses an optical microscope or a scanning electron microscope to obtain a geometric image of sample fibers in the pulping sample, and uses an automatic image processing technique to extract the length, diameter and orientation angle of each sample fiber from the geometric image as the sample fiber characteristic; The optical microscope or scanning electron microscope has a resolution of at least 1 micron / pixel.

3. The intelligent control system for a ground pulping station according to claim 2, characterized in that: The process of extracting the length, diameter and orientation angle of each sample fiber from the geometric image using automatic image processing technology specifically includes: converting the collected geometric image into a grayscale image; applying a filter to smooth the grayscale image and remove noise; using an edge detection algorithm to extract the boundary of the sample fiber from the grayscale image and output an edge image; using a global threshold segmentation method to segment the edge image and using a connectivity analysis algorithm to mark the pixel area of ​​each sample fiber; in the pixel area of ​​each sample fiber, using the minimum circumscribed rectangle to calculate the length of the sample fiber; using the ratio of area and length to calculate the diameter; and obtaining the orientation angle of the sample fiber through a principal component analysis method.

4. The intelligent control system for a ground pulping station according to claim 3, characterized in that: Fiber classification function Use the following formula to express it: ； in, is an integer subscript index; is the number of sample fibers extracted from the geometric image; is the standard deviation of the length of the sample fibers; is the standard deviation of the diameter of the sample fibers; is the standard deviation of the orientation angle of the sample fibers; For the The length of the sample fiber; For the The diameter of the sample fiber; For the Orientation angle of the sample fibers; is the length integral differential; is the integral differential of the orientation angle; is the diameter integral differential; is the distance from the cluster center to the origin of the points corresponding to all sample fiber features in the feature space.

5. The intelligent control system for a ground pulping station according to claim 4, characterized in that: Fiber state evolution function Use the following formula to express it: ； in, For time; , is the diffusion coefficient matrix, is the preset diffusion coefficient value; is the Z-axis coordinate; is the X-axis coordinate; is the Y-axis coordinate; For time The pressure field at that time; is the dynamic viscosity of the fluid; ;in, is the Laplace operator.

6. The intelligent control system for a ground pulping station according to claim 5, characterized in that: The physical property analysis unit constructs a rheological property model of the pulping sample according to the fiber state evolution function through the following formula: ； in, is the beating degree of the pulping sample; is the surface roughness of the pulped sample; For time The temperature field;.

7. The intelligent control system for a ground pulping station according to claim 6, characterized in that: The physical property analysis unit establishes the flocculation kinetic equation of the pulping sample according to the rheological property model through the following formula: ； in, For pulping samples at time The velocity field at time ; is the flocculation diffusion coefficient, is the set value.

8. The intelligent control system for a ground pulping station according to claim 7, characterized in that: The constraints of the global multi-objective optimization function for the pulping sample are: ； 。

Images