A high speed paper cutter control method and system

Abstract

The application discloses a kind of high-speed paper cutter control method and system, method includes: the paper characteristic data of paper to be cut is acquired in real time, paper characteristic data is analyzed by deep learning model, and the corresponding best cutting parameter set is generated, wherein the paper characteristic data includes thickness, humidity and material composition data;During the cutting process of high-speed paper cutter, the actual cutting data in cutting process is monitored in real time by embedded sensor;According to the actual cutting data, the cutting parameters of high-speed paper cutter are dynamically adjusted using adaptive control technology to ensure the consistency and accuracy of cutting quality. Using the embodiment of the application, intelligentization and automation of the cutting process can be realized, and the consistency and accuracy of cutting quality are ensured.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of control, in particular to a high-speed paper cutter control method and system. BACKGROUND

[0002] In modern manufacturing, paper processing involves cutting processes that play a crucial role. High-speed paper cutters are widely used in cutting various types of paper to meet the requirements of production efficiency and precision. However, the complexity of the paper cutting process and the diversity of paper itself pose many challenges in actual operation. Different types of paper have significant differences in thickness, humidity, and material composition, which require the establishment of corresponding cutting parameters for different papers to ensure the stability and consistency of cutting quality.

[0003] Traditional cutting control methods often rely on experience and static parameter-based settings, which are difficult to adapt to changes in paper characteristics. With changes in production environment, static parameters may not cover all possible cutting conditions, resulting in inconsistent cutting quality, and sometimes even paper damage or inaccurate cutting. In addition, the lack of real-time monitoring and feedback mechanisms makes it difficult for production lines to respond quickly when problems occur, thereby affecting overall production efficiency. SUMMARY

[0004] The purpose of the present application is to provide a high-speed paper cutter control method and system to solve the problems in the prior art, and to realize intelligent and automated cutting process to ensure the consistency and accuracy of cutting quality.

[0005] One embodiment of the present application provides a high-speed paper cutter control method, the method comprising:

[0006] Real-time acquisition of paper characteristic data of the paper to be cut, analysis of the paper characteristic data by a deep learning model, and generation of a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity, and material composition data;

[0007] In the cutting process of the high-speed paper cutter, embedded sensors are used to monitor the actual cutting data in the cutting process in real time;

[0008] According to the actual cutting data, the cutting parameters of the high-speed paper cutter are dynamically adjusted using adaptive control technology to ensure the consistency and accuracy of the cutting quality.

[0009] Optionally, the real-time acquisition of paper characteristic data of the paper to be cut, the analysis of the paper characteristic data by a deep learning model, and the generation of a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity, and material composition data, include:

[0010] The thickness, humidity and material composition data are fused and analyzed by using a multi-modal deep learning model. The thickness data in the paper characteristic data is taken as a one-dimensional time series feature, which is input into a sub-model based on a convolutional neural network or a recurrent neural network, for capturing the change rule of the paper thickness. The humidity data and the material composition data are taken as key static features, which are processed by a fully connected neural network to extract the influence of the physical properties of each material composition under different humidity conditions. Finally, a feature fusion layer is used to fuse the data output by the sub-models to generate a high-dimensional vector that comprehensively represents the paper characteristics.

[0011] The high-dimensional vector is input into the last fully connected layer of the multi-modal deep learning model, and a corresponding multi-dimensional parameter vector is generated as the best cutting parameter set through weight mapping and a nonlinear activation function. The best cutting parameter set at least includes cutting speed, tool pressure and tool angle, so as to realize the stability of cutting quality, the minimization of tool wear and the maximization of cutting efficiency.

[0012] Optionally, during the cutting process of the high-speed paper cutting machine, actual cutting data in the cutting process are monitored in real time by using an embedded sensor, including:

[0013] During the cutting process, a distributed embedded sensor system is used to deploy multiple types of sensors, and the data of the sensors are collected in real time in a high-frequency sampling manner, wherein:

[0014] The pressure sensor is installed on the tool carrier to collect the pressure change of the tool on the paper in real time. The acceleration sensor is installed on the tool motion path to capture the motion acceleration of the tool, for analyzing the physical vibration behavior in the cutting process.

[0015] The acoustic sensor is installed around the cutting area to record the high-frequency noise generated in the cutting process, for judging the friction or cutting abnormality.

[0016] The optical sensor monitors the micro-spectrum change of the cutting area in real time, for capturing the slight physical or chemical change of the paper caused by friction in the cutting process.

[0017] Optionally, according to the actual cutting data, an adaptive control technology is used to dynamically adjust the cutting parameters of the high-speed paper cutting machine to ensure the consistency and accuracy of the cutting quality, including:

[0018] The multi-dimensional actual cutting data collected from the sensors are integrated into a state vector, which contains various indicators of the current cutting state.

[0019] By formulating a control rule based on fuzzy logic, an adjustment strategy of the cutting parameters is defined. Based on the current state vector, a fuzzy logic controller automatically evaluates the cutting parameters that need to be adjusted to optimize the cutting quality.

[0020] combining the actual cutting data with historical cutting quality data, utilizing a machine learning algorithm to predict the performance of cutting quality under the current cutting parameter settings;

[0021] adjusting the cutting parameters of the high-speed paper cutter in real-time according to the outputs of the fuzzy logic controller and the machine learning model, wherein when the cutting quality is detected not to meet the set standards, the tool pressure or cutting speed is automatically adjusted to maintain the consistency and accuracy of the cutting.

[0022] Yet another embodiment of the present application provides a high-speed paper cutter control system, the system comprising:

[0023] an analysis module for acquiring paper characteristic data of the paper to be cut in real-time, analyzing the paper characteristic data through a deep learning model, and generating a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity, and material composition data;

[0024] a monitoring module for monitoring actual cutting data in the cutting process in real-time through an embedded sensor during the cutting process of the high-speed paper cutter;

[0025] a control module for dynamically adjusting the cutting parameters of the high-speed paper cutter according to the actual cutting data by utilizing adaptive control technology to ensure the consistency and accuracy of the cutting quality.

[0026] Yet another embodiment of the present application provides a storage medium having a computer program stored therein, wherein the computer program is configured to execute the method described in any of the above embodiments when executed.

[0027] Yet another embodiment of the present application provides an electronic device comprising a memory and a processor, wherein the memory has a computer program stored therein, and the processor is configured to execute the computer program to execute the method described in any of the above embodiments.

[0028] Compared with the prior art, the high-speed paper cutter control method provided by the present application can acquire paper characteristic data of the paper to be cut in real-time, analyze the paper characteristic data through a deep learning model, and generate a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity, and material composition data; monitor actual cutting data in the cutting process in real-time through an embedded sensor during the cutting process of the high-speed paper cutter; and dynamically adjust the cutting parameters of the high-speed paper cutter according to the actual cutting data by utilizing adaptive control technology to ensure the consistency and accuracy of the cutting quality, thereby realizing the intelligentization and automation of the cutting process and ensuring the consistency and accuracy of the cutting quality. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1A hardware structure block diagram of a computer terminal of a high-speed paper cutting machine control method provided by the embodiment of the present application is shown in the figure.

[0030] Figure 2 A flowchart of a high-speed paper cutting machine control method provided by the embodiment of the present application is shown in the figure.

[0031] Figure 3 A structure diagram of a high-speed paper cutting machine control system provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0032] The embodiments described below with reference to the drawings are exemplary and are only used to explain the present application, and cannot be explained as a limitation of the present application.

[0033] The embodiment of the present application first provides a high-speed paper cutting machine control method, which can be applied to electronic equipment, such as a computer terminal, specifically, a general computer, etc.

[0034] The following will be described in detail taking a computer terminal as an example. Figure 1 A hardware structure block diagram of a computer terminal of a high-speed paper cutting machine control method provided by the embodiment of the present application is shown in the figure. Figure 1 As shown in the figure, the computer device includes a processor, a memory and a network interface connected through a system bus, wherein the memory can include a non-volatile storage medium and an internal memory.

[0035] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, which, when executed, can make the processor execute any kind of high-speed paper cutting machine control method.

[0036] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0037] The internal memory provides an environment for the running of the computer program in the non-volatile storage medium, which, when executed by the processor, can make the processor execute any kind of high-speed paper cutting machine control method.

[0038] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 1 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or less components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0039] It should be appreciated that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0040] Referring to Figure 2 Embodiments of the present application provide a high-speed paper cutter control method, which can include the following steps:

[0041] S201, real-time acquisition of paper characteristic data of paper to be cut, analysis of the paper characteristic data by a deep learning model, and generation of a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity and material composition data;

[0042] Real-time acquisition of paper characteristic data of paper to be cut means that before the cutting process starts, the physical and chemical characteristics of the paper to be cut can be comprehensively detected and recorded by sensor technology and data acquisition system. These characteristic data include the thickness, humidity and material composition of the paper. These data can be obtained by different sensors, such as ultrasonic sensors for measuring thickness, humidity sensors for measuring water content of the paper, and chemical analysis instruments for obtaining material composition information. Subsequently, these data will be input into the deep learning model, and the model will analyze these features and generate an optimal cutting parameter set to ensure that it adapts to the cutting requirements of the specific paper.

[0043] The significance of this process is to accurately analyze the characteristics of the paper to be cut by scientific methods, thereby providing reliable data support for the cutting process. By deeply understanding the physical and chemical characteristics of the paper, more suitable cutting parameters can be developed, which not only can improve the cutting quality, but also can reduce the wear of the cutting tool and improve the cutting efficiency. Ultimately, this precise cutting technology will ensure the consistency and high quality of the cutting products, meeting the production requirements.

[0044] Specifically, a multi-modal deep learning model can be used to analyze the thickness, humidity, and material composition data. The thickness data in the paper characteristic data is treated as a one-dimensional time series feature and input into a sub-model based on a convolutional neural network or a recurrent neural network to capture the change pattern of the paper thickness. The humidity data and material composition data are treated as key static features and processed through a fully connected neural network to extract the influence of each material component on the physical properties under different humidity conditions. Finally, a feature fusion layer is used to fuse the data output by the sub-models to generate a high-dimensional vector that comprehensively represents the paper characteristics.

[0045] In this step, a multi-modal deep learning model is used to fuse and analyze the data. The thickness data is treated as a one-dimensional time series feature and captured by a convolutional neural network (CNN) or a recurrent neural network (RNN). For example, a CNN is used to capture local features and focus on the short-term change trend of the thickness. The humidity and material composition data are static features and are processed through a fully connected neural network (FCNN). This method can effectively extract the influence of material composition on paper characteristics under different humidity conditions. Finally, a feature fusion layer is used to integrate the outputs of each sub-model to form a comprehensive high-dimensional vector.

[0046] The significance of this deep learning method lies in its ability to analyze paper characteristics from multiple angles and comprehensively consider various factors, avoiding the limitations of traditional single feature analysis methods. Through this efficient feature extraction and fusion, the actual cutting needs of the paper can be more accurately reflected, laying a solid foundation for subsequent generation of the optimal cutting parameter set.

[0047] In the process of implementing the multi-modal deep learning model, the first step is to preprocess the acquired paper characteristic data. This includes standardizing and normalizing the thickness, humidity, and material composition data to eliminate the influence of different dimensions and ranges on model training. The preprocessed thickness data will be treated as a one-dimensional time series feature, segmented and labeled as sample input into the convolutional neural network (CNN) and recurrent neural network (RNN). CNN will extract local features of thickness data through local perception and weight sharing mechanism to capture the change trend of paper thickness at different time points. RNN will focus on the long-range dependence of time series, especially for the sequence signal of thickness change, which helps the model learn the influence of historical thickness data on current cutting performance.

[0048] In processing humidity and material composition data, a fully connected neural network (FCNN) model is designed to extract these static features. In this stage, humidity data and material composition data will be input into different neuron layers respectively, and FCNN will learn the influence of physical properties under various conditions (such as different humidity and different material combinations) through multiple layers of nonlinear transformation. The key to this process is to choose the appropriate activation function and number of layers to ensure that the network can effectively capture complex feature relationships. In this process, the selection of training data set and the use of data augmentation techniques are also particularly important to improve the generalization ability and robustness of the model.

[0049] Finally, the feature fusion layer will merge the outputs of the above two sub-models to form a comprehensive high-dimensional vector. This vector will fully reflect the comprehensive characteristics of the paper, representing the synergistic effect of thickness, humidity and material composition. By setting a reasonable fusion strategy, such as weighted average or splicing, different sources of feature information are effectively integrated to generate a more comprehensive feature representation, thereby providing sufficient information support for the subsequent processing stage.

[0050] The high-dimensional vector is input into the last fully connected layer of the multi-modal deep learning model, and through weight mapping and nonlinear activation function, a corresponding multi-dimensional parameter vector is generated as the optimal cutting parameter set, which at least includes cutting speed, tool pressure and tool angle, to achieve the stability of cutting quality, the minimization of tool wear and the maximization of cutting efficiency.

[0051] In this step, the high-dimensional vector generated previously is input into the last fully connected layer of the multi-modal deep learning model for processing. The function of the fully connected layer is to map high-dimensional features to cutting parameter space. Using weight mapping combined with nonlinear activation functions such as ReLU or Sigmoid makes the output results have nonlinear characteristics, ensuring the expression ability and adaptability of the model. The multi-dimensional parameter vector generated finally is the optimal cutting parameter set, which includes cutting speed, tool pressure and tool angle, etc. The significance of this process is to convert complex feature information into specific cutting parameters through the trained model, thereby achieving efficient control of the cutting process. By adaptively adjusting the cutting speed, tool pressure and angle, the cutting quality can be effectively improved, and the service life of the tool can be prolonged, thereby maximizing the cutting efficiency, bringing higher economic benefits and lower failure rate to the production line.

[0052] In this step, the last fully connected layer of the multi-modal deep learning model needs to be designed first. This layer's role is to project the high-dimensional vector obtained from the feature fusion layer into the cutting parameter space. Specifically, a fully connected layer is constructed, which associates each dimension of the feature vector with the output parameters, i.e., cutting speed, tool pressure, and tool angle. During the design process, the number of layers and size are crucial. It is generally recommended to use multiple fully connected layers to increase the model's nonlinear expression ability, and use appropriate regularization techniques (such as Dropout) to prevent overfitting.

[0053] Next, the high-dimensional feature vector is mapped to a multi-dimensional parameter vector using the weight mapping mechanism. This process involves training the weights of each connection to ensure that the influence of different features on the final cutting parameters reflects the true situation. Through the backpropagation algorithm, the weights are continuously optimized to reduce the model's loss function until a satisfactory level of accuracy is achieved. In addition, the choice of activation function in the output layer also affects the results. For example, ReLU or Softmax functions can be chosen to ensure the non-negativity and reasonable range of output parameters.

[0054] Finally, through forward propagation, once the high-dimensional vector is input into the fully connected layer, the corresponding multi-dimensional parameter vector is generated. This vector will contain the optimal set of cutting parameters generated for different paper characteristics, which will be applied in real-time to the cutting process. By comparing historical cutting data and quality standards, these parameters can be dynamically adjusted for cutting needs under different conditions, ensuring consistent cutting quality, minimizing tool wear, and maximizing cutting efficiency. In this way, the entire cutting process becomes more intelligent and adaptive, providing an important guarantee for further improving production efficiency.

[0055] S202, during the cutting process of the high-speed paper cutter, real-time monitoring of actual cutting data in the cutting process is performed through embedded sensors;

[0056] The process of real-time monitoring of actual cutting data in the cutting process of the high-speed paper cutter through embedded sensors aims to ensure that data collection and analysis during the cutting process can be closely combined. This process involves multiple types of sensors that can monitor the pressure exerted by the tool on the paper, the acceleration changes of the tool, the noise generated during the cutting process, and the spectral changes of the paper during cutting. This real-time monitoring can provide important feedback information for the cutting process, which helps to understand the impact of different cutting conditions on paper quality.

[0057] The main role of this step is to provide real-time, dynamic state feedback for the cutting of the high-speed paper cutter, ensuring the stability and accuracy of the cutting process. Through high-frequency data collection, it can timely discover and identify possible abnormalities in the cutting process, such as insufficient cutting force, tool wear, or excessive cutting speed, etc. This feedback mechanism will provide data support for subsequent adaptive control strategies, enabling cutting performance to be adjusted according to actual conditions, thereby improving overall cutting quality and efficiency.

[0058] Specifically, during the cutting process, a distributed embedded sensor system can be used to deploy multiple types of sensors, and the data of these sensors are collected in real time with high-frequency sampling;

[0059] In this step, a distributed system composed of multiple embedded sensors is built, aiming to achieve comprehensive monitoring of the cutting process. This system uses high-frequency sampling technology to ensure that all sensors can capture key physical parameters in real time during the cutting process. Through different types of sensors such as pressure sensors, acceleration sensors, acoustic sensors, and optical sensors, multiple cutting parameters can be monitored simultaneously. Such system design enables the capture of rich data during the cutting process, which helps to conduct comprehensive state analysis. The significance of this step is to monitor the cutting conditions in real time through high-frequency, multi-dimensional data collection, enabling operators to obtain dynamic feedback of the cutting process in a timely manner. This comprehensive monitoring capability can detect problems in advance, assess cutting conditions, and provide a solid data foundation for subsequent adaptive control, thereby improving cutting efficiency and quality.

[0060] In implementing the distributed embedded sensor system, first, suitable sensors need to be selected, and multiple sensors need to be arranged according to the actual design of the cutting machine. Each sensor needs to be calibrated to ensure data accuracy. Each sensor is connected to the central processing unit through wireless or wired means, and data is transmitted in real time. The system also needs to have strong data processing capabilities to support high-frequency data collection and storage, ensuring real-time analysis and feedback.

[0061] Among them, the pressure sensor is installed on the tool carrier to collect the pressure changes of the tool on the paper in real time;

[0062] In this phase, the pressure sensor is installed on the tool carrier, and its role is to monitor the pressure changes exerted by the tool on the paper during the cutting process. Through high-frequency data collection, the pressure sensor can record the size and trend of pressure in real time, providing key pressure parameters during the cutting process. These data help to analyze whether the cutting force is maintained at the optimal state, so that adjustments can be made in a timely manner.

[0063] The significance of this step is to ensure that the pressure exerted by the cutter during the cutting process remains within a reasonable range, thereby improving the cutting quality and the integrity of the material. Monitoring pressure changes can help detect cutter wear or other potential issues in a timely manner, preventing cutting failures or paper damage due to improper pressure, and improving overall cutting efficiency and quality.

[0064] To achieve real-time pressure monitoring, a suitable pressure sensor needs to be selected and positioned on the cutter carrier to ensure it accurately reflects the pressure exerted by the cutter on the paper. The sensor needs to be rigorously tested and calibrated to ensure accuracy under different cutting conditions. Data is transmitted in real time to the central control unit through the data acquisition system, which will process and analyze the data and issue adjustment instructions to the control system as needed.

[0065] The acceleration sensor is installed on the cutter's motion path to capture the cutter's motion acceleration for analyzing the physical vibration behavior during cutting;

[0066] In this step, the acceleration sensor is installed on the cutter's motion path to monitor the cutter's motion acceleration in real time during the cutting process. This monitoring can capture the acceleration and deceleration performance of the cutter during movement, thereby analyzing the physical vibration behavior during cutting. By analyzing acceleration data, it can help determine whether the cutter maintains a stable cutting state and whether there is a risk of cutting quality degradation due to vibration.

[0067] The significance of this process is to identify abnormal vibrations that may occur during cutting by analyzing the cutter's motion acceleration, which can affect cutting quality and even cause cutter damage. Real-time monitoring of acceleration can effectively provide early warnings, allowing for targeted adjustment of cutting parameters to maintain cutting stability and consistency, improving production efficiency and cutting effectiveness.

[0068] In specific implementation, a high-sensitivity acceleration sensor should be selected and installed at an appropriate position on the cutter's motion path. The sensor should have high-frequency sampling capability to capture rapidly changing acceleration data. Data is transmitted in real time to the control system through the data acquisition module, which will analyze the motion acceleration in real time and make comprehensive judgments based on other monitoring data (such as pressure, noise, etc.), to timely detect and adjust the cutting state to ensure the stability of the cutting process.

[0069] Acoustic sensors are installed around the cutting area to record high-frequency noise generated during cutting to determine friction or cutting abnormalities;

[0070] In this step, acoustic sensors are installed around the cutting area to monitor and record audio signals, particularly high-frequency noises, in real-time during the cutting process. These noise signals are often associated with friction between the cutter and paper, cutting conditions of the material being cut. By analyzing the data collected by acoustic sensors, it can be determined whether there are problems such as excessive friction, poor cutter contact, or cutting abnormalities during the cutting process. The significance of acoustic monitoring lies in providing a non-contact detection method that can analyze the stability of the cutting process through sound signals. Changes in noise during cutting can indicate potential quality problems, and through timely feedback, operators can quickly adjust cutting parameters to prevent material loss, cutter damage, and production delays caused by cutting abnormalities.

[0071] In specific implementation, appropriate acoustic sensors need to be selected and placed at multiple positions around the cutting area to ensure sufficient acoustic signals are obtained. The system should have high-frequency sampling and real-time data processing capabilities, and through algorithm analysis of the collected acoustic data, potential abnormal noise patterns can be identified. In addition, acoustic data should be analyzed in combination with other sensor data to provide a more comprehensive assessment of the cutting state and ensure smooth cutting process.

[0072] Optical sensors monitor the micro-spectral changes in the cutting area in real time to capture the small physical or chemical changes caused by friction during cutting.

[0073] In this step, optical sensors are used to monitor the spectral changes in the cutting area in real time. By analyzing the interaction of light of different wavelengths with the paper material, small physical or chemical changes caused by friction and cutting can be captured. This monitoring provides in-depth insight into the state of the paper, which can help identify potential quality problems during cutting, such as surface damage or changes in material properties.

[0074] The significance of optical monitoring lies in providing a detailed assessment of the quality of the cutting product. By capturing spectral changes, potential physical and chemical changes can be identified during the cutting process, and measures can be taken to reduce losses in a timely manner. At the same time, this technology can provide rich data for post-analysis, supporting continuous improvement of the cutting process and improving the consistency and stability of cutting quality.

[0075] To implement optical monitoring, appropriate optical sensors need to be selected that can cover important wavelength ranges and ensure that the sensors can respond quickly to spectral changes. The sensors should be placed in the cutting area so that they can monitor light reflection, transmission, and scattering at any time. The system needs to have real-time data acquisition and analysis capabilities, and through spectral analysis algorithms, it can determine the state changes of the paper. Combining multi-dimensional data from sensors provides comprehensive monitoring of cutting quality, providing the basis for subsequent adaptive control.

[0076] S203, dynamically adjusting the cutting parameters of the high-speed cutting machine using adaptive control technology based on the actual cutting data to ensure consistency and accuracy of cutting quality.

[0077] Dynamically adjusting the cutting parameters of the high-speed cutting machine using adaptive control technology based on actual cutting data to ensure consistency and accuracy of cutting quality involves forming a state vector by collecting multi-dimensional cutting data from sensors in real time, such as pressure changes, acceleration, sound, and spectral changes. This state vector is used to comprehensively evaluate the current cutting process state, and then, by using fuzzy logic control rules and machine learning models, the cutting parameters are optimized and adjusted to effectively monitor and adjust the cutting quality. This dynamic adjustment mechanism ensures that the operation of the cutter is always in the best state to adapt to the characteristics of different papers and cutting conditions during the cutting process.

[0078] The significance of this process is that through adaptive control technology, the high-speed cutting machine can respond to changes in real time during the cutting process, improving cutting quality and efficiency. When actual cutting data reflects a downward trend in cutting quality, the system can automatically adjust the cutter pressure or cutting speed to ensure the stability and consistency of the cutting process. Such dynamic adjustment can reduce resource waste and maintenance costs, while improving production efficiency, enabling the high-speed cutting machine to maintain efficient cutting performance in changing working environments.

[0079] Specifically, the multi-dimensional actual cutting data collected from the sensors can be integrated into a state vector, which includes various indicators of the current cutting state.

[0080] In this step, the multi-dimensional cutting data (such as pressure, acceleration, acoustic, and optical data) collected from the embedded sensor system described above is integrated into a comprehensive state vector. This state vector not only includes real-time data from each sensor, but also integrates other key indicators of the cutting process, such as the current cutter wear, cutting speed, and temperature. Through this integration, the current cutting state can be fully reflected, providing a basis for subsequent control decisions. The significance of integrating multi-dimensional data into a state vector is to ensure that all parameters in the cutting process can be observed from a unified perspective. This comprehensive state representation not only enhances the understanding of the cutting process, but also provides accurate information support for subsequent adaptive control, making the adjustment of cutting parameters more scientific and precise, thereby improving the stability and reliability of cutting quality.

[0081] In this step, a data acquisition system needs to be designed first to ensure that the data collected from multiple embedded sensors can be effectively integrated. Each sensor will collect different types of cutting parameters, such as the tool pressure provided by the pressure sensor, the motion acceleration recorded by the acceleration sensor, the noise data of the acoustic sensor, and the spectral changes of the optical sensor, etc. These data will be transmitted in real time to a central processing unit at a high frequency, which needs to process and filter these data in real time to ensure the accuracy and integrity of the data.

[0082] Then, based on these real-time collected data, a state vector needs to be constructed, which should contain all the key feature indicators of the current cutting state. This may include cutting speed, tool pressure, tool motion acceleration, noise level during cutting, and spectral monitoring changes, etc. To ensure the effectiveness of the state vector, standardization or normalization processing can be used to make the values of different data sources on the same order of magnitude, which is convenient for subsequent analysis. The construction of this comprehensive state vector lays the foundation for the subsequent analysis of optimizing cutting parameters.

[0083] Finally, this state vector will be updated in real time to reflect the current cutting conditions. Through the integration algorithm, it can quickly adapt to various dynamic changes in the cutting process, ensuring that the state vector can timely reflect the possible abnormalities of cutting quality when it is updated.

[0084] By formulating control rules based on fuzzy logic, the adjustment strategy of cutting parameters is defined, and based on the current state vector, the fuzzy logic controller automatically evaluates the cutting parameters that need to be adjusted to optimize the cutting quality;

[0085] In this step, the fuzzy logic controller automatically evaluates the adjustment needs of cutting parameters based on the state vector integrated from sensors according to the pre-set control rules. These control rules are based on the experience and expert judgment in the cutting process, covering the adjustment strategies of cutting parameters under different cutting conditions, and can handle uncertainty and fuzziness. Through fuzzy reasoning, the controller can determine the adjustment amplitude and direction of parameters such as tool pressure, speed, and angle. The significance of formulating control rules based on fuzzy logic is that the system can still make reasonable cutting parameter adjustment decisions in the face of uncertain environments. Through fuzzy logic, the complexity of the cutting process can be effectively reduced, and problems caused by different conditions can be dynamically addressed, thereby optimizing cutting quality, reducing errors and failures, and improving the efficiency and reliability of the cutting process.

[0086] In this step, a set of control rules based on fuzzy logic needs to be developed first. These control rules should take into account the relationship between various cutting parameters and cutting quality. For example, through analysis of historical data, a rule set can be created that defines thresholds at which cutting quality is affected when tool pressure is too high, cutting speed is too fast, or noise generated is above normal levels. These rules are usually expressed in the form of "if... then..." and embody the logical relationship of cutting parameter adjustments.

[0087] Next, the fuzzy logic controller will use the data in the current state vector to assess the deviation between the current cutting conditions and the preset standards. The fuzzy logic controller can convert the input real values into fuzzy sets and make inferences according to the set rules to determine which cutting parameters need to be adjusted. In this process, different weight values can be set to reflect the importance of different parameters to cutting quality, thereby optimizing the decision-making of the controller.

[0088] Finally, based on the output of the fuzzy logic controller, a set of adjustment suggestions for cutting parameters will be generated. These suggestions will be presented in specific numerical form, such as suggesting a 2% reduction in tool pressure or a 5% increase in cutting speed. This process ensures that in a dynamic cutting environment, appropriate adjustments can be made based on real-time data feedback to optimize cutting quality.

[0089] Combine actual cutting data with historical cutting quality data to predict the performance of cutting quality under current cutting parameter settings using machine learning algorithms;

[0090] In this step, the system combines real-time actual cutting data obtained from sensors with historical cutting quality data to perform data analysis and model training through machine learning algorithms. By establishing a prediction model, the system can identify the relationship between cutting parameters and cutting quality and predict the performance of cutting quality under the current cutting parameter settings. Such predictions provide an important basis for subsequent cutting parameter adjustments. The significance of combining real-time data with historical data lies in that the system can make decisions based on the current state and also draw on past experience to improve the accuracy of predictions. Through the application of machine learning, potential problems in the cutting process can be identified in advance, allowing necessary adjustments to be made before cutting, ensuring that high cutting quality and production efficiency are maintained at all times.

[0091] In this step, valuable information needs to be extracted from historical cutting quality data first. These historical data can include correlation analysis between different paper types, cutting parameter settings, and cutting results. Therefore, historical data needs to be preprocessed, including data cleaning, missing value processing, and feature selection, so that the data can be effectively used for training of machine learning algorithms.

[0092] Next, the current actual cutting data is compared with historical cutting data, and a prediction model is built based on the machine learning algorithm used. Algorithms such as regression analysis, random forest, or support vector machine can be selected, with the current cutting parameters as input features, and the trained model outputs the cutting quality prediction value under the current cutting settings. At this time, the model will learn the patterns in the historical data, so as to more accurately evaluate the cutting quality performance under the current parameter settings.

[0093] Finally, based on the output of the machine learning model, the system can timely identify potential cutting quality problems. If the predicted cutting quality performance does not meet expectations, the system will immediately issue an alarm and provide a basis for subsequent automatic adjustment of cutting parameters. This process not only improves the timeliness and accuracy of cutting quality monitoring, but also provides data support for the implementation of adaptive control technology.

[0094] According to the output of the fuzzy logic controller and the machine learning model, the cutting parameters of the high-speed paper cutter are adjusted in real time, and when the cutting quality is detected to be inconsistent with the set standard, the tool pressure or cutting speed is automatically adjusted to maintain the consistency and accuracy of the cutting.

[0095] In this step, the system integrates the output of the fuzzy logic controller and the machine learning model to adjust the cutting parameters in real time. When the cutting quality monitored by the sensor does not meet the set standard, the system will automatically adjust the tool pressure or cutting speed according to the adjustment suggestion provided by the fuzzy logic controller, combined with the prediction results of the machine learning model, to ensure that the cutting process is always in an ideal state. The significance of this process is to ensure that the high-speed paper cutter can adapt to different cutting conditions and paper characteristics through real-time feedback and adjustment mechanism, and always maintain the consistency and accuracy of the cutting quality. The automatic adjustment mechanism reduces the need for human intervention, improves production efficiency, and reduces material waste caused by cutting quality fluctuations. In addition, through continuous adjustment and optimization, the service life of the tool can be extended, maintenance costs can be reduced, and overall production efficiency can be improved.

[0096] In this step, the system will adjust the cutting parameters in real time based on the output of the fuzzy logic controller and the machine learning model. First, the suggestions of the fuzzy logic controller and the prediction results of the machine learning model need to be integrated to develop a comprehensive parameter adjustment plan. For example, if the fuzzy logic controller suggests reducing tool pressure and the machine learning model shows a downward trend in cutting quality, the system can process these information comprehensively.

[0097] Subsequently, these parameter adjustments are rapidly executed through the control system. Specifically, the system will automatically adjust the tool pressure, cutting speed, and other relevant parameters to ensure the stability and consistency of the cutting process. This adjustment process must have a high response speed to respond to rapid changes in cutting conditions. A certain adjustment range can be set to avoid cutting abnormalities caused by excessive adjustment.

[0098] Finally, to confirm the adjustment effect, the system also needs to continuously monitor the cutting quality and evaluate the actual results under the new cutting parameters. This feedback mechanism ensures the effectiveness of the adjustment measures and the continuous improvement of cutting quality. If the cutting quality after adjustment still does not meet the standards, the system will restart the evaluation and adjustment process, achieving highly automated and intelligent cutting management.

[0099] It can be seen that the paper sheet characteristic data of the paper to be cut is obtained in real time, the paper sheet characteristic data is analyzed through a deep learning model to generate a corresponding optimal cutting parameter set, wherein the paper sheet characteristic data includes thickness, humidity, and material composition data; during the cutting process of the high-speed paper cutting machine, embedded sensors are used to monitor actual cutting data in the cutting process in real time; based on the actual cutting data, adaptive control technology is used to dynamically adjust the cutting parameters of the high-speed paper cutting machine to ensure the consistency and accuracy of the cutting quality, thereby achieving intelligent and automated cutting process and ensuring the consistency and accuracy of the cutting quality.

[0100] Another embodiment of the present application provides a high-speed paper cutting machine control system, as shown in Figure 3 , the system can include:

[0101] The analysis module 301 is configured to obtain paper sheet characteristic data of the paper to be cut in real time, analyze the paper sheet characteristic data through a deep learning model, and generate a corresponding optimal cutting parameter set, wherein the paper sheet characteristic data includes thickness, humidity, and material composition data.

[0102] The monitoring module 302 is configured to monitor actual cutting data in the cutting process of the high-speed paper cutting machine through embedded sensors in real time.

[0103] The control module 303 is configured to dynamically adjust the cutting parameters of the high-speed paper cutting machine based on the actual cutting data using adaptive control technology to ensure the consistency and accuracy of the cutting quality.

[0104] It can be seen that the paper characteristic data of the paper to be cut is acquired in real time, the paper characteristic data is analyzed through a deep learning model, and a corresponding optimal cutting parameter set is generated, wherein the paper characteristic data includes thickness, humidity and material composition data; in the cutting process of the high-speed cutting machine, actual cutting data in the cutting process is monitored in real time through an embedded sensor; and the cutting parameters of the high-speed cutting machine are dynamically adjusted according to the actual cutting data by using an adaptive control technology, so as to ensure the consistency and accuracy of the cutting quality, thereby realizing the intelligentization and automation of the cutting process and ensuring the consistency and accuracy of the cutting quality.

[0105] The embodiment of the present application also provides a storage medium, wherein the storage medium stores a computer program, and the computer program is arranged to execute the steps in any of the method embodiments.

[0106] Specifically, in the embodiment, the storage medium can be arranged to store a computer program for executing the following steps:

[0107] S201, paper characteristic data of the paper to be cut is acquired in real time, the paper characteristic data is analyzed through a deep learning model, and a corresponding optimal cutting parameter set is generated, wherein the paper characteristic data includes thickness, humidity and material composition data;

[0108] S202, in the cutting process of the high-speed cutting machine, actual cutting data in the cutting process is monitored in real time through an embedded sensor;

[0109] S203, the cutting parameters of the high-speed cutting machine are dynamically adjusted according to the actual cutting data by using an adaptive control technology, so as to ensure the consistency and accuracy of the cutting quality.

[0110] It can be seen that the paper characteristic data of the paper to be cut is acquired in real time, the paper characteristic data is analyzed through a deep learning model, and a corresponding optimal cutting parameter set is generated, wherein the paper characteristic data includes thickness, humidity and material composition data; in the cutting process of the high-speed cutting machine, actual cutting data in the cutting process is monitored in real time through an embedded sensor; and the cutting parameters of the high-speed cutting machine are dynamically adjusted according to the actual cutting data by using an adaptive control technology, so as to ensure the consistency and accuracy of the cutting quality, thereby realizing the intelligentization and automation of the cutting process and ensuring the consistency and accuracy of the cutting quality.

[0111] The embodiment of the present application also provides an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor is arranged to execute the computer program to execute the steps in any of the method embodiments.

[0112] Specifically, the electronic device can further include a transmission device connected to the processor and an input / output device connected to the processor.

[0113] Specifically, in the embodiment, the processor can be configured to execute the following steps by a computer program:

[0114] S201, real-time acquisition of paper characteristic data of paper to be cut, analysis of the paper characteristic data by a deep learning model, and generation of a corresponding optimal cutting parameter set, wherein the paper characteristic data includes thickness, humidity, and material composition data;

[0115] S202, real-time monitoring of actual cutting data in the cutting process by an embedded sensor during the cutting process of the high-speed cutting machine;

[0116] S203, dynamic adjustment of the cutting parameters of the high-speed cutting machine according to the actual cutting data by using adaptive control technology to ensure the consistency and accuracy of the cutting quality.

[0117] It can be seen that the paper characteristic data of the paper to be cut is acquired in real time, the paper characteristic data is analyzed by a deep learning model, and a corresponding optimal cutting parameter set is generated, wherein the paper characteristic data includes thickness, humidity, and material composition data; actual cutting data in the cutting process is monitored in real time by an embedded sensor during the cutting process of the high-speed cutting machine; and the cutting parameters of the high-speed cutting machine are dynamically adjusted according to the actual cutting data by using adaptive control technology to ensure the consistency and accuracy of the cutting quality, so that intelligentization and automation of the cutting process can be realized, and the consistency and accuracy of the cutting quality can be ensured.

[0118] The above embodiments according to the drawings illustrate the structure, features, and effects of the present application. The above description is only a preferred embodiment of the present application, but the present application is not limited by the drawings. Any changes or modifications made in accordance with the concept of the present application, or equivalent embodiments with equivalent changes, are still within the scope of the present application.

Claims

1. A control method for a high-speed paper cutter, characterized in that, The method includes: The paper characteristic data of the paper to be cut is acquired in real time, and the paper characteristic data is analyzed by a deep learning model to generate the corresponding optimal cutting parameter set. The paper characteristic data includes thickness, humidity and material composition data. During the cutting process of the high-speed paper cutter, embedded sensors monitor the actual cutting data in real time. Based on the actual cutting data, the cutting parameters of the high-speed paper cutter are dynamically adjusted using adaptive control technology to ensure the consistency and accuracy of cutting quality. The process involves acquiring real-time paper characteristic data of the paper to be cut, analyzing the paper characteristic data using a deep learning model, and generating a corresponding optimal cutting parameter set. The paper characteristic data includes thickness, moisture content, and material composition data, including: A multimodal deep learning model is used to fuse and analyze thickness, humidity, and material composition data. Thickness data in the paper characteristic data is used as a one-dimensional time series feature and input into a sub-model based on a convolutional neural network or a recurrent neural network to capture the variation pattern of paper thickness. Humidity data and material composition data are used as key static features and are processed by a fully connected neural network to extract the physical property influence of each material component under different humidity conditions. Finally, a feature fusion layer is used to fuse the data output by the sub-model to generate a high-dimensional vector that comprehensively represents the paper characteristics. The high-dimensional vector is input into the last fully connected layer of the multimodal deep learning model. Through weight mapping and nonlinear activation functions, a corresponding multidimensional parameter vector is generated as the optimal cutting parameter set. The optimal cutting parameter set includes at least cutting speed, tool pressure, and tool angle to achieve stable cutting quality, minimize tool wear, and maximize cutting efficiency.

2. The method according to claim 1, characterized in that, During the cutting process of the high-speed paper cutter, the actual cutting data during the cutting process is monitored in real time by embedded sensors, including: During the cutting process, a distributed embedded sensor system is used, deploying multiple types of sensors. Data from these sensors is collected in real-time using a high-frequency sampling method. A pressure sensor is installed on the cutter carrier to collect real-time changes in the pressure exerted by the cutter on the paper; an acceleration sensor is installed along the cutter's motion path to capture the cutter's acceleration and analyze the physical vibration behavior during the cutting process. Acoustic sensors are installed around the cutting area to record high-frequency noise generated during the cutting process, which can be used to determine friction or cutting abnormalities. Optical sensors monitor the microscopic spectral changes in the cutting area in real time to capture the minute physical or chemical changes in paper caused by friction during the cutting process.

3. The method according to claim 2, characterized in that, The step of dynamically adjusting the cutting parameters of the high-speed paper cutter using adaptive control technology based on the actual cutting data to ensure the consistency and accuracy of cutting quality includes: The multidimensional actual cutting data collected from the sensor is integrated into a state vector, which contains various indicators of the current cutting state; By formulating control rules based on fuzzy logic and defining adjustment strategies for cutting parameters, the fuzzy logic controller automatically evaluates the cutting parameters that need to be adjusted based on the current state vector in order to optimize the cutting quality. By combining actual cutting data with historical cutting quality data, machine learning algorithms are used to predict the cutting quality performance under the current cutting parameter settings. Based on the output of the fuzzy logic controller and the machine learning model, the cutting parameters of the high-speed paper cutter are adjusted in real time. When the cutting quality is detected to be inconsistent with the set standard, the blade pressure or cutting speed is automatically adjusted to maintain the consistency and accuracy of the cutting.

4. A high-speed paper cutter control system, characterized in that, The system includes: The analysis module is used to acquire paper characteristic data of the paper to be cut in real time, analyze the paper characteristic data through a deep learning model, and generate the corresponding optimal cutting parameter set. The paper characteristic data includes thickness, humidity and material composition data. The monitoring module is used to monitor the actual cutting data in real time during the cutting process of a high-speed paper cutter using embedded sensors. The control module is used to dynamically adjust the cutting parameters of the high-speed paper cutter based on the actual cutting data using adaptive control technology, so as to ensure the consistency and accuracy of the cutting quality; The analysis module is specifically used for: A multimodal deep learning model is used to fuse and analyze thickness, humidity, and material composition data. Thickness data in the paper characteristic data is used as a one-dimensional time series feature and input into a sub-model based on a convolutional neural network or a recurrent neural network to capture the variation pattern of paper thickness. Humidity data and material composition data are used as key static features and are processed by a fully connected neural network to extract the physical property influence of each material component under different humidity conditions. Finally, a feature fusion layer is used to fuse the data output by the sub-model to generate a high-dimensional vector that comprehensively represents the paper characteristics. The high-dimensional vector is input into the last fully connected layer of the multimodal deep learning model. Through weight mapping and nonlinear activation functions, a corresponding multidimensional parameter vector is generated as the optimal cutting parameter set. The optimal cutting parameter set includes at least cutting speed, tool pressure, and tool angle to achieve stable cutting quality, minimize tool wear, and maximize cutting efficiency.

5. The system according to claim 4, characterized in that, The monitoring module is specifically used for: During the cutting process, a distributed embedded sensor system is used, deploying multiple types of sensors. Data from these sensors is collected in real-time using a high-frequency sampling method. A pressure sensor is installed on the cutter carrier to collect real-time changes in the pressure exerted by the cutter on the paper; an acceleration sensor is installed along the cutter's motion path to capture the cutter's acceleration and analyze the physical vibration behavior during the cutting process. Acoustic sensors are installed around the cutting area to record high-frequency noise generated during the cutting process, which can be used to determine friction or cutting abnormalities. Optical sensors monitor the microscopic spectral changes in the cutting area in real time to capture the minute physical or chemical changes in paper caused by friction during the cutting process.

6. The system according to claim 5, characterized in that, The control module is specifically used for: The multidimensional actual cutting data collected from the sensor is integrated into a state vector, which contains various indicators of the current cutting state; By formulating control rules based on fuzzy logic and defining adjustment strategies for cutting parameters, the fuzzy logic controller automatically evaluates the cutting parameters that need to be adjusted based on the current state vector in order to optimize the cutting quality. By combining actual cutting data with historical cutting quality data, machine learning algorithms are used to predict the cutting quality performance under the current cutting parameter settings. Based on the output of the fuzzy logic controller and the machine learning model, the cutting parameters of the high-speed paper cutter are adjusted in real time. When the cutting quality is detected to be inconsistent with the set standard, the blade pressure or cutting speed is automatically adjusted to maintain the consistency and accuracy of the cutting.

7. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to execute the method of any one of claims 1-3 when it is run.

8. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the method of any one of claims 1-3.

Images