Disc shear blade spacing control method

Abstract

The invention discloses a disc shear blade spacing control method which comprises the following steps: S1, monitoring sound wave signals generated in a shearing process in real time, collecting sound frequency and intensity information generated when shear blades are in contact with a material through the sound wave signals, and indirectly speculating the actual spacing of the shear blades; s2, collecting historical data of the shear blade in different working states, and constructing a deep learning model; s3, a control system is introduced and used for monitoring the distance between the shear blades and adjusting working parameters in real time, and the distance between the shear blades is dynamically adjusted; s4, enabling the control system to have a self-learning function, and performing dynamic adjustment according to a new working environment and a shear blade state so as to cope with changes under different working conditions; and S5, carrying out a performance verification experiment, and testing the feedback accuracy of the acoustic sensor, the prediction precision of the deep learning model and the response speed of the feedback control system. The shearing efficiency, precision and equipment stability can be improved, and abrasion and energy consumption are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of blade spacing control, and particularly to a method for controlling the blade spacing of a disc shear. Background Art

[0002] In modern industrial cutting equipment, disc shear blades are widely used in the cutting process of various materials, such as wood, metal, plastic, etc. However, in the actual cutting process, the contact state between the shear blades and the material is crucial for the cutting effect and the long-term stability of the equipment. The spacing between the shear blades directly affects the cutting accuracy, cutting efficiency, and the wear rate of the equipment. If the spacing is adjusted improperly, it may lead to uneven cutting, excessive wear, and energy consumption waste, thus affecting the product quality and the service life of the equipment. Most of the existing shear blade spacing adjustment technologies rely on manual settings or control systems based on fixed rules, unable to respond to the changes in the working environment in real time, and not fully considering factors such as shear blade wear, load changes, and material characteristics. Therefore, the existing technologies have problems such as insufficient adjustment accuracy, slow response, and poor adaptability.

[0003] In order to improve the cutting quality and extend the service life of the equipment, more and more research has begun to explore how to optimize the shear blade spacing control through intelligent means. Existing technologies have also tried to combine sensor data to detect the contact state between the shear blades and the material, but these methods often rely on a single signal source and mostly lack real-time learning and adaptive functions, unable to effectively cope with complex environments and working condition changes. Summary of the Invention

[0004] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] To solve the above technical problems, the present invention provides the following technical solution: A method for controlling the blade spacing of a disc shear, comprising the following steps:

[0006] A method for controlling the blade spacing of a disc shear, characterized by comprising the steps:

[0007] S1: Arrange a plurality of acoustic sensors in the contact area between the shear blades and the cutting material, and monitor the acoustic wave signals generated during the shearing process in real time. Collect the sound frequency and intensity information generated when the shear blades contact the material through the acoustic wave signals, and these information reflect the contact situation between the shear blades and the material; Use the frequency analysis algorithm to process the acoustic wave signals, extract the contact frequency pattern between the shear blades and the material, and indirectly infer the actual spacing of the shear blades;

[0008] S2: Collect historical data of the scissor blades under different working conditions, construct a deep learning model, and then input this historical data into the model for training, enabling the model to learn and predict the changing trend of the scissor blade spacing. During the training process, optimize the model parameters through the cross-validation method to improve the generalization ability of the model, ensuring that it can handle changes under different working conditions;

[0009] S3: Introduce a control system for real-time monitoring of the scissor blade spacing and adjustment of working parameters. In the control system, fuse the data collected by the acoustic sensor with the prediction results of the deep learning model and the material property data, thereby achieving real-time monitoring of the actual scissor blade spacing. Through the feedback control system, dynamically adjust the scissor blade spacing based on the acoustic signal and the prediction results of the deep learning model;

[0010] S4: Enable the control system to have a self-learning function. During the use of the scissor blades, the control system optimizes the deep learning model on its own according to the dynamic factors of the scissor blades by collecting new data, and adjusts the deep learning model parameters through an online learning mechanism, thereby improving the prediction accuracy of the scissor blade spacing change, enabling the control system to dynamically adjust according to the new working environment and scissor blade state to cope with changes under different working conditions, and ensuring automatic adjustment of the control strategy when the environment changes, always providing the optimal scissor blade distance;

[0011] S5: Conduct performance verification experiments to test the accuracy of the acoustic sensor feedback, the prediction accuracy of the deep learning model, and the response speed of the feedback control system, and evaluate the effect of the scissor blade spacing control method through experimental data, further ensuring the stability and efficiency of the control system. At the same time, conduct system debugging in different types of shearing tasks to ensure that this method can provide stable and accurate spacing control under various loads and material cutting conditions, and continuously optimize the system performance, ultimately achieving precise and efficient scissor blade spacing control.

[0012] As a preferred solution of the disc scissor blade spacing control method described in the present invention, wherein: the historical data in S2 includes the acoustic sensor feedback, scissor blade spacing, cutting load, material type information, and environmental factors. The environmental factors include temperature, humidity, and air pressure. After preprocessing and feature extraction, these historical data serve as the basis for training the deep learning model to improve the prediction accuracy of the model.

[0013] As a preferred solution of the disc scissor blade spacing control method described in the present invention, wherein: the acoustic sensor in S1 adopts a piezoelectric sensor or a micro-microphone array sensor.

[0014] As a preferred solution of the disc scissor blade spacing control method described in the present invention, wherein: the dynamic factors in S4 include the actual wear of the scissor blades, changes in the working load, and changes in the environmental conditions.

[0015] As a preferred solution of the disc shear blade spacing control method described in the present invention, wherein: constructing the deep learning model in S2 includes the following steps:

[0016] S21: By collecting historical data of the shear blades in different working states, including acoustic signal characteristics, environmental variables, and material characteristics, and cleaning, annotating, normalizing, and extracting features from the data, ensure to provide high-quality and representative data input for the deep learning model;

[0017] S22: Based on the temporal characteristics and feature requirements of the data, select a suitable deep learning model architecture, such as LSTM, CNN, or a CNN+LSTM hybrid model, so as to capture temporal dependencies and extract spatial features while meeting the task requirements of shear blade spacing prediction;

[0018] S23: Divide the collected data into a training set, a validation set, and a test set, train the model with the training set and perform cross-validation, tune the hyperparameters, ensure that the model has strong generalization ability on different data sets, and avoid overfitting or underfitting;

[0019] S24: Use the validation set and the test set to evaluate the model, adjust the model based on error metrics such as MSE and MAE, and further improve the prediction accuracy and robustness of the model through optimization algorithms such as regularization and Dropout;

[0020] S25: Integrate the optimized deep learning model into the control system, dynamically adjust the shear blade spacing in real time according to the feedback of the acoustic sensor, and at the same time ensure that the system can continuously optimize the model through online learning to adapt to different working environments and shear blade states;

[0021] S26: Regularly collect new data and update the model to cope with environmental changes and working conditions changes, ensure that the deep learning model maintains high-efficiency performance during long-term use, and maintain its stability and accuracy in practical applications through monitoring and adjustment.

[0022] As a preferred solution of the disc shear blade spacing control method described in the present invention, wherein: fusing the data collected by the acoustic sensor with the prediction results of the deep learning model and the material property data in the control system in S3 includes the following steps:

[0023] S31: First, preprocess the original signal captured by the acoustic sensor to ensure its synchronization with the input data of the deep learning model in terms of time. The preprocessing includes signal denoising, signal normalization, time synchronization, and filtering to remove noise and interference and ensure data quality. Meanwhile, preprocess the material characteristic data and standardize these data for effective combination with the acoustic signal and the deep learning prediction results.

[0024] S32: In the control system, the acoustic sensor data, the prediction results of the deep learning model, and the material characteristic data achieve more precise control of the scissor blade spacing through data fusion technology. Specifically, the material characteristic data will act jointly with the acoustic signal and the prediction values of the deep learning model during the fusion process. In this way, the control system can not only adjust the scissor blade spacing in real time but also dynamically adjust the scissor blade spacing control strategy according to the actual physical characteristics of the material.

[0025] S33: Take the fused scissor blade spacing value as the control signal and transmit it to the actuator module (such as an electric drive or a pneumatic cylinder) in the control system. Then, the control system calculates the error between the ideal spacing and the actual spacing based on the fused signal and adjusts it through a PID controller to precisely adjust the scissor blade spacing. The material characteristic data will affect the parameter adjustment of the controller during this process to ensure the precise adjustment of the scissor blade spacing under different material conditions.

[0026] S34: To cope with the changing working environment and scissor blade state, the control system continuously collects new acoustic data and material characteristic data through an online learning mechanism and updates the deep learning model. During the model adjustment process, it automatically optimizes the data fusion strategy and combines an adaptive fusion strategy to dynamically adjust the weight value or filtering coefficient in the data fusion method as the working environment changes. The change in material characteristics will affect the adjustment of the weight value or filtering coefficient to ensure the accurate combination of the acoustic signal, the deep learning model prediction results, and the material characteristic data, thereby achieving precise control of the scissor blade spacing.

[0027] As a preferred embodiment of the method for controlling the scissor blade spacing of the present invention, wherein: the automatic adjustment of the control strategy when the environment changes in S4 includes the following steps:

[0028] S41: Input the historical data and the sound wave signals collected in real time into the deep learning model. Combine the sensor data and environmental factors to train the model so that it can automatically identify the material type and infer the actual contact situation between the scissor blade and the material based on the physical characteristics of the material, thereby dynamically adjusting the scissor blade spacing.

[0029] S42: The deep learning model automatically adjusts the control strategy by analyzing the physical characteristics of the material and combining the frequency patterns reflected by the acoustic sensor.

[0030] S43: When the environmental conditions change, the deep learning model automatically identifies the impact of these environmental factors on the control of the scissor blade spacing by obtaining the data of the environmental sensors in real time and combining the acoustic signals and material characteristics;

[0031] S44: According to the environmental and material characteristics, the control system automatically updates the scissor blade spacing control strategy by adjusting the data fusion strategy;

[0032] S45: The control system optimizes the deep learning model by continuously collecting and analyzing new data, and dynamically adjusts the control strategy according to the new working environment and the state of the scissor blades.

[0033] As a preferred solution of the disc scissor blade spacing control method described in the present invention, wherein: the dynamic adjustment formula of the scissor blade spacing is as follows:

[0034]

[0035] Wherein:

[0036] D(t): represents the actual spacing of the scissor blades at time t, in millimeters (mm), which is the output value of the formula and is the scissor blade spacing that needs to be dynamically adjusted;

[0037] S(t): the acoustic signal intensity, representing the intensity information of the acoustic wave signal, which is the signal captured by the acoustic sensor and reflects the acoustic intensity when the scissor blades contact the material;

[0038] The predicted value of the scissor blade spacing by the deep learning model, representing the predicted scissor blade spacing value at time t through the deep learning model;

[0039] T(t): the environmental temperature, used to consider the influence of temperature on the scissor blade spacing;

[0040] α: the influence coefficient of temperature on the adjustment of the scissor blade spacing;

[0041] γ: the total influence coefficient of environmental factors, reflecting the comprehensive influence of multiple factors on the adjustment of the scissor blade spacing;

[0042] e -βT(t) : the exponential decay function, representing the influence of environmental temperature on the scissor blade spacing, e is the base of the natural logarithm, β is the temperature decay coefficient, used to control the influence of temperature on the adjustment of the scissor blade spacing, T(t) represents the current temperature, T represents the temperature, and t represents the time point;

[0043] (1 + αe -βT(t) ): represents the part of the influence of temperature on the scissor blade spacing. First, calculate the exponential decay part e -βT(t) , then add 1, and finally multiply by other parts.

[0044] Consider the influence of different material types on the spacing of the scissors blades, δ i The weight coefficient representing the material characteristics, η i Represents the actual value of the characteristic information of the i-th material at the current moment t.

[0045] As a preferred embodiment of the method for controlling the spacing of the disc scissors blades according to the present invention, wherein: the temperature influence coefficient α in the dynamic adjustment formula is dynamically updated through the online learning mechanism of the deep learning model, and is adaptively adjusted according to the temperature, humidity and wear degree of the scissors blades monitored in real time.

[0046] As a preferred embodiment of the method for controlling the spacing of the disc scissors blades according to the present invention, wherein: the control system includes:

[0047] An acoustic signal processing module, responsible for collecting acoustic wave signals from acoustic sensors and preprocessing them to ensure data quality;

[0048] A deep learning model module, used for predicting and dynamically adjusting the spacing of the scissors blades based on historical data and real-time sensor data;

[0049] A material identification module, which detects the physical characteristics of the material in real time by integrating multiple sensors, classifies and identifies the material through a deep learning model, and inputs the identified material information into the control system for subsequent adjustment of the scissors blade spacing;

[0050] A data fusion module, used for fusing the output data of the acoustic signal processing module, the deep learning model module and the material identification module to more accurately control the scissors blade spacing;

[0051] An actuator module, capable of adjusting the actual spacing of the scissors blades according to the fused scissors blade spacing control signal;

[0052] An environment adaptation module, which automatically adjusts the control strategy according to the environmental factors monitored in real time to ensure the optimization of the scissors blade spacing under different environmental conditions;

[0053] A feedback and self-learning module, which collects new data, analyzes the dynamic factors such as scissors blade wear and load change, and optimizes the deep learning model and control strategy, so as to achieve self-adjustment and optimization.

[0054] The beneficial effects of the present invention:

[0055] 1. By combining an acoustic sensor with a deep learning model, the present invention can monitor the contact state between the scissor blades and the material in real time and dynamically adjust the distance between the scissor blades. This method can accurately adjust the distance between the scissor blades according to the wear of the scissor blades, load changes, material properties, and environmental conditions (such as temperature, humidity, etc.), thereby optimizing the cutting effect and reducing unnecessary wear and energy consumption. This intelligent control method ensures the best shearing performance under different working conditions, greatly improving the efficiency and stability of the equipment.

[0056] 2. In the present invention, the control system collects new working data through an online learning mechanism and dynamically adjusts the deep learning model. This not only enables the control system to adapt to the wear of the scissor blades, but also automatically adjusts the control strategy for the distance between the scissor blades according to factors such as the type of material and environmental changes, thereby ensuring the stability and efficiency of the control system during long-term use.

[0057] 3. The present invention combines the acoustic sensor signal with the prediction result of the deep learning model through data fusion technology to ensure high-precision adjustment of the distance between the scissor blades. In addition, by incorporating environmental factors (such as temperature, humidity, air pressure, etc.) and material properties (such as hardness, density) into the control system, the deep learning model can dynamically adjust the distance between the scissor blades according to real-time data. This fusion strategy significantly improves the adaptability of the control system to complex working environments, effectively improving the shearing accuracy and equipment life, and reducing performance fluctuations of the equipment caused by environmental changes or material inconsistencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them:

[0059] Figure 1 is a flowchart of a method for controlling the distance between the disc scissor blades of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0060] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification.

[0061] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0062] Secondly, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an individual or selectively mutually exclusive embodiment with other embodiments.

[0063] Thirdly, the present invention is described in detail in conjunction with schematic diagrams. When detailing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0064] Embodiment 1

[0065] Referring to Figure 1 , the first embodiment of the present invention provides a method for controlling the spacing of disc cutting blades, including the following steps:

[0066] S1: Arrange a plurality of acoustic sensors in the contact area between the cutting blades and the cutting material, and monitor the acoustic wave signals generated during the shearing process in real time. Collect the sound frequency and intensity information generated when the cutting blades contact the material through the acoustic wave signals. These information reflect the contact situation between the cutting blades and the material; use a frequency analysis algorithm (such as the fast Fourier transform FFT) to process the acoustic wave signals, extract the contact frequency pattern between the cutting blades and the material, and indirectly infer the actual spacing of the cutting blades.

[0067] Among them, the acoustic sensors adopt piezoelectric sensors or micro-microphone array sensors. Specifically, when in use, when the cutting blades contact the material, specific frequency and intensity acoustic wave signals will be generated during the shearing process. These acoustic wave signals can reflect the contact situation between the cutting blades and the material. The frequency and intensity of the acoustic wave signals are closely related to factors such as the actual spacing of the cutting blades, the hardness of the material, and the cutting load.

[0068] This step provides accurate real-time data input for the entire control system, ensuring that the subsequent deep learning model can make effective predictions based on the acoustic signals.

[0069] S2: Collect historical data of the cutting blades in different working states, construct a deep learning model, and then input these historical data into the model for training, so that the model can learn and predict the change trend of the cutting blade spacing. During the training process, optimize the model parameters through the cross-validation method to improve the generalization ability of the model to ensure that it can cope with changes in different working states.

[0070] Among them, the historical data includes acoustic sensor feedback, scissor blade spacing, cutting load, material type information, and environmental factors. The environmental factors include temperature, humidity, and air pressure. After preprocessing and feature extraction, these historical data serve as the basis for training the deep learning model to improve the prediction accuracy of the model.

[0071] This step efficiently predicts the scissor blade spacing through the deep learning model, improving the intelligence level of the control system.

[0072] S3: Introduce a control system for real-time monitoring of the scissor blade spacing and adjustment of working parameters. In the control system, fuse the data collected by the acoustic sensor with the prediction results of the deep learning model and the material property data, so as to realize real-time monitoring of the actual scissor blade spacing. Through the feedback control system, based on the acoustic signal and the prediction results of the deep learning model, dynamically adjust the scissor blade spacing; for example, according to the state of the scissor blade in contact with the material reflected by the acoustic signal, the system precisely adjusts the scissor blade spacing through an actuator (such as an electric drive or a pneumatic cylinder). At the same time, the control system uses the deep learning model to predict in advance the spacing change trend caused by scissor blade wear, increased load, or material property changes, ensuring timely adjustment of the scissor blade spacing before actual scissor blade wear or load change, thus guaranteeing the cutting quality.

[0073] This step enhances the response speed and accuracy of the control system through the real-time feedback and prediction mechanism.

[0074] S4: Enable the control system to have a self-learning function. During the use of the scissor blade, the control system collects new data and self-optimizes the deep learning model according to the dynamic factors of the scissor blade (actual wear of the scissor blade, change in working load, change in environmental conditions, etc.), and adjusts the parameters of the deep learning model through an online learning mechanism, thereby improving the prediction accuracy of the scissor blade spacing change, enabling the control system to dynamically adjust according to the new working environment and scissor blade state, to cope with changes under different working conditions, and ensuring automatic adjustment of the control strategy when the environment changes (such as temperature, humidity, material type, etc.), always providing the optimal scissor blade distance.

[0075] This step enhances the adaptive ability of the system and can cope with the challenges of different working environments.

[0076] S5: Conduct performance verification experiments to test the accuracy of the acoustic sensor feedback, the prediction accuracy of the deep learning model, and the response speed of the feedback control system. Evaluate the effectiveness of the scissor blade spacing control method through experimental data, such as indicators like cutting accuracy, equipment life, and energy consumption, to further ensure the stability and efficiency of the control system. At the same time, conduct system debugging in different types of shearing tasks to ensure that this method can provide stable and accurate spacing control under various load and material cutting conditions, continuously optimize the system performance, and ultimately achieve precise and efficient scissor blade spacing control.

[0077] This step ensures the practicality and reliability of the system through multiple rounds of experiments and debugging.

[0078] Specifically, the construction of the deep learning model in S2 includes the following steps:

[0079] S21: Collect historical data of the scissor blades in different working states, including acoustic wave signal characteristics, environmental variables, and material characteristics (such as hardness, density, thickness, etc.), and perform data cleaning, annotation, normalization processing, and feature extraction (extract the characteristics of the audio signal, such as frequency, intensity, time-domain and frequency-domain features (such as mean, standard deviation, FFT, etc.)) to ensure high-quality and representative data input for the deep learning model.

[0080] S22: Based on the temporal characteristics and feature requirements of the data, select a suitable deep learning model architecture, such as LSTM, CNN, CNN + LSTM hybrid model, or deep neural network, to meet the task requirements of scissor blade spacing prediction while capturing temporal dependencies and extracting spatial features.

[0081] S23: Divide the collected data into a training set, a validation set, and a test set, usually in a ratio of 70% for the training set, 15% for the validation set, and 15% for the test set. The validation set is used to adjust hyperparameters during the training process, while the test set is used for the final performance evaluation. Train the model using the training set and optimize the model parameters through the backpropagation and gradient descent algorithms. Use cross-validation techniques to optimize the generalization ability of the model through different data splitting methods. During the training process, adjust the model according to the loss function (such as mean squared error MSE) on the training set and the validation set to avoid overfitting or underfitting.

[0082] S24: Use the validation set and the test set to evaluate the model, adjust the model based on error metrics such as MSE and MAE, and further improve the prediction accuracy and robustness of the model through optimization algorithms such as regularization and Dropout.

[0083] S25: Integrate the optimized deep learning model into the control system to predict the changing trend of the scissor blade spacing in real time. Dynamically adjust the scissor blade spacing in real time according to the feedback of the acoustic sensor, and at the same time ensure that the system can continuously optimize the model through online learning to adapt to different working environments and scissor blade states.

[0084] S26: Regularly collect new data and update the model to cope with environmental changes and working conditions changes, ensure that the deep learning model maintains high efficiency during long-term use, and maintain its stability and accuracy in practical applications through monitoring and adjustment.

[0085] Specifically, in S3, the data collected by the acoustic sensor in the control system is fused with the prediction results of the deep learning model and the material property data, including the following steps:

[0086] S31: First, preprocess the original signal captured by the acoustic sensor to ensure its synchronization with the input data of the deep learning model in time. The preprocessing includes:

[0087] Signal denoising: Apply a filtering algorithm (such as a low-pass filter) to remove high-frequency noise and retain the key signals of the contact between the scissor blade and the material.

[0088] Signal normalization: Normalize the sensor data to a specific range to ensure the consistency of different sensor data.

[0089] Time synchronization: Ensure that the signal data of all sensors is consistent with the time step of the model input. Especially in real-time monitoring, the acoustic signal and the prediction results of the deep learning model need to be processed and fused at the same moment.

[0090] At the same time, preprocess the material property data (such as hardness, density, thickness, etc.) and normalize these data for effective combination with the acoustic signal and the deep learning prediction results.

[0091] S32: In the control system, the acoustic sensor data, the prediction results of the deep learning model, and the material property data achieve more accurate control of the scissor blade spacing through data fusion technology (several methods can be adopted: weighted fusion, Kalman filtering, recurrent neural network (RNN) or LSTM). Specifically, the material property data will act together with the acoustic signal and the prediction value of the deep learning model during the fusion process. In this way, the control system can not only adjust the scissor blade spacing in real time, but also dynamically adjust the scissor blade spacing control strategy according to the actual physical properties of the material (for example, hardness, density, thickness, etc.). For example, materials with greater hardness may require a larger scissor blade spacing predicted by the deep learning model, while materials with greater density may require adjustment of the cutting speed.

[0092] S33: Use the fused scissor blade spacing value as a control signal and transmit it to the actuator module (such as an electric drive or pneumatic cylinder) in the control system. Then, the control system calculates the error between the ideal spacing and the actual spacing based on the fused signal, and adjusts it through a PID controller to precisely adjust the spacing of the scissor blades. The material characteristic data will affect the parameter adjustment of the controller during this process to ensure the precise adjustment of the scissor blade spacing under different material conditions.

[0093] S34: To cope with the changing working environment and scissor blade status, the control system continuously collects new acoustic data and material characteristic data through an online learning mechanism and updates the deep learning model. During the model adjustment process, it automatically optimizes the data fusion strategy and combines an adaptive fusion strategy to dynamically adjust the weight values or filtering coefficients in the data fusion method as the working environment (such as temperature, humidity, or material type) changes. The change in material characteristics will affect the adjustment of the weight values or filtering coefficients to ensure the accurate combination of acoustic signals, the prediction results of the deep learning model, and material characteristic data, thereby achieving precise control of the scissor blade spacing.

[0094] Specifically, the automatic adjustment of the control strategy in S4 when the environment changes includes the following steps:

[0095] S41: Input historical data and the acoustic signals collected in real time into the deep learning model, combine sensor data and environmental factors (such as temperature, humidity, etc.), and train the model to enable it to automatically identify the material type and infer the actual contact situation between the scissor blades and the material based on the physical properties of the material (such as hardness, density, thickness, etc.), thereby dynamically adjusting the scissor blade spacing.

[0096] S42: The deep learning model automatically adjusts the control strategy by analyzing the physical properties of the material and combining the frequency patterns reflected by the acoustic sensor. For example, for materials with higher hardness, the model predicts that the load on the scissor blades is larger, and the spacing needs to be appropriately increased to reduce excessive wear; for materials with higher density, the control strategy may need to adjust the cutting speed and spacing of the scissor blades to improve the cutting efficiency.

[0097] S43: When the environmental conditions (such as temperature, humidity, air pressure, etc.) change, the deep learning model automatically identifies the impact of these environmental factors on the scissor blade spacing control by obtaining the data of environmental sensors in real time and combining acoustic signals and material characteristics. For example, an increase in temperature may cause the scissor blade material to expand or deform, and a change in humidity may affect the friction characteristics of the material. The system automatically adjusts the model parameters through an online learning mechanism to dynamically adapt to these environmental changes.

[0098] S44: According to the environmental and material characteristics, the control system automatically updates the control strategy for the scissor blade spacing by adjusting the data fusion strategy. Specifically, the control system uses the actuator module (such as an electric drive or a pneumatic cylinder) to dynamically adjust the spacing of the scissor blades, ensuring that the scissor blades always maintain the best cutting performance under different environmental conditions and material characteristics, while avoiding excessive wear and energy consumption waste.

[0099] S45: The control system optimizes the deep learning model by continuously collecting and analyzing new data, and dynamically adjusts the control strategy according to the new working environment and the state of the scissor blades. Ensure that the system can always provide the optimal scissor blade spacing when the temperature, humidity, material type, etc. change, so as to achieve efficient and stable cutting performance.

[0100] Specifically, the dynamic adjustment formula for the scissor blade spacing is as follows:

[0101]

[0102] Where:

[0103] D(t): Represents the actual spacing of the scissor blades at time t, in millimeters (mm). This is the output value of the formula and is the scissor blade spacing that needs to be dynamically adjusted.

[0104] S(t): Acoustic signal intensity, representing the intensity information of the acoustic wave signal. It is the signal captured by the acoustic sensor and reflects the acoustic intensity when the scissor blades contact the material.

[0105] The predicted value of the scissor blade spacing by the deep learning model, representing the predicted scissor blade spacing value at time t.

[0106] T(t): Environmental temperature (such as real-time data collected from a temperature sensor), used to consider the influence of temperature on the scissor blade spacing, especially when the material expands or deforms.

[0107] α: Influence coefficient of temperature on the adjustment of scissor blade spacing. α is dynamically updated through the online learning mechanism of the deep learning model and is adaptively adjusted according to the real-time monitored temperature, humidity, and the wear degree of the scissor blades. Thus, the accuracy and adaptability of the scissor blade spacing adjustment are improved. The update of the coefficient is based on the correlation between the predicted error output by the model and the environmental change, and is adjusted through the minimum error optimization algorithm.

[0108] γ: Total influence coefficient of environmental factors, reflecting the comprehensive influence of multiple factors (such as humidity, air pressure) on the adjustment of scissor blade spacing.

[0109] e -βT(t): The exponential decay function represents the influence of environmental temperature on the spacing of the scissor blades. Here, e is the base of the natural logarithm, β is the temperature decay coefficient used to control the impact of temperature on the adjustment of the scissor blade spacing, T(t) represents the temperature at the current moment, T represents the temperature, and t represents the time point.

[0110] (1 + αe -βT(t) ): This represents the part of the influence of temperature on the scissor blade spacing. First, calculate the exponential decay part e -βT(t) , then add 1, and finally multiply by other parts.

[0111] Considering the influence of different material types on the scissor blade spacing, δ i represents the weight coefficient of material characteristics (such as hardness, density, etc.), and η i represents the actual value of the characteristic information of the i-th material at the current moment t. δ i and η i are determined in experiments and are usually positive values, reflecting the influence of the actual characteristics of the material on the spacing.

[0112] The above formula dynamically adjusts the scissor blade spacing by comprehensively considering the spacing values predicted by acoustic signals and deep learning models, and takes into account the influence of factors such as environmental temperature and humidity on the scissor blade spacing. The multiple coefficients and data items in the formula can be dynamically adjusted according to the actual working environment and cutting tasks to ensure the best cutting performance of the scissor blades under different conditions, and avoid excessive wear and energy consumption waste. Through this formula, efficient and precise control of the scissor blade spacing can be achieved.

[0113] Specifically, the control system includes:

[0114] The acoustic signal processing module is responsible for collecting acoustic wave signals from acoustic sensors and preprocessing them to ensure data quality;

[0115] The deep learning model module is used to predict and dynamically adjust the scissor blade spacing based on historical data and real-time sensor data;

[0116] The material identification module integrates multiple sensors (such as acoustic sensors, cameras, infrared sensors, etc.) to detect the physical characteristics of materials in real time, such as hardness, density, thickness, etc., and classifies and identifies the materials through a deep learning model. The material identification module inputs the identified material information into the control system for subsequent adjustment of the scissor blade spacing;

[0117] The data fusion module is used to fuse the output data of the acoustic signal processing module, the deep learning model module, and the material identification module to more precisely control the scissor blade spacing;

[0118] An actuator module that can adjust the actual spacing of the scissors blades according to the fused control signal for the scissors blade spacing; the actuator can be an electric drive, a pneumatic cylinder or a hydraulic system, and can accurately control the movement of the scissors blades;

[0119] An environmental adaptation module that automatically adjusts the control strategy according to the real-time monitored environmental factors (such as temperature, humidity, air pressure, etc.) to ensure the optimization of the scissors blade spacing under different environmental conditions;

[0120] A feedback and self-learning module that collects new data, analyzes dynamic factors such as scissors blade wear and load changes, and optimizes the deep learning model and control strategy, so as to achieve self-adjustment and optimization.

[0121] In summary, by combining an acoustic sensor with a deep learning model, the present invention can real-time monitor the contact state between the scissors blades and the material and dynamically adjust the spacing of the scissors blades. This method can accurately adjust the spacing of the scissors blades according to scissors blade wear, load changes, material properties and environmental conditions (such as temperature, humidity, etc.), so as to optimize the cutting effect, reduce unnecessary wear and energy consumption. This intelligent control method ensures the best shearing performance under different working conditions, greatly improving the efficiency and stability of the equipment. In the present invention, the control system collects new working data through an online learning mechanism and dynamically adjusts the deep learning model. This not only enables the control system to adapt to the wear condition of the scissors blades, but also can automatically adjust the control strategy for the scissors blade spacing according to factors such as the type of material and environmental changes, thus ensuring the stability and high efficiency of the control system during long-term use. The present invention combines the acoustic sensor signal with the prediction result of the deep learning model through data fusion technology to ensure the high precision of the scissors blade spacing adjustment. In addition, by incorporating environmental factors (such as temperature, humidity, air pressure, etc.) and material properties (such as hardness, density) into the control system, the deep learning model can dynamically adjust the scissors blade spacing according to real-time data. This fusion strategy significantly improves the adaptability of the control system to complex working environments, effectively improves the shearing precision and the equipment life, and reduces the performance fluctuations of the equipment caused by environmental changes or material inconsistencies.

[0122] Embodiment 2

[0123] Referring to Tables 1 - 3, this is the second embodiment of the present invention. In order to verify the actual effect of the method for controlling the spacing of the disc scissors blades proposed in the present invention, the experimental data and relevant analysis in the actual environment are used to demonstrate its effectiveness. The following are the experimental design and data presentation methods.

[0124] Experimental purpose: To verify the effect of the method for controlling the spacing of the disc scissors blades by combining an acoustic sensor with a deep learning model in the actual working environment, especially its adaptability to different material properties, environmental changes (such as temperature, humidity, etc.) and scissors blade wear conditions.

[0125] Experimental Design

[0126] Experimental Methods:

[0127] Ⅰ. Test the scissor blade control system under different environmental conditions (temperature, humidity, etc.).

[0128] Ⅱ. Measure the influence of different materials (hardness, density, etc.) on the scissor blade spacing.

[0129] Ⅲ. Monitor the wear condition of the scissor blades and test the dynamic adjustment ability of the control system.

[0130] Ⅳ. Adjust the scissor blade spacing in real time through the fusion strategy of acoustic sensors and deep learning models, and record the control effect.

[0131] Experimental Settings:

[0132] Control System: A deep learning-based controller that combines acoustic sensor data, material property data, and environmental data.

[0133] Measurement Metrics: Scissor blade spacing error, shear quality, equipment energy consumption, equipment wear degree.

[0134] Environmental Variables: Temperature (20°C, 30°C, 40°C), Humidity (30%, 50%, 70%), Material Types (materials with different hardness and density).

[0135] Experimental Results Presentation

[0136] 1. Scissor Blade Spacing Error

[0137] Demonstrate the accuracy of the control system by recording the error between the actual scissor blade spacing and the target spacing.

[0138]

[0139] 2. Scissor Blade Spacing Error

[0140] Evaluate the influence of the precise adjustment of the control system on the shear effect according to the shear quality standard.

[0141]

[0142] 3. Energy Consumption and Equipment Wear

[0143] Record the equipment energy consumption and wear condition to verify the energy-saving and durability of the control system.

[0144]

[0145] Result Analysis

[0146] Scissor blade spacing error: The control system can accurately adjust the scissor blade spacing under different environmental conditions, with an error between 0.1 mm and 0.2 mm, demonstrating the high-precision control ability of the system.

[0147] Cutting quality: By adjusting the scissor blade spacing, the cutting quality is maintained at a high level under different working environments. Especially when the material type and wear condition change, the control system can effectively adapt and maintain a high cutting quality.

[0148] Energy consumption and equipment wear: By dynamically adjusting the control strategy of the scissor blade spacing, unnecessary energy consumption can be significantly reduced, and equipment wear can be effectively slowed down, improving the equipment life.

[0149] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for controlling the spacing between disc scissors blades, characterized in that: Includes steps: S1: Multiple acoustic sensors are arranged in the contact area between the scissor blades and the cut material to monitor the acoustic wave signals generated during the shearing process in real time, and the sound frequency and intensity information generated when the scissor blades contact the material is collected through the acoustic wave signals. This information reflects the contact situation between the scissor blades and the material; the acoustic wave signals are processed using a frequency analysis algorithm to extract the contact frequency pattern between the scissor blades and the material, and indirectly infer the actual spacing between the scissor blades; S2: Collect historical data of the scissor blades under different working conditions, build a deep learning model, and then input these historical data into the model for training, so that the model can learn and predict the changing trend of the scissor blade spacing. During the training process, the model parameters are optimized through cross-validation method to improve the generalization ability of the model to ensure that it can cope with changes in different working conditions; S3: Introduce a control system for real-time monitoring of the spacing between the scissor blades and adjusting working parameters. In the control system, the data collected by the acoustic sensor is integrated with the prediction results of the deep learning model and the material property data to achieve real-time monitoring of the actual spacing between the scissor blades. Through the feedback control system, the spacing between the scissor blades is dynamically adjusted based on the acoustic signal and the prediction results of the deep learning model. S4: The control system is equipped with a self-learning function. During the use of the scissor blades, the control system collects new data, optimizes the deep learning model according to the dynamic factors of the scissor blades, and adjusts the deep learning model parameters through an online learning mechanism, thereby improving the prediction accuracy of the change in the spacing between the scissor blades. The control system can dynamically adjust according to the new working environment and the state of the scissor blades to cope with changes under different working conditions, and ensure that the control strategy is automatically adjusted when the environment changes, always providing the optimal scissor blade distance; S5: Conduct performance verification experiments to test the accuracy of acoustic sensor feedback, the prediction accuracy of the deep learning model, and the response speed of the feedback control system. Evaluate the effectiveness of the scissor blade spacing control method through experimental data to further ensure the stability and efficiency of the control system. At the same time, debug the system in different types of shearing tasks to ensure that the method can provide stable and accurate spacing control under various loads and material cutting conditions, and continuously optimize system performance to ultimately achieve accurate and efficient scissor blade spacing control.

2. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The historical data in S2 include acoustic sensor feedback, scissor blade spacing, cutting load, material type information and environmental factors, and the environmental factors include temperature, humidity and air pressure. After preprocessing and feature extraction, these historical data serve as the basis for training the deep learning model to improve the prediction accuracy of the model.

3. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The acoustic sensor in S1 is a piezoelectric sensor or a micro-microphone array sensor.

4. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The dynamic factors in S4 include actual wear of the scissor blades, changes in workload, and changes in environmental conditions.

5. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The construction of the deep learning model in S2 includes the following steps: S21: Collect historical data of the scissor blades in different working conditions, including acoustic signal characteristics, environmental variables, and material characteristics, and clean, annotate, normalize, and extract features from the data to ensure high-quality and representative data input for the deep learning model; S22: Based on the temporal and feature requirements of the data, select a suitable deep learning model architecture to capture temporal dependencies and extract spatial features while meeting the task requirements of scissor blade spacing prediction; S23: Divide the collected data into training set, validation set and test set, train the model through the training set and perform cross-validation, tune the hyperparameters, ensure that the model has strong generalization ability on different data sets, and avoid overfitting or underfitting; S24: Use the validation set and test set to evaluate the model, adjust the model based on the error index, and further improve the prediction accuracy and robustness of the model through optimization algorithms; S25: Integrate the optimized deep learning model into the control system to dynamically adjust the scissor blade spacing in real time based on acoustic sensor feedback, while ensuring that the system can continuously optimize the model through online learning to adapt to different working environments and scissor blade states; S26: Regularly collect new data and update the model to respond to environmental and working condition changes, ensure that the deep learning model maintains efficient performance during long-term use, and maintain its stability and accuracy in actual applications through monitoring and adjustment.

6. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: In S3, the data collected by the acoustic sensor is integrated with the prediction results of the deep learning model and the material property data in the control system, including the following steps: S31: First, the raw signal captured by the acoustic sensor is preprocessed to ensure that it is synchronized with the input data of the deep learning model in time. The preprocessing includes signal denoising, signal standardization, time synchronization and filtering to remove noise and interference and ensure data quality. At the same time, the material characteristic data is preprocessed and standardized so that it can be effectively combined with the acoustic signal and the deep learning prediction results. S32: In the control system, the acoustic sensor data, the prediction results of the deep learning model and the material property data are combined through data fusion technology to achieve more accurate scissor blade spacing control. Specifically, the material property data will work together with the acoustic signal and the deep learning model prediction value during the fusion process. In this way, the control system can not only adjust the scissor blade spacing in real time, but also dynamically adjust the scissor blade spacing control strategy according to the actual physical properties of the material; S33: The fused scissor blade spacing value is used as a control signal and transmitted to the actuator module in the control system. Then, the control system calculates the error between the ideal spacing and the actual spacing according to the fused signal, and then adjusts it through the PID controller to accurately adjust the spacing of the scissor blades. The material characteristic data will affect the parameter adjustment of the controller in this process to ensure the accurate adjustment of the scissor blade spacing under different material conditions. S34: In order to cope with the ever-changing working environment and scissor blade status, the control system continuously collects new acoustic data and material property data and updates the deep learning model through an online learning mechanism. During the model adjustment process, the data fusion strategy is automatically optimized. At the same time, the adaptive fusion strategy is combined to dynamically adjust the weight value or filter coefficient in the data fusion method as the working environment changes. Changes in material properties will affect the adjustment of the weight value or filter coefficient to ensure the accurate combination of acoustic signals, deep learning model prediction results and material property data, thereby achieving precise control of the scissor blade spacing.

7. The method for controlling the spacing between the circular scissors blades according to claim 1, characterized in that: The automatic adjustment of the control strategy when the environment changes in S4 includes the following steps: S41: Input historical data and real-time collected acoustic wave signals into the deep learning model, combine sensor data and environmental factors, and train the model to enable it to automatically identify the material type and infer the actual contact between the scissor blades and the material based on the physical properties of the material, thereby dynamically adjusting the scissor blade spacing; S42: The deep learning model automatically adjusts the control strategy by analyzing the physical properties of the material and combining the frequency patterns reflected by the acoustic sensor; S43: When environmental conditions change, the deep learning model automatically identifies the impact of these environmental factors on the scissor blade spacing control by acquiring real-time data from environmental sensors, combined with acoustic signals and material characteristics; S44: According to the environment and material characteristics, the control system automatically updates the scissor blade spacing control strategy by adjusting the data fusion strategy; S45: The control system continuously collects and analyzes new data, optimizes the deep learning model, and dynamically adjusts the control strategy according to the new working environment and scissor blade status.

8. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The dynamic adjustment formula of the scissor blade spacing is as follows: in: D(t): represents the actual spacing of the scissor blades at time t, in millimeters (mm). This is the output value of the formula and is the spacing of the scissor blades that needs to be dynamically adjusted. S(t): Acoustic signal intensity, which indicates the intensity information of the acoustic wave signal. It is the signal captured by the acoustic sensor and reflects the intensity of the acoustic wave when the scissor blade contacts the material. The predicted value of the scissor blade spacing by the deep learning model indicates the scissor blade spacing value predicted by the deep learning model at time t; T(t): ambient temperature, used to consider the effect of temperature on the spacing between scissor blades; α: The influence coefficient of temperature on the adjustment of the scissor blade spacing; γ: total influence coefficient of environmental factors, reflecting the comprehensive influence of multiple factors on the adjustment of scissor blade spacing; e -βT ( t ): exponential decay function, which represents the effect of ambient temperature on the spacing between scissor blades. e is the base of natural logarithm, β is the temperature attenuation coefficient, which is used to control the effect of temperature on the adjustment of the spacing between scissor blades. T(t) represents the temperature at the current moment, T represents the temperature, and t represents the time point. (1+αe -βT(t) ): represents the influence of temperature on the spacing between the scissor blades. First, the exponential decay part e is calculated. -βT(t) , then add 1, and finally multiply it with the other parts. Considering the influence of different material types on the scissor blade spacing, δ i Represents the weight coefficient of material properties, η i Represents the actual value of the characteristic information of the i-th material at the current time t.

9. The method for controlling the spacing between the disc scissors blades according to claim 8, characterized in that: The temperature influence coefficient α in the dynamic adjustment formula is dynamically updated through the online learning mechanism of the deep learning model, and is adaptively adjusted according to the real-time monitored temperature, humidity and degree of wear of the scissor blades.

10. The method for controlling the spacing between the disc scissors blades according to claim 1, characterized in that: The control system comprises: The acoustic signal processing module is responsible for collecting acoustic wave signals from acoustic sensors and preprocessing them to ensure data quality; A deep learning model module for predicting and dynamically adjusting the spacing between scissor blades based on historical data and real-time sensor data; The material recognition module integrates multiple sensors to detect the physical properties of materials in real time, and classifies and identifies materials through a deep learning model. The material recognition module inputs the identified material information into the control system for subsequent adjustment of the scissor blade spacing; A data fusion module is used to fuse the output data of the acoustic signal processing module, the deep learning model module, and the material recognition module to more accurately control the scissor blade spacing; The actuator module can adjust the actual spacing of the scissor blades according to the fused scissor blade spacing control signal; Environmental adaptation module, which automatically adjusts the control strategy according to the real-time monitored environmental factors to ensure the optimization of the scissor blade spacing under different environmental conditions; The feedback and self-learning module collects new data, analyzes the dynamic factors of scissor blade wear and load changes, and optimizes the deep learning model and control strategy to achieve self-adjustment and optimization.

Images