Optimized control method and system for deviation rectifying mechanism of circular knife machine
Through the optimization control system of the circular tool machine deviation correction mechanism, the coil material is subjected to multiple deviation correction evaluations using neural network model and simulation module, solving the problem of inaccurate deviation correction results in the existing technology, and achieving a more efficient deviation correction effect.
Patent Information
- Application Number
- CN202411132645.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-26
AI Technical Summary
The prior art can only correct the coil material once, but cannot guarantee the accuracy of the corrective result.
Adoption of an optimization control system of the circular knife machine deviation correction mechanism, including the acquisition module, intelligent training module, simulation module and intelligent deviation correction module, through real-time acquisition of the roll die-cut images and inputting the parameter information of the deviation correction mechanism, the neural network is established using machine learning technology. Model, establish a simulation model for the deviation correction process, and conduct multiple deviation correction evaluations of the coil material based on the model until the deviation correction standard is reached.
Improve the accuracy and efficiency of coil material deviation correction to ensure that the deviation correction results meet the standards.
Smart Images

Figure CN119024766A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of a circular knife machine deviation correction mechanism, and in particular to an optimization control method and system for a circular knife machine deviation correction mechanism. Background Art
[0002] In recent years, with the development of industrial production, industrial machinery and equipment have gradually become richer and more refined. In order to facilitate the cutting and processing of various materials, circular knife machines are often used in the processing of industrial materials to improve processing efficiency. In the process of die-cutting materials using circular knife machines, the cut materials often deviate due to various reasons such as material defects, improper equipment installation and poor process, affecting the processing accuracy of the materials. Therefore, there is a circular knife machine correction mechanism to correct the deviated materials.
[0003] At present, the Chinese invention patent with application number 202022719913.2 provides a correction mechanism for a circular knife machine. Although the coil is corrected by the correction mechanism, correcting the coil only once cannot guarantee the accuracy of the correction result. Summary of the invention
[0004] The technical problem solved by the present invention is that the prior art only performs deviation correction on the coil once and cannot guarantee the accuracy of the deviation correction result.
[0005] In order to solve the above technical problems, in the first aspect, the present invention provides an optimization control system for a circular knife machine correction mechanism, wherein the optimization control system for the circular knife machine correction mechanism comprises an acquisition module, an intelligent training module, a simulation module and an intelligent correction module;
[0006] The acquisition module is used to acquire the coil die-cutting image and input the parameter information of the correction mechanism in real time;
[0007] The intelligent training module is used to train the neural network model on the original correction data using machine learning technology to obtain the trained neural network model;
[0008] The simulation module is used to establish a correction process simulation model based on the coil die-cutting image, coil parameter information, correction mechanism parameter information and the neural network model;
[0009] The intelligent deviation correction module is used to correct the coil according to the deviation correction process simulation model, and to perform deviation correction evaluation on the corrected coil;
[0010] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0011] The acquisition module includes an image acquisition unit and a data entry unit;
[0012] The image acquisition unit is used to acquire the coil die-cutting image in real time;
[0013] The data entry unit is used to enter parameter information of the deviation correction mechanism;
[0014] The parameter information of the correction mechanism includes the guide roller length, guide roller diameter, correction span, correction stroke and maximum speed;
[0015] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0016] The intelligent training module includes a data processing unit, a model training unit and a model storage unit;
[0017] The data processing unit is used to perform data preprocessing and data segmentation on the original correction data set, wherein the data preprocessing includes missing value processing, outlier processing, duplicate value processing and data conversion to obtain the correction data set;
[0018] The data segmentation refers to segmenting the correction data set into a training set and a test set, wherein the training set is used to establish a neural network model, and the test set is used to test and evaluate the neural network model;
[0019] The model training unit is used to establish a neural network model according to the training set and select the best neural network model according to the test set;
[0020] The model storage unit is used to store the neural network model and related parameters;
[0021] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0022] The simulation module includes an analysis unit and a model unit;
[0023] The analysis unit is used to identify the coil die-cutting image using image recognition technology, obtain coil parameter information, and use a neural network model to obtain correction data based on the coil parameter information and correction mechanism parameter information;
[0024] The model unit is used to establish a correction process simulation model according to coil parameter information, correction data and correction mechanism parameter information;
[0025] The coil parameter information includes coil size, coil material, coil speed, coil edge position data and coil center position data;
[0026] The correction data includes correction tension, correction speed, correction direction and correction distance;
[0027] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0028] The intelligent deviation correction module includes a deviation correction unit and an evaluation unit;
[0029] The correction unit is used to accurately correct the coil according to the process simulation model;
[0030] The evaluation unit is used to evaluate the web correction according to the web parameter information after correction, and obtain the correction evaluation result;
[0031] If the deviation correction evaluation result meets the deviation correction standard, the deviation correction is completed and the web is no longer corrected;
[0032] If the correction evaluation result does not meet the correction standard, the collection module is used to collect the coil die-cutting image after correction, and the coil parameter information after correction is obtained. The coil parameter information after correction is fed back to the simulation module, and the neural network model is used to re-establish the correction process simulation model according to the coil parameter information after correction. The coil is corrected again according to the correction process simulation model and the correction result is evaluated again, and this cycle is repeated until the correction evaluation result meets the correction standard.
[0033] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0034] The method of using the neural network model to obtain the correction data according to the coil parameter information and the correction mechanism parameter information includes:
[0035] Compare the edge position data and center position data of the coil with the standard position data of the coil to obtain the correction direction and correction angle of the coil;
[0036] After obtaining the correction direction and correction distance, the correction distance, correction tension and correction speed are obtained according to the correction mechanism parameters and coil parameter information using the neural network model;
[0037] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0038] The coil die-cutting image is identified by using image recognition technology to obtain coil position parameter information and coil material parameter information including:
[0039] Perform image preprocessing on the coil die-cutting image. Image preprocessing includes image segmentation, image enhancement, image binarization and image thinning to remove noise in the coil die-cutting image and obtain a clear coil die-cutting image.
[0040] Extracting image feature points of the preprocessed coil die-cutting image, establishing three-dimensional space coordinates, and obtaining position coordinates of the coil feature points according to the three-dimensional space coordinates;
[0041] Obtaining coil position parameter information according to the position coordinates of coil feature points;
[0042] Use the neural network model to identify the coil material and obtain the coil material parameter information;
[0043] The coil position parameter information includes coil size, coil speed, coil edge position data and coil center position data;
[0044] The coil material parameter information includes the coil material type;
[0045] As a preferred solution of the optimization control system of the circular knife machine correction mechanism described in the present invention, wherein:
[0046] The process of establishing three-dimensional space coordinates is as follows:
[0047] Place two sensors with exactly the same internal parameters in parallel to obtain the roll die-cutting image on the same plane;
[0048] Coordinate systems are established for the two sensors, defined as C1 coordinate system and C2 coordinate system respectively. The coordinates of any spatial point P are (x1, y1, z1) in C1 and (x2, y2, z2) in C2 coordinate system. The coordinates of the image point in the left sensor are (u1, v1) and (u2, v2) in the right sensor. The focal length of the sensor is f, and the distance between the two sensors is d. The three-dimensional coordinates of the spatial point P are obtained by calculation. The calculation expression is:
[0049]
[0050]
[0051]
[0052] Among them, X represents the horizontal coordinate of point P, Y represents the vertical coordinate of point P, and Z represents the vertical coordinate of point P.
[0053] In a second aspect, an optimization control method for a circular knife machine correction mechanism comprises the following steps:
[0054] Step S1: Establish a neural network model using the original correction data set;
[0055] Step S2: Collecting the coil die-cutting image and entering the parameter information of the correction mechanism;
[0056] Step S3: using image recognition technology to obtain coil parameter information according to the coil die-cutting image, and using a neural network model to obtain correction data according to the coil parameter information and correction mechanism parameter information;
[0057] Step S4: establishing a correction process simulation model according to the correction data, coil parameter information and correction mechanism parameter information;
[0058] Step S5: Correcting the web according to the correction process simulation model;
[0059] Step S6: collecting the coil die-cutting image after correction, using image recognition technology to obtain coil parameter information after correction, evaluating the coil correction result according to the correction standard, and obtaining the coil correction evaluation result;
[0060] Step S7: Repeat steps S2 to S6 until the web deflection correction evaluation result meets the deflection correction standard.
[0061] In a third aspect, the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in any one of the methods described above are executed.
[0062] Beneficial effects of the present invention: The present invention establishes a neural network model based on the original correction data set, and uses the neural network model to establish a correction process simulation model based on the coil die-cutting image and the correction mechanism parameter information, and corrects and evaluates the coil according to the correction process simulation model until the coil correction standard is met, so that the coil correction result is more accurate and effective. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Figure 1 A schematic diagram of the basic flow of an optimization control system for a circular knife machine correction mechanism provided by one embodiment of the present invention;
[0064] Figure 2 A schematic diagram of the basic flow chart of an optimization control method for a circular knife machine correction mechanism provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0065] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.
[0066] Example 1, reference Figure 1 , is an embodiment of the present invention, which provides an optimization control system for a circular knife machine correction mechanism, including an acquisition module, an intelligent training module, a simulation module and an intelligent correction module;
[0067] The acquisition module is used to acquire the coil die-cutting image and input the parameter information of the correction mechanism in real time;
[0068] The intelligent training module is used to train the neural network model on the original correction data using machine learning technology to obtain the trained neural network model;
[0069] The simulation module is used to establish a correction process simulation model based on the coil die-cutting image, coil parameter information, correction mechanism parameter information and the neural network model;
[0070] The intelligent deviation correction module is used to correct the coil according to the deviation correction process simulation model, and to perform deviation correction evaluation on the corrected coil;
[0071] In this embodiment, a neural network model is established according to the original correction data set, and a correction process simulation model is established according to the coil die-cutting image and correction mechanism parameter information using the neural network model. The coil is corrected and evaluated according to the correction process simulation model until the coil correction standard is reached, so that the correction result of the coil is more accurate and effective.
[0072] The acquisition module includes an image acquisition unit and a data entry unit;
[0073] The image acquisition unit is used to acquire the coil die-cutting image in real time;
[0074] The data entry unit is used to enter parameter information of the deviation correction mechanism;
[0075] The parameter information of the correction mechanism includes the guide roller length, guide roller diameter, correction span, correction stroke and maximum speed;
[0076] In this embodiment, by collecting the coil die-cutting image and entering the parameter information of the correction mechanism, accurate image support and data support are provided for obtaining correction data and establishing a correction process simulation model;
[0077] The intelligent training module includes a data processing unit, a model training unit and a model storage unit;
[0078] The data processing unit is used to perform data preprocessing and data segmentation on the original correction data set, wherein the data preprocessing includes missing value processing, outlier processing, duplicate value processing and data conversion to obtain the correction data set;
[0079] The data segmentation refers to segmenting the correction data set into a training set and a test set, wherein the training set is used to establish a neural network model, and the test set is used to test and evaluate the neural network model;
[0080] The model training unit is used to establish a neural network model according to the training set and select the best neural network model according to the test set;
[0081] The model storage unit is used to store the neural network model and related parameters;
[0082] The missing value processing includes: using pandas.isnan() to determine whether there are null values in the original data set, and when null values are returned, using DataFrame.dropna() to remove the data containing null values to avoid errors;
[0083] The outlier processing includes: using a determined threshold value to scan the data, identifying data that exceeds the threshold range, identifying these data as outliers, and using drop() to delete the outliers;
[0084] The duplicate value processing includes: using duplicate() to find duplicate data in the data set and define it as a duplicate value, using drop_dupliacate() to delete the duplicate value and retain the first or last one of the duplicate values;
[0085] The process of establishing the neural network model includes:
[0086] Extract features with information content and discriminative ability from the data set through feature engineering, select appropriate machine learning algorithms and neural network models according to the features and recognition targets, adjust the parameters of the neural network model using the training set, and evaluate the performance of the neural network model using the test set and cross-validation method, tune the neural network model, and obtain the best neural network model;
[0087] The feature engineering includes feature selection, feature extraction, feature transformation, feature dimension reduction, feature construction and feature expansion;
[0088] The feature selection and feature extraction are the processes of selecting a feature subset that is useful for the target task from the original data. Common feature selection methods include filtering, packaging, and embedding. The filtering method evaluates the importance of features based on statistical indicators or correlation analysis, and then selects features with high scores. The packaging method embeds the feature selection process into model training and uses model performance as an evaluation indicator for feature subsets. The embedding method regards feature selection as part of model training and selects useful features through regularization techniques or decision trees.
[0089] The feature transformation is a process of mathematically transforming the original features to better meet the needs of the model. Common feature transformation methods include logarithmic transformation, power transformation, normalization, and standardization. These transformations can improve the distribution characteristics of features and reduce the correlation between features, thereby improving the performance of the model;
[0090] The feature dimensionality reduction is the process of converting a high-dimensional feature space into a low-dimensional feature space, aiming to reduce the dimension of the feature and retain as much information as possible. Common feature dimensionality reduction methods include principal component analysis and linear discriminant analysis. These methods achieve data dimensionality reduction through projection or clustering techniques, thereby reducing computational complexity and improving the efficiency of model training.
[0091] The feature construction is a process of creating new features by combining, deriving or transforming original features to improve model performance, including creating new features by mathematical operations, text processing, time series analysis, etc.
[0092] The feature expansion is to increase the number and diversity of features by introducing external data or domain knowledge;
[0093] In this embodiment, a neural network model is established based on the original correction data set, and the neural network model is used to obtain correction data and establish a correction process simulation model, which saves the investment of manpower and material resources and improves the correction efficiency and correction accuracy;
[0094] The simulation module includes an analysis unit and a model unit;
[0095] The analysis unit is used to identify the coil die-cutting image using image recognition technology, obtain coil parameter information, and use a neural network model to obtain correction data based on the coil parameter information and correction mechanism parameter information;
[0096] The model unit is used to establish a correction process simulation model according to coil parameter information, correction data and correction mechanism parameter information;
[0097] The coil parameter information includes coil size, coil material, coil speed, coil edge position data and coil center position data;
[0098] The correction data includes correction tension, correction speed, correction direction and correction distance;
[0099] The image recognition technology is an important branch of the field of computer vision, which involves converting images into digital information and then analyzing and processing the digital information through computer algorithms to identify specific objects or features in the image;
[0100] In this embodiment, the image recognition technology is used to obtain the coil parameter information and the neural network model is used to obtain the correction data. According to the coil parameter information, the correction data and the correction mechanism parameter information, a correction process simulation model is established, which provides a detailed and accurate model reference for precise intelligent correction of the coil.
[0101] The intelligent deviation correction module includes a deviation correction unit and an evaluation unit;
[0102] The correction unit is used to accurately correct the coil according to the process simulation model;
[0103] The evaluation unit is used to evaluate the web correction according to the web parameter information after correction, and obtain the correction evaluation result;
[0104] If the deviation correction evaluation result meets the deviation correction standard, the deviation correction is completed and the web is no longer corrected;
[0105] If the correction evaluation result does not meet the correction standard, the collection module is used to collect the coil die-cutting image after correction, and the coil parameter information after correction is obtained. The coil parameter information after correction is fed back to the simulation module, and the neural network model is used to re-establish the correction process simulation model according to the coil parameter information after correction. The coil is corrected again according to the correction process simulation model and the correction result is evaluated again, and this cycle is repeated until the correction evaluation result meets the correction standard.
[0106] In this embodiment, the web correction results are evaluated multiple times according to the correction standard, and whether to continue the correction is determined according to the web evaluation results, which provides a practical guarantee for the accuracy of the web correction.
[0107] The method of using the neural network model to obtain the correction data according to the coil parameter information and the correction mechanism parameter information includes:
[0108] Compare the edge position data and center position data of the coil with the standard position data of the coil to obtain the correction direction and correction angle of the coil;
[0109] After obtaining the correction direction and correction distance, the correction distance, correction tension and correction speed are obtained according to the correction mechanism parameters and coil parameter information using the neural network model;
[0110] In this embodiment, the correction data is obtained by using a neural network model, which provides detailed and accurate correction data support for establishing an accurate correction process simulation model;
[0111] The coil die-cutting image is identified by using image recognition technology to obtain coil position parameter information and coil material parameter information including:
[0112] Perform image preprocessing on the coil die-cutting image. Image preprocessing includes image segmentation, image enhancement, image binarization and image thinning to remove noise in the coil die-cutting image and obtain a clear coil die-cutting image.
[0113] Extracting image feature points of the preprocessed coil die-cutting image, establishing three-dimensional space coordinates, and obtaining position coordinates of the coil feature points according to the three-dimensional space coordinates;
[0114] Obtaining coil position parameter information according to the position coordinates of coil feature points;
[0115] Use the neural network model to identify the coil material and obtain the coil material parameter information;
[0116] The coil position parameter information includes coil size, coil speed, coil edge position data and coil center position data;
[0117] The coil material parameter information includes the coil material type;
[0118] Image segmentation refers to the process of dividing an image into multiple sub-regions or pixel sets, each of which has certain statistical features or semantic information. Commonly used image segmentation algorithms include threshold-based segmentation algorithms, edge-based segmentation algorithms, region-based segmentation algorithms, and energy-based segmentation algorithms.
[0119] The threshold-based segmentation algorithm divides the pixels in the image into several regions according to their grayscale values, and usually uses a single threshold, multiple thresholds, and adaptive thresholds to perform segmentation;
[0120] The edge-based segmentation algorithm performs segmentation by detecting the edges or contours in the image. Commonly used algorithms include the Canny algorithm and the Sobel algorithm, which have better effects on images with obvious edges.
[0121] The region-based segmentation algorithm divides the pixels in the image into several regions and performs segmentation based on the similarity between the regions. Commonly used algorithms include K-means algorithm, watershed algorithm, etc.
[0122] The energy-based segmentation algorithm performs image segmentation by defining an energy function. Commonly used algorithms include GrabCut algorithm and GraphCut algorithm.
[0123] The image enhancement is a technology that processes the original image to improve its quality, readability and visualization, including image denoising, image sharpening and color enhancement;
[0124] Image denoising refers to a technique for improving the quality and readability of an image by processing image noise. Image denoising algorithms include median filtering, mean filtering, and medium filtering.
[0125] Image sharpening refers to a technique for improving image details and clarity by processing image edges. Sharpening algorithms include Gaussian filtering, Laplace filtering, and Helm filtering.
[0126] Color enhancement refers to a technique for improving the visualization effect of an image by adjusting the image's color saturation, contrast and other properties. Color enhancement algorithms include adaptive contrast adjustment, adaptive color enhancement and adaptive saturation adjustment.
[0127] Image binarization is a common operation in image processing. It converts the pixel values in an image to 0 or 1, thereby converting the image into a binary image containing only two pixel values (usually black and white). Common image binarization methods include global threshold binarization, adaptive threshold binarization, and histogram-based binarization.
[0128] The image thinning refers to the process of reducing the width of the lines of the image from multiple pixels to a unit pixel width. The image thinning algorithm includes an iterative algorithm and a non-iterative algorithm. The iterative algorithm includes a serial algorithm and a parallel algorithm according to whether the pixels are processed in parallel during the operation.
[0129] In the parallel algorithm, the deletion of pixels has nothing to do with the order of pixel values in the image, but only depends on the result of the previous iteration. In the serial algorithm, whether to delete pixels depends not only on the result of the previous iteration, but also on the distribution of the current pixel values.
[0130] Non-iterative thinning algorithms are not pixel-based. They generate a median or center line of the line in a single pass without checking all individual pixels. That is, this algorithm generates a skeleton in one pass; such as distance transform-based methods;
[0131] In this embodiment, the image recognition technology is used to obtain the coil position parameter information and the coil material parameter information, which provides detailed and accurate data support for obtaining the correction data and establishing the correction process simulation model;
[0132] The process of establishing three-dimensional space coordinates is as follows:
[0133] Place two sensors with exactly the same internal parameters in parallel to obtain the roll die-cutting image on the same plane;
[0134] Coordinate systems are established for the two sensors, defined as C1 coordinate system and C2 coordinate system respectively. The coordinates of any spatial point P are (x1, y1, z1) in C1 and (x2, y2, z2) in C2 coordinate system. The coordinates of the image point in the left sensor are (u1, v1) and (u2, v2) in the right sensor. The focal length of the sensor is f, and the distance between the two sensors is d. The three-dimensional coordinates of the spatial point P are obtained by calculation. The calculation expression is:
[0135]
[0136]
[0137]
[0138] Wherein, X represents the horizontal coordinate of point P, Y represents the vertical coordinate of point P, and Z represents the vertical coordinate of point P;
[0139] In this embodiment, three-dimensional space coordinates are established to obtain detailed and accurate position coordinates of the coil, thereby providing accurate coordinate data for obtaining the deviation correction data of the coil, thereby improving the accuracy of the deviation correction of the coil.
[0140] Example 2, reference Figure 2 , which is another embodiment of the present invention, is different from the first embodiment in that it provides an optimization control method for a circular knife machine correction mechanism, comprising the following steps:
[0141] Step S1: Establish a neural network model using the original correction data set;
[0142] Step S2: Collecting the coil die-cutting image and entering the parameter information of the correction mechanism;
[0143] Step S3: using image recognition technology to obtain coil parameter information according to the coil die-cutting image, and using a neural network model to obtain correction data according to the coil parameter information and correction mechanism parameter information;
[0144] Step S4: establishing a correction process simulation model according to the correction data, coil parameter information and correction mechanism parameter information;
[0145] Step S5: Correcting the web according to the correction process simulation model;
[0146] Step S6: collecting the coil die-cutting image after correction, using image recognition technology to obtain coil parameter information after correction, evaluating the coil correction result according to the correction standard, and obtaining the coil correction evaluation result;
[0147] Step S7: Repeat steps S2 to S6 until the web deflection correction evaluation result meets the deflection correction standard.
[0148] Embodiment 3 provides an electronic device, including a processor and a memory, wherein:
[0149] The memory stores a computer program that can be called by the processor;
[0150] The processor executes the above-mentioned optimization control method and system for the circular knife machine correction mechanism in the background by calling the computer program stored in the memory.
[0151] The present invention establishes a neural network model according to the original correction data set, and uses the neural network model to establish a correction process simulation model according to the coil die-cutting image and the correction mechanism parameter information. The coil is corrected and evaluated according to the correction process simulation model until the coil correction standard is met, so that the coil correction result is more accurate and effective.
[0152] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0153] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
Claims
1. An optimization control system for a circular knife machine correction mechanism, characterized in that: It includes acquisition module, intelligent training module, simulation module and intelligent deviation correction module; The acquisition module is used to acquire the coil die-cutting image and input the parameter information of the correction mechanism in real time; The intelligent training module is used to train the neural network model on the original correction data using machine learning technology to obtain the trained neural network model; The simulation module is used to establish a correction process simulation model based on the coil die-cutting image, coil parameter information, correction mechanism parameter information and the neural network model; The intelligent deviation correction module is used to correct the coil according to the deviation correction process simulation model, and to perform deviation correction evaluation on the corrected coil.
2. The optimization control system of the circular knife machine correction mechanism according to claim 1, characterized in that: The acquisition module includes an image acquisition unit and a data entry unit; The image acquisition unit is used to acquire the coil die-cutting image in real time; The data entry unit is used to enter parameter information of the deviation correction mechanism; The parameter information of the correction mechanism includes the guide roller length, the guide roller diameter, the correction span, the correction stroke and the maximum speed.
3. The optimization control system of the circular knife machine correction mechanism according to claim 1, characterized in that: The intelligent training module includes a data processing unit, a model training unit and a model storage unit; The data processing unit is used to perform data preprocessing and data segmentation on the original correction data set, wherein the data preprocessing includes missing value processing, outlier processing, duplicate value processing and data conversion to obtain the correction data set; The data segmentation refers to segmenting the correction data set into a training set and a test set, wherein the training set is used to establish a neural network model, and the test set is used to test and evaluate the neural network model; The model training unit is used to establish a neural network model according to the training set and select the best neural network model according to the test set; The model storage unit is used to store the neural network model and related parameters.
4. The optimization control system of the circular knife machine correction mechanism according to claim 1, characterized in that: The simulation module includes an analysis unit and a model unit; The analysis unit is used to identify the coil die-cutting image using image recognition technology, obtain coil parameter information, and use a neural network model to obtain correction data based on the coil parameter information and correction mechanism parameter information; The model unit is used to establish a correction process simulation model according to coil parameter information, correction data and correction mechanism parameter information; The coil parameter information includes coil size, coil material, coil speed, coil edge position data and coil center position data; The correction data includes correction tension, correction speed, correction direction and correction distance.
5. The optimization control system of the circular knife machine correction mechanism according to claim 1, characterized in that: The intelligent deviation correction module includes a deviation correction unit and an evaluation unit; The correction unit is used to accurately correct the coil according to the process simulation model; The evaluation unit is used to evaluate the web correction according to the web parameter information after correction, and obtain the correction evaluation result; If the deviation correction evaluation result meets the deviation correction standard, the deviation correction is completed and the web is no longer corrected; If the correction evaluation result does not meet the correction standard, the acquisition module is used to collect the die-cutting image of the coil after correction, obtain the coil parameter information after correction, and feed the coil parameter information after correction to the simulation module. The neural network model is used to re-establish the correction process simulation model according to the coil parameter information after correction. The coil is corrected again according to the correction process simulation model and the correction result is evaluated again. This cycle is repeated until the correction evaluation result meets the correction standard.
6. The optimization control system of the circular knife machine correction mechanism according to claim 4, characterized in that: The method of using the neural network model to obtain the correction data according to the coil parameter information and the correction mechanism parameter information includes: Compare the edge position data and center position data of the coil with the standard position data of the coil to obtain the correction direction and correction angle of the coil; After obtaining the correction direction and correction distance, the neural network model is used to obtain the correction distance, correction tension and correction speed according to the correction mechanism parameters and coil parameter information.
7. The optimization control system of the circular knife machine correction mechanism according to claim 4, characterized in that: The coil die-cutting image is identified by using image recognition technology to obtain coil position parameter information and coil material parameter information including: Perform image preprocessing on the coil die-cutting image. Image preprocessing includes image segmentation, image enhancement, image binarization and image thinning to remove noise in the coil die-cutting image and obtain a clear coil die-cutting image. Extracting image feature points of the preprocessed coil die-cutting image, establishing three-dimensional space coordinates, and obtaining position coordinates of the coil feature points according to the three-dimensional space coordinates; Obtaining coil position parameter information according to the position coordinates of coil feature points; Use the neural network model to identify the coil material and obtain the coil material parameter information; The coil position parameter information includes coil size, coil speed, coil edge position data and coil center position data; The coil material parameter information includes the coil material type.
8. The optimization control system of the circular knife machine correction mechanism according to claim 7, characterized in that: The process of establishing three-dimensional space coordinates is as follows: Place two sensors with exactly the same internal parameters in parallel to obtain the roll die-cutting image on the same plane; Coordinate systems are established for the two sensors, defined as C1 coordinate system and C2 coordinate system respectively. The coordinates of any spatial point P are (x1, y1, z1) in C1 and (x2, y2, z2) in C2 coordinate system. The coordinates of the image point in the left sensor are (u1, v1) and (u2, v2) in the right sensor. The focal length of the sensor is f, and the distance between the two sensors is d. The three-dimensional coordinates of the spatial point P are obtained by calculation. The calculation expression is: Among them, X represents the horizontal coordinate of point P, Y represents the vertical coordinate of point P, and Z represents the vertical coordinate of point P.
9. An optimization control method for a circular knife machine correction mechanism, comprising an optimization control system for a circular knife machine correction mechanism as claimed in any one of claims 1 to 8, characterized in that: The following steps are involved: Step S1: Establish a neural network model using the original correction data set; Step S2: Collecting the coil die-cutting image and entering the parameter information of the correction mechanism; Step S3: using image recognition technology to obtain coil parameter information according to the coil die-cutting image, and using a neural network model to obtain correction data according to the coil parameter information and correction mechanism parameter information; Step S4: establishing a correction process simulation model according to the correction data, coil parameter information and correction mechanism parameter information; Step S5: Correcting the web according to the correction process simulation model; Step S6: collecting the coil die-cutting image after correction, using image recognition technology to obtain coil parameter information after correction, evaluating the coil correction result according to the correction standard, and obtaining the coil correction evaluation result; Step S7: Repeat steps S2 to S6 until the web deflection correction evaluation result meets the deflection correction standard.
10. An electronic device, characterized in that: include: A processor and a memory, wherein: The memory stores a computer program that can be called by the processor; The processor executes the optimization control system of the circular knife machine correction mechanism described in any one of claims 1 to 8 in the background by calling the computer program stored in the memory.
Citation Information
Patent Citations
Deviation rectifying mechanism of circular knife machine
CN214269521U
Visual imaging sensor point-light ranging sensor measuring coordinate system unifying method
CN109323650A
High-speed online cross-shaped detecting and deviation correcting device and detecting and deviation correcting method using same
CN110091381A
Current ring target prediction method and system in shield deviation rectification
CN110188947A
Automatic deviation correction control method for cold-rolled strip steel continuous annealing heating furnace
CN112029990A
