A method, system, and apparatus for corrugated linerboard warp sorting and process control

Abstract

The present application relates to the field of corrugated board production, and discloses a corrugated board production line outlet paperboard warping classification and process control method, system and equipment, which is used for real-time acquisition and analysis of the warping type of the paperboard production process, and outputs the production line control parameters according to the warping type. The method comprises the following steps: on-line detecting the vertical plane curve profile image of the paperboard and judging the paperboard type; when the paperboard type is a non-L-shaped paperboard, on-line calculating the warping convexity of the paperboard; if the absolute value of the warping convexity is not greater than a preset warping threshold value, it is determined that the production paperboard quality is qualified; if the absolute value of the warping convexity is greater than the preset warping threshold value, the warping type of the production paperboard is determined and output according to the value of the warping convexity and the preset range value of the warping convexity corresponding to different warping types; the paperboard warping type is input into a pre-trained neural network model to obtain the production line control parameters output by the neural network model, and the production line control parameters are transmitted to the production line control center for adjustment control of the paperboard quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of paperboard production, and particularly relates to a corrugated paperboard warping classification and process control method, system and equipment. BACKGROUND

[0002] Corrugated paper packaging is widely used in downstream consumer industries such as home appliances, electronic products, IT, food and beverage, books, daily chemicals, textiles, and logistics and express companies.

[0003] The quality of the corrugated paperboard production line product is mainly related to the factors such as the quality of the raw paper, the moisture content, and the temperature control in the production process. The warping degree of the outlet paperboard is an important indicator reflecting the quality of the corrugated paperboard production line product. The process control of the corrugated paperboard production line is mainly to monitor the warping degree of the outlet paperboard in real time, and to realize effective feedback control of the quality of the paperboard. Among them, the online detection of the warping degree of the outlet paperboard and the identification of the warping level are the key to realizing the process control of the production line. At present, the corrugated paperboard production lines that have been put into production in China mainly use the method of manual input parameter adjustment to control the product quality by relying on the on-site engineers to judge and identify the outlet paperboard. However, the production speed of the corrugated paperboard production line is generally several hundred meters per minute, and it is difficult to realize the online detection of the warping degree of the outlet paperboard and the identification of the warping level by relying on manual work. Moreover, the rapid development of online shopping and express delivery has brought about the market demand for customization and individualization, which has led to the requirement of small batches of corrugated paperboard production, and the number of orders is large and the material quality changes frequently during the production process, resulting in high labor cost in the corrugated paperboard industry, which affects the automation and intelligent level of the corrugated paperboard manufacturing industry. SUMMARY

[0004] The present application provides a corrugated paperboard warping classification and process control method, system and equipment for real-time acquisition and analysis of the warping type of the produced paperboard, and output of corresponding production line control parameters according to the warping type.

[0005] The first aspect of the present application provides a method for classifying and process controlling the warping of a paperboard at a tile line outlet, the method comprising: collecting a vertical plane curve profile image of a paperboard produced using a line structured light sensor, and determining the type of the paperboard produced according to the vertical plane curve profile image; when the type of the paperboard produced is a non-L-shaped paperboard, calculating the warping convexity of the paperboard produced based on the vertical plane curve profile image; if the absolute value of the warping convexity is not greater than a preset warping threshold value, determining that the quality of the paperboard produced is qualified, and if the absolute value of the warping convexity is greater than the preset warping threshold value, determining and outputting the warping type of the paperboard produced according to the value of the warping convexity and the preset warping convexity range value corresponding to different warping types; inputting the warping type of the paperboard produced into a pre-trained neural network model to obtain the production line control parameter output by the pre-trained neural network model, and outputting the production line control parameter to a production line control center for adjusting and controlling the quality of the paperboard.

[0006] Preferably, the collecting of the vertical plane curve profile image of the paperboard produced using the line structured light sensor and the determination of the type of the paperboard produced according to the vertical plane curve profile image comprise: collecting multiple vertical plane reflected light images of the paperboard produced using the line structured light sensor, and pre-processing the collected vertical plane reflected light images to obtain multiple pre-processed images; extracting and integrating the vertical plane curve profile image of the paperboard produced from the multiple pre-processed images; performing data fitting according to the extracted vertical plane curve profile image to obtain a monomial cubic function of the curve profile horizontal pixel coordinate-curve profile vertical pixel coordinate; calculating the number of extreme points of the monomial cubic function, and when the number of extreme points of the monomial cubic function is 0, determining that the paperboard produced is a normal paperboard, when the number of extreme points of the monomial cubic function is 1, determining that the paperboard produced is a non-S-shaped paperboard, and when the number of extreme points of the monomial cubic function is 2, determining that the paperboard produced is an S-shaped paperboard and issuing an alarm; when the paperboard produced is a non-S-shaped paperboard, calculating the difference of the derivatives on both sides of the paperboard produced based on the monomial cubic function; when the difference exceeds a preset derivative range value, determining that the paperboard produced is an L-shaped paperboard and issuing an alarm, and otherwise, determining that the paperboard produced is a non-L-shaped paperboard.

[0007] Preferably, the collecting of the vertical plane curve profile image of the paperboard produced using the line structured light sensor and the determination of the type of the paperboard produced according to the vertical plane curve profile image comprise: collecting multiple vertical plane reflected light images of the paperboard produced using the line structured light sensor, and pre-processing the collected vertical plane reflected light images to obtain multiple pre-processed images; projecting structured light onto the vertical plane of the paperboard produced; collecting multiple vertical plane reflected light images of the paperboard produced using the line structured light sensor; and pre-processing the collected vertical plane reflected light images to obtain multiple pre-processed images.

[0008] Preferably, the vertical plane curve profile image of the production paperboard is extracted and integrated from the plurality of preprocessed images, comprising: extracting the light stripes in the plurality of preprocessed images to obtain a plurality of groups of light stripes; mapping the position of each group of light stripes to a spatial three-dimensional coordinate by using the camera calibration parameters; and fusing the spatial three-dimensional coordinates of the plurality of groups of light stripes after conversion and alignment to obtain a vertical plane three-dimensional profile of the production paperboard; and processing the vertical plane three-dimensional profile of the production paperboard to obtain the vertical plane curve profile image of the production paperboard.

[0009] Preferably, the monomial cubic function of the curve profile transverse pixel coordinate-curve profile longitudinal pixel coordinate is represented as:

[0010] f1(x) = a1x 3 +b1x 2 +c1x+d

[0011] In the formula, x represents the transverse pixel coordinate of the vertical plane curve profile, f1(x) represents the longitudinal pixel coordinate of the vertical plane curve profile at the corresponding position, and a1, b1, c1, and d represent fitting constants.

[0012] Preferably, the difference δ of the derivatives of the two sides of the production paperboard is represented as:

[0013]

[0014] In the formula, x i represents the transverse pixel coordinate of the vertical plane curve profile, x1, x2, x3...x n ...x m are all pixel horizontal coordinates, m represents the number of all pixel coordinate points, f′1(x) represents the function value of the derivative function at the x i position, and x n is the point that makes f′1(x) = 0.

[0015] Preferably, when the type of the production paperboard is a non-L-shaped paperboard, the warping convexity of the production paperboard is calculated based on the vertical plane curve profile image, comprising: when the type of the production paperboard is a non-L-shaped paperboard, data fitting is performed based on the vertical plane curve profile image to obtain a monomial quadratic function of the vertical plane curve profile transverse pixel coordinate-vertical plane curve profile longitudinal pixel coordinate, and the monomial quadratic function relationship is represented as:

[0016] f(x) = ax 2 +bx+c

[0017] In the formula, x represents the transverse pixel coordinate of the vertical plane curve profile, f(x) represents the longitudinal pixel coordinate of the vertical plane curve profile at the corresponding x position, and a, b, and c represent constants.

[0018] According to the monomial quadratic function, a warping convexity θ corresponding to the produced paperboard is calculated by using the following formula, and the warping convexity θ is represented as:

[0019]

[0020] In the formula, K(x) is an average value of curvatures at positions of three quarter points, f'(x) represents a slope, f'(x) is a derivative of the vertical surface curve profile, f''(x) represents a rate of change of the curvature, f''(x) is a second derivative of the vertical surface curve profile, x a and x b respectively represent pixel coordinates of positions of two end points of the horizontal pixel coordinates of the monomial quadratic function, x1, x2, x3 respectively represent pixel coordinates of positions of three quarter points of the horizontal pixel coordinates of the monomial quadratic function, and are respectively as follows:

[0021]

[0022] Preferably, the method of inputting the warping type of the produced paperboard into the pre-trained neural network model to obtain the production line control parameter output by the pre-trained neural network model, and outputting the production line control parameter to the production line control center to adjust and control the paperboard quality comprises the following steps: establishing an input-output data set by using the paperboard warping type or warping degree and the production line quality control parameter recorded on the corrugated paperboard production line, training the input-output data set by using a neural network model to obtain a pre-trained neural network model, and establishing a corresponding relationship between the production line paperboard warping type and the production line control parameter by the pre-trained neural network model; inputting the warping type of the produced paperboard into the pre-trained neural network model to obtain the production line control parameter output by the pre-trained neural network model; and outputting the production line control parameter to the production line control center to adjust and control the paperboard quality.

[0023] The second aspect of the present application provides a corrugated line outlet paperboard warping classification and process control system, comprising: a collection module, configured to collect a vertical surface curve profile image of a produced paperboard by using a line structured light sensor, and determine a paperboard type of the produced paperboard according to the vertical surface curve profile image; a calculation module, configured to calculate a warping convexity of the produced paperboard based on the vertical surface curve profile image when the type of the produced paperboard is a non-L-shaped paperboard; a judgment module, configured to determine that the quality of the produced paperboard is qualified when an absolute value of the warping convexity is not greater than a warping preset threshold value, and determine and output a warping type of the produced paperboard according to a value of the warping convexity and a warping convexity preset range value corresponding to different warping types when the absolute value of the warping convexity is greater than the warping preset threshold value; and a control parameter output module, configured to input the warping type of the produced paperboard into a pre-trained neural network model to obtain a production line control parameter output by the pre-trained neural network model, and output the production line control parameter to a production line control center to adjust and control the paperboard quality.

[0024] The third aspect of the present application provides a corrugated line outlet paperboard warping classification and process control device, comprising: a memory and at least one processor, the memory storing computer readable instructions, and the memory and the at least one processor being interconnected by a line; the at least one processor invokes the computer readable instructions in the memory to enable the corrugated line outlet paperboard warping classification and process control device to perform each step of the corrugated line outlet paperboard warping classification and process control method described above.

[0025] In the technical solution provided by the present application, the vertical plane curve image of the paperboard is accurately collected, the type of the paperboard is intelligently judged, and the warping convexity is calculated, so that the rapid and accurate evaluation of the paperboard quality is realized. When the warping problem is found, the warping type can be automatically classified, and the production line control parameters are output according to the pre-trained neural network model, so that the production can be timely adjusted, the paperboard quality is ensured to be stable, and the overall production efficiency and intelligent management level are improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 A first flowchart of a corrugated line outlet paperboard warping classification and process control method provided by an embodiment of the present application;

[0027] Figure 2 A structure schematic diagram of a corrugated line outlet paperboard warping classification and process control system provided by an embodiment of the present application;

[0028] Figure 3 A structure schematic diagram of a corrugated line outlet paperboard warping classification and process control device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0029] The terms "first", "second", "third", "fourth" and the like in the description, claims, and drawings of the present application, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of these terms in the present description are to be construed to cover a general order so that "first" and "second" steps are not necessarily performed in that order unless it is expressly specified otherwise. Furthermore, these terms can be interchanged under appropriate circumstances such that the embodiments described herein are, for example, capable of efficient implementation in either order unless a specific or context does not allow that the terms are inter-changeable. Additionally, the term "comprising" or "containing" and variations thereof as used in the present description are intended to cover a non-exclusive inclusion such that a process, method, system, product, or apparatus that comprises a list of steps or units not necessarily limited to those specifically identified, but can include additional steps or units not expressly listed or inherent to such process, method, product, or apparatus.

[0030] For the convenience of understanding, the specific flow of the embodiments of the present application is described below. Please refer to Figure 1 A corrugated line outlet paperboard warping classification and process control method in an embodiment of the present application comprises:

[0031] S101, collecting a vertical plane curve profile image of the production paperboard using a line structured light sensor, and determining a paperboard type of the production paperboard according to the vertical plane curve profile image;

[0032] S102, when the type of the production paperboard is a non-L-type paperboard, calculating a warping convexity of the production paperboard based on the vertical plane curve profile image;

[0033] S103, if an absolute value of the warping convexity is not greater than a warping preset threshold value, determining that the production paperboard is qualified, and if the absolute value of the warping convexity is greater than the warping preset threshold value, determining and outputting a warping type of the production paperboard according to a value of the warping convexity and a warping convexity preset range value corresponding to different warping types;

[0034] S104, inputting the warping type of the production paperboard into a pre-trained neural network model to obtain a production line control parameter output by the pre-trained neural network model, and outputting the production line control parameter to a production line control center.

[0035] It can be understood that the execution subject of the present application can be a paperboard warping classification and process control system at a corrugated paperboard outlet, and can also be a terminal or a server, and the specific execution subject is not limited herein. The server is taken as an example for description of the embodiment of the present application.

[0036] In the embodiment, in step S101, the vertical plane curve profile image of the production paperboard is collected using the line structured light sensor, and the paperboard type of the production paperboard is determined according to the vertical plane curve profile image, which includes: collecting the vertical plane reflection light image of the production paperboard multiple times using the line structured light sensor, and pre-processing the collected vertical plane reflection light image to obtain multiple pre-processed images; extracting and integrating the vertical plane curve profile image of the production paperboard from the multiple pre-processed images; performing data fitting according to the extracted vertical plane curve profile image to obtain a monomial cubic function of the curve profile horizontal pixel coordinate-curve profile vertical pixel coordinate; calculating the number of extreme points of the monomial cubic function, when the number of extreme points of the monomial cubic function is 0, determining that the production paperboard is a normal paperboard, when the number of extreme points of the monomial cubic function is 1, determining that the production paperboard is a non-S-type paperboard, when the number of extreme points of the monomial cubic function is 2, determining that the production paperboard is an S-type paperboard, and issuing an alarm; when the production paperboard is a non-S-type paperboard, calculating the difference of the derivatives on both sides of the production paperboard based on the monomial cubic function; when the difference exceeds a derivative preset range value, determining that the production paperboard is an L-type, and issuing an alarm, otherwise, determining that the production paperboard is a non-L-type.

[0037] In the embodiment, the vertical plane reflection light images of the production paperboard are collected multiple times, and the collected vertical plane reflection light images are preprocessed to obtain multiple preprocessed images, including: projecting structured light to the vertical plane of the production paperboard; collecting the vertical plane reflection light images of the production paperboard multiple times using a structured light sensor; and preprocessing the collected vertical plane reflection light images to obtain multiple preprocessed images.

[0038] Optionally, the structured light can be a light source emitted by a linear laser, and the structured light sensor can be a camera device.

[0039] It can be understood that when the vertical plane reflection light images of the production paperboard are collected, there can be regions without pixel points, and therefore, the camera device is used to collect the vertical plane reflection light images of the production paperboard multiple times, and the images collected multiple times are analyzed and processed.

[0040] In the embodiment, the vertical plane curve contour images of the production paperboard are extracted and integrated from the multiple preprocessed images, including: extracting the light stripes in the multiple preprocessed images to obtain multiple groups of light stripes; mapping the position of each group of light stripes to a spatial three-dimensional coordinate using camera calibration parameters; converting and aligning the spatial three-dimensional coordinates of the multiple groups of light stripes to obtain a vertical plane three-dimensional contour of the production paperboard; and processing the vertical plane three-dimensional contour of the production paperboard to obtain the vertical plane curve contour images of the production paperboard.

[0041] In the embodiment, an image processing algorithm (such as edge detection, threshold segmentation, etc.) is used to identify the light stripes in the images, a Canny edge detector or a Sobel operator is used to detect the edges in the images, and the light stripes are separated from the background by setting a suitable threshold.

[0042] In the embodiment, before the three-dimensional reconstruction is performed, the camera needs to be calibrated to obtain the intrinsic and extrinsic parameters of the camera. Then, the camera calibration parameters and the positions of the light stripes in the images are used to map the positions of the light stripes to the spatial three-dimensional coordinates through the principle of perspective projection.

[0043] In the embodiment, the vertical plane three-dimensional contour of the production paperboard is obtained by converting and aligning the spatial three-dimensional coordinates under different viewing angles. If overlapping regions are encountered, a point cloud fusion algorithm (such as an ICP algorithm) is used to optimize these overlapping regions to ensure the continuity and accuracy of the three-dimensional contour.

[0044] In this embodiment, a slice plane perpendicular to the plane of the paperboard is selected from the vertical face 3D profile of the produced paperboard, which should be located at the expected cross-section position of the paperboard. Then, on the slice plane, a curve profile of the paperboard is extracted using a profile extraction algorithm (such as Canny edge detection, Hough transform, etc.). This profile will reflect the shape and size of the paperboard at this position. Finally, a vertical face curve profile image is generated by smoothing the vertical face 3D profile of the produced paperboard through a filtering algorithm (such as Gaussian filtering, median filtering, etc.). Smoothing the curve-extracted profile through a filtering algorithm (such as Gaussian filtering, median filtering, etc.) can eliminate minor fluctuations caused by noise or measurement errors.

[0045] In this embodiment, the unary cubic function of the curve profile lateral pixel coordinate - curve profile longitudinal pixel coordinate is expressed as:

[0046] f1(x) = a1x 3 +b1x 2 +c1x+d

[0047] In the formula, x represents the curve profile lateral pixel coordinate, f1(x) represents the longitudinal pixel coordinate of the corresponding position of the vertical face curve profile, a1, b1, c1 and d represent the fitting constants.

[0048] The difference δ of the derivatives of the two sides of the produced paperboard is expressed as:

[0049]

[0050] In the formula, xi represents the lateral pixel coordinate of the vertical face curve profile, x1, x2, x3... x n ...x m are all pixel coordinates, m represents the number of all pixel coordinate points, f′1(x) represents the function value of the derivative function at the corresponding x i position, x n is the point that makes f′1(x) = 0.

[0051] It can be understood that by calculating the difference of the derivatives of the two sides of the corrugated paperboard and comparing the calculated difference with the preset range value of the derivative, it is determined whether the paperboard is an L-shaped paperboard or a non-L-shaped paperboard. The reason is that L-shaped paperboard will have a significant change in slope at the bending position (i.e., the corner of L), while non-L-shaped paperboard will not have such a feature.

[0052] The difference of the derivatives of the two sides of the corrugated paperboard is calculated, i.e., the derivative (slope) is calculated on both sides of the suspected L-shaped region (i.e., the region where the corner may exist) of the corrugated paperboard. Then, the difference of the derivatives on both sides is calculated. This difference reflects the degree of change in slope in this region.

[0053] In the embodiment, before formal production, the corrugated paper is trial-produced, after the trial production, the difference between the two sides of the non-L-shaped paperboard derivative during the trial production is taken as the maximum value of the derivative preset range value, and the difference between the two sides of the L-shaped paperboard derivative during the trial production is taken as the minimum value of the derivative preset range value.

[0054] In the embodiment, in step S102, when the type of the production paperboard is the non-L-shaped paperboard, the warping convexity of the production paperboard is calculated based on the vertical plane curve profile image, including:

[0055] When the type of the production paperboard is the non-L-shaped paperboard, data fitting is performed based on the vertical plane curve profile image to obtain a monomial quadratic function of the vertical plane curve profile transverse pixel coordinate-vertical plane curve profile longitudinal pixel coordinate, and the monomial quadratic function relationship is:

[0056] f(x)=ax 2 +bx+c

[0057] In the formula, x represents the transverse pixel coordinate of the vertical plane curve profile, f(x) represents the longitudinal pixel coordinate of the vertical plane curve profile corresponding to the x position, a, b, and c represent constants;

[0058]

[0059] In the formula, K(x) is the average value of the curvatures at the positions of the three quarter points, f'(x) represents the slope, f'(x) is the derivative of the vertical plane curve profile, f''(x) represents the rate of change of the curvature, f''(x) is the second derivative of the vertical plane curve profile, x a and x b respectively represent the pixel coordinates of the positions of the two end points of the transverse pixel coordinate of the monomial quadratic function, x1, x2, and x3 respectively represent the pixel coordinates of the positions of the three quarter points of the transverse pixel coordinate of the monomial quadratic function, and are respectively:

[0060]

[0061] In the embodiment, in step S103, the determination method of the warping convexity preset range value specifically includes: collecting cross-sectional images of non-L-shaped paperboards with different warping directions and warping degrees as samples; identifying the warping direction and the warping degree of the samples, and classifying the samples according to different warping types to obtain a classification result, establishing a sub-image library corresponding to different warping types and warping levels according to the classification result, calculating the warping convexity of each sample in each sub-image library one by one, and determining the warping convexity range value of the warping type and the warping level corresponding to each sub-image library according to the minimum value and the maximum value of the warping convexity of the samples in each sub-image library.

[0062] It can be understood that the calculation method of the warping convexity of each sample in each of the sub-image libraries can adopt the calculation method of the warping convexity θ as described above, which is not described herein.

[0063] Exemplarily, when the warping direction and the warping degree of the sample are identified and the sample is classified according to different warping types and warping levels to obtain a classification result, the classification result formed includes upwarp-type 1, upwarp-type 2, upwarp-type 3, downwarp-type 1, downwarp-type 2, and downwarp-type 3, and then six sub-image libraries are established according to the classification result, each of which corresponds to upwarp-type 1, upwarp-type 2, upwarp-type 3, downwarp-type 1, downwarp-type 2, and downwarp-type 3, and then the warping convexity of each sample in the six sub-image libraries is calculated one by one, and the warping convexity range values of upwarp-type 1, upwarp-type 2, upwarp-type 3, downwarp-type 1, downwarp-type 2, and downwarp-type 3 are determined, including: determining the warping convexity range value corresponding to upwarp-type 1 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to upwarp-type 1 determining the warping convexity range value corresponding to upwarp-type 2 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to upwarp-type 2 determining the warping convexity range value corresponding to upwarp-type 3 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to upwarp-type 3 determining the warping convexity range value corresponding to downwarp-type 1 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to downwarp-type 1 determining the warping convexity range value corresponding to downwarp-type 2 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to downwarp-type 2 determining the warping convexity range value corresponding to downwarp-type 3 according to the minimum value and the maximum value of the warping convexity of the sample in the image library corresponding to downwarp-type 3

[0064] In the embodiment, the warping preset threshold is If the absolute value of the warping convexity is not greater than then it is determined that the production paperboard quality is qualified.

[0065] 9. In this embodiment, in step S104, the warping type of the produced paperboard is input into the pre-trained neural network model to obtain the production line control parameters output by the pre-trained neural network model, and the production line control parameters are output to the production line control center for adjustment to control the quality of the paperboard. Specifically, an input-output data set is established by using the recorded paperboard warping type or warping degree and production line quality control parameters on the corrugated paperboard production line, a neural network model is trained on the input-output data set to obtain a pre-trained neural network model, and the pre-trained neural network model establishes a corresponding relationship between the production line paperboard warping type and the production line control parameters; the warping type of the produced paperboard is input into the pre-trained neural network model to obtain the production line control parameters output by the pre-trained neural network model; and the production line control parameters are output to the production line control center for adjustment to control the quality of the paperboard.

[0066] In this embodiment, different warping types correspond to different production line control parameters, and the pre-trained neural network model learns the corresponding relationship between the warping type and the production line control parameters, and feeds back to the production line control center through prediction to control the product quality.

[0067] In this embodiment, the neural network model uses a convolutional neural network model, and the neural network model includes an input layer, a hidden layer, and an output layer. The input layer receives the warping type of the paperboard as the input feature. The hidden layer contains multiple layers of neurons, which are used to extract useful information in the input feature and learn the complex relationship between the warping type and the production line control parameters. The output layer outputs the predicted production line control parameters.

[0068] Before training the neural network model, an input-output data set is established by using the recorded paperboard warping type or warping degree and production line quality control parameters (such as temperature, pressure, speed, etc.) on the corrugated paperboard production line. The collected data is preprocessed, such as cleaning and normalization, to ensure the quality and consistency of the data.

[0069] In this embodiment, the pre-trained neural network model is obtained by training the neural network model using the preprocessed paperboard production data. During the training process, the model learns the corresponding relationship between different warping types and production line control parameters, and adjusts its internal parameters (such as weights and biases) to minimize the prediction error.

[0070] Optionally, as the production process proceeds, new data can be continuously collected, and the neural network model can be continuously optimized and updated. Specifically, incremental learning, transfer learning, and other methods can be used to achieve this, in order to improve the prediction performance and adaptability of the model.

[0071] In this embodiment, the predicted production line control parameters are fed back to the production line control center in real time, and the production line control center adjusts the operation of the production line according to these parameters, such as adjusting the temperature, pressure, speed, etc., to realize accurate control of the paperboard quality.

[0072] The embodiment provides a corrugated paperboard warping classification and process control method, which realizes rapid and accurate evaluation of paperboard quality by accurately collecting vertical surface curve profile images of paperboard, intelligently judging paperboard types and calculating warping convexity, when warping problems are found, can automatically classify warping types, and output targeted production line control parameters according to a pre-trained neural network model, so as to timely adjust production, ensure stable paperboard quality, and improve overall production efficiency and intelligent management level.

[0073] The above describes the corrugated paperboard warping classification and process control method in the embodiment of the application, and the device in the embodiment of the application is described below, please refer to Figure 2 The corrugated paperboard warping classification and process control system in the embodiment of the application is used to execute the corrugated paperboard warping classification and process control method, and the implementation mode of the corrugated paperboard warping classification and process control system includes:

[0074] The acquisition module 201 is used to collect vertical surface curve profile images of the production paperboard by using a line structured light sensor, and judge the paperboard type of the production paperboard according to the vertical surface curve profile images;

[0075] The calculation module 202 is used to calculate the warping convexity of the production paperboard based on the vertical surface curve profile images when the type of the production paperboard is a non-L-shaped paperboard;

[0076] The judgment module 203 is used to determine that the production paperboard quality is qualified when the absolute value of the warping convexity is not greater than a warping preset threshold value, and determine and output the warping type of the production paperboard according to the value of the warping convexity and the warping convexity preset range value corresponding to different warping types when the absolute value of the warping convexity is greater than the warping preset threshold value;

[0077] The control parameter output module 204 is used to input the warping type of the production paperboard into a pre-trained neural network model to obtain production line control parameters output by the pre-trained neural network model, and output the production line control parameters to a production line control center to adjust and control the paperboard quality.

[0078] In this embodiment, by accurately collecting the vertical surface curve image of the paperboard, the type of the paperboard is intelligently judged and the warping convexity is calculated, the rapid and accurate evaluation of the paperboard quality is realized, when the warping problem is found, the warping type can be automatically classified, and the production line control parameters are output according to the pre-trained neural network model, so that the production is timely adjusted, the paperboard quality is ensured to be stable, and the overall production efficiency and intelligent management level are improved.

[0079] Figure 2 The structure of the illustrated paperboard warping classification and process control system at the linerboard line outlet does not constitute a limitation on the paperboard warping classification and process control system at the linerboard line outlet, and the steps of the paperboard warping classification and process control method provided in each method embodiment can be implemented.

[0080] The above Figure 2 The paperboard warping classification and process control system at the linerboard line outlet in the embodiment of the application is described in detail from the perspective of modular functional entities, and the paperboard warping classification and process control device in the embodiment of the application is described in detail from the perspective of hardware processing.

[0081] Figure 3 is a structural schematic diagram of a paperboard warping classification and process control device provided by the embodiment of the application. The device 300 can have great differences due to different configurations or performances, and can include one or more processors (central processing units, CPUs) 310 (for example, one or more processors) and a memory 320, one or more storage media 330 (for example, one or more mass storage devices) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 can be temporary storage or persistent storage. The programs stored in the storage medium 330 can include one or more modules (not shown in the figure), and each module can include a series of instruction operations in the device 300. Further, the processor 310 can be configured to communicate with the storage medium 330 and execute a series of instruction operations in the storage medium on the device 300.

[0082] The device 300 can also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc.

[0083] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system or device, unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0084] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0085] The above-described embodiments are merely used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of linerboard warp sorting and process control at a linerboard exit, characterized by, The paperboard warping classification and process control method comprises: using a line structured light sensor to collect a vertical plane curve profile image of the production paperboard, and determining the type of the production paperboard according to the vertical plane curve profile image; when the type of the production paperboard is a non-L-shaped paperboard, calculating the warping convexity of the production paperboard based on the vertical plane curve profile image; if the absolute value of the warping convexity is not greater than a preset warping threshold, determining that the quality of the production paperboard is qualified, and if the absolute value of the warping convexity is greater than the preset warping threshold, determining and outputting the warping type of the production paperboard according to the value of the warping convexity and the preset warping convexity range value corresponding to different warping types; inputting the warping type of the production paperboard into a pre-trained neural network model to obtain production line control parameters output by the pre-trained neural network model, and outputting the production line control parameters to a production line control center for adjusting and controlling the quality of the paperboard; the method of using a line structured light sensor to collect a vertical plane curve profile image of the production paperboard, and determining the type of the production paperboard according to the vertical plane curve profile image, comprises:

2. The web line exit paperboard warp sorting and process control method of claim 1 wherein, using a line structured light sensor to collect a vertical plane curve profile image of the production paperboard, and determining the type of the production paperboard according to the vertical plane curve profile image, comprises: collecting a vertical plane reflection light image of the production paperboard multiple times using a line structured light sensor, and pre-processing the collected vertical plane reflection light image to obtain multiple pre-processed images; extracting and integrating the vertical plane curve profile image of the production paperboard from the multiple pre-processed images; performing data fitting on the extracted vertical plane curve profile image to obtain a one-variable cubic function of curve profile horizontal pixel coordinates-curve profile vertical pixel coordinates; 3. The web line exit paperboard warp sorting and process control method of claim 2 wherein, calculating the number of extreme points of the one-variable cubic function, when the number of extreme points of the one-variable cubic function is 0, determining that the production paperboard is a normal paperboard, when the number of extreme points of the one-variable cubic function is 1, determining that the production paperboard is a non-S-shaped paperboard, when the number of extreme points of the one-variable cubic function is 2, determining that the production paperboard is an S-shaped paperboard, and issuing an alarm to remind the staff; when the production paperboard is a non-S-shaped paperboard, calculating the difference of the two sides derivative of the production paperboard based on the one-variable cubic function; when the difference exceeds a preset derivative range value, determining that the production paperboard is an L-shaped paperboard, and issuing an alarm to remind, otherwise, determining that the production paperboard is a non-L-shaped paperboard. collecting a vertical plane reflection light image of the production paperboard multiple times using a line structured light sensor, and pre-processing the collected vertical plane reflection light image to obtain multiple pre-processed images, comprising: projecting structured light onto the vertical plane of the production paperboard; collecting a vertical plane reflection light image of the production paperboard multiple times using a line structured light sensor; pre-processing the collected vertical plane reflection light image to obtain multiple pre-processed images. extracting and integrating the vertical plane curve profile image of the production paperboard from the multiple pre-processed images, comprising: extracting the light stripes in the multiple pre-processed images to obtain multiple groups of light stripes; mapping the position of each group of light stripes to a space three-dimensional coordinate using camera calibration parameters; after conversion and alignment of the space three-dimensional coordinates of the multiple groups of light stripes, fusing to obtain a vertical plane three-dimensional profile of the production paperboard; processing the vertical plane three-dimensional profile of the production paperboard to obtain the vertical plane curve profile image of the production paperboard.

4. The web line exit paperboard warp sorting and process control method of claim 2 wherein, A monomial cubic function of the curve profile transverse pixel coordinate-curve profile longitudinal pixel coordinate is represented as: wherein represents the horizontal pixel coordinate of the vertical plane curve profile, represents the vertical pixel coordinate of the vertical plane curve profile at the corresponding position, represents the fitting constant.

5. The wafer line exit paperboard warp sorting and process control method of claim 4 wherein, The difference between the derivatives of the two sides of the produced paperboard is represented as: wherein represents the horizontal pixel coordinate of the vertical plane curve profile, is the horizontal pixel coordinate of all pixels, represents the number of all pixel coordinate points, represents the function value of the derivative function at the position, is the point of .

6. The wafer line exit paperboard warp sorting and process control method of claim 5 wherein, When the type of the production paperboard is the non-L type paperboard, a warping convexity of the production paperboard is calculated based on the vertical plane curve profile image, comprising: When the type of the production paperboard is the non-L type paperboard, a data fitting is performed based on the vertical plane curve profile image to obtain a monomial quadratic function of the vertical plane curve profile transverse pixel coordinate-vertical plane curve profile longitudinal pixel coordinate, and the monomial quadratic function relationship is: wherein represents the vertical plane curve profile's horizontal pixel coordinate, represents the vertical plane curve profile's longitudinal pixel coordinate, position's vertical plane curve profile, represents a constant; According to the monomial quadratic function, the following formula is used to calculate the warping convexity corresponding to the production of paperboard , warping convexity is expressed as: where C is the average of the curvatures at the three quarter points, denotes the slope, is the derivative of the vertical profile curve, denotes the rate of change of the curvature, is the second derivative of the vertical profile curve, and denote the pixel coordinates of the two end points of the horizontal pixel coordinate of the quadratic function in one variable, respectively, denote the pixel coordinates of the three quarter points of the horizontal pixel coordinate of the quadratic function in one variable, respectively, 。 7. The wafer line exit paperboard warp sorting and process control method of claim 1 wherein, The warping type of the production paperboard is input into the pre-trained neural network model to obtain the production line control parameter output by the pre-trained neural network model, and the production line control parameter is output to the production line control center for adjustment and control of the paperboard quality, comprising: An input-output data set is established through the paperboard warping type or the warping degree and the production line quality control parameter recorded on the corrugated paperboard production line, a neural network model is used to train the input-output data set, a pre-trained neural network model is obtained, and the pre-trained neural network model establishes a corresponding relationship between the production line paperboard warping type and the production line control parameter; The warping type of the production paperboard is input into the pre-trained neural network model to obtain the production line control parameter output by the pre-trained neural network model; The production line control parameter is output to the production line control center for adjustment and control of the paperboard quality.

8. A linerboard warp sorting and process control system at the linerboard exit characterized by, Comprising: The acquisition module is used for acquiring a vertical plane curve profile image of the production paperboard, and judging the paperboard type of the production paperboard according to the vertical plane curve profile image, specifically comprising: using a line structured light sensor to acquire a vertical plane reflected light image of the production paperboard multiple times, and pre-processing the acquired vertical plane reflected light image to obtain multiple pre-processed images; extracting and integrating the vertical plane curve profile image of the production paperboard from the multiple pre-processed images; performing data fitting according to the extracted vertical plane curve profile image to obtain a monomial cubic function of the curve profile transverse pixel coordinate-curve profile longitudinal pixel coordinate; calculating the number of extreme points of the monomial cubic function, when the number of extreme points of the monomial cubic function is 0, determining that the production paperboard is a normal paperboard, when the number of extreme points of the monomial cubic function is 1, determining that the production paperboard is a non-S type paperboard, when the number of extreme points of the monomial cubic function is 2, determining that the production paperboard is an S type paperboard, and issuing an alarm to remind the staff; when the production paperboard is a non-S type paperboard, calculating the difference of the derivatives on both sides of the production paperboard based on the monomial cubic function; when the difference exceeds the preset range value of the derivative, it is determined that the production paperboard is an L type paperboard, and an alarm is issued to remind, otherwise, it is determined that the production paperboard is a non-L type paperboard; The calculation module is used for calculating a warping convexity of the production paperboard based on the vertical plane curve profile image when the type of the production paperboard is the non-L type paperboard; The judgment module is used for determining and outputting the warping type of the production paperboard according to the value of the warping convexity and the warping convexity preset range value corresponding to different warping types when the absolute value of the warping convexity is greater than the warping preset threshold value; A process control module is configured to input a warp type of a produced paperboard into a pre-trained neural network model to obtain a production line control parameter output by the pre-trained neural network model, and output the production line control parameter to a production line control center to adjust and control a paperboard quality.

9. A linerboard warp sorting and process control apparatus for a linerboard machine, characterized by, comprising a memory and at least one processor, the memory having computer readable instructions stored therein; The at least one processor invokes the computer readable instructions in the memory to perform the steps of the method of claim 1-7.

Images