A strip plate shape classification method, device, computer medium and equipment

Abstract

The application mainly relates to a strip plate shape classification method, characterized in that the method comprises the following steps: acquiring actual plate shape deviation values of each detection area of a reference strip and acquiring process parameter data of each detection area of the reference strip; based on the actual plate shape deviation values, determining plate shape defect types corresponding to each detection area of the strip, taking the plate shape defect types as training label data, taking the process parameter data as training feature data, and obtaining multiple groups of training sample data; constructing an initial plate shape classification model, training the initial plate shape classification model based on the multiple groups of training sample data, and obtaining a plate shape classification model; and based on process parameter data of a target area in a to-be-detected strip, determining a plate shape defect type of the target area in the to-be-detected strip through the plate shape classification model. The application can solve the problems of experience-based errors and omissions in manual strip plate shape classification to a certain extent.

Description

Technical Field

[0001] This application relates to the field of cold-rolled strip shape control, and more specifically, to a strip shape classification method, apparatus, computer medium, and device. Background Technology

[0002] In the production of cold-rolled strip steel, shape pattern recognition of the rolled strip is a crucial component of the cold continuous rolling strip shape control system. Under current technology, manual classification of strip steel is prone to problems such as misjudgment and omissions due to empiricism. Therefore, those skilled in the art urgently need a method to solve these empiricism problems associated with manual judgment. Summary of the Invention

[0003] The embodiments of this application provide a method, apparatus, computer medium, and device for classifying the shape of steel strips, which can, to a certain extent, solve the empirical problems such as misjudgment and omission in manual classification of steel strip shapes.

[0004] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.

[0005] According to one aspect of this application, a strip steel shape classification method is provided, the method comprising: acquiring actual shape deviation values ​​in various detection areas of a reference strip steel, and acquiring process parameter data in various detection areas of the reference strip steel; determining the shape defect type corresponding to each detection area of ​​the strip steel based on the actual shape deviation values, and obtaining multiple sets of training sample data by using the shape defect type as training label data and the process parameter data as training feature data; constructing an initial shape classification model, and training the initial shape classification model based on the multiple sets of training sample data to obtain a shape classification model; and determining the shape defect type of the target area in the strip steel to be inspected by using the shape classification model based on the process parameter data of the target area in the strip steel to be inspected.

[0006] In some embodiments of this application, the actual shape deviation value of each detection area includes the actual shape deviation value corresponding to each detection point within each detection area. The method for obtaining the actual shape deviation value of each detection area in the reference strip includes: obtaining the actual residual stress of each detection point within each detection area of ​​the reference strip; calculating the average value of the actual residual stress of each detection area based on the actual residual stress of each detection point; and calculating the actual shape deviation value of each detection point based on the actual residual stress of each detection point and the average value of the actual residual stress of each detection area.

[0007] In some embodiments of this application, the number of actual shape deviation values ​​for each detection area of ​​the reference strip is multiple. The method for determining the shape defect type corresponding to each detection area of ​​the strip based on the actual shape deviation values ​​includes: obtaining a preset shape deviation model, which includes curve models corresponding to each shape defect type and defect coefficients matching each curve model; for each detection area, fitting the preset shape deviation model based on multiple actual shape deviation values ​​to obtain the values ​​of each defect coefficient, which are used as defect coefficient values ​​corresponding to each shape defect type; obtaining shape deviation weights corresponding to each shape defect type, which are used to characterize the average shape deviation corresponding to each shape defect type; productting the defect coefficient values ​​and shape deviation weights corresponding to each shape defect type to obtain a defect reference value corresponding to each shape defect type, which is used to characterize the severity of the corresponding defect type in each detection area; and determining the shape defect type corresponding to each detection area of ​​the strip based on the defect reference values.

[0008] In some embodiments of this application, the method for determining the plate shape defect type corresponding to each detection area of ​​the strip steel based on the defect reference value includes: for each detection area, if there is a defect reference value higher than a predetermined threshold, then the plate shape defect type corresponding to the defect coefficient value with the largest absolute value is determined as the plate shape defect type corresponding to the detection area; among the multiple defect reference values, if there is no defect reference value higher than the predetermined threshold, then it is determined that the detection area has no plate shape defect.

[0009] In some embodiments of this application, the method for constructing an initial plate shape classification model includes: constructing an initial plate shape classification model based on a GBDT model, wherein the GBDT model includes at least one decision tree, and the decision tree is a classification tree.

[0010] In some embodiments of this application, the method of training the initial board shape classification model based on the multiple sets of training sample data to obtain the board shape classification model includes: dividing the multiple sets of training samples into a training set and a test set according to a preset ratio; training the initial board shape classification model using the training set to obtain the trained board shape classification model; and verifying the effect of the trained board shape classification model using the test set to obtain a board shape classification model that meets preset conditions.

[0011] In some embodiments of this application, the process parameter data based on the target area in the strip to be tested includes at least: the tension and forward slip value of the target area in the strip to be tested, as well as the rolling force, bending force and roll shifting amount acting on the target area in the strip to be tested.

[0012] According to one aspect of this application, a strip steel shape classification device is provided. The device includes: an acquisition unit, configured to acquire actual shape deviation values ​​in various detection areas of a reference strip steel, and to acquire process parameter data in various detection areas of the reference strip steel; a determination unit, configured to determine the shape defect type corresponding to each detection area of ​​the strip steel based on the actual shape deviation values, and to obtain multiple sets of training sample data using the shape defect type as training label data and the process parameter data as training feature data; a construction unit, configured to construct an initial shape classification model, and to train the initial shape classification model based on the multiple sets of training sample data to obtain a shape classification model; and a classification unit, configured to determine the shape defect type of the target area in the strip steel to be inspected based on the process parameter data of the target area in the strip steel to be inspected, using the shape classification model.

[0013] According to one aspect of this application, a computer-readable storage medium is provided, wherein at least one piece of program code is stored therein, the at least one piece of program code being loaded and executed by a processor to perform the operations performed by the strip shape classification method as described above.

[0014] According to one aspect of this application, a computer device is provided, characterized in that the computer device includes one or more processors and one or more memories, the one or more memories storing at least one piece of program code, the at least one piece of program code being loaded and executed by the one or more processors to perform the operations performed by the strip shape classification method described above.

[0015] Based on the above technical solution, this application has at least the following beneficial effects:

[0016] This application first constructs multiple sets of training sample data based on actual production conditions. By using these training samples to train an initial classification model, it combines data-driven non-mechanistic modeling with process knowledge and experience, utilizing prior knowledge to save training samples for the data-driven model. Simultaneously, the data-driven model compensates for characteristics that the original model could not interpret. This significantly improves the interpretability and application scope of the final strip shape classification model, thereby addressing, to some extent, the empirical problems of misclassification and omissions that easily occur in manual strip shape classification.

[0017] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0019] Figure 1 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown;

[0020] Figure 2 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown;

[0021] Figure 3 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown;

[0022] Figure 4 The figure shows curve model diagrams for four types of plate shape deviation;

[0023] Figure 5 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown;

[0024] Figure 6 A simplified diagram of a strip-shaped sorting device according to one embodiment of this application is shown;

[0025] Figure 7 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown. Detailed Implementation

[0026] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art.

[0027] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0028] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0029] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0030] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such uses of these terms can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described.

[0031] Next, this application will provide a detailed description of the technical solution provided in conjunction with the accompanying drawings.

[0032] Figure 1 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown. The method may include steps S101-S104:

[0033] Step S101: Obtain the actual shape deviation value of each detection area of ​​the reference strip, and obtain the process parameter data of each detection area of ​​the reference strip.

[0034] Step S102: Based on the actual strip shape deviation value, determine the strip shape defect type corresponding to each detection area of ​​the strip, and use the strip shape defect type as training label data and the process parameter data as training feature data to obtain multiple sets of training sample data.

[0035] Step S103: Construct an initial plate shape classification model, and train the initial plate shape classification model based on the multiple sets of training sample data to obtain a plate shape classification model.

[0036] Step S104: Based on the process parameter data of the target area in the strip to be inspected, determine the type of plate shape defect in the target area of ​​the strip to be inspected through the plate shape classification model.

[0037] Based on the above technical solution, this application first constructs multiple sets of training sample data according to the actual production situation. By using the training samples to train the initial classification model, the data-driven non-mechanistic modeling can be combined with process knowledge and experience. Prior knowledge is used to save training samples for the data-driven model. At the same time, the data-driven model is used to compensate for the characteristics that the original model could not explain. This greatly improves the interpretability and application scope of the final plate shape classification model.

[0038] Figure 2 A simplified flowchart of a strip shape classification method according to an embodiment of this application is shown. The actual shape deviation value of each detection area may include the actual shape deviation value corresponding to each detection point in each detection area. The method for obtaining the actual shape deviation value of each detection area of ​​the reference strip may include steps S201-S203:

[0039] Step S201: Obtain the actual residual stress of the strip at each detection point in each detection area of ​​the reference strip.

[0040] Step S202: Based on the actual residual stress of the plate at each detection point, calculate the average value of the actual residual stress of the plate in each detection area.

[0041] Step S203: Based on the actual plate shape residual stress at each detection point and the average value of the actual plate shape residual stress in each detection area, calculate the actual plate shape deviation value at each detection point.

[0042] In this application, after the strip passes through the strip shaper, the scattered values ​​of the residual stress of the strip shape at each detection point in the bandwidth direction can be obtained. The actual residual stress of the strip shape at each detection point is processed to calculate the actual strip shape deviation value, which is in IU. The physical meaning of 1 IU is that after the strip tension is released, each meter of strip will have an extension of 10 μm.

[0043] In another embodiment of this application, the specific calculation process can be as follows:

[0044]

[0045]

[0046] Where i represents the number of each detection point, Δσ(i) is the deviation of plate residual stress at each detection point within the detection area, and σ(i) is the actual plate residual stress value at each detection point; The average residual stress within the detection area; L is the ideal plate length; ΔL(i) is the plate length deviation; η(i) is the elongation, also known as the plate deviation value; E m It is the elastic modulus.

[0047] The construction of training data will now be described in detail with reference to specific embodiments and accompanying drawings.

[0048] Figure 3 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown. The method involves determining the strip shape defect type corresponding to each detection area based on the actual strip shape deviation values, and may include steps S301-S305:

[0049] Step S301: Obtain a preset plate shape deviation model, which includes curve models corresponding to each plate shape defect type and defect coefficients matching each curve model.

[0050] Step S302: For each detection area, the preset plate shape deviation model is fitted based on multiple actual plate shape deviation values ​​to obtain the values ​​of each defect coefficient, which are used as the defect coefficient values ​​corresponding to each plate shape defect type.

[0051] Step S303: Obtain the plate shape deviation weight corresponding to each plate shape defect type. The plate shape deviation weight is used to characterize the average plate shape deviation corresponding to each plate shape defect type.

[0052] Step S304: Take the product of the defect coefficient value and the plate shape deviation weight value corresponding to each plate shape defect type to obtain the defect reference value corresponding to each plate shape defect type. The defect reference value is used to characterize the severity of the corresponding defect type in each detection area.

[0053] Step S305: Based on the defect reference value, determine the plate shape defect type corresponding to each detection area of ​​the strip steel.

[0054] In this application, due to the limitations of the regulatory agency, there are eight common board pattern defects, such as left-side wave, right-side wave, middle wave, double-sided wave, left three-quarter wave, right three-quarter wave, quarter wave, and side-middle composite wave. These can be divided into four pairs of opposite board pattern defects, and the specific curve model can be:

[0055] Right-side wave: Y1 = δ1(x) = x;

[0056] Left-side wave: Y2 = -δ1(x) = -x;

[0057] Two-sided waves:

[0058] Mid-wave:

[0059] Left three-wave:

[0060] Right three-wave:

[0061] Complex Waves in the Middle and Sides:

[0062] Quarter wave:

[0063] Figure 4 The curve model diagrams for four types of plate shape deviation are shown.

[0064] For the preset plate shape deviation model, the following formula can be used: y=λ1δ1+λ2δ2+λ3δ3+λ4δ4;

[0065] λ1, λ2, λ3, and λ4 are the defect coefficients that match each curve model, x is the coordinate of the discrete point after the strip width is normalized, and y is the strip shape deviation value at the coordinate of the discrete point.

[0066] Next, based on the actual plate shape deviation values ​​of multiple discrete point coordinates in each detection area, the equation y = λ1δ1 + λ2δ2 + λ3δ3 + λ4δ4 is fitted to obtain the values ​​of each defect coefficient, which serve as the defect coefficient values ​​corresponding to each plate shape defect type. For example, if there are 80 discrete point coordinates in each detection area, and each coordinate has a corresponding actual plate shape deviation value, substituting the actual plate shape deviation values ​​of the 80 discrete points into y = λ1δ1 + λ2δ2 + λ3δ3 + λ4δ4, the specific values ​​of λ1, λ2, λ3, and λ4 can be fitted. Therefore, λ1, λ2, λ3, and λ4 can be used as the defect coefficient values ​​for the four pairs of plate shape defects mentioned above. The specific plate shape defect in a pair of plate shape defect types can be determined based on the sign of the finally obtained defect coefficient values. For example, for λ1, it can be a or -a.

[0067] When λ1 is a, a is the defect coefficient value of the right wave; when λ1 is -a, a is the defect coefficient value of the left wave.

[0068] In this embodiment, the method for obtaining the plate shape deviation weights corresponding to each plate shape defect type, wherein the plate shape deviation weights are used to characterize the average plate shape deviation corresponding to each plate shape defect type, can be as follows:

[0069] The evaluation index for plate shape quality is the average plate shape deviation, measured in IU. The larger the average plate shape deviation, the greater the plate shape defect. The calculation method is as follows:

[0070]

[0071] Where f represents the average plate shape deviation, y i This represents the plate shape deviation at the i-th position in the bandwidth direction. This represents the average value of all plate shape deviations in the bandwidth direction, and n represents the number of sampling points in the bandwidth direction.

[0072] Substituting the curve models corresponding to the four pairs of plate shape defect types into the above formula, we can obtain the four average plate shape deviations. These four average plate shape deviations are used as the plate shape deviation weights k = [0.5063, 0.3947, 0.3336, 0.2968].

[0073] Finally, the type of strip shape defect corresponding to each detection area can be determined based on the product of the strip shape deviation weight and λ1, λ2, λ3 and λ4.

[0074] In this embodiment, the method for determining the plate shape defect type corresponding to each detection area of ​​the strip steel based on the defect reference value may include:

[0075] For each detection area, if there is a defect reference value higher than a predetermined threshold, the plate shape defect type corresponding to the defect coefficient value with the largest absolute value is determined as the plate shape defect type corresponding to the detection area; if there is no defect reference value higher than the predetermined threshold among the multiple defect reference values, it is determined that the detection area has no plate shape defect.

[0076] In this embodiment, if the calculated defect reference value is greater than 3 IU, it can be determined that a plate-shaped defect has appeared in a detection area. Then, based on the maximum absolute value of the four coefficients λ1, λ2, λ3, and λ4, the plate-shaped defect type corresponding to the detection area is determined. For example, if λ1 is -5, λ2 is 4, λ3 is 3, and λ4 is 1, the plate-shaped defect type can be determined as a left-side wave according to the method described.

[0077] In some embodiments of this application, the method for constructing an initial plate shape classification model includes: constructing an initial plate shape classification model based on a GBDT model, wherein the GBDT model includes at least one decision tree, and the decision tree is a classification tree.

[0078] It's important to note that GBDT is an iterative algorithm, a gradient boosting algorithm using decision trees (CART) as base learners. Furthermore, the decision trees are regression trees, not classification trees. Each tree learns from the residuals of all previous trees, and each new training iteration aims to improve upon the previous result. The conclusions of all trees are summed to arrive at the final answer. GBDT combines weak learners to form a strong learner, giving it a natural advantage in discovering various discriminative features and feature combinations, making it an algorithm with strong generalization capabilities.

[0079] For GBDT multi-class classification problems, assuming there are K classification categories, the commonly used loss function is:

[0080]

[0081] Among them, y k Let y be a sign function, and let y be a sign function if and only if the sample is classified as k. k =1; p k f(x) represents the probability that model f(x) determines x to belong to the k-th class, i.e., the softmax function, whose expression is:

[0082]

[0083] For multi-class classification problems, GBDT employs a "one-to-many" strategy during training. GBDT converts the output class into a one-hot vector, and a regression tree is built for each class in each training round. Therefore, when GBDT training is complete, there should be M*K regression trees. Furthermore, the training loop order is to fit the first set of trees for each of the K classes before proceeding to the second round of K-class fitting; it is not possible to train M trees for one class before learning another. Therefore, the derivation formula for the negative gradient error of the class k corresponding to the i-th sample in the m-th round can be derived.

[0084]

[0085] Differentiation yields:

[0086]

[0087] From the above, we can directly derive the negative gradient error of the i-th sample in the m-th round corresponding to class k:

[0088]

[0089] As can be seen from the above formula, the error here is actually the difference between the true probability of sample i corresponding to category k and the predicted probability in m-1 rounds.

[0090] The method for calculating the tree nodes of the generated decision tree is as follows:

[0091]

[0092] Where φ(y) k ,f k (x))=-y k logp k (x). Expanding using Taylor's formula and substituting the derivative with an approximation, we get:

[0093]

[0094] At this point, the k-th class is initialized, let φ′(y i The initial value can be obtained by setting c) = 0.

[0095]

[0096] This completes the derivation of the GBDT multi-class classification algorithm. In summary, the steps of the GBDT multi-class classification algorithm with the log-likelihood function as the loss function can be shown in Table 1:

[0097] Table 1: Algorithm steps of GBDT multi-class classification

[0098]

[0099]

[0100] The maximum tree depth for this GBDT can be 12 layers, the minimum number of samples required for further partitioning of internal nodes can be 30, and the maximum number of features can be 300.

[0101] Figure 5 A simplified flowchart of a strip shape classification method according to one embodiment of this application is shown. The method of training the initial strip shape classification model based on the multiple sets of training sample data to obtain the strip shape classification model may include steps S501-S503:

[0102] Step S501: Divide the multiple sets of training samples into a training set and a test set according to a preset ratio.

[0103] Step S502: Train the initial plate shape classification model using the training set to obtain the trained plate shape classification model.

[0104] Step S503: Verify the effect of the trained board shape classification model using the test set to obtain a board shape classification model that meets the preset conditions.

[0105] In this application, considering the on-site process requirements, a value less than 3 IU is considered to have no plate shape defect; if the value is greater than 3 IU, the plate shape is classified into labels [-4, -3, -2, -1, 0, 1, 2, 3, 4] based on its corresponding plate shape defect coefficient, corresponding to quarter waves, right third waves, middle waves, left waves, no waves, right waves, double-sided waves, left third waves, and edge-middle composite waves, respectively. 80% of the data samples are used as the training set, and 20% as the test set. Upsampling and downsampling are performed to address uneven distribution of the training set samples. The final distribution of positive and negative samples in the multi-class data is shown in Table 2. Due to data limitations, edge-middle waves, double-sided waves, left third waves, and right third waves are not included.

[0106] Table 2 Data Sample Distribution Table

[0107] Label -4 -2 -1 0 1 training set 4400 1960 328 8230 128 test set 123 110 90 1138 123 total 4523 2070 418 9368 251

[0108] In some embodiments of this application, the process parameter data based on the target area in the strip to be tested may include at least: the tension and forward slip value of the target area in the strip to be tested, as well as the rolling force, bending force and roll shifting amount acting on the target area in the strip to be tested.

[0109] Next, an embodiment of the apparatus of this application will be described with reference to the accompanying drawings.

[0110] Figure 6 A simplified diagram of a steel plate-shaped sorting device according to one embodiment of this application is shown. The device includes: an acquisition unit 601, a determination unit 602, a construction unit 603, and a sorting unit 604, with the following specific configuration:

[0111] The acquisition unit 601 is used to acquire the actual shape deviation value of each detection area of ​​the reference strip, as well as the process parameter data of each detection area of ​​the reference strip.

[0112] The determination unit 602 is used to determine the type of strip shape defect corresponding to each detection area of ​​the strip based on the actual strip shape deviation value, and to obtain multiple sets of training sample data by using the strip shape defect type as training label data and the process parameter data as training feature data.

[0113] The construction unit 603 is used to construct an initial plate shape classification model and train the initial plate shape classification model based on the multiple sets of training sample data to obtain a plate shape classification model.

[0114] The classification unit 604 is used to determine the type of plate shape defect in the target area of ​​the strip under test based on the process parameter data of the target area in the strip under test through the plate shape classification model.

[0115] Figure 7 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown.

[0116] It should be noted that, Figure 7 The computer system 700 of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0117] like Figure 7As shown, the computer system 700 includes a Central Processing Unit (CPU) 701, which can perform various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) 702 or programs loaded from storage portion 708 into Random Access Memory (RAM) 703, such as performing the methods described in the above embodiments. The RAM 703 also stores various programs and data required for system operation. The CPU 701, ROM 702, and RAM 703 are interconnected via a bus 704. An Input / Output (I / O) interface 705 is also connected to the bus 704.

[0118] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0119] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 709, and / or installed from removable medium 711. When the computer program is executed by central processing unit (CPU) 701, it performs various functions defined in the system of this application.

[0120] It should be noted that the computer-readable medium shown in the embodiments of this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such transmitted data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.

[0121] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0122] The units described in the embodiments of this application can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the specific unit itself.

[0123] In another aspect, this application also provides a computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the first aspect or various optional implementations thereof.

[0124] In another aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.

[0125] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0126] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, touch terminal, or network device, etc.) to execute the method according to the embodiments of this application.

[0127] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0128] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for classifying strip steel plates, characterized in that, The method includes: Obtain the actual shape deviation values ​​in each inspection area of ​​the reference strip, and obtain the process parameter data in each inspection area of ​​the reference strip; Obtain a preset plate shape deviation model, which includes curve models corresponding to each plate shape defect type and defect coefficients matching each curve model; For each detection area, the preset plate shape deviation model is fitted based on multiple actual plate shape deviation values ​​to obtain the values ​​of each defect coefficient, which are used as the defect coefficient values ​​corresponding to each plate shape defect type. Obtain the plate shape deviation weights corresponding to each plate shape defect type, wherein the plate shape deviation weights are used to characterize the average plate shape deviation corresponding to each plate shape defect type; The defect coefficient value and the plate shape deviation weight value corresponding to each plate shape defect type are multiplied to obtain the defect reference value corresponding to each plate shape defect type. The defect reference value is used to characterize the severity of the corresponding defect type in each detection area. Based on the defect reference value, the plate shape defect type corresponding to each detection area of ​​the strip is determined; Using the plate shape defect type as training label data and the process parameter data as training feature data, multiple sets of training sample data are obtained; An initial board shape classification model is constructed, and the initial board shape classification model is trained based on the multiple sets of training sample data to obtain the board shape classification model; Based on the process parameter data of the target area in the strip to be inspected, the type of plate shape defect in the target area of ​​the strip to be inspected is determined by the plate shape classification model.

2. The method according to claim 1, characterized in that, The actual shape deviation value of each detection area includes the actual shape deviation value corresponding to each detection point in each detection area. Obtaining the actual shape deviation value of each detection area in the reference strip includes: Obtain the actual residual stress of the strip at each detection point within each detection area of ​​the reference strip; Based on the actual residual stress of the plate at each detection point, calculate the average value of the actual residual stress of the plate in each detection area; Based on the actual residual stress of the plate at each detection point and the average value of the actual residual stress of the plate in each detection area, the actual plate shape deviation value of each detection point is calculated.

3. The method according to claim 1, characterized in that, The step of determining the type of plate shape defect corresponding to each detection area of ​​the strip steel based on the defect reference value includes: For each detection area, if there is a defect reference value higher than a predetermined threshold, the plate shape defect type corresponding to the defect coefficient value with the largest absolute value is determined as the plate shape defect type corresponding to the detection area. If there is no defect reference value higher than a predetermined threshold among the multiple defect reference values, then it is determined that there is no plate-shaped defect in the detection area.

4. The method according to claim 1, characterized in that, The construction of the initial plate shape classification model includes: An initial plate-shaped classification model is constructed based on the GBDT model, wherein the GBDT model includes at least one decision tree, and the decision tree is a classification tree.

5. The method according to claim 1, characterized in that, The step of training the initial plate shape classification model based on the multiple sets of training sample data to obtain the plate shape classification model includes: According to a preset ratio, the multiple sets of training samples are divided into a training set and a test set; The initial board shape classification model is trained using the training set to obtain the trained board shape classification model; The effectiveness of the trained board shape classification model is verified using the test set to obtain a board shape classification model that meets the preset conditions.

6. The method according to claim 1, characterized in that, The process parameter data based on the target area in the strip to be tested includes at least: the tension and forward slip value of the target area in the strip to be tested, as well as the rolling force, bending force and roll shifting amount acting on the target area in the strip to be tested.

7. A strip steel plate-shaped sorting device, implemented based on the method as described in claim 1, characterized in that, The device includes: The acquisition unit is used to acquire the actual shape deviation values ​​in each detection area of ​​the reference strip, as well as the process parameter data in each detection area of ​​the reference strip. The determination unit is used to determine the type of strip shape defect corresponding to each detection area of ​​the strip based on the actual strip shape deviation value, and to obtain multiple sets of training sample data by using the strip shape defect type as training label data and the process parameter data as training feature data. The construction unit is used to construct an initial board shape classification model, and to train the initial board shape classification model based on the multiple sets of training sample data to obtain a board shape classification model; The classification unit is used to determine the type of plate shape defect in the target area of ​​the strip steel to be inspected based on the process parameter data of the target area in the strip steel to be inspected, through the plate shape classification model.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one piece of program code, which is loaded and executed by a processor to perform the operations performed by the strip shape classification method as described in any one of claims 1 to 6.

9. A computer device, characterized in that, The computer device includes one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, and the at least one piece of program code is loaded and executed by the one or more processors to perform the operations performed by the strip shape classification method as described in any one of claims 1 to 6.

Images