Transformer silicon steel sheet separation method, device and equipment

Abstract

The present invention discloses a transformer silicon steel sheet separation method, device and equipment. When a separation device separates the silicon steel sheets of a transformer, a real-time image and a real-time separation force are obtained during the separation process, thereby obtaining a quality impact area on the silicon steel sheets during the separation process, and then obtaining a quality impact parameter for evaluating the quality impact area; and the quality impact parameter is input into a corresponding separation force prediction model to obtain a reference separation force, so that when the real-time separation force does not match the reference separation force, the real-time separation force is adjusted. Since the separation force of the separation device is adjusted in time during the separation process of the silicon steel sheets, the separation force applied in real time meets the standard separation force, thereby reducing the adverse effects on the silicon steel sheets during the separation process.

Description

Technical Field

[0001] The present invention relates to the technical field of power grids, and in particular to a method, device and equipment for separating silicon steel sheets of a transformer. Background Art

[0002] In the green disassembly and recycling process of distribution transformers, the control of the separation force of the core laminations is a key technical problem. The separation force applied to the core laminations directly affects the reuse value of the silicon steel sheets. If the separation force is too small, the insulating material between the laminations is difficult to remove effectively, and the residual insulating material will affect the reuse of the silicon steel sheets; if the separation force is too large, the silicon steel sheets will be deformed and the surface coating will be damaged, reducing the reuse value of the silicon steel sheets. Therefore, how to accurately control the separation force to maximize the integrity of the silicon steel sheets while removing the insulation is a technical problem that needs to be overcome urgently. Although the commonly used chemical immersion method can effectively remove insulation, it is easy to cause corrosion to the surface coating of the silicon steel sheets. Therefore, in order to avoid corrosion to the surface coating of the silicon steel sheets during the disassembly process, the existing technology often uses a mechanical stripping method. Although the damage to the silicon steel sheets is small, the separation device uses fixed operating parameters during the stripping process of the silicon steel sheets, which is prone to problems such as deformation of the silicon steel sheets and damage to the surface coating, thereby affecting the reuse value of the silicon steel sheets. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide a transformer silicon steel sheet separation method, device and equipment. During the separation process of the silicon steel sheets, the separation force of the separation device is adjusted in time so that the separation force applied in real time meets the standard separation force, thereby reducing the adverse effects on the silicon steel sheets during the separation process.

[0004] To achieve the above object, an embodiment of the present invention provides a transformer silicon steel sheet separation method, comprising:

[0005] Obtain real-time images and real-time separation force of the transformer silicon steel sheets during separation;

[0006] Performing image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet;

[0007] generating a quality impact parameter using the quality impact region;

[0008] Inputting the mass influencing parameter into a separation force prediction model corresponding to the model of the transformer to obtain a corresponding reference separation force;

[0009] When the real-time separation force does not match the reference separation force, the real-time separation force is adjusted.

[0010] As an improvement of the above solution, the method for constructing the separation force prediction model includes:

[0011] Applying a plurality of preset separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence;

[0012] After applying the preset separation force each time, obtaining a surface image of the silicon steel sheet;

[0013] Analyzing the surface image to obtain quality influencing parameters of the silicon steel sheet corresponding to different preset separation forces;

[0014] For each preset separation force, the minimum value among the mass influencing parameters is taken as the standard mass influencing parameter;

[0015] A separation force prediction model between the standard quality influencing parameter and the preset separation force is constructed; wherein different types of transformers correspond to a separation force prediction model respectively.

[0016] As an improvement of the above solution, the quality influencing parameters include the deformation degree value, coating shedding degree value and insulating material residual degree value of the silicon steel sheet; then, the separation force prediction model includes:

[0017] A first mathematical model of the deformation degree value and the first separation force;

[0018] A second mathematical model of the coating shedding degree value and the second separation force;

[0019] A third mathematical model of the insulating material residual degree value and the third separation force.

[0020] As an improvement of the above solution, the number of the reference separation forces is the same as the number of the separation force prediction models; then, adjusting the real-time separation force includes:

[0021] determining a target separation force according to the reference separation force;

[0022] The real-time separation force is adjusted based on the target separation force; wherein the target separation force is determined in one of the following ways:

[0023] Taking any value of the reference separation force as the target separation force;

[0024] Taking the average value of the reference separation force as the target separation force;

[0025] The middle value of the reference separation force is taken as the target separation force.

[0026] As an improvement of the above solution, when the real-time separation force does not match the reference separation force, the method further includes:

[0027] Recording the continuous cumulative number of times that the real-time separation force does not match the reference separation force;

[0028] When the continuous accumulated number of times is greater than a preset number threshold, adjusting the position of the separation device;

[0029] When the continuous accumulated number of times is less than or equal to the number threshold, the current position of the separation device is kept unchanged.

[0030] As an improvement of the above solution, after adjusting the real-time separation force, the method further includes:

[0031] When it is detected that all silicon steel sheets of the current transformer have been separated, a multivariate monitoring data set of the transformer during the silicon steel sheet separation process is obtained;

[0032] When the multivariate monitoring data set meets a preset data standard, storing the multivariate monitoring data set;

[0033] All stored multivariate monitoring data sets are analyzed regularly to obtain at least one standard monitoring data set corresponding to each transformer model; wherein the standard monitoring data set is used for initial program setting of the separation device when the silicon steel sheets of the transformer are separated next time.

[0034] As an improvement of the above scheme, the multivariate monitoring data set includes operating parameters of the separation device and quality influencing parameters of each silicon steel sheet after separation; wherein the operating parameters include operating power, real-time separation force, position, separation speed and separation time.

[0035] As an improvement to the above solution, the periodic analysis of all stored multivariate monitoring data sets includes:

[0036] For each type of transformer, the stored multivariate monitoring data set belonging to the same type is sequentially input into the preset target optimization model;

[0037] The multivariate monitoring data set corresponding to the minimum value output by the target optimization model is used as the standard monitoring data set.

[0038] As an improvement to the above solution, the periodic analysis of all stored multivariate monitoring data sets includes:

[0039] Inputting the multivariate monitoring data set into a pre-trained quality prediction model so that the quality prediction model outputs a corresponding quality grade;

[0040] When the quality level meets the preset standard level, the multivariate monitoring data set corresponding to the quality level is determined to be the standard monitoring data set.

[0041] As an improvement to the above solution, before regularly analyzing all stored multivariate monitoring data sets, the method further includes:

[0042] Determine the resource recovery rate of the current transformer after completing the separation of silicon steel sheets and the energy loss of the separation device;

[0043] When the resource recovery rate is greater than or equal to a preset resource recovery threshold, and the lost energy consumption is less than or equal to a preset energy consumption threshold, retaining the multivariate monitoring data set corresponding to the current transformer;

[0044] When the resource recovery rate is less than the resource recovery threshold, or the lost energy consumption is greater than the energy consumption threshold, the multivariate monitoring data set corresponding to the current transformer is deleted.

[0045] To achieve the above purpose, an embodiment of the present invention further provides a transformer silicon steel sheet separation device, comprising:

[0046] A data acquisition module, used to obtain real-time images and real-time separation force of silicon steel sheets of the transformer during separation;

[0047] A quality impact area acquisition module, used for performing image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet;

[0048] A quality impact parameter generation module, used to generate quality impact parameters using the quality impact area;

[0049] A reference separation force determination module, used for inputting the quality influencing parameter into a separation force prediction model corresponding to the model of the transformer to obtain a corresponding reference separation force;

[0050] The separation force adjustment module is used to adjust the real-time separation force when the real-time separation force does not match the reference separation force.

[0051] To achieve the above objectives, an embodiment of the present invention further provides a transformer silicon steel sheet separation device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the transformer silicon steel sheet separation method as described in any of the above embodiments is implemented.

[0052] Compared with the prior art, the transformer silicon steel sheet separation method, device and equipment disclosed in the present invention pre-establishes a mathematical model of silicon steel sheet separation force and deformation, coating integrity, and insulation material retention, and combines machine vision technology to collect separation process images in real time and perform regional identification, thereby obtaining the quality influencing parameter as an evaluation index, and then calculates the reference separation force and compares it with the real-time separation force, and dynamically adjusts the real-time separation force. During the process of separating silicon steel sheets, the real-time separation force of the separation device is adjusted in time, so that the separation force applied by the separation device in real time meets the standard separation force, thereby reducing the adverse effects on the silicon steel sheets during the separation process. In addition, by collecting multivariate monitoring data of transformers of different models during the silicon steel sheet separation process, the better multivariate monitoring data is selected for the initial program setting of the next transformer of the same model when separating silicon steel sheets, so that the separation device is as consistent as possible with expectations during the initial operation, reducing the adverse effects on the silicon steel sheets in the initial separation stage, thereby reducing the number of subsequent adjustments to the separation force, and improving the separation efficiency of silicon steel sheets. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of a transformer silicon steel sheet separation method provided by an embodiment of the present invention;

[0054] Figure 2 is a flow chart of a method for constructing a first mathematical model provided by an embodiment of the present invention;

[0055] Figure 3 is a flow chart of a method for constructing a second mathematical model provided by an embodiment of the present invention;

[0056] Figure 4 is a flow chart of a method for constructing a third mathematical model provided by an embodiment of the present invention;

[0057] Figure 5 is another flow chart of a transformer silicon steel sheet separation method provided by an embodiment of the present invention;

[0058] Figure 6 This is a structural block diagram of a transformer silicon steel sheet separation device provided by an embodiment of the present invention;

[0059] Figure 7 It is a structural block diagram of a transformer silicon steel sheet separation device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] A transformer is a device that uses the principle of electromagnetic induction to change AC voltage. Its main components are the primary coil, the secondary coil and the iron core (magnetic core). The iron core is the main magnetic circuit part of the transformer. The iron core is usually made of hot-rolled or cold-rolled silicon steel sheets with a high silicon content and a coating coated with insulating paint on the surface, and the silicon steel sheets need to be staggered and overlapped. In addition, in order to further ensure the insulation of the silicon steel sheets between the stacks, insulating materials need to be added between the silicon steel sheets. When recycling the transformer for secondary use, it is necessary to use a separation device to separate the silicon steel sheets of the transformer. The separation force applied by the separation device to the silicon steel sheet stack directly affects the reuse value of the silicon steel sheets. In order to reduce the adverse effects on the silicon steel sheets during the separation process, an embodiment of the present invention provides a transformer silicon steel sheet separation method.

[0062] See also Figure 1 , Figure 1 1 is a flow chart of a transformer silicon steel sheet separation method provided by an embodiment of the present invention, wherein the transformer silicon steel sheet separation method comprises:

[0063] S1. Obtaining real-time images and real-time separation force of silicon steel sheets of transformer during separation process;

[0064] S2, performing image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet;

[0065] S3, generating quality impact parameters using the quality impact area;

[0066] S4, inputting the quality influencing parameter into a separation force prediction model corresponding to the model of the transformer to obtain a corresponding reference separation force;

[0067] S5. When the real-time separation force does not match the reference separation force, adjust the real-time separation force.

[0068] Exemplarily, the quality influencing parameters include the deformation degree value, coating shedding degree value and insulating material residual degree value of the silicon steel sheet. Real-time images of the silicon steel sheet separation process of the transformer are collected in real time by machine vision technology, and the deformation area, surface coating shedding area and insulating material residual area of ​​the silicon steel sheet are identified and located by image segmentation algorithm to obtain the quality influencing area of ​​the silicon steel sheet, calculate the area proportion of each area, and obtain the three quality influencing parameters of the silicon steel sheet deformation degree value, coating shedding degree value and insulating material residual degree value. Then these three quality influencing parameters are respectively input into the separation force prediction model corresponding to the transformer model to obtain the corresponding reference separation force. If the real-time separation force matches the reference separation force, it means that the real-time separation force applied by the current separation device is reasonable, and the deformation, surface coating shedding and insulating material residual caused to the silicon steel sheet are also within the normal range. At this time, there is no need to adjust the real-time separation force; if the real-time separation force does not match the reference separation force, it means that the real-time separation force applied by the current separation device is unreasonable, and the real-time separation force needs to be adjusted in time. It can be adjusted according to the reference separation force so that the adjusted real-time separation force matches the reference separation force.

[0069] It should be noted that since the silicon steel sheets are stacked together, and since the silicon steel sheets of the same transformer are basically the same in size, thickness, coating amount, and insulating material amount (stacked according to the same procedure), during the silicon steel sheet separation process, after each silicon steel sheet is separated, the separated silicon steel sheet can be image recognized based on the acquired real-time image. It can be understood that for silicon steel sheets other than the first layer, silicon steel sheets are stacked on both the upper and lower surfaces. Therefore, after a silicon steel sheet is separated, image recognition is required for both the upper and lower surfaces of the silicon steel sheet, so as to obtain the quality influence area of ​​each silicon steel sheet, and then obtain the corresponding quality influence parameter (when two sides are involved, the total is taken), and then determine whether it is necessary to adjust the real-time separation force of the next silicon steel sheet of the current transformer based on the quality influence parameter.

[0070] In addition, the separation device can separate each silicon steel sheet of the transformer by means of adsorption, clamping, etc. For example, when the adsorption method is used, the separation device can use at least one mechanical arm with a suction cup to adsorb on the surface of the silicon steel sheet; or, when the clamping method is used, the separation device can use the opposite side clamping method to clamp the side of the silicon steel sheet (since the thickness of the silicon steel sheet is relatively small, the size of the clamping arm can also be set to be relatively small, but it should have a large contact area with the side of the silicon steel sheet so that there are enough force points to separate the silicon steel sheet). Furthermore, in order to ensure that when the separation device separates the upper silicon steel sheet, the weight of the silicon steel sheet on the lower layer will not affect the separation of the upper silicon steel sheet, the separation device can use a fixing device to fix the silicon steel sheet that does not need to be separated temporarily.

[0071] In an embodiment of the present invention, a mathematical model of the separation force and deformation, coating integrity, and insulation material retention of silicon steel sheets is pre-established, and the separation process image is collected in real time in combination with machine vision technology for regional identification, thereby obtaining an evaluation index of a quality influencing parameter, and then calculating a reference separation force and comparing it with the real-time separation force, and dynamically adjusting the real-time separation force. During the process of separating silicon steel sheets, the real-time separation force of the separation device is adjusted in time, so that the separation force applied by the separation device in real time meets the standard separation force, thereby reducing the adverse effects on the silicon steel sheets during the separation process.

[0072] Specifically, in step S1, a real-time image and a real-time separation force of silicon steel sheets of the transformer during separation are obtained.

[0073] For example, an industrial camera is used to obtain real-time images of transformer silicon steel sheets during the separation process, such as an image acquisition frequency of ten frames per second (this value is only an example and can be changed according to actual conditions), a resolution of 1920×1080 pixels, a ring-shaped LED light source is configured to provide uniform lighting, and the image is saved in PNG lossless compression format. At the same time, the real-time separation force applied by the separation device corresponding to the real-time image is recorded.

[0074] Specifically, in step S2, image recognition is performed on the real-time image to obtain the quality impact area of ​​the silicon steel sheet.

[0075] Exemplarily, before performing image recognition on the real-time image, since there are many images acquired, in order to speed up the recognition efficiency, invalid images, such as images with low clarity and wrong shooting angles, can be eliminated, and then the remaining real-time images can be preprocessed, such as removing noise through median filters and Gaussian filters, using contrast-limited adaptive histogram equalization methods to improve image contrast, correcting image angles and sizes, and other image processing methods. After the real-time image is preprocessed, the real-time image can be recognized by a neural network. The embodiment of the present invention provides an example of a neural network model (the existing technology of using a neural network model for image recognition can be applicable), such as using an improved U-Net convolutional neural network to recognize the preprocessed real-time image, and dividing the image into three categories: silicon steel sheet deformation area, surface coating shedding area, and insulating material residual area. For example, the network structure can adopt a 5-layer encoder-decoder architecture, use a 3×3 convolution kernel, use batch normalization and ReLU activation function in each layer, and use Softmax in the last layer to output the category probability distribution of each pixel, thereby obtaining the deformation area of ​​the silicon steel sheet, the surface coating shedding area, and the insulating material residual area between the laminates.

[0076] Specifically, in step S3, the quality impact parameter is generated using the quality impact region.

[0077] For example, the number of pixels in each area is calculated according to the segmentation result, and the area ratio is obtained by dividing it by the total number of pixels in the image. The area ratio is used to represent the three quality influencing parameters: the degree of deformation, the degree of coating shedding, and the degree of residual insulation material. For example, if the calculated percentage of the deformed area of ​​the silicon steel sheet is 8.5%, the percentage of the coating shedding area is 3.2%, and the percentage of the residual insulation material area is 94.6%, with 100 as the unit, the corresponding deformation degree value is 8.5, the coating shedding degree value is 3.2, and the residual insulation material value is 94.6.

[0078] Specifically, in step S4, the quality influencing parameter is input into a separation force prediction model corresponding to the type of the transformer to obtain a corresponding reference separation force.

[0079] Exemplarily, since there are at least one or more differences in the number of stacked silicon steel sheets, size, coating thickness, and insulation material thickness of transformers of different models, when the separation device applies the same separation force to transformers of different models, the adverse effects on the silicon steel sheets of transformers of different models may also be different. Therefore, in the embodiment of the present invention, corresponding separation force prediction models are provided for transformers of different models. Specifically, the separation force prediction model includes:

[0080] 1) A first mathematical model of the deformation degree value and the first separation force;

[0081] 2) a second mathematical model of the coating shedding degree value and the second separation force;

[0082] 3) A third mathematical model of the insulating material residual degree value and the third separation force.

[0083] Exemplarily, for the three mathematical models, after the deformation degree value, the coating shedding degree value and the insulating material residual degree value are respectively input into the corresponding mathematical models, three corresponding reference separation forces can be obtained.

[0084] Specifically, in step S5, when the real-time separation force does not match the reference separation force, the real-time separation force is adjusted.

[0085] For example, when the real-time separation force matches any of the three reference separation forces, it indicates that the real-time separation force applied by the current separation device is appropriate and does not need to be adjusted. When the real-time separation force does not match any of the three reference separation forces, it indicates that the real-time separation force applied by the current separation device is inappropriate. If this separation force is continued to be used, it will cause greater adverse effects on the silicon steel sheet. Therefore, the real-time separation force needs to be adjusted. At this time, a target separation force is first determined based on the reference separation force, and then the real-time separation force is adjusted based on the target separation force. There are several ways to determine the specific target separation force (one of the methods can be selected to determine the target separation force in specific applications):

[0086] 1) Taking any value of the reference separation force as the target separation force;

[0087] 2) taking the average value of the reference separation force as the target separation force;

[0088] 3) The middle value of the reference separation force is taken as the target separation force.

[0089] Specifically, in step S4, the method for constructing the separation force prediction model includes: applying a number of preset separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence; obtaining a surface image of the silicon steel sheet after each application of the preset separation force; analyzing the surface image to obtain quality influencing parameters of the silicon steel sheet corresponding to different preset separation forces; for each preset separation force, taking the minimum value of the quality influencing parameters as the standard quality influencing parameter; constructing a separation force prediction model between the standard quality influencing parameter and the preset separation force; wherein, different models of transformers correspond to a separation force prediction model respectively.

[0090] It should be noted that, since the quality influencing parameters in the present invention include three values: the deformation degree value, the coating shedding degree value and the insulating material residual degree value of the silicon steel sheet, there are three model construction methods for the first mathematical model, the second mathematical model and the third mathematical model respectively.

[0091] See also Figure 2 , Figure 2 is a flow chart of a method for constructing a first mathematical model provided by an embodiment of the present invention, wherein the method for constructing the first mathematical model includes:

[0092] S11, applying a plurality of first separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence;

[0093] S12, acquiring a surface image of the silicon steel sheet after applying the first separation force each time;

[0094] S13, analyzing the surface image to obtain deformation degree values ​​of the silicon steel sheet corresponding to different first separation forces;

[0095] S14, for each first separation force, taking the minimum value of the deformation degree values ​​as the standard deformation degree value;

[0096] S15. Constructing a first mathematical model between the standard deformation degree value and the first separation force; wherein different types of transformers correspond to a first mathematical model respectively.

[0097] For example, assuming that the total number of transformer models is 50, and each model provides at least 20 transformers, the transformers of the same model are grouped together for actual separation force measurement, and for each transformer, several first separation forces are applied in sequence, such as applying 10 first separation forces with the same interval and increasing in sequence, and then the deformation area is obtained by using image recognition technology, and then the corresponding deformation degree value is generated. Since the transformer needs to apply 10 first separation forces at this time, if the number of silicon steel sheets of the same transformer is insufficient, a new transformer of the same model can continue to be used. After all transformers are measured, the first separation force is used as the classification basis, and the deformation degree values ​​corresponding to the same model and the same first separation force are divided into a group, that is, at this time, there are 20 deformation degree values ​​corresponding to one first separation force and the same model transformer, and the minimum deformation degree value is selected from the 20 deformation degree values ​​as the standard deformation degree value, which means that the deformation of the silicon steel sheet is minimal when the first separation force is applied. After obtaining the standard deformation degree values ​​corresponding to all first separation forces (a total of 10), a first mathematical model of the first separation force and the deformation degree value is constructed.

[0098] Specifically, the first mathematical model satisfies the following formula:

[0099] (1);

[0100] in, is the deformation degree value; is the deformation degree influencing factor, which is a preset fixed value; The first separation force must satisfy: , is the minimum separation force that the separation device can exert, The maximum separation force that the separation device can exert.

[0101] Exemplarily, the separation force is directly proportional to the degree of deformation of the silicon steel sheet. As the first separation force increases, the value of the degree of deformation increases.

[0102] Furthermore, since the deformation degree of the silicon steel sheet can be simulated by a model, the implementation of the present invention also provides another method of constructing a first mathematical model, including: obtaining structural parameters and material property parameters of several different types of transformers; constructing different types of simulated transformers based on the structural parameters and the material property parameters; applying several first separation forces to the simulated transformer in sequence to obtain deformation degree values ​​of the silicon steel sheet of the simulated transformer corresponding to different first separation forces; obtaining the deformation degree value of the silicon steel sheet corresponding to each first separation force, and constructing a first mathematical model between the deformation degree value and the first separation force; wherein, different types of transformers correspond to a first mathematical model respectively.

[0103] Exemplarily, in the implementation of the present invention, the structural parameters and material property parameters of several transformers of different models are obtained to construct a simulated transformer, and the structural parameters and material property parameters can be obtained by querying the transformer model database, and the structural parameters and material property parameters corresponding to different transformer models are pre-stored in the database. The structural parameters include parameters such as the number, thickness, size and gap of the silicon steel sheets of the transformer, and the material property parameters include parameters such as yield strength, magnetic flux density, surface coating adhesion and insulating material adhesion. Several first separation forces are applied to the transformer model, such as applying a radial separation force of 100-1000N, each 100N is a loading step, and each loading step needs to separate a corresponding silicon steel sheet (due to the limited number of silicon steel sheets of a transformer, the separation can be repeated, that is, after separating the silicon steel sheets once, the model can restore the original shape and continue to use different separation forces for separation), and the deformation degree of the silicon steel sheet is analyzed by the stress distribution diagram (such as generated by von Mises), and the deformation degree value of the silicon steel sheet of the simulated transformer corresponding to different first separation forces is obtained. It should be noted that, since the relationship between the separation force and the deformation degree of the silicon steel sheet is obtained by applying separation force to the simulated transformer, the separation process of the simulated transformer does not need to consider the influence of external factors, such as the working loss of the separation device (the actual output is not equal to the ideal output), environmental (temperature, humidity, etc.) influence, image recognition error influence, etc. Therefore, the relationship between the separation force and the deformation degree of the silicon steel sheet obtained by simulation is the most ideal situation, that is, the corresponding deformation degree of the silicon steel sheet can be minimized under the appropriate separation force.

[0104] See also Figure 3 , Figure 3 is a flow chart of a method for constructing a second mathematical model provided by an embodiment of the present invention, wherein the method for constructing the second mathematical model includes:

[0105] S21, applying a plurality of second separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence;

[0106] S22, acquiring a surface image of the silicon steel sheet after applying the second separation force each time;

[0107] S23, analyzing the surface image to obtain coating shedding degree values ​​of the silicon steel sheet corresponding to different second separation forces;

[0108] S24, for each second separation force, taking the minimum value of the coating shedding degree value as the standard coating shedding degree value;

[0109] S25, constructing a second mathematical model between the standard coating shedding degree value and the second separation force; wherein different types of transformers correspond to a second mathematical model respectively.

[0110] Exemplarily, since the coating integrity of the silicon steel sheet during separation cannot be simulated by a simulated transformer, the embodiment of the present invention adopts an actual measurement method to obtain the relationship between the coating integrity and the separation force. For example, in the embodiment of the present invention, it is assumed that the total number of transformer models is 50, and at least 20 transformers are provided for each model. The transformers of the same model are grouped together to measure the separation force. For each transformer, several second separation forces are applied in sequence, such as applying 10 second separation forces that increase in sequence at the same interval, and then the surface coating shedding area is obtained by using image recognition technology, and then the corresponding coating shedding degree value is generated. After all transformers are measured, the coating shedding degree values ​​corresponding to the same model and the same second separation force are divided into a group based on the second separation force, that is, at this time, there are 20 coating shedding degree values ​​corresponding to one second separation force and the same model transformer. The minimum coating shedding degree value is selected from the 20 coating shedding degree values ​​as the standard coating shedding degree value. This standard coating shedding degree value indicates that the coating shedding effect on the silicon steel sheet is minimized when the second separation force is applied. After obtaining all the standard coating shedding degree values ​​corresponding to the second separation forces (10 in total), a second mathematical model of the second separation force and the coating shedding degree value is constructed.

[0111] Specifically, the second mathematical model satisfies the following formula:

[0112] (2);

[0113] in, is the standard coating shedding degree value; For the second separation force, it is necessary to satisfy: ; , , are coating influence factors, which are preset fixed values.

[0114] Exemplarily, the separation force is directly proportional to the coating shedding of the silicon steel sheet. As the separation force gradually increases, more coating shedding occurs. Conversely, as the separation force gradually decreases, less coating shedding occurs.

[0115] See also Figure 4 , Figure 4 is a flow chart of a method for constructing a third mathematical model provided in an embodiment of the present invention, wherein the method for constructing the third mathematical model comprises:

[0116] S31, applying a plurality of third separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence;

[0117] S32, acquiring a surface image of the silicon steel sheet after applying the third separation force each time;

[0118] S33, analyzing the surface image to obtain insulating material residual degree values ​​of the silicon steel sheet corresponding to different third separation forces;

[0119] S34, for each third separation force, taking the minimum value of the insulating material residual degree values ​​as the standard insulating material residual degree value;

[0120] S35. Construct a third mathematical model between the standard insulating material residual degree value and the third separation force; wherein different types of transformers correspond to a third mathematical model respectively.

[0121] Since the residual degree of insulating material of silicon steel sheets during separation cannot be simulated by a simulated transformer, the embodiment of the present invention adopts an actual measurement method to obtain the relationship between the residual degree of insulating material and the separation force. For example, in the embodiment of the present invention, it is assumed that the total number of transformer models is 50, and at least 20 transformers are provided for each model. The transformers of the same model are grouped together for actual measurement of separation force. For each transformer, a number of third separation forces are applied in sequence, such as applying 10 third separation forces that increase in sequence at the same interval, and then the insulating material residual area is obtained by using image recognition technology, and then the corresponding insulating material residual degree value is generated. After all transformers are measured, the insulating material residual degree values ​​corresponding to the same model and the same third separation force are divided into a group based on the third separation force, that is, at this time, there are 20 insulating material residual degree values ​​corresponding to one third separation force and the same model transformer. The minimum insulating material residual degree value is selected from the 20 insulating material residual degree values ​​as the standard insulating material residual degree value. This standard insulating material residual degree value indicates that the insulating material residual effect on the silicon steel sheet is the smallest when the third separation force is applied. After obtaining all the standard insulating material residual degree values ​​corresponding to the third separation forces (10 in total), a third mathematical model of the third separation force and the insulating material residual degree value is constructed.

[0122] Specifically, the third mathematical model satisfies the following formula:

[0123] (3);

[0124] in, is the residual degree value of the standard insulating material; The third separation force must satisfy: ; is the material residue influence factor, which is a preset fixed value.

[0125] Exemplarily, the separation force is inversely proportional to the degree of residual insulating material of the silicon steel sheet. As the separation force gradually increases, less insulating material remains. Conversely, as the separation force gradually decreases, more insulating material remains.

[0126] In an embodiment of the present invention, by constructing a relationship model between the separation force and the minimum mass influence parameter when separating silicon steel sheets in transformers of different types, it is possible to generate a suitable reference separation force based on the mass influence parameters obtained by actual measurement when the separation device actually separates the silicon steel sheets, thereby adjusting the real-time separation force of the separation device and minimizing the adverse effects of the real-time separation force applied by the separation device on the silicon steel sheets.

[0127] Specifically, when the real-time separation force does not match the reference separation force, the method further includes:

[0128] Recording the continuous cumulative number of times that the real-time separation force does not match the reference separation force;

[0129] When the continuous accumulated number of times is greater than a preset number threshold, adjusting the position of the separation device;

[0130] When the continuous accumulated number of times is less than or equal to the number threshold, the current position of the separation device is kept unchanged.

[0131] Exemplarily, during the separation of silicon steel sheets, the initial position of the separation device is fixed, which refers to the placement position of the mechanical arm in the separation device. When the adsorption method is used, the default is that the suction cup of the mechanical arm is adsorbed on the initial position of the silicon steel sheet. Assuming the number threshold is 5, if the real-time separation force is adjusted 6 times in a row, it means that the real-time separation force that continues to be adjusted subsequently cannot match the reference separation force. At this time, it may be caused by the incorrect placement of the separation device, that is, the initial position of the suction cup is set incorrectly, so it is necessary to adjust the initial position of the suction cup, such as moving a fixed value in any direction of the front, back, left, or right. The specific value and direction of movement can be pre-set according to the empirical value, which is obtained by summarizing historical experience. The present invention does not make specific restrictions on this. After adjusting the position of the separation device, when separating the next silicon steel sheet, if the real-time separation force matches the reference separation force, then continue to use this position to separate the next silicon steel sheet, and the continuous cumulative number is reset to zero; after adjusting the position of the separation device, when separating the next silicon steel sheet, if the real-time separation force still matches the reference separation force, then adjust the position of the separation device again, and the continuous cumulative number is +1.

[0132] In the embodiment of the present invention, when the separation device cannot match the reference separation force even after adjusting the real-time separation force, the position of the separation device is adjusted in time to minimize the adverse effects of the separation device on the silicon steel sheet.

[0133] See also Figure 5 , Figure 5 is another flow chart of a transformer silicon steel sheet separation method provided by an embodiment of the present invention. In step S5, after adjusting the real-time separation force, the method further includes:

[0134] S6. When it is detected that all silicon steel sheets of the current transformer have been separated, a multivariate monitoring data set of the transformer during the silicon steel sheet separation process is obtained;

[0135] S7. When the multivariate monitoring data set meets a preset data standard, storing the multivariate monitoring data set;

[0136] S8. Regularly analyze all stored multivariate monitoring data sets to obtain at least one standard monitoring data set corresponding to each transformer model; wherein the standard monitoring data set is used for initial program setting of the separation device when the silicon steel sheets of the transformer are separated next time.

[0137] Exemplarily, the multivariate monitoring data set includes the operating parameters of the separation device and the quality influencing parameters of each silicon steel sheet after separation; wherein the operating parameters include operating power, real-time separation force, position, separation speed and separation time. In the embodiment of the present invention, after each transformer is separated, the multivariate monitoring data set of the transformer is collected to expand the database. After obtaining a multivariate monitoring data set, a data standard evaluation is required to evaluate whether the multivariate monitoring data set is complete and valid. If there is a lack of real-time separation force, or there are multiple records of real-time separation force that are obviously wrong (such as exceeding the separation force threshold of the separation device), that is, the record of the real-time separation force is determined to be invalid, then the multivariate monitoring data set can be deleted. If all data are complete and valid, the multivariate monitoring data set is stored. The embodiment of the present invention also regularly analyzes all stored multivariate monitoring data sets, and selects a relatively excellent standard monitoring data set for each type of transformer. This standard monitoring data set can be used as the initial program setting for the next start of the separation device, such as setting the initial program of the separation device according to the operating power, real-time separation force, position, separation speed and separation time in the standard monitoring data set.

[0138] Specifically, the embodiment of the present invention provides two methods for determining a standard monitoring data set, the first method is achieved by constructing a target optimization model, and the second method is achieved by training a quality prediction model.

[0139] In a first embodiment, the periodic analysis of all stored multivariate monitoring data sets includes: for each type of transformer, sequentially inputting the stored multivariate monitoring data sets belonging to the same type into a preset target optimization model; and taking the multivariate monitoring data set corresponding to the minimum value output by the target optimization model as the standard monitoring data set.

[0140] Exemplarily, the target optimization model aims to minimize the transformer deformation degree, coating shedding degree and insulation material residual degree, as well as the energy consumption evaluation value of the separation device. All transformers of each model are input into formula (4) for calculation, and then the standard monitoring data set corresponding to the smallest value (or two values) is selected. The operating parameters of the separation device in the standard monitoring data set (if there are more than two standard monitoring data sets, one of them is selected) are used to set the initial procedure for separating the transformer of this model next time.

[0141] Specifically, the target optimization model satisfies the following formula:

[0142] (4);

[0143] in, For the The optimization objective of the target optimization model corresponding to each type of transformer is: , is the total number of models; For the Model No. The average deformation value corresponding to each transformer is: , For the Total number of transformers of different models; For the Model No. The average coating shedding degree value corresponding to each transformer; For the Model No. The average insulation material residual value corresponding to each transformer; For the separation device in the separation Model No. The energy consumption assessment value corresponding to the transformer.

[0144] It is worth noting that since the transformer has multiple silicon steel sheets and each silicon steel sheet corresponds to a quality influencing parameter, the average value in formula (4) refers to the average value of the quality influencing parameters corresponding to all silicon steel sheets.

[0145] Specifically, the calculation process of the energy consumption evaluation value satisfies the following formula:

[0146] (5);

[0147] in, , , are the first weight, the second weight and the third weight respectively; For the separation device in the separation The average separation force of silicon steel sheets is , For the Total number of silicon steel sheets per transformer; For the separation device in the separation The average separation speed of silicon steel sheets; For the separation device in the separation The distance of position movement when one silicon steel sheet is formed; For the separation device in the separation Operating power per silicon steel sheet; is the first separation time, is the second separation time, and The difference is the separation device in the separation The separation time of each transformer is 2000.

[0148] It should be noted that the first weight, the second weight and the third weight can be preset according to experience values, such as satisfying < < This is because when the separation device is running, the operating power is the most representative of the energy consumption of the separation device, so its corresponding third weight is Should be set to the maximum, such as =0.7. The real-time separation force and separation speed of the separation device are secondary considerations, and the operating loss of the separation device caused by them is also relatively large, so the first weight Less than the third weight ,like =0.25. Since the position adjustment of the separation device rarely occurs, the operating loss of the separation device is also small, so the corresponding second weight The value of is set to the minimum, such as =0.05.

[0149] In a second embodiment, the periodic analysis of all stored multivariate monitoring data sets includes: inputting the multivariate monitoring data sets into a pre-trained quality prediction model so that the quality prediction model outputs a corresponding quality grade; when the quality grade meets a preset standard grade, determining the multivariate monitoring data set corresponding to the quality grade as a standard monitoring data set.

[0150] Exemplarily, the training method of the quality prediction model includes: using a multivariate monitoring data set acquired historically as a sample data set; labeling the sample data set with a quality grade, and dividing the sample data set into a training set and a test set; using the training set to train a support vector machine model, and using the test set to test the support vector machine model, and outputting a trained quality prediction model. After obtaining the trained quality prediction model, by inputting the multivariate monitoring data set, the quality grade of this multivariate monitoring data set can be obtained, such as the quality grade is divided into excellent, good, medium, and poor, and the standard grade is "excellent", so all multivariate monitoring data sets with a quality grade of "excellent" corresponding to different types of transformers can be screened out as standard monitoring data sets.

[0151] It is worth noting that when all stored multivariate monitoring data sets are analyzed regularly, there is no need to analyze the multivariate monitoring data sets that have been analyzed in the previous cycle again. Instead, the screening process of the standard monitoring data set in any of the above two ways is performed for the multivariate monitoring data sets newly added to the database.

[0152] In an embodiment of the present invention, multivariate monitoring data of transformers of different types during the silicon steel sheet separation process is collected, and the better multivariate monitoring data is screened out for the initial program setting of the next time the same type of transformer is used to separate the silicon steel sheets, so that the separation device can meet expectations as much as possible during the initial operation, reduce the adverse effects on the silicon steel sheets in the initial separation stage, and then reduce the number of subsequent adjustments to the separation force, thereby improving the separation efficiency of the silicon steel sheets.

[0153] Specifically, before the periodic analysis of all stored multivariate monitoring data sets, the method also includes: determining the resource recovery rate of the current transformer after completing the separation of silicon steel sheets and the loss energy consumption of the separation device; when the resource recovery rate is greater than or equal to the preset resource recovery threshold, and the loss energy consumption is less than or equal to the preset energy consumption threshold, retaining the multivariate monitoring data set corresponding to the current transformer; when the resource recovery rate is less than the resource recovery threshold, or the loss energy consumption is greater than the energy consumption threshold, deleting the multivariate monitoring data set corresponding to the current transformer.

[0154] For example, as time goes by, there are more multi-source monitoring data sets stored in the database. Before regularly analyzing all the stored multivariate monitoring data sets, some multivariate monitoring data sets that cause large losses to resource recovery or separation devices can be eliminated, thereby improving data processing efficiency. At this time, the resource recovery rate of the current transformer after completing the separation of silicon steel sheets and the loss energy consumption of the separation device are determined. The resource recovery rate can be obtained by calculating the metal recovery (gram weight) ratio of the transformer before and after separation. For example, if the resource recovery rate is 92.5% (metal recovery 74kg / total 80kg), the loss energy consumption can be obtained according to the above formula (5) It is worth noting that the values ​​of the resource recovery threshold and the energy consumption threshold can be preset according to empirical values, and the present invention does not make any specific limitation on this.

[0155] Compared with the prior art, the transformer silicon steel sheet separation method disclosed in the present invention pre-establishes a mathematical model of silicon steel sheet separation force and deformation, coating integrity, and insulation material retention, and combines machine vision technology to collect separation process images in real time and perform regional identification, thereby obtaining the quality influencing parameter as an evaluation index, and then calculates the reference separation force and compares it with the real-time separation force, and dynamically adjusts the real-time separation force. During the process of separating silicon steel sheets, the real-time separation force of the separation device is adjusted in time, so that the separation force applied by the separation device in real time meets the standard separation force, thereby reducing the adverse effects on the silicon steel sheets during the separation process. In addition, by collecting multivariate monitoring data of transformers of different models during the silicon steel sheet separation process, the better multivariate monitoring data is selected for the initial program setting of the next transformer of the same model when separating silicon steel sheets, so that the separation device is as consistent as possible with expectations during the initial operation, reducing the adverse effects on the silicon steel sheets in the initial separation stage, thereby reducing the number of subsequent adjustments to the separation force, and improving the separation efficiency of silicon steel sheets.

[0156] See also Figure 6 , Figure 6 1 is a structural block diagram of a transformer silicon steel sheet separation device 100 provided in an embodiment of the present invention, wherein the transformer silicon steel sheet separation device 100 comprises:

[0157] A data acquisition module 11 is used to acquire real-time images and real-time separation forces of silicon steel sheets of the transformer during separation;

[0158] The quality impact area acquisition module 12 is used to perform image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet;

[0159] A quality impact parameter generating module 13, configured to generate quality impact parameters using the quality impact region;

[0160] A reference separation force determination module 14, configured to input the quality influencing parameter into a separation force prediction model corresponding to the transformer model to obtain a corresponding reference separation force;

[0161] The separation force adjustment module 15 is used to adjust the real-time separation force when the real-time separation force does not match the reference separation force.

[0162] Specifically, the method for constructing the separation force prediction model includes:

[0163] Applying a plurality of preset separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence;

[0164] After applying the preset separation force each time, obtaining a surface image of the silicon steel sheet;

[0165] Analyzing the surface image to obtain quality influencing parameters of the silicon steel sheet corresponding to different preset separation forces;

[0166] For each preset separation force, the minimum value among the mass influencing parameters is taken as the standard mass influencing parameter;

[0167] A separation force prediction model between the standard quality influencing parameter and the preset separation force is constructed; wherein different types of transformers correspond to a separation force prediction model respectively.

[0168] Specifically, the quality influencing parameters include the deformation degree value, coating shedding degree value and insulating material residual degree value of the silicon steel sheet; then, the separation force prediction model includes:

[0169] A first mathematical model of the deformation degree value and the first separation force;

[0170] A second mathematical model of the coating shedding degree value and the second separation force;

[0171] A third mathematical model of the insulating material residual degree value and the third separation force.

[0172] Specifically, the first mathematical model satisfies the following formula:

[0173] (1);

[0174] in, is the deformation degree value; is the deformation degree influencing factor, which is a preset fixed value; The first separation force must satisfy: is the minimum separation force that the separation device can exert, It is the maximum separation force that the separation device can exert.

[0175] Specifically, the second mathematical model satisfies the following formula:

[0176] (2);

[0177] in, is the standard coating shedding degree value; is the second separation force; , , are coating influence factors, which are preset fixed values.

[0178] Specifically, the third mathematical model satisfies the following formula:

[0179] (3);

[0180] in, is the residual degree value of the standard insulating material; is the third separation force; is the material residue influence factor, which is a preset fixed value.

[0181] Specifically, the method for constructing the first mathematical model includes:

[0182] Obtain structural parameters and material property parameters of several transformers of different models;

[0183] Constructing different types of simulated transformers based on the structural parameters and the material property parameters;

[0184] applying a plurality of first separation forces to the simulated transformer in sequence to obtain deformation degree values ​​of the silicon steel sheet of the simulated transformer corresponding to different first separation forces;

[0185] The deformation degree value of the silicon steel sheet corresponding to each first separation force is obtained, and a first mathematical model between the deformation degree value and the first separation force is constructed; wherein different types of transformers correspond to a first mathematical model respectively.

[0186] Specifically, the separation force adjustment module 15 is specifically used to: determine a target separation force according to the reference separation force; and adjust the real-time separation force based on the target separation force; wherein the target separation force is determined in one of the following ways:

[0187] Taking any value of the reference separation force as the target separation force;

[0188] Taking the average value of the reference separation force as the target separation force;

[0189] The middle value of the reference separation force is taken as the target separation force.

[0190] Specifically, the transformer silicon steel sheet separation device 100 further includes:

[0191] The position adjustment module is used to record the continuous cumulative number of times that the real-time separation force does not match the reference separation force; when the continuous cumulative number is greater than a preset number threshold, adjust the position of the separation device; when the continuous cumulative number is less than or equal to the number threshold, keep the current position of the separation device unchanged.

[0192] Specifically, the transformer silicon steel sheet separation device 100 further includes:

[0193] A multivariate monitoring data set acquisition module is used to acquire a multivariate monitoring data set of the transformer during the silicon steel sheet separation process when it is detected that all silicon steel sheets of the current transformer have been separated;

[0194] A multivariate monitoring data set screening module, used for storing the multivariate monitoring data set when the multivariate monitoring data set meets a preset data standard;

[0195] The standard monitoring data set acquisition module is used to regularly analyze all stored multivariate monitoring data sets to obtain at least one standard monitoring data set corresponding to each transformer model; wherein the standard monitoring data set is used for the initial program setting of the separation device when the silicon steel sheets of the transformer are separated next time.

[0196] Specifically, the multivariate monitoring data set includes operating parameters of the separation device and quality influencing parameters of each silicon steel sheet after separation; wherein the operating parameters include operating power, real-time separation force, position, separation speed and separation time.

[0197] Specifically, the standard monitoring data set acquisition module is used to: for each type of transformer, sequentially input the stored multivariate monitoring data sets belonging to the same type into a preset target optimization model; and take the multivariate monitoring data set corresponding to the minimum value output by the target optimization model as the standard monitoring data set.

[0198] Specifically, the target optimization model satisfies the following formula:

[0199] (4);

[0200] in, For the The optimization objective of the target optimization model corresponding to each type of transformer is: , is the total number of models; For the Model No. The average deformation value corresponding to each transformer is: , For the Total number of transformers of different models; For the Model No. The average coating shedding degree value corresponding to each transformer; For the Model No. The average insulation material residual value corresponding to each transformer; For the separation device in the separation Model No. The energy consumption assessment value corresponding to the transformer.

[0201] Specifically, the calculation process of the energy consumption evaluation value satisfies the following formula:

[0202] (5);

[0203] in, , , are the first weight, the second weight and the third weight respectively; For the separation device in the separation The average separation force of silicon steel sheets is , For the Total number of silicon steel sheets per transformer; For the separation device in the separation The average separation speed of silicon steel sheets; For the separation device in the separation The distance of position movement when one silicon steel sheet is formed; For the separation device in the separation Operating power per silicon steel sheet; is the first separation time, is the second separation time, and The difference is the separation device in the separation The separation time of each transformer is 2000.

[0204] Specifically, the standard monitoring data set acquisition module is used to: input the multivariate monitoring data set into a pre-trained quality prediction model so that the quality prediction model outputs a corresponding quality grade; when the quality grade meets the preset standard grade, determine the multivariate monitoring data set corresponding to the quality grade as the standard monitoring data set.

[0205] Specifically, the training method of the quality prediction model includes: using a multivariate monitoring data set acquired historically as a sample data set; labeling the sample data set with a quality grade, and dividing the sample data set into a training set and a test set; using the training set to train a support vector machine model, and using the test set to test the support vector machine model, and outputting the trained quality prediction model.

[0206] Specifically, the transformer silicon steel sheet separation device 100 further includes:

[0207] A resource recovery rate and energy loss determination module, used to determine the resource recovery rate of the current transformer after the silicon steel sheet separation is completed and the energy loss of the separation device;

[0208] The multivariate monitoring data set screening module is also used to retain the multivariate monitoring data set corresponding to the current transformer when the resource recovery rate is greater than or equal to a preset resource recovery threshold, and the lost energy consumption is less than or equal to a preset energy consumption threshold; when the resource recovery rate is less than the resource recovery threshold, or the lost energy consumption is greater than the energy consumption threshold, delete the multivariate monitoring data set corresponding to the current transformer.

[0209] It is worth noting that the working process of each module in the transformer silicon steel sheet separation device 100 described in the embodiment of the present invention can refer to the working process of the transformer silicon steel sheet separation method described in the above embodiment, and will not be repeated here.

[0210] See also Figure 7 , Figure 7 1 is a structural block diagram of a transformer silicon steel sheet separation device 200 provided in an embodiment of the present invention, wherein the transformer silicon steel sheet separation device 200 comprises a processor 21, a memory 22, and a computer program stored in the memory 22 and executable on the processor 21. When the processor 21 executes the computer program, the steps in each of the above transformer silicon steel sheet separation method embodiments are implemented.

[0211] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory 22 and executed by the processor 21 to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the transformer silicon steel sheet separation device 200.

[0212] The transformer silicon steel sheet separation device 200 may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art may understand that the schematic diagram is only an example of the transformer silicon steel sheet separation device 200 and does not constitute a limitation on the transformer silicon steel sheet separation device 200. The transformer silicon steel sheet separation device 200 may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, the transformer silicon steel sheet separation device 200 may also include input and output devices, network access devices, buses, etc.

[0213] The processor 21 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor 21 is the control center of the transformer silicon steel sheet separation device 200, and uses various interfaces and lines to connect various parts of the entire transformer silicon steel sheet separation device 200.

[0214] The memory 22 can be used to store the computer program and / or module. The processor 21 realizes various functions of the transformer silicon steel sheet separation device 200 by running or executing the computer program and / or module stored in the memory 22 and calling the data stored in the memory 22. The memory 22 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory 22 can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0215] Wherein, if the module / unit integrated in the transformer silicon steel sheet separation device 200 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor 21. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0216] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for separating transformer silicon steel sheets, characterized in that: include: Obtain real-time images and real-time separation force of the transformer silicon steel sheets during separation; Performing image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet; generating a quality impact parameter using the quality impact region; Inputting the mass influencing parameter into a separation force prediction model corresponding to the model of the transformer to obtain a corresponding reference separation force; When the real-time separation force does not match the reference separation force, adjusting the real-time separation force; Wherein, the method for constructing the separation force prediction model includes: Applying a plurality of preset separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence; After applying the preset separation force each time, obtaining a surface image of the silicon steel sheet; Analyzing the surface image to obtain quality influencing parameters of the silicon steel sheet corresponding to different preset separation forces; For each preset separation force, the minimum value among the mass influencing parameters is taken as the standard mass influencing parameter; Constructing a separation force prediction model between the standard quality influencing parameter and the preset separation force; wherein different types of transformers correspond to a separation force prediction model respectively; The quality influencing parameters include the deformation degree value, coating shedding degree value and insulating material residual degree value of the silicon steel sheet; then, the separation force prediction model includes: A first mathematical model of the deformation degree value and the first separation force; A second mathematical model of the coating shedding degree value and the second separation force; A third mathematical model of the insulating material residual degree value and the third separation force.

2. The transformer silicon steel sheet separation method according to claim 1, characterized in that: The number of the reference separation forces is the same as the number of the separation force prediction models; then, adjusting the real-time separation force comprises: determining a target separation force according to the reference separation force; The real-time separation force is adjusted based on the target separation force; wherein the target separation force is determined in one of the following ways: Taking any value of the reference separation force as the target separation force; Taking the average value of the reference separation force as the target separation force; The middle value of the reference separation force is taken as the target separation force.

3. The transformer silicon steel sheet separation method according to claim 1, characterized in that: When the real-time separation force does not match the reference separation force, the method further comprises: Recording the continuous cumulative number of times that the real-time separation force does not match the reference separation force; When the continuous accumulated number of times is greater than a preset number threshold, adjusting the position of the separation device; When the continuous accumulated number of times is less than or equal to the number threshold, the current position of the separation device is kept unchanged.

4. The transformer silicon steel sheet separation method according to claim 1, characterized in that: After adjusting the real-time separation force, the method further includes: When it is detected that all silicon steel sheets of the current transformer have been separated, a multivariate monitoring data set of the transformer during the silicon steel sheet separation process is obtained; When the multivariate monitoring data set meets a preset data standard, storing the multivariate monitoring data set; All stored multivariate monitoring data sets are analyzed regularly to obtain at least one standard monitoring data set corresponding to each transformer model; wherein the standard monitoring data set is used for initial program setting of the separation device when the silicon steel sheets of the transformer are separated next time.

5. The transformer silicon steel sheet separation method according to claim 4, characterized in that: The multivariate monitoring data set includes operating parameters of the separation device and quality influencing parameters of each silicon steel sheet after separation; wherein the operating parameters include operating power, real-time separation force, position, separation speed and separation time.

6. The transformer silicon steel sheet separation method according to claim 4, characterized in that: The periodic analysis of all stored multivariate monitoring data sets includes: For each type of transformer, the stored multivariate monitoring data set belonging to the same type is sequentially input into the preset target optimization model; The multivariate monitoring data set corresponding to the minimum value output by the target optimization model is used as the standard monitoring data set.

7. The transformer silicon steel sheet separation method according to claim 4, characterized in that: The periodic analysis of all stored multivariate monitoring data sets includes: Inputting the multivariate monitoring data set into a pre-trained quality prediction model so that the quality prediction model outputs a corresponding quality grade; When the quality level meets the preset standard level, the multivariate monitoring data set corresponding to the quality level is determined to be the standard monitoring data set.

8. The transformer silicon steel sheet separation method according to claim 4, characterized in that: Before periodically analyzing all stored multivariate monitoring data sets, the method further includes: Determine the resource recovery rate of the current transformer after completing the separation of silicon steel sheets and the energy loss of the separation device; When the resource recovery rate is greater than or equal to a preset resource recovery threshold, and the lost energy consumption is less than or equal to a preset energy consumption threshold, retaining the multivariate monitoring data set corresponding to the current transformer; When the resource recovery rate is less than the resource recovery threshold, or the lost energy consumption is greater than the energy consumption threshold, the multivariate monitoring data set corresponding to the current transformer is deleted.

9. A transformer silicon steel sheet separation device, characterized in that: include: A data acquisition module, used to obtain real-time images and real-time separation force of silicon steel sheets of the transformer during separation; A quality impact area acquisition module, used for performing image recognition on the real-time image to obtain the quality impact area of ​​the silicon steel sheet; A quality impact parameter generation module, used to generate quality impact parameters using the quality impact area; A reference separation force determination module, used for inputting the quality influencing parameter into a separation force prediction model corresponding to the model of the transformer to obtain a corresponding reference separation force; a separation force adjustment module, configured to adjust the real-time separation force when the real-time separation force does not match the reference separation force; Wherein, the method for constructing the separation force prediction model includes: Applying a plurality of preset separation forces to the silicon steel sheets of at least two transformers corresponding to each model in sequence; After applying the preset separation force each time, obtaining a surface image of the silicon steel sheet; Analyzing the surface image to obtain quality influencing parameters of the silicon steel sheet corresponding to different preset separation forces; For each preset separation force, the minimum value among the mass influencing parameters is taken as the standard mass influencing parameter; Constructing a separation force prediction model between the standard quality influencing parameter and the preset separation force; wherein different types of transformers correspond to a separation force prediction model respectively; The quality influencing parameters include the deformation degree value, coating shedding degree value and insulating material residual degree value of the silicon steel sheet; then, the separation force prediction model includes: A first mathematical model of the deformation degree value and the first separation force; A second mathematical model of the coating shedding degree value and the second separation force; A third mathematical model of the insulating material residual degree value and the third separation force.

10. A transformer silicon steel sheet separation device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the transformer silicon steel sheet separation method according to any one of claims 1 to 8 when executing the computer program.

Images