A control method, device and control system for a fully automatic magnet dispensing machine

Abstract

The present invention relates to the field of dispensing equipment control technology, and in particular to a control method, device and control system for a fully automatic magnet dispensing machine. Calculate the structural similarity index between the actual state image and the ideal state image; if the structural similarity index is greater than the preset index threshold, correct the preset dispensing parameters, and use the corrected dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed; based on the optimal dispensing parameters, control the dispensing machine to dispense the magnet surface to be dispensed, obtain the real-time dispensing parameters of the dispensing machine at a preset time node, and compare and analyze the real-time dispensing parameters with the optimal dispensing parameters at the corresponding time node; if the real-time dispensing state of the dispensing machine is an abnormal state, regulate the dispensing machine. The automated control system automatically corrects the dispensing parameters in combination with the defect situation to ensure accurate coating and bonding of the glue, thereby improving product quality and consistency.

Description

Technical Field

[0001] The present invention relates to the technical field of glue dispensing equipment control, and in particular to a control method, device and control system for a fully automatic magnet glue dispensing machine. Background Art

[0002] The fully automatic magnet dispensing machine is a highly automated production equipment, which is mainly used for the process of accurately applying adhesive (usually glue) to the surface of magnets in the electronics, machinery or other industrial fields. This equipment realizes a series of automated operations from material loading, positioning, dispensing to curing by integrating precise mechanical systems, sensors and computer control programs. In the magnet manufacturing process, such as melting, pressing, sintering and other links, if the equipment accuracy is not high and the process control is not strict, defects may occur. Defects will cause the surface of the magnet to be uneven, resulting in the influence of the uniform distribution of glue during dispensing, thereby affecting the bonding strength and consistency of the magnet after dispensing. For example, areas with concave or crack defects may prevent the glue from being fully covered, resulting in weak or uneven bonding. Therefore, when dispensing magnets, if the defects in the magnets to be dispensed are not considered, the scrap rate of the finished dispensing products will be greatly increased, resulting in high processing costs. In addition, the current dispensing machine's control system has a low level of intelligence during the dispensing process, and it is difficult to adjust equipment parameters according to real-time dispensing conditions to adapt to dynamic changes in working conditions. This is also one of the reasons why the scrap rate of finished dispensing products and processing costs remain high. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a control method, device and control system for a fully automatic magnet dispensing machine.

[0004] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention discloses a control method for a fully automatic magnet dispensing machine, comprising the following steps:

[0006] Acquire an ideal state image of the magnet surface to be dispensed with glue, and acquire an actual state image of the magnet surface to be dispensed with glue, and calculate a structural similarity index between the actual state image and the ideal state image;

[0007] If the structural similarity index is greater than the preset index threshold, it means that there is no defect in the magnet surface to be glued, and the preset glue dispensing parameters of the glue dispensing machine are directly called up, and the preset glue dispensing parameters are used as the optimal glue dispensing parameters for the magnet surface to be glued;

[0008] If the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain the corrected glue dispensing parameters, and the corrected glue dispensing parameters are used as the optimal glue dispensing parameters for glue dispensing on the magnet surface to be glued;

[0009] Based on the optimal glue dispensing parameters, the glue dispensing machine is controlled to perform glue dispensing on the magnet surface to be glued, and during the actual glue dispensing work of the glue dispensing machine, the real-time glue dispensing parameters of the glue dispensing machine are obtained at a preset time node, and the real-time glue dispensing parameters are compared and analyzed with the optimal glue dispensing parameters at the corresponding time node to obtain the real-time glue dispensing state of the glue dispensing machine;

[0010] If the real-time dispensing state of the dispensing machine is in a normal state, the dispensing machine is not regulated; if the real-time dispensing state of the dispensing machine is in an abnormal state, the dispensing machine is regulated.

[0011] Furthermore, in a preferred embodiment of the present invention, if the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain the corrected glue dispensing parameters, specifically:

[0012] For the actual state image, the gradient strength and gradient direction of each pixel in the image are obtained, a neighborhood size is preset and a window is defined, the neighborhood window is slid in the image, the gradient co-occurrence between the gradient direction in each window and the reference direction is counted, and a gradient co-occurrence matrix of the actual state image is generated according to the gradient co-occurrence;

[0013] For the ideal state image, the gradient strength and gradient direction of each pixel in the image are obtained, a neighborhood size is preset and a window is defined, the neighborhood window is slid in the image, the gradient co-occurrence between the gradient direction in each window and the reference direction is counted, and a gradient co-occurrence matrix of the ideal state image is generated according to the gradient co-occurrence;

[0014] Respectively obtaining the elements at the same matrix position in the gradient co-occurrence matrix of the actual state image and the gradient co-occurrence matrix of the ideal state image, and calculating the difference between the elements at the same matrix position in the two matrices; and comparing the difference between the elements at the same matrix position in the two matrices with a preset difference threshold;

[0015] If the difference between the elements at the same matrix position in the two matrices is greater than a preset difference threshold, the corresponding position node in the actual state image is marked as a defective position node; if the difference between the elements at the same matrix position in the two matrices is not greater than the preset difference threshold, the corresponding position node in the actual state image is marked as a non-defective position node;

[0016] If a position node in the actual state image is a defective position node, the position node is rendered with the first color; if a position node in the actual state image is a non-defective position node, the position node is rendered with the second color; and so on, until all position nodes in the actual state image are rendered, and a defective position rendering image is obtained.

[0017] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0018] Obtaining a preset dispensing path of the dispensing machine, integrating the preset dispensing path into the defect position rendering, analyzing the intersection area of ​​the preset dispensing path and the first color in the defect position rendering to obtain a plurality of areas to be corrected;

[0019] Performing image segmentation processing on the actual state image to obtain sub-region images of each area to be corrected; and constructing a regional three-dimensional model map of each area to be corrected based on the corresponding sub-region images;

[0020] The model volume value of the three-dimensional model map of each area is calculated based on the gridding method to obtain the defect volume value of each area to be corrected; the amount of repair glue required to repair each area to be corrected is obtained according to the defect volume value of each area to be corrected, and the amount of repair glue of each area to be corrected is obtained;

[0021] According to the preset dispensing path, the preset dispensing time period corresponding to the dispensing machine dispensing glue on each area to be corrected during the actual working process is obtained, and the preset dispensing time period of each area to be corrected is obtained;

[0022] Obtain the preset dispensing amount of the dispensing machine when it works in each preset dispensing time period, and obtain the repair glue amount of each area to be corrected; correct the preset dispensing amount of the dispensing machine when it works in each preset dispensing time period according to the repair glue amount of each area to be corrected, and obtain the corrected dispensing amount of the dispensing machine when it works in each preset dispensing time period;

[0023] For the intersection area between the preset dispensing path and the second color, the preset dispensing parameters of the intersection area are not modified.

[0024] Furthermore, in a preferred embodiment of the present invention, the real-time dispensing parameters are compared and analyzed with the optimal dispensing parameters at the corresponding time nodes, and the real-time dispensing state of the dispensing machine is obtained by analysis, specifically:

[0025] Obtain the optimal dispensing parameters of the dispensing machine at a preset time node, and calculate the difference between various real-time dispensing parameters of the dispensing machine at the preset time node and the optimal dispensing parameters;

[0026] If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as an abnormal dispensing parameter;

[0027] If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is not greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as a normal dispensing parameter;

[0028] Analyze whether there are abnormal dispensing parameters of the dispensing machine at the preset time node. If so, the real-time dispensing state of the dispensing machine is defined as an abnormal state; if not, the real-time dispensing state of the dispensing machine is defined as a normal state.

[0029] Furthermore, in a preferred embodiment of the present invention, if the real-time dispensing state of the dispensing machine is abnormal, the dispensing machine is regulated and processed, specifically:

[0030] Obtaining work log information of the glue dispensing machine, and obtaining working condition characteristic images corresponding to various scrapped glue dispensing working conditions of the glue dispensing machine according to the work log information;

[0031] Construct a knowledge graph, import the working condition feature images corresponding to various scrapped dispensing working conditions of the dispensing machine into the knowledge graph; and regularly update the knowledge graph;

[0032] If the real-time dispensing state of the dispensing machine is abnormal, obtain the real-time dispensing working condition image of the dispensing magnet surface; calculate the cosine similarity between the real-time dispensing working condition image and each working condition feature image in the knowledge graph based on the cosine similarity algorithm to obtain a plurality of cosine similarities; compare the plurality of cosine similarities with the preset cosine similarity threshold;

[0033] If there is a situation where at least one cosine similarity is greater than a preset cosine similarity threshold, the dispensing machine is controlled to stop dispensing, and the magnet product being dispensed in the dispensing machine is immediately scrapped.

[0034] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0035] Obtain each execution unit in the dispensing machine, and obtain the functional information of each execution unit; based on the functional information of each execution unit, evaluate the linear correlation between each dispensing parameter and each execution unit;

[0036] If the linear correlation between a certain dispensing parameter and a certain execution unit is greater than the preset correlation, the execution unit is calibrated as the characteristic execution unit of the dispensing parameter; and so on, the characteristic execution units of various dispensing parameters are obtained;

[0037] Constructing a knowledge base, and importing the feature execution units of various dispensing parameters into the knowledge base;

[0038] If each cosine similarity is not greater than a preset cosine similarity threshold, the abnormal dispensing parameters are obtained, the abnormal dispensing parameters are imported into the knowledge base for pairing, a feature execution unit of the abnormal dispensing parameters is obtained, and the feature execution unit of the abnormal dispensing parameters is defined as a suspicious feature execution unit;

[0039] Acquire real-time electrical parameters of the suspicious feature execution unit, import the real-time electrical parameters into a Bayesian network for fault prediction, and obtain the fault probability of the suspicious feature execution unit;

[0040] If the failure probability of the suspicious feature execution unit is greater than the preset failure probability, the dispensing machine is controlled to stop dispensing, and failure information is generated, and the failure information is sent to a preset terminal for display;

[0041] If the failure probability of the suspicious feature execution unit is not greater than the preset failure probability, obtaining the preset electrical parameters of the suspicious feature execution unit, and calculating the electrical parameter value between the real-time electrical parameters and the preset electrical parameters;

[0042] If the electrical parameter value is not greater than the preset electrical parameter difference, the suspicious feature execution unit is marked as a normal feature execution unit; if the electrical parameter value is greater than the preset electrical parameter difference, the real-time electrical parameter is adjusted based on the electrical parameter value.

[0043] The second aspect of the present invention discloses a control system for a fully automatic magnet dispensing machine, the control system includes a memory and a processor, the memory stores a control method program for the fully automatic magnet dispensing machine, when the control method program for the fully automatic magnet dispensing machine is executed by the processor, any one of the control method steps for the fully automatic magnet dispensing machine is implemented.

[0044] The third aspect of the present invention discloses a control device for a fully automatic magnet dispensing machine, comprising:

[0045] An image acquisition module is used to acquire an ideal state image of the magnet surface to be dispensed with glue, and acquire an actual state image of the magnet surface to be dispensed with glue, and calculate a structural similarity index between the actual state image and the ideal state image;

[0046] A parameter retrieving module is used for directly retrieving the preset dispensing parameters of the dispensing machine if the structural similarity index is greater than a preset index threshold, indicating that there are no defects in the magnet surface to be dispensed, and using the preset dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed;

[0047] A parameter correction module, for correcting the preset dispensing parameters to obtain corrected dispensing parameters if the structural similarity index is not greater than a preset index threshold, indicating that there are defects in the magnet surface to be dispensed, and using the corrected dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed;

[0048] An analysis module is used to control the glue dispensing machine to perform glue dispensing on the magnet surface to be glued based on the optimal glue dispensing parameters, and obtain the real-time glue dispensing parameters of the glue dispensing machine at a preset time node during the actual glue dispensing operation of the glue dispensing machine, compare and analyze the real-time glue dispensing parameters with the optimal glue dispensing parameters at the corresponding time node, and obtain the real-time glue dispensing state of the glue dispensing machine by analysis;

[0049] The control module is used for not controlling the glue dispensing machine if the real-time glue dispensing state of the glue dispensing machine is in a normal state; and for controlling the glue dispensing machine if the real-time glue dispensing state of the glue dispensing machine is in an abnormal state.

[0050] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: the automatic control system automatically corrects the dispensing parameters by combining the defect conditions, thereby ensuring the accurate coating and bonding of the glue and improving product quality and consistency; the automatic control system can quickly adjust the dispensing parameters according to the real-time dispensing conditions, adapt to different production needs and working condition changes, improve production flexibility, effectively improve product yield and improve economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0052] Figure 1 It is a flow chart of the overall method of a control method of a fully automatic magnet dispensing machine;

[0053] Figure 2 It is a partial method flow chart of a control method of a fully automatic magnet dispensing machine;

[0054] Figure 3 The system block diagram of a control system for a fully automatic magnet dispensing machine. DETAILED DESCRIPTION

[0055] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0056] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0057] like Figure 1 As shown, the first aspect of the present invention discloses a control method for a fully automatic magnet dispensing machine, comprising the following steps:

[0058] S102: obtaining an ideal state image of the magnet surface to be dispensed with glue, and obtaining an actual state image of the magnet surface to be dispensed with glue, and calculating a structural similarity index between the actual state image and the ideal state image;

[0059] S104: If the structural similarity index is greater than a preset index threshold, it means that there is no defect in the magnet surface to be glued, and the preset glue dispensing parameters of the glue dispensing machine are directly called up, and the preset glue dispensing parameters are used as the optimal glue dispensing parameters for glue dispensing on the magnet surface to be glued;

[0060] S106: If the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain corrected glue dispensing parameters, and the corrected glue dispensing parameters are used as the optimal glue dispensing parameters for glue dispensing on the magnet surface to be glued;

[0061] S108: Based on the optimal glue dispensing parameters, the glue dispensing machine is controlled to perform glue dispensing on the magnet surface to be glued, and during the actual glue dispensing operation of the glue dispensing machine, the real-time glue dispensing parameters of the glue dispensing machine are obtained at a preset time node, and the real-time glue dispensing parameters are compared and analyzed with the optimal glue dispensing parameters at the corresponding time node to obtain the real-time glue dispensing state of the glue dispensing machine;

[0062] S110: If the real-time dispensing state of the dispensing machine is normal, the dispensing machine is not regulated; if the real-time dispensing state of the dispensing machine is abnormal, the dispensing machine is regulated.

[0063] It should be noted that the ideal state image of the magnet surface to be glued refers to the ideal pre-treatment state of the magnet surface before the glue dispensing process, that is, the surface should be smooth and flat, without defects such as dents, cracks, scratches, etc., so that the glue can be evenly applied to ensure consistent bonding effect. The structural similarity index is a quantitative method for evaluating the visual similarity between two images. It is based on the characteristics of the human visual system's perception of image structure and regards the image as consisting of three parts: brightness, contrast, and structure. When the structural similarity index is close to 1, it means that the two images are very similar in brightness, contrast, and structure; the lower the value, the greater the difference. If the structural similarity index is greater than the preset index threshold, it means that the actual state image is highly similar to the ideal state image, which means that there are no defects (such as dents, cracks, scratches, etc.) in the magnet surface to be glued. At this time, the preset glue dispensing parameters of the glue dispensing machine are directly called up, and the preset glue dispensing parameters are used as the optimal glue dispensing parameters for the magnet surface to be glued. Among them, the preset dispensing parameters are preset parameters designed and planned by relevant technical personnel based on the ideal state of the magnet surface to be dispensed (that is, there are no defects). The preset dispensing parameters include but are not limited to dispensing speed, dispensing pressure, dispensing amount per unit time, needle tip diameter, dispensing temperature and humidity.

[0064] like Figure 2 As shown, further, in a preferred embodiment of the present invention, if the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain the corrected glue dispensing parameters, specifically:

[0065] S202: for the actual state image, obtain the gradient intensity and gradient direction of each pixel in the image, preset the neighborhood size and define a window, slide the neighborhood window in the image, count the gradient co-occurrence between the gradient direction in each window and the reference direction, and generate a gradient co-occurrence matrix of the actual state image according to the gradient co-occurrence;

[0066] S204: for the ideal state image, obtain the gradient intensity and gradient direction of each pixel in the image, preset the neighborhood size and define a window, slide the neighborhood window in the image, count the gradient co-occurrence between the gradient direction in each window and the reference direction, and generate a gradient co-occurrence matrix of the ideal state image according to the gradient co-occurrence;

[0067] S206: respectively obtaining the elements at the same matrix position in the gradient co-occurrence matrix of the actual state image and the gradient co-occurrence matrix of the ideal state image, and calculating the difference between the elements at the same matrix position in the two matrices; and comparing the difference between the elements at the same matrix position in the two matrices with a preset difference threshold;

[0068] S208: if the difference between the elements at the same matrix position in the two matrices is greater than a preset difference threshold, the corresponding position node in the actual state image is marked as a defective position node; if the difference between the elements at the same matrix position in the two matrices is not greater than the preset difference threshold, the corresponding position node in the actual state image is marked as a non-defective position node;

[0069] S210: If a position node in the actual state image is a defective position node, the position node is rendered with a first color; if a position node in the actual state image is a non-defective position node, the position node is rendered with a second color; and so on, until all position nodes in the actual state image are rendered, and a defective position rendering image is obtained.

[0070] It should be noted that if the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be dispensed, and at this time, the actual state image and the ideal state image are converted into a gradient co-occurrence matrix. Then, the elements at the same matrix position in the gradient co-occurrence matrix of the actual state image and the gradient co-occurrence matrix of the ideal state image are compared respectively. If the difference between the elements at the same matrix position in the two matrices is greater than the preset difference threshold, the corresponding position node in the actual state image is marked as a defective position node. If a position node in the actual state image is a defective position node, the position node is rendered with the first color; if a position node in the actual state image is a non-defective position node, the position node is rendered with the second color, thereby obtaining a defective position rendering diagram, wherein the first color and the second color should be different colors that are easy to distinguish, such as the first color can be red, and the second color is green. In this way, the defective position area in the magnet surface to be dispensed can be quickly analyzed, and the defective position rendering diagram of the magnet surface to be dispensed can be obtained, so that the dispensing parameters can be corrected in combination with the preset dispensing path in the future, and the yield rate of the dispensing finished product can be improved.

[0071] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0072] Obtaining a preset dispensing path of the dispensing machine, integrating the preset dispensing path into the defect position rendering, analyzing the intersection area of ​​the preset dispensing path and the first color in the defect position rendering to obtain a plurality of areas to be corrected;

[0073] Performing image segmentation processing on the actual state image to obtain sub-region images of each area to be corrected; and constructing a regional three-dimensional model map of each area to be corrected based on the corresponding sub-region images;

[0074] The model volume value of the three-dimensional model map of each area is calculated based on the gridding method to obtain the defect volume value of each area to be corrected; the amount of repair glue required to repair each area to be corrected is obtained according to the defect volume value of each area to be corrected, and the amount of repair glue of each area to be corrected is obtained;

[0075] According to the preset dispensing path, the preset dispensing time period corresponding to the dispensing machine dispensing glue on each area to be corrected during the actual working process is obtained, and the preset dispensing time period of each area to be corrected is obtained;

[0076] Obtain the preset dispensing amount of the dispensing machine when it works in each preset dispensing time period, and obtain the repair glue amount of each area to be corrected; correct the preset dispensing amount of the dispensing machine when it works in each preset dispensing time period according to the repair glue amount of each area to be corrected, and obtain the corrected dispensing amount of the dispensing machine when it works in each preset dispensing time period;

[0077] For the intersection area between the preset dispensing path and the second color, the preset dispensing parameters of the intersection area are not modified.

[0078] It should be noted that, according to the position of each area to be corrected, the actual state image is segmented in combination with the image segmentation algorithm to obtain the sub-region image of each area to be corrected, and then the three-dimensional reconstruction technology is used to reconstruct the regional three-dimensional model map of each area to be corrected based on the sub-region image. At this time, the defect three-dimensional model map of each position area defect in the magnet surface to be dispensed is obtained, such as the three-dimensional model map of the concave defect in a certain position area; then by calculating the volume value of the corresponding regional three-dimensional model map, the amount of repair glue required to repair the corresponding area to be corrected can be determined. Specifically, if there are defects such as concave, cracks, scratches, etc. in a certain area, the corresponding defect position can be "filled" by spraying more glue in these areas, thereby realizing the function of "repairing" defects in the dispensing process, thereby avoiding the existence of defects affecting the uniform distribution of glue, affecting the bonding strength and consistency of the magnet after dispensing. Through this method, the preset dispensing parameters of the dispensing machine can be corrected according to the defect situation, eliminating the influence of defects on the uniform distribution of glue after dispensing, realizing intelligent dispensing processing, and effectively improving the product yield.

[0079] Furthermore, in a preferred embodiment of the present invention, the real-time dispensing parameters are compared and analyzed with the optimal dispensing parameters at the corresponding time nodes, and the real-time dispensing state of the dispensing machine is obtained by analysis, specifically:

[0080] Obtain the optimal dispensing parameters of the dispensing machine at a preset time node, and calculate the difference between various real-time dispensing parameters of the dispensing machine at the preset time node and the optimal dispensing parameters;

[0081] If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as an abnormal dispensing parameter;

[0082] If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is not greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as a normal dispensing parameter;

[0083] Analyze whether there are abnormal dispensing parameters of the dispensing machine at the preset time node. If so, the real-time dispensing state of the dispensing machine is defined as an abnormal state; if not, the real-time dispensing state of the dispensing machine is defined as a normal state.

[0084] It should be noted that after obtaining the optimal gluing parameters, the gluing machine is controlled to perform gluing on the magnet surface to be glued based on the optimal gluing parameters, and during the actual gluing work of the gluing machine, the real-time gluing parameters of the gluing machine are obtained at the preset time node, and the real-time gluing parameters are compared and analyzed with the optimal gluing parameters at the corresponding time node. If the difference between a certain real-time gluing parameter and the optimal gluing parameter is greater than the preset threshold value, the real-time gluing parameter of the gluing machine is marked as an abnormal gluing parameter, otherwise, the real-time gluing parameter is a normal gluing parameter; if the gluing machine has abnormal gluing parameters at the preset time node, the real-time gluing state of the gluing machine is defined as an abnormal state, otherwise, the real-time gluing state is defined as a normal state.

[0085] Furthermore, in a preferred embodiment of the present invention, if the real-time dispensing state of the dispensing machine is abnormal, the dispensing machine is regulated and processed, specifically:

[0086] Obtaining work log information of the glue dispensing machine, and obtaining working condition characteristic images corresponding to various scrapped glue dispensing working conditions of the glue dispensing machine according to the work log information;

[0087] Construct a knowledge graph, import the working condition feature images corresponding to various scrapped dispensing working conditions of the dispensing machine into the knowledge graph; and regularly update the knowledge graph;

[0088] If the real-time dispensing state of the dispensing machine is abnormal, obtain the real-time dispensing working condition image of the dispensing magnet surface; calculate the cosine similarity between the real-time dispensing working condition image and each working condition feature image in the knowledge graph based on the cosine similarity algorithm to obtain a plurality of cosine similarities; compare the plurality of cosine similarities with the preset cosine similarity threshold;

[0089] If there is a situation where at least one cosine similarity is greater than a preset cosine similarity threshold, the dispensing machine is controlled to stop dispensing, and the magnet product being dispensed in the dispensing machine is immediately scrapped.

[0090] It should be noted that the work log information of the dispensing machine refers to the records and data generated during the operation of the dispensing machine, including the execution of the dispensing task, the operator's operation record, the equipment operation status, abnormal alarms, dispensing parameter settings, production plans and other information. These log information can be used to monitor the production process, track problems and failures, evaluate production efficiency, optimize production plans, and perform fault diagnosis and maintenance when necessary. By analyzing and utilizing work log information, production managers can better understand the operation of the dispensing machine and improve production efficiency and quality. Among them, scrap dispensing conditions include glue overflow or contamination of non-dispensing areas, inappropriate curing temperature resulting in uneven curing and agglomeration, etc.

[0091] It should be noted that if the real-time dispensing state of the dispensing machine is abnormal, the real-time dispensing condition image of the magnet surface being dispensed is obtained at this time, and the cosine similarity between the real-time dispensing condition image and each condition feature image in the knowledge graph is calculated based on the cosine similarity algorithm. If there is at least one cosine similarity greater than the preset cosine similarity threshold, it means that the semi-finished product being dispensed at this time has become scrapped. At this time, even if the semi-finished product continues to be dispensed, the finished product is still scrap. In this way, the dispensing machine is controlled to stop dispensing, and the magnet product being dispensed in the dispensing machine is immediately scrapped. In this way, the scrapped semi-finished products can be identified in time and scrapped in time, thereby reducing the waste of processing resources and saving processing costs.

[0092] Furthermore, in a preferred embodiment of the present invention, the following steps are also included:

[0093] Obtain each execution unit in the dispensing machine, and obtain the functional information of each execution unit; based on the functional information of each execution unit, evaluate the linear correlation between each dispensing parameter and each execution unit;

[0094] If the linear correlation between a certain dispensing parameter and a certain execution unit is greater than the preset correlation, the execution unit is calibrated as the characteristic execution unit of the dispensing parameter; and so on, the characteristic execution units of various dispensing parameters are obtained;

[0095] Constructing a knowledge base, and importing the feature execution units of various dispensing parameters into the knowledge base;

[0096] If each cosine similarity is not greater than a preset cosine similarity threshold, the abnormal dispensing parameters are obtained, the abnormal dispensing parameters are imported into the knowledge base for pairing, a feature execution unit of the abnormal dispensing parameters is obtained, and the feature execution unit of the abnormal dispensing parameters is defined as a suspicious feature execution unit;

[0097] Acquire real-time electrical parameters of the suspicious feature execution unit, import the real-time electrical parameters into a Bayesian network for fault prediction, and obtain the fault probability of the suspicious feature execution unit;

[0098] If the failure probability of the suspicious feature execution unit is greater than the preset failure probability, the dispensing machine is controlled to stop dispensing, and failure information is generated, and the failure information is sent to a preset terminal for display;

[0099] If the failure probability of the suspicious feature execution unit is not greater than the preset failure probability, obtaining the preset electrical parameters of the suspicious feature execution unit, and calculating the electrical parameter value between the real-time electrical parameters and the preset electrical parameters;

[0100] If the electrical parameter value is not greater than the preset electrical parameter difference, the suspicious feature execution unit is marked as a normal feature execution unit; if the electrical parameter value is greater than the preset electrical parameter difference, the real-time electrical parameter is adjusted based on the electrical parameter value.

[0101] It should be noted that the execution unit of the glue machine refers to the parts or components used to actually perform the dispensing operation. They are responsible for controlling the flow and distribution of glue to ensure accurate dispensing effects, including: dispensing valves, needles, pressure controllers, nozzles, control systems, motion systems, temperature control systems, etc. If the linear correlation between a certain dispensing parameter and a certain execution unit is greater than the preset correlation, the execution unit is calibrated as the characteristic execution unit of the dispensing parameter. If the dispensing pressure parameter is abnormal, it is often related to the pressure controller, and the pressure controller is the characteristic execution unit of the dispensing pressure parameter.

[0102] If all cosine similarities are not greater than the preset cosine similarity threshold, it means that although abnormal dispensing parameters have appeared in the dispensing machine, the semi-finished product in the dispensing process has not appeared scrapped. At this time, the corresponding suspicious feature execution unit is matched in the knowledge base, and then the real-time electrical parameters of the suspicious feature execution unit are obtained, and the electrical parameters include voltage, current, power, etc. The real-time electrical parameters are imported into the Bayesian network for fault prediction to obtain the failure probability of the suspicious feature execution unit. If the failure probability of the suspicious feature execution unit is greater than the preset failure probability, it means that the execution unit has failed, then the dispensing machine is controlled to stop dispensing, and fault information is generated, and the fault information is sent to the preset terminal for display, so as to timely control the dispensing machine to stop production, avoid the phenomenon of producing a large number of defective products due to equipment failure, and realize intelligent fault diagnosis. If the electrical parameter value is not greater than the preset electrical parameter difference, it means that the dispensing parameter abnormality is not related to the execution unit, and the dispensing parameter abnormality may be caused by accidental factors (such as vibration) or other execution units, then the suspicious feature execution unit is marked as a normal feature execution unit; if the electrical parameter value is greater than the preset electrical parameter difference, it means that the dispensing parameter abnormality is related to the execution unit, then the real-time electrical parameters are adjusted based on the electrical parameter value, so as to adjust the parameters of the execution unit to an appropriate range, thereby realizing intelligent control.

[0103] Among them, the Bayesian network is a probabilistic graph model used to represent the probabilistic dependencies between variables. It consists of a set of nodes (representing random variables) and directed edges (representing the dependencies between variables), each node represents a random variable, and each edge represents the probabilistic dependencies between variables. Bayesian networks can be used to infer the probability distribution of unknown variables, perform probabilistic reasoning and prediction, and help decision-making and problem solving. By observing the values ​​of known variables and combining the probability distribution in the Bayesian network, the probability distribution of other variables can be inferred, thereby helping to understand the behavior and relationships of complex systems. The principle of importing the real-time electrical parameters of the execution unit into the Bayesian network for fault prediction is to use the Bayesian network model to analyze the electrical parameter data of the execution unit and infer the probability that the execution unit may have a fault. First, the real-time electrical parameter data of the execution unit, such as current, voltage, power, etc., are collected as observed variables. Then, a Bayesian network model is constructed based on these electrical parameter data, considering the potential dependencies between different electrical parameters. Finally, by learning and training the known fault data, the Bayesian inference method is used to calculate the probability of fault in each execution unit, and the execution units with suspicious characteristics are identified, thereby achieving fault prediction and timely maintenance.

[0104] In summary, the automated control system automatically corrects the dispensing parameters based on the defect conditions to ensure accurate coating and bonding of the glue, thereby improving product quality and consistency; the automated control system can quickly adjust the dispensing parameters according to real-time dispensing conditions to adapt to different production needs and working condition changes, improve production flexibility, and effectively improve product yield and economic benefits.

[0105] In addition, the method further comprises the following steps:

[0106] After the glue dispensing on the magnet surface to be glued is completed, the actual glue shape image of the magnet sealing surface is obtained, and the actual glue shape three-dimensional model diagram is constructed according to the actual glue state image; and the standard glue shape three-dimensional model diagram of the magnet sealing surface is obtained;

[0107] The degree of overlap between the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram is calculated based on the Euclidean distance algorithm; if the degree of overlap is greater than a preset degree of overlap, the magnet after glue dispensing is calibrated as a qualified product in terms of appearance characteristics;

[0108] If the overlap is not greater than the preset overlap, the magnet after glue dispensing is marked as a defective product in terms of appearance characteristics, and a virtual space is constructed, and the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram are imported into the virtual space;

[0109] The actual glue morphology 3D model image and the standard glue morphology 3D model image are registered in a virtual space to obtain an intersection 3D model image; the intersection 3D model image is analyzed to find out the non-overlapping area between the actual glue morphology 3D model image and the standard glue morphology 3D model image, which is defined as the glue dispensing abnormal area;

[0110] According to the preset dispensing path, a preset dispensing time node for dispensing the abnormal dispensing area is obtained, the real-time dispensing parameters of the dispensing machine are obtained, and a Markov chain is introduced. Based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine, the state transition probability of the dispensing machine at the preset dispensing time node is calculated;

[0111] If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node.

[0112] The following steps are also included:

[0113] If the overlap is greater than a preset overlap, the glue on the magnet sealing surface is scanned and detected by an ultrasonic detector, and ultrasonic data fed back by the glue on the magnet sealing surface is obtained;

[0114] Performing feature extraction processing on the ultrasonic data to obtain bubble feature data inside the glue on the magnet sealing surface, the bubble feature data including bubble size and bubble shape, and obtaining the bubble concentration value of the glue on the magnet sealing surface according to the bubble feature data;

[0115] If the bubble concentration value is not greater than the preset concentration value, the magnet after dispensing is calibrated as a qualified product with internal characteristics;

[0116] If the bubble concentration value is greater than the preset concentration value, the magnet after dispensing is calibrated as an unqualified product with internal characteristics, and a three-dimensional model diagram of the actual characteristics of the glue in the magnet sealing surface is constructed based on the ultrasonic data;

[0117] Constructing a second virtual space, importing the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image into the second virtual space for analysis, and defining the area where the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image do not overlap as the abnormal glue dispensing area;

[0118] According to the preset dispensing path, a preset dispensing time node for dispensing the abnormal dispensing area is obtained, the real-time dispensing parameters of the dispensing machine are obtained, and a Markov chain is introduced. Based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine, the state transition probability of the dispensing machine at the preset dispensing time node is calculated;

[0119] If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node.

[0120] It should be noted that by analyzing the actual glue surface and internal morphological characteristics of the magnet sealing surface, it is possible to determine whether the quality of the finished dispensing product is qualified. If it is unqualified, the dispensing abnormal area is further analyzed, and the real-time dispensing parameters of the dispensing machine at the corresponding dispensing time node are obtained according to the dispensing abnormal area. Then, by real-time monitoring of the parameter data of the dispensing machine, the state space and state transfer matrix are established; based on the Markov chain, the dependency between the current state and the previous state is considered, and the probability distribution of the dispensing machine state at a given time node is calculated; finally, combined with the actual dispensing machine operation data and parameter changes, the possible state of the dispensing machine at the preset time node is inferred, and the state transition probability is calculated, so as to perform fault prediction and production adjustment, and thus obtain the equipment abnormal time node. In general, if the dispensing machine finds that the state transition probability deviates abnormally from expectations at a specific time node, it means that the dispensing machine has signs of abnormality or failure. By analyzing the purpose of the abnormal time node, the relevant technical personnel can adjust the dispensing machine operation strategy according to the occurrence time of the corresponding abnormal time node, and optimize the dispensing machine operation parameters in a targeted manner to improve the reliability of the dispensing machine, reduce the risk of failure, and ensure the stability and efficiency of the production process.

[0121] like Figure 3 As shown, the second aspect of the present invention discloses a control system of a fully automatic magnet dispensing machine, the control system includes a memory 20 and a processor 80, the memory 20 stores a control method program of the fully automatic magnet dispensing machine, when the control method program of the fully automatic magnet dispensing machine is executed by the processor 80, any one of the control method steps of the fully automatic magnet dispensing machine is implemented.

[0122] The third aspect of the present invention discloses a control device for a fully automatic magnet dispensing machine, comprising:

[0123] An image acquisition module is used to acquire an ideal state image of the magnet surface to be dispensed with glue, and acquire an actual state image of the magnet surface to be dispensed with glue, and calculate a structural similarity index between the actual state image and the ideal state image;

[0124] A parameter retrieving module is used for directly retrieving the preset dispensing parameters of the dispensing machine if the structural similarity index is greater than a preset index threshold, indicating that there are no defects in the magnet surface to be dispensed, and using the preset dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed;

[0125] A parameter correction module, for correcting the preset dispensing parameters to obtain corrected dispensing parameters if the structural similarity index is not greater than a preset index threshold, indicating that there are defects in the magnet surface to be dispensed, and using the corrected dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed;

[0126] An analysis module is used to control the glue dispensing machine to perform glue dispensing on the magnet surface to be glued based on the optimal glue dispensing parameters, and obtain the real-time glue dispensing parameters of the glue dispensing machine at a preset time node during the actual glue dispensing operation of the glue dispensing machine, compare and analyze the real-time glue dispensing parameters with the optimal glue dispensing parameters at the corresponding time node, and obtain the real-time glue dispensing state of the glue dispensing machine by analysis;

[0127] The control module is used for not controlling the glue dispensing machine if the real-time glue dispensing state of the glue dispensing machine is in a normal state; and for controlling the glue dispensing machine if the real-time glue dispensing state of the glue dispensing machine is in an abnormal state.

[0128] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0129] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0130] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0131] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0132] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0133] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A control method for a fully automatic magnet dispensing machine, characterized in that: The following steps are involved: Acquire an ideal state image of the magnet surface to be dispensed with glue, and acquire an actual state image of the magnet surface to be dispensed with glue, and calculate a structural similarity index between the actual state image and the ideal state image; If the structural similarity index is greater than the preset index threshold, it means that there is no defect in the magnet surface to be glued, and the preset glue dispensing parameters of the glue dispensing machine are directly called up, and the preset glue dispensing parameters are used as the optimal glue dispensing parameters for the magnet surface to be glued; If the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain the corrected glue dispensing parameters, and the corrected glue dispensing parameters are used as the optimal glue dispensing parameters for glue dispensing on the magnet surface to be glued; Based on the optimal glue dispensing parameters, the glue dispensing machine is controlled to perform glue dispensing on the magnet surface to be glued, and during the actual glue dispensing work of the glue dispensing machine, the real-time glue dispensing parameters of the glue dispensing machine are obtained at a preset time node, and the real-time glue dispensing parameters are compared and analyzed with the optimal glue dispensing parameters at the corresponding time node to obtain the real-time glue dispensing state of the glue dispensing machine; If the real-time dispensing status of the dispensing machine is normal, the dispensing machine will not be regulated; If the real-time dispensing state of the dispensing machine is abnormal, the dispensing machine is regulated; The following steps are also included: Construct a three-dimensional model diagram of the actual glue shape, and obtain a three-dimensional model diagram of the standard glue shape of the magnet sealing surface; Calculate the degree of overlap between the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram; if the degree of overlap is greater than a preset degree of overlap, calibrate the magnet after glue dispensing as a qualified product with good appearance characteristics; If the overlap is not greater than the preset overlap, the magnet after glue dispensing is marked as a defective product in terms of appearance characteristics, and a virtual space is constructed, and the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram are imported into the virtual space; The actual glue morphology 3D model image and the standard glue morphology 3D model image are registered in a virtual space to obtain an intersection 3D model image; the intersection 3D model image is analyzed to find out the non-overlapping area between the actual glue morphology 3D model image and the standard glue morphology 3D model image, which is defined as the glue dispensing abnormal area; Obtain a preset dispensing time node for dispensing the dispensing abnormal area, obtain the real-time dispensing parameters of the dispensing machine, and calculate the state transition probability of the dispensing machine at the preset dispensing time node based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine; If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node; If the overlap is greater than a preset overlap, ultrasonic data fed back by the glue on the magnet sealing surface is obtained; Performing feature extraction processing on the ultrasonic data to obtain feature data of bubbles inside the glue on the magnet sealing surface; If the bubble concentration value is not greater than the preset concentration value, the magnet after dispensing is calibrated as a qualified product with internal characteristics; If the bubble concentration value is greater than the preset concentration value, the magnet after dispensing is calibrated as an unqualified product with internal characteristics, and a three-dimensional model diagram of the actual characteristics of the glue in the magnet sealing surface is constructed based on the ultrasonic data; Constructing a second virtual space, importing the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image into the second virtual space for analysis, and defining the area where the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image do not overlap as the abnormal glue dispensing area; According to the preset dispensing path, a preset dispensing time node for dispensing the abnormal dispensing area is obtained, and the real-time dispensing parameters of the dispensing machine are obtained. Based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine, the state transition probability of the dispensing machine at the preset dispensing time node is calculated; If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node.

2. The control method of a fully automatic magnet dispensing machine according to claim 1 is characterized in that: If the structural similarity index is not greater than the preset index threshold, it means that there are defects in the magnet surface to be glued, and the preset glue dispensing parameters are corrected to obtain the corrected glue dispensing parameters, which are specifically: For the actual state image, the gradient strength and gradient direction of each pixel in the image are obtained, a neighborhood size is preset and a window is defined, the neighborhood window is slid in the image, the gradient co-occurrence between the gradient direction in each window and the reference direction is counted, and a gradient co-occurrence matrix of the actual state image is generated according to the gradient co-occurrence; For the ideal state image, the gradient strength and gradient direction of each pixel in the image are obtained, a neighborhood size is preset and a window is defined, the neighborhood window is slid in the image, the gradient co-occurrence between the gradient direction in each window and the reference direction is counted, and a gradient co-occurrence matrix of the ideal state image is generated according to the gradient co-occurrence; Respectively obtaining the elements at the same matrix position in the gradient co-occurrence matrix of the actual state image and the gradient co-occurrence matrix of the ideal state image, and calculating the difference between the elements at the same matrix position in the two matrices; and comparing the difference between the elements at the same matrix position in the two matrices with a preset difference threshold; If the difference between the elements at the same matrix position in the two matrices is greater than a preset difference threshold, the corresponding position node in the actual state image is marked as a defective position node; if the difference between the elements at the same matrix position in the two matrices is not greater than the preset difference threshold, the corresponding position node in the actual state image is marked as a non-defective position node; If a certain position node in the actual state image is a defective position node, the position node is rendered with the first color; If a certain position node in the actual state image is a non-defective position node, the position node is rendered with a second color; This process is repeated in this way until all position nodes in the actual state image are rendered, and a defect position rendering image is obtained.

3. The control method of a fully automatic magnet dispensing machine according to claim 2 further comprises the following steps: Obtaining a preset dispensing path of the dispensing machine, integrating the preset dispensing path into the defect position rendering, analyzing the intersection area of ​​the preset dispensing path and the first color in the defect position rendering to obtain a plurality of areas to be corrected; Performing image segmentation processing on the actual state image to obtain sub-region images of each area to be corrected; and constructing a regional three-dimensional model map of each area to be corrected based on the corresponding sub-region images; Calculate the model volume value of the three-dimensional model map of each area based on the grid method to obtain the defect volume value of each area to be corrected; According to the defect volume value of each area to be corrected, the amount of repair glue required to repair each area to be corrected is obtained, and the amount of repair glue for each area to be corrected is obtained; According to the preset dispensing path, the preset dispensing time period corresponding to the dispensing machine dispensing glue on each area to be corrected during the actual working process is obtained, and the preset dispensing time period of each area to be corrected is obtained; Obtaining the preset dispensing amount of the dispensing machine when it works in each preset dispensing time period, and obtaining the repair glue amount of each area to be corrected; According to the repair glue amount of each area to be corrected, the preset glue dispensing amount of the glue dispensing machine when working in each preset glue dispensing time period is corrected to obtain the corrected glue dispensing amount of the glue dispensing machine when working in each preset glue dispensing time period; For the intersection area between the preset dispensing path and the second color, the preset dispensing parameters of the intersection area are not modified.

4. The control method of a fully automatic magnet dispensing machine according to claim 1 is characterized in that: The real-time dispensing parameters are compared and analyzed with the optimal dispensing parameters at the corresponding time node, and the real-time dispensing status of the dispensing machine is obtained by analysis, specifically: Obtain the optimal dispensing parameters of the dispensing machine at a preset time node, and calculate the difference between various real-time dispensing parameters of the dispensing machine at the preset time node and the optimal dispensing parameters; If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as an abnormal dispensing parameter; If the difference between a certain real-time dispensing parameter and the optimal dispensing parameter is not greater than a preset threshold, the real-time dispensing parameter of the dispensing machine is marked as a normal dispensing parameter; Analyze whether there are abnormal dispensing parameters of the dispensing machine at the preset time node. If so, the real-time dispensing state of the dispensing machine is defined as an abnormal state; if not, the real-time dispensing state of the dispensing machine is defined as a normal state.

5. The control method of a fully automatic magnet dispensing machine according to claim 1 is characterized in that: If the real-time dispensing status of the dispensing machine is abnormal, the dispensing machine is regulated and processed, specifically: Obtaining work log information of the glue dispensing machine, and obtaining working condition characteristic images corresponding to various scrapped glue dispensing working conditions of the glue dispensing machine according to the work log information; Construct a knowledge graph, import the working condition feature images corresponding to various scrapped dispensing working conditions of the dispensing machine into the knowledge graph; and regularly update the knowledge graph; If the real-time dispensing state of the dispensing machine is abnormal, obtain the real-time dispensing working condition image of the dispensing magnet surface; calculate the cosine similarity between the real-time dispensing working condition image and each working condition feature image in the knowledge graph based on the cosine similarity algorithm to obtain a plurality of cosine similarities; compare the plurality of cosine similarities with the preset cosine similarity threshold; If there is a situation where at least one cosine similarity is greater than a preset cosine similarity threshold, the dispensing machine is controlled to stop dispensing, and the magnet product being dispensed in the dispensing machine is immediately scrapped.

6. The control method of a fully automatic magnet dispensing machine according to claim 5 is characterized in that: The following steps are also included: Obtain each execution unit in the dispensing machine, and obtain the functional information of each execution unit; based on the functional information of each execution unit, evaluate the linear correlation between each dispensing parameter and each execution unit; If the linear correlation between a certain dispensing parameter and a certain execution unit is greater than the preset correlation, the execution unit is calibrated as the characteristic execution unit of the dispensing parameter; By analogy, the characteristic execution units of various dispensing parameters are obtained; Constructing a knowledge base, and importing the feature execution units of various dispensing parameters into the knowledge base; If each cosine similarity is not greater than a preset cosine similarity threshold, the abnormal dispensing parameters are obtained, the abnormal dispensing parameters are imported into the knowledge base for pairing, a feature execution unit of the abnormal dispensing parameters is obtained, and the feature execution unit of the abnormal dispensing parameters is defined as a suspicious feature execution unit; Acquire real-time electrical parameters of the suspicious feature execution unit, import the real-time electrical parameters into a Bayesian network for fault prediction, and obtain the fault probability of the suspicious feature execution unit; If the failure probability of the suspicious feature execution unit is greater than the preset failure probability, the dispensing machine is controlled to stop dispensing, and failure information is generated, and the failure information is sent to a preset terminal for display; If the failure probability of the suspicious feature execution unit is not greater than the preset failure probability, obtaining the preset electrical parameters of the suspicious feature execution unit, and calculating the electrical parameter value between the real-time electrical parameters and the preset electrical parameters; If the electrical parameter value is not greater than the preset electrical parameter difference, the suspicious feature execution unit is marked as a normal feature execution unit; if the electrical parameter value is greater than the preset electrical parameter difference, the real-time electrical parameter is adjusted based on the electrical parameter value.

7. A control system for a fully automatic magnet dispensing machine, characterized in that: The control system includes a memory and a processor, wherein the memory stores a control method program for a fully automatic magnet dispensing machine. When the control method program for the fully automatic magnet dispensing machine is executed by the processor, the control method steps for the fully automatic magnet dispensing machine as described in any one of claims 1 to 6 are implemented.

8. A control device for a fully automatic magnet dispensing machine, characterized in that: include: An image acquisition module is used to acquire an ideal state image of the magnet surface to be dispensed with glue, and acquire an actual state image of the magnet surface to be dispensed with glue, and calculate a structural similarity index between the actual state image and the ideal state image; A parameter retrieving module is used for directly retrieving the preset dispensing parameters of the dispensing machine if the structural similarity index is greater than a preset index threshold, indicating that there are no defects in the magnet surface to be dispensed, and using the preset dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed; A parameter correction module, for correcting the preset dispensing parameters to obtain corrected dispensing parameters if the structural similarity index is not greater than a preset index threshold, indicating that there are defects in the magnet surface to be dispensed, and using the corrected dispensing parameters as the optimal dispensing parameters for dispensing the magnet surface to be dispensed; An analysis module is used to control the glue dispensing machine to perform glue dispensing on the magnet surface to be glued based on the optimal glue dispensing parameters, and obtain the real-time glue dispensing parameters of the glue dispensing machine at a preset time node during the actual glue dispensing operation of the glue dispensing machine, compare and analyze the real-time glue dispensing parameters with the optimal glue dispensing parameters at the corresponding time node, and obtain the real-time glue dispensing state of the glue dispensing machine by analysis; The control module is used for not controlling the dispensing machine if the real-time dispensing state of the dispensing machine is normal; If the real-time dispensing state of the dispensing machine is abnormal, the dispensing machine is regulated; Also includes: Construct a three-dimensional model diagram of the actual glue shape, and obtain a three-dimensional model diagram of the standard glue shape of the magnet sealing surface; Calculate the degree of overlap between the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram; if the degree of overlap is greater than a preset degree of overlap, calibrate the magnet after glue dispensing as a qualified product with good appearance characteristics; If the overlap is not greater than the preset overlap, the magnet after glue dispensing is marked as a defective product in terms of appearance characteristics, and a virtual space is constructed, and the actual glue morphology three-dimensional model diagram and the standard glue morphology three-dimensional model diagram are imported into the virtual space; The actual glue morphology 3D model image and the standard glue morphology 3D model image are registered in a virtual space to obtain an intersection 3D model image; the intersection 3D model image is analyzed to find out the non-overlapping area between the actual glue morphology 3D model image and the standard glue morphology 3D model image, which is defined as the glue dispensing abnormal area; Obtain a preset dispensing time node for dispensing the dispensing abnormal area, obtain the real-time dispensing parameters of the dispensing machine, and calculate the state transition probability of the dispensing machine at the preset dispensing time node based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine; If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node; If the overlap is greater than a preset overlap, ultrasonic data fed back by the glue on the magnet sealing surface is obtained; Performing feature extraction processing on the ultrasonic data to obtain feature data of bubbles inside the glue on the magnet sealing surface; If the bubble concentration value is not greater than the preset concentration value, the magnet after dispensing is calibrated as a qualified product with internal characteristics; If the bubble concentration value is greater than the preset concentration value, the magnet after dispensing is calibrated as an unqualified product with internal characteristics, and a three-dimensional model diagram of the actual characteristics of the glue in the magnet sealing surface is constructed based on the ultrasonic data; Constructing a second virtual space, importing the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image into the second virtual space for analysis, and defining the area where the actual feature three-dimensional model image and the standard glue morphology three-dimensional model image do not overlap as the abnormal glue dispensing area; According to the preset dispensing path, a preset dispensing time node for dispensing the abnormal dispensing area is obtained, and the real-time dispensing parameters of the dispensing machine are obtained. Based on the Markov chain and combined with the real-time dispensing parameters of the dispensing machine, the state transition probability of the dispensing machine at the preset dispensing time node is calculated; If the state transition probability is greater than the preset state transition probability, the preset dispensing time node is defined as the equipment abnormality time node.

Images