Glass cup laser engraving device

Abstract

The application relates to the technical field of glass product engraving, in particular to a glass cup laser engraving device. The laser engraving device comprises a workbench, a positioning base is arranged on the workbench, and a positioning groove is arranged on the positioning base; a support is installed on the workbench, a driving motor is installed on one side of the support, and a first bearing seat is installed on the other side of the support; the output end of the driving motor is connected with a driving shaft in a matched mode, the other end of the driving shaft is connected with the first bearing seat in a matched mode, and a first driving pulley and a second driving pulley are installed on the driving shaft; by adjusting the motion parameters of the five-axis linkage mechanical arm and the setting of the positioning and clamping system, various specifications of glass cups can be effectively engraved, and the universality and production efficiency of the device are improved.

Description

Technical Field

[0001] The invention relates to the technical field of glass product engraving, in particular to a glass cup laser engraving device. Background Art

[0002] With the continuous development of laser technology, laser engraving has gradually been applied to the decorative processing of glass cups. Laser engraving uses a high-energy-density laser beam to illuminate the glass surface, causing localized melting, vaporization, or chemical changes in the material, thereby forming a preset pattern or text. Compared with traditional processes, laser engraving offers significant advantages such as non-contact processing, high precision, high speed, flexible and changeable patterns, and environmental friendliness. It can achieve the fine engraving of complex patterns without causing mechanical stress damage to the glass, thereby improving the product yield rate.

[0003] However, existing laser engraving equipment for glass cups still has some shortcomings in practical applications. First, for glass cups of different shapes and sizes, existing equipment often requires the replacement of fixtures or complex parameter adjustments, which is cumbersome to operate and affects production efficiency. Secondly, the clamping method used by some equipment may not be stable enough, which can easily cause the glass cup to shift or vibrate during the engraving process, affecting the engraving accuracy. In addition, for glass cups with complex curved surfaces, existing engraving equipment has difficulty in achieving all-round, high-precision engraving, which limits its scope of application. Finally, equipment that lacks real-time monitoring and feedback control systems cannot dynamically adjust parameters according to the actual situation during the engraving process, making it difficult to ensure the stability and consistency of the engraving quality. Summary of the Invention

[0004] The present invention overcomes the shortcomings of the prior art and provides a glass laser engraving device.

[0005] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0006] The present invention discloses a laser engraving device for a glass cup, comprising a workbench, a positioning base provided on the workbench, and a positioning groove provided on the positioning base; a bracket installed on the workbench, a drive motor installed on one side of the bracket, and a first bearing seat installed on the other side; an output end of the drive motor is cooperatively connected to a drive shaft, the other end of the drive shaft is cooperatively connected to the first bearing seat, and a first driving pulley and a second driving pulley are installed on the drive shaft;

[0007] The workbench is also symmetrically provided with a first bidirectional threaded slide rod and a second bidirectional threaded slide rod, and a first slider and a second slider are slidably connected between the first bidirectional threaded slide rod and the second bidirectional threaded slide rod; a first clamping block is fixedly installed on the first slider, and a second clamping block is fixedly installed on the second slider;

[0008] A first driven pulley is provided at one end of the first bidirectional threaded slide rod, and a second driven pulley is provided at one end of the second bidirectional threaded slide rod; the first driving pulley and the first driven pulley are connected via a first belt transmission, and the second driving pulley and the second driven pulley are connected via a second belt transmission;

[0009] The workbench is also provided with a five-axis linkage mechanical arm, on which a laser engraving head is installed.

[0010] Furthermore, the cross-sectional shape of the positioning groove is the same as the cross-sectional shape of the bottom of the glass to be engraved.

[0011] Furthermore, the first bidirectional threaded slide rod is rotatably mounted on the workbench through a group of second bearing seats; the second bidirectional threaded slide rod is rotatably mounted on the workbench through a group of third bearing seats.

[0012] Furthermore, the first clamping block and the second clamping block are both provided with pressure sensors, and the pressure sensors are communicatively connected to the driving motor.

[0013] The present invention also discloses a control method for a glass laser engraving device, which is applied to any one of the glass laser engraving devices described above. The control method comprises the following steps:

[0014] Acquiring etching feature data of the laser engraving device under various preset engraving parameters, and constructing a dictionary tree search map based on the etching feature data of the laser engraving device under various preset engraving parameters;

[0015] Acquire real-time engraving parameters of a laser engraving device at a preset time node, and acquire predicted etching feature data of the glass at the current preset time node based on the real-time engraving parameters and in combination with a dictionary tree search graph;

[0016] The laser engraving equipment is adjusted and optimized based on the predicted etching feature data of the glass at the current preset time node.

[0017] Furthermore, etching feature data of the laser engraving device under various preset engraving parameters is obtained, and a dictionary tree search map is constructed based on the etching feature data of the laser engraving device under various preset engraving parameters, specifically:

[0018] Testing the laser engraving device to obtain etching feature data of the laser engraving device under various preset engraving parameters;

[0019] Constructing a dictionary tree, converting various preset engraving parameters into branch nodes of the dictionary tree, and converting various etching feature data into leaf nodes of the dictionary tree;

[0020] According to the etching feature data of the laser engraving device obtained by the test under various preset engraving parameters, a directed description relationship between various preset engraving parameters and various etching feature data is established;

[0021] According to the directed description relationship between various preset engraving parameters and various etching feature data, each branch node and each leaf node are feature-combined to form a dictionary tree search map of various preset engraving parameters and various etching feature data;

[0022] Among them, the preset engraving parameters include the laser power, moving speed and moving direction of the laser engraving head; the etching feature data includes the etching depth, etching width and etching shape corresponding to the corresponding position node of the glass cup after the laser engraving head acts on the corresponding preset engraving parameters.

[0023] Furthermore, the real-time engraving parameters of the laser engraving device are obtained at a preset time node, and the predicted etching feature data of the glass at the current preset time node is obtained based on the real-time engraving parameters and combined with the dictionary tree search graph, specifically:

[0024] Acquire real-time engraving parameters of a laser engraving device at a preset time node, and import the real-time engraving parameters into the dictionary tree search graph;

[0025] Converting the real-time engraving parameters into the root node of the dictionary tree, introducing a cosine similarity algorithm, and calculating the cosine similarity between the root node and each branch node in the dictionary tree search graph based on the cosine similarity algorithm;

[0026] Constructing a size sorting table, importing the cosine similarities between the root node and each branch node in the dictionary tree search graph into the size sorting table for size sorting, and sorting to obtain the highest cosine similarity;

[0027] The branch node corresponding to the highest cosine similarity is calibrated, the leaf node belonging to the calibrated branch node is extracted, and the feature information of the extracted leaf node is interpreted to obtain the etching feature data corresponding to the extracted leaf node;

[0028] The etching feature data corresponding to the extracted leaf node is output as the predicted etching feature data of the glass at the preset time node.

[0029] Furthermore, the laser engraving equipment is adjusted and optimized based on the predicted etching feature data of the glass at the current preset time node, specifically:

[0030] Obtaining a preset engraving process plan for the glass, and obtaining preset engraving feature data range information of the glass at a current preset time node according to the preset engraving process plan;

[0031] determining the maximum preset engraving feature data and the minimum preset engraving feature data of the glass cup at the current preset time node according to the preset engraving feature data range information;

[0032] taking the maximum preset engraving feature data of the glass cup at the current preset time node as the upper boundary condition and taking the minimum preset engraving feature data of the glass cup at the current preset time node as the lower boundary condition;

[0033] setting a definition domain according to the upper boundary condition and the lower boundary condition, and judging whether the predicted etching feature data of the glass cup at the current preset time node is completely within the definition domain;

[0034] if the predicted etching feature data of the glass cup at the current preset time node is completely within the definition domain, the real-time engraving parameters of the laser engraving equipment are not controlled and the real-time engraving parameters of the laser engraving equipment are continuously monitored at the next preset time node;

[0035] if the predicted etching feature data of the glass cup at the current preset time node is partially or completely outside the definition domain, a particle swarm optimization algorithm is introduced, the real-time engraving parameters of the laser engraving equipment are iteratively optimized based on the particle swarm optimization algorithm, the real-time engraving parameters of the laser engraving equipment at the next preset time node are adjusted again, and the predicted etching feature data at the next time node is made to fall within the definition domain again.

[0036] Further, if the predicted etching feature data of the glass cup at the current preset time node is partially or completely outside the definition domain, a particle swarm optimization algorithm is introduced, the real-time engraving parameters of the laser engraving equipment are iteratively optimized based on the particle swarm optimization algorithm, the real-time engraving parameters of the laser engraving equipment at the next preset time node are adjusted again, and the predicted etching feature data at the next time node is made to fall within the definition domain again, specifically:

[0037] determining the particle population size in the particle swarm optimization algorithm, each particle corresponding to a set of real-time engraving parameters of the laser engraving equipment, including the laser power, the moving speed and the moving direction of the laser engraving head;

[0038] initializing the position and velocity vectors of the particles, wherein the position vector represents a specific combination of engraving parameters and the velocity vector represents the adjustment direction and amplitude of the parameters;

[0039] constructing a fitness function according to the adaptability of the predicted etching feature data and the definition domain, calculating the fitness value of each particle, and the fitness value reflects the effectiveness of the engraving parameter combination represented by the particle in making the predicted etching feature data fall within the definition domain;

[0040] For each particle, its current fitness value is compared with the individual historical optimal fitness value. If it is better, the individual optimal position is updated; at the same time, the global optimal position is updated by comparing with the entire particle group;

[0041] Based on the particle's current position, velocity, individual optimal position, and global optimal position, continue to update the particle's velocity and position vector to generate a new engraving parameter combination;

[0042] Continue to iterate the above steps until the predicted etching feature data completely falls within the definition domain, and set the engraving parameters corresponding to the global optimal position as the real-time engraving parameters of the next preset time node, thereby ensuring that the predicted etching feature data at the next time node falls within the definition domain again.

[0043] This invention addresses the technical deficiencies of the prior art and has the following beneficial effects: First, by adjusting the motion parameters of the five-axis linkage robot arm and the settings of the positioning and clamping system, it can effectively engrave glass cups of various specifications, improving the versatility and production efficiency of the equipment. Second, this control method constructs a dictionary tree search map by pre-acquiring etching feature data under different engraving parameters, and then predicts the etching feature data based on real-time engraving parameters, thereby achieving timely adjustment and optimization of the laser engraving equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, without paying any creative work, they can also obtain drawings of other embodiments based on these drawings.

[0045] Figure 1 is a schematic diagram of the first three-dimensional structure of the device;

[0046] Figure 2 for Figure 1 Enlarged schematic diagram of AA in the middle;

[0047] Figure 3 is a second three-dimensional structural schematic diagram of the device;

[0048] Figure 4 It is a side view structural diagram of the device;

[0049] The reference signs are explained as follows: 101, workbench; 102, positioning base; 103, positioning groove; 104, support; 105, driving motor; 106, first bearing seat; 107, driving shaft; 108, first driving pulley; 109, second driving pulley; 201, first bidirectional threaded slide rod; 202, second bidirectional threaded slide rod; 203, first sliding block; 204, second sliding block; 205, first clamping block; 206, second clamping block; 207, first driven pulley; 208, second driven pulley; 209, first belt; 301, second belt; 302, five-axis linkage mechanical arm; 303, laser engraving head; 304, second bearing seat; 305, third bearing seat. DETAILED DESCRIPTION

[0050] In order to facilitate the understanding of the present application, the present application will be described more fully below with reference to the accompanying drawings. The preferred embodiments of the present application are shown in the drawings. However, the present application can be realized in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided so that the disclosure of the present application can be more thoroughly and completely understood.

[0051] The present application discloses a kind of glass cup laser engraving equipment, as Figure 1 As shown, the laser engraving equipment includes workbench 101, the positioning base 102 is arranged on the workbench 101, and the positioning groove 103 is arranged on the positioning base 102;The support 104 is installed on the workbench 101, the driving motor 105 is installed on one side of the support 104, and the first bearing seat 106 is installed on the other side;The output end of the driving motor 105 is connected with the driving shaft 107, the other end of the driving shaft 107 is connected with the first bearing seat 106, and the first driving pulley 108 and the second driving pulley 109 are installed on the driving shaft 107;

[0052] As shown in Figure 2 , Figure 3 , Figure 4 As shown, the workbench 101 is also provided with the first bidirectional threaded slide rod 201 and the second bidirectional threaded slide rod 202 symmetrically, and the first sliding block 203 and the second sliding block 204 are slidably connected between the first bidirectional threaded slide rod 201 and the second bidirectional threaded slide rod 202;The first clamping block 205 is fixedly installed on the first sliding block 203, and the second clamping block 206 is fixedly installed on the second sliding block 204;

[0053] A first driven pulley 207 is provided at one end of the first bidirectional threaded slide 201, and a second driven pulley 208 is provided at one end of the second bidirectional threaded slide 202; the first driving pulley 108 and the first driven pulley 207 are connected to each other via a first belt 209, and the second driving pulley 109 and the second driven pulley 208 are connected to each other via a second belt 301;

[0054] A five-axis linkage robot arm 302 is also provided on the workbench 101 , and a laser engraving head 303 is installed on the five-axis linkage robot arm 302 .

[0055] The cross-sectional shape of the positioning groove 103 is the same as the cross-sectional shape of the bottom of the glass to be engraved.

[0056] The first bidirectional threaded slide 201 is rotatably mounted on the workbench 101 via a set of second bearing seats 304 ; the second bidirectional threaded slide 202 is rotatably mounted on the workbench 101 via a set of third bearing seats 305 .

[0057] The first clamping block 205 and the second clamping block 206 are both provided with pressure sensors, and the pressure sensors are in communication connection with the driving motor 105 .

[0058] The positioning base 102 is detachable, and users can install positioning bases of different shapes and specifications according to specific requirements, such as positioning bases with positioning grooves 103 of different diameters or various shapes. The shapes of positioning grooves include but are not limited to square, circular, triangular, and elliptical.

[0059] It should be noted that the glass to be engraved is placed in the positioning groove 103 of the positioning base 102. Since the cross-sectional shape of the positioning groove 103 is the same as that of the bottom of the glass, the preliminary positioning of the glass on the horizontal plane is accurate, preventing unnecessary translation in subsequent operations. The driving motor 105 is started, and the driving motor 105 drives the driving shaft 107 to rotate. The first driving pulley 108 and the second driving pulley 109 on the driving shaft 107 rotate synchronously with the driving shaft 107. The first driving pulley 108 drives the first driven pulley 207 at one end of the first bidirectional threaded slide rod 201 to rotate through the first belt 209, and the second driving pulley 109 drives the second driven pulley 208 at one end of the second bidirectional threaded slide rod 202 to rotate through the second belt 301. Since the first bidirectional threaded slide rod 201 and the second bidirectional threaded slide rod 202 are bidirectional threaded structures, as they rotate, the first slide block 203 and the second slide block 204 move towards each other on the slide rod. The first clamping block 205 on the first slide block 203 and the second clamping block 206 on the second slide block 204 also move towards each other, clamping the glass in the positioning groove 103. During clamping, the pressure sensors on the first clamping block 205 and the second clamping block 206 monitor the clamping pressure in real time. When the pressure reaches the set value, the pressure sensor transmits a signal to the driving motor 105, and the driving motor 105 stops rotating, thereby ensuring that the clamping force can firmly fix the glass, and the glass is not damaged due to excessive pressure.

[0060] The laser engraving head 303 is moved to the initial engraving position above the glass by the five-axis linkage mechanical arm 302. The five-axis linkage can realize the precise positioning of the laser engraving head 303 in space to adapt to the complex curved surface engraving requirements of the glass. The laser engraving head 303 emits a laser beam to engrave the glass according to the set pattern, text or texture, etc. During the engraving process, the five-axis linkage mechanical arm 302 adjusts the position, angle and other parameters of the laser engraving head 303 in real time according to the requirements of the engraving path, to ensure the accuracy and quality of the engraving.

[0061] After the engraving is completed, the driving motor 105 is reversed. The first driving pulley 108 and the second driving pulley 109 are reversed, and the first bidirectional threaded slide rod 201 and the second bidirectional threaded slide rod 202 are reversed through the belt, so that the first slide block 203 and the second slide block 204 move away from each other, and the first clamping block 205 and the second clamping block 206 loosen the clamping of the glass. Then the completed glass can be taken out of the positioning groove 103 by an automatic unloading mechanical arm or manually, thereby completing an engraving process.

[0062] Overall, the positioning groove 103, designed to match the shape of the glass's bottom, ensures precise initial positioning of the glass on a horizontal plane, providing a stable foundation for subsequent clamping and engraving operations. The bidirectional threaded slide rod structure, combined with the pressure sensor-based clamping system, ensures a secure clamping of the glass. The bidirectional threaded slide rod enables the first and second clamping blocks 205 and 206 to move toward each other, ensuring even clamping of the glass from both sides. The pressure sensor precisely controls the clamping force, preventing damage to the glass due to excessive clamping force, and improving the device's adaptability to glasses of varying materials and thicknesses. The use of a five-axis robotic arm 302 enables the laser engraving head 303 to perform complex movements in space, precisely engraving the glass according to design requirements. High-precision engraving is achieved on both curved and flat surfaces, meeting the requirements for engraving complex patterns and fine textures. By adjusting the motion parameters of the five-axis robotic arm 302 and the settings of the positioning and clamping system, effective engraving operations can be performed on glasses of various sizes, improving the device's versatility and production efficiency.

[0063] The present invention also discloses a control method for a glass laser engraving device, which is applied to any one of the glass laser engraving devices described above. The control method comprises the following steps:

[0064] Acquiring etching feature data of the laser engraving device under various preset engraving parameters, and constructing a dictionary tree search map based on the etching feature data of the laser engraving device under various preset engraving parameters;

[0065] Acquire real-time engraving parameters of a laser engraving device at a preset time node, and acquire predicted etching feature data of the glass at the current preset time node based on the real-time engraving parameters and in combination with a dictionary tree search graph;

[0066] The laser engraving equipment is adjusted and optimized based on the predicted etching feature data of the glass at the current preset time node.

[0067] It should be noted that etching feature data of the laser engraving device under various preset engraving parameters is obtained. A dictionary tree search graph is then constructed based on this data. A dictionary tree search graph is a data structure that effectively organizes and stores the relationship between these etching feature data and engraving parameters. In this graph, different engraving parameter combinations and their corresponding etching feature data are stored in a tree structure, facilitating subsequent query and analysis. At a preset time point, the real-time engraving parameters of the laser engraving device are obtained. These real-time engraving parameters reflect the device's current operating status, such as actual operating parameters such as laser power and engraving speed. Next, based on these real-time engraving parameters and the previously constructed dictionary tree search graph, predicted etching feature data for the glass at the current preset time point is obtained. This utilizes the previously stored relationship between engraving parameters and etching feature data. Using the known real-time engraving parameters, the dictionary tree search graph is used to search for the corresponding etching feature data, thereby predicting the current engraving effect. Finally, the laser engraving device is adjusted and optimized based on the predicted etching feature data. If the predicted etching feature data shows that unsatisfactory engraving results may occur, such as etching depth that is too deep or too shallow, or unclear etching patterns, etc., the engraving parameters of the equipment can be adjusted in a timely manner, such as adjusting the laser power, changing the engraving speed, etc., to avoid a large range of poor engraving effects. This control method can construct a dictionary tree search map by pre-acquiring etching feature data under different engraving parameters, and then predict the etching feature data in combination with real-time engraving parameters, thereby achieving timely adjustment and optimization of the laser engraving equipment. It can effectively improve the quality of laser engraving of glass cups and avoid poor etching effects caused by improper engraving parameter settings, such as uneven etching depth, unclear patterns, etc. At the same time, this control method based on data structure and prediction can improve the degree of automation of the equipment, reduce manual intervention, improve production efficiency, and can adapt to different engraving requirements and the characteristics of glass cups, enhancing the versatility and flexibility of the equipment.

[0068] Furthermore, etching feature data of the laser engraving device under various preset engraving parameters is obtained, and a dictionary tree search map is constructed based on the etching feature data of the laser engraving device under various preset engraving parameters, specifically:

[0069] Testing the laser engraving device to obtain etching feature data of the laser engraving device under various preset engraving parameters;

[0070] Constructing a dictionary tree, converting various preset engraving parameters into branch nodes of the dictionary tree, and converting various etching feature data into leaf nodes of the dictionary tree;

[0071] According to the etching feature data of the laser engraving device obtained by the test under various preset engraving parameters, a directed description relationship between various preset engraving parameters and various etching feature data is established;

[0072] According to the directed description relationship between various preset engraving parameters and various etching feature data, each branch node and each leaf node are feature-combined to form a dictionary tree search map of various preset engraving parameters and various etching feature data;

[0073] Among them, the preset engraving parameters include the laser power, moving speed and moving direction of the laser engraving head; the etching feature data includes the etching depth, etching width and etching shape corresponding to the corresponding position node of the glass cup after the laser engraving head acts on the corresponding preset engraving parameters.

[0074] It should be noted that the laser engraving equipment must first be tested to obtain etching feature data under different preset engraving parameters. The preset engraving parameters here specifically include the laser power, movement speed, and movement direction of the laser engraving head. These parameters are key factors affecting the engraving effect. Different power, speed, and direction will result in different effects between the laser and the glass surface. During the testing process, these preset engraving parameters were varied, and the etching characteristics of the glass surface were recorded for each parameter combination. This includes etching depth, width, and shape. Etching depth reflects the extent of glass material removed by the laser, etching width reflects the lateral extent of the engraving mark, and etching shape represents the outline of the final pattern on the glass surface. When constructing the dictionary tree, the various preset engraving parameters are converted into branch nodes of the dictionary tree, and the various etching feature data are converted into leaf nodes of the dictionary tree. This structure is similar to a tree, with branch nodes representing the branches leading to leaf nodes. Just like in a decision tree, different decision conditions (here, preset engraving parameters) lead to different outcomes (here, etching feature data). Based on the test data, a directed descriptive relationship is constructed between the various preset engraving parameters and the various etching feature data. This means that it is clear how different engraving parameters affect the final etching features, which is a description of cause and effect. For example, a specific laser power, moving speed and direction (preset engraving parameters) will result in a specific etching depth, width and shape (etching feature data). Finally, based on this directed description relationship, each branch node and each leaf node are feature combined to form a dictionary tree search map of various preset engraving parameters and various etching feature data. This map fully presents the mapping relationship between the preset engraving parameters and the etching feature data, just like a lookup table. The etching feature data that may be generated can be quickly located through given engraving parameters.

[0075] The dictionary tree search graph constructed through the above series of operations clearly displays the relationship between the laser engraving equipment's preset engraving parameters and the etching feature data. This allows for rapid prediction of the expected etching effect based on the current engraving parameters during the actual engraving process, providing a basis for optimizing the engraving process. This relationship graph helps improve the accuracy and controllability of laser engraving, enabling better adjustment of engraving parameters to achieve the ideal etching depth, width, and shape. It also provides effective data support for equipment automation control and process optimization.

[0076] Furthermore, the real-time engraving parameters of the laser engraving device are obtained at a preset time node, and the predicted etching feature data of the glass at the current preset time node is obtained based on the real-time engraving parameters and combined with the dictionary tree search graph, specifically:

[0077] Acquire real-time engraving parameters of a laser engraving device at a preset time node, and import the real-time engraving parameters into the dictionary tree search graph;

[0078] Converting the real-time engraving parameters into the root node of the dictionary tree, introducing a cosine similarity algorithm, and calculating the cosine similarity between the root node and each branch node in the dictionary tree search graph based on the cosine similarity algorithm;

[0079] Constructing a size sorting table, importing the cosine similarities between the root node and each branch node in the dictionary tree search graph into the size sorting table for size sorting, and sorting to obtain the highest cosine similarity;

[0080] The branch node corresponding to the highest cosine similarity is calibrated, the leaf node belonging to the calibrated branch node is extracted, and the feature information of the extracted leaf node is interpreted to obtain the etching feature data corresponding to the extracted leaf node;

[0081] The etching feature data corresponding to the extracted leaf node is output as the predicted etching feature data of the glass at the preset time node.

[0082] It should be noted that the real-time engraving parameters of the laser engraving device are acquired at preset time points. These real-time engraving parameters reflect the device's current operating status, such as actual operating parameter values ​​such as laser power, engraving speed, and movement direction. These real-time engraving parameters are then imported into a previously constructed dictionary tree search graph. This dictionary tree search graph contains the relationships between various preset engraving parameters and etching feature data, and serves as an important basis for predicting etching feature data. After converting the real-time engraving parameters into the root node of the dictionary tree, the cosine similarity algorithm is introduced. The cosine similarity algorithm is used to measure the similarity between two vectors. Here, the root node (representing the real-time engraving parameter) and each branch node in the dictionary tree search graph (representing different preset engraving parameters) are treated as vectors. The cosine similarity between these two nodes is calculated to determine the degree of similarity between the real-time engraving parameters and the preset engraving parameters. A size ranking table is constructed, and the calculated cosine similarity values ​​between the root node and each branch node are imported into it for ranking. After sorting, the highest cosine similarity value is obtained. The branch node corresponding to this highest cosine similarity value represents the preset engraving parameter combination that is most similar to the current real-time engraving parameters. The branch node corresponding to the highest cosine similarity is calibrated, and the leaf nodes belonging to the calibrated branch node are extracted. These leaf nodes contain etching feature data related to the branch node (preset engraving parameters). The feature information in the extracted leaf nodes is interpreted, that is, the etching feature data stored in the leaf nodes is converted into an understandable and usable form. Finally, the corresponding etching feature data in the interpreted leaf nodes is output as the predicted etching feature data of the glass at the preset time point. This predicted etching feature data reflects the possible etching effects of the glass, such as etching depth, etching width, and etching shape, under the current real-time engraving parameters. Based on the real-time engraving parameters of the laser engraving equipment, the dictionary tree search graph and the cosine similarity algorithm can accurately predict the etching feature data of the glass at the current moment. This helps to understand the possible etching effects in advance during the engraving process, allowing the engraving parameters to be adjusted in time to achieve the desired engraving effect. This improves the accuracy and controllability of laser engraving, reduces engraving failures or poor quality caused by improper parameter settings, and provides strong technical support for the efficient and precise operation of laser engraving equipment.

[0083] Furthermore, the laser engraving equipment is adjusted and optimized based on the predicted etching feature data of the glass at the current preset time node, specifically:

[0084] Obtaining a preset engraving process plan for the glass, and obtaining preset engraving feature data range information of the glass at a current preset time node according to the preset engraving process plan;

[0085] Determining the maximum preset engraving feature data and the minimum preset engraving feature data of the glass at the current preset time node according to the preset engraving feature data range information;

[0086] The maximum preset engraving feature data of the glass at the current preset time node is used as the upper boundary condition, and the minimum preset engraving feature data of the glass at the current preset time node is used as the lower boundary condition;

[0087] Setting a definition domain according to the upper boundary condition and the lower boundary condition, and determining whether the predicted etching feature data of the glass at the current preset time node completely falls within the definition domain;

[0088] If the predicted etching feature data of the glass at the current preset time node completely falls within the definition domain, the real-time engraving parameters of the laser engraving device are not adjusted, and the real-time engraving parameters of the laser engraving device are continuously monitored at the next preset time node;

[0089] If the predicted etching feature data of the glass cup at the current preset time node partially or completely falls outside the definition domain, the particle swarm optimization algorithm is introduced to iteratively optimize the real-time engraving parameters of the laser engraving device based on the particle swarm optimization algorithm, and the real-time engraving parameters of the laser engraving device at the next preset time node are readjusted so that the predicted etching feature data at the next time node falls within the definition domain again.

[0090] It should be noted that the preset engraving process plan for the glass is first obtained. This plan specifies the ideal engraving effect for the glass at each preset time point, that is, the preset engraving feature data range information. For example, there may be an ideal range for etching depth, and corresponding required ranges for etching width and shape. Based on this range information, the maximum and minimum preset engraving feature data at the current preset time point can be determined. These two data represent the upper and lower limits allowed by the process plan, respectively. The maximum preset engraving feature data is used as the upper boundary condition, and the minimum preset engraving feature data is used as the lower boundary condition, thus establishing a definition domain. This definition domain is the range of values ​​of the predicted etching feature data that meets the preset engraving process plan at the current preset time point. It is then determined whether the predicted etching feature data for the glass at the current preset time point completely falls within this definition domain. If the predicted etching feature data completely falls within the definition domain, it indicates that the current real-time engraving parameter settings of the laser engraving device are reasonable and meet the requirements of the preset engraving process plan. Therefore, there is no need to adjust the real-time engraving parameters of the laser engraving device; only the real-time engraving parameters need to be monitored at the next preset time point to ensure that the engraving process continues to meet the requirements. If the predicted etching feature data falls partially or completely outside the domain of definition, this means that the current engraving parameter settings may result in an engraving result that does not conform to the preset process plan. This is when the particle swarm optimization algorithm, an optimization algorithm based on swarm intelligence, is introduced. The particle swarm optimization algorithm is used to iteratively optimize the real-time engraving parameters of the laser engraving device. During the iterative process, the predicted etching feature data is recalculated by continuously adjusting the real-time engraving parameters of the laser engraving device (such as laser power, engraving speed, movement direction, etc.) until the predicted etching feature data at the next preset time node falls back into the domain of definition, thereby returning the engraving process to the track that conforms to the preset process plan.

[0091] Through the above operations, this adjustment and optimization process effectively monitors and adjusts the laser engraving equipment's real-time engraving parameters according to the preset engraving process plan. While ensuring that the engraving process meets process requirements, it reduces unnecessary parameter adjustments and improves the stability and efficiency of the equipment. When the predicted etching feature data does not meet the requirements, the particle swarm optimization algorithm can intelligently adjust the engraving parameters to ensure that the engraving result is as close to the preset process plan as possible, thereby improving the quality and yield of the glass laser engraving and ensuring the accuracy and controllability of the entire engraving process.

[0092] Furthermore, if the predicted etching feature data of the glass at the current preset time node partially or completely falls outside the domain of definition, a particle swarm optimization algorithm is introduced to iteratively optimize the real-time engraving parameters of the laser engraving device based on the particle swarm optimization algorithm, and the real-time engraving parameters of the laser engraving device at the next preset time node are readjusted so that the predicted etching feature data at the next time node falls within the domain of definition again. Specifically,

[0093] Determine the particle population size in the particle swarm optimization algorithm, where each particle corresponds to a set of real-time engraving parameters of the laser engraving device, including the laser power, moving speed, and moving direction of the laser engraving head;

[0094] Initialize the particle's position and velocity vectors, where the position vector represents a specific engraving parameter combination, and the velocity vector represents the parameter adjustment direction and amplitude;

[0095] A fitness function is constructed based on the adaptability of the predicted etching feature data and the definition domain, and the fitness value of each particle is calculated. The fitness value reflects the effectiveness of the engraving parameter combination represented by the particle in making the predicted etching feature data fall into the definition domain.

[0096] For each particle, its current fitness value is compared with the individual historical optimal fitness value. If it is better, the individual optimal position is updated; at the same time, the global optimal position is updated by comparing with the entire particle group;

[0097] Based on the particle's current position, velocity, individual optimal position, and global optimal position, continue to update the particle's velocity and position vector to generate a new engraving parameter combination;

[0098] Continue to iterate the above steps until the predicted etching feature data completely falls within the definition domain, and set the engraving parameters corresponding to the global optimal position as the real-time engraving parameters of the next preset time node, thereby ensuring that the predicted etching feature data at the next time node falls within the definition domain again.

[0099] It's important to note that the particle swarm optimization algorithm first determines the size of the particle population. Each particle here corresponds to a set of real-time engraving parameters for the laser engraving device. These parameters include the laser power, movement speed, and movement direction of the laser engraving head, factors that significantly influence the engraving effect. This means that multiple particles are used to explore the space of different engraving parameter combinations to find the optimal combination that ensures the predicted etching feature data falls within the domain of definition. The particle's position and velocity vectors are initialized. The position vector represents a specific engraving parameter combination and determines each particle's initial position in the parameter space, specifically, the initial set of laser power, movement speed, and movement direction values. The velocity vector represents the direction and magnitude of parameter adjustment and determines how the particle moves within the parameter space at each iteration—that is, how the engraving parameter values ​​are changed. A fitness function is constructed based on the compatibility of the predicted etching feature data with the domain of definition. This fitness function is key to evaluating the quality of particles, reflecting the effectiveness of the engraving parameter combination represented by the particle in ensuring that the predicted etching feature data falls within the domain of definition. By calculating the fitness value of each particle, the degree to which the engraving parameter combination distance corresponding to each particle meets the requirements (making the predicted etching feature data fall into the definition domain) can be measured.

[0100] For each particle, its current fitness value is compared with the individual historical optimal fitness value. If the current fitness value is better, the individual optimal position is updated, which means that a better engraving parameter combination for the particle has been found. At the same time, the global optimal position is updated by comparing in the entire particle swarm. This global optimal position represents the most effective engraving parameter combination currently found in the entire particle swarm to make the predicted etching feature data fall into the definition domain. Based on the particle's current position, speed, individual optimal position and global optimal position, the particle's speed and position vectors continue to be updated. This step is the core operation of the particle swarm optimization algorithm. Through the comprehensive use of this information, particles can move in the direction where it is more likely to find the optimal solution (making the predicted etching feature data fall into the definition domain), thereby generating a new engraving parameter combination.

[0101] The above steps are iterated continuously, constantly adjusting the particle position and velocity, and exploring different combinations of engraving parameters until the predicted etching feature data completely falls within the domain of definition. Once this condition is met, the engraving parameters corresponding to the global optimal position are set as the real-time engraving parameters for the next preset time node, ensuring that the predicted etching feature data at the next time node falls within the domain of definition again.

[0102] This iterative optimization process, based on a particle swarm optimization algorithm, intelligently adjusts the laser engraving equipment's real-time engraving parameters. When the predicted etching feature data falls partially or completely outside the domain of definition, multiple particles explore and optimize the engraving parameter combination space to find the parameter combination that best brings the predicted etching feature data back into the domain of definition. This improves the laser engraving equipment's adaptability to the engraving process requirements, effectively avoiding undesirable engraving results due to improper parameter settings. This improves the quality and precision of glass laser engraving and ensures that the engraving process proceeds according to the preset process plan.

[0103] In addition, in practical applications, the control method may further include the following steps:

[0104] Obtaining a working calendar of the laser engraving device, obtaining a fault operating condition event of the laser engraving device according to the working calendar, obtaining fault characteristic data of the laser engraving device when the fault operating condition event occurs, and obtaining normal characteristic data of the laser engraving device during a preset time period before the fault operating condition event occurs;

[0105] A fuzzy evaluation algorithm is introduced, and based on the fuzzy evaluation algorithm, fuzzy evaluation is performed on the fault characteristic data of the laser engraving device when the fault condition event occurs and the normal characteristic data obtained during a preset time period before the fault condition event occurs, so as to obtain the fuzziness of the fault transfer characteristics of the laser engraving device when various fault condition events occur;

[0106] Constructing a fault transfer feature fuzziness matrix based on the fault transfer feature fuzziness of various fault conditions of the laser engraving device; constructing a Bayesian network model, and embedding the fault transfer feature fuzziness matrix into the Bayesian network model for encoding learning;

[0107] Acquiring real-time operating parameters of the laser engraving device at a preset time node, importing the real-time operating parameters into the Bayesian network model for evaluation, and acquiring the fuzziness of the fault transfer characteristics of the laser engraving device at the current preset time node;

[0108] If the fuzziness of the fault transfer feature of the laser engraving device at the current preset time node is greater than a preset threshold, the laser engraving device is controlled to shut down and a fault warning message is generated;

[0109] If the fuzziness of the fault transfer feature of the laser engraving device at the current preset time node is not greater than a preset threshold, further obtaining the fuzziness of the fault transfer feature of the laser engraving device at several future time nodes based on the Bayesian network model;

[0110] If the fuzziness of the fault transfer feature of at least one future time node of the laser engraving device is greater than a preset threshold, the laser engraving device is controlled to shut down and a fault warning message is generated.

[0111] It should be noted that the laser engraving equipment's operating calendar is first obtained to identify faulty operating conditions. For each faulty operating condition, fault signature data at the time of failure and normal signature data from a preset period before the failure are collected. This data forms the basis for subsequent equipment failure analysis. Fault signature data includes abnormalities in equipment operating parameters (such as laser power fluctuations and abnormal robotic arm motion), while normal signature data reflects the equipment's normal operating state. A fuzzy evaluation algorithm is introduced to process these fault and normal signature data. The fuzzy evaluation algorithm can handle uncertainty and ambiguity in the data. This algorithm determines the ambiguity of the fault transition characteristics for various faulty operating conditions of the laser engraving equipment. This ambiguity quantifies the fuzzy characteristics of the transition from normal to faulty state, reflecting the probability and tendency of the fault. A fault transition characteristic ambiguity matrix is ​​constructed based on the calculated fault transition characteristic ambiguities for various faulty operating conditions. This matrix systematically organizes the ambiguity information for different fault scenarios. A Bayesian network model is then constructed, embedding the fault transition characteristic ambiguity matrix for encoding learning. A Bayesian network model is a probabilistic graphical model that can represent probabilistic relationships between variables. Through this embedded learning approach, the model can learn the complex relationship between fault signature data, normal signature data, and the ambiguity of the fault transition signature. The real-time operating parameters of the laser engraving device at a preset time are obtained and evaluated within a Bayesian network model. This yields the device's fault transition signature ambiguity at the current preset time. This ambiguity is then compared with a preset threshold. If it exceeds the threshold, the device is currently at a high risk of failure. The device is then shut down and a fault warning message is generated to prevent possible damage or poor engraving results. If the fault transition signature ambiguity at the current preset time is not greater than the threshold, the Bayesian network model is used to further determine the device's fault transition signature ambiguity at several future time points. If the fault transition signature ambiguity at at least one future time point exceeds the threshold, the device is also shut down and a fault warning message is generated. This step enables the prediction of future device failures, allowing proactive measures to prevent them, reducing downtime and repair costs, and improving device reliability and production efficiency. Through these steps, this control method comprehensively analyzes the fault conditions of the laser engraving device. By collecting data on both faults and normal states and building a fault prediction system using fuzzy evaluation algorithms and Bayesian network models, we can not only accurately assess the current risk of equipment failure but also predict the likelihood of future failures. This helps to promptly identify potential equipment failures, allowing for proactive downtime and early warning measures to prevent the impact of equipment failures on production, such as interruptions in engraving and reduced product quality. This improves equipment stability, reliability, and production efficiency, while reducing maintenance costs.

[0112] In addition, in practical applications, the control method may further include the following steps:

[0113] When the laser engraving head completes engraving of a certain position node of the glass and moves to the next position node for engraving, the CCD camera acquires the actual engraving working condition image of the position node that has been engraved;

[0114] Performing matrix conversion processing on the actual engraving working condition image to obtain the actual engraving working condition gray level co-occurrence matrix of the certain position node; calculating the structural similarity index between the actual engraving working condition gray level co-occurrence matrix and the preset engraving working condition gray level co-occurrence matrix;

[0115] If the structural similarity index is not greater than a preset index value, the element values ​​of the gray-level co-occurrence matrix of the actual engraving working condition and the element values ​​of the same matrix position in the preset engraving working condition gray-level co-occurrence matrix are compared, and the matrix positions with different element values ​​are marked to obtain the defect position area of ​​the node at the certain position;

[0116] determining whether the defect location area of ​​the certain location node has extended to the non-engraving area; if the defect location area of ​​the certain location node has extended to the non-engraving area, segmenting the area image corresponding to the defect location area in the actual engraving working condition image based on a watershed algorithm;

[0117] Perform feature recognition on the regional image corresponding to the defect location area to obtain the defect type of the defect location area. If the defect type of the defect location area is a concave defect (such as a crack or a erosion hole), the laser engraving device is controlled to stop and the current engraved semi-finished product is scrapped; if the defect type of the defect location area is a convex defect (such as welding slag or a bulge), the laser engraving device is controlled to continue working and the current engraved semi-finished product is marked as a product to be repaired;

[0118] If the defective position area of ​​the certain position node does not extend to the non-engraving area, the laser engraving device is controlled to continue working, and the current engraved semi-finished product is marked as a product to be repaired.

[0119] It should be noted that when the laser engraving head completes the engraving of a position node on the glass and moves to the next position node, the CCD camera is used to obtain the actual engraving working condition image of the position node that has been engraved. The CCD camera can capture the actual surface condition of the position node after engraving, and provide original image data for subsequent quality inspection. The obtained actual engraving working condition image is subjected to matrix conversion processing to obtain the actual engraving working condition grayscale co-occurrence matrix. The grayscale co-occurrence matrix can reflect the spatial correlation characteristics of the pixel grayscale in the image. Then, the structural similarity index between the actual engraving working condition grayscale co-occurrence matrix and the preset engraving working condition grayscale co-occurrence matrix is ​​calculated. The preset engraving working condition grayscale co-occurrence matrix is ​​a pre-set matrix representing the ideal engraving effect. By comparing the structural similarity indexes of the two, the degree of difference between the actual engraving effect and the ideal effect can be preliminarily judged. If the structural similarity index is not greater than the preset index value, it indicates that there may be problems with the actual engraving effect. At this point, the gray-level co-occurrence matrix of the actual engraving condition is compared with the element values ​​of the preset gray-level co-occurrence matrix for the same matrix position. Matrix positions with different element values ​​are marked. These marked positions constitute the defect location area for that node. This step precisely identifies the areas where the actual engraving differs from the ideal state by comparing matrix elements. The next step is to determine whether the defect location area has extended into non-engraved areas. This determination is used to determine the severity and scope of the defect. If the defect extends into non-engraved areas, it will have a more serious impact on the quality and structural integrity of the entire engraving.

[0120] If the defect location area has extended to the non-engraving area, the regional image corresponding to the defect location area is segmented in the actual engraving working condition image based on the watershed algorithm. The watershed algorithm is an image segmentation algorithm that can divide the image according to different feature areas. The segmented regional image is then subjected to feature recognition to determine the defect type. If it is a concave defect (such as cracks, corrosion holes), this type of defect will seriously affect the strength and appearance of the glass, so the laser engraving equipment is controlled to stop and the current engraved semi-finished product is scrapped; if it is a convex defect (such as welding slag, bulge), although it also affects the appearance, it has less impact on the product structure than the concave defect, so the laser engraving equipment is controlled to continue working and the current engraved semi-finished product is marked as a product to be repaired.

[0121] If the defective location area of ​​the certain position node does not extend to the non-engraving area, it means that the impact range of the defect is relatively small. At this time, the laser engraving device is controlled to continue working, and the current engraved semi-finished product is marked as a product to be repaired.

[0122] The above series of operations enables real-time quality inspection of engraved nodes during the laser engraving process. By comparing actual engraving conditions with preset conditions, potential defect areas can be accurately identified. The handling of the semi-finished engraved product is determined based on whether the defect extends into non-engraved areas and the type of defect. This helps to promptly identify and address quality issues during the engraving process, reducing scrap rates and improving production efficiency. It also enables appropriate responses to defects of varying types and severity, ensuring the quality of laser engraved products and the stability of the production process.

[0123] The above description of the preferred embodiments of the present invention is provided as a guide, and while the description is relatively specific and detailed, it should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A glass laser engraving device, characterized in that: The laser engraving device includes a workbench, a positioning base is provided on the workbench, and a positioning groove is provided on the positioning base; a bracket is installed on the workbench, a driving motor is installed on one side of the bracket, and a first bearing seat is installed on the other side; the output end of the driving motor is cooperatively connected to the driving shaft, the other end of the driving shaft is cooperatively connected to the first bearing seat, and a first driving pulley and a second driving pulley are installed on the driving shaft; The workbench is also symmetrically provided with a first bidirectional threaded slide rod and a second bidirectional threaded slide rod, and a first slider and a second slider are slidably connected between the first bidirectional threaded slide rod and the second bidirectional threaded slide rod; a first clamping block is fixedly installed on the first slider, and a second clamping block is fixedly installed on the second slider; A first driven pulley is provided at one end of the first bidirectional threaded slide rod, and a second driven pulley is provided at one end of the second bidirectional threaded slide rod; the first driving pulley and the first driven pulley are connected via a first belt transmission, and the second driving pulley and the second driven pulley are connected via a second belt transmission; The workbench is also provided with a five-axis linkage robotic arm, on which a laser engraving head is installed; Also included is a control method for the glass laser engraving device: Acquiring etching feature data of the laser engraving device under various preset engraving parameters, and constructing a dictionary tree search map based on the etching feature data of the laser engraving device under various preset engraving parameters; Acquire real-time engraving parameters of a laser engraving device at a preset time node, and acquire predicted etching feature data of the glass at the current preset time node based on the real-time engraving parameters and in combination with a dictionary tree search graph; Adjust and optimize the laser engraving equipment based on the predicted etching feature data of the glass at the current preset time node; The etching feature data of the laser engraving device under various preset engraving parameters are obtained, and a dictionary tree search map is constructed based on the etching feature data of the laser engraving device under various preset engraving parameters, specifically: Testing the laser engraving device to obtain etching feature data of the laser engraving device under various preset engraving parameters; Constructing a dictionary tree, converting various preset engraving parameters into branch nodes of the dictionary tree, and converting various etching feature data into leaf nodes of the dictionary tree; According to the etching feature data of the laser engraving device obtained by the test under various preset engraving parameters, a directed description relationship between various preset engraving parameters and various etching feature data is established; According to the directed description relationship between various preset engraving parameters and various etching feature data, each branch node and each leaf node are feature-combined to form a dictionary tree search map of various preset engraving parameters and various etching feature data; Among them, the preset engraving parameters include the laser power, moving speed and moving direction of the laser engraving head; the etching feature data includes the etching depth, etching width and etching shape corresponding to the corresponding position node of the glass cup after the laser engraving head acts on the corresponding preset engraving parameters.

2. The glass laser engraving device according to claim 1, characterized in that: The cross-sectional shape of the positioning groove is the same as the cross-sectional shape of the bottom of the glass to be engraved.

3. The glass laser engraving device according to claim 1, characterized in that: The first bidirectional threaded slide rod is rotatably mounted on the workbench through a group of second bearing seats; the second bidirectional threaded slide rod is rotatably mounted on the workbench through a group of third bearing seats.

4. The glass laser engraving device according to claim 1, characterized in that: The first clamping block and the second clamping block are both provided with pressure sensors, and the pressure sensors are communicatively connected with the driving motor.

5. The glass laser engraving device according to claim 1, characterized in that: The workbench is also provided with a CCD camera, which is used to detect the clamping and engraving conditions of the glass.

6. The glass laser engraving device according to claim 1, characterized in that: At a preset time point, the real-time engraving parameters of the laser engraving device are obtained. Based on the real-time engraving parameters and in combination with the dictionary tree search graph, the predicted etching feature data of the glass at the current preset time point is obtained, specifically: Acquire real-time engraving parameters of a laser engraving device at a preset time node, and import the real-time engraving parameters into the dictionary tree search graph; Converting the real-time engraving parameters into the root node of the dictionary tree, introducing a cosine similarity algorithm, and calculating the cosine similarity between the root node and each branch node in the dictionary tree search graph based on the cosine similarity algorithm; Constructing a size sorting table, importing the cosine similarities between the root node and each branch node in the dictionary tree search graph into the size sorting table for size sorting, and sorting to obtain the highest cosine similarity; The branch node corresponding to the highest cosine similarity is calibrated, the leaf node belonging to the calibrated branch node is extracted, and the feature information of the extracted leaf node is interpreted to obtain the etching feature data corresponding to the extracted leaf node; The etching feature data corresponding to the extracted leaf node is output as the predicted etching feature data of the glass at the preset time node.

7. The glass laser engraving device according to claim 1, characterized in that: The laser engraving equipment is adjusted and optimized based on the predicted etching feature data of the glass at the current preset time node, specifically: Obtaining a preset engraving process plan for the glass, and obtaining preset engraving feature data range information of the glass at a current preset time node according to the preset engraving process plan; Determining the maximum preset engraving feature data and the minimum preset engraving feature data of the glass at the current preset time node according to the preset engraving feature data range information; The maximum preset engraving feature data of the glass at the current preset time node is used as the upper boundary condition, and the minimum preset engraving feature data of the glass at the current preset time node is used as the lower boundary condition; Setting a definition domain according to the upper boundary condition and the lower boundary condition, and determining whether the predicted etching feature data of the glass at the current preset time node completely falls within the definition domain; If the predicted etching feature data of the glass at the current preset time node completely falls within the definition domain, the real-time engraving parameters of the laser engraving device are not adjusted, and the real-time engraving parameters of the laser engraving device are continuously monitored at the next preset time node; If the predicted etching feature data of the glass cup at the current preset time node partially or completely falls outside the definition domain, the particle swarm optimization algorithm is introduced to iteratively optimize the real-time engraving parameters of the laser engraving device based on the particle swarm optimization algorithm, and the real-time engraving parameters of the laser engraving device at the next preset time node are readjusted so that the predicted etching feature data at the next time node falls within the definition domain again.

8. The glass laser engraving device according to claim 7, characterized in that: If the predicted etching feature data of the glass at the current preset time node partially or completely falls outside the definition domain, the particle swarm optimization algorithm is introduced to iteratively optimize the real-time engraving parameters of the laser engraving device based on the particle swarm optimization algorithm, and the real-time engraving parameters of the laser engraving device are readjusted at the next preset time node so that the predicted etching feature data at the next time node falls within the definition domain again. Specifically, Determine the particle population size in the particle swarm optimization algorithm, where each particle corresponds to a set of real-time engraving parameters of the laser engraving device, including the laser power, moving speed, and moving direction of the laser engraving head; Initialize the particle's position and velocity vectors, where the position vector represents a specific engraving parameter combination, and the velocity vector represents the parameter adjustment direction and amplitude; A fitness function is constructed based on the adaptability of the predicted etching feature data and the definition domain, and the fitness value of each particle is calculated. The fitness value reflects the effectiveness of the engraving parameter combination represented by the particle in making the predicted etching feature data fall into the definition domain. For each particle, its current fitness value is compared with the individual historical optimal fitness value. If it is better, the individual optimal position is updated; at the same time, the global optimal position is updated by comparing with the entire particle group; Based on the particle's current position, velocity, individual optimal position, and global optimal position, continue to update the particle's velocity and position vector to generate a new engraving parameter combination; Continue to iterate the above steps until the predicted etching feature data completely falls within the definition domain, and set the engraving parameters corresponding to the global optimal position as the real-time engraving parameters of the next preset time node, thereby ensuring that the predicted etching feature data at the next time node falls within the definition domain again.

Images