Laser-waterjet processing device and laser beam control method thereof

Abstract

This laser-waterjet processing device comprises: a water spray device for spraying a waterjet; a laser device for emitting a laser beam; a camera for photographing the laser beam formed on the waterjet; and an artificial intelligence controller. The artificial intelligence controller comprises: an artificial intelligence learning unit for storing an artificial intelligence training result for an image feature for each focal position of the laser beam; a focal position determination unit for determining the focal position of the laser beam by using an image of the laser beam, obtained by the camera, and the artificial intelligence training result; and a feedback control unit for controlling a driving unit by using the image of the laser beam, obtained by the camera, and a representative image of a reference focal position, stored in the artificial intelligence training result.

Description

Laser-waterjet processing device and laser beam control method thereof

[0001] The present invention relates to a laser-waterjet processing technology, and more particularly, to a laser-waterjet processing device capable of automatically aligning the focus position of a laser beam and a laser beam control method thereof.

[0002] A laser-waterjet processing device, combining laser and waterjet technologies, is being developed for processing high-strength, difficult-to-machine materials. The device essentially comprises a water jet nozzle, a laser generator, and an optical system. The device can process high-strength, difficult-to-machine materials like silicon carbide (SiC) into desired shapes without cracking and can achieve a long depth of cut.

[0003] However, thermal deformation of the optical system caused by high-power lasers can alter the laser beam's focus, potentially damaging key components like the spray nozzle or window, making long-term continuous operation difficult. Furthermore, if the laser is continuously irradiated while the spray nozzle or window is damaged, heat generation can pose a safety risk.

[0004] Embodiments of the present invention aim to provide a laser-waterjet processing device having a monitoring technology capable of automatically aligning the focus position of a laser beam and a method for controlling the laser beam thereof.

[0005] According to one embodiment, a laser-waterjet processing device includes a waterjet device, a laser device, a camera, and an artificial intelligence controller. The waterjet device includes a window for passing a laser beam and a nozzle for spraying a waterjet. The laser device includes a laser generator for emitting a laser beam, an optical system for changing the path of the laser beam to guide it to the window and focusing the laser beam, and a driving unit coupled to the optical system for adjusting the three-axis position of the laser beam focus. The camera photographs the laser beam focused on the waterjet. The artificial intelligence controller includes an artificial intelligence learning unit for storing artificial intelligence learning results for image features according to the focus position of the laser beam, a focus position determination unit for determining the focus position of the laser beam using the image of the laser beam acquired by the camera and the artificial intelligence learning results, and a feedback control unit for controlling the driving unit using the image of the laser beam acquired by the camera and a representative image of a reference focus position stored in the artificial intelligence learning results.

[0006] The optical system may include a first mirror and a second mirror for changing a path of a laser beam, a beam splitter, and at least one lens for focusing the laser beam. The beam splitter may be positioned between the camera and the window on the Z-axis. The driving unit may include a first driving unit for X-axis position adjustment coupled to the first mirror, a second driving unit for Y-axis position adjustment coupled to the second mirror, and a third driving unit for Z-axis position adjustment coupled to the at least one lens.

[0007] The artificial intelligence learning unit may include a classification task unit that categorizes and classifies images of a laser beam by focus position, an artificial intelligence learning model unit that performs artificial intelligence learning based on image characteristics by focus position, an artificial intelligence learning optimization unit that optimizes the artificial intelligence learning model so that new images are classified into the most similar category when input, and an artificial intelligence learning model storage unit that stores the optimized artificial intelligence learning model.

[0008] The focus position determination unit may include an image storage unit that stores an image of a laser beam acquired by a camera, an artificial intelligence learning model input unit that inputs the stored image into an artificial intelligence learning model, and a category output unit that receives and outputs information on a specific category classified by the artificial intelligence learning model.

[0009] The feedback control unit may include an image comparison unit that compares an image of a laser beam acquired by a camera with a representative image of a reference focus position stored in an artificial intelligence learning model, and a driving control unit that calculates a difference in the three-axis coordinates of the two images when the two images to be compared do not match, and outputs a control signal corresponding to the calculation result to the driving unit to adjust the focus position of the laser beam.

[0010] A method for controlling a laser beam of a laser-waterjet processing device according to one embodiment includes the steps of performing artificial intelligence learning on image features for each focus position of a laser beam, determining the focus position of the laser beam using an image of the laser beam acquired by a camera during an actual processing process and the artificial intelligence learning result, comparing the image of the laser beam acquired by the camera with a representative image of a reference focus position stored in an artificial intelligence learning model to determine whether they match, and, if they do not match, feedback controlling a driving unit coupled with an optical system.

[0011] The performance of artificial intelligence learning may include the steps of capturing multiple images of a laser beam by focus position, categorizing and classifying the captured images of the laser beam by focus position, extracting features from multiple images belonging to each category, learning the features of an image representing each category, optimizing an artificial intelligence learning model so that when a new image is input, it is classified into the most similar category, and storing the optimized artificial intelligence learning model.

[0012] Determining the focus position of a laser beam may include the steps of using a camera to capture a laser beam focused on a water jet, storing an image of the laser beam acquired by the camera, inputting the stored image into an artificial intelligence learning model, and receiving and outputting information on a specific category classified by the artificial intelligence learning model.

[0013] Feedback control may include steps of comparing an image of a laser beam acquired by a camera with a representative image of a reference focus position stored in an artificial intelligence learning model, determining whether the two images to be compared match, maintaining the focus position of the laser beam if the two images match, calculating the difference in the three-axis coordinates of the two images if the two images do not match, and outputting a control signal corresponding to the calculation result to a driving unit to adjust the focus position of the laser beam to match the reference focus position.

[0014] According to embodiments, when processing a workpiece using a laser and waterjet, an artificial intelligence controller can be used to automatically align the laser beam's focus with a reference focus. This can improve the processing precision of the workpiece, suppress thermal deformation of the optical system, prevent damage to key components, and increase continuous operation time.

[0015] FIG. 1 is a schematic diagram of a laser-waterjet processing device according to one embodiment.

[0016] Figure 2 is a schematic diagram of an artificial intelligence controller among the laser-waterjet processing devices illustrated in Figure 1.

[0017] FIG. 3 is a process flow diagram showing a laser beam control method of a laser-waterjet processing device according to one embodiment.

[0018] Figures 4 to 8 are photographs showing examples of laser beam images according to focus position.

[0019] Hereinafter, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. The present invention may be implemented in various different forms and is not limited to the embodiments described herein.

[0020] FIG. 1 is a schematic diagram of a laser-waterjet processing device according to one embodiment.

[0021] Referring to FIG. 1, the laser-waterjet processing device according to the present embodiment includes a water jet device (100) that sprays a water jet (10), a laser device (200) that irradiates a laser beam (20) toward the water jet (10), a camera (300) that photographs the laser beam (20) focused on the water jet (10), and an artificial intelligence controller (400) that determines the focus position of the laser beam (20) and moves the laser beam (20) to a reference focus position.

[0022] The water jet device (100) sprays a high-pressure water stream, a water jet (10), toward the object to be processed (50). The water jet (10) guides a laser beam (20). That is, the water jet (10) functions to guide the laser beam (20) like an optical fiber. Specifically, the water jet device (100) includes a water chamber (110) that accommodates water, and a window (120) and a nozzle (130) installed in the water chamber (110).

[0023] The water chamber (110) may be a high-pressure chamber that stores high-pressure water, and a supply pipe (not shown) for supplying water may be connected to the side wall of the water chamber (110). The window (120) is formed of a transparent window that transmits the laser beam (20) and may be positioned on the upper side of the water chamber (110). The window (120) may be made of a transparent material that does not cause distortion or refraction of the laser beam (20), such as crystal or tempered glass.

[0024] The nozzle (130) may be configured as a tube of a certain length and may be located at the lower side of the water chamber (110). Water in the water chamber (110) passes through the hollow of the nozzle (130) and is ejected in the form of a jet toward the object to be processed (50). A pressure reducing filter (not shown) may be installed in the nozzle (130) so that the water jet (10) can be emitted at a certain pressure. The window (120) and the nozzle (130) may be arranged to face each other along the Z-axis direction (vertical direction in FIG. 1).

[0025] The laser device (200) includes a laser generator (210) that emits a laser beam (20), an optical system (220) that focuses the laser beam (20) by changing the path of the laser beam (20) and guiding it to a window (120), and a driving unit (230) that is coupled to the optical system (220) and precisely controls the X-axis, Y-axis, and Z-axis positions of the focus of the laser beam (20).

[0026] The optical system (220) may include a plurality of mirrors (221, 222) and at least one lens (223, 224). The driving unit (230) may include a first driving unit (231) and a second driving unit (232) coupled to the plurality of mirrors (221, 222), and a third driving unit (233) coupled to at least one lens (223, 224). The first to third driving units (231, 232, 233) may be formed of a typical micro-actuator using piezoelectricity, etc.

[0027] In FIG. 1, an example in which the optical system (220) is composed of a first mirror (221), a second mirror (222), a beam splitter (225), a first lens (223), and a second lens (224) is illustrated, but the configuration of the optical system (220) is not limited to the illustrated example.

[0028] In the optical system (220) illustrated in FIG. 1, the first mirror (221) can refract the laser beam (20) emitted from the laser generator (210) at a right angle, and the second mirror (222) can refract the laser beam (20) reflected from the first mirror (221) at a right angle again. The laser beam (20) reflected from the second mirror (222) can be refracted at a right angle again by the beam splitter (225) and directed toward the window (120). The first lens (223) and the second lens (224) can be aligned between the beam splitter (225) and the window (120) to focus the laser beam (20) directed toward the window (120).

[0029] The first driving unit (231) can be coupled to the first mirror (221), and can adjust the X-axis position of the focus of the laser beam (20) by slightly changing the position of the first mirror (221) in one of the X-axis direction and the Y-axis direction, for example, in the X-axis direction. The second driving unit (232) can be coupled to the second mirror (222), and can adjust the Y-axis position of the focus of the laser beam (20) by slightly changing the position of the second mirror (222) in the other of the X-axis direction and the Y-axis direction, for example, in the Y-axis direction.

[0030] The third driving unit (233) can be coupled to each of the plurality of lenses (223, 224), and can adjust the Z-axis position (focus position) of the focus of the laser beam (20) by minutely changing the position of each of the plurality of lenses (223, 224) in the Z-axis direction. In Fig. 1, the reference focus position of the laser beam (20) is shown as P1. The focus of the laser beam (20) can be located at the center of the hollow inlet portion of the nozzle (130).

[0031] A camera (300) may be positioned above a beam splitter (225) and captures a laser beam (20) focused on a water jet (10). The camera (300) may include a CCD (Charge Coupled Device) sensor (310) that converts light into charge to obtain an image, a focusing lens (320) that adjusts the focus of the captured image, and a lens driving unit (not shown).

[0032] The artificial intelligence controller (400) determines the focus position of the laser beam (20) from the laser beam image acquired by the camera (300) based on artificial intelligence learning, and, if the laser beam (20) deviates from the reference focus position, operates any one of the first to third driving units (331, 332, 333) to perform feedback control to move the laser beam (20) to the reference focus position. The artificial intelligence controller (400) is a computer program and may be configured with a database for storing data, a processor storing an algorithm, etc.

[0033] Specifically, the artificial intelligence controller (400) includes an artificial intelligence learning unit (410) that stores artificial intelligence learning results for image features according to the focus position of the laser beam (20), a focus position determination unit (420) that determines the focus position of the laser beam (20) using the image of the laser beam (20) acquired by the camera (300) and the artificial intelligence learning results, and a feedback control unit (430) that compares the image of the laser beam (20) acquired by the camera (300) with a representative image of the reference focus position stored in the artificial intelligence learning model and feedback-controls the driving unit (320) when the laser beam (20) deviates from the reference focus position.

[0034] Figure 2 is a schematic diagram of an artificial intelligence controller among the laser-waterjet processing devices illustrated in Figure 1.

[0035] Referring to FIGS. 1 and 2, the artificial intelligence learning unit (410) may include a classification task unit (411) that categorizes and classifies images of a laser beam by focus position, an artificial intelligence (AI) learning model unit (412) that performs artificial intelligence learning based on image features by focus position, an artificial intelligence (AI) learning optimization unit (413) that optimizes the artificial intelligence learning model so that when a new image is input, it is classified into the most similar category, and an artificial intelligence (AI) learning model storage unit (414) that stores the optimized artificial intelligence learning model.

[0036] The classification task unit (411) assumes the Z-axis coordinate of the reference focus position as 0, and can categorize the focus position in specific ㎛ units along the +Z-axis direction (upward direction based on FIG. 1) and -Z-axis direction (downward direction based on FIG. 1). The categorization can be done by folder-by-folder classification or category classification. The artificial intelligence (AI) learning model unit (412) can extract features from multiple images belonging to each category, and learn the features of images representing each category.

[0037] The artificial intelligence (AI) learning optimization unit (413) may include, for example, processes of receiving a new image with missing information about the focus location, classifying the input image into a specific category based on the artificial intelligence learning model, providing information about the focus location to determine the validity of the category classification, and reflecting the judgment result to the artificial intelligence learning model. The process of changing the configuration of each layer within the algorithm constituting the artificial intelligence learning model and determining the number of learning repetitions may also be included in the artificial intelligence (AI) learning optimization unit (413).

[0038] The focus position determination unit (420) may include an image storage unit (421) that stores an image of a laser beam acquired by a camera in a specific folder during a process of processing an actual processing object, an artificial intelligence (AI) learning model input unit (422) that inputs the new stored image into an artificial intelligence learning model, and a category output unit (423) that receives and outputs information on a specific category classified by the artificial intelligence learning model. That is, the focus position determination unit (420) may determine the focus position of a laser beam from an image of the laser beam captured by the camera based on the artificial intelligence learning model.

[0039] The feedback control unit (430) may include an image comparison unit (431) that compares an image of a laser beam acquired by a camera with a representative image of a reference focus position stored in an artificial intelligence learning model, and a driving control unit (432) that calculates the difference in three-axis coordinates of the two images when the two images do not match and outputs a control signal corresponding to the calculation result to at least one of the first to third driving units.

[0040] For example, if the X-axis coordinate and Y-axis coordinate of two images are the same, the driving control unit (432) can calculate the difference in the Z-axis coordinate and output a control signal to the third driving unit (233). The third driving unit (233) can move the lenses (223, 224) according to the control signal to finely adjust the focus of the laser beam (20) in the Z-axis direction.

[0041] The artificial intelligence controller (400) can determine the focus position of the laser beam (20) in real time by the aforementioned configuration, and when the laser beam (20) deviates from the reference focus position, can control at least one of the first to third driving units (231, 232, 233) so that the laser beam (20) can maintain the reference focus position. Therefore, the laser-waterjet processing device of the present embodiment can automatically align the focus position of the laser beam (20) to improve the processing precision of the processing target (50).

[0042] FIG. 3 is a process flow diagram showing a laser beam control method of a laser-waterjet processing device according to one embodiment.

[0043] Referring to FIGS. 1 to 3, the laser beam control method of the laser-waterjet processing device according to the present embodiment includes the steps of performing artificial intelligence learning on image features for each focus position of the laser beam (20) (S100), determining the focus position of the laser beam (20) using the image of the laser beam (20) acquired by the camera (300) and the artificial intelligence learning result (S200), comparing the image of the laser beam (20) acquired by the camera (300) with a representative image of a reference focus position stored in an artificial intelligence learning model to determine whether they match, and feedback controlling the driving unit (320) if the two images do not match (S30).

[0044] The process (S100) of performing artificial intelligence learning may include a step (S110) of capturing multiple images of a laser beam (20) according to focus position, and categorizing and classifying the captured images of the laser beam (20) according to focus position. In the classification step (S110), assuming the reference focus position to be 0, the focus position can be categorized in specific ㎛ units along the +Z-axis direction (upward direction based on FIG. 1) and the -Z-axis direction (downward direction based on FIG. 1). The categorization may be performed by classification by folder or by category.

[0045] The process of performing artificial intelligence learning (S100) may include a step (S120) of preprocessing data such as image size and color to suit the artificial intelligence learning model, and a step (S130) of dividing the data set. The data set may be divided into a learning set, a test set, and a validation set.

[0046] The process of performing artificial intelligence learning (S100) may include a step of extracting features from multiple images belonging to each category and learning the features of an image representing each category (S140), a step of optimizing an artificial intelligence learning model so that when a new image is input, it is classified into the most similar category (S150), and a step of storing the optimized artificial intelligence learning model (S160).

[0047] The step (S150) of optimizing the artificial intelligence learning model may include, for example, detailed processes of inputting a new image with missing information about the focus location, classifying the input image into a specific category based on the artificial intelligence learning model, providing information about the focus location to determine the validity of the category classification, and reflecting the judgment result to the artificial intelligence learning model. In addition, the step (S150) of optimizing the artificial intelligence learning model may include a process of changing the configuration of each layer within the algorithm constituting the artificial intelligence learning model and determining the number of learning iterations.

[0048] The process (S200) for determining the focus position of a laser beam may include a step (S210) of photographing a laser beam (20) focused on a water jet (10) using a camera (300) in a process of processing an actual processing object (50), storing an image of the laser beam (20) acquired by the camera (300) in a specific folder, a step (S220) of inputting a new image stored in the folder into an artificial intelligence learning model, and a step (S230) of receiving and outputting information on a specific category classified by the artificial intelligence learning model. That is, the focus position of the laser beam (20) can be determined as a specific category classified by the artificial intelligence learning model.

[0049] The feedback control process (S300) may include a step (S310) of comparing an image of a laser beam (20) acquired by a camera (300) with a representative image of a reference focus position stored in an artificial intelligence learning model, and a step (S320) of determining whether the two images match. If the two images match, the laser beam (20) maintains the reference focus position, and thus feedback control is omitted (S330).

[0050] The feedback control process (S300) may include a step (S340) of calculating the difference in the three-axis coordinates of the two images when the two images do not match, and outputting a control signal corresponding to the calculation result to at least one of the first to third driving units (231, 232, 233) to adjust the focus position of the laser beam to match the reference focus position.

[0051] For example, if the X-axis coordinate and Y-axis coordinate of two images are the same, the difference in the Z-axis coordinate can be calculated and a control signal can be output to the third driving unit (233). The third driving unit (233) can move the lenses (223, 224) according to the control signal to finely adjust the focus of the laser beam (20) in the Z-axis direction. As a result, the focus position of the laser beam (20) can be automatically aligned in real time to improve the processing precision of the processing target (50).

[0052] Figures 4 to 8 are photographs showing examples of laser beam images according to focus positions. Figure 4 shows an image of a laser beam whose focus position matches the reference focus position. Figures 5 and 6 show images of a laser beam whose focus position is spaced 50 μm and 100 μm apart in the +Z-axis direction from the reference focus position, respectively. Figures 7 and 8 show images of a laser beam whose focus position is spaced 50 μm and 100 μm apart in the -Z-axis direction from the reference focus position, respectively.

[0053] The laser beam images shown in Figures 4 to 8 differ slightly in shape, size, and edge blurring, but these differences are difficult to clearly distinguish with the naked eye. However, using the aforementioned artificial intelligence controller and control method, images can be clearly distinguished according to focus position, enabling precise focus position measurement.

[0054] Although the preferred embodiments of the present invention have been described above, the present invention is not limited thereto, and various modifications can be made within the scope of the patent claims, the detailed description of the invention, and the attached drawings, and it is obvious that this also falls within the scope of the present invention.

[0055] - Explanation of symbols -

[0056] 10: Waterjet 20: Laser Beam

[0057] 100: Water spray device 110: Water chamber

[0058] 200: Laser device 210: Laser generator

[0059] 220: Optical system 221, 222: First and second mirrors

[0060] 223, 224: Lens 225: Beam splitter

[0061] 230: Drive unit 231, 232, 233: 1st to 3rd drive units

[0062] 300: Camera 310: CCD sensor

[0063] 320: Focusing lens 400: Artificial intelligence controller

[0064] 410: Artificial Intelligence Learning Unit 420: Focus Position Determination Unit

[0065] 430: Feedback control unit

Claims

1. A water jet device including a window through which a laser beam passes and a nozzle that sprays a water jet; A laser device comprising a laser generator that emits a laser beam, an optical system that focuses the laser beam and guides it to the window, and a driving unit coupled to the optical system that adjusts the focal position of the laser beam; A camera for photographing a laser beam formed on the water jet; and A laser-waterjet processing device comprising an artificial intelligence controller including an artificial intelligence learning unit that stores artificial intelligence learning results on features of images according to focus positions of a laser beam, a focus position determining unit that determines the focus position of the laser beam using the image of the laser beam acquired by the camera and the artificial intelligence learning result, and a feedback control unit that controls the driving unit using the image of the laser beam acquired by the camera and a representative image of the reference focus position stored in the artificial intelligence learning result.

2. In paragraph 1, The optical system includes a first mirror and a second mirror and a beam splitter for changing the path of the laser beam, and at least one lens for focusing the laser beam. A laser-waterjet processing device wherein the beam splitter is positioned between the camera and the window on the Z-axis.

3. In paragraph 2, A laser-waterjet processing device, wherein the driving unit includes a first driving unit coupled to the first mirror for X-axis position adjustment, a second driving unit coupled to the second mirror for Y-axis position adjustment, and a third driving unit coupled to the at least one lens for Z-axis position adjustment.

4. In any one of paragraphs 1 to 3, The above artificial intelligence learning unit, A classification task that categorizes and classifies images of laser beams by focus position; An artificial intelligence learning model section that performs artificial intelligence learning based on image features by focus location; An artificial intelligence learning optimization unit that updates the artificial intelligence learning model by classifying a newly input image into one of multiple categories classified by the classification task unit; and A laser-waterjet processing device including an artificial intelligence learning model storage unit for storing the updated artificial intelligence learning model.

5. In paragraph 4, The above focus position determination unit, An image storage unit for storing an image of a laser beam acquired by the above camera; An artificial intelligence learning model input unit that inputs the saved image into the artificial intelligence learning model; and A laser-waterjet processing device including a category output unit that receives and outputs information on one of the categories derived by inputting the stored image among the plurality of categories into the artificial intelligence learning model.

6. In paragraph 5, The above feedback control unit, An image comparison unit that compares the image of the laser beam acquired by the camera with the representative image stored in the artificial intelligence learning model; and A laser-waterjet processing device including a driving control unit that calculates a positional difference between the two images when the two images to be compared do not match, and outputs a control signal corresponding to the calculation result to the driving unit to adjust the focal position of the laser beam.

7. Perform artificial intelligence learning on image features according to the focus position of the laser beam, During the actual processing, the focus position of the laser beam is determined using the image of the laser beam acquired by the camera and the results of artificial intelligence learning. A method for controlling a laser beam of a laser-waterjet processing device, the method comprising: comparing an image of a laser beam acquired by the above camera with a representative image of a reference focus position stored in an artificial intelligence learning model to determine whether they match, and if they do not match, controlling a driving unit coupled with an optical system by feedback.

8. In paragraph 7, The above artificial intelligence learning is performed, Multiple images of the laser beam are captured at each focus position, and the captured images of the laser beam are categorized and classified by focus position. Extract features from multiple images belonging to each of multiple categories, and learn the features of images representing each category. When a new image is input, the AI ​​learning model is updated so that it is classified into one of the above multiple categories. A method for controlling a laser beam of a laser-waterjet processing device storing the above-mentioned updated artificial intelligence learning model.

9. In paragraph 8, Determination of the focus position of the above laser beam is as follows: Using the above camera, a laser beam focused on a water jet is captured, and an image of the laser beam acquired by the camera is stored. The above saved image is input into the artificial intelligence learning model, A method for controlling a laser beam of a laser-waterjet processing device, which receives and outputs information on a specific category into which the stored image is classified through the artificial intelligence learning model.

10. In paragraph 9, The above feedback control is, Compare the image of the laser beam acquired by the above camera with the representative image of the reference focus position stored in the above artificial intelligence learning model, Determine whether the two images being compared match, If the two images above match, the focus position of the laser beam is maintained, A laser beam control method of a laser-waterjet processing device, wherein the method calculates the positional difference between the two images when the two images do not match, and outputs a control signal corresponding to the calculation result to the driving unit to adjust the focus position of the laser beam so that it matches the reference focus position.

Images