Method and apparatus for monitoring and regulating a laser machining process

Abstract

The invention relates to a method and a device for monitoring and regulating a laser machining process on a workpiece (13). The method and the device are characterized in that a recording (16) of a process zone of the workpiece (13) is preprocessed (17, 18) using an AI method, disturbances are eliminated from the recording (16), geometric workpiece properties are detected, and an input signal (19) for an image processing algorithm (20) is generated therefrom, which algorithm calculates correction values for at least one control parameter (21) of the laser path.

Description

Background Technology

[0001] In laser processing, proper alignment of the processing laser beam with the workpiece is crucial for high-quality machining. Therefore, various monitoring and control systems with sensors are used in laser processing systems for object identification and distance measurement between the optical focusing system for the processing laser beam and the workpiece.

[0002] By using a line laser positioned in front of the processing laser beam and an imaging device that captures both the reflection of the line laser beam and the process zone created by the processing laser beam in the workpiece, distance measurement can be performed using an image processing device, and height differences within the workpiece can also be detected. Therefore, during laser welding, the transition from the top plate to the bottom plate can be detected, and the processing laser can be guided along this transition line to produce a precise weld. However, this solution is limited to monitoring in the feed direction of the processing laser beam.

[0003] In contrast, when using an OCT scanner for monitoring, image recording can detect height differences within the workpiece and distance measurements in all spatial directions, regardless of the feed direction of the processing laser. However, due to the use of an OCT scanner, this monitoring system is relatively complex and expensive.

[0004] From EP2365890B1, a monitoring system for a laser processing system is known. This monitoring system includes an OCT scanner, a camera with HDR functionality, and an image processing unit. The camera records the process area multiple times with different exposure times and calculates a high-contrast image based on the recordings. This image is then supplied as a measurement to the image processing unit. The image processing unit compares the measurement values ​​with key values ​​determined through data simplification processing. Alternatively, parameters can also be calculated using a neural network. The determined parameters enable the classification of the laser processing process. Furthermore, the path and / or distance from the processing laser to the workpiece can be readjusted based on the determined parameters.

[0005] The purpose of this invention is to monitor and regulate the laser processing process using a simple structural method. Summary of the Invention

[0006] This objective is achieved through a method for monitoring and controlling the laser processing on a workpiece, the method comprising the following steps: - Use an image recording device to record the environment of the process zone (Prozesszone) in the workpiece generated by the processing laser beam; - Use artificial intelligence methods in image processing devices to eliminate interfering factors in the recording, such as process luminescence, smoke, or material splashes; - Use artificial intelligence methods to detect geometric workpiece characteristics in the records, such as workpiece edges and / or height jumps; - Using artificial intelligence methods, input signals for downstream image processing algorithms are generated from records that have been cleaned of interfering factors and contain information about the characteristics of the geometric workpiece; - Use image processing algorithms to calculate the correction value of at least one control parameter of the path of the machining laser beam on the workpiece.

[0007] The method according to the invention allows for the monitoring and adjustment of the processing laser beam path using only a few devices, which are typically already present on the processing head of the laser processing system. No additional line laser or OCT scanner is required. Process light emitted from the laser processing can be used as indirect illumination for recording.

[0008] According to the invention, the monitoring and adjustment are independent of the feed direction of the processing laser beam, and therefore, especially in the laser welding process, weld position adjustment can be performed even in cases of complex weld geometry.

[0009] The artificial intelligence approach also enables continuous improvement of the preprocessing of the recordings of the process zone, thus allowing robust conclusions about the position of the machining laser beam relative to the workpiece edges, welds, and other elements of the workpiece to be drawn according to the method of the invention. By preprocessing the recordings of the process zone and its surrounding environment, an input signal unaffected by all interfering factors can be provided to the actual image processing algorithm. Based on this input signal, the image processing algorithm can detect geometrical characteristics of interest in the workpiece, such as height jumps, and calculate correction values ​​for at least one control parameter of the path used for the machining laser beam. The parameterization of the image processing algorithm can be more easily selected compared to methods according to the prior art.

[0010] Additional advantages arise when artificial intelligence is used not only to eliminate interference in the recordings, but also to enhance the information about the geometric characteristics of the workpiece in the input signal used by the image processing algorithm. This enables the image processing algorithm to identify these workpiece characteristics with greater determinism and, when necessary, generate corresponding correction values ​​for the control parameters of the laser path.

[0011] A further advantage is that the parameters of the image processing algorithm do not need to be changed by the processed input signal as in the case of methods according to existing techniques. Preprocessing the records allows the parameters of the image processing algorithm to remain unchanged by ensuring the uniqueness of the information in the input signal. In this case, parameterization can also be significantly simpler than using known methods.

[0012] Further enhancement of the input signal can be achieved by using an image recording device to generate multiple records with different exposure times and using a high dynamic range (HDR) method to compute the multiple records with each other to generate an image with high contrast values, using artificial intelligence methods to eliminate interference factors from the image and detect geometric workpiece characteristics.

[0013] The recording is preferably performed using an image recording device coaxial with the processing laser beam.

[0014] The robustness of the method can also be improved by incorporating information about the workpiece shape and the laser beam path into the preprocessing of the recordings in the process area. In this way, the AI ​​method can limit the search space in the recordings for interfering factors and the geometric characteristics of the workpiece of interest.

[0015] Artificial intelligence methods can be preferably used to compute single-layer or multi-layer neural networks for the detection of interfering factors and geometric characteristics of workpieces. However, AI methods can also be based on other methods, such as linear regression.

[0016] The method according to the invention can be used in all laser processing procedures. However, it has particular advantages when used for monitoring and regulating laser welding processes. In these processes, accurate detection of the joint line between two workpieces is crucial. If a top plate is welded to a bottom plate, the joint line can be detected by height variations within the workpiece.

[0017] The present invention also relates to an apparatus for performing a method according to the invention for monitoring and controlling a laser processing process on a workpiece, the apparatus comprising: an image recording device arranged coaxially with the processing laser beam of the laser processing system, and an image processing device, characterized in that the image processing device is designed to perform preprocessing on the recording from the image recording device using an artificial intelligence-based method, wherein, by means of the preprocessing, interference factors in the recording are removed and information about the geometric characteristics of the workpiece that are critical to the processing process is enhanced before the image processing device subjectes the preprocessed recording to an image processing algorithm and thereby generates control parameters for the path of the processing laser beam.

[0018] Process light emitted from laser processing can be used as indirect illumination for recording. However, additional light sources can also be provided.

[0019] The image processing device can preferably compute a neural network and image processing algorithm for artificial intelligence methods, and record the image processing process performed by the algorithm. In this way, the allocation of new workpieces can be achieved.

[0020] Image recording devices can advantageously be camera devices with HDR functionality.

[0021] The device can also be relatively easily retrofitted into existing laser processing systems.

[0022] The present invention also includes a processing head for a laser processing system having the means according to the invention. Attached Figure Description

[0023] In the following description, preferred exemplary embodiments of the device according to the invention will be described in more detail with reference to the accompanying drawings. Detailed Implementation

[0024] The only accompanying drawing shows a schematic block diagram of the laser processing system, excluding further details, of the processing head 10 and the image processing device 11. The processing head 10 guides the processing laser beam 12 onto the workpiece 13. Light 14 reflected from the workpiece 12 is detected by an image recording device 15, preferably an HDR camera, via a recording 16, which is then forwarded to the image processing device 11.

[0025] In the image processing apparatus 11, the recording 16 from the image recording apparatus 15 is preprocessed 17, wherein interfering factors such as reflections from splashes, smoke, or process luminescence are eliminated from the recording 16 by means of artificial intelligence methods. The preprocessed recording 16' then undergoes further processing 18, wherein workpiece edges, height variations, and other geometric workpiece characteristics in the recording are identified through artificial intelligence processing. This generates input signals 19 for an image processing algorithm 20, which uses these signals 19 to generate control signals 21 for the processing head 10. These signals contain the recording 16' free of interfering factors and have enhanced information about the geometric workpiece characteristics. These control signals 21 undergo further error control 22 before being forwarded to the processing head 10. The control signals 21 can be specifically used to correct the optical axis of the processing head 10.

[0026] List of reference numerals 10 processing heads 11 Image processing device 12. Processing laser beams 13 Workpieces 14 Reflected light 15 Image recording device 16, 16' record 17 Preprocessing 18. Processing 19 Input Signal 20 Image Processing Algorithms 21 Control Parameters 22 Error Control

Claims

1. A method for monitoring and regulating laser processing on a workpiece, the method comprising the following steps: - The environment of the process zone generated by the processing laser beam (12) in the workpiece (13) is recorded using an image recording device (15); - Artificial intelligence methods are used in the image processing device (11) to eliminate interfering factors in the recording (16), such as process luminescence, smoke or material splashes; - Use the artificial intelligence method to detect geometric workpiece characteristics such as workpiece edges and / or height jumps in the records (16, 16'); - The artificial intelligence method is used to generate an input signal (19) for the downstream image processing algorithm (20) from the record (16) which has been cleared of interference factors and contains information about the characteristics of the geometric workpiece. - The image processing algorithm (20) is used to calculate the correction value of at least one control parameter (21) of the path of the processing laser beam (12) on the workpiece (13).

2. The method according to claim 1, characterized in that, The process light emitted during the laser processing is used as indirect illumination for the recording (16).

3. The method according to claim 1 or 2, characterized in that, The artificial intelligence method is used to enhance the information about the geometric workpiece characteristics in the input signal (19) used for the image processing algorithm (20).

4. The method according to any one of the preceding claims, characterized in that, The parameters of the image processing algorithm (20) are not changed by the preprocessed input signal (19).

5. The method according to any one of the preceding claims, characterized in that, The image recording device (15) generates multiple records (16) with different exposure times, and the multiple records are computed to each other using a high dynamic range method (HDR method) to generate an image (16) with a high contrast value. The artificial intelligence method is used to remove the interference factors from the image and detect the geometric workpiece characteristics.

6. The method according to any one of the preceding claims, characterized in that, The record (16) is recorded coaxially with the processing laser beam (12) using the image recording device (15).

7. The method according to any one of the preceding claims, characterized in that, When preprocessing the records in the process area, the artificial intelligence method takes into account information about the shape of the workpiece (13) and the path of the laser beam (12).

8. The method according to any one of the preceding claims, characterized in that, The artificial intelligence method described above is used to compute at least one neural network for identifying the interference factors and the geometric characteristics of the workpiece.

9. The method according to any one of the preceding claims, characterized in that, The method described herein is used for monitoring and regulating the laser welding process.

10. An apparatus for performing a method for monitoring and controlling a laser processing process on a workpiece (13) according to any one of the preceding claims, the apparatus comprising: an image recording device (15) arranged coaxially with a processing laser beam (12) of a laser processing system, and an image processing device (11), characterized in that, The image processing device (11) is designed to perform preprocessing (17) on the recording (16) from the image recording device (15) using an artificial intelligence-based method. With the preprocessing, interference factors in the recording (16) are removed and information about the workpiece (13)’s geometric characteristics that are crucial to the processing process is enhanced before the image processing device subjects the preprocessed recording (19) to an image processing algorithm (20) and thereby generates control parameters (21) for the path of the processing laser beam (12).

11. The apparatus according to claim 10, characterized in that, The image processing device (11) calculates the neural network for the artificial intelligence method and the image processing algorithm (20).

12. The apparatus according to claim 10 or 11, characterized in that, The image recording device (15) is a camera device with HDR function.

13. The apparatus according to any one of claims 10 to 12, characterized in that, The device can be retrofitted into existing laser processing systems.

14. A processing head for a laser processing system, characterized in that, The processing head has the means according to any one of claims 10 to 13.

Images