Method and device for monitoring a laser processing operation to be performed on a workpiece and laser processing head having such a device

1. A method for monitoring a laser processing operation to be performed on a workpiece, having the following steps:

detecting at least two current, measured values by using at least one sensor that monitors the laser processing operation,

determining at least two current features from the at least two current measured values, the at least the two current features jointly representing a current fingerprint in a feature space,

providing a predetermined point set in the feature space, wherein the predetermined point set is defined inside the feature space using a learning process,

classifying a process state of the laser processing operation by detecting a position of the current fingerprint relative to the predetermined point set in the feature space, the current fingerprint being a current fingerprint for an observed instant, wherein the at least one sensor comprises at least one photodiode and at least one camera unit for imaging a laser processing zone, and wherein for workpiece illumination an illumination unit with optics is used to focus onto the at least one camera unit details on a workpiece surface, to obtain the at least two current measured values from the at least one camera unit regarding the melted workpiece area, and

closed-loop controlling at least one process parameter of an actuator in such a way that, when the current fingerprint leaves the predetermined point set of the feature space, the actuator is activated so that the change in the process parameter of the actuator corresponds to a gradient in the feature space that, starting from the fingerprint, extends from said fingerprint in a direction towards the predetermined point set in the feature space.

2. The method as claimed in claim 1 , wherein the at least one photodiode is provided with filters for specific wavelengths.

3. The method as claimed in claim 1 , wherein the step of determining the at least two current features from the at least two current measured values comprises a method for data reduction or dimension reduction, and that the classifying comprises support vector machines or a support vector classification.

4. The method as claimed in claim 2 , wherein for determining the gradient of the feature space a reference run with a varied process parameter is performed, wherein in the process an upper and a lower specification boundary occurs and is overshot.

5. The method as claimed in claim 1 , wherein camera images captured by the at least one camera unit are recorded with different exposure times and processed with each other using a high dynamic range (HDR) method.

6. The method as claimed in claim 2 , wherein the at least one actuator is selected from a group that comprises a controller for controlling the laser power, a controller for controlling the processing velocity of the processing head relative to the workpiece, a controller for controlling the focal point of the processing laser beam, a controller for controlling the distance of the processing head from the workpiece, a controller for controlling the lateral offset, and a controller for controlling the supply of a process gas.

7. A device for performing a method for monitoring a laser processing operation to be performed on a workpiece in accordance with claim 1 , comprising:

a sensor unit including at least one photodiode and at least one camera unit for imaging the laser processing zone that are configured to detect at least two current measured values from the at least one camera unit regarding a melted workpiece area,

an illumination unit configured to illuminate the workpiece to be processed, with optics for focusing onto the at least one camera unit details on a workpiece surface,

a data processing unit configured to determine at least two current features from the at least two current measured values in order to create a current fingerprint in a feature space,

a memory unit configured to store a predetermined point set inside the feature space, wherein the predetermined point set is defined inside the feature space using a learning process,

a classification unit configured to evaluate a process state of the laser processing operation by detecting the position of the current fingerprint relative to the predetermined point set in the feature space, the current fingerprint being a current fingerprint for an observed instant, and

a control unit configured to control at least one process parameter of an actuator in such a way that, when the current fingerprint leaves the point set of the feature space, the actuator is activated so that the change in the process parameter of the actuator corresponds to a gradient in the feature space that, starting from the fingerprint, extends from said fingerprint in a direction towards the predetermined point set in the feature space.

8. The device as claimed in claim 7 , wherein the illumination unit comprises light-emitting diodes or lasers.

9. The device as claimed in claim 8 , wherein the at least one actuator is selected from a group that comprises a controller for controlling the laser power, a controller for controlling the processing velocity of the processing head relative to the workpiece, a controller for controlling the focal point of the processing laser beam, a controller for controlling the distance of the processing head from the workpiece, a controller for controlling the lateral offset, and a controller for controlling the supply of a process gas.

10. A laser processing head that irradiates a laser beam for processing a workpiece, having the device as claimed in claim 7 .