Monitoring laser machining process to be carried out on workpiece, comprises detecting actual measuring values by sensor, which monitors the laser machining process, and determining actual characteristic values from actual measuring values

摘要

The method comprises detecting two actual measuring values by a sensor, which monitors a laser machining process, determining two actual characteristic values from the two actual measuring values using neural networks, where the actual characteristic values mutually represent an actual fingerprint in a characteristic value range, providing a predetermined point set in the characteristic value range, and classifying the laser machining process by detecting the position of the actual fingerprint relative to the predetermined point set in the characteristic value range. The method comprises detecting two actual measuring values by a sensor, which monitors a laser machining process, determining two actual characteristic values from the two actual measuring values using neural networks, where the actual characteristic values mutually represent an actual fingerprint in a characteristic value range, providing a predetermined point set in the characteristic value range, classifying the laser machining process by detecting the position of the actual fingerprint relative to the predetermined point set in the characteristic value range, regulating a process parameter of an associated actuator, so that the actuator is activated when the actual fingerprint exits the predetermined point set of the characteristic value range. The change of the associated process parameter corresponds to gradients in the characteristic value range, which extends itself from the fingerprint in the direction of the point set in the characteristic value range. The determination of the actual characteristic value from the actual measuring value comprises a process for data reduction or dimension reduction as a main component analysis, multidimensional scaling, support vector machine or support vector classification, or industrial standard organization-manufacturing automation protocol process. The predetermined point set is fixed within the characteristic value range using a learning process. The gradient field of the characteristic value range is determined in different areas at the points in the characteristic value range in dependent of the process parameter, where the points in the characteristic value range are representative for the respective area with respect to the gradients. The gradients of the characteristic value range are determined by variation of the process parameter at a predetermined point of the characteristic value range in dependent of the process parameter. The sensor comprises a photodiode with filter for the determined wavelengths, body- and airborne receiver, and a camera unit with a corresponding surface illumination. The actuator comprises a controller of the laser capacity, a speed controller of a machining head relative to a workpiece, a controller of the focal position of the machining laser beam, a controller of the distance of the machining head to the workpiece, and a controller of lateral offsets. Independent claims are included for: (1) a device for monitoring a laser machining process to be carried out on a workpiece; and (2) a laser machining head for processing a workpiece.