Computer-implemented method for determining a dynamic laser beam shape, control unit, computer program product, and computer-readable storage medium

Abstract

In one aspect, the present invention relates to a control unit (100) for determining a dynamic laser beam shape to control a laser cutting machine (L), the laser cutting machine (L) being provided with a dynamic beam shaping module for changing the shape of the laser beam, for example, the control unit (100) having: a cutting plan interface (101) configured to receive a cutting plan for a part to be processed for cutting out a workpiece, wherein each part is defined by a cutting profile, wherein the cutting profile is segmented into a set of cutting segments, and wherein each workpiece is characterized by a characteristic index selected from the group including a material index and / or a thickness index; an interface (102) with a shape storage device (ShS) having a stored set of dynamic laser beam shapes, in particular more than two dynamic laser beam shapes; a processor (P) configured to iteratively and automatically calculate, for each cutting segment of all parts to be cut out from the workpiece, an assignment of a dynamic laser beam shape from the set of dynamic laser beam shapes stored in the shape storage device (ShS), wherein the calculation of the assignment (S3) is based on the characteristic index of the workpiece and is specific to the respective cutting segment; and wherein the processor (P) is further configured to provide control instructions (CI) via an output interface (103) for controlling the laser cutting machine (L) to execute the received cutting plan by specifically applying the determined dynamic laser beam shape for each cutting segment.

Description

Technical Field

[0001] The present invention relates to laser cutting by means of a laser cutting machine provided with a dynamic beam shaping module or at least one other optical module for dynamically changing the shape of a laser beam. In particular, the present invention relates to a method, a control unit, a computer program, and a computer-readable storage medium for determining a dynamic laser beam shape during laser cutting. Background Art

[0002] In laser cutting applications, high quality and performance are key factors therein.

[0003] Generally, the laser cutting process can be optimized for opposite requirements such as productivity and quality, especially. The higher the feed rate of the laser cutting head, the higher the productivity, but the quality may be reduced because the laser beam acts on the material with a certain energy distribution defined by, for example, spot size, laser beam shape, and laser power. The key factor in cutting is to convert the absorbed laser energy into heat to melt the material. The energy coupling is determined by many factors and interacts with, for example, the cutting notch conditions.

[0004] To optimize the above-mentioned requirements, in the prior art, it is known to influence the beam profile by, for example, changing the intensity distribution, spot size, laser beam shape, and focal position. The first option to obtain such a modification is static beam shaping (SBS), mainly by spatial methods. By using SBS, the laser beam is provided before the start of processing and does not change anymore. The second option is to modify a laser beam with dynamic characteristics. In this case, the characteristics of the laser beam may change during processing by dynamic beam shaping (DBS). Alternatively, some spatial modulation methods can also be performed dynamically, for example, by adaptive optical techniques. In this regard, reference is made to "Dynamic beam shaping for thick sheet metal cutting", Cindy Goppold, Thomas Pinder, Patrick Herwig, IWS, in: Lasers in Manufacturing Conference 2017.

[0005] When using DBS, by the spatio-temporal distribution of laser energy on the material surface, the key challenge is solved: to have a sufficient notch size with a reduced spot size to obtain the available laser energy. For this purpose, the high laser energy of the small spot size oscillates periodically and is superimposed on the feed rate. Thus, the energy distribution is around the generated cutting notch and thereby acts as an artificial larger spot. At this time, the notch width is enlarged, which obtains an unobstructed melt ejection. In addition, because the interaction time of the laser beam with the material is reduced, this distribution prevents heat accumulation.

[0006] For applying DBS, a laser cutting machine is equipped with a dynamic laser beam shaping module. Example embodiments with such DBS applications are described in WO2019 145536A1.

[0007] It is known from US2010 / 0059490 A1 to rapidly modify a laser processing beam and in particular its spatial intensity distribution. SUMMARY OF THE INVENTION

[0008] The object of the invention is to improve the quality of the cutting result and to distinguish between different types of cutting segments when selecting or determining an appropriate dynamic laser beam shape.

[0009] This object is achieved by the appended independent claims. Other advantageous embodiments and features are mentioned in the dependent claims and in the following description.

[0010] In one aspect, the invention relates to a computer-implemented method for determining a dynamic laser beam shape for laser cutting of a workpiece by means of a laser cutting machine, the laser cutting machine comprising at least one optical module for dynamically changing the shape of the laser beam. The method may at least comprise:

[0011] - receiving a cutting plan to be processed for cutting out parts of a workpiece, wherein each part is defined by a cutting profile, the cutting profile comprising a set of cutting segments, and wherein each (to-be-cut) workpiece is characterized by a characteristic index selected from the group consisting of a material index and / or a thickness index;

[0012] - providing a shape storage device having a set of dynamic laser beam shapes, in particular more than two dynamic laser beam shapes;

[0013] - by a processor: iteratively and automatically calculating, for each cutting segment of all parts to be cut out from the workpiece, an assignment of a dynamic laser beam shape from the set of dynamic laser beam shapes stored and accessed in the shape storage device, wherein the calculation of the assignment is based on or takes into account or considers the characteristic index of the workpiece, and wherein the calculation of the assignment is specific to the respective cutting segment;

[0014] - by a processor: providing control instructions for controlling the laser cutting machine to execute the received cutting plan by specifically applying the determined dynamic laser beam shape for each cutting segment.

[0015] Typically, a laser cutting machine may include more than one optical model that helps or enables the dynamic change of the laser beam. For example, two (2) galvanometer scanning mirrors can be used, one for movement along the X and one for movement along the Y. Alternatively or additionally, 3D beam shaping can be achieved by means of a 2-axis module for X / Y variation and / or a Z wobble module for movement in the direction of the beam axis. Alternatively or additionally, a CIVAN laser system can be used to shape the beam by, for example, interconnecting 32 individual optical modules.

[0016] The assignment is segment-specific. Different dynamic laser beam shapes will be assigned to different types of segments (e.g., straight lines, curves). In particular, segments to be cut at different speeds are subject to different processing, and in particular, different dynamic laser beam shapes are assigned to such different segments cut at different speeds. Typically, dynamic laser beam shapes are assigned to (certain) segments, while (certain) segments are not assigned to dynamic laser beam shapes because the segments are fixed according to the geometry defined in the cutting plane.

[0017] In a preferred embodiment, the method may further include an intermediate verification step. After the assignment (segment - dynamic laser beam shape) has been calculated by a processor, in particular by an algorithm (assignment algorithm), the assignment can be provided on a user interface for verification purposes. In the case of detecting a verification signal, the assignment can be applied. Otherwise, a correction process can be initiated. The correction process can be performed online (e.g., during the cutting process) or offline (e.g., independent of and / or not during the cutting process) by means of an algorithm and can be configured to calculate the deviation from the assignment. For example, the user can manually select a different assignment (e.g., another dynamic laser beam shape for a specific segment). Alternatively, different assignments can be suggested algorithmically by, for example, considering historical data of other assignments or statistical evaluations (e.g., averages).

[0018] In a preferred embodiment of the present invention, the types of cutting segments are selected from the group including the following:

[0019] Straight lines;

[0020] Circles or circular segments with a configurable specific radius;

[0021] Corners with a configurable angle,

[0022] Parametric curves;

[0023] Penetrations;

[0024] Introductions;

[0025] Exits and / or

[0026] Engravings.

[0027] In another preferred embodiment of the invention, the step of (automatically) calculating the assignment is performed by a trained model, in particular a neural network model, which provides specific segments as input and specific dynamic laser beam shapes as output.

[0028] In another preferred embodiment of the present invention, the model is trained with training data including:

[0029] - Selected dynamic laser beam shape for test cutting;

[0030] - Characteristics of the workpiece being tested (material type and / or material thickness);

[0031] - the type of cut section being tested;

[0032] - A selected dynamic beam shape for a test cut segment with an evaluation data set used as annotation data labels, in particular for a quality assessment or another assessment (to be explained in more detail below).

[0033] Preferably, the model is trained by performing the following steps:

[0034] - preselecting a dynamic laser beam shape from a set of dynamic laser beam shapes;

[0035] -Perform a cutting segment specific test cut with a pre-selected dynamic laser beam shape;

[0036] - performing an evaluation of the results of the test cuts by providing an evaluation data set for each test cut;

[0037] -Adjust the weights of the model, e.g., to optimize the objective function used to evaluate the dataset.

[0038] The model has been trained to assign specific dynamic laser beam shapes to specific types of segments. The learning algorithm or training algorithm is configured to automatically find the "best" assignment or an assignment based on a "best" evaluation. The training algorithm can be based on an evaluation data set.

[0039] The evaluation can be a quality evaluation, a performance evaluation, an energy consumption evaluation, a process stability evaluation, a burr height evaluation, a roughness evaluation, a feed rate evaluation, a kerf width evaluation, a gas consumption evaluation, a contour error evaluation, a tilt angle / rectangularity evaluation, a flatness cutting edge evaluation, a heat affected zone evaluation. Regarding the process stability evaluation, the following example is given: it is possible to have a setting that produces very good quality, but the setting is unstable and small changes in the system and / or material will result in worse quality. Therefore, in a preferred embodiment, process stability is considered in the evaluation and evaluation data set. In a preferred embodiment of the present invention, more than one type of evaluation is performed and a combination of different evaluations is provided, such as a quality evaluation and a performance evaluation and an energy consumption evaluation.

[0040] Typically, the evaluation can be performed automatically by a machine or automatically by software with the aid of a sensory automatic evaluation unit. The sensory automatic evaluation unit can include a process optical system, in particular a camera device and / or a diode. The optical system can be attached to the cutting head and directed at the processing area on the workpiece being processed.

[0041] Alternatively or additionally, the evaluation can be performed manually with the aid of user input received on the human-machine interface. The evaluation data set can include a configurable share of manually or automatically jointly setting different evaluation criteria, including the above-mentioned quality evaluation, performance evaluation, and / or process stability evaluation and / or other evaluations. Since the different evaluation criteria are interdependent, in the case of manually setting the configurable shares of different evaluation criteria, the interdependence can be adjusted on the user interface selection buttons that can be set on the human-machine interface HMI.

[0042] In addition, the above-mentioned two options (manual or automatic evaluation) can be combined such that the two modes can be used as verification steps. First, an automatic evaluation is provided and output on the human-machine interface. Second, this automatic evaluation can be verified by the user. The user can accept or reject the automatic evaluation, and in the case of rejection, a user input signal indicating a manual evaluation can be provided.

[0043] According to a preferred embodiment, the automatically determined dynamic laser beam shape for each cutting segment in the cutting section is determined specific to certain types of cutting machines. The types of cutting machines include inertia indicators (such as laser cutting heads and corresponding participants for moving the laser cutting head), size indicators of the laser cutting machine, and / or machine characteristics.

[0044] According to another preferred embodiment, in a set (of different) dynamic laser beam shapes, the dynamic laser beam shape is dynamically changed by generating a focal oscillation shape based on the spatio-temporal distribution of laser energy on the material surface and / or the focal plane with respect to:

[0045] The frequencies in the X and Y directions and preferably in the X, Y, and Z directions;

[0046] The amplitudes in the X and Y directions and preferably in the X, Y, and Z directions; and / or

[0047] The phase shift in the Y direction with respect to the X direction, preferably the phase shift in the Y and Z directions with respect to the X direction.

[0048] Alternatively or additionally, the dynamic laser beam shape is changed by the oscillation of the focus in the X and Y directions (with respect to frequency and / or amplitude), which can be combined with a wobble (in the Z direction) as the laser beam diameter changes.

[0049] Alternatively or additionally, contour errors can be determined (estimated and / or measured) and these determined contour errors can be compensated. Compensation of the contour errors can be performed by means of an assigned correction calculation. Measuring the contour errors can be performed by means of, for example, a coaxial camera device.

[0050] For each position or time step, due to the previous measurements in the X and Y directions (c x (t), c y (t)), the contour deviation is known.

[0051] There are at least two methods for compensating contour errors (offset, DBS amplitude).

[0052] Method offset:

[0053] This contour error value is used as the offset value for the dynamic laser beam shape. Since the contour error varies with the part / time, each position / time step has a different offset value.

[0054] Example:

[0055] · The dynamic laser beam shape is a Lissajous figure LF_1

[0056]

[0057] · The Lissajous figure for contour error compensation is LF_2

[0058]

[0059] DBS amplitude:

[0060] Instead of using an offset, the amplitude of the dynamic laser beam shape can be adjusted over the part / time.

[0061] Example:

[0062]

[0063] According to another preferred embodiment, the transition region is determined by a linear function, a non-linear function, and / or a logarithmic function and / or other transition functions. The transition region can be interpreted in a spatially and / or time-related sense. The transition region can be defined as the transition between two consecutive segments of the workpiece. The transition region can be, for example, the region between a curved segment and a straight segment and between a straight segment and another type of segment. The division of the workpiece contour to be cut into segments can be defined according to the cutting plan.

[0064] According to another preferred embodiment, the assignment of a type of dynamic laser beam shape to a specific type of cutting segment depends on the predicted contour error within the segment, where the predicted contour error is provided by a contour prediction algorithm. The contour prediction algorithm takes into account the inertia of the laser cutting head. In cases where "overshoot" (movement, deviation from the target contour based on inertia) is suspected, which may lead to contour distortion or incorrectness, the "overshoot" can be compensated for by correspondingly adjusting the dynamic beam shaping module.

[0065] According to another preferred embodiment, the (dynamic) laser beam shape is implemented as a Lissajous shape. A set of Lissajous figures can be stored in a shape storage device. As an advantage of providing a specific shape storage device that can be accessed independently, the set of shapes can be continuously modified and extended even during the application of the method for determining the dynamic laser beam shape.

[0066] According to another preferred embodiment, the method includes receiving cutting requirements via a user interface, the cutting requirements being selected from the group consisting of: burr height, roughness, feed rate, cut width, energy consumption, gas consumption, process stability, contour error, tilt angle / rectangularity, flatness of the cutting edge, and / or heat affected zone. Control instructions are generated by considering the received cutting requirements.

[0067] According to another preferred embodiment, the (dynamic) laser beam shape is determined based on user input data. The user input data can be or can include quality requirements, material requirements, and / or other process conditions. Generally, the assignment step is performed automatically. However, a semi-automatic mode can also be applied in another preferred embodiment of the present invention to verify the automatic assignment. In this case, the system makes a computer-generated (automatic) suggestion, for example, one Lissajous figure for each segment, and displays it to the user, who can verify the automatic decision or can "veto" the suggestion and can manually select another figure or shape.

[0068] According to another preferred embodiment, each dynamic laser beam shape in the dynamic laser beam shapes includes a geometric data set indicating the geometry and / or a time-related data set indicating how the geometry is to be performed, in particular indicating speed and / or acceleration and / or jerk.

[0069] Alternatively additionally, the calculation assignment is based on or takes into account the geometric data set and / or the time-related data set. Considering the above-mentioned aspects, the efficiency can be improved by providing a preselection of assignment candidates, such that the (final) calculation can be performed on a smaller data set.

[0070] Alternatively or additionally, the calculation assignment is performed in a cutting speed-related manner. This can improve the quality and can prevent a high speed difference between consecutive cutting segments.

[0071] So far, the present invention has been described with respect to the claimed method. Features, advantages or alternative embodiments herein may be assigned to other claimed objects (e.g., a computer program or a control unit), and vice versa. In other words, features described or claimed in the context of a method may be used to improve an apparatus or device, and vice versa. In this case, the functional features of the method are respectively embodied by the structural units of the apparatus or device or system, and vice versa. Generally, in computer science, software implementations and corresponding hardware implementations (e.g., as an embedded system) are equivalent. Thus, for example, a method step for "receiving" data (e.g., a cutting plan) can be performed using an interface for receiving data and corresponding instructions. To avoid redundancy, although the device can also be used in alternative embodiments described with reference to the method, these embodiments are not explicitly described for the device anymore.

[0072] According to another aspect, the present invention relates to a control unit configured to perform the method for determining a dynamic laser beam shape for controlling a laser cutting machine as described above, the laser cutting machine being provided with at least one optical module (in particular a dynamic beam shaping module) for changing the shape of the laser beam, the control unit having:

[0073] - a cutting plan interface configured to receive a to-be-processed cutting plan for cutting out a part of a workpiece, wherein each part is defined by a cutting profile, wherein the cutting profile includes a set of cutting segments (or can be defined by a set of cutting segments), and wherein each workpiece is characterized by a characteristic index selected from the group consisting of a material index and / or a thickness index;

[0074] - an interface to a shape storage device having a stored set of (different) dynamic laser beam shapes (in particular more than two (2));

[0075] - a processor configured to iteratively access the assignment storage device for each of the cutting segments in the cutting profile to automatically assign (or calculate the assignment of) a specific dynamic laser beam shape to each of the cutting segments of the cutting profile according to the received cutting plan, wherein the automatic assignment is specific to the respective cutting segment and based on the characteristic index of the workpiece;

[0076] - and wherein the processor is further configured to provide control instructions via or at an output interface for controlling the laser cutting machine to execute the received cutting plan by specifically applying the determined dynamic laser beam shape for each cutting segment.

[0077] In another aspect, the present invention relates to a computer program comprising computer program code which, when executed by a processor, causes the control unit to perform the steps of the method as described above.

[0078] In yet another aspect, the present invention relates to a computer-readable storage medium in which the computer program as mentioned above is stored.

[0079] In the following, definitions of terms used in the present application are given.

[0080] The workpiece will be understood as the material to be cut. The workpiece can be a flat workpiece, such as a metal plate with different properties, or the workpiece can be a tubular workpiece or other closed or open 3D profiles with different cross-sections (rectangular or square), such as a U-shaped profile or a V-shaped profile. The workpiece can be a metal workpiece. The workpiece can be, for example, metal plates of different types and / or with different thicknesses. Generally, the cutting plane defines the part that needs to be cut out from the workpiece. The part to be cut out from the workpiece can have a specific profile, and the specific profile may be different between different parts. For example, a first set of circular parts needs to be completely cut out from the workpiece, and a second set of rectangular parts needs to be cut out. Each part to be cut out can be defined by a cutting profile. The cutting profile can include a set or multiple cutting segments. For example, a rectangular part can include a first segment as a first straight line segment, followed by a second segment as a corner segment (radius), followed by a third segment as a second straight line segment, followed by a fourth segment as a corner segment, followed by a fifth segment as a first straight line segment, and so on. The cutting profile defines the shape of the part from a top view.

[0081] A set of cutting segments can be sorted in a queue. The order is defined by the movement direction of the cutting head. The queue is an ordered list of cutting segments, and the cutting segments are cut one by one.

[0082] The cutting profile can include a set of segments and at least include a single segment. For example, if a circular part needs to be cut out from the workpiece, the circular part generally includes a single (circular) segment. If a more complex profile needs to be cut, the profile can be divided into a set of segments. For example, if a square profile needs to be cut, the square profile can be divided into four equal straight line segments representing the four sides of the square and four corner or curve segments representing the four corners of the square.

[0083] The segmentation of a portion of a contour into a set of segments can be automatically performed by a segmentation algorithm. In this regard, it must be noted that the segmentation algorithm need not be interpreted as a "normal" segmentation algorithm in image processing, such as a segmentation algorithm for segmenting organs in an anatomical body part known from medical image processing. The segmentation algorithm here involves subdividing a two-dimensional contour into different portions or parts. In a very simple and preferred embodiment, the segmentation algorithm can simply access the cutting plan and process the data therein to determine the segmentation of the contour to be cut into segments in a sequence. Alternatively or additionally, in view of the inertia of the laser cutting head arrangement when moving along the contour, the segmentation algorithm can take into account the differences between the individual contour portions for the laser cutting process. Thus, the segmentation algorithm takes into account the inertia of the laser cutting head arrangement. Alternatively or additionally, the segmentation algorithm can also take into account other laser cutting characteristics. For example, the segmentation algorithm can take into account the energy input per unit length, the energy distribution within the workpiece, and / or the dynamic capabilities of the machine axes. The segmentation algorithm can be based on artificial intelligence (AI), in particular on a machine learning module that has been trained to find the segmentation for a contour by taking into account laser cutting machine characteristics (inertia, mass, rate, speed, etc.).

[0084] As described above, straight portions can be cut at a higher speed than corner portions. Thus, such a contour can be segmented into straight segments and corner segments. Alternatively or additionally, the step of segmentation (i.e., the segmentation) can be manually performed via a user input on a human-machine interface.

[0085] The cutting segments are typically provided in an ordered sequence or queue of cutting segments. Each cutting segment in the ordered queue of cutting segments undergoes an assignment of a particular dynamic laser beam shape.

[0086] The cutting segments are sorted, for example, the first cutting segment is before the second cutting segment, the second cutting segment is before the third cutting segment, etc. The cutting segments and / or the queue of cutting segments can be defined in the cutting plan or can be calculated based on the data within the cutting plan. The queue is in particular based on the cutting direction of the laser cutting head moving on the surface of the workpiece to be cut. Depending on the cutting plan to be executed, the cutting segments represent geometrically different types of cuts. The cutting segments can be, for example, straight lines, curves with different radii (parametric curves), which can represent corners, circles or circular segments, penetrations, introductions, exits, and / or engravings.

[0087] Typically, different segments represent or relate to different types or modes of laser cutting. Thus, the first segment can be cut with a first set of cutting parameters (e.g., a first speed), while the second segment can be cut with a second set of cutting parameters (e.g., a second speed of the cutting head). The speeds and rates within this document can be interpreted to refer to the same content / meaning. A segment can be a straight line, a curve with different radii, a corner, a penetration, an introduction, an end cut, etc. A contour can be divided or segmented into different types of segments and / or segments of the same type. Thus, the contour to be cut includes a plurality of such equal segments and / or different segments. For example, if a rectangle is to be cut into a contour, the following segments can be defined: 2 long straight segments, 2 short straight segments, 4 (four) 90° corner segments. In the case of a circle to be cut, each circular contour can have only a single segment.

[0088] A segment is related to the contour to be cut out of the workpiece. Thus, the segment is also (indirectly) related to the workpiece having workpiece characteristic indicators (i.e., the material indicator and / or thickness indicator and / or other characteristics of the workpiece). Thus, the segment is related to the characteristic indicators (of its workpiece). Thus, for example, the first segment of the first contour of the first workpiece can be different from the first segment of the first contour of the second workpiece. Even more, the first segment of the first contour of the first workpiece can be different from the first segment of the second contour of the first workpiece.

[0089] Thus, in another advantageous embodiment, the processor performs the step of automatically calculating the assignment by accessing a trained model, in particular a neural network model. The neural network can be, for example, a convolutional neural network (CNN).

[0090] The neural network has been trained to iteratively determine a dynamic laser beam shape for a specific type of segment for all cutting segments. For example, for segment type 1, the neural network or machine learning model is trained to provide a first dynamic laser beam shape, and for segment type 2 to provide another second dynamic laser beam shape, and for segment type 3, again to provide the first dynamic laser beam shape, and so on.

[0091] The trained neural network model can be stored, for example, on a cloud-based server that exchanges data with the control unit. Alternatively or additionally, the neural network model can be stored locally on the storage device (such as its controller) of the laser machine. (Deep) machine learning algorithms are data- and computation-intensive and are therefore preferably computed on a graphics processing unit (GPU) or tensor processing unit (TPU) or a network of processors. Each layer of the neural network can be computed on a powerful, massively parallel processor (especially a multi-core or multi-core processor). The computing unit is preferably designed as or includes a graphics card or the other hardware modules mentioned above.

[0092] The step of automatically "iteratively calculating the assignment of a dynamic laser beam shape from a set of dynamic beam shapes for each cutting segment for all parts to be cut out from the workpiece (segment-beam shape)" (hereinafter referred to as "the step of calculating the assignment") can be performed with the aid of artificial intelligence and / or machine learning algorithms, based on a trained model, in particular a neural network model. The trained model provides or determines a dynamic laser beam shape (output) for a specific type of segment (e.g., straight line, curve with different diameters) (input).

[0093] Alternatively or additionally, the machine learning model is trained to recognize the assignment between the segment (type) and the dynamic laser beam shape, in particular to recognize which properties / features of the cutting segment, in particular which chemical, spatial, and / or temporal properties, are relevant to the determination of the dynamic laser beam shape, especially in the absence of prior analysis of properties (or feature - feature extraction).

[0094] The step of automatically "calculating the assignment" can be implemented by a featureless extractor (or featureless) process, in particular an algorithm. Alternatively or additionally, the step of automatically "calculating the assignment" can be implemented by a so-called end-to-end algorithm. In this context, "end-to-end" means that the raw data (i.e., the acquired contour segments) can be used without substantial preprocessing, and in particular, without manual determination of the features in the cutting segments and their processing, and then the raw data is further processed (e.g., classified) using machine learning algorithms (hereinafter also simply referred to as ML algorithms) to obtain the result. In this context, "without substantial preprocessing" means except for edge preprocessing such as histogram equalization, image depth reduction, and / or region of interest (ROI) cropping.

[0095] In particular, end-to-end algorithms or methods do not require separate preprocessing of the raw data to extract "features" that are important for learning. Compared to classical ML methods with prior feature extraction, in the possible solutions presented herein, not only is the classifier trained by the algorithm, but preferably also the feature extractor is trained in the same step. This means that the algorithm independently computes or learns a representation from the input data (cut segments), and thus also computes or learns "features". In order to identify such relationships, the algorithm must independently find the best representation of the input data to classify it. The fact that characteristic values ("features") do not have to be extracted in the preferred embodiments of the present invention is advantageous in several respects. First, the work of algorithm development can be simplified because there is no need to detect, determine, and extract important features. In addition, it is advantageous that in the development of "featureless" algorithms (end-to-end algorithms), there is no risk that the most important features containing the most information may be overlooked. Finally, extremely important information is often also contained in the representation of very complex, superimposed, or difficult-to-understand signals, images, or cut sequence features, which makes optimal feature analysis difficult. Therefore, it is not surprising that the deep learning method implemented herein without any feature extraction is superior to the method based on a feature extractor.

[0096] The neural network can be trained with a training algorithm based on annotated or partially annotated training data, which includes an evaluation of the cutting results in the case of applying a dynamic laser beam shape to the corresponding segments. The training algorithm can be a supervised learning method or a semi-supervised learning method. The training algorithm can be based on historical data. A reinforcement learning method can also be used to update or adjust the model. Reinforcement learning makes it possible to find a solution to this complex problem without (prior) knowledge and initial data about the laser cutting process and transition phases. In addition, reinforcement learning eliminates the need for time-consuming collection and processing of training data.

[0097] In order to map the dynamic laser beam shape to a certain segment type in the sequence of cut segments, in this case, a CNN or a deep neural network (DNN) can be applied to learn time-related features. The so-called gated recurrent unit (GRU) or long short-term memory network (LSTM) can in particular be applied in combination with a CNN.

[0098] The laser cutting machine is configured to apply a laser beam to a workpiece to thermally separate the workpiece material by laser radiation. The workpiece can be a tubular workpiece or a flat workpiece with a cutting length of up to 12 meters and a width of 2 to 3 meters. The laser cutting machine can be configured for 3D metal sheets, such as bent parts, cast parts, or printed (additive manufacturing) parts or welded parts.

[0099] The laser cutting machine is equipped with at least one optical module. The optical module can be implemented as a dynamic beam shaping module. Alternatively or additionally, the optical module can include a mirror-based system for dynamically changing the laser beam shape during cutting and thus dynamically changing the laser beam shape during the movement of the laser cutting head over the workpiece according to the cutting plan and / or the cutting profile to be cut.

[0100] As previously mentioned, the optical module can be implemented as a dynamic laser beam shaping module. The optical module can include, for example, laser scanner optics (e.g., as described in WO2019145536A1) or lens optics actuated or oscillated in the x and y directions perpendicular to the laser beam axis (e.g., as described in WO 2019 / 145536 A1). The technical purpose of the dynamic laser beam shaping (module) DBSM is to improve quality and / or performance. In some aspects, DBSM can be used to provide energy over a larger area on the workpiece. In particular, for a wider cutting kerf width (which makes subsequent automation / categorization of parts easier), a dynamic laser beam shape with a higher amplitude and / or a different focus position can be selected and used to enlarge the spot size. Additionally, due to the shorter interaction time with the high-power laser machine (less heat accumulation), DBSM is used to provide less damaged material properties, which is the main improvement. Regarding quality improvement: Among the standard cutting parameters (focus position, laser power, gas pressure,...), using an additional dynamic laser beam shape allows the corresponding selection of cutting and dynamic laser beam shaping parameters to achieve better quality and / or performance with a higher-dimensional parameter space. Additionally, due to the ability to change the dynamic laser beam shape, the overall quality (also in corners and at transitions) and / or performance can be improved, and no compromise settings are required for straight lines and corners.

[0101] Preferably, the dynamic laser beam shape is adjusted during a transition phase (e.g., cutting speed-related or rate-related). The dynamic beam shaping can be applied with different beam shaping frequencies. The beam shaping frequency can be in the range between 100 Hz and one or more megahertz, and preferably in the range between 100 Hz and 900 kHz or between 100 Hz and a few hundred kilohertz. Preferably, the beam shaping frequency can vary in each of the above-mentioned directions (i.e., in the X direction and the Y direction, and even in the Z direction). It must be noted that in a preferred embodiment, the beam shaping frequency can be set differently in each of these directions, such that, for example, a first beam shaping frequency of 200 Hz can be applied in the X direction, and a second beam shaping frequency of 900 kHz can be used in the Y direction. The beam shaping frequency settings for each direction can be set and configured independently of each other on the user interface.

[0102] Alternatively, or in addition to changing the beam shaping frequency as mentioned above, the laser pulse frequency can also be changed. The laser pulse frequency can vary between 0 kHz and 5 kHz. Generally, the laser pulse frequency is only used for specific applications, such as engraving, pulsed cutting (e.g., when penetrating, bending, turning), such that the energy input can be adjusted by pulse modulation. In some aspects, changing the pulse frequency of the laser beam can be used to reduce the energy to be provided to the workpiece. The setting of the laser pulse frequency can be configured on a man-machine or user interface HMI. The HMI can provide the settings for configuring the beam shaping frequency for each direction either in a combined form or separately, and additionally provide the settings for configuring the laser pulse frequency.

[0103] A cutting plan is a digital representation of the part to be cut out from the workpiece, the movement direction, and the control instructions for executing the cutting plan, such as a sequence defining which part of the cutting plan is to be cut next and / or which segment within a contour is to be cut next. The cutting plan is typically represented in a digital file format, such as in an XML-based format.

[0104] The shape storage device can be implemented as a memory structure or a database. The shape storage device is configured to store different dynamic laser beam shapes. Dynamic laser beam shapes can, for example, include different Lissajous figures. Alternatively or additionally, other patterns and / or shapes can also be stored and accessed.

[0105] Automatically calculating the assignment of a specific dynamic laser beam shape for a specific cutting segment means automatically assigning the dynamic laser beam shape to a specific type of segment. Preferably, the determination or calculation of the assignment can be performed algorithmically by means of applying an assignment algorithm. In a simple form, the assignment algorithm can be implemented as a lookup table in a predefined data structure, the lookup table having a pre-computed assignment between specific dynamic laser beam shapes and physical properties selected from a list including: material type, material thickness, laser power, feed rate of a specific segment, and objectives (e.g., performance, quality, burr height, roughness, robustness, etc.). For example, a setting with physical properties of material type M, thickness T, and laser power P is associated with a dynamic laser beam shape S.

[0106] Alternatively or additionally, the allocation algorithm accesses a rule base for a specific segment to find the corresponding associated dynamic laser beam shape. A set of rules is stored in the rule base. The rule base can be implemented in a shape storage device or can be executed by accessing a shape storage device. For example, the rules can represent the following settings: for a cut with a wide cut width, (individually) lateral oscillation perpendicular to the cutting direction is appropriate. For example: in order to increase the cutting speed, (individually) longitudinal oscillation in the cutting direction is most suitable. Thus, a set of configurable rules is provided in the rule base. By applying these rules, the allocation algorithm is configured to determine the allocation.

[0107] Alternatively or additionally, the allocation algorithm can be configured to, for example, map or allocate a specific dynamic laser beam shape to a certain (workpiece contour) segment, also taking into account the previous and / or subsequent segments and / or their dynamic laser beam shapes, respectively. This helps to improve the cutting quality by providing a smooth transition between the shapes of two consecutive segments.

[0108] In a more complex embodiment, the allocation algorithm can be based on a trained neural network that has been trained to find the optimal dynamic laser beam shape for each type of segment as described above.

[0109] Control instructions are used to control the laser cutting machine by applying the determined dynamic laser beam shape for a specific segment.

[0110] The control unit and / or processor is an electronic module, or can be a software module implemented in a hardware module with a processor, or can be a hardware module (such as an FPGA; ASIC as described below). In the context of the present invention, a "processor" can be understood to mean, for example, a machine or an electronic circuit. In particular, the processor can be a central processing unit (CPU), a microprocessor or a microcontroller, such as an application-specific integrated circuit or a digital signal processor, possibly in combination with a memory unit for storing program instructions, etc. The processor can also be, for example, an IC (integrated circuit), in particular an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit), or, for example, a multi-chip module (such as a 2.5D or 3D multi-chip module, in which several so-called dies are directly or via an interposer connected to each other), or a DSP (digital signal processor) or a GPU (graphics processing unit). The processor can also be a virtualized processor, a virtual machine or a soft CPU. For example, the processor can also be a programmable processor, such as an FPGA or an ASIC, which is equipped with configuration steps for executing the method according to the present invention, or is configured in such a way that the programmable processor implements the features of the method, components, modules or other aspects and / or partial aspects of the present invention. The allocation tool and / or the conversion tool can be part of a controller for controlling a laser cutting machine, or can be a separate module data-connected to the controller.

[0111] The order of the steps of the method according to the present invention described in this specification does not necessarily reflect the chronological order according to which the steps are executed. For example, the steps of providing the allocation tool and providing the conversion tool can be executed in a different order.

[0112] Based on the following description and embodiments, which will be described in more detail in the context of the drawings, the above-mentioned characteristics, features and advantages of the present invention and the ways of realizing them become clearer and easier to understand. The following description does not limit the present invention to the included embodiments. In different drawings, the same components or parts can be labeled with the same reference numerals. Generally, the drawings are not to scale. Description of the Drawings

[0113] Figure 1 is a schematic representation of a control unit for determining a dynamic laser beam shape for a set of cutting segment types;

[0114] Figure 2 is a schematic representation of a training algorithm for training the assignment of a dynamic laser beam shape to a set of cutting segment types;

[0115] Figure 3 is a flowchart of a method for determining a dynamic laser beam shape for each segment of a cutting profile according to a preferred embodiment of the present invention;

[0116] Figure 4 are examples of cutting profiles with segments of different types;

[0117] Figure 5 is another example of a cutting profile with its segments. DETAILED DESCRIPTION

[0118] Figure 1 shows a schematic representation of a control unit 100 configured to perform a method for determining a dynamic laser beam shape for laser cutting by means of a laser cutting machine L.

[0119] The laser machine L has a cutting head CH which may include at least one optical module. The optical module may be implemented as a dynamic laser beam shaping module DBSM or another type of optical module for dynamically changing the laser beam shape. Furthermore, the laser machine is equipped with a controller PLC. The control unit 100 may be provided as a separate entity for data exchange with the laser cutting machine L and in particular with the controller PLC of the laser cutting machine L. Alternatively or additionally, the control unit 100 may be implemented directly on the controller PLC of the laser cutting machine L.

[0120] The control unit 100 includes a cutting plan interface 101 configured to receive a pending cutting plan for cutting out parts of a workpiece on the laser cutting machine L. According to the cutting plan, each part is defined by a cutting profile. The cutting profile itself is segmented or subdivided into a set of different types of cutting segments, such as straight segments, curved segments with a first radius and another curved segment with a second radius, corner segments, lead-in segments, etc. Typically, since the parts are to be cut out from the workpiece, the parts share the characteristics of the workpiece (before cutting, the parts and the workpiece are not separated). The workpiece has characteristics. In particular, the workpiece is made of a certain type of material and has a certain type of thickness. Furthermore, other characteristics may be processed. The characteristics are encoded by characteristic indicators. Thus, the type of material is encoded in a material indicator and / or the thickness is encoded in a thickness indicator.

[0121] The control unit 100 further includes an interface 102 interfacing with a shape storage device ShS. The shape storage device ShS is configured to store a set of different dynamic laser beam shapes, in particular more than two such different dynamic laser beam shapes.

[0122] The control unit 100 also includes a processor P, which is configured to iteratively and automatically calculate, for each cutting segment of all the parts to be cut out from a workpiece, the assignment of a specific dynamic laser beam shape from a set of dynamic laser beam shapes stored in the shape storage device ShS. The calculation of the assignment is based on the characteristic metrics of the relevant workpiece. The characteristic metrics can be derived from the cutting plan. The automatic calculation of the assignment "cutting segment type - selected dynamic laser beam shape" is performed for each type of cutting segment. In other words, the first type of cutting segment (e.g., a straight segment) is assigned to the first dynamic laser beam shape, and the second type of cutting segment (e.g., a corner segment) is assigned to the second dynamic laser beam shape, and the third type of cutting segment (e.g., an introduction segment) is assigned again to the first dynamic laser beam shape, and so on. The relationship between the segment type and the dynamic laser beam shape type can be an n:m relationship. The assignment can be provided by executing an assignment algorithm. The assignment algorithm can access a rule database R-DB. The rule database is configured to store rules for determining the relationship between the segment type and the dynamic laser beam shape type. Alternatively or additionally, the assignment algorithm can apply machine learning algorithms and / or neural networks.

[0123] The processor P is configured to provide control instructions CI on an output interface 103, which connects the control unit 100 to the laser cutting machine L. The provided control instructions CI are configured to control the laser cutting machine L to execute the received cutting plan by specifically applying the determined dynamic laser beam shape for each cutting segment or each type of cutting segment.

[0124] The laser cutting machine is configured to machine / cut materials and thicknesses such as the following:

[0125] - Various steel alloys, such as steel, mild steel, 0.8 mm to 30 mm, and thinner and thicker materials can also be cut;

[0126] - Chrome steel (also known as stainless steel), 0.8 mm to 30 mm, and thinner and thicker materials can also be cut;

[0127] - Aluminum, 0.8 mm to 30 mm, and thinner and thicker materials can also be cut;

[0128] - Non-ferrous metals, such as copper, 0.8 mm to 15 mm, or brass, 0.8 mm to 15 mm, both of which also have thinner or thicker variants.

[0129] The material characteristics can include:

[0130] - Physical properties, in particular including the material composition (Fe, C, Si, S,...), can be provided via a material certificate, which can be encoded as a digital code (barcode, QR code) and provided together with the workpiece or with the cutting plan;

[0131] - Melting temperature;

[0132] - Surface tension;

[0133] - Viscosity;

[0134] - Absorption coefficient;

[0135] - Heat capacity;

[0136] - Thermal conductivity;

[0137] - Solid density, melt density;

[0138] - Actual temperature of the metal sheet;

[0139] - Surface quality (roughness of the material);

[0140] - Crystal structure and / or

[0141] - Rolling direction.

[0142] Figure 2Shows modules or methods for training an allocation algorithm. It can be seen that training is performed on a processing entity, which can be different from the processor and / or control unit. Preferably, the processing entity on which training is performed is a separate unit. The processing entity includes an input interface 21, which is configured to receive a tuple that includes a specific dynamic laser beam shape i (represented by its metrics, such as oscillation frequencies in different directions) and a specific type of segment a. Both the dynamic laser beam shape and the segment type are provided as a result dataset of the allocation algorithm. In other words, the allocation algorithm matches the dynamic laser beam shape i with the type of segment a. The result of the allocation algorithm is forwarded to the processing entity via the output interface 23. After performing the cutting of a specific segment with the determined specific dynamic laser beam shape allocated by the allocation algorithm, an evaluation of the cutting result is performed. The evaluation can be, for example, a quality evaluation. The evaluation is encoded in an evaluation dataset. The evaluation dataset is provided to the processing entity by means of the input interface 22. The processing entity is configured to evaluate the evaluation dataset and optionally to compare the evaluation dataset with a reference evaluation dataset to provide an evaluation of the allocation via the output interface 23. The evaluation of the allocation relates to the received specific dynamic laser beam shape i and the received specific segment a. The training dataset and / or weights can be fed back to the allocation algorithm to train and / or calibrate the training dataset and / or weights. Generally, if the allocation between the dynamic laser beam shape and the segment type has a positive evaluation of the allocation, the allocation between the dynamic laser beam shape and the segment type will be rewarded, otherwise if the allocation between the dynamic laser beam shape and the segment type has a negative evaluation of the allocation, the allocation between the dynamic laser beam shape and the segment type will be penalized.

[0143] Figure 3 Is a flowchart of a method for determining a segment-specific dynamic laser beam shape for a laser cutting profile according to a cutting plan according to a preferred embodiment of the present invention.

[0144] After the method starts, in step S1, a cutting plan is received. The cutting plan encodes the profile of the part to be cut out from the workpiece. The cutting plan also includes workpiece characteristics that indicate the type of material and / or the thickness of the material. The characteristics of the workpiece are the same as the characteristics of the part to be cut out from the workpiece.

[0145] In step S2, a cutting segment from a set of cutting segments received together with the cutting plan in the previous step is used to access the shape storage device ShS. A set of dynamic laser beam shapes, in particular more than two and preferably a certain amount of dynamic laser beam shapes, are stored in the shape storage device. In principle, the frequencies and / or amplitudes in the X / Y / Z directions can be arbitrarily selected within the adjustment range. For pierce-in, curved segments, and straight segments, this amount can be at least greater than 3, and up to several hundred or one thousand dynamic laser beam shapes, and in particular in the range between 3 and 100.

[0146] Step S3 involves iteratively and automatically calculating, for each cutting segment of all the parts to be cut out from the workpiece, the assignment of a dynamic laser beam shape from the set of dynamic laser beam shapes accessed in the shape storage device ShS. The calculation of the assignment S3 is based on the characteristic indicators of the workpiece. The calculation of the assignment is specific to the corresponding type of cutting segment.

[0147] Step S4 involves providing control instructions CI for controlling the laser cutting machine L to execute the received cutting plan by specifically applying the determined dynamic laser beam shape for each cutting segment. After this, the method can be repeated or can be ended.

[0148] Figure 4 is an example of a cutting profile with different types of segments. As can be seen in the example shown in Figure 4 the cutting profile includes a part of a circular element with a circular cutout and a small circular cutting segment above the cutout and in the upper right part of the circular cutting profile. The transition zones and segments are depicted by the left hatched pattern in Figure 4

[0149] The reference numeral 81a (on the right hand side of Figure 4 ) depicts the "pierce-in" cutting segment for the outer circular cutting profile. The reference numeral 81a2 represents the dynamic laser beam shaping transition zone from segment 81a to segment 82a. The latter segment 82a represents the "linear introduction" cutting segment. The reference numeral 82a3 depicts the dynamic laser beam shaping transition zone between segment 82a and segment 83. The latter segment 83 represents the "straight line" cutting segment, where the laser cutting head may move faster in the "straight line" cutting segment than in the corner segments. The reference numeral 834 represents the dynamic laser beam shaping transition zone or phase between segment 83 and segment 84. Segment 84 is the "right turn corner" cutting segment. The reference numeral 845 represents the dynamic laser beam shaping transition zone or phase between segment 84 and segment 85. Segment 85 refers to another straight line cutting segment. The reference numeral 856 represents the dynamic laser beam shaping transition zone or phase between segment 85 and segment 86. Segment 86 represents the "left turn corner" cutting segment.

[0150] As can be seen in Figure 4 the profile also includes inFigure 4 Another circular cutting segment denoted by reference numeral 87 in the figure. This circular contour has Figure 4 a "penetrating" cutting segment denoted by reference numeral 81b in the figure. The dynamic laser beam shaping transition region between segment 81b and segment 82b is Figure 4 denoted by reference numeral 81b2b in the figure.

[0151] Figure 5 is another example of a cutting contour having different types of segments. In this example, the contour includes a straight segment 51. Moving in a clockwise manner, the next segment in the contour is segment 52, which is a corner segment. Then, segment 53 is another straight segment, followed by corner segment 54, followed by another corner segment 55 with another radius. Segment 56 is a straight segment, followed by Figure 5 a series of corner segments or radius segments denoted by reference numerals 56-1, 56-2, 56-3, and 56-4 in the figure. As can be seen in the figure, the radii between the different 56 segments are different. After these radius segments 56-1 to 56-4, segment 57 is provided, followed by different radius segments 58, 59, and then straight segment 60. Subsequently, the laser cutting head H must move to corner segment 61 to close the closed contour structure again.

[0152] In the absence of an explicit description, the various embodiments or their various aspects and features described with respect to the figures may be combined with or exchanged for one another without limiting or broadening the scope of the described invention, provided that such combination or exchange is meaningful and within the meaning of the present invention. In any applicable case, the advantages described with respect to a particular embodiment of the present invention or with respect to a particular figure are also advantages of other embodiments of the present invention.

Claims

1. A computer-implemented method for determining a dynamic laser beam shape for laser cutting by means of a laser cutting machine (L), the laser cutting machine comprising at least one optical module for dynamically changing the shape of the laser beam, the method comprising the following method steps: - Receiving (S1) a cutting plan for a part to be processed for cutting out a workpiece, wherein, each part is defined by a cutting profile, the cutting profile comprising a set of cutting segments, wherein each workpiece is characterized by a characteristic index selected from the group comprising a material index and a thickness index; - Providing (S2) a shape storage device (ShS), wherein the shape storage device is configured to store a set of dynamic laser beam shapes; - By a processor (P): accessing the shape storage device (ShS) to automatically iteratively calculate (S3) the assignment of a dynamic laser beam shape from the set of dynamic laser beam shapes for each cutting segment of all the parts to be cut out from the workpiece, wherein the assignment is segment-specific such that different dynamic laser beam shapes will be assigned to different types of segments, and wherein the type of cutting segment is selected from the group comprising: straight line; circle or circular segment with a configurable specific radius; corner with a configurable angle; parametric curve; penetration; introduction; lead-out; and engraving, and wherein calculating the assignment (S3) is based on the characteristic index of the workpiece and is specific to the respective cutting segment; - By the processor (P): providing (S4) control instructions (CI) on an output interface (103) for controlling the laser cutting machine (L) to execute the received cutting plan by specifically applying the determined dynamic laser beam shape to each cutting segment.

2. The method according to claim 1, wherein the shape storage device is configured to store more than two dynamic laser beam shapes.

3. The method according to claim 1, wherein, calculating the assignment is performed by a trained model that takes a specific type of segment as input and a specific dynamic laser beam shape as output.

4. The method according to claim 3, wherein calculating the assignment is performed by a neural network model.

5. The method according to claim 3, wherein, the model is trained with training data comprising: - Selected dynamic laser beam shapes for test cuts; - The characteristic index of the workpiece of the test cut; - The type of cutting segment of the test cut; - Selected dynamic laser beam shapes for test cutting segments for an evaluation data set having an evaluation data label.

6. The method according to claim 5, wherein, The evaluation data set includes configurable shares for setting common different evaluation criteria, and the different evaluation criteria include quality evaluation, demand evaluation, performance evaluation, energy consumption evaluation, process stability evaluation, burr height evaluation, roughness evaluation, feed rate evaluation, cut width evaluation, gas consumption evaluation, contour error evaluation, tilt angle / rectangularity evaluation, flatness cutting edge evaluation, and heat affected zone evaluation, and wherein the different evaluation criteria are interdependent, and the interdependence is adjusted on the user interface selection button set on the human-machine interface HMI.

7. The method according to any one of claims 3 to 6, in, The model is trained by performing the following steps: - preselecting a dynamic laser beam shape from the set of dynamic laser beam shapes; -Perform a cutting segment-specific test cut with a preselected dynamic laser beam shape; - performing an evaluation of the results of the test cuts by providing an evaluation data set for each test cut; - Adjust the weights of the model.

8. The method according to claim 7, in, Adjusting the weights of the model optimizes the objective function for the evaluation data set.

9. The method according to claim 5 or 6, in, The evaluation data set is provided one or both of manually by means of using a user interface and automatically by a sensory automatic evaluation unit.

10. The method according to any one of claims 1 to 6, in, The automatically determined dynamic laser beam shape for each of the cutting segments is determined specifically to the type of cutting machine.

11. The method according to any one of claims 1 to 6, in, In the set of dynamic laser beam shapes, the dynamic laser beam shape is dynamically changed by generating a focus oscillation shape by spatiotemporal distribution of laser energy on one or both of the material surface and the focal plane with respect to one or more of the following: - frequency in the X and Y directions; - amplitude in the X and Y directions; and - Phase shift in the Y direction compared to the X direction.

12. The method according to claim 11, in, The frequency also includes the frequency in the Z direction.

13. The method according to claim 11, in, The amplitude also includes the amplitude in the Z direction.

14. The method according to claim 11, in, The phase shift also includes a phase shift in the Z direction compared to the X direction.

15. The method according to any one of claims 1 to 6, in, The transition region is determined by one or more of: a linear transition function, a non-linear transition function, a logarithmic transition function, and other transition functions.

16. The method according to any one of claims 1 to 6, in, The dynamic laser beam shape is implemented as a Lissajous shape.

17. The method according to any one of claims 1 to 6, in, The method comprises: -Receiving a cutting requirement via a user interface, the cutting requirement being selected from the group comprising: burr height, roughness, feed rate, cut width, energy consumption, gas consumption, process stability, profile error, tilt angle / rectangularity, flatness of the cutting edge, and heat affected zone; -Wherein, the control instruction (CI) is generated by considering the received cutting requirement.

18. The method according to any one of claims 1 to 6, wherein, The shape of the laser beam is determined based on user input data.

19. The method according to any one of claims 1 to 6, wherein, Each dynamic laser beam shape in the dynamic laser beam shapes includes a geometric data set indicating a geometry and a time-related data set indicating how to execute the geometry.

20. The method according to claim 19, wherein, The time-related data set indicates one or more of speed, acceleration, and jerk.

21. A control unit (100) for controlling a laser cutting machine (L), the control unit being configured to perform the method for determining a dynamic laser beam shape according to any one of claims 1 to 20, the laser cutting machine being provided with at least one optical module for changing the shape of the laser beam, the control unit having: -A cutting plan interface (101), the cutting plan interface being configured to receive a cutting plan for a part to be processed for cutting out a workpiece, wherein, Each part is defined by a cutting profile, the cutting profile including a set of cutting segments, wherein each workpiece is characterized by a characteristic index selected from the group comprising a material index and a thickness index; -An interface (102) with a shape storage device (ShS) having a stored set of dynamic laser beam shapes; -A processor (P), the processor being configured to automatically iteratively calculate, for all parts to be cut out from the workpiece, an assignment of a dynamic laser beam shape from the set of dynamic laser beam shapes stored in the shape storage device (ShS) for each cutting segment in the cutting segments, wherein the assignment is segment-specific such that different dynamic laser beam shapes will be assigned to different types of segments, and wherein the type of cutting segment is selected from the group comprising: straight line; circle or circular segment with a configurable specific radius; corner with a configurable angle; parametric curve; penetration; introduction; lead-out; and engraving, and wherein the calculation of the assignment (S3) is based on the characteristic index of the workpiece and is specific to the corresponding cutting segment; -And wherein the processor (P) is further configured to provide a control instruction (CI) via an output interface (103) for controlling the laser cutting machine (L) to execute the received cutting plan by specifically applying the determined dynamic laser beam shape to each cutting segment.

22. The control unit according to claim 21, wherein, The laser cutting machine is provided with a dynamic beam shaping module (DBSM).

23. The control unit according to claim 21, wherein, The number of the stored set of dynamic laser beam shapes is more than two.

24. A computer program product comprising computer program code which, when executed by a processor, causes a control unit to perform the steps of the method according to any one of claims 1 to 20.

25. A computer-readable storage medium storing the computer program code of the computer program product according to claim 24.

Images