Computer-implemented method, computer program, data carrier signal and manufacturing system for generating control data relating to at least one manufacturing machine

Abstract

The invention relates to a computer-implemented method (40), to a computer program, to a data carrier signal and to a manufacturing system (201) for generating control data relating to at least one manufacturing machine (1), wherein a plurality of steps (42) to (54) for manufacturing process control are included for this purpose and point cloud data relating to a workpiece (7) are registered (43), wherein the registration (43) of the point cloud data relating to the workpiece (7) is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data that is formed from the difference matrix, wherein the difference matrix is formed from the difference between the point cloud data coordinates and the centre of gravity coordinates of the point cloud.

Description

[0001] Description

[0002] Computer-implemented method, computer program, data carrier signal and

[0003] Manufacturing system for generating control data for at least one manufacturing machine

[0004] The invention relates to a computer-implemented method, a computer program, a data carrier signal, and a manufacturing system for generating control data for at least one manufacturing machine. The present computer-implemented method is not only, according to the classic definition, an invention comprising a computer, a computer network, or another programmable device in which at least one feature is implemented entirely or partially with a computer program, but also, in the age of cloud computing, an invention that additionally and / or alternatively comprises a method for generating control data for at least one manufacturing machine, said method running on at least one processor and / or controlled by at least one processor and / or initiated by at least one processor, wherein the method comprises steps that are executed on the at least one and / or another processor.

[0005] Many industries are striving to integrate the measurement technology required for quality assurance in production into the production line or even into the actual production machines. For example, US Pat. No. 6,969,821 B2 discloses a method or a production machine for the production of turbine blades in which the measurement technology for quality assurance is already integrated into the production machine.

[0006] Furthermore, US 10,220,566 B2, US 10,532,513 B2, US 11,104,064 B2 and US 2021 / 0379823 A1 disclose a measurement technology for an additive manufacturing process (AM) that is integrated into the production machine.

[0007] There are several reasons for this approach. Firstly, the measurement technology, which today is largely based on measurement room infrastructure (controlled temperature, clean / grey room, low vibration, etc.), is expensive. Secondly, the measurement rooms are remote from production, thus complicating workpiece logistics. Furthermore, the remote measurement rooms generate significant latency, meaning that the measurement results obtained there at a later time cannot be used for effective production control. Furthermore, the information from the measurement rooms is not prepared in such a way that it can be used by production in an automated form for process control.

[0008] Therefore, remote measuring rooms are unsuitable for online recording and online control of process states, especially in potentially global manufacturing networks, and cannot support process development or process start-ups for the increasing demand for small-series and unique production.

[0009] Furthermore, the measuring rooms take up large areas that could otherwise be used for production. Highly qualified personnel are also required for the measuring rooms. The level of automation in the use of measuring rooms is also low, thus causing additional personnel costs.

[0010] Accordingly, US 10,180,667 B2 discloses a metrology system integrated into the production machine, in which the measurement results are interpreted by a trained AI. WO 2018 / 204410 also discloses a trained AI in which measurement results from metrology sensors or coordinate measuring machines are processed via a cloud computing network or management system for one or more production machines.

[0011] The fundamental problem with all measuring technologies integrated into the production line is to compare the point cloud recorded from a workpiece with target specifications, usually test plans derived from CAD data. To do this, the recorded point clouds of the workpiece must first be registered, i.e. aligned based on the CAD model, rotated, and translated into the computer or CAD coordinate system for the actual comparison. This is particularly true when the workpieces are moved uncoordinatedly at arbitrary locations and with arbitrary orientations on a conveyor belt past the measurement sensors for point cloud measurement, or when the workpieces are held in different poses within the capture range of the measurement sensors for point cloud measurement, for example by robots.US 11,049,236 B2 discloses a special solution that is particularly suitable for the registration of flat workpieces moving on a conveyor belt.

[0012] Based on the objective described at the beginning, namely a measurement technology integrated into production, it is the object of the present invention, in view of the state of the art, to provide a robust method suitable for all types of workpieces and globally applicable to all possible production machines, which enables a quick target / actual comparison based on test plans or CAD models with point clouds recorded from the workpiece to generate new control commands for a production machine.

[0013] This object is achieved according to the invention by a computer-implemented method for generating control data of at least one production machine, which method runs on at least one processor and / or is controlled by at least one processor and / or is initiated by at least one processor, wherein the method comprises the following steps and at least one of the following steps of the method is carried out on the at least one and / or another processor:

[0014] • Obtaining point cloud data of a workpiece;

[0015] • Registering the point cloud data of the workpiece;

[0016] • Fitting geometric surface and / or form elements to the point cloud data of the workpiece;

[0017] • Obtaining a test plan and / or CAD model of the workpiece and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data of at least one production machine for manufacturing the workpiece;

[0018] • Comparing the test plan and / or CAD model of the workpiece with the fitted surfaces and / or form elements and identifying deviations;

[0019] • Evaluating or assessing the deviations identified in the comparison, if necessary based on specified tolerances of the workpiece and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data; • Generating new target control data and / or new infrastructure or environmental data of the production machine for the continued production of the workpiece or for the production of a new workpiece depending on the evaluation;

[0020] • Transferring the new target control data and / or, if applicable, the new infrastructure or environmental data to the at least one production machine and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine and / or the at least one further production machine as a function of evaluation; wherein the registration of the point cloud data of the workpiece is based on a singular value decomposition (S VD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud.

[0021] According to a further aspect of the present invention, the object is achieved by a computer program according to the invention which comprises instructions which, when the program is executed by at least one processor, cause this at least one processor and / or further processors to carry out the method according to the invention according to one of the embodiments described herein.

[0022] According to another aspect of the present invention, the object is achieved by a data carrier signal according to the invention which transmits the computer program according to the invention in whole or in part according to one of the embodiments described herein.

[0023] According to a further aspect of the present invention, the object is achieved by a manufacturing system according to the invention, which comprises a computer program according to the invention according to one of the embodiments described herein and at least one processor.

[0024] Furthermore, according to another aspect of the present invention, the present invention is achieved by a manufacturing system according to the invention comprising at least one processor and at least one memory, wherein the at least one processor exchanges data with the at least one memory and the manufacturing system is configured to:

[0025] • Obtaining point cloud data of a workpiece;

[0026] • Registering the point cloud data of the workpiece;

[0027] • Fitting geometric surface and / or form elements to the point cloud data of the workpiece;

[0028] • Obtaining a test plan and / or CAD model of the workpiece and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data of at least one production machine for manufacturing the workpiece;

[0029] • Comparing the test plan and / or CAD model of the workpiece with the fitted surfaces and / or form elements and identifying deviations;

[0030] • Evaluate or assess the deviations identified in the comparison, if necessary based on specified tolerances of the workpiece and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data;

[0031] • Generation of new target control data and / or new infrastructure or environmental data of the production machine for the continued production of the workpiece or for the production of a new workpiece depending on the evaluation;

[0032] • Transferring the new target control data and / or, if applicable, the new infrastructure or environmental data to the at least one production machine and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine and / or the at least one further production machine depending on the evaluation; wherein the registration of the point cloud data of the workpiece is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors, a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud. According to the invention, it was recognized that the data obtained from the statistics orThe covariance matrix known from image processing is particularly advantageous for detecting the alignment of point clouds in space of any workpiece, even unknown workpieces, in an unknown position and orientation, when the point clouds of the workpieces are only incomplete, for example due to occlusions or visible shadows. According to the invention, the covariance matrix is ​​formed with respect to a "statistical" expected value in X, Y and Z in relation to the center of gravity of the point cloud in X, Y and Z. The method known from statistics or image processing for forming a covariance matrix and for determining the main axes orThe invention demonstrates superiority over other algorithms in robustly and reliably determining the principal axes of the spatial orientation of a workpiece relative to the measuring device, even when information about the workpiece is only incomplete. Similar to the cognitive abilities of a human being who, for example, can grasp the orientation of a workpiece in the hands of a colleague in less than a second, even if the workpiece is largely obscured, the covariance matrix or the underlying difference matrix can be used to determine the orientation of the three principal axes of a partially obscured workpiece in less than a second using a standard computer used in industry.In this respect, orientation determination using the covariance matrix is ​​superior to other methods for registering point clouds or measurement data. These other methods either fail fundamentally due to the incomplete point cloud or require too much time or computing power due to complex computational operations.

[0033] A good overview of the known methods for registering point clouds is provided by Xiaoshui Huang et al.: “A comprehensive survey on point cloud registration”, published under: arXiv:2103.02690v2 [cs.CV] 5 Mar 2021. An explanation of the singular value decomposition (SVD) of a covariance matrix as a method of multivariate statistics can be found under the keyword “Principal Component Analysis” (PCA) in Wikipedia at: https: / / de.wikipedia.org / wiki / Hauptkomponentenanalyse or in the tutorial by Lindsay I Smith: “A tutorial on Principal Components Analysis” at: https: / / web.archive.Org / web / 20210309115736 / http: / / www.cs.otago.ac.nz / cosc453 / student_tut orials / prmcipal_components.pdf .

[0034] A complete registration of point clouds with a CAD model or with a coordinate system of the CAD or computer model involves two different steps. Firstly, the point cloud must be aligned according to the principal axes of the model or the coordinate axes of the computer, for which the point cloud data must be rotated or panned into the corresponding coordinate system. Secondly, the rotated or panned point cloud data must be offset from the coordinate origin so that its center of gravity or another reference point coincides with either the coordinate origin or a point of the CAD model. The singular value decomposition (S The difference matrix or the determination of the singular vectors of the difference matrix and / or the determination of the eigenvectors of the covariance matrix derived from the difference matrix now ensures the first part of the registration by determining the necessary rotation matrix for aligning the point cloud data according to the coordinate axes of the recliner or the CAD model. Most known registration algorithms, however, attempt to complete both registration steps simultaneously by iteratively minimizing a distance criterion between the point cloud data and the CAD model. This usually requires a complete point cloud of the entire workpiece.

[0035] In one embodiment of the invention, in a further step, scalar products of the point cloud data are calculated with at least two determined singular vectors and / or eigenvectors, and / or at least two coordinates are determined in the X and / or Y and / or Z directions of the coordinates of the point cloud data rotated into the computer or CAD coordinate system. Both equivalent procedures—calculating scalar products and determining the coordinates of the rotated point cloud data—can be accomplished very quickly and efficiently.

[0036] The scalar products formed with the singular vectors or eigenvectors naturally correspond to the scalar products of the point cloud data rotated or tilted into the model coordinate system with the respective unit vectors of the coordinate axes of the model coordinate system of the CAD model or the computer model, and thus to the coordinates of the point cloud data in the computer or CAD coordinate system. The scalar products formed with the singular vectors or eigenvectors thus represent projections of the point cloud data along the singular / eigenvectors or coordinate axes and thus represent a measure of the workpiece's extent. In this respect, a measure of the workpiece's extent can be determined very quickly and efficiently using the scalar products or the coordinates.

[0037] In a further embodiment of the invention, the most probable initial value and / or final value of the expansion of the workpiece in the direction of a singular vector and / or an eigenvector and / or in a coordinate direction is determined with the aid of the scalar products and / or the coordinates by considering a density or frequency distribution of the scalar products in the direction of the corresponding singular vector and / or eigenvector and / or by considering a density or frequency distribution of the coordinates.

[0038] A cluster or high density of scalar products or coordinates can only be caused by an edge surface or an end surface of the workpiece. This is because, on the one hand, all scalar products of the point cloud data of a surface with the direction of the surface normal, or all point cloud data of a surface perpendicular to a coordinate axis, have almost the same value of the scalar product or coordinates. On the other hand, the relative number of such surface points compared to the total number of point cloud data is very high. Therefore, an examination of the density or frequency distribution of the determined scalar products or coordinates is useful for reliably and robustly determining the position of surfaces of the workpiece in the direction of the singular / eigenvectors or in the coordinate direction.

[0039] Those values ​​of the density or frequency distribution of the determined scalar products or coordinates which provide the largest or lowest “amount” of the scalar product in relation to the corresponding singular / eigenvector or which provide the largest or lowest coordinate value, represent the end or the beginning of the workpiece in the direction of the considered singular / eigenvector or in the coordinate direction. This determined “amount” of the scalar product or

[0040] Coordinate value then represents an initial value and / or final value of the extension of the workpiece in the direction of the associated singular / eigenvector or in the coordinate direction.

[0041] In one embodiment of the invention, when considering the density or frequency distribution of the scalar products and / or the coordinates, the evaluation is carried out on the basis of mean values, threshold values, half-widths, threshold widths and / or gradients of the local density or frequency distributions. With the help of such methods for peak evaluation, the most probable initial value and / or final value of the workpiece can be safely and reliably determined by determining an averaged “scalar product absolute value” or an averaged “coordinate value” of the peak in question. A simple possibility for this is, for example, to determine the most probable initial value and / or final value by calculating the median of the values ​​of the frequency distribution and then using the points (scalar products or coordinates) as the start or end points.

[0042] The end point of the workpiece is determined at which the frequency distribution exceeds half of the median value for the first time from the positive or negative direction of the eigenvector or the coordinate direction under consideration.

[0043] In a further embodiment of the invention, the most probable initial value and / or final value of the workpiece's dimensions is determined by determining the local maximum and / or the local minimum of the first partial derivative of the density or frequency distribution in the direction of the corresponding singular vector and / or eigenvector and / or in the coordinate direction. In the event of ambiguities, the first extremum in the direction under consideration is used. Using this peak evaluation method, the most probable initial value and / or final value of the workpiece can be safely and reliably determined, for example, in point cloud data obtained using a computed tomography measurement (voxel).

[0044] In one embodiment, the most probable initial value and / or final value of the workpiece's expansion in the direction of a singular vector and / or eigenvector and / or in the coordinate direction is determined by evaluating the density or frequency distribution of the largest or smallest scalar products or the largest or smallest coordinates, whereby scalar products or coordinates that deviate from the largest or smallest scalar products or coordinates by more than 5%, and in particular by more than 1%, are disregarded for this evaluation. This distance requirement represents a simple and effective method for eliminating so-called "outliers" that would otherwise complicate or distort the determination of the initial value and / or final value.

[0045] In a further embodiment of the invention, the registration of point cloud data of the workpiece is carried out through parallelization by or with the participation of at least one graphics processor. Such parallelization, for example, on the front-end computer of the production machine or, for example, in the cloud, can further save computing time.

[0046] In one embodiment of the invention, the point cloud data of the workpiece are measured by at least one sensor as coordinates of points of the workpiece and further processed by at least one processor to form point cloud data of the workpiece, wherein the point cloud data represent at least a partial area of ​​the workpiece to be manufactured and wherein the point cloud data are further processed by this and / or at least one further processor in accordance with the further method steps and / or are transferred to other processors for further processing.This allows the actual situation of the workpiece in the current processing situation to be recorded by means of a sensor in or on the production machine, whereby the measurement data from the sensor is pre-processed, taking into account any calibrations, to form point cloud data of the workpiece by a processor, for example a processor of the front-end computer of the production machine, a processor of the higher-level factory control system and / or also, for example, by a processor in the cloud.

[0047] In a further embodiment of the invention, the test plan and / or the CAD model of the workpiece and, if applicable, the actual control data and / or, if applicable, the available infrastructure or environmental data of the production machine are preprocessed and transferred by at least one processor of the production machine and / or by at least one processor of the factory control system superordinate to the production machine. This enables the recording of the target situation of the workpiece based on test plans and / or CAD data, and the recording of the actual situation of the control data of the production machine orthe actual situation of the machining head and / or the manufacturing environment with the aid of which the manufacturer of the workpiece intends to manufacture the workpiece within tolerance during machining with the at least one manufacturing machine, whereby a pre-processing of these target data and the intended actual situations is carried out by a processor, for example a processor of the front-end computer of the manufacturing machine, a processor of the higher-level factory control and / or also by, for example, a processor in the cloud.

[0048] In one embodiment of the invention, the new target control data and / or, if applicable, the new infrastructure or environmental data are further processed by at least one processor of the at least one production machine and / or the at least one further production machine and / or by at least one processor of the factory control system superordinate to the at least one production machine and / or the at least one further production machine for the production of the new workpiece. As a result, the results of the invention are fed into the production machines and / or factory control system in the form of new target data for the further production of the workpiece or for the production of new workpieces.

[0049] Further features and advantages of the invention will become apparent from the following description of exemplary embodiments of the invention, with reference to the figures, which illustrate details essential to the invention, and from the claims. The individual features can be implemented individually or in combination in a variant of the invention.

[0050] It is clear that the features mentioned above and those to be explained below can be used not only in the combination specified in each case, but also in isolation or in other combinations without departing from the scope of the present invention.

[0051] In addition, it is understood that the features defined in the dependent claims for the computer-implemented method can also be used in the same or equivalent manner as features for the computer program according to the invention, the data carrier signal according to the invention and as system features for the manufacturing system according to the invention, without these being listed again separately here as corresponding features.

[0052] The present invention will now be explained in more detail using embodiments and drawings.

[0053] Showing:

[0054] Fig. 1 A two-dimensional representation of point cloud measurement data of a transmission shaft of a gearbox for electric bicycles,

[0055] Fig. 2 The arrangement of the two-dimensional point cloud from Fig. 1 along the determined direction of the principal singular vector of the difference matrix or the principal eigenvector of the covariance matrix in the horizontal direction as a result of the singular value decomposition (SVD) of the difference matrix formed from the point cloud data of Fig. 1,

[0056] Fig. 3 The density or frequency distribution of the scalar products of point cloud data and singular eigenvectors in the direction of the principal singular vector or principal eigenvector of the point cloud data from Fig. 1 or in the X-direction,

[0057] Fig. 4 The arrangement of the two-dimensional point cloud from Fig. 1 along the zero line in the horizontal direction, by taking into account the most probable initial value and / or final value of the extension of the transmission wave from Fig. 1 in the direction of the main singular vector or main eigenvector or in the X-direction by considering a density or frequency distribution of the scalar products of point cloud data and singular / eigenvectors in the direction of the corresponding main singular vector or main eigenvector or in the X-direction, Fig. 5 A flow chart of the method according to the invention,

[0058] Fig. 6A A schematic representation of the number and arrangement of devices,

[0059] Fig. 6B A diagram of an example environment in which systems and methods according to the invention may be implemented, and

[0060] Fig. 6C A diagram of example components of a manufacturing system.

[0061] Figure 1 shows a two-dimensional representation of point cloud measurement data from a transmission shaft (7) of an electric bicycle gearbox. The two-dimensional representation of the original three-dimensional point cloud of measurement data from the transmission shaft (7) as an exemplary workpiece (7) of the invention was intentionally chosen to illustrate the registration process, since the rotation and fitting of a three-dimensional point cloud in black and white are difficult to represent on paper.

[0062] The registration method according to the invention generally starts from a three-dimensional point cloud of measurement data, for which, for each measurement data point in the point cloud, three numerical values ​​are available for the coordinates of each recorded measurement data point in the X, Y, and Z directions with respect to the sensor coordinate system. This assumes that the sensor head is sufficiently calibrated to record the measurement data in its sensor coordinate system and that the numerical values ​​for the coordinates correspond to the error-prone absolute values ​​in a metric system within the error tolerance of the sensor head. The measurement data can, for example, be coordinates of points on the surface of the workpiece (7).

[0063] However, measurement data resulting from a computer tomographic image of the workpiece using a CT measuring device can also be registered using the registration method according to the invention.

[0064] Without loss of generality, it is assumed in the following to explain the procedure that each measurement data point x 1 has three coordinate values ​​(x, y, z) with respect to the sensor coordinate system and that the corresponding point cloud of the measurement data is therefore defined by the set of measurement data points {x 1 , ... , x n} is represented.

[0065] Furthermore, a center of gravity vector S with the coordinates (S x , S y , S z ), where S x is given by the mean value of all X-coordinates of the measured data points x* and accordingly S y is given by the mean value of all Y-coordinates of the measured data points x*, and S z is given by the mean value of all Z-coordinates of the measured data points x*.

[0066] Furthermore, a difference matrix M can then be created, which is valid for all measurement points x 1 represents the deviation of the measuring points relative to the center of gravity S, with

[0067] From this difference matrix M the covariance matrix X can then be determined as

[0068] The orientation of the point data cloud can then be determined by determining the three eigenvectors i) 1 , v 2 , D 3 with the three largest eigenvalues ​​of the covariance matrix X, since it can be shown that the largest eigenvalue li corresponds to an eigenvector v 1 , which represents the direction of the best-fit line for the point cloud under consideration. Accordingly, the eigenvectors w 1 and w 2, which belong to the first two largest eigenvalues ​​Xi and M, form the “best-fit” plane for the considered point cloud. Therefore, a 3x3 rotation matrix Q can be calculated based on the three eigenvectors v 1 , v 2 , v 3 with the corresponding eigenvalues ​​li > MM sorted by size, with the help of which the point cloud data x present in the sensor coordinate system 1 can be aligned with the computer or CAD coordinate system to x' 1 x'.

[0069] The 3x3 rotation matrix Q is given by

[0070] The calculation of all eigenvectors v k the covariance matrix X can be done using a normal singular value decomposition (SVD) of M:

[0071] The matrix Z of the singular value decomposition has the form: with the singular values ​​oi, 02, 03

[0072] II

[0073] The singular values ​​Ne Oi are linked to the eigenvalues ​​Xi by m 2 = (n-1) Xi.

[0074] The 3x3 matrix V of the singular value decomposition of M contains the eigenvectors v k of £ as columns, as follows starting from with M = UZV T shown: follows:

[0075] This equation can now be multiplied from the right with the matrix V, which gives

[0076] Use of V T V = I follows: This last result is equivalent to the eigenvalue equation £ v" = Xi v', from which it follows that the columns of V are the eigenvectors of If the eigenvectors in the columns of V are then also ordered by the magnitude of the eigenvalues ​​according to Xi > I2 > h, then the matrix V corresponds to the desired rotation matrix Q for rotating the point data cloud from the sensor coordinate system to the computer coordinate system. The eigenvectors of the covariance matrix ]T for the three largest eigenvalues ​​are identical to the singular vectors of the difference matrix M.

[0077] The calculation of the singular value decomposition of the difference matrix MGR" 13 , n > 3, is usually carried out in two steps:

[0078] First, in a first step, the difference matrix M is transformed into a bidiagonal matrix B using a so-called Householder transformation:

[0079] UB T M VB = B , where B is a bidiagonal matrix and ÜB, VB are orthogonal matrices as a result of the Householder transformations.

[0080] In a second step, the superdiagonal elements, i.e. all elements directly above and to the right of the main diagonal, are set to zero in an iterative procedure using an algorithm according to Golub and Kahn:

[0081] Uz T B Vz = Z = diag(oi , 02, 03) , where the matrices Uz and Vz are orthogonal.

[0082] With Z the singular values ​​01, 02, 03 are also present.

[0083] The calculation of V is then carried out by:

[0084] V = VB VZ . The numerical effort for calculating the matrices Z and V can be estimated as 2*n*3 2 + 11*3 3 Calculation steps, which thus scale linearly with the number n of measurement points in the point cloud. Thus, 10 7Measurement points can be easily processed on a modern industrial PC in under a second, and the rotation matrix Q required for registration can be determined. To further reduce time, it is possible to run the algorithm in parallel on multiple processors, for example, on one or more graphics cards. However, especially in cloud computing, the computing operations can also be distributed across multiple processors.

[0085] In addition to the singular value decomposition of the difference matrix M outlined above, there are numerous other numerical methods that can be used to iteratively determine the singular vectors of the difference matrix M and / or the eigenvectors of the covariance matrix J. Ready-made solutions for this are available, for example, in open source libraries such as https: / / eigen.tuxfamily.org / dox / group SVD Module.html or specialized libraries for GPGPU https: / / github.com / scrose / SVDSolver or for distributed computing systems https: / / github.com / ecrc / ksvd . In addition, alternative calculation methods according to the so-called Krylov-Schur approach and the so-called truncated SVD are available; see https: / / doi.org / 10.1016 / jdaa.2011.07.022 and https: / / arxiv.org / pdf / 2009.00761.

[0086] Figure 2 shows the arrangement of the two-dimensional point cloud from Figure 1 along the determined direction of the principal singular vector of the difference matrix or principal eigenvector of the covariance matrix in the horizontal direction as a result of the singular value decomposition (SVD) of the difference matrix formed from the point cloud data from Figure 1. This arrangement of the two-dimensional point cloud along the principal singular vector or principal eigenvector or in the X direction corresponds to the rotation of the point cloud from Figure 1 using the rotation matrix Q in the computer coordinate system. The principal singular vector or principal eigenvector is aligned parallel to the X axis of the computer coordinate system after the rotation using the rotation matrix Q.

[0087] Now that the two-dimensional point cloud data from Fig. 1 have been oriented along the X-direction of the computer coordinate system in the first step of the registration process, see Fig. 2, the point cloud data must be positioned correctly along the X-direction and the Y-direction according to the second step of the registration process.

[0088] There are two ways to do this: firstly, to place the center of gravity S or another arithmetic mean or median of the point cloud on the coordinate origin (centering) or secondly, to place an end surface resulting from the point cloud on a specific target coordinate, which may be obtained, for example, from the CAD data or the test plans.

[0089] The determination of such an end surface of the point cloud data comprises a calculation of the scalar products of the point cloud data from Fig. 1 with the determined singular vectors and / or eigenvectors. Alternatively, the X-coordinate of the end surface of the point cloud data from Fig. 2 rotated into the computer coordinate system can also be determined. With the help of the scalar products and / or the coordinates, the most probable initial value and / or final value of the extension of the workpiece (7) in the direction of a singular vector and / or an eigenvector or in the coordinate direction is determined by considering a density or frequency distribution of the scalar products in the direction of the corresponding singular vector and / or eigenvector or by considering a density or frequency distribution of the coordinates.

[0090] Fig. 3 shows schematically the density or frequency distribution D of the scalar products < xn | v 1 > the point cloud data x n with the principal singular vector and / or principal eigenvector v normalized to length 1 1 , i.e. the number or density of the scalar products < x n | w 1 > / 1 v 1 |, which result in a certain value x, plotted against this value x. It should be noted that the general scalar product of a vector with a direction vector normalized to length 1 represents nothing other than the length projection of this vector in the direction of the direction vector.

[0091] Since the principal singular vector and / or principal eigenvector v 1 of the point cloud data of Fig. 1 after the rotation of the point cloud data according to Fig. 2 is aligned parallel to the x-axis of Fig. 2, this length projection of the point cloud data from Fig. 1 is calculated by means of the scalar product along the direction of the main singular vector and / or main eigenvector D 1This is equivalent to the length projection of the point cloud data from Fig. 2 using the scalar product along the x-direction (1, 0, 0). In other words, the density or frequency distribution D in Fig. 3 simply indicates how often or frequently a certain x-value occurs as an x-coordinate in the point cloud data of Fig. 2.

[0092] It should be noted that the actual 3D point cloud data of the transmission shaft (7) in Figures 1, 2, and 4 naturally also include data for the left front surface, which is not visible in the two-dimensional images of Figures 1, 2, and 4, but is present in the underlying 3D data set. The number of these data for the front surface is significantly larger than the number of data for an outer cylinder with a width of Ax. Therefore, the density or frequency distribution D in Figure 3 for the left front surface has a large peak compared to the rest of the cylinder shell.

[0093] With a corresponding procedure with regard to the second principal singular vector and / or principal eigenvector, which points in the Y direction after the rotation of the point data cloud of Fig. 1 in Fig. 2, a camel or Bactrian camel curve with two humps or peaks results for the density or frequency distribution.

[0094] The positions of one peak in the X direction and the two peaks in the Y direction can now be determined using all known methods of peak evaluation. For example, the evaluation can be performed using mean values, thresholds, half-widths, threshold widths, and / or gradients of the local density or frequency distributions.

[0095] A simple threshold method for peak evaluation in one direction can, for example, be carried out by determining an average density D over all measured points considered in the direction considered using the mean or median, by then considering at which point in the X or Y direction, depending on which direction is currently being considered, the threshold value of D / 2 is exceeded.

[0096] In addition, the end surface of the point cloud data can also be determined by determining the local maximum and / or the local minimum of the first partial derivative of the density or frequency distribution in the direction of the corresponding singular vector and / or eigenvector. A corresponding method is disclosed in patent US 8,045,806 B2, in particular with regard to point data clouds from CT measurement images (voxels). The entire disclosure content of patent US 8,045,806 B2 is hereby incorporated by reference into the content of this application; this applies in particular to Figure 6 and the associated figure description; however, in the event of a conflict, the present description takes precedence over the description of patent US 8,045,806 B2. The method shown in US 8,045,806 B2 is not only restricted to use with CT data, but can be applied to all forms of point data clouds and / or density or frequency data.Frequency distributions of scalar products of point data clouds are applied.

[0097] In a further embodiment, the most probable initial value and / or final value of the expansion of the workpiece (7) in the direction of a singular vector and / or eigenvector and / or in a coordinate direction can be determined by evaluating the density or frequency distribution of the absolute values ​​of the largest and / or smallest scalar products or the absolute values ​​of the largest and / or smallest coordinates in the direction of the corresponding singular vector and / or eigenvector or in the coordinate direction, wherein, for example, scalar products are disregarded for this evaluation which deviate from the absolute values ​​of the largest and / or smallest scalar products or the absolute values ​​of the largest and / or smallest coordinates by more than 5%, in particular by more than 1%.

[0098] Fig. 4 shows such an arrangement of the two-dimensional point cloud from Fig. 1 along the zero line in the horizontal direction and, viewed from the vertical direction, symmetrical with respect to the x-axis. The most probable initial values ​​and / or final values ​​of the extension of the transmission wave from Fig. 1 in the direction of the two main singular vectors or main eigenvectors, or the most probable initial values ​​and / or final values ​​of the extension of the transmission wave from Fig. 2 in the x- and y-direction were determined by examining a density or frequency distribution of the scalar products of point cloud data and singular / eigenvectors in the direction of the two main singular vectors or main eigenvectors, or by examining a density or frequency distribution of the coordinates in the x- and y-direction. The density distribution has one peak in the x-direction according to Fig. 3 and two peaks in the y-direction according to a camel hump or camel hump pattern.Bactrian camel hump curve. It should be noted here that the registration method described with reference to Figures 1 to 4 was only carried out using a partial point cloud of measurement data, which only includes the head of a transmission shaft (7) and the beginning of the collar of the transmission shaft (7) with the beginnings of a gear rim (on the right in the image of Figures 1, 2, and 4). This means that the present registration method according to the invention is capable of reliably and robustly performing the necessary rotation and offset of point cloud data in the computer coordinate system, even if only small partial areas of the workpiece under consideration are available as point cloud data due, for example, to shadowing.In this respect, the present registration method is superior to other registration methods with regard to inline inspection of workpieces in the clock cycle, since it uses statistical methods and therefore leaves out a geometric consideration of the registration problem.

[0099] Furthermore, it should be noted that the term “computer coordinate system” used in relation to the description of Figures 1 to 4 should not be understood in a restrictive sense, but is intended to describe any coordinate system in which the point data clouds recorded in the sensor coordinate system are to be viewed, such as, for example, the CAD coordinate system.

[0100] Fig. 5 shows a diagram of the method according to the invention with the steps:

[0101] • Obtaining 42 point cloud data of a workpiece 7;

[0102] • Registering 43 the point cloud data of the workpiece 7;

[0103] • Fitting 44 of geometric surface and / or form elements to the point cloud data of the workpiece 7;

[0104] • Obtaining 46 a test plan and / or CAD model of the workpiece 7 and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data of at least one production machine 1 for producing the workpiece 7;

[0105] • Comparing 48 the test plan and / or CAD model of the workpiece 7 with the fitted surfaces and / or form elements and determining deviations; • Evaluating or assessing 50 the deviations determined in the comparison, if necessary based on specified tolerances of the workpiece 7 and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data;

[0106] • Generation 52 of new target control data and / or of new infrastructure or environmental data, for the further production of the workpiece 7 or for the production of a new workpiece 7 depending on the evaluation 50;

[0107] • Transferring 54 the new target control data and / or, if applicable, the new infrastructure or environmental data to the at least one production machine 1 and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine 1 and / or the at least one further production machine depending on the evaluation 50; wherein the registration 43 of the point cloud data of the workpiece 7 is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud.

[0108] It should be noted that all of the aforementioned steps 42-54 of the method according to the invention can, for example, run entirely in Microsoft's Azure cloud and thus independently of the control computer of the production machine 1, whereby these steps can also be handled in the cloud in separate software modules, whereby different programs or program parts can also be encapsulated and called within the software modules for the respective purpose of the software module.

[0109] However, it is equally conceivable that individual or even all steps are carried out locally on site as individual software modules or as a complete software package on the at least one production machine 1 and / or on at least one further production machine and / or on a factory control system superordinate to the at least one production machine 1 and / or the at least one further production machine.Furthermore, it is conceivable that the method according to the invention is started and / or controlled from a remote computer, a smartphone app, the control computer of the at least one production machine 1 and / or the control computer of the at least one further production machine and / or a control computer of the factory control system superordinate to the at least one production machine 1 and / or the at least one further production machine, and subsequently the processing of all or individual steps 42-54 takes place, for example, in the Azure cloud and / or in the remote computer and / or in a smartphone app and / or in the control computer of the at least one production machine 1 and / or in the control computer of the at least one further production machine and / or in a control computer of the factory control system superordinate to the at least one production machine 1 and / or the at least one further production machine.

[0110] In this respect, there is a computer-implemented method 40 for generating control data of at least one production machine 1, which method 40 runs on at least one processor, is controlled by at least one processor, and / or is initiated by at least one processor, wherein the method 40 comprises steps 42-54 and at least one of the steps 42-54 of the method 40 is executed on the at least one and / or another processor.

[0111] In step 42, "Obtaining point cloud data of a workpiece," point cloud data from at least one sensor in the sensor coordinate system are obtained, representing coordinates of at least one part of the workpiece under consideration. These can be surface coordinates, for example, recorded using an optical sensor. However, CT data can also be considered as corresponding coordinate data for a part of the workpiece under consideration.

[0112] "Receiving point data clouds" can be understood as any form of data transfer, such as "receiving" point data clouds. However, the active "retrieval" of such point data clouds or the "storage" or "transfer" of such point data clouds from or to a computer, hard drive, other storage medium, and / or in the cloud should also be understood as "receiving point cloud data." These exemplary lists of the "receiving" feature are not to be considered exhaustive, as every conceivable form of data transfer is intended to be covered by the "receiving" feature.

[0113] The point cloud data from step 42 can have been measured as coordinates of points on the workpiece (7) by at least one sensor 202 before step 42 and further processed by at least one processor into point cloud data of the workpiece (7), wherein the point cloud data represent at least a partial area of ​​the workpiece (7) to be manufactured and wherein the point cloud data are further processed by this and / or at least one further processor according to the further method steps and / or are transferred to other processors for further processing. The sensor 202 can be arranged within or near the at least one production machine 1, see, for example, Fig.6C, and the at least one processor may be a processor of the sensor 202, a processor of the manufacturing machine, a processor of the higher-level factory control, a processor of the cloud, or another processor that further processes the sensor signals of the sensor 202 into point data clouds.

[0114] The step 43 “Registration of the point cloud data of the workpiece 7” with the feature that the registration 43 of the point cloud data of the workpiece 7 is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference of the point cloud data coordinates and the center of gravity coordinates of the point cloud, has already been explained in detail above with reference to Figures 1 to 4.

[0115] Step 44, "Fitting geometric surface and / or form elements to the point cloud data of workpiece 7," involves fitting equivalent geometric elements (surface and / or form elements) to individual partial point clouds of the point cloud data using an algorithm (usually a Gaussian BestFit algorithm). A wide variety of software packages from various manufacturers can be used for this purpose, such as the Caligo® or Calypso® software packages from Carl Zeiss Industrielle Messtechnik GmbH, as well as the GOM Inspect® software package from Carl Zeiss GOM GmbH. These software packages can be called in an encapsulated manner for this purpose by other programs and can thus be used on front-end computers, on factory control computers, or even in the cloud.Possible algorithms for the “fitting of geometric surface and / or form elements to the point cloud data of the workpiece 7” are also described in many different ways in the relevant literature, see for example “Least Squares Orthogonal Distance Fitting of Curves and Surfaces in Space” by Sung Joon Ahn, Springer-Verlag, ISSN 0302-9743, ISBN 3-540-23966-9.

[0116] Step 46, "Receiving a test plan and / or CAD model of the workpiece 7 and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data of the at least one production machine 1 for producing the workpiece 7," involves the transfer of the data considered relevant for the production of the workpiece 7. This includes the test plan and / or the CAD data of the workpiece 7 to be manufactured and thus the target data of the workpiece 7 to be manufactured, but if applicable also the intended actual control data of the at least one production machine 1 for producing the workpiece 7 to be manufactured and / or, if applicable, the existing infrastructure or environmental data of the production machine 1 for producing the workpiece 7 to be manufactured."Test plan and / or CAD model of workpiece 7" refers to all possible geometric target data of the workpiece that are suitable for comparison with the recorded actual data. This can also include data derived from the test plan and / or CAD model and / or any design drawings.

[0117] "Receiving" can be understood as any form of data transfer, such as "receiving" data. However, the active "retrieval" of such data or the "storage" or "transfer" of such data from or to a computer, hard drive, other storage medium, and / or in the cloud should also be understood as "receiving." These examples of the characteristic "receiving" are not to be considered exhaustive, as every conceivable form of data transfer is intended to be covered by the characteristic "receiving."A test plan can be created based on CAD data from a CAD model or other production data using a wide variety of software packages from a wide variety of manufacturers, such as the Caligo® or Calypso® software package from Carl Zeiss Industrielle Messtechnik GmbH and the GOM Inspect® software package from Carl Zeiss GOM GmbH. These software packages can be called up in an encapsulated manner for this purpose by other programs and can therefore be used on front-end computers, on factory control computers or in the cloud, in addition to or independently of step 44, for test plan creation and / or for transmitting CAD data from a CAD model in step 46.

[0118] Accordingly, the test plan and / or the CAD model of the workpiece (7) and, if applicable, the actual control data and / or, if applicable, the available infrastructure or environmental data of the at least one production machine (1) can be pre-processed and transferred by at least one processor of the at least one production machine (1) and / or by at least one processor of the factory control system superordinate to the production machine (1).

[0119] It should also be noted that step 46 can take place before, at the same time as or after steps 42, 43 and 44 and thus independently of steps 42, 43 and 44, see also the illustration in Fig. 5.

[0120] In step 48 “Comparing the test plan and / or CAD model of the workpiece 7 with the fitted surfaces and / or form elements and determining deviations”, a pure target / actual comparison is first carried out for the workpiece, taking into account the test plan or < vmodels and, if applicable, associated tolerances. A wide variety of software packages from a variety of manufacturers can be used for this purpose, such as the Caligo® or Calypso® software packages from Carl Zeiss Industrielle Messtechnik GmbH, as well as the GOM Inspect® software package from Carl Zeiss GOM GmbH. These software packages can be called in an encapsulated manner by other programs for this purpose and can thus be used on front-end computers, on factory control computers, or in the cloud, in addition to or independently of step 44, and in addition to or independently of step 46, also for the nominal / actual comparison of the workpiece according to step 48.

[0121] In step 50 "Evaluating or assessing 50 the deviations identified in the comparison, for example based on any specified tolerances of the workpiece 7 and, if applicable, the actual control data and / or, if applicable, the infrastructure or environmental data," an assessment is first carried out to determine whether the identified deviations of the workpiece 7 are acceptable, for example based on any specified tolerances, or whether the identified deviations are no longer acceptable. In the latter case, a check is then carried out to determine whether the identified deviations can be reworked or tolerated, for example, in the sense of a set pairing, or whether the workpiece 7 should be assessed as scrap. In the case of a reworkable deviation, the necessary new target control data of the at least one production machine 1 and / or the at least one further production machine are determined.Furthermore, if the actual control data and / or any infrastructure or environmental data of the production machine 1 are available, it is checked whether the deviations identified from the actual control data and / or infrastructure or environmental data show a certain correlation, so that new target control data and / or new infrastructure data for the production of new workpieces 7 can be derived on the basis of this correlation.

[0122] A very simple feedback model for step 50, for example, simply uses the deviations detected on the workpiece as a negative offset to the original actual control data and generates new target control data from this. For example, if the original actual control data contained a workpiece length of 100 mm for workpiece 7, and workpiece 7 is subsequently 50 μm too long (detected deviation), then the new target data simply specifies a workpiece length of 100 mm - 50 μm. However, such a simple feedback model for step 50 only works as long as a reasonably linear relationship between control data and the resulting dimensions of workpiece 7 can be assumed.

[0123] A typical error in turning applications, for example, is tool wear. The resulting error often manifests itself in deviations in diameters. This is usually corrected by CNC operators, usually by manually entering values ​​into an offset table that a CNC manufacturing machine uses to determine the correct position of the tool center point. In practice, this adjusts the tool feed, which influences the manufactured diameter and allows for correction. The measured error is often multiplied by a factor to correct the offset table. In practice, these factors originate from the specialist knowledge of the CNC operators and are determined empirically. The need for factors is determined by the physical deformations of the component during machining. This includes, for example, thermal and force-related processes.

[0124] The application of the method according to the invention now allows for fine adjustment of the factors in the offset table using the simple feedback model for step 50, since the influence can be directly observed by measuring the workpiece 7. Once the found factors and compensations have been determined from the offset table, production control can be carried out almost autonomously and without manual intervention by the CNC operator.

[0125] There are also reasons other than tool wear that make correction necessary. For example, the rotation axis shifts during the warm-up phase of a CNC manufacturing machine. This error can also be corrected by adjusting the

[0126] Offset table can be compensated. Thus, the method of adjusting the offset table using the simple feedback model of step 50 can be applied to many error patterns.

[0127] This type of compensation is also not limited to lathes, as milling machines, for example, are also affected by an incorrect position of the tool center point. A simple feedback model for step 50 of the method according to the invention can also be used to improve workpiece quality in other manufacturing processes such as laser cladding or 3D printing.

[0128] In addition, machine learning algorithms or trained AI systems can be used to derive or establish new target control data from detected deviations. So-called reinforced learning algorithms, in particular, can be used to provide an effective feedback model for step 50 in complex causal relationships. For example, US 10,180,667 B2 discloses a measurement technology integrated into a production machine in which the measurement results are interpreted by a trained AI, with the AI ​​then deriving or establishing new target control data.

[0129] Other feedback models based on a model simulation of the physical processes in the production machine or based on a black-box approach for the production machine are also conceivable as feedback models for step 50 of the method according to the invention. Therefore, the present invention is not limited to a specific feedback model for step 50; rather, all possible feedback models can be used for step 50 that can derive or establish meaningful new target control data.

[0130] In step 52 “Generation of new target control data and / or new infrastructure or environmental data for the further production of the workpiece 7 or for the production of a new workpiece 7 depending on the evaluation 50”, the new target control data and / or new infrastructure or environmental data defined or derived in step 50 for the further production of the workpiece 7 or for the production of a new workpiece 7 are compiled or made available for transfer to the at least one production machine 1 and / or to the at least one further production machine and / or to the factory control system higher than the at least one production machine 1 and / or the at least one further production machine.

[0131] In step 54 “Transfer of the new target control data and / or, if applicable, the new infrastructure or environmental data to at least one production machine 1 and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine 1 and / or the at least one further production machine depending on the evaluation 50”, the new target control data and / or, if applicable, the new infrastructure or environmental data are transferred accordingly.

[0132] "Transferring" can be understood as any form of data transfer, such as "sending" data. However, the active "storage" of such data or the "transfer" of such data from or to a computer, hard drive, other storage medium, and / or in the cloud should also be understood as "transferring." These examples of the "transfer" characteristic are not to be considered exhaustive, as every conceivable form of data transfer is intended to be covered by the "transfer" characteristic.

[0133] The new target control data and / or, if applicable, the new infrastructure or environmental data can then be further processed by at least one processor of the at least one production machine 1 and / or the at least one further production machine and / or by at least one processor of the factory control system superordinate to the at least one production machine 1 and / or the at least one further production machine for the production of the new workpiece 7.

[0134] However, it goes without saying that in the case of a positive evaluation in step 50, no new target control data and / or no new infrastructure or environmental data need to be generated and subsequently transferred for the further production of the workpiece 7 or for the production of a new workpiece 7, since the workpiece 7 currently measured for the evaluation fulfills the required specification.

[0135] In a further embodiment of the invention, during step 46, actual control data and / or infrastructure or environmental data can be obtained continuously during the production of the new workpiece 7 or only at specific times during the production of the workpiece 7. In particular, the actual control data and / or, if applicable, the available infrastructure or environmental data of the at least one production machine 1 for producing the workpiece 7 can be acquired continuously or only at specific times during the production of the workpiece 7 on or in the vicinity of the at least one production machine 1.

[0136] In this case, machine data of the at least one production machine 1, such as the tool type, the tool number and the service life of a tool used in the processing head of the at least one production machine 1, the current feed rate of the processing head, any accelerations of the processing head, the current power requirement of the processing head, the currently used coolant quantity, the laser parameters of a laser processing head, the deposition parameters of a 3D printing head, the temperature and / or the humidity of the processing environment inside and / or outside the at least one production machine 1, etc., can be recorded or obtained as actual control data and / or infrastructure or environmental data. These data, which are recorded or obtained continuously or sporadically during the production of the workpiece 7, en thus represent a chronologically ordered recording of a plurality of process parameters for the production of the workpiece 7, wherein for each process parameter a corresponding process parameter sequence is recorded or obtained in step 46.

[0137] In a further embodiment of the invention, in an assignment step 46a (not shown in detail in Fig. 5), this plurality of process parameter sequences of the process parameters used to manufacture the workpiece 7 can be assigned to the manufacturing steps originally intended for production (mapping data). In this case, each actual control command within the original actual control data of the at least one manufacturing machine 1 is assigned process parameters at a time or time period at which the actual control command is executed on the at least one manufacturing machine 1.

[0138] In this further embodiment of the invention, an evaluation can then be carried out in step 50 “Evaluating or assessing 50 the deviations determined in the comparison on the basis of any specified tolerances of the workpiece 7 and, if applicable, the actual control data and / or, if applicable, the infrastructure or environmental data”, in which, in addition to the above-described consideration of the deviations determined, the mapping data of the previously described step 46a are also additionally and / or alternatively included or taken into account.

[0139] In this embodiment of the invention, which can be subsumed under the main claims of the invention and is therefore fully covered by the claimed scope of protection of the main claims, the computer-implemented method 40 for generating control data for at least one production machine 1, which method runs on at least one processor, is controlled by at least one processor, and is initiated by at least one processor, comprises the following steps, wherein at least one of the following steps of the method 40 is executed on the at least one and / or another processor:

[0140] • Obtaining 42 point cloud data of a workpiece 7;

[0141] • Registering 43 the point cloud data of the workpiece 7;

[0142] • Fitting 44 of geometric surface and / or form elements to the point cloud data of the workpiece 7;

[0143] • Obtaining 46 a test plan and / or CAD model of the workpiece 7 and actual control data and / or, if applicable, existing infrastructure or environmental data from at least one production machine 1 for producing the workpiece 7, wherein the actual control data and / or the possibly existing infrastructure or environmental data represent process parameters which are obtained continuously or only at specific times during the production of the workpiece 7 in the context of step 46;

[0144] • Assigning 46a (mapping) a plurality of process parameter sequences of the process parameters obtained in step 46 to the individual manufacturing steps required for the production of the workpiece 7 in order to generate corresponding mapping data by assigning process parameters to each actual control command within the original actual control data of the at least one manufacturing machine 1 at a time or during a time period at which the actual control command is executed;

[0145] • Comparing 48 the test plan and / or CAD model of the workpiece 7 with the fitted surfaces and / or form elements and determining deviations;

[0146] • Evaluating or assessing 50 the deviations determined in the comparison, if necessary based on specified tolerances of the workpiece 7 and the actual control data and / or if necessary the infrastructure or environmental data, wherein additionally and / or alternatively the mapping data of step 46a are included or taken into account;

[0147] • Generating 52 new target control data and / or new infrastructure or environmental data for the further production of the workpiece 7 or for the production of a new workpiece 7 depending on the evaluation 50; • Transferring the new target control data and / or, if applicable, the new infrastructure or environmental data.Environmental data to the at least one production machine 1 and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine 1 and / or the at least one further production machine as a function of the evaluation 50; wherein the registration 43 of the point cloud data of the workpiece 7 is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud.

[0148] It goes without saying that in the present invention and in the previously described further embodiment of the invention, step 52 for generating new target control data and / or new infrastructure or environmental data for the further production of the workpiece 7 or for the production of a new workpiece 7 depending on the evaluation 50 does not have to be carried out explicitly, but can also be a direct result of the evaluation step 50, which results implicitly from the evaluation of the deviations determined in the comparison, if necessary based on predetermined tolerances of the workpiece 7 and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data, whereby the mapping data of step 46a can also be included or taken into account additionally and / or alternatively.

[0149] In this respect, steps 50 and 52 do not have to be carried out separately and in isolation from one another within the scope of the present invention, but can also be carried out jointly, simultaneously or sequentially, interwoven or combined within a single step, without thereby departing from the scope of the present invention.

[0150] Furthermore, it goes without saying that the generation 52 of new target control data and / or new infrastructure or environmental data for the production of a new workpiece 7 depending on the evaluation 50 and the transfer of the new target control data and / or the possibly new infrastructure or environmental data to the at least one production machine 1 and / or to at least one further production machine and / or to a factory control system higher than the production machine 1 and / or the at least one further production machine depending on the evaluation 50 also includes a further, separate production machine which does not necessarily have to correspond to the design of the production machine 1 and which produces a new workpiece 7 with the new target control data in the same factory or at another location in another country.In this case, factory control can also operate across locations or even across countries with regard to the production machines.

[0151] Furthermore, it is understood that the at least one production machine 1 and the at least one further production machine represent terms that are interchangeable with one another, so that the workpiece 7 to be manufactured can also be produced on the at least one further production machine, wherein corresponding actual data of the workpiece 7 can then also be recorded on this further production machine using, for example, optical sensors, and wherein new target control data for manufacturing a new workpiece 7 for the production machine 1 can then also be generated based on the invention.

[0152] Furthermore, within the scope of the further embodiment of the invention outlined above, by logging the multitude of process parameter sequences during the production steps, it is possible to create a historical logbook of the production of one or more workpieces 7. This type of logbook enables time-resolved error analysis or production monitoring. In this respect, the production machine is no longer a black box that delivers a finished workpiece 7, which can then be qualified after production, thereby deriving empirical or historical production rules for the production of the workpiece 7.

[0153] Therefore, with the further embodiment of the invention, very high demands on the necessary manufacturing accuracy of the production machinery can be avoided due to ignorance or fear, as well as dependence on the individual knowledge of the production machine operators. The logbook can also be used to establish an indirect or virtual in-situ measurement technique for the workpiece 7, since it is possible to assign process parameter sequences and control data to the geometric target data of the workpiece 7.By means of the mapping, in which process parameters are assigned to each actual control command within the original actual control data of the at least one production machine 1 at a time or during a time period at which the actual control command is executed, the complex cause-effect relationship between the production process, in which many different causes can accumulate to a dimensional error of the workpiece 7, and the production result can be selectively investigated.

[0154] The more information is available, be it through a correspondingly dense recording or monitoring of the process parameters during the production of a workpiece 7 or be it through the use of a corresponding amount of information from previous production processes, the better the cause-effect relationship can be trained or mapped using an algorithm, such as a machine learning algorithm with, for example, reinforced learning.

[0155] In this respect, new target control data can be generated, which can be used to further manufacture the workpiece 7 or to produce new workpieces 7 with fewer errors or deviations. Thus, the invention enables time-resolved advance error correction during the manufacturing process by monitoring and evaluating the current process parameters in a time-resolved manner and by providing time-resolved generation of new target control data as an improved feedback method for the industrial production of workpieces 7.

[0156] Fig. 6B shows a diagram of an example environment 200 in which the systems and / or methods according to the invention can be implemented. As shown in Fig. 6B, the environment 200 can include a manufacturing system 210 with at least one manufacturing machine 1, a control device 220 with at least one processor, and a network 230. Devices of the environment 200 can be connected to one another via wired connections, wireless connections, or a combination of wired and wireless connections. The control device 220 includes one or more devices capable of receiving, storing, generating, processing, and / or providing information related to the control or configuration of the manufacturing system 210. For example, the control device 220 can include a server, a computer, a portable device, a cloud computing device, and / or the like.In some implementations, the controller 220 may receive and / or provide information from one or more other devices of the environment 200, such as the manufacturing system 210.

[0157] The network 230 includes one or more wired and / or wireless networks. For example, the network 230 may be a cellular network (e.g., a Long-Term Evolution (LTE) network, a Code Division Multiple Access (CDMA) network, a 3G network, a 4G network, a 5G network, another type of next-generation network, etc.), a public land mobile network (PLMN), a local area network (LAN), a wide area network (WAN), a metropolitan area network (MAN), a telephone network (e.g., the Public Switched Telephone Network (PSTN)), a private network, an ad hoc network, an intranet, the Internet, a fiber optic network, a cloud computing network, or the like, and / or a combination of these or other types of networks.

[0158] The arrangement of devices and networks shown in Fig. 6A serves as an example. In practice, there may be additional devices and / or networks, fewer devices and / or networks, different devices and / or networks, or differently arranged devices and / or networks than those in Fig. 6A. Furthermore, two or more of the devices shown in Fig. 6A may be implemented within a single device, or a single device shown in Fig. 6A may be implemented as multiple, distributed devices. Additionally or alternatively, a number of devices (e.g., one or more devices) of the environment 200 may comprise one or more execute actions described as being performed by another set of devices in environment 200.

[0159] Fig. 6A shows a diagram of example components of a device 300. The device 300 may correspond to the manufacturing system 210 and / or the control device 220. In some implementations, the manufacturing system 210 and / or the control device 220 may include one or more devices 300 and / or one or more components of the device 300. As shown in Fig. 6A, the device 300 may include a bus 310, a processor 320, a memory 330, a storage component 340, an input component 350, an output component 360, and a communications interface 370.

[0160] The bus 310 includes a component that enables communication between the components of the device 300.

[0161] Processor 320 is implemented in hardware, firmware, or a combination of hardware and software. Processor 320 is a central processing unit (CPU), a graphics processing unit (GPU), an accelerated processing unit (APU), a microprocessor, a microcontroller, a digital signal processor (DSP), a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), or another type of processing component. In some implementations, processor 320 includes one or more processors that can be programmed to perform a function.

[0162] The memory 330 includes one or more memories, such as random access memory (RAM), read-only memory (ROM), and / or another type of dynamic or static storage device (e.g., flash memory, magnetic memory, and / or optical memory) that stores information and / or instructions for use by the processor 320.

[0163] The storage component 340 stores information and / or software related to the operation and use of the device 300. For example, the storage component 340 may include a hard disk (e.g., a magnetic disk, an optical disk, a magneto-optical disk, and / or a solid-state disk), a compact disc (CD), a digital versatile disc (DVD), a floppy disk, a cartridge, a magnetic tape, and / or another type of non-transitory computer-readable media, along with a corresponding drive.

[0164] The input component 350 comprises a component that enables the device 300 to receive information, for example, via user input (e.g., a touchscreen display, a keyboard, a mouse, a button, a switch, and / or a microphone). Additionally or alternatively, the input component 350 may also comprise at least one sensor for acquiring information. These may be, for example, triangulation sensors or fringe projection sensors for acquiring surface coordinates of a workpiece 7, or even sensors for acquiring the infrastructure or environmental data of the production machine 1.

[0165] The output component 360 includes a component that provides output information to the device 300 (e.g., a display, a speaker, and / or one or more light-emitting diodes (LEDs)).

[0166] The communication interface 370 includes a transceiver-like component (e.g., a transceiver and / or a separate receiver and transmitter) that enables the device 300 to communicate with other devices, for example, via a wired connection, a wireless connection, or a combination of wired and wireless connections. The communication interface 370 may enable the device 300 to receive information from and / or forward information to another device. For example, the communication interface 370 may include an Ethernet interface, an optical interface, a coaxial interface, an infrared interface, a radio frequency (RF) interface, a Universal Serial Bus (USB) interface, a Wi-Fi interface, a cellular network interface, or the like.

[0167] The device 300 can perform one or more of the method steps 42-54 described above, as well as the entire inventive method. The device 300 can perform these method steps or the method based on the processor 320 executing software instructions stored by a non-transitory computer-readable medium, such as the memory 330 and / or the storage component 340. A computer-readable medium is defined herein as a non-transitory storage device. A storage device includes storage space within a single physical storage device or storage space distributed across multiple storage devices. Software instructions can be read into the memory 330 and / or the storage component 340 from another computer-readable medium or from another device via the communication interface 370.When software instructions stored in memory 330 and / or storage component 340 are executed, this may cause processor 320 to perform one or more of the method steps 42-54 described above. Additionally or alternatively, hardwired circuitry may be used instead of or in combination with software instructions to perform one or more of the described method steps. Therefore, the implementations described herein are not limited to any particular combination of hardware circuitry and software.

[0168] Fig. 6C shows an example of a production system 210, which comprises at least one production machine 1 and at least one control device 220 and optionally a factory control for controlling the at least one manufacturing machine 1 and / or for controlling at least one further manufacturing machine. The control device 220 can be connected locally via data lines 222, via a network 230, and / or via a bus 310 to the controller 212 of the manufacturing machine 1 and / or the sensor 202 of the manufacturing system 210.

[0169] In Figure 6C, a workpiece 7 that has already been machined or is yet to be machined is located on the workpiece table 118 of the production machine 1. The production head 106 of the production machine 1 is used for subtractive, i.e., machining, or additive machining of the workpiece 7. A movable arm and / or robot 108 is used to move the sensor 202 relative to the workpiece 7. Additionally or alternatively, a movable arm and / or robot 108 could also be used to move the workpiece relative to the sensor 202. In the example of Figure 6C, the sensor 202 has an optical illumination beam 206 and a measurement beam 204 reflected by the workpiece. This can therefore be a triangulation or fringe projection sensor 202. However, the present invention is not limited to a specific sensor type; thus, all sensor types that can generate a point cloud can be used.Furthermore, the present invention is not limited to measuring and evaluating only a completely machined workpiece 7. The present invention can be used, for example, to evaluate partial machining steps and / or the machining progress in situ, since the registration method can process incomplete partial point clouds of the workpiece in a very short time. For such a production-accompanying evaluation of the

[0170] Workpiece 7, either sensor 202 and workpiece 7 can be moved relative to each other during processing or during a processing break by at least one corresponding actuator 108 or there can be enough sensors 202 to cover the areas of interest of the entire workpiece 7.

[0171] It is understood that the exemplary embodiments shown in the figures and described in the description serve merely to schematically illustrate the principle and embodiment of the present invention. Various functional and structural modifications are possible without departing from the scope of the present invention.

Claims

Patent claims:

1. A method (40) for generating control data for at least one production machine (1) that is computer-implemented and / or runs on at least one processor and / or is controlled by at least one processor and / or is initiated by at least one processor, wherein the method (40) comprises the following steps and at least one of the following steps of the method (40) is executed on the at least one and / or another processor: • Obtaining (42) point cloud data of a workpiece (7); • Registering (43) the point cloud data of the workpiece (7); • Fitting (44) geometric surface and / or form elements to the point cloud data of the workpiece (7); • Obtaining (46) a test plan and / or C AD model of the workpiece (7) and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data from at least one production machine (1) for producing the workpiece (7); • Comparing (48) the test plan and / or CAD model of the workpiece (7) with the fitted surfaces and / or form elements and determining deviations; • Evaluating or assessing (50) the deviations determined in the comparison, if necessary based on specified tolerances of the workpiece (7) and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data; • Generating (52) new target control data and / or new infrastructure or environmental data for the further production of the workpiece (7) or for the production of a new workpiece (7) depending on the evaluation (50); • Transferring the new target control data and / or, if applicable, the new infrastructure or environmental data to the at least one production machine (1) and / or to at least one further production machine and / or to one of the production machines (1) and / or the at least one further Production machine higher-level factory control depending on the evaluation (50); wherein the registration (43) of the point cloud data of the workpiece (7) is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud.

2. Method (40) according to claim 1, wherein the registration (43) also comprises a calculation of the scalar products of the point cloud data with at least two determined singular vectors and l or eigenvectors and / or wherein the registration (43) also comprises a determination of at least two coordinates in the X and / or Y and / or Z direction of the coordinates of the point cloud data rotated into the computer or CAD coordinate system.

3. Method (40) according to claim 2, wherein with the aid of the scalar products and / or the coordinates, the most probable initial value and / or final value of the extension of the workpiece (7) in the direction of a singular vector and / or an eigenvector and / or in a coordinate direction is determined by considering a density or frequency distribution of the scalar products in the direction of the corresponding singular vector and / or eigenvector and / or by considering a density or frequency distribution of the coordinates.

4. Method (40) according to claim 3, wherein when considering the density or frequency distribution of the scalar products and / or the coordinates, the evaluation is carried out on the basis of mean values, threshold values, half-widths, threshold widths and / or gradients of the local density or frequency distributions.

5. Method (40) according to claim 3, wherein the most probable initial value and / or final value of the extension of the workpiece (7) is determined by determining the local maximum and / or the local minimum of the first partial derivative of the density or frequency distribution in the direction of the corresponding singular vector and / or eigenvector and / or in the coordinate direction, whereby in case of ambiguities the first extremum in the direction considered is used.

6. Method (40) according to claim 3 or 4, wherein the most probable initial value and / or final value of the extension of the workpiece (7) in the direction of a singular vector and / or eigenvector and / or in a coordinate direction is determined by an evaluation of the density or frequency distribution of the scalar products or coordinates with the largest or smallest absolute values, and wherein scalar products or coordinates which deviate from the scalar products or coordinates with the largest or smallest absolute values ​​by more than 5%, in particular by more than 1%, are disregarded for this evaluation.

7. Method (40) according to one of the preceding claims, wherein the registration (43) of point cloud data of the workpiece (7) is carried out on the basis of parallelization by or with the participation of at least one graphics processor.

8. Method (40) according to one of the preceding claims, wherein the point cloud data of the workpiece (7) are measured by at least one sensor as coordinates of points of the workpiece (7) and further processed by at least one processor to form point cloud data of the workpiece (7), wherein the point cloud data represent at least a partial area of ​​the workpiece (7) to be manufactured and wherein the point cloud data are further processed by this and / or at least one further processor in accordance with the further method steps and / or are transferred to other processors for further processing.

9. Method (40) according to one of the preceding claims, wherein the test plan and / or the CAD model of the workpiece (7) and optionally the actual control data and / or optionally the available infrastructure or environmental data of the at least one production machine (1) are processed by at least one processor of the at least one production machine (1) and / or by at least one processor of the factory control system superordinate to the at least one production machine (1) preprocessed and handed over.

10. Method (40) according to one of the preceding claims, wherein the new target control data and / or optionally the new infrastructure or environmental data are further processed by at least one processor of the at least one production machine (1) and / or the at least one further production machine and / or by at least one processor of the factory control system superordinate to the at least one production machine (1) and / or the at least one further production machine for the production of the new workpiece (7).

11. A computer program comprising instructions which, when the program is executed by at least one processor, cause this at least one processor and / or further processors to carry out the method according to one of claims 1 to 10.

12. Data carrier signal that transmits the computer program in whole or in part.

13. Manufacturing system (210) comprising a computer program according to claim 11 and at least one processor.

14. A manufacturing system (210) comprising at least one processor and at least one memory, wherein the at least one processor exchanges data with the at least one memory and the manufacturing system (210) is configured to: • Obtaining (42) point cloud data of a workpiece (7); • Registering (43) the point cloud data of the workpiece (7); • Fitting (44) geometric surface and / or form elements to the point cloud data of the workpiece (7); • Obtaining (46) a test plan and / or CAD model of the workpiece (7) and, if applicable, actual control data and / or, if applicable, existing infrastructure or environmental data of at least one production machine (1) for producing the workpiece (7); • Comparing (48) the test plan and / or CAD model of the workpiece (7) with the fitted surfaces and / or form elements and determining deviations; • Evaluating or assessing (50) the deviations determined in the comparison, if necessary based on specified tolerances of the workpiece (7) and, if necessary, the actual control data and / or, if necessary, the infrastructure or environmental data; • Generating (52) new target control data and / or new infrastructure or environmental data for the further production of the workpiece (7) or for the production of a new workpiece (7) depending on the evaluation (50); • Transferring the new target control data and / or, if applicable, the new infrastructure or environmental data to the at least one production machine (1) and / or to at least one further production machine and / or to a factory control system higher than the at least one production machine (1) and / or the at least one further production machine depending on the evaluation (50); wherein the registration (43) of the point cloud data of the workpiece (7) is based on a singular value decomposition (SVD) of a difference matrix of the point cloud data with determination of the associated singular vectors and / or on a determination of the eigenvectors of a covariance matrix of the point cloud data formed from the difference matrix, wherein the difference matrix is ​​formed from the difference between the point cloud data coordinates and the center of gravity coordinates of the point cloud.

15. Manufacturing system (210) according to claim 14, wherein the registration (43) also comprises a calculation of the scalar products of the point cloud data with at least two determined singular vectors and / or eigenvectors and / or wherein the registration (43) also comprises a determination of at least two coordinates in the X and / or Y and / or Z direction of the coordinates of the data entered into the computer or CAD coordinate system rotated point cloud data.

Images