Laser processing system

Abstract

To efficiently perform the sorting operation of parts cut by laser cutting, the system includes: a control device (10) that controls a laser oscillator (1) that emits a laser and a drive unit (4) that moves the position of the workpiece irradiated by the laser, i.e., the irradiation point, along the processing path to perform the processing of cutting the workpiece into parts and remaining material; a detection unit (6) that determines a timing signal by observing the state of the workpiece being processed in a time sequence; a processing state evaluation unit (7) that determines evaluation information by evaluating the processing state based on the timing signal for each interval divided by the processing path; an evaluation information storage unit (8) that associates the boundary, i.e., the contour line, between the parts and the remaining material with the evaluation information and stores it as contour line evaluation information; and a sorting operation determination unit (9) that determines a sorting control command based on the contour line evaluation information, which is a command to control the sorting operation of taking out parts from the processed position of the workpiece and moving them to a target position.

Description

Technical Field

[0001] The present invention relates to a laser processing system for sorting components cut out in a laser cutting process to a desired location. Background Technology

[0002] In a laser processing apparatus that cuts the workpiece into parts and remaining material, if the parts, which are to be sorted, are arranged according to their type and processing status after processing, the efficiency of subsequent processes can be improved.

[0003] Patent Document 1 discloses a production monitoring system for monitoring the differentiation of products. In this system, monitoring data is obtained from the processing machine for each product. Furthermore, products whose evaluation information obtained from the monitoring data falls outside a threshold range are differentiated from other products. A prescribed notification is given when differentiating from other products, and the monitoring data falling outside the threshold range is displayed.

[0004] Patent Document 1: Japanese Patent Application Publication No. 2019-46293 Summary of the Invention

[0005] In the production monitoring system described in Patent Document 1, sorting is performed even when the monitoring data of the product to be sorted is poor. Moreover, when the production monitoring system described in Patent Document 1 is applied to the sorting of parts cut by laser cutting, the device may stop if it is unable to sort parts that have not been separated from the remaining material due to poor processing. As a result, a laser processing system that can efficiently perform the sorting operation of parts cut by laser cutting cannot be provided.

[0006] The present invention was made in view of the above circumstances, and its object is to provide a laser processing system capable of efficiently performing sorting operations on parts cut out by laser cutting.

[0007] The laser processing system of the present invention comprises: a control unit that controls a laser oscillator that emits laser light and a drive unit that moves the position of the workpiece irradiated by the laser, i.e., the irradiation point, along a processing path to perform processing that cuts the workpiece into parts and remaining material parts; a detection unit that determines a timing signal by observing the state of the workpiece during processing according to a time sequence; a processing state evaluation unit that evaluates the processing state based on the timing signal for each interval divided by the processing path and determines evaluation information as evaluation information; an evaluation information storage unit that associates the boundary, i.e., the contour line, between the parts and the remaining material parts with the evaluation information and stores it as contour line evaluation information; and a sorting action determination unit that determines a sorting control command based on the contour line evaluation information, the sorting control command being a command that controls a sorting action that removes a part from the position where the workpiece is processed and moves it to a target position.

[0008] The effects of the invention

[0009] According to the present invention, a laser processing system is provided that can efficiently perform sorting operations on parts cut out by laser cutting processing. Attached Figure Description

[0010] Figure 1 This is a diagram illustrating an example of the structure of the laser processing system in Embodiment 1.

[0011] Figure 2 This is a diagram showing an example of the workpiece in this embodiment.

[0012] Figure 3 This is a flowchart illustrating an example of a sorting operation in this embodiment.

[0013] Figure 4 This is a diagram showing an example of the workpiece in Embodiment 1.

[0014] Figure 5 This is a diagram illustrating an example of contour evaluation information in Implementation 1.

[0015] Figure 6 This is a diagram showing an example of the placement of components in Embodiment 1.

[0016] Figure 7 This is a flowchart illustrating an example of the operation of the sorting action determination unit in Embodiment 1.

[0017] Figure 8 This diagram illustrates a structural example where the processing circuit of the laser processing system in Embodiment 1 is composed of a processor and a memory.

[0018] Figure 9This diagram illustrates a structural example where the processing circuit of the laser processing system in Embodiment 1 is constructed using dedicated hardware.

[0019] Figure 10 This is a diagram illustrating an example of the structure of the laser processing system in Embodiment 2.

[0020] Figure 11 This is a flowchart illustrating an example of the operation when the program generation unit in Embodiment 2 generates a processing program.

[0021] Figure 12 This is a flowchart illustrating an example of the actions taken when the program generation unit in Implementation 2 generates a sorting program.

[0022] Figure 13 This is a diagram illustrating an example of the structure of the laser processing system in Embodiment 3.

[0023] Figure 14 This is a block diagram illustrating an example of the structure of the processing status evaluation unit in Embodiment 3.

[0024] Figure 15 This is a block diagram illustrating an example of the structure of the learning unit in Embodiment 3.

[0025] Figure 16 This is a diagram illustrating an example of the structure of the neural network in Implementation Method 3.

[0026] Figure 17 This is a block diagram illustrating an example of the structure of the processing status evaluation unit in Embodiment 3.

[0027] Figure 18 This is a diagram illustrating an example of the structure of the laser processing system in Embodiment 4.

[0028] Figure 19 This is a block diagram illustrating an example of the structure of the processing status evaluation unit in Embodiment 4. Detailed Implementation

[0029] The embodiments will now be described in detail with reference to the accompanying drawings. Furthermore, the embodiments described below are illustrative, and the scope of the present invention is not limited to the embodiments described below. Additionally, the embodiments described below can be appropriately combined and implemented.

[0030] Implementation Method 1

[0031] Figure 1 This is a diagram illustrating an example of the structure of the laser processing system in Embodiment 1 of the present invention. Figure 1The laser processing system 1000 shown includes: a laser processing apparatus 100 that performs laser processing; and a sorting apparatus 200 that sorts parts cut out by the processing performed by the laser processing apparatus 100. The laser processing apparatus 100 includes a laser oscillator 1 that emits a laser L, an optical path 2 that guides the laser L, a processing head 3 equipped with an optical system 31 that focuses the laser L, and a tray 5 that holds the workpiece w. Furthermore, the laser processing apparatus 100 includes a detection unit 6 that determines a timing signal ts, a processing status evaluation unit 7 that determines evaluation information EI, an evaluation information storage unit 8 that stores contour line evaluation information E, and a sorting action determination unit 9 that determines the sorting control command sc. The laser processing apparatus 100 also includes a control unit 10 that determines the oscillator command oc and the drive command dc, and a drive unit 4 that moves the laser L to the workpiece w at the position of the irradiation point along the processing path. Furthermore, the tray 5 moves the processed workpiece w, which has been processed by moving between the laser processing device 100 and the sorting device 200, toward the sorting device 200.

[0032] The sorting device 200 includes a hand control device 12 that determines the hand drive command hd based on the sorting control command sc, a hand drive unit 13 that drives the hand 14 based on the hand drive command hd, and a hand 14 mechanically connected to the hand drive unit 13. Furthermore, the sorting device 200 has a part placement area 16 for moving parts from their cut-out position. The hand 14 grasps the parts on the tray 5, moving the parts towards the part placement area 16.

[0033] Laser oscillator 1 emits laser light L based on oscillator command OC. Examples of laser oscillators 1 include solid-state lasers, fiber lasers, gas lasers, and diode lasers. Oscillator command OC is a command issued by control device 10 to control laser oscillator 1. Oscillator command OC can indicate laser output, pulse width, duty cycle, peak intensity, current, and operating timing of laser oscillator 1. Laser light L emitted from laser oscillator 1 enters processing head 3 through optical path 2. Optical path 2 can be a space including reflectors, lenses, etc., for guiding light. Alternatively, optical path 2 can be an optical fiber connecting laser oscillator 1 and processing head 3. Furthermore, if optical path 2 includes devices such as a light shield or a beam-changing mechanism, oscillator command OC or drive command dc (described later) can include instructions to instruct the operation of such devices.

[0034] An optical system 31 is provided in the processing head 3 to focus the laser L. For example... Figure 1As shown, a nozzle 32 can be provided to eject the laser L and processing gas from inside the processing head 3 toward the workpiece w. The workpiece w is placed on the tray 5. The drive unit 4 changes the position of at least one of the workpiece w or the processing head 3, and changes the relative position of the workpiece w and the processing head 3 based on the drive command dc. As the drive unit 4, a servo control device with a motor and a position detector can be used, and the motor can be a linear motor. In addition, the aforementioned servo control device can be a drive mechanism using a motor and gears. Moreover, the aforementioned drive mechanism is not limited to a linear axis, and a drive mechanism with a rotary axis can also be used, such as in a multi-joint robot. The structure of the tray 5, processing head 3, etc., is not limited as long as it can change the relative position of the workpiece w and the processing head 3 in accordance with the drive command dc. Figure 1 The structure shown.

[0035] The control device 10 determines the oscillator command OC and the drive command DC, and controls the laser oscillator 1 and the drive unit 4. The control device 10 moves the laser L along the processing path to the position of the workpiece w, i.e., the irradiation point, and performs processing to cut the workpiece w into parts and remaining material parts r. Here, the part of the processed workpiece w that is to be used is called a part. The part other than the part, i.e., the part not to be used, is called the remaining material part r. Furthermore, the distinction between the remaining material part r and the part does not consider reuse. Hereinafter, the oscillator command OC and the drive command DC are sometimes collectively referred to as commands. The processing path can be specified as long as the trajectory of the irradiation point can be specified; for example, the change in the relative position between the workpiece w and the irradiation point can be specified. For example, the path defined by the trajectory of the irradiation point relative to the surface of the workpiece w can be defined as the processing path. The processing path can be the actual trajectory of the irradiation point, or it can be a trajectory in the program or a trajectory in the command. Here, the surface of the workpiece w on which the laser is incident is called the surface. The surface opposite to the surface is called the back surface. Here, a portion of the machining path and the outline of the component are referred to as the outline line. The outline line is the boundary between the component and the remaining material portion. Furthermore, the portion used to evaluate the machining status of a portion of the machining path and the outline line is called the evaluation path. In other words, the portion used by the machining status evaluation section 7, described later in the machining path, to determine the evaluation information EI related to the outline line is called the evaluation path. The evaluation path sometimes includes portions other than the outline line, that is, portions other than the boundary between the component and the remaining material portion.

[0036] This example illustrates the case where the evaluation path includes portions outside the outline. Figure 2 This is a diagram showing an example of the workpiece w in this embodiment. Figure 2The workpiece w is set as a rectangular mild steel material with a constant thickness. Figure 2 (a) The processing path x before processing is shown by a dashed line. Figure 2 (b) shows evaluation paths y1 and y2, and parts q1 and q2. The processing status evaluation unit 7, described later, determines evaluation information EI related to the contour lines of parts q1 and q2 based on the timing signals ts obtained through evaluation paths y1 and y2, respectively. That is, the processing status evaluation unit 7 can determine the evaluation information EI by obtaining the timing signal ts during processing of the evaluation path. The irradiation point is along the processing path x, from... Figure 2 (a) The laser L moves from point A to point B. At point A, the laser L begins to irradiate the workpiece w, performing a perforation to form a hole through the workpiece w. Next, the laser L irradiates the workpiece w while the irradiation point moves from point A to point B. During this process, the irradiation point traces a circular trajectory around part q1. It also traces a semi-circular trajectory around part q2. As described above, the workpiece w is cut into part q1, part q2, and the remaining material part r. Furthermore, the output of the laser L can be changed midway through the movement of the irradiation point along the processing path x. Additionally, the output of the laser L during perforation and the output of the laser L during cutting can also be changed. Figure 2 As shown in the description, the contour lines of evaluation path y1 and part q1 are not identical. Similarly, the contour lines of evaluation path y2 and part q2 are not identical. That is, the evaluation path includes not only the boundary between the part and the remaining material section, but also a portion of the processing path that is close to the boundary between the part and the remaining material section. Furthermore, in Figure 2 In the example, the outline of the part is part of the evaluation path, but conversely, the evaluation path can also be part of the outline.

[0037] also, Figure 1The control device 10 determines the oscillator command OC and drive command DC based on the processing program MPa and processing conditions PC, but it can also determine the oscillator command OC and drive command DC based on external inputs such as from the operator. Here, the processing program MPa is a program that defines the shape of the part. The processing conditions PC are conditions related to laser processing. The processing conditions PC include, for example, information related to the workpiece w, the laser oscillator 1, the optical path 2, the processing head 3, the drive unit 4, and the workpiece w. Examples of information related to the workpiece w include the material, size, shape, and surface condition of the workpiece w. In addition, examples of information related to the laser oscillator 1 include the output, the current value of the current supplied to the laser oscillator 1, the pulse repetition frequency, the duty cycle of the pulse output, the pulse waveform, and the wavelength. Furthermore, examples of information related to the optical path 2, the processing head 3, and the drive unit 4 include information related to the optical system 31, information related to the focal position of the laser L, the focusing diameter of the laser L, the distance between the workpiece w and the processing head 3, the type of processing gas, the pressure of the processing gas, the orifice diameter of the nozzle 32, the shape of the nozzle 32, the processing speed, and the environment during processing. Here, the processing speed can be defined as the relative speed between the processing head 3 and the workpiece w. Additionally, examples of the environment during processing include temperature and humidity.

[0038] in addition, Figure 1 The control device 10 determines the position information li. Position information li is information that specifies a position in the machining path. Position information li can be a combination of position data pd and identifier id. Position data pd can contain multiple structural elements, each corresponding to a position in the machining path. Furthermore, identifier id can be information specifying one or more structural elements from among the multiple structural elements. Additionally, identifier id can specify the range of structural elements from among the multiple structural elements. For example, the aforementioned structural elements can be set as data points corresponding to their respective positions in the machining path, and the data that lists the data points can be set as position data pd. Examples of data points include the number attached to the positions included in the machining path, the position in the machining path represented by coordinates, the distance value from the reference point in the machining path, and the time when the irradiation point passes through each position in the machining path. Identifier id can be information specifying a portion of the data points from the list.

[0039] Here, a portion of the data points can be a single location within the processing path, multiple discrete locations, or a range between two locations. Furthermore, the location data pd can also be instructions or programs such as oscillator commands OC, drive commands DC, processing procedures, or sorting procedures described later. Moreover, the structural elements of the location data pd can be set as commands or instructions, and the identifier id is information that specifies a portion of the structural elements. Furthermore, the location information li is preferably associated with the oscillator command OC or drive command DC. Figure 1 The control device 10 preferably generates position information li in association with oscillator command oc or drive command dc. By associating position information li with oscillator command oc or drive command dc, the parts, contours, etc., formed by the oscillator command oc and drive command dc can be more accurately specified. Furthermore, when a position on the machining path is specified using position information li, the position information li and drive command dc can be associated with each other when the drive command dc indicating the machining at that position can be determined.

[0040] Here, the methods for using the identifier `id` and the location data `pd` are unrestricted. The location data `pd` can be stored in any way that allows for the saving of a list of structural elements.

[0041] The structural elements are specified within. Additionally, the identifier `id` can be derived from the stored structural elements.

[0042] The list specifies one or more structural elements. For example, if the position data pd has a data number and is a column of data recorded in the storage device, the identifier id can be a signal that specifies the data number and the range of the data number. Alternatively, if the position data pd is an oscillator command OC, a drive command DC, or a machining program MPA, the identifier id can be a signal that specifies the desired position in the machining path, machining-related commands or instructions, etc. Furthermore, either or both of the position data pd and the identifier id can be pre-input into the control device 10, the evaluation information storage unit 8, the sorting action determination unit 9, the hand control device 12, etc. Moreover, it can also be configured to use a structure that only retrieves the identifier id.

[0043] In addition, Figure 1In the example, the position information li is determined by the control device 10, but the position information li can also be generated outside the laser processing system 1000. For example, the machining program mpa is configured to be generated based on CAD (Computer Aided Design) shape data specifying the shape of the part to be processed, and the identifier id related to the outline of the part is assigned to the machining program mpa outside the laser processing system 1000. Moreover, the machining program mpa can be used as position data pd. Here, the assigned identifier id can be read by the control device 10. Furthermore, CAM (Computer Aided Manufacturing) software can be used when creating the machining program mpa. In addition, the CAM software can generate a machining program mpa that can be read by the control device 10 based on the CAD shape data.

[0044] The detection unit 6 observes the processing status during laser processing according to a time sequence and determines the observation results as a time sequence signal ts. Examples of the processing status include intensity, wavelength, vibration frequency, or combinations thereof, related to the light emitted by irradiating the workpiece w with laser L (processing light), the sound emitted by irradiating the workpiece w with laser L (processing sound), and the reflected light from the workpiece w (reflected light). Here, processing sound may include ultrasound. Examples of the detection unit 6 include photodiodes, CCD (Charge Coupled Device) sensors, CMOS (Complementary Metal Oxide Semiconductor) sensors, spectrum splitters, sound sensors, vibration sensors, and combinations thereof.

[0045] Furthermore, the detection unit 6 detects the state of the laser processing apparatus 100 based on the processing state, thereby enabling a more detailed evaluation of the processing state. Examples of the state of the laser processing apparatus 100 include the atmosphere surrounding or inside the apparatus, the position, speed, and acceleration of the workpiece w or the processing head 3, and the angular velocity of the motor included in the drive unit 4. Examples of sensors for observing the state of the laser processing apparatus 100 include temperature sensors, humidity sensors, acceleration sensors, gyroscope sensors, distance sensors, and position detectors. Using these sensors, evaluation information EI can be determined based on the processing sound, processing light, and the movement of the workpiece. The detection unit 6 can be composed of one or more types of sensors. Furthermore, the detection unit 6 can be positioned in various locations. For example, the sensors constituting the detection unit 6 can be positioned inside or outside the processing head 3, on the surface side or the back side of the workpiece w, etc. Additionally, the detection unit 6 can be configured such that the light of the object being detected is transmitted to the sensor via an optical fiber. By using the detection equipment described above to construct the detection unit 6, evaluation information can be determined based on processing sounds, processing light, the movement state of the workpiece, etc. Therefore, the processing status can be observed, monitored, or judged more accurately.

[0046] The processing status evaluation unit 7 determines the evaluation information EI based on the result of evaluating the processing status according to the timing signal ts. The evaluation information storage unit 8 associates the evaluation information EI with the contour line based on the position information li and stores it as contour line evaluation information E. Here, as mentioned above, when the contour line of the part and the evaluation path of the part are different, the processing status evaluation unit 7 determines the evaluation information EI related to the contour line of the part based on the timing signal ts related to the evaluation path of the part. Regarding the processing status, the processing status evaluation unit 7 can determine the evaluation information EI based on a predetermined benchmark, and the evaluation information EI includes a judgment result j indicating whether it is better or worse than the benchmark. An example of judgment result j is given. If the part and the remaining material part r are in a state that can be separated, the judgment result j is considered good; if the part and the remaining material part r are in a state that cannot be separated, the judgment result j is considered bad, and it is determined whether the separation between the part and the workpiece w can be performed earlier or more accurately. Here, being able to separate and not being able to separate are respectively the states that the part can be taken out and the state that it cannot be taken out in the sorting device 200 described later. The processing status evaluation unit 7 can divide the processing path into multiple intervals and determine evaluation information EI for each interval. An interval can be obtained by dividing the processing path into units of a certain distance or a certain processing time. Alternatively, an interval can be obtained by dividing the processing path at locations where the direction of travel of the irradiation point changes. Alternatively, an interval can be obtained by dividing the processing path at points where the shape of the processing path changes. Alternatively, the interval can be set in the following way: that is, an evaluation value is determined for each position on the processing path. Furthermore, the processing path can be divided into intervals at locations where the evaluation value changes by a certain amount exceeding a threshold. Alternatively, the methods described above can be combined to set the intervals.

[0047] The line along the outline of the machined part in the processing path is called the outer perimeter of the part, and the line surrounding the portion of the outer perimeter and the excavated portion is called the inner perimeter. Furthermore, the outer and inner perimeter portions of the contour lines are called the outer perimeter contour line and the inner perimeter contour line, respectively. For the outer and inner perimeter contour lines, the processing status evaluation unit 7 determines evaluation information EI, and the evaluation information storage unit 8 can store the contour line evaluation information E. As described above, the sorting action determination unit 9, described later, can distinguish between the processing status related to the inner perimeter contour line and the processing status related to the outer perimeter contour line to determine the sorting action.

[0048] Furthermore, the processing status evaluation unit 7 can determine evaluation information EI, including the degree of processing defects. Additionally, the processing status evaluation unit 7 can determine evaluation information EI, including information that distinguishes between different types of processing defects. Moreover, the sorting device 200 can perform a sorting operation that more finely classifies parts, improving the efficiency of subsequent processes. Examples of the aforementioned processing defects include burning, scraping, slag, damage, upper surface roughness, middle surface roughness, lower surface roughness, oxide film detachment, and burning. Here, burning can refer to the phenomenon where the workpiece w burns more intensely near the irradiation point of the laser L than during normal processing, resulting in through holes, dents, etc., in the workpiece w. Scraping can refer to the phenomenon where molten material from the workpiece w is blown towards the surface of the workpiece before the laser reaches the back side of the workpiece w. Slag can refer to the phenomenon where molten material adheres to the part. Damage can refer to dents, protrusions, etc., that occur relative to the processing path. Here, damage can occur on the cross-section, surface, back side, etc., of the part. Top surface roughness, middle surface roughness, and bottom surface roughness can refer to the roughness of the surface, cross-section, and back surface of the workpiece w, respectively. Oxide film detachment can refer to the detachment of the oxide film formed on the cross-section of the workpiece w during processing, exposing the unoxidized surface. Burning can refer to the discoloration of the cross-section of the processed part. As described above, by classifying the processing status into defective items, it is also possible to perform sorting for each defective item.

[0049] The processing status evaluation unit 7 can evaluate the degree of defect for each type of processing defect using characteristic quantities, and can also evaluate the degree of processing defect using characteristic quantities. Furthermore, the processing status evaluation unit 7 can determine the evaluation information EI using the processing condition pc. Additionally, the processing status evaluation unit 7 can also calculate one or more characteristic quantities based on the time sequence signal ts, and determine the evaluation information EI based on the calculated characteristic quantities. For example, as characteristic quantities, the average value and standard deviation of a measurement value can be calculated, and the evaluation information EI representing the degree of processing defect can be determined based on the aforementioned average value and standard deviation. Furthermore, for example, the evaluation value of the evaluation information EI can be set to a value between 0 and 1; the worse the processing status, the larger the evaluation value can be.

[0050] Furthermore, the processing state evaluation unit 7 can obtain feature quantities from the time-series signal ts through statistical analysis, frequency analysis, filter bank analysis, wavelet transform, or a combination thereof. Additionally, when determining the evaluation information EI, the processing state evaluation unit 7 can use a classifier method for the feature quantities obtained from the time-series signal ts. Examples of classifier methods include linear discriminant analysis, logistic regression, support vector machines, correlation vector machines, and decision trees. Furthermore, the processing state evaluation unit 7 can use regression methods when determining the evaluation information EI. Examples of regression methods include linear regression, multinomial regression, Bayesian linear regression, and Gaussian process regression. Additionally, clustering methods can be used when evaluating the processing state. Examples of clustering methods include the K-means algorithm, Gaussian mixture distribution, and Bernoulli mixture distribution. Furthermore, the processing state evaluation unit 7 can evaluate the processing state using neural network methods. Examples of neural networks include deep neural networks, convolutional neural networks, and recurrent neural networks. Furthermore, known methods such as classification algorithms, clustering methods, regression algorithms, or combinations thereof can be used.

[0051] The sorting action determination unit 9 determines the sorting control instruction sc based on the contour evaluation information E. The sorting control instruction sc is an instruction to the sorting device 200 to perform sorting actions. Figure 1 In the example, the sorting action determination unit 9 determines the sorting action for M components from component p1 to component pM. The hand control device 12 acquires position information li and determines the hand drive command hd based on the sorting control command sc and the position information li. The component that will be the object of the sorting action is called the sorting object component. The sorting action determination unit 9 can acquire evaluation information EI related to multiple intervals contained in the outline of the sorting object component as outline evaluation information E. Moreover, the sorting control command sc can be determined based on the average value of multiple evaluation information EI. In addition, the sorting action determination unit 9 can acquire the worst evaluation information among the multiple evaluation information EI as outline evaluation information E and determine the sorting control command sc. As described above, the sorting action determination unit 9 can determine the sorting control command sc at the outline based on the evaluation result of the interval with the worst processing state. Moreover, it can avoid the stopping of the laser processing system 1000, unnecessary sorting actions, etc.

[0052] Furthermore, when the processing status evaluation unit 7 determines the evaluation information EI, including the judgment result j, for each interval, the sorting action determination unit 9 can determine the sorting control command sc based on the proportion of intervals with good judgment result j within the intervals contained by the outline of the sorting target component. For example, the sorting action determination unit 9 can perform a sorting action only for the sorting target component if the proportion of intervals with good evaluation information EI within the intervals contained by the sorting target component exceeds a predetermined value. Alternatively, the sorting action determination unit 9 can determine to perform a sorting action only if none of the intervals with bad judgment result j are contained within the intervals contained by the outline of the sorting target component. Additionally, when the processing status evaluation unit 7 determines the degree of defect for each item of processing defect as evaluation information EI, the sorting action determination unit 9 can determine to execute a classified sorting control command sc for each item of processing defect. Furthermore, the sorting action determination unit 9 can obtain position information li and outline evaluation information E, determine the sorting control command sc based on the position information li, and the hand control device 12 can determine the hand drive command hd based on the sorting control command sc.

[0053] Display unit 11 acquires display information di and displays it to operators, etc. Display information di can be information related to laser processing, such as information related to the laser processing device 100, information related to sorting operations, information related to the sorting device 200, etc. Figure 1 As shown, the display unit 11 can obtain information related to the processing status from the control device 10 and information related to the sorting status from the sorting device 200. Furthermore, the display unit 11 can be a human-machine interface (HMI) for operating the control device 10. Here, the HMI is a connection between a person and a machine that transmits instructions, information, etc. For example, in... Figure 1 In this system, the processing program MPa and processing condition PC are input from an external source to the control device 10. However, the display unit 11 can also be a touch panel, through which the specification of the processing program MPa, the input of the processing condition PC, and the display of the processing status can be performed. Furthermore, the display unit 11 can be installed in the sorting device 200, or it can be omitted from the laser processing system 1000. Information related to processing and sorting can be displayed to operators, supervisors, and users through the display unit 11, allowing them to be aware of the process.

[0054] The hand drive unit 13 changes the position of the hand 14 based on the hand drive command hd. Furthermore, the hand drive unit 13 switches between the state where the hand 14 grasps an object (grasping state) and the state where it does not grasp an object (non-grasping state) based on the hand drive command hd. Here, the position where the part is cut from the workpiece w through processing is called the gripping position. The position that will become the target for the movement of the part in the sorting operation is called the sorting position. Examples of the hand drive unit 13 include servo control devices including motors and position detectors, and robotic arms. Examples of the hand drive unit 13 include portal frames with hands 14 capable of position changes in three mutually orthogonal axes, and vertical multi-joint robots with hands 14. Examples of the hand 14 include vacuum suction cups, clamping members, grippers, permanent magnets, and electromagnets. Furthermore, in this invention, actions such as adsorption by magnets and vacuum adsorption are also called gripping.

[0055] After laser processing, tray 5 transports the workpiece w from the laser processing unit 100 to the sorting unit 200. Figure 1 In the diagram, a tray 5 is shown in the laser processing apparatus 100 and the sorting apparatus 200, respectively, for performing processing and sorting operations. There may be one or more trays 5. Alternatively, the sorting apparatus 200 may be configured to be located inside the laser processing apparatus 100, with a hand 14 provided in the sorting apparatus 200 grasping the parts on the tray 5 and moving them toward the part placement area 16. Figure 1 The diagram shows a cross-section of the remaining material section r and component pn+1 in the sorting device 200. Component pn is held by hand 14 during the sorting operation. Component pn+1 is shown in its state after processing is complete, before the sorting operation. Components p1 to pn-1, having completed processing and sorting, are placed in component placement area 16. Components pn+2 to pm are not yet in the sorting area. Figure 1 The workpiece w is shown in cross-section, therefore it is not illustrated. Furthermore, the sizes, shapes, etc., of parts p1 to pM can be different from each other or the same.

[0056] Figure 3 This is a flowchart illustrating an example of the sorting operation in this embodiment. In step S101, the hand drive unit 13 moves the hand 14 towards the gripping position. In step S102, the hand 14 grips the component in the gripping position. In step S103, the hand drive unit 13 maintains the gripping state and moves the hand 14 from the gripping position to the sorting position. In step S104, the hand 14 releases the gripping state and becomes a non-gripping state. The component is positioned at the sorting position of the component placement area 16. This is an example of the sorting operation.

[0057] Figure 4This is a diagram showing an example of the workpiece w in this embodiment. Figure 4 In the workpiece w, three components are shown: p1, p2, and p3. Components p1 and p2 have outer peripheral contour lines but no inner peripheral contour lines. Component p3 has both outer and inner peripheral contour lines. The identifiers id for the outer peripheral contour lines of components p1, p2, and p3 are set to No.1, No.2, and No.3, respectively. The identifier id for the inner peripheral contour line of component p3 is set to No.3-1. Component p3 connects to a portion of the remaining material portion r via the contour line identified by No.3. Additionally, component p3 connects to another portion of the remaining material portion r via the contour line identified by No.3-1.

[0058] Figure 5 This is a diagram illustrating an example of contour evaluation information in this embodiment. Figure 5 The evaluation information EI includes the judgment result j. The good evaluation information EI for the interval is defined as follows: when performing a sorting operation, the remaining material portion r and the component can be separated at the outline of the object. Conversely, the bad evaluation information EI is defined as follows: the remaining material portion r and the component cannot be separated. Figure 4 The intervals represented by the thin solid lines in the shown processing path are designated as having a "Good" evaluation information (EI). Conversely, the intervals represented by the thick solid lines are designated as having a "Poor" evaluation information (EI). Figure 5 The image shows the identifier `id` and the corresponding evaluation information `EI`. Figure 5 In the text, regarding the overall shape, a well-processed contour line is recorded as "pass." Conversely, a contour line that includes poorly processed areas is recorded as "failure." Figure 5 In the evaluation information EI corresponding to No.1 and No.3, it is pass; and the evaluation information EI corresponding to No.2 and No.3-1 is failure.

[0059] Figure 6 This diagram shows an example of the component placement location in this embodiment. Figure 6 The execution is shown in the figure. Figure 4 The state of the components contained in the workpiece w after the sorting action. Figure 6 The component placement area 16 shown has a qualified component placement area 161 for qualified components and a defective component placement area 162 for defective components. Component p1, which is considered a pass, is placed in the qualified component placement area 161. Component p3, which is considered a failure, is placed in the defective component placement area 162. Component p2 has a poor evaluation information (EI) for its outer perimeter contour line and cannot be removed; therefore, it does not move to component placement area 16. Figure 6Not shown. Furthermore, the method of placing components 16 is not limited, as long as components can be arranged separately. For example, it could be a multi-tiered shelf, multiple boxes, etc. Additionally, the remaining material section r can also be included in the sorting operation, categorized into three types: qualified products, unqualified products, and remaining material section r. This allows the remaining material section r to be easily discarded.

[0060] Figure 7 This is a flowchart illustrating an example of the operation of the sorting action determination unit 9 in this embodiment. (Using...) Figure 7 This illustrates the operation flow of the laser processing system 1000. Figure 7 In the example operation, before starting the operation process, the evaluation information storage unit 8 stores contour line evaluation information E. Furthermore, the sorting operation determination unit 9 determines the sorting operation for M components from component p1 to component pM. Here, a component number is assigned to the component to be sorted, and the component with component number i is called pi. Additionally, if a component has an inner periphery and an outer periphery, the evaluation information storage unit 8 stores evaluation information EI related to both the inner and outer periphery contour lines as contour line evaluation information E. In step S111, the sorting operation determination unit 9 initializes the component number i. For example, it can be done as follows: Figure 7 The part number i is set to 0. In step S112, the sorting action determination unit 9 increments the part number i by 1. In step S113, the sorting action determination unit 9 reads the contour evaluation information E, i.e., the contour evaluation information Ei, of part pi from the evaluation information storage unit 8. Next, in step S114, the sorting action determination unit 9 performs a judgment on whether the contour evaluation information E of the outer perimeter contour in the contour evaluation information E of part pi is good or bad. If the judgment is good, proceed to step S116. On the other hand, if the judgment is bad, proceed to step S115. As an example, regarding Figure 4 The operation of the sorting action determination unit 9 is explained for the workpiece w, in the case of executing step S114. Regarding parts p1 and p3, the outline evaluation information E of the outer perimeter contour is judged as good, and the process proceeds to step S116. On the other hand, regarding part p2, the outline evaluation information E of the outer perimeter contour is judged as bad, and the process proceeds to step S115.

[0061] In step S115, the sorting action determination unit 9 decides to stop the sorting action related to component pi and proceeds to step S121. In step S116, the sorting action determination unit 9 performs a determination on whether component pi has an inner circumferential contour line. The sorting action determination unit 9 can obtain the processing program MPa in advance and perform the determination in step S116 based on the processing program MPa. In step S116, if the sorting action determination unit 9 determines that component pi has an inner circumferential contour line, it proceeds to step S118. On the other hand, if the sorting action determination unit 9 determines in step S116 that component pi does not have an inner circumferential contour line, it proceeds to step S117. In step S117, the sorting action determination unit 9 decides not to perform a determination related to the inner circumferential contour line for component pi and proceeds to step S121. As an example, regarding component pi... Figure 4 The operation of the workpiece w shown in step S116 will be explained. Regarding component p1, the sorting operation determination unit 9 determines that it does not have an inner circumferential outline and proceeds to step S117, where it decides not to perform a determination related to the inner circumferential outline. Regarding component p3, the sorting operation determination unit 9 determines that it has an inner circumferential outline and proceeds to step S118.

[0062] In step S118, the sorting action determination unit 9 determines whether the contour evaluation information E related to the inner circumferential contour line is qualified, i.e., whether the processing state is qualified. In step S118, if the contour evaluation information E related to the inner circumferential contour line is determined to be good, the process proceeds to step S119. In step S119, the sorting action determination unit 9 considers part pi as having good contour evaluation information E related to the inner circumferential contour line, i.e., good inner circumference, and decides to perform sorting, proceeding to step S121. In step S118, if the contour evaluation information E related to the inner circumferential contour line is determined to be bad, the process proceeds to step S120. In step S120, the sorting action determination unit 9 considers part pi as having bad contour evaluation information E related to the inner circumferential contour line, i.e., bad inner circumference, and decides to perform sorting, proceeding to step S121. As an example, regarding... Figure 4 The operation of step S118 related to the workpiece w shown will be explained. The sorting operation determination unit 9 determines that the contour evaluation information E related to the inner circumferential contour of part p3 is defective and proceeds to step S120. Moreover, in step S120, the sorting operation determination unit 9 considers part p3 as having defective contour evaluation information E related to the inner circumferential contour and decides to sort it, proceeding to step S121.

[0063] In step S121, the sorting action determination unit 9 determines whether the part number i is greater than or equal to M. As mentioned earlier, M is the largest value among the part numbers i. In step S121, if it is determined that the part number i is less than M, the process proceeds to step S112. Furthermore, the sorting action determination unit 9 repeats the operation flow from step S112 to step S121 until it is determined in step S121 that the part number i is greater than or equal to M. In step S121, if it is determined that the part number i is greater than or equal to M, the sorting action determination unit 9 completes the determination of the sorting action related to the workpiece w. The above is related to... Figure 7 The following is an explanation of an example of the operation flow of the sorting action determination unit 9. The sorting action determination unit 9 can determine the sorting actions related to parts p1 to pM based on the contour evaluation information E, and determine the sorting control instructions sc, including instructions to instruct these sorting actions. The sorting control instructions sc can be determined separately for each part. Furthermore, in Figure 7 The flowchart illustrates the case where a component has one inner perimeter contour line, but in the case where a component has multiple inner perimeter contour lines, conditional branches related to each inner perimeter contour line can be built in.

[0064] In this embodiment, the sorting action determination unit 9 determines the sorting control command sc based on the contour evaluation information E. Therefore, it is possible to associate the evaluation of the processing status of the component and its contour with the decision of the sorting control command sc. Thus, for example, if the remaining material part r and the component cannot be separated, it can be determined not to perform the sorting action for that component. Moreover, even if the remaining material part and the component cannot be separated due to defects, the sorting device 200 can continue the sorting action. Furthermore, it is possible to reduce the assistance required by the operator and perform laser processing and sorting actions more automatically or autonomously. In addition, the amount of inspection required after laser processing can be reduced in order to perform the sorting action.

[0065] Furthermore, in this embodiment, the processing status evaluation unit 7 determines the evaluation information EI, and the evaluation information storage unit 8 stores the evaluation information EI and the contour line together as contour line evaluation information. Therefore, the sorting action determination unit 9 can reflect the processing status of each contour line and the processing status of each component in the classification of the sorting action. Moreover, the laser processing system 1000 can perform efficient sorting.

[0066] Furthermore, the processing status evaluation unit 7 can determine contour line evaluation information E, including information on identifying the inner and outer contour lines. Moreover, the sorting action determination unit 9 can reflect the information on identifying the inner and outer contour lines of the parts in the decision of the sorting action. Additionally, the processing status evaluation unit 7 determines evaluation information EI for each item of processing defect, and the sorting action determination unit 9 determines sorting control instructions sc for classifying parts for each item of processing defect. Furthermore, the processing status evaluation unit 7 can determine contour line evaluation information E, including information indicating whether the remaining material part r and the part can be separated, and the sorting action determination unit 9 determines sorting control instructions sc for parts that cannot be separated, indicating that the sorting action will not be performed. Additionally, the processing status evaluation unit 7 can determine evaluation information EI, including the degree of processing defect, and the sorting action determination unit 9 determines sorting control instructions sc, which instruct the parts to be classified according to the degree of processing defect. Furthermore, the processing status evaluation unit 7 can determine the contour evaluation information E, which includes information indicating the degree to which post-processing, such as additional processing, can mitigate processing defects occurring in each component. Moreover, the sorting action determination unit 9 can determine sorting control instructions sc that indicate the classification of components undergoing post-processing and those not. This can improve the efficiency of subsequent processes involving post-processing.

[0067] Through the aforementioned structure and actions, the laser processing system 1000 can perform sorting more efficiently. Furthermore, it can perform more precise sorting. Additionally, it can determine whether to perform a sorting action beforehand, thus creating a sorting device capable of more automated or autonomous operations. Moreover, the laser processing system 1000 can reduce the time and workload required for subsequent processes.

[0068] Figure 8 This diagram illustrates a structural example where the processing circuit of the laser processing system 1000 in this embodiment is composed of a processor 10001 and a memory 10002. When the processing circuit is composed of a processor 10001 and a memory 10002, each function of the processing circuit of the laser processing system 1000 is implemented through software, firmware, or a combination of both. The software, firmware, etc., are described as programs and stored in the memory 10002. In the processing circuit, the programs stored in the memory 10002 are read from the processor 10001 and executed, thereby implementing each function. That is, the processing circuit has a memory 10002, which stores the programs that ultimately execute the processing of the laser processing system 1000. Furthermore, these programs can be described as the sequence and method by which a computer executes the laser processing system 1000.

[0069] Here, processor 10001 can be a CPU (Central Processing Unit), processing device, arithmetic device, microprocessor, microcomputer, or DSP (Digital Signal Processor), etc. Memory 10002 can be, for example, non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable ROM), or EEPROM (Electrically EPROM). Alternatively, memory 10002 can be a disk, floppy disk, optical disk, compact disc, mini-disk, or DVD (Digital Versatile Disc), etc.

[0070] Figure 9 This diagram illustrates a structural example where the processing circuit of the laser processing system 1000 in this embodiment is constructed using dedicated hardware. When the processing circuit is constructed using dedicated hardware, Figure 9 The processing circuit 10003 shown can be, for example, a single circuit, a composite circuit, a programmable processor, a parallel programmable processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof. The functions of the laser processing system 1000 can be implemented by the processing circuit 10003 for each function, or multiple functions can be combined and implemented by the processing circuit 10003. The location where the processing circuit is installed is not limited to inside the laser processing apparatus 100 or the sorting device 200. For example, the processing circuit can be located away from the laser processing apparatus 100 and the sorting device 200, or it can be connected to the laser processing apparatus 100 and the sorting device 200 via a network. Furthermore, for example, the processing status evaluation unit 7, the evaluation information storage unit 8, the sorting action determination unit 9, the control device 10, and the hand control device 12 can be configured as the processing circuit. Moreover, the processing status evaluation unit 7 can be located inside or near the laser processing apparatus 100 and the sorting device 200. Furthermore, the structural elements of the laser processing system 1000 are not limited to... Figure 1 In this way, for example, the laser oscillator 1, optical path 2, processing head 3, drive unit 4, tray 5, sorting device 200, etc. can be set outside the laser processing system 1000.

[0071] As explained above, using the position information li enables the evaluation of the position in the processing path, the processing status, and the determination of the sorting action to correspond. Furthermore, in conjunction with the contour evaluation information E, the sorting action can be performed more efficiently or accurately. Moreover, according to this embodiment, a laser processing system capable of efficiently performing the sorting action of parts cut by laser cutting can be provided. Additionally, a laser processing system capable of automatically sorting parts without interrupting the operation can be provided even in the event of a defect during laser cutting. Furthermore, by associating the position information li with the oscillator command oc or the drive command dc, the contour line and evaluation information can be correlated, allowing for accurate determination of the sorting action.

[0072] Implementation method 2.

[0073] Figure 10 This is a diagram illustrating an example of the structure of the laser processing system in this embodiment. Based on the structural elements of the laser processing system 1000 of Embodiment 1, the laser processing system 1000a further includes a program generation unit 15, and instead of the sorting action determination unit 9, it has a sorting action determination unit 9a. Figure 1 The sorting action decision unit 9 determines the sorting control command sc based on the contour evaluation information E. On the other hand, Figure 10 The sorting action decision unit 9a determines the sorting control command sc based on the contour evaluation information E and the sorting procedure sp. Figure 10 In the explanation, regarding and Figure 1 The same or corresponding structural elements are labeled with the same number.

[0074] The program generation unit 15 generates a machining program MPA and a sorting program SP based on the machining shape data MSD. Furthermore, the program generation unit 15 assigns an identifier id to both the machining program MPA and the sorting program SP. Figure 10 In the laser processing system 1000a shown, the processing program MPa, the sorting program SP, and the identifier id are used as position information li. The processing program MPa and the sorting program SP are used respectively, thereby the evaluation information storage unit 8 and the sorting action determination unit 9a associate the sorted object component with the evaluation results related to the outline of the sorted object component.

[0075] Here, the program generation unit 15 can be, for example, a processing device that operates based on CAM (Computer-Aided Manufacturing) software. Additionally, the machining shape data msd can be, for example, data referred to as CAD (Computer-Aided Design) data. The machining shape data msd can be data describing the shape of a part cut from the workpiece w through machining. Furthermore, the machining program mpa, sorting program sp, and oscillator command oc or drive command dc are preferably associated.

[0076] Figure 11 This is a flowchart illustrating an example of the operation when the program generation unit 15 in this embodiment generates the processing program MPA. In step S201, the program generation unit 15 specifies the number of parts to be processed and the positions of the parts in the workpiece w based on the processing shape data MSD. In step S202, the program generation unit 15 generates a processing path based on the number and positions of the parts specified in step S201. For example, the program generation unit 15 can set a coordinate system in the workpiece w and generate the processing path using the set coordinate system. In step S203, the program generation unit 15 divides the processing path into contour lines. The program generation unit 15 also determines the evaluation path, intervals, etc. In step S204, the program generation unit 15 assigns an identifier id to each contour line. Furthermore, the processing program MPA may include information for identifying the inner and outer contour lines. Here, the program generation unit 15 can assign the identifier id to each part that becomes a sorting object using the same method as the control device 10 in Embodiment 1. Steps S202 to S204 can be executed in a different order, or they can be executed simultaneously. The above is an example of the action flow of the machining program MPA generated by the program generation unit 15.

[0077] Furthermore, in this embodiment, the combination of processing procedure MPa and identifier id, and the combination of sorting procedure SP and identifier id are used as location information li. Figure 10 In this context, the identifier `id` is assigned to the machining program `MPA`, therefore, the identifier `id` is not illustrated based on the machining program `MPA`. When the control device 10 determines the oscillator command `OC` and the drive command `DC` based on the machining program `MPA`, it associates the machining program `MPA` with the identifier `id`, the oscillator command `OC`, or the drive command `DC`. In other words, the control device 10 determines the oscillator command `OC` or the drive command `DC`, thereby performing machining according to the machining program `MPA` and the identifier `id`. That is, the component and its outline are formed according to the machining program `MPA` and the identifier `id`.

[0078] Figure 12This diagram illustrates an example of the operation when the program generation unit 15 generates the sorting program sp in this embodiment. In step S211, the program generation unit 15 determines the position and quantity of the parts based on the processing shape data msd. In step S212, the program generation unit 15 determines the gripping position for each part. In step S212, for example, the program generation unit 15 can set a coordinate system on the workpiece w and use the coordinate system to determine the position of each part. In step S213, the program generation unit 15 determines the sorting position of each part based on the position of each part determined in step S211. For example, the gripping position can be set as the position of hand 14 when gripping a part. In step S213, the program generation unit 15 determines the sorting position.

[0079] In step S214, the program generation unit 15 assigns an identifier id to the portion of the sorting program sp that instructs the processing of each contour line. That is, the contour line is assigned an identifier id. Here, the identifier id assigned to the sorting program sp is preferably associated with the identifier id assigned to the processing program mpa. In other words, in the processing program mpa and the sorting program sp, it is preferable to assign the same identifier id to the portions corresponding to the same contour line. By assigning an identifier id to the contour line, it is possible to associate the component performing the sorting operation with the processing status related to the contour line containing the component, which is then used to determine the sorting operation. In step S215, the program generation unit 15 sets a conditional branch corresponding to the contour line evaluation information E for each contour line in the sorting program sp. For example, the conditional branch corresponding to the contour line evaluation information E can be configured to set multiple sorting operation options in the sorting program sc, and the sorting operation determination unit 9a selects a sorting operation from the options accordingly with the contour line evaluation information E. By setting the conditional branch, the sorting operation can be changed accordingly with the contour line evaluation information E. Additionally, the sorting procedure (sp) can include information for identifying the inner and outer perimeter contours. Furthermore, conditional branches can be set for each component. The above is an example of the actions that generate the sorting procedure. Figure 12 The sequence of generating the sorting procedure sp shown is merely an example. The order of each step from step S211 to step S215 can be changed as needed, and some or all of them can be executed simultaneously.

[0080] Furthermore, in this embodiment, the sorting action determination unit 9a uses the sorting procedure sp and the identifier id as location information li. Figure 10 In this process, the sorting procedure sp is assigned an identifier id, therefore there is no graphic identifier id based on the sorting procedure sp. When deciding on the sorting action, the sorting action decision unit 9a uses the sorting procedure sp and the identifier id to associate the sorted object component, outline, and outline evaluation information E.

[0081] Additionally, the evaluation information storage unit 8 obtains the processing program mpa and identifier id generated by the program generation unit 15 and uses them as position information li. Based on the position information li, the evaluation information storage unit 8 associates the evaluation information EI with the contour line and stores it as contour line evaluation information E. Furthermore, the sorting action determination unit 9a obtains the sorting program sp and identifier id generated by the program generation unit 15 and uses them as position information li. Based on the position information li, the sorting action determination unit 9a determines the sorting control instruction sc. Furthermore, in Figure 10 In the example, the sorting action decision unit 9a obtains the sorting procedure sp and the identifier id as location information li, but as in embodiment 1... Figure 1 As shown in the example, the hand control device 12 can also obtain the position information li and determine the hand drive command hd based on the sorting control command sc and the position information li.

[0082] As described above, according to this embodiment, a laser processing system capable of efficiently performing the sorting operation of parts cut by laser cutting can be provided. Furthermore, a laser processing system capable of automatically sorting parts without interrupting the operation even in the event of a defect during laser cutting can be provided. Additionally, the laser processing system 1000a includes a program generation unit 15, which generates a processing program MPA and a sorting program SP based on processing shape data MSD and processing conditions PC. Figure 10 In the laser processing system 1000a shown, the processing program MPa, the sorting program SP, and the identifier id are used as position information li. Furthermore, by specifying the position on the processing path using the position information li, the contour line of the part, the evaluation result of the part's contour line, and the sorting action of that part are associated, enabling the evaluation of the processing status and the decision-making of the sorting action to be carried out in a coordinated manner. Moreover, in accordance with the contour line evaluation information E, the sorting action can be performed more efficiently or more accurately.

[0083] Implementation method 3.

[0084] Figure 13 This is a diagram illustrating an example of the structure of the laser processing system in this embodiment. Figure 13 The laser processing system 1000b shown includes a laser processing device 100b and a sorting device 200a. The laser processing device 100b replaces the one in Embodiment 1. Figure 1 The processing status evaluation unit 7 shown has a processing status evaluation unit 7a. Additionally, the sorting device 200a replaces... Figure 1 The hand control device 12 shown has a hand control device 12a. Figure 13 In the description, regarding the implementation method 1 Figure 1 Structural elements that are the same as or correspond to structural elements are labeled with... Figure 1 Same label.

[0085] Figure 14 This is a block diagram illustrating an example of the structure of the processing status evaluation unit in this embodiment. Figure 14 The processing status evaluation unit 7a shown includes a processing signal storage unit 71 for storing processing signals ps, a sorting determination unit 72 for determining sorting determination results sj, and a determination result storage unit 73 for storing contour line determination results cj. The processing status evaluation unit 7a also includes a status quantity observation unit 74 for observing status quantities sq, a learning unit 75 for performing learning, and an intention determination unit 76 for determining evaluation information EI.

[0086] The processing signal storage unit 71 acquires timing signal ts and position information li, and associates timing signal ts or feature quantities generated based on timing signal ts with contour lines based on position information li, storing it as processing signal ps. The sorting determination unit 72 determines whether the removal of a component is successful or unsuccessful based on gripping state information GS. The sorting determination unit 72 can acquire a manual drive command hd when associating gripping state information GS with a component, and use the manual drive command hd based on or replacing position information li. The determination result is called sorting determination result sj. Here, the action of moving a component from its gripping position by gripping it with hand 14 is called removal. Cases where component removal fails include situations where the component and remaining material r are joined due to the adhesion of molten material, etc., or where the component and remaining material r are not completely severed. The determination result storage unit 73 associates sorting determination result sj with contour lines based on position information li, storing it as contour line determination result cj. The judgment result storage unit 73 preferably associates the sorting judgment result sj with the outline of the part that is the judgment object or the outer periphery outline of the part that is the judgment object.

[0087] Figure 13 The hand control device 12a is in Figure 1Based on the operation of the hand control device 12, the gripping state information gs is determined. The gripping state information gs is the information used by the sorting determination unit 72 to determine whether the retrieval was successful. The operation of the hand control device 12a and the operation of the sorting determination unit 72 are illustrated below. The hand control device 12a stores the weight of the component as the component weight in advance, calculates the difference between the weight actually gripped by the hand 14 and the component weight, and sets the ratio of the absolute value of the difference to the stored component weight as the gripping state information gs. The sorting determination unit 72 can determine that the retrieval of the component by the hand 14 was successful only if the value of the gripping state information gs is less than or equal to a predetermined threshold. Here, the weight actually gripped by the hand 14 can also be estimated based on the driving force, torque, etc. generated by the hand drive unit 13. In addition, the component weight can also be calculated based on the volume and specific gravity calculated by the shape of the component. For example, the component weight can be calculated based on the plate thickness of the workpiece w, the specific gravity of the workpiece w, and the area of ​​the part enclosed by the outline of the component. Furthermore, as another example, the weight that hand 14 can grasp is set to a value between the weight of the component and the total weight of the component and the remaining material part r. Moreover, the information indicating whether hand 14 is in a grasping state can be determined as grasping state information gs.

[0088] Furthermore, the sorting and determination unit 72 can obtain gripping state information gs from devices other than the hand control device 12a. Examples of sensors that determine the gripping state information gs include: For instance, if the gripping mechanism of the hand 14 is a vacuum suction cup, a sensor that detects the airflow through the vacuum suction cup can be used. Alternatively, a switch that toggles between the energized and de-energized states of the hand 14 in gripping and non-gripping states can be used as the aforementioned sensor. Additionally, a distance sensor that detects the distance between the gripping mechanism and the component of the hand 14 can be used. Furthermore, a proximity sensor that detects the presence of an object within a predetermined distance range can be used. Alternatively, a light source such as an LED (Light Emitting Diode) and a light sensor that detects the presence or absence of an object by reflecting light emitted from the light source onto the object can be used. Finally, a camera that captures images of the gripping mechanism of the hand 14 can also be used as the aforementioned sensor.

[0089] The state quantity observation unit 74 observes the state quantity sq. Here, the state quantity sq includes the processing signal ps and the contour determination result cj, which are quantities related to laser processing or sorting operations. The learning unit 75, based on the state quantity sq, estimates the contour determination result cj according to the processing signal ps, or performs learning for decision-making. The learning unit 75 can perform learning using various learning algorithms. For example, known learning algorithms such as reinforcement learning, teacher-assisted learning, teacherless learning, and semi-teacher-assisted learning can be applied. As an example, the case where the learning unit 75 learns the estimation function Es, which outputs the contour determination result cj relative to the input of the processing signal ps, will be explained. Figure 15 This is a block diagram illustrating an example of the structure of the learning unit 75 in this embodiment. The learning unit 75 includes an error calculation unit 751 that calculates the error er and an estimation function update unit 752 that updates the estimation function Es based on the error er. Here, the error er is the error between the estimated value of the contour line determination result cj and the measured value of the contour line determination result cj. The estimation function update unit 752 updates the estimation function Es based on the state variable sq in a manner that the error er decreases.

[0090] Furthermore, the learning unit 75 can utilize a neural network model to learn the relationship between the contour determination result cj and the processing signal ps through teacher-guided learning. Here, by providing a large set of input and result (label) data to the learning device, the model that learns the features existing in these data sets and infers the result based on the input is called teacher-guided learning. Figure 16 This diagram illustrates an example of the structure of the neural network in this embodiment. The neural network consists of an input layer composed of multiple neurons, an intermediate layer (hidden layer) composed of multiple neurons, and an output layer composed of multiple neurons. The intermediate layer can be one layer or more than or equal to two layers.

[0091] For ease of understanding, Figure 16 The neural network has 3 inputs and 3 layers. If multiple inputs are fed into the input layer consisting of X1 to X3, the input value is multiplied by weights W1 consisting of w11 to w16, and the resulting value is fed into the intermediate layer consisting of Y1 and Y2. Furthermore, the input value to the intermediate layer is multiplied by weights W2 consisting of w21 to w24, and the resulting value is output from the output layer consisting of Z1 and Z2. This output varies depending on the values ​​of weights W1 and W2. Figure 16The neural network shown learns the relationship between the processing signal ps and the contour determination result cj using a dataset created based on the state variable sq, through teacher-guided learning. Specifically, the learning unit 75 inputs the processing signal ps into the input layer and adjusts the weights W1 and W2 in a way that reduces the error between the contour determination result cj output from the output layer and the actual obtained contour determination result cj. When the learning unit 75 performs learning using the neural network model, the intention determination unit 76 can obtain the neural network model obtained from the learning unit 75 as a result of the learning process.

[0092] The state quantity observation unit 74 can observe quantities other than the processing signal ps and the contour determination result cj, and use them as state quantities sq. For example, quantities related to the state of the workpiece w, the state of the laser processing device 100, the state of the sorting device 200, etc., can be included in the state quantity sq. Figure 17 This is a block diagram illustrating an example of the structure of the processing status evaluation unit in this embodiment. Figure 17 China regarding Figure 14 Same or corresponding structural elements, labeled with Figure 14 Same label. Figure 17 The laser processing system 1000b includes a processing surface image acquisition and evaluation unit 80. The processing surface image acquisition and evaluation unit 80 captures images of the processing surface of the component, analyzes the captured images, and evaluates the state of the processing surface. Furthermore, the evaluation result is determined as processing surface image information msi. Examples of the state of the processing surface include the presence or absence of processing defects, the type of processing defect (i.e., the item of the processing defect), and the degree of processing defect. A state quantity observation unit 74 observes the processing surface image information msi based on the processing signal ps and the contour line determination result cj, and obtains a state quantity sq. Furthermore, a learning unit 75 performs learning based on the state quantity sq containing the processing surface image information msi, to determine the contour line determination result cj according to the processing signal ps. The learning unit 75 can use the processing surface image information msi during learning. Moreover, the intention determination unit 76 can more accurately determine the contour line determination result cj and evaluation information EI by utilizing the learning result using the processing surface image information msi. Furthermore, the processing signal ps related to the component performing the sorting operation in the state quantity sq is preferably obtained before the sorting operation is determined. Additionally, the processing surface image capture and evaluation unit 80 can be a structural element of the laser processing system 1000b. Furthermore, the capturing, analysis, and input of images of the processing surface by the processing surface image capture and evaluation unit 80 can be partially performed by the operator.

[0093] The learning unit 75 can learn a model that outputs a contour determination result cj in response to the input of the processing signal ps. Examples of models learned by the learning unit 75 as described above include a classification model that categorizes success and failure, and a regression model that sets success to 1 and failure to 0. Alternatively, the processing condition pc can be used as part of the state variable sq to learn a model that allows for more accurate determination. For example, the state variable observation unit 74 can observe the processing condition pc associated with the contour and use it as part of the state variable sq.

[0094] As described above, according to this embodiment, a laser processing system can be provided that can efficiently perform the sorting operation of parts cut by laser cutting. Furthermore, during the sorting operation, remaining material and parts can be efficiently separated. Additionally, a laser processing system can be provided that can automatically or autonomously sort parts without interrupting the sorting operation even in the event of a defect during laser cutting.

[0095] Furthermore, the laser processing system 1000b includes a learning unit 75. The learning unit 75 performs learning based on a state variable sq, which includes the processing signal ps and the contour determination result cj, related to the laser processing or sorting operation. This learning determines the contour determination result cj based on the processing signal ps. Therefore, the evaluation information EI can be determined in advance or accurately using the learning results. Additionally, if the relationship between the contour determination result cj and the processing signal ps is learned, the success of sorting operations, such as picking, can be more accurately estimated based on the state variable sq. Furthermore, the processing signal storage unit 71 stores the processing signal ps, allowing the learning unit 75 to perform learning using more data.

[0096] Implementation method 4.

[0097] Figure 18 This is a diagram illustrating an example of the structure of the laser processing system in this embodiment. Figure 18 The laser processing system 1000c shown includes a laser processing apparatus 100c and a sorting device 200. The laser processing apparatus 100c replaces the one in Embodiment 1. Figure 1 The processing status evaluation unit 7 shown includes a processing status evaluation unit 7c. The processing status evaluation unit 7c is used for... Figure 13 and Figure 14 The explanation refers to a well-trained learning machine after learning. Figure 19 This diagram illustrates an example of the processing status evaluation unit in this embodiment. The processing status evaluation unit 7c includes a processing signal storage unit 71, a processing status quantity observation unit 77, and an intention determination unit 76a. Figure 19 In the middle, to and Figure 14 The same or corresponding structural elements are labeled with the same number.

[0098] Figure 19 The processing state quantity observation unit 77 includes a processing signal ps and observes quantities related to processing or sorting as processing state quantity v. Furthermore, the processing state quantity v may include quantities contained in the state quantity sq, excluding the determination result cj and the processing signal ps. Next, the intention determination unit 76a has the function of executing... Figure 14 The learning unit 75 updates the estimation function Es. This trained estimation function Es or the intention determination unit 76a can be set as a trained learning machine. The intention determination unit 76a determines the judgment result cj and evaluation information EI based on the processing state quantity v and the processing signal ps. Furthermore, as in Embodiment 3... Figure 17 As shown, a trained learning machine can be constructed, which performs learning using the processing surface image information msi from the processing surface image capture evaluation unit 80.

[0099] As described above, according to this embodiment, a laser processing system can be provided that can efficiently perform the sorting operation of parts cut by laser cutting. Furthermore, during the sorting operation, remaining material and parts can be efficiently separated. Additionally, a laser processing system can be provided that can automatically or autonomously sort parts without interrupting the sorting operation even in the event of a defect during laser cutting.

[0100] Furthermore, the laser processing system 1000c has a trained learning machine, namely the intention determination unit 76a. Therefore, the processing state evaluation unit 7c does not perform learning, and can use the learning results to determine the contour judgment result cj and evaluation information E based on the processing state quantity v. Moreover, by using the trained learning machine, the learning results can be applied to other laser processing systems different from the laser processing system that performs the learning. Furthermore, it is possible to provide a laser processing system that can save the time of each laser processing system performing learning independently and can achieve efficient sorting operations.

[0101] Explanation of the label

[0102] 1. Laser oscillator; 4. Drive unit; 6. Detection unit; 7, 7a, 7c. Processing status evaluation unit; 8. Evaluation information storage unit; 9, 9a. Sorting action determination unit; 10. Control device; 11. Display unit; 15. Program generation unit; 71. Processing signal storage unit; 72. Sorting judgment unit; 73. Judgment result storage unit; 74. Status quantity observation unit; 75. Learning unit; 76, 76a. Intention determination unit; 77. Processing status quantity observation unit; 1000, 1000a, 1000b, 1000c. Laser processing system; dc. Drive command; E. Contour line evaluation information; EI. Evaluation information; id. Identifier; L. Laser; li. Position information; mpa. Processing program; oc. Oscillator command; pd. Position data; ps. Processing signal; r. Remaining material unit; sc. Sorting control command; sj. Sorting judgment result; sq. Status quantity; ts. Timing signal; v. Processing status quantity; w. Workpiece.

Claims

1. A laser processing system, characterized in that, have: A control device controls a laser oscillator that emits a laser and a drive unit that moves the position of the workpiece irradiated by the laser, i.e., the irradiation point, along the processing path to perform processing that cuts the workpiece into parts and remaining material. The detection department determines the time sequence signal by observing the state of the workpiece during the processing in a time sequence. The processing status evaluation unit determines the evaluation information by evaluating the processing status based on the time sequence signal for each interval into which the processing path is divided. An evaluation information storage unit associates the boundary, i.e., the contour line, between the component and the remaining material portion with the evaluation information and stores it as contour line evaluation information; and The sorting action determination unit determines the sorting control instruction based on the contour evaluation information. This sorting control instruction is an instruction to control the sorting action of taking the part out from the position where the workpiece is processed and moving it to the target position.

2. The laser processing system according to claim 1, characterized in that, The evaluation information storage unit uses position information—specifically, location information—to specify the position within the processing path, thus associating the contour line with the evaluation information. The location information is data containing multiple structural elements, including location data that associates each structural element with multiple locations in the processing path, and identifiers that specify one or more of the multiple structural elements.

3. The laser processing system according to claim 2, characterized in that, The laser oscillator emits the laser based on oscillator commands, and the drive unit moves the irradiation point along the processing path based on drive commands. The control device determines the oscillator command and the drive command in a manner that associates the position information with at least one of the oscillator command and the drive command.

4. The laser processing system according to claim 3, characterized in that, It has a program generation department that generates processing programs and sorting programs. The control device determines the oscillator command and the drive command based on the processing program. The sorting action decision unit determines the sorting control command based on the sorting procedure. The evaluation information storage unit uses the processing procedure as the location data.

5. The laser processing system according to any one of claims 1 to 4, characterized in that, When the component has an outer peripheral contour (i.e., an outer peripheral contour line) and an inner peripheral contour (i.e., an inner peripheral contour line), The processing status evaluation section determines the evaluation information based on the outer peripheral contour line and the inner peripheral contour line, respectively. The evaluation information storage unit associates the outer peripheral contour line and the inner peripheral contour line with the evaluation information and stores them as contour line evaluation information.

6. The laser processing system according to any one of claims 1 to 4, characterized in that, The sorting action decision unit determines whether to perform a sorting action or not based on the contour evaluation information for each of the components.

7. The laser processing system according to any one of claims 1 to 4, characterized in that, If the remaining material portion and the component can be separated at the outline, the processing status evaluation unit determines that the processing status is good with respect to the outline; if the remaining material portion and the component cannot be separated at the outline, the processing status is poor with respect to the outline, and determines the evaluation information including the determination result.

8. The laser processing system according to any one of claims 1 to 4, characterized in that, The processing status evaluation unit determines the evaluation information, which includes information on evaluating the processing status of at least one of the following processing defects: combustion, scraping, slag, damage, upper surface roughness, middle surface roughness, lower surface roughness, oxide film detachment, and burning.

9. The laser processing system according to any one of claims 1 to 4, characterized in that, The detection unit includes at least one of the following: a sound sensor, a light sensor, a camera, a vibration sensor, an accelerometer, and a gyroscope sensor.

10. The laser processing system according to any one of claims 1 to 4, characterized in that, It has a display unit that displays information related to the sorting action or the processing.

11. The laser processing system according to any one of claims 1 to 4, characterized in that, The processing status evaluation unit has: A processing signal storage unit stores at least one of the timing signal or a feature quantity extracted from the timing signal as a processing signal by associating it with the contour line. The sorting and determination unit determines the sorting result by judging whether the sorting action of the component is successful or not. The determination result storage unit associates the sorting determination result with the contour line and stores it as a contour line determination result; The state quantity observation unit observes quantities that include the processing signal and the contour determination result and are related to the processing or sorting action, and uses them as state quantities. The learning unit, based on the state quantity, performs learning to determine the contour determination result according to the processing signal; as well as The intention-determination unit determines the evaluation information based on the results learned by the learning unit and according to the processing signal.

12. The laser processing system according to any one of claims 1 to 4, characterized in that, The processing status evaluation unit has: A processing signal storage unit stores at least one of the timing signal or a feature quantity extracted from the timing signal as a processing signal by associating it with the contour line. A processing status quantity observation unit observes quantities containing the processing signal and related to laser processing or sorting operations as processing status quantities; and The trained learning machine has learned to determine the evaluation information based on the processing signal, based on quantities related to the processing or sorting action, including the sorting determination result obtained from the judgment of success or failure of the sorting action of the component and the processing signal.

Images