# US 12,766,642 B2 · Laser processing apparatus and laser processing method > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,766,642 B2 - Document ID: US12766642B2 - Status: Granted - Kind: B2 - Application number: 18/281,565 - Filed: 2021-04-06 - Published: 2024-09-26 - Granted: 2026-10-20 - Earliest claimed: 2021-04-06 - Assignees: Mitsubishi Electric Corporation - Applicants: Mitsubishi Electric Corporation - Inventors: Kenta Fujii; Kyohei Ishikawa; Nobuaki Tanaka; Motoaki Nishiwaki - Web page: https://patentgrep.com/patent/US12766642B2 ## CPC - B23K 31/125: {Weld quality monitoring} - B23K 26/03: Observing, e.g. monitoring, the workpiece - B23K 26/032: {using optical means} - B23K 26/705: {Beam measuring devices} - B23K 26/707: {for monitoring laser beam transmission optics} - B23K 31/006: {relating to using of neural networks} - G05B 13/027: {using neural networks only} - G05B 2219/45041: Laser cutting - B23K: SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM - G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G) ## Abstract A laser processing apparatus includes a control device controlling laser processing in accordance with processing condition; a detection unit observing processing state on a time-series basis during laser processing and outputting a signal corresponding to observation result as a time-series signal; a parameter estimation unit estimating and outputting, based on first time-series signal data including the time-series signal and/or a first feature computed from the time-series signal and preset corresponding parameter information indicating relationship between the first time-series signal data and at least one parameter value included in the processing condition, a parameter estimate corresponding to the first time-series signal data; a condition change amount determination unit comparing a target parameter value and the parameter estimate and determining a change amount for the processing condition based on comparison result; and a parameter update unit updating the processing condition based on the change amount and the processing condition. ## Claims ### Claim 1 A laser processing apparatus comprising: control circuitry configured to control laser processing in accordance with a processing condition for the laser processing; a detector configured to observe, on a time-series basis, at least one of sound or light generated during execution of the laser processing as a processing state during execution of the laser processing and output a signal corresponding to an observation result as a time-series signal; parameter estimation circuitry configured to estimate, on a basis of first time-series signal data and corresponding parameter information, a parameter value, included in the processing condition and targeted for update, that corresponds to the first time-series signal data, and output the estimated parameter value as a parameter estimate, the first time-series signal data including at least one of the time-series signal or a first feature computed from the time-series signal, the corresponding parameter information being preset in the parameter estimation circuitry and being information indicating correspondence between the first time-series signal data and the parameter value included in the processing condition and targeted for update; condition change amount determination circuitry configured to compare a target parameter value that is a target value for the parameter value and the parameter estimate and determine an amount of change for the parameter value of the processing condition to bring the parameter estimate closer to the target parameter value, on a basis of a comparison result; and parameter update circuitry configured to update the processing condition by changing the parameter value on a basis of the amount of change and the processing condition, wherein the control circuitry is configured to control the laser processing in accordance with the processing condition including the parameter value updated by the parameter update circuitry, and wherein during the laser processing, the detector, the parameter estimation circuitry, the condition change amount determination circuitry, and the parameter update circuitry are configured to sequentially and repeatedly perform detection of the time-series signal, output of the parameter estimate based on the corresponding parameter information that is preset, determination of the amount of change, and update of the parameter value of the processing condition. ### Claim 2 (depends on 1) The laser processing apparatus according to claim 1, comprising: processing state evaluation circuitry configured to output an evaluation value evaluating the processing state on a basis of second time-series signal data including at least one of the time-series signal or a second feature computed from the time-series signal; target change amount determination circuitry configured to compare the evaluation value and a target evaluation value that is a preset target value for the evaluation value and determine a parameter target change amount that is an amount of change for the target parameter value on the basis of the comparison result; and target value update circuitry configured to update the target parameter value on a basis of the parameter target change amount. ### Claim 3 (depends on 1) The laser processing apparatus according to claim 1, wherein the corresponding parameter information is information in which correspondence is set between the first time-series signal data detected by the detector when the laser processing is performed using a plurality of the processing conditions having different values specified for a parameter to be updated and the parameter value used when the first time-series signal data is detected. ### Claim 4 (depends on 3) The laser processing apparatus according to claim 3, wherein the corresponding parameter information is a function indicating correspondence between the first time-series signal data and the parameter value. ### Claim 5 (depends on 4) The laser processing apparatus according to claim 4, wherein the function is a regression model trained by regression learning to output the parameter value that corresponds to the first time-series signal data when the first time-series signal data is input, and the parameter estimation circuitry takes an output obtained by inputting the first time-series signal data detected by the detector during processing to the regression model as the parameter estimate. ### Claim 6 (depends on 3) The laser processing apparatus according to claim 3, wherein the parameter estimation circuitry is configured to learn correspondence between the first time-series signal data detected by the detector when the laser processing is performed using the plurality of the processing conditions having different values specified for the parameter and the parameter value that corresponds to the processing condition in generating the corresponding parameter information. ### Claim 7 (depends on 3) The laser processing apparatus according to claim 3, wherein the corresponding parameter information is learned information representing a relationship between the parameter value and the first time-series signal data stored in a database, the first time-series signal data and the parameter value being obtained by performing the laser processing using a plurality of parameter values within a configurable range of the parameter value. ### Claim 8 (depends on 7) The laser processing apparatus according to claim 7, wherein the database further stores processing results, each processing result being a determination result of the laser processing performed using a corresponding one of the plurality of parameter values within the configurable range, and the target parameter value is selected from among the parameter values based on the processing results, each processing result including at least one of a determination of whether a processed result is good or bad, a symptom of a defect, or a degree of defectiveness. ### Claim 9 (depends on 3) The laser processing apparatus according to claim 3, wherein the corresponding parameter information includes information in which correspondence is set between the first time-series signal data detected by the detector when the laser processing is performed using the plurality of processing conditions having different values specified for the parameter to be updated, and a set of parameter values used when the first time-series signal data is detected, wherein the first time-series signal data and the set of parameter values are included in a selected data set selected based on a processing result recorded for the laser processing. ### Claim 10 (depends on 9) The laser processing apparatus according to claim 9, wherein the corresponding parameter information is a function indicating correspondence between the first time-series signal data and the parameter value. ### Claim 11 (depends on 10) The laser processing apparatus according to claim 10, wherein the function is a regression model trained by regression learning to output the parameter value that corresponds to the first time-series signal data when the first time-series signal data is input, and the parameter estimation circuitry takes an output obtained by inputting the first time-series signal data detected by the detector during processing to the regression model as the parameter estimate. ### Claim 12 (depends on 9) The laser processing apparatus according to claim 9, wherein the parameter estimation circuitry is configured to learn correspondence between the first time-series signal data detected by the detector when the laser processing is performed using the plurality of the processing conditions having different values specified for the parameter and the parameter value that corresponds to the processing condition in generating the corresponding parameter information. ### Claim 13 (depends on 9) The laser processing apparatus according to claim 9, wherein the processing result includes at least one of a determination of whether a processed result is good or bad, a symptom of a defect, or a degree of defectiveness. ### Claim 14 (depends on 13) The laser processing apparatus according to claim 13, wherein the corresponding parameter information is a function indicating correspondence between the first time-series signal data and the parameter value. ### Claim 15 (depends on 13) The laser processing apparatus according to claim 13, wherein the parameter estimation circuitry is configured to learn correspondence between the first time-series signal data detected by the detector when the laser processing is performed using the plurality of the processing conditions having different values specified for the parameter and the parameter value that corresponds to the processing condition in generating the corresponding parameter information. ### Claim 16 (depends on 1) The laser processing apparatus according to claim 1, wherein the condition change amount determination circuitry is configured to determine that a value obtained by multiplying a difference between the target parameter value and the parameter estimate by a preset value greater than 0 and smaller than or equal to 1 be an amount of change for the processing condition. ### Claim 17 (depends on 1) The laser processing apparatus according to claim 1, wherein the condition change amount determination circuitry is configured to change over from the amount of change determined for the processing condition to an upper limit preset when the amount of change for the processing condition is larger than the upper limit, and change over from the amount of change determined for the processing condition to a lower limit preset when the amount of change for the processing condition is smaller than the lower limit. ### Claim 18 (depends on 1) The laser processing apparatus according to claim 1, wherein the parameter update circuitry is configured to compute the parameter value of the processing condition on the basis of the amount of change and the processing condition, change over from the parameter value computed to an upper limit preset when the parameter value is larger than the upper limit, and change over from the parameter value computed to a lower limit preset when the parameter value is smaller than the lower limit. ### Claim 19 (depends on 1) The laser processing apparatus according to claim 1, wherein a parameter included in the processing condition includes at least one of output intensity of a laser light, an output frequency of the laser light, an output duty ratio of the laser light, processing speed, focal position of the laser light relative to a workpiece, pressure of a processing gas, a height where a nozzle is to be as measured from the workpiece, or beam magnification of the laser light. ### Claim 20 A laser processing method for laser processing to be performed in accordance with a processing condition for the laser processing, the laser processing method comprising: a detection of observing, on a time-series basis, at least one of sound or light generated during execution of the laser processing as a processing state during execution of the laser processing and outputting a signal corresponding to an observation result as a time-series signal; a parameter estimation of estimating, on a basis of first time-series signal data and corresponding parameter information, a parameter value, included in the processing condition and targeted for update, that corresponds to the first time-series signal data, and outputting the estimated parameter value as a parameter estimate, the first time-series signal data including at least one of the time-series signal or a first feature computed from the time-series signal, the corresponding parameter information being preset and information indicating a relationship correspondence between the first time-series signal data and at least the parameter value included in the processing condition and targeted for update; a condition change amount determination of comparing a target parameter value that is a target value for the parameter value and the parameter estimate and determining an amount of change for the parameter value of the processing condition to bring the parameter estimate closer to the target parameter value, on a basis of a comparison result; and a parameter update of updating the processing condition by changing the parameter value on a basis of the amount of change and the processing condition, wherein the laser processing is controlled in accordance with the processing condition including the parameter value updated in the parameter update, and wherein during the laser processing, detection of the time-series signal, output of the parameter estimate based on the corresponding parameter information that is preset, determination of the amount of change, and update of the parameter value of the processing condition are performed sequentially and repeatedly in the detection, the parameter estimation, the condition change amount determination, and the parameter update.