Galvanized sheet welding collaborative control system based on vapor disturbance perception

Abstract

The application discloses a galvanized sheet welding collaborative control system based on vapor disturbance sensing, and relates to the technical field of welding control.The system comprises a vapor monitoring, collecting and processing module, which collects welding vapor monitoring data and performs preprocessing on the welding vapor monitoring data;an escape channel inversion identification module, which calculates a vapor stagnation pressure escape mismatch value, determines a vapor disturbance enhanced section, and outputs a vapor escape channel inversion result;an overpressure curve fusion generation module, which generates each correction curve, fuses the overlapping correction curves according to the vapor stagnation pressure escape mismatch value proportion, and outputs overpressure type welding collaborative control parameters;and a welding parameter matching execution module, which reads the overpressure type welding collaborative control parameters and outputs target control parameters to corresponding welding execution mechanisms.The problems that the zinc vapor stagnation pressure escape position is difficult to accurately identify in the existing galvanized sheet welding process, the welding parameter adjustment is delayed, and the local weld forming quality is difficult to stably control are solved.

Description

Technical Field

[0001] This invention relates to the field of welding control technology, specifically to a collaborative control system for galvanized sheet welding based on steam disturbance sensing. Background Technology

[0002] With the increasing demand for automobile manufacturing, home appliance housings, box-type structural components, and thin-plate coverings, galvanized steel sheets are widely used in welding applications due to their combination of base strength and surface corrosion resistance. Because the galvanized layer is prone to vaporization under the heat of welding, the welding process of galvanized steel sheets typically requires consideration of workpiece positioning accuracy, welding heat input stability, droplet transfer state, and coating vapor discharge conditions. Therefore, various technical solutions have been proposed to optimize the structure of galvanized steel sheet welding equipment, control the welding process, and improve weld quality.

[0003] For example, application CN121670236A discloses a multi-dimensional precision welding device and method for galvanized steel sheets, relating to the field of welding technology for galvanized steel sheets. This multi-dimensional precision welding device for galvanized steel sheets includes: a welding robot; a rotating platform for supporting the galvanized steel sheet workpiece to be welded and capable of active rotation; and positioning plates positioned above the rotating platform, with multiple sets evenly arranged circumferentially along the platform. The positioning plates are divided into an inner ring group and an outer ring group, and the two sets of positioning plates can move radially towards or away from each other to spatially align and clamp multiple galvanized steel sheets, assembling them into a prefabricated box structure.

[0004] For example, application CN120619524A discloses a welding control method, device, and storage medium for galvanized steel sheets in the field of welding technology. The method includes: periodically acquiring the arc duration and short-circuit release current; when the arc duration is greater than a preset standard arc duration and the short-circuit release current is greater than a preset standard short-circuit release current, it is determined to be a waveform for suppressing droplet transition; when the waveform is a waveform for suppressing droplet transition, the actual output voltage within a unit cycle is controlled to be less than a preset voltage, and the wire feeding speed during the arc and short-circuit stages is adjusted to promote droplet transition.

[0005] However, existing technologies focus more on the positioning and clamping of galvanized sheet workpieces, welding trajectory matching, or welding parameter adjustment based on arc waveform. They lack a coordinated control mechanism to address the stagnant pressure escape phenomenon caused by zinc vapor being affected by lap gaps, weld center offset, clamping force, and local channel contraction during the welding of lap-jointed galvanized sheets. This mechanism is insufficient for understanding the theoretical escape location from the observed vapor plume position, constructing a virtual escape channel, and calculating the degree of vapor stagnant pressure escape mismatch. When the local lap gap narrows or the clamping force causes the vapor discharge channel to close, zinc vapor easily forms instantaneous back pressure near the keyhole, inducing spatter, porosity, keyhole fluctuations, or local weld collapse. If adjustments are still made according to fixed welding parameters or a single waveform state, it is difficult to identify the enhanced vapor disturbance section in time and generate pressure-relief welding coordinated control parameters in advance.

[0006] Therefore, in order to address the above problems, there is an urgent need for a collaborative control system for galvanized sheet welding based on steam disturbance perception. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a collaborative control system for galvanized sheet welding based on vapor disturbance sensing. This system solves the problem that the location of zinc vapor stagnation and escape during the welding process of existing galvanized sheets is difficult to accurately identify, resulting in delayed adjustment of welding parameters and difficulty in controlling the local forming quality of the weld.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a collaborative control system for galvanized sheet welding based on steam disturbance perception, comprising: a steam monitoring and processing module, which collects welding steam monitoring data and performs noise removal, fluctuation smoothing, missing data completion, timestamp alignment, and numerical standardization on the welding steam monitoring data, and outputs preprocessed welding steam monitoring data; an escape channel inversion and identification module, which inversely calculates the theoretical escape position value based on the preprocessed welding steam monitoring data, generates a virtual escape channel record, calculates the steam stagnation pressure escape mismatch value based on the virtual escape channel record, determines the steam disturbance enhancement section, and outputs the steam escape channel inversion result; and a pressure relief curve. The line fusion generation module generates various welding control corrections based on the inversion results of the steam escape channel, and determines the corresponding control start position value according to the welding travel speed value and each response delay time. Based on the corresponding control start position value, segment end position value and each welding control correction, it generates various correction curves, and fuses the overlapping correction curves according to the proportion of steam hysteresis escape mismatch value. After the change rate is limited and the original control parameters are superimposed, the pressure relief welding collaborative control parameters are output. The welding parameter matching execution module reads the pressure relief welding collaborative control parameters, matches the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputs the target control parameters to the corresponding welding execution mechanism.

[0009] Further, the specific steps for collecting welding steam monitoring data and performing noise removal, fluctuation smoothing, missing data completion, timestamp alignment, and numerical standardization on the welding steam monitoring data are as follows: Welding steam monitoring data is collected during the welding process of galvanized sheets. This data includes the galvanized sheet number, weld number, target weld length, lap width, lap gap, weld centerline offset, clamping force, laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, distance from welding torch to sheet surface, weld position, steam plume height, steam plume offset angle, steam plume brightness, number of spatter particles, spatter particle image displacement, keyhole bright area, welding sound pressure level, and data acquisition. The data collection process includes: collecting timestamps; performing instantaneous spike noise removal using median filtering for the collected welding steam monitoring data; smoothing continuous sampling fluctuations using exponential moving average; completing short-term missing records using linear interpolation; generating a standard sampling time series according to a preset unified sampling period; mapping welding steam monitoring data from different sources to the standard sampling time series using a combination of nearest neighbor timestamp matching and linear interpolation compensation; performing multi-source acquisition sequence synchronization processing on the welding steam monitoring data; and performing numerical standardization processing on the welding steam monitoring data using robust standardization based on median and interquartile range, outputting the preprocessed welding steam monitoring data.

[0010] Furthermore, the specific steps for back-deriving the theoretical initiation position value based on the preprocessed welding steam monitoring data are as follows: Read the preprocessed welding steam monitoring data, extract the data records of the same weld according to the weld number, and sort them in ascending order according to the weld position value; take each weld position value as the current observation position, multiply the distance value from the welding torch to the plate surface by the tangent value of the steam plume offset angle value to obtain the steam lateral back-pushing distance value, and then subtract the steam lateral back-pushing distance value from the current observation position to obtain the theoretical initiation position value.

[0011] Further, the specific steps for generating a virtual escape channel record are as follows: at each theoretical escape position value, call the corresponding lap gap value, lap width value, weld centerline offset value, and clamping force value; use the lap gap value as the channel height of the virtual escape channel, use the absolute value of the lap width value minus the weld centerline offset value as the effective channel width of the virtual escape channel, and use the clamping force value as the closing load of the virtual escape channel to generate a virtual escape channel record.

[0012] Further, the specific steps for calculating the steam hysteresis escape mismatch value based on the virtual escape channel record are as follows: Read the virtual escape channel record and the preprocessed welding steam monitoring data at the corresponding location; use a logistic function normalization algorithm to perform dimensionless positive interval mapping processing on the steam plume height value, steam plume brightness value, spatter particle number value, spatter particle image displacement value, welding sound pressure level value, fixture clamping force value, channel height, and effective channel width value, respectively, to obtain the steam plume height mapping value, steam plume brightness mapping value, spatter particle number mapping value, spatter particle image displacement mapping value, welding sound pressure level mapping value, fixture clamping force mapping value, channel height mapping value, and effective channel width mapping value; add one to the steam plume height mapping value to obtain the plume height amplification term; take the square root of the steam plume brightness mapping value and add one to obtain... The following terms are used to calculate the amplification terms: plume brightness amplification term; calculate the tangent of the absolute value of the vapor plume offset angle and add one to obtain the offset angle amplification term; multiply the splash particle number mapping value with the splash particle image displacement mapping value, add one to the product, take the natural logarithm, and add one again to obtain the splash disturbance amplification term; add one to the welding sound pressure level mapping value to obtain the sound pressure disturbance amplification term; add one to the clamping force mapping value to obtain the clamping closure amplification term; multiply the plume height amplification term, plume brightness amplification term, offset angle amplification term, splash disturbance amplification term, sound pressure disturbance amplification term, and clamping closure amplification term to obtain the vapor hysteresis response term; add the channel height mapping value to the minimum constant, and then multiply it by the sum of the effective channel width mapping value and the minimum constant to obtain the escape channel constraint term; divide the vapor hysteresis response term by the escape channel constraint term to obtain the vapor hysteresis escape mismatch value.

[0013] Further, the specific steps for determining the enhanced steam disturbance section and outputting the steam escape channel inversion results are as follows: Read the virtual escape channel records and steam stagnation mismatch values ​​corresponding to each theoretical escape position value, and compare adjacent virtual escape channel records in ascending order of theoretical escape position values; when the effective channel width and channel height of the later virtual escape channel record are both less than those of the earlier virtual escape channel record, mark the theoretical escape position value corresponding to the later virtual escape channel record as a channel contraction node; when the closing load of the later virtual escape channel record is greater than that of the earlier virtual escape channel record, and the channel height of the later virtual escape channel record is less than that of the earlier virtual escape channel record, mark the theoretical escape position value corresponding to the later virtual escape channel record as a channel compression node; from the channel contraction nodes and channel compression nodes, select nodes whose corresponding steam stagnation mismatch values ​​are greater than the mismatch threshold, and use the selected nodes as nodes to be checked; read the continuous values ​​following the welding direction of the nodes to be checked. N weld position values ​​are used as the verification distance interval. Any two adjacent weld position values ​​within the verification distance interval form adjacent sampling point pairs. The increase in steam plume height, the change in keyhole bright area, and the increase in the number of spatter particles are calculated for each adjacent sampling point pair. When there is an adjacent sampling point pair that simultaneously satisfies the following conditions: the increase in steam plume height is greater than the plume height increase threshold, the change in keyhole bright area is greater than the keyhole area change threshold, and the increase in the number of spatter particles is greater than the spatter increase threshold, the node to be verified is determined as a steam stagnation release node. The steam stagnation release nodes are arranged in ascending order of theoretical escape position values. The source position distance between adjacent steam stagnation release nodes is calculated. When the source position distance is less than or equal to the node merging distance threshold, the corresponding adjacent steam stagnation release nodes are assigned to the same steam disturbance enhancement segment. The minimum theoretical escape position value within the same steam disturbance enhancement segment is used as the segment start position value, and the maximum theoretical escape position value is used as the segment end position value. The steam escape channel inversion result is output.

[0014] Furthermore, the specific steps for generating welding control corrections based on the vapor escape channel inversion results and determining the corresponding control start position values ​​according to the welding travel speed value and each response delay time are as follows: Read the vapor escape channel inversion results and preprocessed welding vapor monitoring data; subtract the mismatch threshold from the vapor hysteresis escape mismatch value to obtain the hysteresis exceedance; multiply the hysteresis exceedance by the laser power correction coefficient, welding speed correction coefficient, oscillation amplitude correction coefficient, oscillation frequency correction coefficient, shielding gas flow correction coefficient, and fixture clamping force correction coefficient to obtain the laser output power correction and welding travel speed correction. The correction values ​​are: laser oscillation amplitude correction value, laser oscillation frequency correction value, shielding gas flow rate correction value, and fixture clamping force correction value. The welding travel speed value is multiplied by the pre-calibrated laser power response delay time, laser oscillation response delay time, shielding gas flow rate response delay time, and fixture clamping response delay time to obtain the laser forward distance value, oscillation forward distance value, shielding gas forward distance value, and fixture forward distance value. The section start position value is subtracted from each forward distance value to obtain the laser control start position value, oscillation control start position value, shielding gas control start position value, and fixture control start position value.

[0015] Furthermore, the specific steps for generating each correction curve based on the corresponding control start position value, segment end position value, and each welding control correction amount are as follows: Using the control start position value to the segment end position value as the corresponding correction interval, a cubic Bézier curve algorithm is used to perform position continuity processing on each correction amount; for any correction interval, the correction amount corresponding to the start position value of the correction interval is set to 0, and the correction amount corresponding to the segment end position value is set to the corresponding correction amount; the product of the correction interval length and the first Bézier smoothing coefficient is used as the first position offset; the start position value of the correction interval plus the first position offset is used as the position coordinate of the first intermediate control point, and 0 is set as the correction value of the first intermediate control point. Positive coordinates are used to obtain the first intermediate control point; the product of the corresponding correction amount and the second Bezier smoothing coefficient is used as the second correction amount offset; the segment termination position value minus the first position offset is used as the position coordinate of the second intermediate control point, and the corresponding correction amount minus the second correction amount offset is used as the correction amount coordinate of the second intermediate control point to obtain the second intermediate control point; based on the correction interval start position value, segment termination position value, 0, corresponding correction amount, first intermediate control point and second intermediate control point, the laser output power correction curve, welding travel speed correction curve, laser oscillation amplitude correction curve, laser oscillation frequency correction curve, shielding gas flow rate correction curve and fixture clamping force correction curve are generated.

[0016] Further, the specific steps for merging the overlapping correction curves according to the proportion of vapor stagnation pressure escape mismatch, and outputting the pressure relief welding collaborative control parameters after rate of change limiting and superposition of the original control parameters are as follows: Read the correction curves corresponding to adjacent vapor disturbance enhancement sections. When adjacent correction curves overlap at the weld location, take the correction curve with the larger vapor stagnation pressure escape mismatch value at the overlapping location as the main correction curve, and merge the correction curve with the smaller vapor stagnation pressure escape mismatch value into the main correction curve according to the mismatch value proportion to obtain the fused correction curve; where the mismatch value proportion is the ratio of the smaller vapor stagnation pressure escape mismatch value to the larger vapor stagnation pressure escape mismatch value; calculate the correction curves corresponding to adjacent weld locations. The ratio of the absolute value of the positive difference to the weld position difference yields the unit position change rate. When the unit position change rate exceeds the change rate limit, the correction amount corresponding to the next weld position point is truncated and corrected according to the change rate limit, resulting in a limit correction curve. The limit correction curve is then superimposed with the laser output power value, welding travel speed value, laser oscillation amplitude value, laser oscillation frequency value, shielding gas flow rate value, and fixture clamping force value of the corresponding weld position point to obtain the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target fixture clamping force value, outputting the pressure relief welding collaborative control parameters.

[0017] Further, the specific steps for reading the pressure-relief welding collaborative control parameters, matching the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputting the target control parameters to the corresponding welding actuator are as follows: Read the pressure-relief welding collaborative control parameter table according to the weld number; search for the position interval in the pressure-relief welding collaborative control parameter table according to the current weld position value fed back by the encoder of the welding motion mechanism within the preset control cycle; perform linear interpolation on the pressure-relief welding collaborative control parameters at both ends of the position interval to obtain the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target clamping force value corresponding to the current welding position; output the target laser output power value to the laser welding power source; output the target welding travel speed value to the welding motion mechanism; output the target laser oscillation amplitude value and target laser oscillation frequency value to the laser oscillation mechanism; output the target shielding gas flow rate value to the shielding gas supply mechanism; output the target clamping force value to the clamping mechanism; and repeat the parameter matching and output process when the current weld position value is updated.

[0018] The present invention has the following beneficial effects:

[0019] (1) The galvanized sheet welding collaborative control system based on steam disturbance perception pushes the steam plume deviation state observed at the welding site back to the theoretical starting position value in the overlapping area, so that the steam anomaly no longer stays in the apparent position judgment in the camera field of view, but can be traced back to the actual possible weld source position of steam, thereby improving the accuracy of steam disturbance source positioning and reducing misjudgment caused by plume deviation, welding gun height change or observation angle.

[0020] (2) The galvanized sheet welding collaborative control system based on steam disturbance perception constructs the overlap gap, effective channel width and clamping force of the fixture into a virtual escape channel record, and calculates the steam stagnation escape mismatch value together with the dynamic response of steam plume, spatter and sound pressure. This allows the steam disturbance judgment to consider both "whether the channel is allowed to discharge" and "whether the steam has become stagnant and enhanced", avoiding control based solely on the number of spatter, welding sound or plume brightness, thus improving the reliability of steam stagnation identification.

[0021] (3) The galvanized sheet welding collaborative control system based on steam disturbance perception establishes a positional correlation between local overlap gap narrowing, clamp closing and subsequent plume rise, keyhole fluctuation and spatter increase by combining the channel contraction node, channel compression node and backward steam release response through the combined verification. This makes the determination of the enhanced steam disturbance section not dependent on single-point threshold alarm, but forms a continuous section identification result along the welding direction, which improves the ability to make early judgments on the risk sections of porosity, spatter and molten pool collapse.

[0022] (4) The galvanized sheet welding collaborative control system based on steam disturbance perception converts the steam escape channel inversion results into collaborative correction quantities of laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate and clamping force. The system adopts continuous correction curve, overlapping curve fusion and rate of change limiting processing to enable welding control parameters to be smoothly adjusted around the steam disturbance enhancement section, avoiding heat input fluctuation, shielding gas disturbance and clamping load impact caused by single parameter mutation, thereby improving the pressure relief regulation stability and weld formation consistency in the galvanized sheet welding process. Attached Figure Description

[0023] Figure 1 This is a structural diagram of a collaborative control system for galvanized sheet welding based on steam disturbance sensing.

[0024] Figure 2 This is a sorting diagram of vapor stagnation pressure escape mismatch coupling;

[0025] Figure 3 This is a schematic diagram of the generation of the correction curve for pressure relief welding collaborative control. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see Figures 1-3 This invention provides a technical solution: a collaborative control system for galvanized sheet welding based on steam disturbance perception, comprising: a steam monitoring and processing module, which collects welding steam monitoring data and performs noise removal, fluctuation smoothing, missing data completion, timestamp alignment, and numerical standardization on the welding steam monitoring data, and outputs preprocessed welding steam monitoring data; an escape channel inversion and identification module, which inversely calculates the theoretical escape position value based on the preprocessed welding steam monitoring data, generates a virtual escape channel record, calculates the steam hysteresis escape mismatch value based on the virtual escape channel record, determines the steam disturbance enhancement section, and outputs the steam escape channel inversion result; and a pressure relief curve fusion generation module. The module generates various welding control correction values ​​based on the inversion results of the steam escape channel, and determines the corresponding control start position value according to the welding travel speed value and each response delay time. Based on the corresponding control start position value, the segment end position value and each welding control correction value, it generates various correction curves, and merges the overlapping correction curves according to the proportion of steam hysteresis escape mismatch value. After the change rate is limited and the original control parameters are superimposed, the pressure relief welding collaborative control parameters are output. The welding parameter matching execution module reads the pressure relief welding collaborative control parameters, matches the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputs the target control parameters to the corresponding welding execution mechanism.

[0028] Specifically, the steps for collecting welding steam monitoring data and performing noise removal, fluctuation smoothing, missing data completion, timestamp alignment, and numerical standardization on the data are as follows: Welding steam monitoring data is collected during the welding process of galvanized sheets. The data is attributed to the same galvanized sheet number and the same weld number, with the data collection timestamp as the time index and the weld position value as the position index. The structural assembly status, process execution status, steam plume status, spatter disturbance status, keyhole change status, and acoustic disturbance status during the welding process are recorded synchronously, forming an original monitoring sequence with continuous weld position and sampling time relationships. The welding steam monitoring data includes the galvanized sheet number, weld number, target weld length value, etc. The data includes lap width, lap gap, weld centerline offset, clamping force, laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, distance from welding torch to plate surface, weld position, steam plume height, steam plume offset angle, steam plume brightness, number of spatter particles, spatter particle image displacement, keyhole bright area, welding sound pressure level, and data acquisition timestamp. The galvanized sheet number is generated by combining the production batch code, sheet loading sequence number, and welding task number, and is read by a barcode scanner when the galvanized sheet enters the welding station. The weld number is generated by the preset weld sequence in the welding task file and is written to the current sampling record when the welding path is loaded. The target weld length... The degree value is calculated from the weld start-point and weld end-point coordinates in the welding task file; the lap width value is calculated by the pre-weld visual measurement device after extracting the edge of the lap boundary of the two galvanized plates; the lap gap value is obtained by the laser displacement sensor scanning the height difference between the upper plate edge and the lower plate surface along the weld direction; the weld centerline offset value is calculated by the pre-weld visual measurement device after identifying the lap centerline and the preset welding trajectory centerline; the clamping force value is collected by the pressure sensor built into the clamping head; the laser output power value is read from the power feedback interface of the laser welding power supply; the welding travel speed value is calculated from the unit time displacement fed back by the encoder of the welding motion mechanism; the laser oscillation amplitude value is fed back by the laser oscillation mechanism controller; the laser oscillation... The frequency value is obtained from feedback by the laser oscillation mechanism controller; the shielding gas flow rate value is collected by the mass flow meter on the shielding gas supply pipeline; the distance from the welding torch head to the plate surface is collected by the laser rangefinder sensor placed beside the welding torch head; the weld position value is obtained by the cumulative displacement conversion of the encoder of the welding motion mechanism; the vapor plume height value is obtained by the high-speed vision camera acquiring the image of the welding area and calculating it based on the calibration distance between the upper boundary of the bright area of ​​the molten pool and the highest connected domain pixel of the vapor plume; the vapor plume offset angle value is obtained by the high-speed vision camera identifying the direction of the vapor plume main axis and calculating the angle between the vapor plume main axis and the plate surface normal; the vapor plume brightness value is obtained by the high-speed vision camera statistically calculating the grayscale mean in the vapor plume connected domain and normalizing it to the grayscale range of 0 to 255.The spatter particle count is obtained by identifying the number of bright particle connected regions leaving the molten pool boundary within a single sampling period using a high-speed vision camera; the spatter particle image displacement is calculated from the coordinate difference of the center point of the same spatter particle in adjacent image frames; the keyhole bright area is calculated by thresholding the bright area at the center of the weld molten pool using a high-speed vision camera; the welding sound pressure level is calculated by collecting sound pressure signals from an acoustic sensor near the welding torch; the data acquisition timestamp is written by a unified clock on the industrial control computer, with a time resolution of 1ms to 10ms; among these, the target welding length value is used to limit the effective range of weld position values ​​under the same weld number, and the overlap width value, overlap gap value, and weld centerline offset value are used to characterize the geometry of the vapor escape space within the overlap area. The clamping force value is used to characterize the degree of pressure closure in the lap area; the laser output power value, welding travel speed value, laser oscillation amplitude value, laser oscillation frequency value, and shielding gas flow rate value are used to characterize the current welding heat input and welding execution status; the distance from the welding torch to the plate surface, the vapor plume height value, and the vapor plume offset angle value are used to subsequently deduce the theoretical escape position value; the vapor plume brightness value, spatter particle number value, spatter particle image displacement value, keyhole bright area value, and welding sound pressure level value are used to characterize the visual disturbance, spatter disturbance, keyhole disturbance, and acoustic disturbance generated when zinc vapor is released after being subjected to hysteresis; for the collected welding vapor monitoring data, the sampling record order is first established according to the galvanized plate number, weld number, and data acquisition timestamp, and then the same weld... The weld position values ​​under the weld number are continuously checked, and sampling records with values ​​less than 0 or greater than the target weld length are removed, while valid sampling records corresponding to the weld position values ​​are retained. A median filtering algorithm is used to perform instantaneous spike noise removal. Specifically, a sliding window is set according to the data acquisition timestamp under the same weld number. The sliding window length is 5 to 9 sampling points, corresponding to a time range of 20ms to 90ms. The sliding window length is determined by the welding travel speed and the visual acquisition frame rate. The median value within the window is calculated for the steam plume height, steam plume brightness, number of spatter particles, image displacement of spatter particles, bright area of ​​the keyhole, and welding sound pressure level. Any values ​​deviating from the median value beyond the spike detection threshold are removed. The sampled values ​​are replaced with the median of the window; the peak determination threshold is 3 to 5 times the absolute median difference within the window, and the specific value is obtained by statistical analysis of the no-load data collected by the welding equipment and the historical data collected by qualified welds; for the spatter particle count, if the increase of a single sampling point relative to the previous sampling point exceeds 50 and the subsequent sampling point falls back to within 120% of the previous sampling point, the current sampling point is identified as an instantaneous peak; for the welding sound pressure level, if the deviation of a single sampling point relative to the median of the window exceeds 12dB to 20dB and the duration is less than 30ms, the current sampling point is identified as an instantaneous peak; through this processing, the instantaneous image flickering, spatter obstruction, and sound pressure jumps during the welding process are not directly included in the subsequent calculation of the vapor stagnation escape mismatch value;An exponential moving average algorithm was used to perform continuous sampling fluctuation smoothing on the welding steam monitoring data. Specifically, the current sampled value was weighted with the smoothed value of the previous time step according to the progressive order of data acquisition timestamps, generating smoothed steam plume height, smoothed steam plume brightness, smoothed spatter particle image displacement, smoothed keyhole bright area, and smoothed welding sound pressure level. The smoothing weight coefficient was set between 0.20 and 0.40, with the specific value determined by the natural fluctuation range of the steam plume height and keyhole bright area values ​​in the historical data of qualified welds. When the natural fluctuation range was greater than 15% of the corresponding mean, the coefficient was set to 0.20 to 0.30; when the natural fluctuation range was not greater than 15% of the corresponding mean, the coefficient was set to 0. The value is adjusted from 0.30 to 0.40 to reduce high-frequency jitter in continuous sampling while preserving the trend characteristics of continuously rising steam plumes, increasing spatter, and continuously changing keyhole bright areas. A linear interpolation algorithm is used to perform short-term missing record completion processing on the welding steam monitoring data. Specifically, under the same weld number, short-term missing intervals are identified where the data acquisition timestamp interval is greater than the standard sampling interval but less than the missing record completion duration threshold. The standard sampling interval is set to 5ms to 20ms, determined by the frame rate of the high-speed vision camera, the sampling frequency of the acoustic sensor, and the maximum time interval in the welding execution record refresh cycle. The missing record completion duration threshold is set to 50ms to 120ms, determined by the welding travel speed value and the length of the minimum identifiable steam disturbance segment. The missing interval is then read. The system collects adjacent valid sampling records before and after the missing interval, and calculates the lap gap value, clamping force value, distance from welding torch to plate surface value, steam plume height value, steam plume offset angle value, steam plume brightness value, spatter particle number value, spatter particle image displacement value, keyhole bright area value, and welding sound pressure level value at the missing location according to the distance ratio of the data acquisition timestamp. When the duration of the missing interval is greater than the missing completion time threshold, the sampling record corresponding to the missing interval is marked as an invalid record and is not included in the subsequent theoretical initiation position value back calculation, so that there are no short-term breakpoints when calculating the theoretical initiation position value in ascending order of weld position value. The welding steam monitoring data is processed by multi-source acquisition sequence synchronization using a combination of nearest neighbor timestamp matching and linear interpolation compensation. Specifically, a standard sampling time series is generated using a preset unified sampling period. Each standard sampling time point in the standard sampling time series is used as a synchronization benchmark. The steam plume height, steam plume offset angle, steam plume brightness, spatter particle number, spatter particle image displacement, and keyhole bright area value obtained from visual acquisition are uniformly mapped to the same standard sampling time point along with the laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, fixture clamping force, and welding sound pressure level value obtained from acoustic acquisition in the welding execution record. When there is a deviation between the data acquisition timestamps from different acquisition sources and the standard sampling time point, the sampling record with the smallest absolute time difference is used as the candidate alignment record.The allowable deviation threshold for timestamps is set between 5ms and 15ms, determined by the synchronization error of the industrial control computer's unified clock and the exposure cycle of the high-speed vision camera. When the absolute value of the time difference between the candidate alignment record and the standard sampling time point is not greater than the allowable deviation threshold, the candidate alignment record is used as the alignment record for the corresponding standard sampling time point. When the absolute value of the time difference between the candidate alignment record and the standard sampling time point is greater than the allowable deviation threshold, the valid sampling records adjacent to the standard sampling time point are read, and linear interpolation compensation is performed according to the time distance ratio between the standard sampling time point and the adjacent valid sampling records to obtain the compensation data at the corresponding standard sampling time point. This compensation is then applied based on the welding progress. The speed value is converted into weld position offset by the sampling time difference. The allowable threshold for weld position offset is 0.05mm to 0.30mm, determined by the minimum control resolution of the weld and the encoder resolution of the welding motion mechanism. When the weld position offset is not greater than the allowable threshold, the corresponding sampling record is retained as a valid synchronous record at the same weld position value, ensuring consistency of process parameters, assembly parameters, and steam disturbance response at the same weld position value. A robust normalization method based on median and interquartile range is used to perform numerical normalization on the welding steam monitoring data. Specifically, the median and interquartile range of each numerical data point under the same weld number are calculated separately, and each sampling... The value is subtracted from the corresponding median and then divided by the corresponding interquartile range to obtain the standardized result against abnormal fluctuations. When the interquartile range is less than the minimum scale threshold, the minimum scale threshold is used as the denominator in the calculation. The minimum scale threshold is taken as 0.5% to 2.0% of the corresponding data engineering range and is determined by the resolution and equipment repeatability test data in the sensor calibration certificate. For values ​​such as lap width, lap gap, weld centerline offset, clamping force, laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, distance from welding torch to plate surface, weld position, vapor plume height, vapor plume offset angle, vapor plume brightness, and flight path, the standardization result against abnormal fluctuations is obtained. The number of spatter particles, the displacement of spatter particle images, the area of ​​the bright area of ​​the keyhole, and the welding sound pressure level are each formed into a standardized numerical sequence. Among them, the engineering quantity values ​​after noise removal, fluctuation smoothing, missing data completion, and synchronous processing of multi-source acquisition sequences are used for back-calculation of theoretical escape position values, generation of virtual escape channel records, calculation of control start position values, and output of target control parameters. The standardized numerical sequence is used for dimensionless mapping and multi-index coupling evaluation in the calculation of steam stagnation escape mismatch value, so that the geometric position calculation, channel size calculation, and control parameter output maintain the consistency of engineering dimensions, and at the same time, data of different dimensions and magnitudes can participate in the calculation of steam stagnation escape mismatch value.Finally, the welding steam monitoring data, after noise removal, fluctuation smoothing, missing data completion, multi-source acquisition sequence synchronization processing, and numerical standardization, are reordered according to galvanized plate number, weld number, data acquisition timestamp, and weld location value, and the preprocessed welding steam monitoring data is output.

[0029] In this implementation plan, by uniformly defining the sources, acquisition methods, sampling scales, window ranges, threshold values, and anomaly handling rules for welding steam monitoring data, a traceable correspondence is established between galvanized sheet numbers, weld numbers, weld location values, data acquisition timestamps, and various structural assembly states, process execution states, and steam disturbance responses. This avoids location mismatches, timing mismatches, and scale deviations in subsequent calculations caused by data from different acquisition sources and with different dimensions. Consequently, stable, continuous, and comparable basic data are provided for back-calculating theoretical escape location values, generating virtual escape channel records, and calculating steam stagnation escape mismatch values, thereby improving the accuracy of steam disturbance source location and the generation of subsequent pressure relief welding collaborative control parameters.

[0030] Specifically, the steps for inferring the theoretical origin position value based on preprocessed welding steam monitoring data are as follows: Read the preprocessed welding steam monitoring data, extract data records for the same weld seam according to weld seam number, and sort them in ascending order of weld seam position value; while extracting data records for the same weld seam, simultaneously read the galvanized plate number, target welding length value, distance from welding torch to plate surface, steam plume offset angle value, data acquisition timestamp, and weld seam position value, and perform position continuity checks on the data records under the same weld seam number to ensure that the weld seam position value is within the range of 0 to the target welding length value, and that the distance between adjacent weld seam position values ​​is no greater than the preset position sampling interval. The position sampling interval is set from 0.05mm to 0.30mm and is determined by the encoder resolution of the welding motion mechanism, the welding travel speed, and the data acquisition timestamp interval. When the interval between adjacent weld position values ​​is greater than the preset position sampling interval, position completion is performed based on the data acquisition timestamps of adjacent valid data records and the weld position value, so that the subsequent theoretical escape position value is calculated according to the continuous weld trajectory. Each weld position value is used as the current observation position, and the distance from the welding torch to the plate surface is multiplied by the tangent of the steam plume offset angle to obtain the steam lateral pushback distance. The distance from the welding torch to the plate surface is processed by noise removal, fluctuation smoothing, and missing data elimination. The engineering quantity values ​​after completion and multi-source acquisition sequence synchronization processing are used. The steam plume offset angle value is the engineering angle value after noise removal, fluctuation smoothing, missing data completion, and multi-source acquisition sequence synchronization processing. The steam plume offset angle value does not use the numerically standardized result in the tangent function calculation. The steam plume offset angle value is the angle between the steam plume centerline and the plate surface normal in the vertical plane where the weld travel direction is located. The steam plume centerline is obtained from the steam plume image acquired by the high-speed vision camera through brightness threshold segmentation, plume connectivity region extraction, region skeleton line fitting, and centerline direction calculation. Before calculating the steam plume offset angle value, the camera intrinsic parameter calibration results are used. The camera extrinsic calibration results, plate reference calibration points, and weld centerline calibration points are used to transform the steam plume centerline in the image coordinate system to a welding space coordinate system composed of the plate surface normal, weld travel direction, and plate surface transverse direction. The projection direction of the steam plume centerline in the plane formed by the weld travel direction and the plate surface normal is taken as the angle calculation direction, so that the steam plume offset angle value has a unified reference system. The steam plume offset angle value is uniformly converted into radians before participating in the tangent function calculation, and an angle sign rule is established according to the welding travel direction. The steam plume offset in the welding travel direction is recorded as a positive angle, and the steam plume offset in the opposite direction of the welding travel direction is recorded as a negative angle.The calculation principle of the lateral back-pull distance of steam is that after the steam plume escapes upward from its actual starting point in the overlap area, it will be tilted due to the influence of welding heat flux, shielding gas flow, plate gap, and welding movement direction. The distance from the welding torch tip to the plate surface represents the vertical projection distance of the steam plume from its starting point to the observation height. The tangent of the steam plume offset angle represents the proportion of the steam plume's horizontal offset relative to the plate surface normal in the weld travel direction. Multiplying the two can convert the plume tilt observed in the image into a back-pull distance along the weld direction. Since the stagnant escape position of zinc vapor is usually located near the position where the overlap gap is limited, the effective channel width is narrowed, and the clamping force of the fixture creates a closing effect, the steam plume height and steam plume brightness... The values ​​for steam emission and spatter particle quantity are apparent responses formed after steam escapes and reaches the observation area. The current observation position corresponding to the apparent response has a spatial offset along the weld direction relative to the actual restricted emission position. Therefore, by back-calculating the theoretical emission position value using the steam lateral back-distance value, the observed steam disturbance response can be correlated back to the channel-restricted source position in the overlapping area. Compared to using only the current observation position as the subsequent adjustment position, the theoretical emission position value can locate the source position of the restricted release of zinc steam in advance, so that the subsequent virtual emission channel record, steam stagnation emission mismatch value, and control start position value are all calculated around the source position. This allows the pressure relief welding collaborative control parameters to be pre-adjusted before the steam disturbance enhancement section reaches the actual welding execution position. When the absolute value of the steam plume offset angle is less than the lower limit of angle stability, the lateral back-pushing distance of the steam is recorded as 0. The lower limit of angle stability is set to 1° to 3° and is determined by the angle recognition error of the high-speed vision camera and the natural oscillation amplitude of the plume in the historical data of qualified welds. When the absolute value of the steam plume offset angle is greater than the upper limit of angle confidence, the current data record is marked as an offset angle anomalous record. The upper limit of angle confidence is set to 55° to 65° and is determined by the effective observation range of the high-speed vision camera's field of view, the distance from the welding torch to the plate surface, and the steam plume height. For offset angle anomalous records, the steam plume offset angle values ​​of adjacent valid data records are replaced according to the distance ratio of the data acquisition timestamp to prevent the tangent function from being in a large angle state. The image recognition error is magnified; then, the theoretical starting position value is obtained by subtracting the steam lateral back-pushing distance value from the current observation position; when the theoretical starting position value is less than 0, it is corrected to 0; when the theoretical starting position value is greater than the target welding length value, it is corrected to the target welding length value; finally, the theoretical starting position value is bound and stored with the galvanized plate number, weld number, current observation position, distance from welding torch to plate surface, steam plume offset angle, steam lateral back-pushing distance, and data acquisition timestamp, forming a theoretical starting position sequence arranged in ascending order of weld position value. This provides a position index for subsequently calling the lap gap value, lap width value, weld centerline offset value, and clamping force value at the theoretical starting position value.

[0031] In this implementation scheme, by incorporating the weld position value, data acquisition timestamp, distance from the welding torch to the plate surface, and steam plume offset angle value into the same position backtracking chain, and by defining the binding relationship between the angle symbol, the angle confidence range, the position boundary, and the theoretical escape position value, the tilt offset exhibited by the steam plume at the observation position can be converted into source position information within the overlap area. This avoids source position offset caused by judging the source of steam disturbance solely based on the current observation position, thereby improving the matching accuracy between the theoretical escape position value and subsequent overlap gap value, overlap width value, weld centerline offset value, and clamping force value, providing a stable position basis for the generation of virtual escape channel records.

[0032] Specifically, the steps for generating virtual escape channel records are as follows: At each theoretical escape position value, the corresponding lap gap value, lap width value, weld centerline offset value, and clamping force value are retrieved. During retrieval, the galvanized sheet number, weld number, and theoretical escape position value are used as indexes to search for the data record with the smallest position distance in the pre-processed welding steam monitoring data. When the theoretical escape position value does not completely coincide with the existing weld position value, the weld position values ​​adjacent to the theoretical escape position value are read. Based on the positional ratio of the theoretical escape position value to the adjacent weld position values, linear interpolation is performed on the lap gap value, lap width value, weld centerline offset value, and clamping force value to obtain the corresponding lap gap value, lap width value, weld centerline offset value, and clamping force value at the theoretical escape position value. When the distance between the theoretical escape position value and the nearest weld position value is greater than the allowable deviation for position matching, the corresponding data record is marked as a position matching anomaly record. The allowable deviation for position matching is between 0.05mm and 0.30m. The value of m is determined by the encoder resolution of the welding motion mechanism, the minimum control resolution of the weld, and the sampling interval of the pre-processed weld position value. The lap gap value is used as the channel height of the virtual escape channel, the absolute value of the lap width value minus the weld centerline offset value is used as the effective channel width of the virtual escape channel, and the clamping force value of the fixture is used as the closing load of the virtual escape channel to generate a virtual escape channel record. Among them, the channel height is used to characterize the vertical gap between the upper and lower galvanized plates that allows zinc vapor to escape along the lap gap, the effective channel width is used to characterize the lateral channel range that can still participate in vapor discharge after the weld centerline offset, and the closing load is used to characterize the strength of the compressive closing effect of the clamping force value on the lap gap value. When the absolute value of the weld centerline offset value is greater than the lap width value, the effective channel width is corrected to 0 so that no effective lateral escape space is formed when the weld centerline deviates significantly from the lap area. When the lap gap value is less than the effective lower limit of the gap, the channel height is marked as the compression height, and the effective lower limit of the gap is taken as 0.02mm to 0.0.05mm, and determined by the resolution of the laser displacement sensor, the surface roughness of the galvanized sheet, and the minimum stable gap in the historical data of qualified welds; when the clamping force of the clamp is greater than the upper limit of the clamping load, the closed load is marked as a strong clamping load. The upper limit of the clamping load is taken as 70% to 90% of the rated clamping force of the clamp, and determined by the calibration curve of the clamping mechanism and the lap gap compression test results corresponding to the thickness of the galvanized sheet; when generating the virtual escape channel record, the theoretical escape position value, the current observation position, the channel height, the effective channel width, the closed load, the lap gap value, the lap width value, the weld centerline offset value, the clamping force value, the galvanized sheet number, the weld number, and the data acquisition timestamp are included. Binding storage ensures that each virtual escape channel record corresponds to a steam plume observation record and the geometric stress state of the overlap area. The technical principle behind this process is that zinc vapor, during the welding of lap-jointed galvanized sheets, primarily relies on the overlap gap to form a pressure relief path. The overlap width and weld centerline offset jointly determine the effective lateral range that vapor can enter along the overlap direction. The clamping force of the fixture determines whether this range will further close under welding heat input. Therefore, by constructing virtual escape channels through channel height, effective channel width, and closing load, the directly observable vapor escape space can be transformed into a calculable virtual escape channel record, providing geometric and stress constraints for subsequent calculations of vapor stagnation and escape mismatch values.

[0033] In this implementation scheme, a positional correspondence is established between the theoretical escape position value and the lap gap value, lap width value, weld centerline offset value, and clamping force value. This correspondence is then converted into channel height, effective channel width, and closing load. This allows the zinc vapor escape space, which cannot be directly observed, to be quantified as a virtual escape channel record. This avoids the loss of channel constraints caused by judging the vapor discharge conditions solely based on the apparent state of the vapor plume. Consequently, the ability of subsequent vapor stagnation and escape mismatch values ​​to characterize the geometrically restricted and closed states of the lap area is improved.

[0034] Specifically, the steps for calculating the vapor hysteresis escape mismatch value based on the virtual escape channel record are as follows: Read the virtual escape channel record and the preprocessed welding vapor monitoring data at the corresponding location. During reading, use the galvanized plate number, weld number, theoretical escape position value, and data acquisition timestamp as a joint index. Bind the channel height, effective channel width, and closure load in the virtual escape channel record to the vapor plume height, vapor plume brightness, vapor plume offset angle, spatter particle number, spatter particle image displacement, and welding sound pressure level values ​​in the preprocessed welding vapor monitoring data. This ensures that the vapor disturbance response and escape channel constraint corresponding to the same theoretical escape position value are included in the same calculation. The logistic function normalization algorithm is used to perform dimensionless positive interval mapping processing on the steam plume height, steam plume brightness, spatter particle number, spatter particle image displacement, welding sound pressure level, fixture clamping force, channel height, and effective channel width values, respectively, to obtain the mapped values ​​for steam plume height, steam plume brightness, spatter particle number, spatter particle image displacement, welding sound pressure level, fixture clamping force, channel height, and effective channel width. The logistic function normalization algorithm uses a negative exponent form of 1 divided by 1 plus the natural constant e. The negative exponent input is the result of subtracting the corresponding data baseline value from the current data value and then dividing by the corresponding data scale value. In this study, the steam plume height, steam plume brightness, number of spatter particles, image displacement of spatter particles, welding sound pressure level, and clamping force are positively mapped; the larger the value, the closer the mapped value is to one. Channel height and effective channel width are used to characterize the openness of the escape channel, also mapped to the positive range of zero to one; the larger the value, the closer the mapped value is to one. The corresponding data baseline value is the median of historical data collected from qualified welds under the same weld number, and the corresponding data scale value is the interquartile range of historical data collected from qualified welds under the same weld number. When the interquartile range is less than one percent of the corresponding data engineering range, one percent of the corresponding data engineering range is used as the corresponding data scale value to prevent scale inaccuracies. Too small a value will cause abrupt changes in the mapping result; the data engineering range for the steam plume height value is 0.5 mm to 30 mm, the data engineering range for the steam plume brightness value is 0 to 255 grayscale values, the data engineering range for the number of spatter particles is 0 to 200, the data engineering range for the image displacement value of spatter particles is 0.1 mm to 20 mm, the data engineering range for the welding sound pressure level value is 60 dB to 140 dB, the data engineering range for the clamping force value is 100 N to 3000 N, the data engineering range for the channel height is 0.02 mm to 0.50 mm, and the data engineering range for the effective channel width value is 0 mm to 30 mm; add one to the steam plume height mapping value to obtain the plume height amplification term;The plume height amplification term characterizes the degree of vertical emission enhancement of zinc vapor above the plate surface. A larger plume height mapping value indicates more intense vapor release and a greater contribution to the vapor hysteresis response term. The plume brightness amplification term is obtained by taking the square root of the plume brightness mapping value and adding one. Taking the square root of the plume brightness mapping value is used to compress excessive amplification caused by high-brightness image areas, ensuring that the plume brightness value reflects the increasing trend of zinc vapor concentration, while reducing the influence of instantaneous strong light from the welding heat source on the calculation results. The tangent value is calculated by taking the absolute value of the plume offset angle value and adding one to obtain the offset angle amplification term. The plume offset angle value uses radians in the tangent function calculation, and the absolute value is used to eliminate the plume's tendency to move towards the welding surface. The sign difference between the offset in the direction of travel and the offset in the opposite direction of welding means that the offset angle amplification term only represents the degree to which the steam plume deviates from the normal to the plate surface. The tangent function is used to convert the angle offset into a horizontal offset ratio. When the absolute value of the steam plume offset angle is less than 1 to 3 degrees, the offset angle amplification term is recorded as one. When the absolute value of the steam plume offset angle is greater than 55 to 65 degrees, the steam plume offset angle value of the adjacent valid data record is used for replacement to prevent the tangent function from amplifying visual recognition errors under large angle conditions. The spatter particle number mapping value is multiplied by the spatter particle image displacement mapping value, the product is incremented by one, the natural logarithm is taken, and then incremented by one again to obtain the spatter disturbance amplification term. Among them, the spatter particle number mapping value is used for... The frequency of spatter occurrence is characterized by the image displacement mapping value of the spatter particles, which is used to characterize the motion intensity of the spatter particles after they leave the molten pool. Multiplying the two values ​​ensures that a significant spatter disturbance contribution is only achieved when both high-frequency and high-displacement spatter are enhanced. The natural logarithm is used to suppress the sudden increase in the vapor stagnation response term due to extreme spatter records. Adding one to the welding sound pressure level mapping value yields the sound pressure disturbance amplification term. This term characterizes the acoustic abrupt change in intensity caused by the release of zinc vapor stagnation, ensuring that weld locations with insignificant visual disturbances but enhanced acoustic disturbances are still included in the vapor stagnation response term. Adding one to the clamping force mapping value yields the clamping closure amplification term. This term characterizes the effect of the clamping force on the vapor stagnation response term when it exerts a closing effect on the lap gap. The greater the pressure amplification factor, the higher the risk of channel closure. Multiplying the plume height amplification term, plume brightness amplification term, offset angle amplification term, splash disturbance amplification term, sound pressure disturbance amplification term, and pressure closure amplification term yields the vapor stagnation response term. The reason for using product coupling is that zinc vapor stagnation release typically manifests simultaneously as plume rise, increased brightness, plume offset, intensified splashing, increased sound pressure, and enhanced clamp closure. Product coupling amplifies the vapor stagnation response term when multiple responses are enhanced together, and reduces the probability of misjudgment when a single response is abnormal. Adding the channel height mapping value to the minimum constant, and then multiplying it by the sum of the effective channel width mapping value and the minimum constant, yields the escape channel constraint term; the minimum constant is taken as... to The value is determined by the computational accuracy and minimum resolution of the data mapping to prevent division by zero errors when the channel height mapping value is zero or the effective channel width mapping value is zero. The escape channel constraint term characterizes the ability of zinc vapor to be released through the overlap gap and the transverse channel. The smaller the channel height mapping value and the smaller the effective channel width mapping value, the more obvious the channel height and transverse range restrictions are, and the smaller the escape channel constraint term is. Dividing the vapor hysteresis response term by the escape channel constraint term yields the vapor hysteresis escape mismatch value. The larger the vapor hysteresis escape mismatch value, the more likely there is a mismatch at the same theoretical escape position. A strong vapor disturbance response, coupled with insufficient channel height and effective channel width to provide a stable pressure relief path, indicates that this location is more likely to form a zinc vapor stagnation and escape mismatch state. Finally, the vapor stagnation and escape mismatch value is bound and stored with the galvanized plate number, weld number, theoretical escape location value, current observation location, vapor stagnation response term, escape channel constraint term, channel height, effective channel width, closing load, and data acquisition timestamp to form a vapor stagnation and escape mismatch value sequence arranged in ascending order of theoretical escape location value, providing a quantitative basis for subsequent determination of the vapor disturbance enhancement section.

[0035] The specific formula for calculating the vapor hysteresis escape mismatch value is as follows:

[0036] ;

[0037] In the formula, This indicates the vapor stagnation pressure escape mismatch value. This represents the vapor plume height mapping value. This represents the brightness mapping value of the vapor plume. This indicates the angle of deviation of the steam plume. This represents the mapping value for the number of splashed particles. This represents the displacement mapping value of the splash particle image. This represents the welding sound pressure level mapping value. This indicates the clamping force mapping value of the fixture. Indicates the channel height mapping value. Indicates the effective channel width mapping value. This represents a very small constant.

[0038] In this implementation scheme, the channel height, effective channel width, closure load, and steam plume height, steam plume brightness, steam plume offset angle, number of spatter particles, image displacement of spatter particles, and welding sound pressure level values ​​recorded in the virtual escape channel are coupled and evaluated under the same theoretical escape position value. This allows the steam stagnation escape mismatch value to simultaneously reflect the zinc vapor release intensity and the degree of restriction of the escape channel. It avoids the loss of channel conditions caused by judging the stagnation state solely based on the steam disturbance response, and also avoids the loss of disturbance response caused by judging the pressure relief capacity solely based on the overlap gap value. This improves the completeness and stability of the steam stagnation escape mismatch state identification and provides a more reliable quantitative basis for the determination of the enhanced steam disturbance section.

[0039] In Example 1, Table 1 is a data table of vapor stagnation pressure escape mismatch values, listing exemplary values ​​used to illustrate the calculation process of vapor stagnation pressure escape mismatch values ​​for five nodes to be checked. Specifically: in node 1 to be checked, the vapor plume height mapping value is 0.62, the vapor plume brightness mapping value is 0.58, the vapor plume offset angle value is 8, the number of spatter particles mapping value is 0.42, the spatter particle image displacement mapping value is 0.35, the welding sound pressure level mapping value is 0.50, the fixture clamping force mapping value is 0.64, the channel height mapping value is 0.46, the effective channel width mapping value is 0.52, and the vapor stagnation pressure escape mismatch value is 37.9089; in node 2 to be checked, the vapor plume height mapping value is 0.75, and the vapor plume brightness mapping value is 0.7. 0, steam plume offset angle value is 12, spatter particle number mapping value is 0.60, spatter particle image displacement mapping value is 0.48, welding sound pressure level mapping value is 0.66, fixture clamping force mapping value is 0.72, channel height mapping value is 0.33, effective channel width mapping value is 0.41, steam stagnation escape mismatch value is 102.4986; in node 3 to be checked, steam plume height mapping value is 0.44, steam plume brightness mapping value is 0.39, steam plume offset angle value is 5, spatter particle number mapping value is 0.28, spatter particle image displacement mapping value is... The mapping values ​​are as follows: 0.22 for welding sound pressure level, 0.36 for welding sound pressure level, 0.48 for clamping force, 0.62 for channel height, 0.67 for effective channel width, and 13.0229 for vapor stagnation escape mismatch. In node 4 to be checked, the mapping values ​​are as follows: 0.86 for vapor plume height, 0.82 for vapor plume brightness, 16 for vapor plume offset angle, 0.75 for spatter particle number, 0.62 for spatter particle image displacement, 0.78 for welding sound pressure level, 0.83 for clamping force, and 0.48 for channel height. The value is 0.24, the effective channel width mapping value is 0.31, and the vapor stagnation escape mismatch value is 273.8921; in node 5 to be checked, the vapor plume height mapping value is 0.69, the vapor plume brightness mapping value is 0.64, the vapor plume offset angle value is 10, the number of spatter particles mapping value is 0.51, the spatter particle image displacement mapping value is 0.40, the welding sound pressure level mapping value is 0.57, the fixture clamping force mapping value is 0.68, the channel height mapping value is 0.39, the effective channel width mapping value is 0.45, and the vapor stagnation escape mismatch value is 63.4599.

[0040] Table 1. Vapor hysteresis escape mismatch data.

[0041]

[0042] like Figure 2As shown, the coupling and sorting relationship of vapor hysteresis escape mismatch values ​​is illustrated. The upper part is the mismatch intensity sorting area for the nodes to be checked, arranged in descending order of vapor hysteresis escape mismatch value as node 4, node 2, node 5, node 1, and node 3, with the mismatch degree changing from strong to weak indicated by a red to green directional marker. The lower part is the coupling calculation logic area for vapor hysteresis escape mismatch values. The left side shows the vapor hysteresis response factors, including vapor plume height mapping value, vapor plume brightness mapping value, plume offset angle value, splash particle number mapping value, and splash particle image displacement mapping value. The values ​​of the welding sound pressure level and the clamping force are used to characterize the combined effects of zinc vapor stagnation, spray deviation, spatter disturbance, sound pressure disturbance, and clamp closure on the vapor release state during the welding process of galvanized sheets. The right side is the escape channel constraint factor, including the channel height mapping value and the effective channel width mapping value, which are used to characterize the degree of restriction of the virtual escape channel. The middle part is the vapor stagnation escape mismatch value. By coupling the vapor stagnation response factor with the escape channel constraint factor, the vapor stagnation escape mismatch value is formed. The larger the vapor stagnation escape mismatch value, the stronger the vapor stagnation response and the more restricted the escape channel is at the corresponding node to be checked.

[0043] Specifically, the steps for determining the enhanced steam disturbance section and outputting the steam escape channel inversion results are as follows: Read the virtual escape channel records and steam stagnation mismatch values ​​corresponding to each theoretical escape position value, and compare adjacent virtual escape channel records in ascending order of theoretical escape position values. During reading, use the galvanized sheet number, weld number, theoretical escape position value, and data acquisition timestamp as association fields to bind the channel height, effective channel width, and closing load in each virtual escape channel record to the steam stagnation mismatch value under the same theoretical escape position value. Invalid records with duplicate theoretical escape position values, missing data acquisition timestamps, empty channel heights, empty effective channel widths, or empty closing loads are removed. When the same theoretical escape position value... When multiple virtual escape channel records exist, the virtual escape channel record with the largest vapor hysteresis escape mismatch value is selected as the representative record of the theoretical escape position value, so that subsequent segment identification prioritizes retaining the most obvious positional feature of vapor hysteresis. When comparing adjacent virtual escape channel records in ascending order of theoretical escape position value, the effective channel width difference, channel height difference, and closure load difference of the subsequent virtual escape channel record relative to the previous virtual escape channel record are calculated respectively. Among them, the effective channel width difference is the subsequent effective channel width minus the previous effective channel width, the channel height difference is the subsequent channel height minus the previous channel height, and the closure load difference is the subsequent closure load minus the previous closure load. When the effective channel width and... When the channel height is less than the previous virtual escape channel record, the theoretical escape position value corresponding to the subsequent virtual escape channel record is marked as a channel contraction node. During the channel contraction node marking process, it is further required that the effective channel width difference is less than the negative width change threshold, and the channel height difference is less than the negative height change threshold. The negative width change threshold is taken as 1% to 5% of the effective channel width data engineering range, and the negative height change threshold is taken as 1% to 5% of the lap gap value data engineering range. The specific values ​​are determined by the natural fluctuation amplitude between adjacent theoretical escape position values ​​in the historical data of qualified welds, ensuring that minor changes caused by measurement jitter are not mistakenly marked as channel contraction nodes. When the closing load of the subsequent virtual escape channel record... When the load is greater than the closing load recorded in the previous virtual escape channel, and the channel height recorded in the subsequent virtual escape channel is less than the channel height recorded in the previous virtual escape channel, the theoretical escape position value corresponding to the subsequent virtual escape channel record is marked as the channel compression node. In the process of marking the channel compression node, it is further required that the difference in closing load is greater than the load increase threshold, and the difference in channel height is less than the negative height change threshold. The load increase threshold is taken as two to eight percent of the engineering range of the clamping force value data. The specific value is determined by the calibration curve of the clamping mechanism, the lap gap compression test results corresponding to the thickness of the galvanized sheet, and the historical data collected from qualified welds. The position where the clamping force value increases and causes the channel height to decrease is identified as the compression risk position.From the channel contraction nodes and channel compression nodes, nodes with corresponding steam stagnation escape mismatch values ​​greater than the mismatch threshold are selected, and these selected nodes are used as nodes to be checked. The N consecutive weld position values ​​behind the node to be checked along the welding direction are read as the check distance interval. Any two adjacent weld position values ​​within the check distance interval are used to form adjacent sampling point pairs. The steam plume height increase, keyhole bright area change, and spatter particle number increase are calculated for each adjacent sampling point pair. The mismatch threshold is taken as 1.5 to 3 times the median of the steam stagnation escape mismatch value corresponding to the historical data of qualified welds under the same weld number. The specific value is determined by the historical data of qualified welds and the historical data of welds with spatter defects. Historical data of welds with porosity defects were statistically analyzed to determine N; N ranged from 3 to 10, with the specific value determined by the welding travel speed, the sampling interval of weld position values, and the delay distance required for zinc vapor to be released from the stagnant pressure position to the observed enhanced response; the verification distance interval was used to verify whether enhanced vapor release occurred behind the channel contraction node and the channel compression node. The technical principle is that after zinc vapor is restricted in the overlapping area, it usually does not immediately manifest as the maximum visual disturbance at the geometrically restricted position, but will form a plume rise, keyhole bright area fluctuation, and increased spatter over a short distance behind with the welding travel direction; for each adjacent sampling point pair, the vapor plume height increase is the vapor plume height value corresponding to the subsequent weld position value minus the previous one. The weld position value corresponds to the steam plume height value. The change in the bright area of ​​the keyhole is the absolute value of the difference between the bright area of ​​the keyhole corresponding to the subsequent weld position value and the bright area of ​​the keyhole corresponding to the previous weld position value. The increase in the number of spatter particles is the difference between the number of spatter particles corresponding to the subsequent weld position value and the number of spatter particles corresponding to the previous weld position value. When there is an adjacent sampling point pair that simultaneously satisfies the following conditions: the increase in steam plume height is greater than the plume height increase threshold, the change in the bright area of ​​the keyhole is greater than the keyhole area change threshold, and the increase in the number of spatter particles is greater than the spatter number increase threshold, the node to be checked is determined as the steam stagnation release node. Among them, the plume height increase threshold is taken from 0.5 mm to 3 mm, and the keyhole area change threshold is taken as 0. The threshold for the increase in the number of spatter particles is set between 5 and 30, with the specific values ​​determined by the natural incremental distribution of the steam plume height, keyhole bright area, and spatter particle number values ​​in the historical data of qualified welds, as well as the abnormal incremental distribution in the historical data of defective welds. The judgment method that requires simultaneous satisfaction of the three types of responses is adopted because a single increase in steam plume height may come from shielding gas disturbance, a single change in keyhole bright area may come from laser output power fluctuation, and a single increase in the number of spatter particles may come from surface contamination. Therefore, limiting the three types of responses to synchronous enhancement on the same adjacent sampling point pair can make the steam stagnation release node closer to the comprehensive disturbance caused by the restricted release of zinc vapor.The vapor stagnation release nodes are arranged in ascending order of their theoretical initiation positions. The source-position distance between adjacent vapor stagnation release nodes is calculated. When the source-position distance is less than or equal to the node merging distance threshold, the corresponding adjacent vapor stagnation release nodes are assigned to the same vapor disturbance enhancement segment. The minimum theoretical initiation position value within the same vapor disturbance enhancement segment is used as the segment's starting position value, and the maximum theoretical initiation position value is used as the segment's ending position value. The vapor escape channel inversion result is output. The source-position distance is the distance obtained by subtracting the theoretical initiation position value of the preceding vapor stagnation release node from the theoretical initiation position value of the subsequent vapor stagnation release node. The node merging distance threshold is set between 0.5 mm and 5 mm, and the specific value is determined by the weld position value sampling interval, weld pool length, zinc vapor diffusion length along the lap gap, and the qualified weld... The interval between adjacent disturbances in the historical data of the weld seam is determined; when the source distance between adjacent steam stagnation release nodes is greater than the node merging distance threshold, the next steam stagnation release node is taken as the starting point of the new steam disturbance enhancement section; when outputting the steam escape channel inversion results, the galvanized plate number, weld seam number, section start position value, section end position value, number of steam stagnation release nodes in the section, maximum steam stagnation escape mismatch value in the section, average steam stagnation escape mismatch value in the section, minimum channel height in the section, minimum effective channel width in the section, maximum closing load in the section, and data acquisition timestamp are bound and stored, so that the steam escape channel inversion results can simultaneously express the source range, channel restriction degree, and steam release intensity, providing the section boundary and mismatch degree basis for the subsequent generation of various welding control corrections;

[0044] In this implementation scheme, by establishing a continuous verification relationship between the channel contraction node, the channel compression node, the vapor stagnation and escape mismatch value and the increase in vapor plume height, keyhole bright area area change, and spatter particle number behind the welding direction, the enhanced vapor disturbance section no longer relies on single-point anomaly judgment, but can simultaneously reflect the correlation between source channel restriction, stagnation intensity exceeding limit, and enhanced release response. This improves the reliability of the statistical results of the section start position value, section end position value, and vapor stagnation and escape mismatch value within the section, providing a more accurate spatial boundary and strength basis for the subsequent generation of welding control correction values.

[0045] Specifically, the steps for generating welding control corrections based on the vapor escape channel inversion results and determining the corresponding control start position values ​​based on the welding travel speed and response delay times are as follows: Read the vapor escape channel inversion results and preprocessed welding vapor monitoring data. During reading, use the galvanized sheet number and weld number as task indices, and the segment start position value and segment end position value as segment boundary indices. Extract the maximum vapor stagnation pressure escape mismatch value, the average vapor stagnation pressure escape mismatch value, the minimum channel height, the minimum effective channel width, and the maximum closing load within the segment from the vapor escape channel inversion results. Also, read the actual executed or feedback excitation values ​​within the corresponding segment from the preprocessed welding vapor monitoring data. The optical output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, and fixture clamping force are read and used as the original control parameters, serving as the basis for generating welding control corrections. These original control parameters characterize the welding process execution state before correction within the enhanced steam disturbance zone and serve as the benchmark for superimposing subsequent target control parameters with each correction. The steam stagnation pressure escape mismatch value is subtracted from the mismatch threshold to obtain the stagnation pressure exceedance. The maximum steam stagnation pressure escape mismatch value within the zone is used to characterize the strongest stagnation pressure release risk in the current enhanced steam disturbance zone. The mismatch threshold is the same threshold used previously when determining the steam stagnation pressure release node. The determination of the enhanced steam disturbance zone and the generation of welding control corrections maintain the same judgment benchmark. When the steam stagnation pressure escape mismatch value is not greater than the mismatch threshold, the stagnation pressure excess is set to zero. When the steam stagnation pressure escape mismatch value is greater than the mismatch threshold, the difference between the two is retained as the stagnation pressure excess, so that the welding control correction is generated only for steam stagnation states that exceed the normal pressure relief capacity. The stagnation pressure excess is multiplied by the laser power correction coefficient, welding speed correction coefficient, oscillation amplitude correction coefficient, oscillation frequency correction coefficient, shielding gas flow correction coefficient, and fixture clamping force correction coefficient to obtain the laser output power correction, welding travel speed correction, laser oscillation amplitude correction, laser oscillation frequency correction, shielding gas flow correction, and fixture clamping force correction. Among them, the laser power correction coefficient is negative to reduce the laser output power value when the hysteresis exceeds the limit, thereby reducing the instantaneous vaporization intensity of the zinc layer; the welding speed correction coefficient is positive to increase the welding travel speed value when the hysteresis exceeds the limit, thereby shortening the heat treatment time per unit length; the oscillation amplitude correction coefficient is positive to expand the laser oscillation amplitude value, thereby dispersing the heat input laterally in the overlapping area; the oscillation frequency correction coefficient is positive to increase the laser oscillation frequency value, thereby making the molten pool disturbance more continuous and promoting the discharge of zinc vapor; the shielding gas flow rate correction coefficient is positive to increase the shielding gas flow rate value, thereby enhancing the vapor plume discharge capability; and the clamping force correction coefficient is negative to reduce the clamping force value, thereby allowing the channel height to recover.The laser power correction factor is taken as -0.01 to -0.20 kW per unit of hysteresis exceeding the limit; the welding speed correction factor is taken as 0.5 to 10 mm / s per unit of hysteresis exceeding the limit; the oscillation amplitude correction factor is taken as 0.01 to 0.20 mm per unit of hysteresis exceeding the limit; the oscillation frequency correction factor is taken as 1 to 20 Hz per unit of hysteresis exceeding the limit; the shielding gas flow rate correction factor is taken as 0.1 to 2.0 L / min per unit of hysteresis exceeding the limit; and the fixture clamping force correction factor is taken as -5 to -100 N per unit of hysteresis exceeding the limit. Specific values ​​are derived from historical data collected on qualified welds, welds with spatter defects, welds with porosity defects, and welding process experiments. The calibration results were obtained; after obtaining the various welding control corrections, the laser output power correction, welding travel speed correction, laser oscillation amplitude correction, laser oscillation frequency correction, shielding gas flow rate correction, and fixture clamping force correction were respectively subjected to amplitude limiting processing to ensure that the corrected laser output power value was between 500 watts and 6000 watts, the corrected welding travel speed value was between 10 mm / s and 200 mm / s, the corrected laser oscillation amplitude value was between 0.10 mm and 3.00 mm, the corrected laser oscillation frequency value was between 20 Hz and 500 Hz, and the corrected shielding gas flow rate value was between 5 liters / min and 30 liters / min. The clamping force of the corrected fixture is between 100 N and 3000 N to prevent the correction from exceeding the range of the welding equipment. The welding travel speed is multiplied by the pre-calibrated laser power response delay time, laser oscillation response delay time, shielding gas flow response delay time, and fixture clamping response delay time to obtain the laser forward distance, oscillation forward distance, shielding gas forward distance, and fixture forward distance. Each response delay time refers to the time required for different welding actuators to reach the target change state from receiving the control command, including the laser power response delay time, laser oscillation response delay time, shielding gas flow response delay time, and fixture clamping response delay time. Clamping response delay time; laser power response delay time is used to characterize the time required for the laser welding power source to reach the target change state after receiving the power correction command; laser oscillation response delay time is used to characterize the time required for the laser oscillation mechanism to reach the target change state after receiving the oscillation correction command; shielding gas flow response delay time is used to characterize the time required for the shielding gas supply mechanism to reach the target change state after receiving the flow correction command; clamping response delay time is used to characterize the time required for the clamping mechanism to reach the target change state after receiving the clamping force correction command.Each response delay time was obtained through equipment step response calibration. Specifically, step control commands were sent to the laser welding power source, laser oscillation mechanism, shielding gas supply mechanism, and clamping mechanism, respectively. The time intervals from the issuance of the control command to when the laser output power value, laser oscillation amplitude value, laser oscillation frequency value, shielding gas flow rate value, and clamping force value reached 90% of the target change were recorded as the corresponding response delay times. The laser power response delay time was set to 5 milliseconds to 50 milliseconds, the laser oscillation response delay time to 10 milliseconds to 80 milliseconds, the shielding gas flow rate response delay time to 50 milliseconds to 300 milliseconds, and the clamping response delay time to 1 millisecond. The timeframe ranges from 0.00 milliseconds to 800 milliseconds, with the specific value determined by equipment calibration records and on-site no-load response test data. The technical principle behind using the forward distance value is that there is a response lag between the welding actuator receiving the control command and the actual change in parameters. If the control command is issued only at the starting position of the section, the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target clamping force will only take effect after the welding position has passed the starting position of the section. Therefore, the response delay time needs to be converted into a forward distance value along the weld direction based on the welding travel speed value. The starting position value of the section is then subtracted from... By removing each forward distance value, we obtain the laser control starting position value, the oscillation control starting position value, the shielding gas control starting position value, and the fixture control starting position value. Among these, the laser control starting position value corresponds to the early activation position of the laser output power correction and welding travel speed correction; the oscillation control starting position value corresponds to the early activation position of the laser oscillation amplitude correction and laser oscillation frequency correction; the shielding gas control starting position value corresponds to the early activation position of the shielding gas flow correction; and the fixture control starting position value corresponds to the early activation position of the fixture clamping force correction. When the laser control starting position value, the oscillation control starting position value, the shielding gas control starting position value, and the fixture control starting position value are all adjusted, the starting position is determined. When the control start position value is less than zero, the corresponding control start position value is corrected to zero. When the corresponding control start position value is greater than the target welding length value, the corresponding control start position value is corrected to the target welding length value. Finally, the galvanized sheet number, weld number, section start position value, section end position value, hysteresis limit, laser output power correction, welding travel speed correction, laser oscillation amplitude correction, laser oscillation frequency correction, shielding gas flow correction, fixture clamping force correction, laser control start position value, oscillation control start position value, shielding gas control start position value, and fixture control start position value are bound and stored to form a welding control correction record.

[0046] In this implementation scheme, the segment start position value, segment end position value, and steam stagnation escape mismatch value in the inversion results of the steam escape channel are converted into laser output power correction, welding travel speed correction, laser oscillation amplitude correction, laser oscillation frequency correction, shielding gas flow correction, and fixture clamping force correction. These are then combined with the laser power response delay time, laser oscillation response delay time, shielding gas flow response delay time, and fixture clamping response delay time to determine the corresponding control start position value. This allows the welding control correction record to simultaneously reflect the correspondence between steam stagnation intensity, weld spatial position, and actuator response lag, avoiding control commands taking effect only after the steam disturbance enhancement segment arrives, thus preventing adjustment lag. This improves the ability of the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow value, and target fixture clamping force value to proactively intervene in the risk of zinc steam stagnation release.

[0047] Specifically, the steps for generating each correction curve based on the corresponding control start position value, section end position value, and each welding control correction amount are as follows: Read the galvanized sheet number, weld number, section start position value, section end position value, laser control start position value, oscillation control start position value, shielding gas control start position value, fixture control start position value, laser output power correction amount, welding travel speed correction amount, laser oscillation amplitude correction amount, laser oscillation frequency correction amount, shielding gas flow rate correction amount, and fixture clamping force correction amount from the welding control correction record. Establish the correspondence between the correction amount and the control start position value according to different execution objects. Among these, the laser output power correction amount and the welding... The travel speed correction corresponds to the laser control starting position value, the laser oscillation amplitude correction and laser oscillation frequency correction correspond to the oscillation control starting position value, the shielding gas flow correction corresponds to the shielding gas control starting position value, and the fixture clamping force correction corresponds to the fixture control starting position value. The correction interval is defined as the distance from each control starting position value to the end position value of the section. A cubic Bézier curve algorithm is used to perform positional continuity processing on each correction. Specifically, the correction intervals for the laser output power correction curve and the welding travel speed correction curve are from the laser control starting position value to the end position value of the section, while the correction intervals for the laser oscillation amplitude correction curve and the laser oscillation frequency correction curve are from the oscillation control starting position value to the end position value of the section. The correction interval for the protective gas flow correction curve is from the protective gas control start position value to the section end position value, and the correction interval for the clamping force correction curve is from the clamp control start position value to the section end position value. When any control start position value is less than zero, the corresponding control start position value is corrected to zero. When any control start position value is greater than the section end position value, the corresponding correction interval is marked as an invalid correction interval, and the full interval correction amount of the corresponding correction curve is set to 0. For any correction interval, the correction amount corresponding to the correction interval start position value is set to 0, and the correction amount corresponding to the section end position value is set to the corresponding correction amount. The correction interval start position value is used to indicate the start of the control command. The intervention position, with the segment termination position value indicating the location where the correction amount reaches the target amplitude, and the curve correction amount corresponding to any weld position value within the correction interval is obtained by interpolation of a cubic Bézier curve, so that the welding control parameters smoothly transition from the uncorrected state to the target corrected state along the weld direction; the product of the correction interval length and the first Bézier smoothing coefficient is used as the first position offset, and the first position offset is added to the correction interval start position value to obtain the position coordinates of the first intermediate control point, with 0 as the correction amount coordinates of the first intermediate control point, thus obtaining the first intermediate control point; wherein, the correction interval length is the segment termination position value minus the correction interval start position value, and the first Bézier smoothing coefficient is taken as 0.20 to 0.40, the specific value is obtained by jointly calibrating the speed response stability of the welding motion mechanism, the power adjustment slope of the laser welding power source, the oscillation adjustment slope of the laser oscillation mechanism, the flow adjustment slope of the shielding gas supply mechanism, and the load adjustment slope of the clamping mechanism; the correction coordinate of the first intermediate control point is set to 0, so that the correction curve maintains a small rate of change near the starting position value of the correction interval, avoiding abrupt changes in the laser output power value, welding travel speed value, laser oscillation amplitude value, laser oscillation frequency value, shielding gas flow rate value, and clamping force value at the control start position; the product of the corresponding correction amount and the second Bessel smoothing coefficient is used as the second correction amount offset; the second Bessel smoothing coefficient is taken as 0.20 to 0.50, and the specific value is determined by the influence of parameter changes on the stability of the molten pool, the stability of the steam plume, and the spatter variation amplitude in the welding process test. A larger value is taken when the welding process is sensitive to parameter abrupt changes, and a smaller value is taken when the welding process has a high tolerance for parameter changes; the position coordinate of the second intermediate control point is obtained by subtracting the first position offset from the segment termination position value, and the corresponding correction amount is used as the product of the first position offset and the first position offset. The second intermediate control point is obtained by subtracting the offset of the second correction from the positive value. This second intermediate control point controls the convergence slope of the correction curve as it approaches the end position value of the segment, ensuring the correction value gradually approaches the corresponding correction value and avoiding abrupt changes in the correction value before and after the end position value of the segment. Based on the starting position value of the correction interval, the end position value of the segment, 0, the corresponding correction value, the first intermediate control point, and the second intermediate control point, the following correction curves are generated: laser output power correction curve, welding travel speed correction curve, laser oscillation amplitude correction curve, laser oscillation frequency correction curve, shielding gas flow rate correction curve, and fixture clamping force correction curve. Specifically, the starting position value of the correction interval and 0 are used as the starting control point, the end position value of the segment and the corresponding correction value are used as the ending control point, and the first and second intermediate control points are used as curve shape control points. The correction value corresponding to each weld position value on the curve is calculated according to the cubic Bézier curve parameters increasing from 0 to 1, and curve sampling points are output according to a preset position resolution, ranging from 0.05 mm to 0.30 mm, and determined by the encoder resolution of the welding motion mechanism and the minimum control resolution of the weld; when the corresponding correction amount is negative, the curve correction amount changes smoothly in the negative direction, and when the corresponding correction amount is positive, the curve correction amount changes smoothly in the positive direction, so that the laser output power correction curve, welding travel speed correction curve, laser oscillation amplitude correction curve, laser oscillation frequency correction curve, shielding gas flow rate correction curve, and fixture clamping force correction curve all maintain positional continuity; after generation, each correction curve is bound and stored with the galvanized plate number, weld number, segment start position value, segment end position value, corresponding control start position value, corresponding correction amount, first Bessel smoothing coefficient, second Bessel smoothing coefficient, and data acquisition timestamp, forming a correction curve record for subsequent curve fusion and target parameter generation; among which, overlap A correction curve refers to a segment of the correction curve located within a shared position range when adjacent steam disturbance enhancement sections share this range along the weld seam direction. The shared position range is determined by the segment termination position value of the correction curve corresponding to the preceding section and the control start position value of the correction curve corresponding to the following section. When the control start position value of the correction curve corresponding to the following section is less than the segment termination position value of the correction curve corresponding to the preceding section, the curve segments within this position range are defined as overlapping correction curves. Overlapping correction curves characterize the situation where adjacent steam disturbance enhancement sections simultaneously generate correction effects at the same weld seam location due to response delay shift, smooth transition of the Bézier curve, and close section distance. They also serve as input for subsequent curve fusion based on the proportion of steam hysteresis escape mismatch.

[0048] In this implementation scheme, by establishing a curve mapping relationship between the laser control starting position value, the oscillation control starting position value, the shielding gas control starting position value, and the fixture control starting position value and the corresponding welding control correction value, the laser output power correction curve, welding travel speed correction curve, laser oscillation amplitude correction curve, laser oscillation frequency correction curve, shielding gas flow rate correction curve, and fixture clamping force correction curve can continuously change along the weld position value. This avoids abrupt changes in various welding execution parameters before and after the vapor disturbance enhancement section, thereby improving the stability and executability of the pressure relief welding collaborative control parameters in the actual execution process and reducing the secondary disturbances caused by parameter mutations to the molten pool stability, vapor plume state, and spatter particle quantity.

[0049] In Example 2, a steam disturbance enhancement section corresponding to the weld number is used as an example for illustration. The starting position of this steam disturbance enhancement section is 150.0 mm, and the ending position is 190.0 mm; the welding travel speed is 20.0 mm / s; the laser response delay time is 0.15 s, corresponding to a laser forward movement distance of 3.0 mm, and the laser control starting position is 147.0 mm; the oscillation response delay time is 0.25 s, corresponding to an oscillation forward movement distance of 5.0 mm, and the oscillation control starting position is 145.0 mm; the shielding gas response delay time is 0.35 s, corresponding to a shielding gas forward movement distance of 7.0 mm, and the shielding gas control starting position is 143.0 mm; the fixture response delay time is 0.45 s, corresponding to a fixture forward movement distance of 9.0 mm, and the fixture control starting position is 141.0 mm. The welding control corrections calculated based on the vapor stagnation pressure escape mismatch value and various correction coefficients are as follows: laser output power correction is -180.0W, welding travel speed correction is -2.5mm / s, laser oscillation amplitude correction is 0.18mm, laser oscillation frequency correction is 12.0Hz, shielding gas flow rate correction is 3.5L / min, and fixture clamping force correction is -45.0N.

[0050] like Figure 3As shown, the horizontal axis represents the weld position value, and the yellow dashed box represents the steam disturbance enhancement section determined by the inversion result of the steam escape channel. The starting position of the section is 150.0 mm, and the ending position is 190.0 mm. Because the laser welding power source, laser oscillation mechanism, shielding gas supply mechanism, and fixture clamping mechanism have different response delay times, the control parameters are not corrected simultaneously from the starting position value of the section. Instead, the forward movement distance is calculated based on the welding travel speed value and the corresponding response delay time, and then the correction process begins from the starting position values ​​of the laser control, oscillation control, shielding gas control, and fixture control, respectively. In the figure, the laser output power correction curve and the welding travel speed correction curve both start at 147.0 mm, the laser oscillation amplitude correction curve and the laser oscillation frequency correction curve both start at 145.0 mm, the shielding gas flow rate correction curve starts at 143.0 mm, and the fixture clamping force correction curve starts at 141.0 mm, demonstrating the coordinated control relationship of different actuators moving forward and misaligning according to their response delays. The six correction curves in the figure correspond to six types of control variables in the generation process of pressure-relief welding collaborative control parameters. The laser output power correction curve and the welding travel speed correction curve show negative corrections, used to reduce the instantaneous heat input and welding advancement intensity in the steam stagnation section; the laser oscillation amplitude correction curve, the laser oscillation frequency correction curve, and the shielding gas flow rate correction curve show positive corrections, used to expand the molten pool disturbance adjustment range and enhance steam exhaust guidance; the clamping force correction curve shows a negative correction, used to reduce the degree of lap gap compression. Each correction curve changes continuously between the corresponding control start position value and the section end position value, and reaches the corresponding correction amount at the section end position value, thereby converting the steam disturbance enhancement section into the target control parameter for the synchronous action of multiple actuators. Figure 3 This invention demonstrates that it does not rely solely on adjusting fixed welding parameters, but rather generates a pressure-relief welding collaborative control curve distributed along the weld position based on the inversion results of the steam escape channel and the response delay of the actuator.

[0051] Specifically, the steps for fusing the overlapping correction curves according to the proportion of vapor hysteresis escape mismatch value, and outputting the pressure relief welding collaborative control parameters after rate of change limiting and superposition of the original control parameters are as follows: Read the correction curves corresponding to adjacent vapor disturbance enhancement sections. During reading, use the galvanized sheet number, weld number, section start position value, section end position value, and corresponding control start position value as associated fields. Read the laser output power correction curve, welding travel speed correction curve, laser oscillation amplitude correction curve, laser oscillation frequency correction curve, shielding gas flow rate correction curve, and fixture clamping force correction curve respectively. Arrange each correction curve into a sequence of curve sampling points in ascending order according to the weld position value. When adjacent correction curves overlap at the weld position, the overlapping position... The correction curve with the larger steam stagnation pressure escape mismatch value is used as the main correction curve. Correction curves with smaller steam stagnation pressure escape mismatch values ​​are merged into the main correction curve according to their mismatch values ​​to obtain a fusion correction curve. The overlapping position is the shared weld position range formed when the termination weld position value of the previous correction curve is not less than the starting weld position value of the subsequent correction curve. Within the overlapping position, the curve correction amounts corresponding to the two correction curves are read according to the same weld position value. The mismatch value ratio is the ratio of the smaller steam stagnation pressure escape mismatch value to the larger steam stagnation pressure escape mismatch value. Specifically, during fusion, at each overlapping position, the curve correction amount of the main correction curve is used as the main correction amount, and the curve correction amount of the correction curve with the smaller steam stagnation pressure escape mismatch value is multiplied by the mismatch value ratio. As the incorporated correction amount, the main correction amount is added to the incorporated correction amount to obtain the fusion correction amount at the overlapping position. When the correction amounts corresponding to the two correction curves are in opposite directions, the incorporated correction amount is first calculated according to the mismatch value ratio, and then algebraically superimposed with the main correction amount to ensure that the steam disturbance enhancement section with a large steam stagnation escape mismatch value maintains the dominant regulation role at the overlapping position. When the correction curves corresponding to multiple steam disturbance enhancement sections overlap at the same weld position, they are fused sequentially according to the steam stagnation escape mismatch value in descending order. First, the correction curve with the largest steam stagnation escape mismatch value is selected as the main correction curve. Then, the remaining correction curves are incorporated into the main correction curve according to the ratio of their respective steam stagnation escape mismatch value to the largest steam stagnation escape mismatch value to obtain the fusion correction amount for that weld. The fusion correction amount at the location; the technical principle of this fusion process is that adjacent steam disturbance enhancement sections may overlap spatially due to response delay compensation and curve smoothing. If a single correction curve is directly selected, the hysteresis effect of adjacent sections will be lost. If all correction curves are directly added, it is easy to cause excessive amplification of the control quantity. Therefore, fusion according to the proportion of steam hysteresis escape mismatch value can enable the correction curve to reflect the contribution intensity of the dominant hysteresis section and the adjacent hysteresis section at the same time. Calculate the ratio of the absolute value of the correction amount difference corresponding to the adjacent weld position points to the weld position difference to obtain the unit position change rate. When the unit position change rate is greater than the change rate limit value, the correction amount corresponding to the next weld position point is truncated and corrected according to the change rate limit value to obtain the limit correction curve.In this process, adjacent weld positions are selected sequentially from the fusion correction curves in ascending order of weld position values. The correction difference is the fusion correction amount for the subsequent weld position minus the fusion correction amount for the preceding weld position. The weld position difference is the weld position value for the subsequent weld position minus the weld position value for the preceding weld position. The rate of change limits are set separately for different controlled objects: the rate of change limit for the laser output power correction curve is from 5 W / mm to 80 W / mm; the rate of change limit for the welding travel speed correction curve is from 0.2 mm / s / mm to 5 mm / s / mm; and the rate of change limit for the laser oscillation amplitude correction curve is from 0.005 mm / s / mm to 0.0 mm / s / mm. The rate of change limit for the laser oscillation frequency correction curve is set at 5 mm / mm, ranging from 0.5 Hz / mm to 10 Hz / mm; the rate of change limit for the shielding gas flow correction curve is set at 0.05 L / min / mm to 0.5 L / min / mm; and the rate of change limit for the fixture clamping force correction curve is set at 2 N / mm to 30 N / mm. The specific values ​​are determined by the laser welding power supply speed adjustment slope, the welding motion mechanism speed adjustment capability, the laser oscillation mechanism response slope, the shielding gas supply flow adjustment slope, the fixture clamping mechanism load adjustment slope, and the welding process test results. When the rate of change per unit position exceeds the rate of change limit, if the fusion correction amount corresponding to the subsequent weld position is greater than that of the previous weld... The amplitude correction amount corresponding to the position point is set as follows: the amplitude correction amount corresponding to the subsequent weld position point is set as the amplitude correction amount corresponding to the previous weld position point plus the product of the rate of change amplitude limit and the weld position difference. If the fusion correction amount corresponding to the subsequent weld position point is less than the amplitude correction amount corresponding to the previous weld position point, the amplitude correction amount corresponding to the subsequent weld position point is set as the amplitude correction amount corresponding to the previous weld position point minus the product of the rate of change amplitude limit and the weld position difference. When the unit position change rate is not greater than the rate of change amplitude limit, the fusion correction amount corresponding to the subsequent weld position point is retained as the amplitude correction amount. The technical principle of this amplitude limiting process is that the fusion correction curve may form a curve at a local weld position after overlapping. A steep slope, when directly executed, will cause rapid changes in the laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, and fixture clamping force. The rate of change limiting can constrain the changes in welding control parameters within the range that the actuator can follow. By superimposing the limiting correction curve with the laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, and fixture clamping force at the corresponding weld position, the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target fixture clamping force are obtained, and the pressure relief welding collaborative control parameters are output.During the specific overlay process, the laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, and clamping force values ​​corresponding to the pre-processed welding steam monitoring data are read at each weld location. The corresponding amplitude limiting correction values ​​at the same weld location are then overlaid to generate the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target clamping force values. After generating the target values, equipment range constraints are applied, limiting the target laser output power to 500 watts to 6000 watts, the target welding travel speed to 10 mm / s to 200 mm / s, and the target laser oscillation amplitude to... The parameters are limited to 0.10 mm to 3.00 mm, the target laser oscillation frequency is limited to 20 Hz to 500 Hz, the target shielding gas flow rate is limited to 5 L / min to 30 L / min, and the target clamping force is limited to 100 N to 3000 N. Finally, the galvanized sheet number, weld number, weld position, target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, target clamping force, corresponding vapor stagnation escape mismatch, corresponding section start position, corresponding section end position, and data acquisition timestamp are bound and stored to form a pressure-relief welding collaborative control parameter arranged in ascending order of weld position value.

[0052] In this implementation scheme, by proportionally fusing the steam stagnation and escape mismatch values ​​of the correction curves corresponding to adjacent steam disturbance enhancement sections, and by applying unit position change rate constraints to the fused correction curves, the pressure relief welding collaborative control parameters can take into account the influence intensity of multiple steam disturbance enhancement sections at the same weld position value. At the same time, it avoids excessive superposition and local abrupt changes in laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, and fixture clamping force, thereby improving the coordination, stability, and equipment executability of the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target fixture clamping force during continuous welding.

[0053] Specifically, the steps for reading the pressure-relief welding collaborative control parameters, matching the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputting the target control parameters to the corresponding welding actuator are as follows: Read the pressure-relief welding collaborative control parameters output after the aforementioned overlapping correction curve is fused according to the proportion of vapor hysteresis escape mismatch value, limited by the rate of change, and superimposed with the original control parameters. Establish a pressure-relief welding collaborative control parameter table according to the galvanized sheet number, weld number, and weld position value. The pressure-relief welding collaborative control parameter table includes the galvanized sheet number, weld number, weld position value, target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, and target protection... The data includes gas flow rate, target clamping force, corresponding steam stagnation pressure escape mismatch, corresponding section start position, corresponding section end position, and data acquisition timestamp. The pressure relief welding collaborative control parameter table is read according to the weld number. During reading, the completeness of the galvanized sheet number, weld number, weld position value, target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, target clamping force value, and data acquisition timestamp is verified. A control parameter index is established in ascending order of weld position value, ensuring the current welding position can be found within the pressure relief welding collaborative control parameter table corresponding to the same weld number. The welding motion mechanism is then used to... The encoder feeds back the current weld position value within a preset control cycle. The position range is then searched in the pressure-relief welding collaborative control parameter table. Linear interpolation is performed on the pressure-relief welding collaborative control parameters at both ends of the position range to obtain the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target clamping force corresponding to the current welding position. The preset control cycle ranges from 1 millisecond to 10 milliseconds and is jointly determined by the encoder feedback frequency of the welding motion mechanism, the laser welding power supply communication cycle, the laser oscillation mechanism response cycle, the shielding gas supply mechanism flow rate adjustment cycle, and the clamping mechanism load adjustment cycle. The current weld position value is determined by the welding motion mechanism... The encoder pulse count is converted by multiplying the encoder pulse increment by the single pulse displacement equivalent to obtain the current position increment. Then, the current position increment is added to the weld start position value to obtain the current weld position value. When searching for a position interval, the weld position value that is less than and closest to the current weld position value is taken as the front position value of the interval, and the weld position value that is greater than and closest to the current weld position value is taken as the back position value of the interval. When the current weld position value is consistent with a weld position value in the pressure relief welding collaborative control parameter table, the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target clamping force value corresponding to the weld position value are directly read.When the current weld position value is between the front and rear positions of the interval, the ratio of the difference between the current weld position value and the front position value to the difference between the rear position value and the front position value is calculated to obtain the position interpolation ratio. Then, the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target clamping force value corresponding to the front position value are used as the front parameters. The target laser output power value, target welding travel speed value, and target laser oscillation amplitude value corresponding to the rear position value are used as the front parameters. The target laser oscillation frequency, target shielding gas flow rate, and target clamping force are used as back-end parameters. These parameters are obtained by multiplying the front-end parameters by the position interpolation ratio and the difference between the back-end and front-end parameters, respectively, to determine the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target clamping force corresponding to the current welding position. When the current weld position value is less than the minimum weld position value in the pressure-relief welding collaborative control parameter table, the pressure-relief welding collaborative control parameter corresponding to the minimum weld position value is read as the current welding position value. Set the corresponding target control parameters; when the current weld position value is greater than the maximum weld position value in the pressure relief welding collaborative control parameter table, read the pressure relief welding collaborative control parameter corresponding to the maximum weld position value as the target control parameter corresponding to the current welding position; the technical principle of this linear interpolation process is that the pressure relief welding collaborative control parameters are generated according to discrete weld position values, while the current weld position value fed back by the encoder of the welding motion mechanism may fall between adjacent discrete positions during continuous movement. Through position interval search and linear interpolation, the target control parameters can be made to change continuously with the weld position value, avoiding target laser... The output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target fixture clamping force value change between adjacent sampling positions; the target laser output power value is output to the laser welding power source, the target welding travel speed value is output to the welding motion mechanism, the target laser oscillation amplitude value and the target laser oscillation frequency value are output to the laser oscillation mechanism, the target shielding gas flow rate value is output to the shielding gas supply mechanism, and the target fixture clamping force value is output to the fixture clamping mechanism, and the parameter matching and output process is repeated when the current weld position value is updated;Before output, the target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, and target clamping force are calibrated to ensure the target laser output power is between 500 watts and 6000 watts, the target welding travel speed is between 10 mm / s and 200 mm / s, the target laser oscillation amplitude is between 0.10 mm and 3.00 mm, the target laser oscillation frequency is between 20 Hz and 500 Hz, the target shielding gas flow rate is between 5 L / min and 30 L / min, and the target clamping force is between 100 N and 3000 N. If any target control parameter exceeds the corresponding equipment range, the target control parameter is corrected to the corresponding equipment range boundary value before output. During output, the target laser output power value is converted... The system converts the target welding travel speed value into a speed setting command recognizable by the welding motion mechanism, the target laser oscillation amplitude and frequency values ​​into oscillation setting commands recognizable by the laser oscillation mechanism, the target shielding gas flow rate into a flow rate setting command recognizable by the shielding gas supply mechanism, and the target clamping force into a load setting command recognizable by the clamping mechanism. Within each preset control cycle, it records the current weld position, target laser output power, target welding travel speed, target laser oscillation amplitude, target laser oscillation frequency, target shielding gas flow rate, target clamping force, and command output timestamp, forming a welding collaborative control execution record. This record is used to trace the execution status of the pressure-relief welding collaborative control parameters during the actual welding process.

[0054] In this implementation plan, by clarifying the source, field content, and interpolation calling method of the pressure-relief welding collaborative control parameter table, the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target clamping force value corresponding to the current welding position can be traceably associated with the galvanized plate number, weld number, weld position value, and corresponding steam stagnation pressure escape mismatch value. This avoids parameter breakpoints and position mismatches caused by direct calling of discrete control parameters during continuous welding, thereby improving the continuity, accuracy, and closed-loop traceability capability of the pressure-relief welding collaborative control parameters in actual welding execution.

[0055] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0056] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A collaborative control system for galvanized sheet welding based on steam disturbance sensing, characterized in that, Includes the following modules: The steam monitoring acquisition and processing module collects welding steam monitoring data and performs noise removal, fluctuation smoothing, missing data completion, timestamp alignment and numerical standardization on the welding steam monitoring data, and outputs the pre-processed welding steam monitoring data. The steam escape channel inversion identification module infers the theoretical steam escape position value based on the preprocessed welding steam monitoring data, generates a virtual escape channel record, calculates the steam stagnation pressure escape mismatch value based on the virtual escape channel record, determines the steam disturbance enhancement section, and outputs the steam escape channel inversion result. The specific steps for back-deriving the theoretical initiation position value based on the preprocessed welding steam monitoring data are as follows: Read the preprocessed welding steam monitoring data, extract the data records of the same weld by weld number, and sort them in ascending order by weld position value; take each weld position value as the current observation position, multiply the distance value from the welding torch to the plate surface by the tangent value of the steam plume offset angle value to obtain the steam lateral back-pushing distance value, and then subtract the steam lateral back-pushing distance value from the current observation position to obtain the theoretical initiation position value; The pressure relief curve fusion generation module generates various welding control corrections based on the inversion results of the steam escape channel, and determines the corresponding control start position value according to the welding travel speed value and each response delay time; it generates various correction curves based on the corresponding control start position value, the segment end position value and each welding control correction, and fuses the overlapping correction curves according to the proportion of steam stagnation pressure escape mismatch value, and outputs the pressure relief welding collaborative control parameters after the change rate is limited and the original control parameters are superimposed; The welding parameter matching and execution module reads the pressure relief welding collaborative control parameters, matches the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputs the target control parameters to the corresponding welding execution mechanism.

2. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 1, characterized in that: The specific steps for collecting welding steam monitoring data and performing noise removal, fluctuation smoothing, missing data completion, timestamp alignment, and numerical standardization on the welding steam monitoring data are as follows: Welding vapor monitoring data was collected during the welding process of galvanized steel sheets. The welding vapor monitoring data included galvanized steel sheet number, weld number, target weld length, lap width, lap gap, weld centerline offset, clamping force, laser output power, welding travel speed, laser oscillation amplitude, laser oscillation frequency, shielding gas flow rate, distance from welding torch to plate surface, weld position, vapor plume height, vapor plume offset angle, vapor plume brightness, number of spatter particles, spatter particle image displacement, keyhole bright area, welding sound pressure level, and data acquisition timestamp. For the collected welding steam monitoring data, median filtering was used to remove instantaneous spike noise. The welding steam monitoring data is processed by exponential moving average to smooth continuous sampling fluctuations; short-term missing records are filled in using linear interpolation; a standard sampling time series is generated according to a preset unified sampling period, and welding steam monitoring data from different sources are mapped to the standard sampling time series using a combination of nearest neighbor timestamp matching and linear interpolation compensation, and multi-source acquisition sequence synchronization is performed on the welding steam monitoring data; the welding steam monitoring data is numerically standardized using robust standardization based on median and interquartile range, and the preprocessed welding steam monitoring data is output.

3. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 2, characterized in that: The specific steps for generating the virtual escape channel record are as follows: At each theoretical escape position, the corresponding lap gap value, lap width value, weld centerline offset value, and clamping force value are called; the lap gap value is used as the channel height of the virtual escape channel, the absolute value of the lap width value minus the weld centerline offset value is used as the effective channel width of the virtual escape channel, and the clamping force value is used as the closing load of the virtual escape channel to generate a virtual escape channel record.

4. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 3, characterized in that: The specific steps for calculating the vapor hysteresis escape mismatch value based on the virtual escape channel record are as follows: Read the virtual escape channel record and the preprocessed welding steam monitoring data at the corresponding position. Use a logic function normalization algorithm to perform dimensionless positive interval mapping processing on the steam plume height value, steam plume brightness value, spatter particle number value, spatter particle image displacement value, welding sound pressure level value, fixture clamping force value, channel height and effective channel width value respectively to obtain the steam plume height mapping value, steam plume brightness mapping value, spatter particle number mapping value, spatter particle image displacement mapping value, welding sound pressure level mapping value, fixture clamping force mapping value, channel height mapping value and effective channel width mapping value; Add one to the steam plume height mapping value to obtain the plume height amplification term; take the square root of the steam plume brightness mapping value and add one to obtain the plume brightness amplification term; calculate the tangent of the absolute value of the steam plume offset angle value and add one to obtain the offset angle amplification term; multiply the spatter particle number mapping value and the spatter particle image displacement mapping value, add one to the product, take the natural logarithm, and add one again to obtain the spatter disturbance amplification term; add one to the welding sound pressure level mapping value to obtain the sound pressure disturbance amplification term; add one to the clamping force mapping value to obtain the clamping closure amplification term; multiply the plume height amplification term, plume brightness amplification term, offset angle amplification term, spatter disturbance amplification term, sound pressure disturbance amplification term, and clamping closure amplification term to obtain the steam hysteresis response term; add the channel height mapping value to the minimum constant, and then multiply by the sum of the effective channel width mapping value and the minimum constant to obtain the escape channel constraint term; divide the steam hysteresis response term by the escape channel constraint term to obtain the steam hysteresis escape mismatch value.

5. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 4, characterized in that: The specific steps for determining the enhanced vapor disturbance section and outputting the vapor escape channel inversion results are as follows: Read the virtual escape channel records and vapor stagnation escape mismatch values ​​corresponding to each theoretical escape position value, and compare adjacent virtual escape channel records in ascending order of theoretical escape position values; when the effective channel width and channel height of the later virtual escape channel record are both less than those of the earlier virtual escape channel record, mark the theoretical escape position value corresponding to the later virtual escape channel record as a channel contraction node; when the closing load of the later virtual escape channel record is greater than that of the earlier virtual escape channel record, and the channel height of the later virtual escape channel record is less than that of the earlier virtual escape channel record, mark the theoretical escape position value corresponding to the later virtual escape channel record as a channel compression node. From the channel contraction nodes and channel compression nodes, nodes with corresponding steam stagnation escape mismatch values ​​greater than the mismatch threshold are selected, and these selected nodes are used as nodes to be checked. The N consecutive weld position values ​​behind the node to be checked along the welding travel direction are read as the check distance interval, and any two adjacent weld position values ​​within the check distance interval are used to form adjacent sampling point pairs. The steam plume height rise, keyhole bright area change, and spatter particle number increase of each adjacent sampling point pair are calculated. When there is an adjacent sampling point pair that simultaneously satisfies the following conditions: steam plume height rise is greater than the plume height rise threshold, keyhole bright area change is greater than the keyhole area change threshold, and spatter particle number increase is greater than the spatter number increase threshold, the node to be checked is determined as a steam stagnation release node. Arrange the steam stagnation release nodes in ascending order of their theoretical initiation positions. Calculate the source-location distance between adjacent steam stagnation release nodes. When the source-location distance is less than or equal to the node merging distance threshold, assign the corresponding adjacent steam stagnation release nodes to the same steam disturbance enhancement segment. Use the minimum theoretical initiation position value within the same steam disturbance enhancement segment as the segment's starting position value and the maximum theoretical initiation position value as the segment's ending position value. Output the steam escape channel inversion results.

6. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 5, characterized in that: The specific steps for generating each welding control correction based on the vapor escape channel inversion results, and determining the corresponding control start position value according to the welding travel speed value and each response delay time are as follows: Read the inversion results of the steam escape channel and the preprocessed welding steam monitoring data, subtract the mismatch threshold from the steam stagnation pressure escape mismatch value to obtain the stagnation pressure excess; multiply the stagnation pressure excess by the laser power correction coefficient, welding speed correction coefficient, oscillation amplitude correction coefficient, oscillation frequency correction coefficient, shielding gas flow rate correction coefficient, and fixture clamping force correction coefficient respectively to obtain the laser output power correction, welding travel speed correction, laser oscillation amplitude correction, laser oscillation frequency correction, shielding gas flow rate correction, and fixture clamping force correction. Multiply the welding travel speed value by the pre-calibrated laser power response delay time, laser oscillation response delay time, shielding gas flow response delay time, and fixture clamping response delay time to obtain the laser forward movement distance value, oscillation forward movement distance value, shielding gas forward movement distance value, and fixture forward movement distance value; subtract each forward movement distance value from the section start position value to obtain the laser control start position value, oscillation control start position value, shielding gas control start position value, and fixture control start position value.

7. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 6, characterized in that: The specific steps for generating each correction curve based on the corresponding control start position value, section end position value, and each welding control correction amount are as follows: Using the control start position value to the segment end position value as the corresponding correction interval, a cubic Bézier curve algorithm is used to perform position continuity processing on each correction amount. For any correction interval, the correction amount corresponding to the start position value of the correction interval is set to 0, and the correction amount corresponding to the segment end position value is set to the corresponding correction amount. The product of the correction interval length and the first Bézier smoothing coefficient is used as the first position offset. The start position value of the correction interval plus the first position offset is used as the position coordinate of the first intermediate control point, and 0 is used as the correction amount coordinate of the first intermediate control point to obtain the first intermediate control point. The product of the corresponding correction amount and the second Bézier smoothing coefficient is used as the second correction amount offset. The segment end position value minus the first position offset is used as the position coordinate of the second intermediate control point, and the corresponding correction amount minus the second correction amount offset is used as the correction amount coordinate of the second intermediate control point to obtain the second intermediate control point. Based on the starting position value of the correction interval, the ending position value of the segment, 0, the corresponding correction amount, the first intermediate control point, and the second intermediate control point, the laser output power correction curve, the welding travel speed correction curve, the laser oscillation amplitude correction curve, the laser oscillation frequency correction curve, the protective gas flow rate correction curve, and the fixture clamping force correction curve are generated.

8. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 7, characterized in that: The specific steps for fusing the overlapping correction curves according to the proportion of vapor hysteresis escape mismatch value, and then outputting the pressure relief welding collaborative control parameters after limiting the rate of change and superimposing the original control parameters are as follows: Read the correction curves corresponding to adjacent steam disturbance enhancement sections. When adjacent correction curves overlap at the weld position, take the correction curve with larger steam stagnation and escape mismatch value at the overlapping position as the main correction curve, and merge the correction curve with smaller steam stagnation and escape mismatch value into the main correction curve according to the mismatch value ratio to obtain the fusion correction curve. The mismatch ratio is the ratio of the smaller steam stagnation pressure escape mismatch to the larger steam stagnation pressure escape mismatch; the ratio of the absolute value of the correction difference corresponding to adjacent weld positions to the weld position difference is calculated to obtain the unit position change rate. When the unit position change rate is greater than the change rate limit, the correction amount corresponding to the next weld position is truncated and corrected according to the change rate limit to obtain the limit correction curve. The limiting correction curve is superimposed with the laser output power value, welding travel speed value, laser oscillation amplitude value, laser oscillation frequency value, shielding gas flow rate value, and fixture clamping force value at the corresponding weld position point to obtain the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target fixture clamping force value, and output the pressure relief welding collaborative control parameters.

9. The galvanized sheet welding collaborative control system based on steam disturbance sensing according to claim 8, characterized in that: The specific steps for reading the pressure-relief welding collaborative control parameters, matching the target control parameters corresponding to the current welding position according to the weld number and weld position value, and outputting the target control parameters to the corresponding welding actuator are as follows: The pressure relief welding collaborative control parameter table is read according to the weld number. Based on the current weld position value fed back by the encoder of the welding motion mechanism within the preset control cycle, the position interval is searched in the pressure relief welding collaborative control parameter table. The pressure relief welding collaborative control parameters at both ends of the position interval are linearly interpolated to obtain the target laser output power value, target welding travel speed value, target laser oscillation amplitude value, target laser oscillation frequency value, target shielding gas flow rate value, and target clamping force value corresponding to the current welding position. The target laser output power value is output to the laser welding power source, the target welding travel speed value is output to the welding motion mechanism, the target laser oscillation amplitude value and the target laser oscillation frequency value are output to the laser oscillation mechanism, the target shielding gas flow rate value is output to the shielding gas supply mechanism, and the target clamping force value is output to the clamping mechanism. The parameter matching and output process is repeated when the current weld position value is updated.

Images