CN-122007702-A - Intelligent detection system for laser welding quality of new energy automobile battery module

CN122007702ACN 122007702 ACN122007702 ACN 122007702ACN-122007702-A

Abstract

The invention belongs to the technical field of welding quality detection, and discloses an intelligent detection system for the laser welding quality of a new energy automobile battery module; the welding quality detection system comprises a data acquisition module, a welding compensation module, a reflection elimination module, a virtual combination area analysis module, a virtual combination early warning module and an overall evaluation module, wherein the data acquisition module acquires laser welding data in real time and carries out data cleaning, the welding compensation module carries out welding stage judgment and carries out phase compensation based on stage judgment results, the reflection elimination module carries out spectral response analysis and carries out reflection interference elimination based on spectral analysis results, the virtual combination area analysis module carries out three-dimensional section analysis, extracts geometric parameters of fusion welding errors, identifies a welding virtual combination area and carries out marking, the virtual combination early warning module carries out virtual combination area shielding to obtain virtual combination early warning information, the overall evaluation module carries out overall quality evaluation to obtain welding quality evaluation records and sends the welding quality evaluation records to a preset control end, and the data reliability and engineering adaptability of welding quality detection are improved.

Inventors

CHEN FENG
CHEN BIN
Ling Baiwei
PAN MINGCHUN

Assignees

深圳市宇盛光电有限公司

Dates

Publication Date: 20260512
Application Date: 20251224

Claims (10)

1. New energy automobile battery module laser welding quality intelligent detection system, its characterized in that includes: The data acquisition module acquires laser welding data in real time and performs data cleaning to obtain a welding multisource monitoring data set; The welding compensation module is used for judging welding phases based on the welding multi-source monitoring data set, and carrying out phase compensation on the welding multi-source monitoring data set based on a phase judgment result to obtain a reversing correction monitoring data set; the reflection elimination module is used for carrying out spectral response analysis on the direction-changing correction monitoring data set, carrying out reflection interference elimination based on a spectral analysis result and outputting a melting width accurate data set; The virtual fusion area analysis module is used for carrying out three-dimensional section analysis based on the fusion width accurate data set and extracting fusion welding error geometric parameters; The virtual fusion early warning module is used for shielding a virtual fusion area of the fusion welding mark data set to obtain virtual fusion early warning information; And the whole evaluation module is used for carrying out overall quality evaluation by combining the fusion welding mark data set and the virtual fusion warning information, obtaining a welding quality evaluation record and sending the welding quality evaluation record to a preset control end, and each module is connected in a wired and/or wireless mode.
2. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 1, wherein the method for performing the welding stage judgment comprises: Grouping welding multisource monitoring data sets based on a preset sampling period to obtain a periodic welding data subset; Calculating the motion direction angle of adjacent welding path coordinates in a welding path coordinate sequence, calculating a first derivative of the motion direction angle to obtain an angle change rate, and drawing an angle change rate curve based on the angle change rate; And simultaneously identifying the welding direction of the corresponding penetration data section, and if the average change rate is higher than a fluctuation threshold value and the welding direction is reversed in the corresponding time section, determining the time section corresponding to the current candidate reversing node as a swinging reversing stage.
3. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 2, wherein the mode of performing phase compensation comprises: if the welding speed of the sampling point is smaller than a set reversing threshold value and the front and rear welding directions of the sampling point are reversed, judging the sampling point as a reversing point, and constructing a local response window based on the reversing point as a center; Calculating the maximum difference value of the reflected signal intensity of the local response window, and if the maximum difference value is larger than a preset proportion, determining the reversing point corresponding to the window as a collapse interference point; calculating the penetration average value of penetration data in a local response window in adjacent front and rear time sections to respectively obtain a front average value and a rear average value; calculating the relative position proportion of a collapse interference point corresponding window in the swing reversing stage, and carrying out sectional weighting on a front average value and a rear average value based on the relative position proportion to obtain a penetration phase compensation value; And replacing the penetration data corresponding to the collapse interference point by using the penetration phase compensation value, and integrating the corrected penetration data and the rest data to obtain a reversing correction monitoring data set.
4. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 3, wherein the method for performing spectral response analysis comprises: Extracting structural light welding image data positioned in a welding line area in the reversing correction monitoring data set, and screening structural gray image data of a plurality of wavelength channels; Counting gray values of different wavelength channels in each structure gray image, and calculating pixel gray ratio of adjacent wavelength channels in the same structure gray image; Calculating the change slope of the continuous pixel gray ratio, and identifying a pixel region with continuous descending trend of the change slope in the structural gray image, wherein if the average value of the pixel gray ratio of the pixel region is lower than a preset reflection unbalance threshold value, the region is judged to be a spectral response attenuation region; And extracting welded temperature data, judging whether the welded temperature of the spectral response attenuation region is higher than a preset temperature standard or not based on the welded temperature data, and determining the corresponding spectral response attenuation region as an oxidation interference region if the welded temperature is higher than the preset temperature standard, namely a spectral analysis result.
5. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 4, wherein the method for eliminating reflection interference comprises the following steps: identifying continuous pixel segments with the oxidation interference area coincident with the welding line area, counting the gray values of the continuous pixel segments, and drawing a transverse gray gradient change curve based on the gray values; calculating the gradient change rate of a transverse gray level gradient change curve, and judging the area where the corresponding continuous pixel section is located as an edge error area if the gradient change rate is lower than a preset abrupt threshold; the method comprises the steps of extracting a weld joint center line coordinate sequence of a normal region, carrying out matching fitting on an edge error region based on the weld joint center line coordinate sequence, and outputting a reasonable offset interval; The method comprises the steps of carrying out edge recognition on a compensation pixel segment again to obtain a compensation welding seam edge contour, calculating a transverse fusion width value based on the compensation welding seam edge contour, carrying out fusion width value fitting on a corresponding oxidation interference area based on the transverse fusion width value, outputting the fusion width value of the area after adjustment, and integrating other data to obtain a fusion width accurate data set.
6. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 5, wherein the method for performing three-dimensional section analysis comprises: dividing a melting width accurate data set into equidistant sampling section data based on a welding path coordinate sequence, and extracting melting depth data, melting width data and structural light welding image data in the equidistant sampling section data; constructing a cross-section coordinate system of each equidistant sampling section based on penetration data and penetration data, and generating a contour map of a welding seam cross section by combining gray information of a welding seam area in the structured light welding image data; and carrying out geometric analysis on the profile map to obtain fusion welding error set parameters including fusion width, fusion depth, sidewall taper, bottom closure degree and weld symmetry deviation rate.
7. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 6, wherein the method for identifying the welding virtual joint area comprises: Sequentially reading fusion welding error set parameters corresponding to each equidistant sampling section, and calculating the ratio of the maximum penetration to the maximum fusion width to obtain the welding transverse-longitudinal ratio; calculating the average value of the taper of the side walls at two sides of the welding section to obtain average taper, and fitting the minimum curvature radius based on the profile diagram of the welding section; constructing a judging condition and setting a corresponding threshold value, wherein the judging condition comprises that the welding transverse-longitudinal ratio is smaller than the corresponding ratio threshold value, the average taper is larger than the corresponding threshold value, the bottom closure degree is smaller than the corresponding closure threshold value, the weld symmetry deviation rate is higher than the corresponding ratio threshold value, and the minimum curvature radius is smaller than the corresponding curvature threshold value; if the continuous adjacent virtual combination area exists or the adjacent equidistant sampling section continuously appears to meet any two judging conditions, and the penetration change rate is lower than a preset penetration change threshold value, the area formed by the corresponding equidistant sampling section is judged to be a continuous virtual combination section; And marking the coordinates of the single virtual joint area and the continuous virtual joint section, and integrating all the data to obtain a fusion welding marking data set.
8. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 7, wherein the mode of shielding the virtual fusion area comprises: extracting a welding path coordinate sequence in a fusion welding mark data set, identifying overlap section coordinates of a welding line area, and integrating the overlap section coordinates into a process special section coordinate set; the method comprises the steps of obtaining a coordinate set of a process special section, reading a welding time stamp corresponding to the coordinate set of the process special section, extracting laser welding power and welding speed corresponding to the welding time stamp, carrying out heat energy calculation on the process special section based on the laser welding power and the welding speed, and outputting heat energy density; And when the potential remelting section overlaps with the marked virtual merging area or the structural parameter fluctuation of the potential remelting section is larger than a preset parameter fluctuation threshold, marking the corresponding potential remelting section as a virtual melting area and shielding, and outputting virtual melting early warning information.
9. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 8, wherein the mode for performing curvature analysis and judgment comprises: And calculating structural parameters of structural light welding image data corresponding to the waste heat sections, and if the change angle of the expansion angle of the boundary section is larger than the expansion angle threshold value, the lower boundary average gray value is smaller than the preset gray scale proportion of the gray average value of the rest areas and the upper boundary curvature change rate is higher than the preset fluctuation threshold value, judging the corresponding waste heat section as a potential remelting section.
10. The intelligent detection system for laser welding quality of a battery module of a new energy automobile according to claim 9, wherein the method for performing overall quality assessment comprises: Acquiring a zone number and a welding path coordinate of a mark in a fusion welding mark data set, and comparing specific parameters of a corresponding zone with a history normal welding record to judge a risk zone; and integrating the section number of the risk section, the welding path coordinates, the welding quality grade and the corresponding specific parameters to obtain a welding quality evaluation record.

Description

Intelligent detection system for laser welding quality of new energy automobile battery module Technical Field The invention relates to the technical field of welding quality detection, in particular to an intelligent detection system for the laser welding quality of a new energy automobile battery module. Background The manufacturing quality of the power battery module serving as a core component directly determines the safety performance and the service life of the whole vehicle, wherein laser welding is widely applied as a main technical means in a connecting process, welding quality detection is used as a core link for guaranteeing the reliability of the battery module and directly influences the safety and the stability of the battery module, and a conventional welding quality detection method commonly used in engineering practice at present can realize a conventional defect identification function, but is limited in more complex welding scenes and still faces a plurality of technical bottlenecks. In the actual welding scene, swing laser welding is often adopted, in the process, a laser welding head is decelerated to zero at a reversing point and moves reversely, so that a short-time collapse phenomenon occurs in a welding hole at the position, depth data acquired by an OCT probe at the moment shows periodic instantaneous fluctuation, the conventional welding quality detection method is lack of sensing capability on the relation between penetration and a moving path, the natural shallow melting phenomenon is easily misjudged to be insufficient in edge penetration, evaluation precision is influenced, in addition, when a welding line contour is scanned, the surface of an aluminum alloy or copper alloy part is often accompanied with generation of a color oxide film at high temperature, diffuse reflection inhibition is generated on a main laser wave band by a color ribbon, the area is misjudged to be a concave boundary by the conventional welding quality detection method, and further, excessive estimation deviation of the fusion width is caused, meanwhile, in the welding process, partial gaps at the lap joint position can form virtual empty fusion welding pinholes, the fusion penetration reflected by an OCT signal is similar to conform to the fusion penetration requirement, in fact, the conventional welding quality detection method is lack of recognition capability on the geometrical characteristics of a hole shape structure and a side wall, on the other hand, the condition that the tail lap joint area of a welding seam is often damaged by secondary heating, the primary lap joint area is often caused, the surface is always damaged, the surface of the primary position is connected, the conventional welding quality is not is easily ignored, and the lap joint quality is not stable, and the lap joint quality is easily detected, and the lap joint quality of the lap joint quality is easily is not easy to be ignored. In view of the above, the present invention proposes an intelligent detection system for laser welding quality of a battery module of a new energy automobile to solve the above-mentioned problems. Disclosure of Invention In order to overcome the defects in the prior art and achieve the purposes, the invention provides the following technical scheme that the intelligent detection system for the laser welding quality of the new energy automobile battery module comprises: The data acquisition module acquires laser welding data in real time and performs data cleaning to obtain a welding multisource monitoring data set; The welding compensation module is used for judging welding phases based on the welding multi-source monitoring data set, and carrying out phase compensation on the welding multi-source monitoring data set based on a phase judgment result to obtain a reversing correction monitoring data set; the reflection elimination module is used for carrying out spectral response analysis on the direction-changing correction monitoring data set, carrying out reflection interference elimination based on a spectral analysis result and outputting a melting width accurate data set; The virtual fusion area analysis module is used for carrying out three-dimensional section analysis based on the fusion width accurate data set and extracting fusion welding error geometric parameters; The virtual fusion early warning module is used for shielding a virtual fusion area of the fusion welding mark data set to obtain virtual fusion early warning information; And the whole evaluation module is used for carrying out overall quality evaluation by combining the fusion welding mark data set and the virtual fusion warning information, obtaining a welding quality evaluation record and sending the welding quality evaluation record to a preset control end, and each module is connected in a wired and/or wireless mode. Further, the method for judging the welding stage includes: Grouping welding multisource monitoring data sets