A method and system for controlling weld penetration in laser-arc hybrid welding

Abstract

The present application relates to the technical field of welding process, disclose a kind of laser electric arc composite welding weld penetration control method and system, the method includes: based on the infrared thermal radiation signal of welding pool constructs pool three-dimensional temperature field;Based on the specific element spectral line intensity of the plasma optical spectrum signal in the welding pool, the pool metal vapor concentration is calibrated;The pool three-dimensional temperature field and the pool metal vapor concentration are coupled analysis, and the pool state characteristics of the welding pool are obtained;Based on the pool state characteristics dynamic division laser-arc energy distribution ratio;Based on the laser-arc energy distribution ratio, the offset position of laser focal point in weld depth direction is adjusted, and the arc voltage waveform is adjusted by inverter power supply;The present application can improve the efficiency of a kind of laser electric arc composite welding weld penetration control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding process, in particular to a weld penetration control method and system for laser-arc hybrid welding. BACKGROUND

[0002] In the process of laser-arc hybrid welding, weld penetration is a key parameter affecting welding quality, and accurate control of penetration is crucial to ensure the mechanical properties and reliability of the welded joint. However, the existing technology in the weld penetration control of laser-arc hybrid welding faces the problem of incomplete and inaccurate acquisition of molten pool state information, which cannot effectively construct the three-dimensional temperature field of the molten pool and calibrate the metal vapor concentration, making it difficult to accurately grasp the dynamic change characteristics of the molten pool, and thus the penetration control precision cannot be guaranteed.

[0003] The existing technology also has the defects of insufficient analysis of molten pool state characteristics and lack of dynamic adaptability of energy distribution strategy, which cannot realize reasonable distribution of laser and arc energy based on the flowability and stability parameters of the molten pool, and cannot adjust the laser focal point position and arc voltage waveform in real time according to the molten pool oscillation state, resulting in poor weld penetration control effect in the welding process, which is difficult to meet the demand of high-quality welding. SUMMARY

[0004] The present application provides a weld penetration control method and system for laser-arc hybrid welding, which mainly aims to solve the problem of low efficiency in weld penetration control of laser-arc hybrid welding.

[0005] To achieve the above-mentioned purpose, the present application provides a weld penetration control method for laser-arc hybrid welding, which comprises:

[0006] S1. Constructing a three-dimensional temperature field of the molten pool based on the infrared thermal radiation signal of the welding molten pool;

[0007] S2. Calibrating the metal vapor concentration of the molten pool based on the specific element spectral line intensity of the plasma spectrum signal in the welding molten pool;

[0008] S3. Coupling analysis of the three-dimensional temperature field of the molten pool and the metal vapor concentration of the molten pool to obtain the molten pool state characteristics of the welding molten pool;

[0009] S4. Dynamically dividing the laser-arc energy distribution ratio based on the molten pool state characteristics;

[0010] S5. Adjusting the offset position of the laser focal point in the weld depth direction based on the laser-arc energy distribution ratio, and adjusting the arc voltage waveform through the inverter power supply.

[0011] In a preferred embodiment, the three-dimensional temperature field of the molten pool is constructed based on the infrared thermal radiation signal of the welding molten pool, which comprises:

[0012] Multispectral infrared thermal imaging was performed on the weld pool to obtain the multi-band thermal radiation distribution on the surface of the weld pool;

[0013] Based on the multi-band thermal radiation distribution, the depth characteristic spectrum is separated by color difference to obtain the thermal radiation gradient in the direction of the molten pool depth.

[0014] The thermal radiation gradient is mapped onto a spatial filter array to obtain the three-dimensional voxel temperature distribution of the molten pool;

[0015] The isothermal boundary of the three-dimensional voxel temperature distribution is reconstructed to obtain the three-dimensional temperature field of the molten pool.

[0016] In a preferred embodiment, the calibration of the molten pool metal vapor concentration based on the intensity of specific elemental spectral lines in the plasma spectral signal of the weld pool includes:

[0017] The plasma spectral signal is acquired by dual-path beam splitting to obtain synchronous light intensity signals of characteristic spectral lines and reference spectral lines;

[0018] The synchronous light intensity signal is differentially amplified to obtain the relative intensity ratio of the elemental spectral lines of the molten pool;

[0019] The relative intensity ratio of the element spectral lines is linearly calibrated by photoelectric conversion to obtain the molten pool metal vapor concentration gradient value;

[0020] Based on the steam concentration gradient value, the boundary of the concentration distribution field is reconstructed to obtain the molten pool metal steam concentration field.

[0021] In a preferred embodiment, the step of coupling the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics includes:

[0022] By synchronizing the dynamic distribution data of the three-dimensional temperature field of the molten pool with the real-time monitoring data of the metal vapor concentration of the molten pool, a coupled dataset of the molten pool is obtained;

[0023] Based on the coupled dataset, the temperature gradient and the steam concentration gradient are vector-superimposed to obtain the interaction characteristics of energy transfer and mass migration inside the molten pool.

[0024] Based on the interaction features, the convection intensity distribution and vaporization suppression region boundary in the molten pool fluid are identified to obtain the molten pool state characteristics of the welding molten pool.

[0025] In a preferred embodiment, the identification of the convection intensity distribution and the vaporization suppression region boundary in the molten pool fluid based on the interaction feature obtains the molten pool state feature of the welding molten pool, including:

[0026] Collecting the plasma light ripple pattern and the molten pool surface ripple direction;

[0027] Determining the vaporization suppression region boundary according to the contraction width of the plasma light ripple pattern;

[0028] Identifying the convection intensity distribution in the molten pool fluid based on the convergence direction of the molten pool surface ripple direction, and taking the identified convection intensity distribution as the molten pool state feature of the welding molten pool.

[0029] In a preferred embodiment, the dynamic division of the laser-arc energy distribution ratio based on the molten pool state feature includes:

[0030] Based on the molten pool state feature, orthogonal physical parameters are separated to obtain the flowability parameter and the stability parameter of the molten pool;

[0031] Superimposing the flowability parameter and the stability parameter to obtain the dynamic coupling effect of laser energy and arc energy;

[0032] Adaptive distribution is performed on the real-time change trend of the dynamic coupling relationship to obtain the laser-arc energy distribution ratio.

[0033] In a preferred embodiment, the orthogonal physical parameter separation based on the molten pool state feature obtains the flowability parameter and the stability parameter of the molten pool, including:

[0034] Mapping the heat flow vector in the molten pool state feature to the heat transfer dominant axis to obtain a first orthogonal component representing energy transfer;

[0035] Mapping the melt flow vector in the molten pool state feature to the material migration dominant axis to obtain a second orthogonal component representing material transport;

[0036] Taking the module length of the first orthogonal component as the molten pool flowability parameter, and taking the direction consistency of the second orthogonal component as the molten pool stability parameter.

[0037] In a preferred embodiment, the adjustment of the offset position of the laser focal point in the weld depth direction based on the laser-arc energy distribution ratio includes:

[0038] Real-time monitoring of the molten pool oscillation state of the weld molten pool obtains the oscillation frequency of the molten pool;

[0039] Perform dynamic response analysis on the degree of deviation of the oscillation frequency from the preset reference frequency to obtain a molten pool depth control signal of the molten pool.

[0040] Perform displacement instruction transmission on the servo motor based on the molten pool depth control signal to obtain a depth direction displacement of the molten pool.

[0041] Adjust the laser focal point position based on the depth direction displacement to obtain an offset position of the weld seam in the depth direction.

[0042] In a preferred embodiment, the arc voltage waveform is adjusted by the inverter power supply, comprising:

[0043] Extract the frequency variation in the oscillation frequency to obtain an oscillation frequency variation of the molten pool.

[0044] Map the modulation response characteristic to the oscillation frequency variation to obtain an arc voltage waveform parameter of the molten pool.

[0045] Reconstruct the inverter power supply switch sequence based on the arc voltage waveform parameter to obtain a modulated arc voltage waveform.

[0046] To solve the above problems, the application also provides a weld seam penetration control system for laser-arc hybrid welding, comprising:

[0047] A molten pool three-dimensional temperature field construction module constructs a molten pool three-dimensional temperature field based on an infrared thermal radiation signal of the welding molten pool.

[0048] A molten pool metal vapor concentration calibration module calibrates the molten pool metal vapor concentration based on the specific element spectral line intensity of the plasma spectrum signal in the welding molten pool.

[0049] A comprehensive molten pool state feature module couples and analyzes the molten pool three-dimensional temperature field and the molten pool metal vapor concentration to obtain a molten pool state feature of the welding molten pool.

[0050] A laser-arc energy distribution ratio module dynamically divides the laser-arc energy distribution ratio based on the molten pool state feature.

[0051] An arc voltage waveform adjustment module adjusts the offset position of the laser focal point in the depth direction of the weld seam based on the laser-arc energy distribution ratio, and adjusts the arc voltage waveform by the inverter power supply.

[0052] 1. By constructing the molten pool three-dimensional temperature field based on the infrared thermal radiation signal of the welding molten pool, calibrating the molten pool metal vapor concentration based on the specific element spectral line intensity of the plasma spectrum signal, and coupling and analyzing the molten pool state characteristics, the temperature distribution, metal vapor concentration and their interaction characteristics of the molten pool can be comprehensively and accurately obtained, which provides accurate state basis for the penetration control, thereby effectively improving the accuracy of the weld penetration control.

[0053] 2. The laser-arc energy distribution ratio is dynamically divided based on the molten pool state characteristics, and the laser focal point position in the weld depth direction is adjusted accordingly, and the arc voltage waveform is adjusted through the inverter power supply, so that the energy distribution ratio, focal point position and arc voltage waveform and other parameters can accurately match the change of the molten pool state, the accuracy of the parameters is significantly improved, and the weld penetration control effect is optimized. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 A flowchart of a weld penetration control generation method for laser-arc hybrid welding is provided for an embodiment of the present application.

[0055] Figure 2 A functional module diagram of a weld penetration control generation system for laser-arc hybrid welding is provided for an embodiment of the present application.

[0056] The implementation of the present application, functional characteristics and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0057] It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0058] The embodiments of the present application provide a weld penetration control generation method for laser-arc hybrid welding. The execution subject of the weld penetration control generation method for laser-arc hybrid welding includes but is not limited to at least one of electronic devices such as a server, a terminal and the like which can be configured to execute the method provided by the embodiments of the present application. In other words, the weld penetration control generation method for laser-arc hybrid welding can be executed by software or hardware installed in a terminal device or a server device. The server includes but is not limited to a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be a stand-alone server, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content distribution networks (CDN), and big data and artificial intelligence platforms, etc. basic cloud computing services.

[0059] Referring to Figure 1 Fig. 1 is a flowchart of a weld penetration control method for laser-arc hybrid welding according to an embodiment of the present application. In this embodiment, the reference information generation method based on artificial intelligence and smart home includes:

[0060] S1. Constructing a molten pool three-dimensional temperature field based on an infrared thermal radiation signal of the molten pool;

[0061] S2. Calibrating a molten pool metal vapor concentration based on a specific element spectral line intensity of a plasma spectrum signal in the molten pool;

[0062] S3. Coupling analysis of the molten pool three-dimensional temperature field and the molten pool metal vapor concentration to obtain a molten pool state feature of the molten pool;

[0063] S4. Dynamically dividing a laser-arc energy distribution ratio based on the molten pool state feature;

[0064] S5. Adjusting a laser focal point offset position in a weld depth direction based on the laser-arc energy distribution ratio, and adjusting an arc voltage waveform through an inverter power supply.

[0065] S1. Constructing a molten pool three-dimensional temperature field based on an infrared thermal radiation signal of the molten pool;

[0066] In this embodiment of the present application, the molten pool three-dimensional temperature field is constructed based on the infrared thermal radiation signal of the molten pool, including:

[0067] Performing multispectral infrared thermal imaging acquisition on the molten pool to obtain a multi-band thermal radiation distribution of the molten pool surface;

[0068] Performing chromatic aberration separation on a depth feature spectrum based on the multi-band thermal radiation distribution to obtain a thermal radiation gradient in a depth direction of the molten pool;

[0069] Mapping the thermal radiation gradient to a spatial filter array to obtain a three-dimensional voxel temperature distribution of the molten pool;

[0070] Reconstructing an isothermal boundary of the three-dimensional voxel temperature distribution to obtain a molten pool three-dimensional temperature field of the molten pool.

[0071] Specifically, multispectral infrared thermal imaging acquisition is performed on the molten pool.

[0072] Further, the multispectral infrared thermal imaging equipment is aligned to the molten pool surface, and the thermal radiation signals of the molten pool surface under different infrared wavebands are synchronously acquired through multiple spectral channels of the equipment.

[0073] Further, these thermal radiation signals are processed through signal amplification and analog-to-digital conversion to be converted into digital image data, thereby obtaining a multi-band thermal radiation distribution of the molten pool surface.

[0074] Furthermore, this distribution contains information on the thermal radiation intensity of different regions on the surface of the molten pool at different wavelengths, providing accurate basic data for subsequent depth feature analysis and temperature field construction.

[0075] Specifically, color difference separation is performed on depth characteristic spectra based on multi-band thermal radiation distribution.

[0076] Furthermore, we carefully analyzed the spectral information corresponding to different wavelengths in the multi-band thermal radiation distribution and established the mapping relationship between wavelength and molten pool depth.

[0077] Furthermore, by taking advantage of the significant differences in the infrared absorption and reflection characteristics of molten pool materials at different depths, and by using spectral color difference analysis technology to set specific wavelength screening thresholds, characteristic spectra related to the depth of the molten pool can be separated from the overall spectrum.

[0078] Furthermore, by using gradient calculation methods, the rate of change of thermal radiation intensity at different locations along the depth of the molten pool is calculated, and the thermal radiation gradient along the depth of the molten pool is obtained. This gradient can accurately reflect the trend of molten pool temperature change with depth, providing key parameters for the construction of a three-dimensional temperature field.

[0079] Specifically, the thermal radiation gradient is mapped onto the spatial filter array.

[0080] Furthermore, a three-dimensional spatial filtering array model is constructed, in which each node of the array corresponds to a tiny voxel in the molten pool space, and the size of the voxel is set according to the precision requirements of the welding process.

[0081] Furthermore, based on the thermal radiation variation law in the depth direction represented by the thermal radiation gradient, the thermal radiation gradient value at each depth location is precisely assigned to the corresponding spatial filter array node according to the preset mapping rules.

[0082] Furthermore, the discrete gradient values ​​are smoothed and three-dimensional interpolated to convert them into a continuous three-dimensional spatial distribution, thereby obtaining the three-dimensional voxel temperature distribution of the molten pool. This distribution clearly describes the temperature conditions at each spatial location inside the molten pool in units of voxels, realizing the voxelized characterization of the temperature field.

[0083] Specifically, isothermal boundaries are reconstructed for the three-dimensional voxel temperature distribution.

[0084] Furthermore, all voxels in the three-dimensional voxel temperature distribution are traversed, and voxel nodes with the same temperature value are extracted according to the set temperature interval threshold.

[0085] Furthermore, by utilizing boundary detection technology in 3D image processing, and through edge tracking and contour extraction algorithms, these nodes with the same temperature are connected into a continuous curved surface.

[0086] Furthermore, the curved surface is filled and optimized to eliminate holes and burrs, thus constructing isothermal surfaces corresponding to different temperature values.

[0087] Furthermore, all isothermal surfaces are finally combined in order of temperature gradient to obtain the three-dimensional temperature field of the molten pool. This temperature field intuitively shows the temperature distribution state in the three-dimensional space inside the molten pool, providing an accurate temperature basis for subsequent molten pool state analysis and energy distribution control.

[0088] In summary, multispectral infrared thermal imaging acquires multi-band thermal radiation distribution on the surface of the molten pool, providing accurate data for depth feature analysis, improving the accuracy of temperature field construction, and enhancing the accuracy of weld penetration control.

[0089] In summary, based on the color difference separation of multi-band thermal radiation distribution, the thermal radiation gradient in the depth direction is obtained, which reflects the temperature change trend, provides key parameters for the construction of a three-dimensional temperature field, and enables dynamic adjustment of melting depth control.

[0090] In summary, the thermal radiation gradient is mapped onto the spatial filter array to obtain the three-dimensional voxel temperature distribution, realizing the voxelization characterization of the temperature field, laying the foundation for isothermal boundary reconstruction, ensuring accurate representation of the molten pool temperature, and assisting in the setting of melt depth control parameters.

[0091] In summary, the isothermal boundary reconstruction of the three-dimensional voxel temperature distribution yields the three-dimensional temperature field of the molten pool, which intuitively displays the internal temperature state and provides a basis for molten pool state analysis and energy distribution. This enables energy distribution and focus adjustment to conform to the molten pool state, thereby improving the molten depth control effect.

[0092] S2. Calibrate the metal vapor concentration in the weld pool based on the intensity of specific element spectral lines in the plasma spectral signal in the weld pool;

[0093] In this embodiment of the invention, the calibration of the molten pool metal vapor concentration based on the intensity of specific elemental spectral lines in the plasma spectral signal of the weld pool includes:

[0094] The plasma spectral signal is acquired by dual-path beam splitting to obtain synchronous light intensity signals of characteristic spectral lines and reference spectral lines;

[0095] The synchronous light intensity signal is differentially amplified to obtain the relative intensity ratio of the elemental spectral lines of the molten pool;

[0096] The relative intensity ratio of the element spectral lines is linearly calibrated by photoelectric conversion to obtain the molten pool metal vapor concentration gradient value;

[0097] Based on the steam concentration gradient value, the boundary of the concentration distribution field is reconstructed to obtain the molten pool metal steam concentration field.

[0098] Specifically, the plasma spectrum signal is collected by dual-path beam splitting.

[0099] Further, the plasma spectrum signal is split into two light signals by a beam splitter using a dual-path beam splitting system in the spectrum collection device.

[0100] Further, one of the light signals passes through a filter of a specific wavelength to obtain a light signal containing a characteristic spectrum line of a target metal element, and the other light signal passes through a filter of a reference wavelength to obtain a reference spectrum line light signal as a reference.

[0101] Further, the light intensity of the two light signals is collected in real time by a synchronous collection device to obtain a synchronous light intensity signal of the characteristic spectrum line and the reference spectrum line, which contains light intensity data of the characteristic spectrum line and the reference spectrum line at the same time.

[0102] Specifically, the synchronous light intensity signal is subjected to differential amplification processing.

[0103] Further, the light intensity signal of the characteristic spectrum line and the light intensity signal of the reference spectrum line are input into a differential amplifier, and the differential amplifier performs subtraction operation on the two signals to obtain a difference value of the two signals.

[0104] Further, the difference value signal is then amplified to enhance the intensity of the signal for subsequent analysis.

[0105] Further, the relative intensity ratio of the element spectrum line of the molten pool obtained after processing reflects the proportional relationship between the light intensity of the characteristic spectrum line and the light intensity of the reference spectrum line, and eliminates the influence of background light intensity and other interference factors.

[0106] Specifically, the relative intensity ratio of the element spectrum line is subjected to photoelectric conversion linear calibration.

[0107] Further, the light signal corresponding to the relative intensity ratio of the element spectrum line is first converted into an electrical signal by a photodetector, and the size of the electrical signal is proportional to the relative intensity ratio of the light intensity.

[0108] Further, a standard sample with a known metal vapor concentration is then used to establish a linear relationship model between the relative intensity ratio of the element spectrum line and the metal vapor concentration.

[0109] Further, the measured electrical signal is substituted into the linear relationship model to calculate the corresponding metal vapor concentration value, and further to obtain a molten pool metal vapor concentration gradient value of the molten pool, which represents the change amplitude of the metal vapor concentration at different positions in the molten pool.

[0110] Specifically, the concentration distribution field is reconstructed based on the vapor concentration gradient value.

[0111] Furthermore, the spatial range of the concentration distribution field is determined based on the geometry and size of the molten pool.

[0112] Furthermore, based on the steam concentration gradient value, the concentration value is gradually filled inward from the boundary of the molten pool.

[0113] Furthermore, by calculating the concentration changes at adjacent locations point by point and combining the gradient values ​​to determine the concentration distribution trend, a continuous concentration distribution field is constructed.

[0114] Furthermore, during the construction process, the boundary of the concentration distribution field is modified and optimized to ensure that the concentration value at the boundary conforms to the actual physical laws, and finally the molten pool metal vapor concentration field is obtained. This concentration field intuitively shows the three-dimensional distribution of metal vapor concentration in the molten pool.

[0115] In summary, dual-beam splitting of plasma spectral signals yields characteristic and reference spectral line synchronous light intensity signals, providing a precise concentration data basis for melt depth control.

[0116] In summary, differential amplification of the synchronous light intensity signal yields the relative intensity ratio of elemental spectral lines, eliminating interference to improve concentration calibration accuracy and enabling more accurate control of fluxing depth.

[0117] In summary, the photoelectric conversion linearly calibrates the intensity ratio, obtains the metal vapor concentration gradient value, and quantifies the concentration distribution to support dynamic adjustment of melting depth control.

[0118] In summary, the concentration distribution field is reconstructed based on the concentration gradient value, which intuitively presents the three-dimensional distribution of molten pool metal vapor and provides a basis for energy allocation for melting depth control.

[0119] S3. Couple the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics of the welding molten pool;

[0120] In this embodiment of the invention, the step of coupling the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics includes:

[0121] By synchronizing the dynamic distribution data of the three-dimensional temperature field of the molten pool with the real-time monitoring data of the metal vapor concentration of the molten pool, a coupled dataset of the molten pool is obtained;

[0122] Based on the coupled dataset, the temperature gradient and the steam concentration gradient are vector-superimposed to obtain the interaction characteristics of energy transfer and mass migration inside the molten pool.

[0123] Based on the interaction features, the convection intensity distribution and vaporization suppression region boundary in the molten pool fluid are identified to obtain the molten pool state characteristics of the welding molten pool.

[0124] identifying, based on the interaction feature, a convection intensity distribution and a vaporization inhibition region boundary in the molten pool fluid to obtain a molten pool state feature of the welding molten pool, including:

[0125] collecting a plasma light stripe pattern and a molten pool surface wave direction;

[0126] determining a vaporization inhibition region boundary according to the contraction width of the plasma light stripe pattern;

[0127] identifying a convection intensity distribution in the molten pool fluid based on the convergence direction of the molten pool surface wave direction, and taking the identified convection intensity distribution as the molten pool state feature of the welding molten pool.

[0128] Specifically, the dynamic distribution data of the molten pool three-dimensional temperature field and the real-time monitoring data of the molten pool metal vapor concentration are synchronized.

[0129] Further, by means of time stamp matching, the molten pool three-dimensional temperature field data and the molten pool metal vapor concentration data at the same time are associated to form a time-aligned data set.

[0130] Further, in the data synchronization process, the temperature field data and the concentration data are spatially mapped.

[0131] Further, ensure that both are analyzed in the same spatial coordinate system, so as to obtain a coupled data set of the molten pool, which contains temperature and concentration information at the same time and at the same spatial position.

[0132] Specifically, based on the coupled data set, the temperature gradient and the vapor concentration gradient are vector superimposed.

[0133] Further, at each spatial point in the coupled data set, the gradient vector of the temperature field and the gradient vector of the concentration field are calculated respectively, and the two vectors respectively represent the change rate and direction of temperature and concentration at the point.

[0134] Further, the two gradient vectors are superimposed according to the same proportion to obtain a new vector, which represents the comprehensive action direction and strength of energy transfer and mass transfer in the molten pool.

[0135] Further, the interaction feature of energy transfer and mass transfer in the molten pool is obtained, which reflects the synergistic effect of temperature and concentration change.

[0136] Specifically, based on the interaction feature, the convection intensity distribution and the vaporization inhibition region boundary in the molten pool fluid are identified.

[0137] Further, by analyzing the size and direction of the interaction feature vector, the flow trend and intensity of the fluid in the molten pool are determined, the size of the interaction feature vector represents the size of the convection intensity, and the vector direction represents the direction of the fluid flow, thereby obtaining the convection intensity distribution.

[0138] Further, according to the specific combination conditions of temperature and concentration in the interaction feature, the area with high vapor concentration and large temperature gradient change in the molten pool is identified, which corresponds to the area where the vaporization process is inhibited.

[0139] Further, the boundaries of these areas are determined by the boundary detection algorithm, and finally the molten pool state feature of the welding molten pool is obtained, which includes the information of the convection intensity distribution and the boundary of the vaporization inhibition area in the molten pool.

[0140] Specifically, the plasma light fringe pattern and the molten pool surface wave direction are collected.

[0141] Further, a high-speed camera is aligned with the welding molten pool area, and by setting a specific optical filter, the light emission pattern of the plasma light fringe and the wave pattern formed by the reflection of the molten pool surface light are captured respectively.

[0142] Further, the frame synchronization processing is performed on the collected image sequence to ensure that the plasma light fringe and the molten pool wave in each frame of image correspond to the molten pool state at the same time, thereby obtaining the image data containing the plasma light fringe pattern and the molten pool surface wave direction.

[0143] Specifically, the contraction width of the plasma light fringe pattern is used to determine the boundary of the vaporization inhibition area.

[0144] Further, the collected plasma light fringe image is subjected to grayscale processing to enhance the contrast of the fringe edge.

[0145] Further, the contour of the fringe is identified by edge detection technology, and the width of the fringe at different positions is calculated.

[0146] Further, when the fringe width is obviously contracted, the coordinate point of the position is recorded.

[0147] Further, the coordinate points of these contraction positions are connected into a continuous curve, which is the boundary of the vaporization inhibition area, because the contraction of the plasma light fringe reflects the range of the area where the vaporization process of the molten pool surface is inhibited.

[0148] Specifically, the convection intensity distribution in the molten pool fluid is identified based on the convergence direction of the molten pool surface wave direction, and the identified convection intensity distribution is used as the molten pool state feature of the welding molten pool.

[0149] Further, the molten pool surface ripple image is directionally filtered to extract the main direction information of the ripples.

[0150] Further, the convergence point and divergence point of the ripples are determined by analyzing the change trend of the adjacent ripple directions.

[0151] Further, in the convergence direction of the ripples, the fluid flow speed is faster, corresponding to higher convection intensity; and in the divergence direction of the ripples, the fluid flow speed is slower, corresponding to lower convection intensity.

[0152] Further, according to the convergence degree and direction of the ripple direction, the molten pool is divided into different convection intensity regions, so as to obtain the distribution of the convection intensity in the molten pool fluid, and the distribution is taken as the molten pool state feature of the welding molten pool.

[0153] In summary, the synchronous temperature field and concentration real-time data acquisition coupled dataset provides a multi-dimensional fusion analysis basis for the penetration control, and improves the state judgment accuracy.

[0154] In summary, the vector superposition temperature and concentration gradient obtain interactive features, reveal the synergistic effect of energy transfer and mass transfer, and provide a dynamic basis for the penetration control parameter adjustment.

[0155] In summary, the convection intensity and vaporization inhibition region are identified based on the interactive features, the molten pool state features are accurately obtained, the laser-arc energy distribution is more suitable for the actual needs of the molten pool, and the penetration control effect is optimized.

[0156] In summary, the plasma light stripe and the ripple direction of the molten pool are collected to provide a basis for the molten pool surface dynamic feature data for the penetration control.

[0157] In summary, the vaporization inhibition region boundary is determined according to the light stripe contraction width, and the vaporization state of the molten pool surface is accurately defined to help optimize the energy distribution.

[0158] In summary, the convection intensity distribution is identified according to the ripple convergence direction, the molten pool fluid dynamic features are obtained, the laser-arc energy distribution is adapted to the molten pool state, and the penetration control precision is improved.

[0159] S4, dynamically dividing the laser-arc energy distribution ratio based on the molten pool state features;

[0160] In the embodiment of the present application, the dynamic division of the laser-arc energy distribution ratio based on the molten pool state features comprises:

[0161] Based on the molten pool state features, orthogonal physical parameters are separated to obtain the flowability parameters and stability parameters of the molten pool;

[0162] Superimpose the flow parameter and the stability parameter to obtain a dynamic coupling effect of laser energy and arc energy.

[0163] Adaptively distribute the real-time change trend of the dynamic coupling relationship to obtain a laser-arc energy distribution ratio.

[0164] The orthogonal physical parameter separation based on the molten pool state feature obtains the flow parameter and the stability parameter of the molten pool, including:

[0165] Map the heat flow vector in the molten pool state feature to the heat transfer dominant axis to obtain a first orthogonal component representing energy transfer;

[0166] Map the molten material flow vector in the molten pool state feature to the material migration dominant axis to obtain a second orthogonal component representing material transport;

[0167] The modulus of the first orthogonal component is taken as the molten pool flow parameter, and the direction consistency of the second orthogonal component is taken as the molten pool stability parameter.

[0168] Specifically, the orthogonal physical parameter separation is based on the molten pool state feature.

[0169] Further, the convection intensity distribution and the vaporization suppression region boundary in the molten pool state feature are quantitatively analyzed, and the vector direction and the speed amplitude of the convection intensity distribution are converted into the flowability index of the molten pool by image processing technology. The greater the index value, the stronger the flow ability of the molten pool fluid.

[0170] Further, through the boundary stability evaluation method, the fluctuation amplitude and the maintenance time of the vaporization suppression region boundary are mapped into the stability index of the molten pool. The greater the index value, the more stable the boundary state.

[0171] Further, a two-dimensional orthogonal coordinate system is then constructed, with the flowability index as the horizontal axis and the stability index as the vertical axis, and the two indexes are projected onto the corresponding coordinate axes.

[0172] Further, the correlation between the two is eliminated by coordinate transformation to separate the two physical parameters, and finally the flow parameter and the stability parameter of the molten pool are obtained. The two parameters are independent of each other and describe the physical properties of the molten pool from different dimensions.

[0173] Specifically, the flow parameter and the stability parameter are superimposed.

[0174] Further, according to the physical mechanism of the effect of laser energy and arc energy on the molten pool, a preset weight coefficient is determined through preliminary welding process test, wherein the laser energy weight is positively correlated with the flowability parameter, and the arc energy weight is positively correlated with the stability parameter.

[0175] Further, the fluidity parameter and the stability parameter are linearly superimposed according to preset weights, and a dynamic adjustment mechanism is set in the superimposition process: when the fluidity parameter is greater than a set threshold, the weight proportion of the stability parameter is automatically reduced to avoid excessive energy allocation to stability control; when the stability parameter is greater than a set threshold, the weight proportion of the fluidity parameter is correspondingly reduced to ensure that the energy allocation meets the molten pool state demand.

[0176] Further, by the weighted superimposition method, a parameter that comprehensively reflects the dynamic coupling effect of the laser energy and the arc energy acting on the molten pool is obtained, and the parameter can reflect the interaction strength and balance relationship of the two energy forms in the molten pool.

[0177] Specifically, the real-time change trend of the dynamic coupling relationship is adaptively allocated.

[0178] Further, the dynamic coupling effect parameter is high-frequency sampled by a real-time data acquisition system, and the change slope and direction of the parameter over time are identified by a trend analysis algorithm.

[0179] Further, when the coupling effect parameter is monitored to change in the direction of enhancing fluidity, the system automatically triggers the energy allocation adjustment program, reduces the laser energy output and increases the arc energy output according to a preset proportion, to enhance the stability of the molten pool; when the coupling effect parameter changes in the direction of enhancing stability, the laser energy output is increased and the arc energy output is reduced according to an opposite proportion, to maintain the necessary fluidity of the molten pool.

[0180] Further, by continuously monitoring the change of the coupling parameter and adjusting the energy allocation in real time, the laser-arc energy allocation ratio is always matched with the dynamic change of the molten pool, and finally the laser-arc energy allocation ratio that can optimally meet the current welding demand of the molten pool state is obtained, to ensure the weld penetration control effect.

[0181] Specifically, the heat flow vector in the molten pool state feature is mapped to the heat transfer dominant axis.

[0182] Further, the three-dimensional temperature field data of the molten pool are analyzed point by point, the temperature gradient vector of each space point is calculated by comparing the temperature values of adjacent systems, and the direction of the maximum temperature gradient of the entire molten pool is determined, which is set as the direction of the heat transfer dominant axis.

[0183] Further, for each heat flow vector, the vector decomposition method is used to decompose it into a component along the direction of the heat transfer dominant axis and a component perpendicular to the axis.

[0184] Further, by coordinate transformation, the vertical component is discarded, and only the component along the axis direction is reserved, so as to obtain the first orthogonal component representing energy transmission.

[0185] Further, the component can accurately reflect the action intensity of heat flow in the dominant transmission direction, and provide key data for subsequent analysis.

[0186] Specifically, the melt flow vector in the molten pool state feature is mapped to the material migration dominant axis.

[0187] Further, the spatial analysis is performed on the molten pool metal vapor concentration field data, and the metal vapor concentration difference at different positions is calculated.

[0188] Further, the direction with the most significant concentration change, i.e., the direction with the largest vapor concentration gradient, is determined as the material migration dominant axis direction.

[0189] Further, each melt flow vector is decomposed and split into a component along the material migration dominant axis direction and a component perpendicular to the axis.

[0190] Further, by vector projection, the vertical component is discarded, and only the component along the axis direction is reserved, so as to obtain the second orthogonal component representing material transport. The component can clearly reflect the transport capacity of solute in the dominant migration direction, and provide an important basis for molten pool state analysis.

[0191] Specifically, the module length of the first orthogonal component is taken as the molten pool fluidity parameter, and the direction consistency of the second orthogonal component is taken as the molten pool stability parameter.

[0192] Further, the module length of the first orthogonal component is directly calculated, and the vector length value of the component is taken as the molten pool fluidity parameter. The numerical value directly reflects the ability of heat flow to drive fluid flow in the molten pool, and the larger the value, the stronger the fluidity.

[0193] Further, for the second orthogonal component, a specific analysis region is first defined in the molten pool, and then the cosine value of the angle between the direction of the second orthogonal component of each point in the region and the material migration dominant axis direction is calculated.

[0194] Further, all the cosine values are arithmetically averaged, and the average value is taken as the molten pool stability parameter. The closer the parameter is to 1, the more concentrated the melt flow direction, and the more stable the molten pool state, which provides a reliable stability index for energy distribution.

[0195] In summary, the orthogonal separation of the molten pool state feature obtains the fluidity and stability parameters, which provide independent quantitative indicators for energy distribution and improve the precision of molten depth control.

[0196] Overall, the two-parameter superposition laser-arc energy dynamic coupling effect is added, and the synergistic effect of the reaction energy is combined to make the energy distribution more in line with the actual needs of the molten pool.

[0197] Overall, the adaptive allocation dynamic coupling relationship changes in real time, and the energy ratio is adjusted in real time to enhance the dynamic response capability and accuracy of the penetration control.

[0198] Overall, the heat flow vector is mapped to the heat transfer dominant axis and the first orthogonal component to quantify the energy transfer intensity and provide accurate indicators for the molten pool flowability parameters to help the dynamic adaptation of the energy distribution of the penetration control.

[0199] Overall, the molten flow vector is mapped to the material migration dominant axis and the second orthogonal component to represent the stability of material transport through direction consistency to provide a basis for state evaluation for the penetration control.

[0200] Overall, the first orthogonal component module length is used as the flowability parameter, and the second orthogonal component direction consistency is used as the stability parameter to realize the orthogonal separation of the molten pool state, lay a quantitative foundation for the dynamic allocation of laser-arc energy, and improve the precision of the penetration control.

[0201] S5, based on the laser-arc energy distribution ratio, adjust the offset position of the laser focal point in the weld depth direction, and adjust the arc voltage waveform through the inverter power supply.

[0202] In the embodiment of the present application, the offset position of the laser focal point in the weld depth direction based on the laser-arc energy distribution ratio comprises:

[0203] The molten pool oscillation state of the weld pool is monitored in real time to obtain the oscillation frequency of the molten pool;

[0204] The deviation degree of the oscillation frequency from the preset reference frequency is analyzed to obtain the penetration control signal of the molten pool;

[0205] Based on the penetration control signal, the displacement command of the servo motor is transmitted to obtain the depth direction displacement of the molten pool;

[0206] Based on the depth direction displacement, the laser focal point position is adjusted to obtain the offset position in the weld depth direction.

[0207] The arc voltage waveform is adjusted through the inverter power supply, comprising:

[0208] The frequency change amount in the oscillation frequency is extracted to obtain the oscillation frequency change amount of the molten pool;

[0209] The modulation response characteristic is mapped to the oscillation frequency change amount to obtain the arc voltage waveform parameter of the molten pool;

[0210] Reconstruct the inverter power switch sequence based on the arc voltage waveform parameters to obtain the modulated arc voltage waveform.

[0211] Specifically, the molten pool oscillation state of the weld pool is monitored in real time to obtain the oscillation frequency of the molten pool.

[0212] Further, a high-precision oscillation sensor is installed on the welding equipment, and the sensing end of the sensor is aligned with the surface of the weld pool to ensure that the sensor can accurately capture the vibration signals of the molten pool.

[0213] Further, the sensor is used to collect the vibration signals of the molten pool surface in real time, and the signals are transmitted to a signal processing unit.

[0214] Further, the collected vibration signals are filtered to remove environmental noise and irrelevant frequency interference using a band-pass filter, and effective signals related to the molten pool oscillation are retained.

[0215] Further, the vibration signals in the time domain are converted into frequency domain signals by a frequency spectrum analysis method, and the energy-concentrated frequency components are identified in the frequency domain, from which the main frequency components of the molten pool oscillation are extracted to obtain the oscillation frequency of the molten pool, which accurately reflects the dynamic vibration characteristics of the molten pool during welding.

[0216] Specifically, the deviation degree of the oscillation frequency from the preset reference frequency is dynamically analyzed to obtain the molten depth control signal of the molten pool.

[0217] Further, the real-time monitored molten pool oscillation frequency is compared with the preset reference frequency, and the frequency difference between the two is calculated.

[0218] Further, a grading standard for the deviation degree is established according to the size and sign of the frequency difference, for example, a plurality of different deviation intervals are set, and each interval corresponds to a different control signal intensity level.

[0219] Further, when the frequency difference is within the allowed normal range, a control signal is generated to maintain the current molten depth state; when the frequency difference exceeds the normal range, a corresponding adjustment control signal is generated according to the direction and degree of deviation, thereby obtaining the molten depth control signal of the molten pool, which clearly indicates the direction and specific amplitude of the molten depth adjustment required.

[0220] Specifically, the servo motor is transmitted with displacement instructions based on the molten depth control signal to obtain the depth direction displacement of the molten pool.

[0221] Further, the melt depth control signal is input into the controller of the servo motor, the controller decodes the input signal, analyzes the direction and size information of the displacement, and then amplifies the signal to generate a corresponding displacement instruction pulse.

[0222] Further, after receiving the instruction pulse, the servo motor starts the encoder feedback system to monitor the position of the motor rotor in real time through the encoder and accurately control the rotation angle of the motor rotor.

[0223] Further, through a transmission mechanism such as a ball screw or a gear rack mechanism, the rotational motion of the motor is converted into linear motion to drive the relevant components to produce displacement in the depth direction of the molten pool, and the size and direction of the displacement are strictly determined by the melt depth control signal, ensuring the accuracy and timeliness of the displacement and meeting the real-time requirements of melt depth control.

[0224] Specifically, the laser focal point position is adjusted based on the depth direction displacement to obtain the offset position in the depth direction of the weld.

[0225] Further, the depth direction displacement generated by the servo motor is transmitted to the adjustment mechanism of the laser focusing system through mechanical connection, which is usually composed of precision guide rails and lens supports and can realize high-precision linear movement.

[0226] Further, when the adjustment mechanism receives the displacement signal, it drives the laser focusing lens to move smoothly in the depth direction of the weld to change the focusing position of the laser beam.

[0227] Further, the actual position of the laser focal point is monitored in real time by the position sensor installed in the focusing system, and the actual position is compared with the target position. If there is a deviation, fine adjustment is performed to ensure that the focusing lens moves to the precise position and finally obtains the offset position in the depth direction of the weld.

[0228] Further, the position accurately reflects the adjustment result of the laser focal point in the depth direction, which can well adapt to the current molten pool state and realize accurate control of the weld penetration.

[0229] Specifically, the frequency change amount in the oscillation frequency is extracted to obtain the oscillation frequency change amount of the molten pool.

[0230] Further, time series analysis is performed on the real-time monitored molten pool oscillation frequency, and the oscillation frequency is sampled at a fixed sampling interval at consecutive time points to ensure the time consistency of the sampling data.

[0231] Further, the frequency difference between adjacent sampling points is calculated to obtain the original data of the frequency change. Through low-pass filtering processing, the appropriate cut-off frequency is set to remove the high-frequency noise in the frequency difference, which may be caused by environmental interference or equipment error.

[0232] Further, the frequency change trend of the real reaction molten pool state change is retained to obtain the oscillation frequency change amount of the molten pool, which can sensitively reflect the dynamic change of the internal energy state of the molten pool, and provide an accurate basis for subsequent voltage waveform adjustment.

[0233] Specifically, the modulation response characteristic is mapped to the oscillation frequency change amount to obtain the arc voltage waveform parameter of the molten pool.

[0234] Further, a modulation response characteristic database is established in advance through a large number of welding experiments and data analysis.

[0235] Further, the database records in detail the mapping relationship between different oscillation frequency change amounts and corresponding optimal arc voltage waveform parameters.

[0236] Further, according to the currently obtained oscillation frequency change amount, a linear interpolation or nearest neighbor search method is used in the database to find the most matched mapping entry.

[0237] Further, the corresponding arc voltage waveform parameter is extracted to ensure that the arc energy input and the molten pool oscillation state form the best coupling effect, so that the arc energy can better adapt to the dynamic change of the molten pool.

[0238] Specifically, the inverter power supply switching sequence is reconstructed based on the arc voltage waveform parameter to obtain the modulated arc voltage waveform.

[0239] Further, the extracted arc voltage waveform parameter is input to the control unit of the inverter power supply, and the processor built in the control unit performs calculation and processing according to the waveform parameter to generate a corresponding pulse width modulation signal.

[0240] Further, the pulse signal is encoded by a logic circuit to convert the digital signal into a driving signal capable of controlling the power switching device, and the on and off timing of the power switching device in the inverter power supply is reconstructed to form a specific switching sequence.

[0241] Further, the power switching device accurately acts according to the reconstructed switching sequence to convert the direct current power supply input into alternating current output with specific waveform characteristics, and after filtering and shaping processing, the modulated arc voltage waveform is finally obtained, which can accurately match the energy demand of the current state of the molten pool, and realize precise regulation and control of the arc energy.

[0242] In summary, the oscillation frequency is obtained by monitoring the oscillation state of the molten pool in real time, dynamic data support is provided for the penetration control, and timely response of the focal point adjustment to the molten pool change is ensured.

[0243] In summary, the penetration control signal is obtained by analyzing the degree of deviation of the oscillation frequency from the reference frequency, the frequency difference is converted into a precise regulation instruction, and the pertinence of the penetration control is improved.

[0244] In summary, the depth direction displacement is generated by driving the servo motor based on the control signal, the millimeter-level precise displacement is realized through the encoder feedback and the transmission mechanism, and the focal point adjustment accuracy is ensured.

[0245] In summary, the laser focal point position is adjusted according to the depth displacement and is corrected in real time, the laser energy acting depth dynamically adapts to the molten pool state, and the accuracy of the weld penetration control is directly improved.

[0246] In summary, the oscillation frequency change amount is extracted, real-time dynamic basis is provided for the arc voltage waveform adjustment, and the energy output adapts to the molten pool state.

[0247] In summary, the arc voltage waveform parameters are obtained by mapping the modulation response characteristics to the frequency change amount, and the voltage waveform is accurately matched with the molten pool oscillation.

[0248] In summary, the inverter power supply switching sequence is reconstructed, the modulated arc voltage waveform is generated, the arc energy dynamic regulation is realized, and the penetration control precision is improved.

[0249] As shown in Figure 2 Fig. 1 is a functional module diagram of a weld penetration control system for laser-arc hybrid welding according to an embodiment of the present application.

[0250] The weld penetration control system 100 for laser-arc hybrid welding according to the present application can be installed in an electronic device. According to the functions implemented, the weld penetration control system 100 for laser-arc hybrid welding can include a molten pool three-dimensional temperature field construction module 101, a molten pool metal vapor concentration calibration module 102, a comprehensive molten pool state feature module 103, a laser-arc energy distribution ratio module 104, and an arc voltage waveform adjustment module 105. The modules according to the present application can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can complete a fixed function, and are stored in the memory of the electronic device.

[0251] In the present embodiment, the functions of each module / unit are as follows:

[0252] The molten pool three-dimensional temperature field construction module 101 constructs a molten pool three-dimensional temperature field based on an infrared thermal radiation signal of the welding molten pool.

[0253] The molten pool metal vapor concentration calibration module 102 calibrates the molten pool metal vapor concentration based on the specific element spectral line intensity of the plasma spectrum signal in the welding molten pool;

[0254] The comprehensive molten pool state feature module 103 analyzes the molten pool three-dimensional temperature field and the molten pool metal vapor concentration to obtain the molten pool state feature of the welding molten pool;

[0255] The laser-arc energy distribution ratio module 104 dynamically divides the laser-arc energy distribution ratio based on the molten pool state feature;

[0256] The arc voltage waveform adjustment module 105 adjusts the offset position of the laser focal point in the weld depth direction based on the laser-arc energy distribution ratio, and adjusts the arc voltage waveform through the inverter power supply.

[0257] In several embodiments provided in the present application, it should be understood that the disclosed method and system can be implemented in other ways. For example, the above-described system embodiments are only illustrative, for example, the division of the modules is only a logical function division, and actual implementation can have another division way.

[0258] The modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. According to actual needs, part or all of the modules can be selected to achieve the purpose of the embodiment scheme.

[0259] In addition, each functional module in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The above integrated unit can be realized in the form of hardware or in the form of hardware plus software function modules.

[0260] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0261] The embodiments of the present application can acquire and process related data based on artificial intelligence technology. Among them, artificial intelligence is to use digital computers or digital computer controlled machines to simulate, extend and expand human intelligence, perceive environment, acquire knowledge and use knowledge to obtain the best results.

[0262] Finally, it should be noted that the above examples are merely intended to illustrate the technical solutions of the present application and not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A method for controlling weld penetration depth in laser-arc hybrid welding, characterized in that, The method includes: S1. Construct a three-dimensional temperature field of the weld pool based on the infrared thermal radiation signal of the weld pool; S2. The concentration of molten pool metal vapor is calibrated based on the intensity of specific element spectral lines in the plasma spectral signal of the weld pool; The calibration of the molten pool metal vapor concentration based on the intensity of specific elemental spectral lines in the plasma spectral signal of the weld pool includes: The plasma spectral signal is acquired by dual-path beam splitting to obtain synchronous light intensity signals of characteristic spectral lines and reference spectral lines; The synchronous light intensity signal is differentially amplified to obtain the relative intensity ratio of the elemental spectral lines of the molten pool; The relative intensity ratio of the element spectral lines is linearly calibrated by photoelectric conversion to obtain the molten pool metal vapor concentration gradient value; Based on the steam concentration gradient value, the boundary of the concentration distribution field is reconstructed to obtain the molten pool metal steam concentration field; S3. Couple the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics of the welding molten pool; S4. Dynamically divide the laser-arc energy distribution ratio based on the molten pool state characteristics; S5. Adjust the offset position of the laser focus in the weld depth direction based on the laser-arc energy distribution ratio, and adjust the arc voltage waveform through the inverter power supply.

2. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 1, characterized in that, The construction of the three-dimensional temperature field of the weld pool based on the infrared thermal radiation signal of the weld pool includes: Multispectral infrared thermal imaging was performed on the weld pool to obtain the multi-band thermal radiation distribution on the surface of the weld pool; Based on the multi-band thermal radiation distribution, the depth characteristic spectrum is separated by color difference to obtain the thermal radiation gradient in the direction of the molten pool depth. The thermal radiation gradient is mapped onto a spatial filter array to obtain the three-dimensional voxel temperature distribution of the molten pool; The isothermal boundary of the three-dimensional voxel temperature distribution is reconstructed to obtain the three-dimensional temperature field of the molten pool.

3. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 1, characterized in that, The process of coupling the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics includes: By synchronizing the dynamic distribution data of the three-dimensional temperature field of the molten pool with the real-time monitoring data of the metal vapor concentration of the molten pool, a coupled dataset of the molten pool is obtained; Based on the coupled dataset, the temperature gradient in the three-dimensional temperature field of the molten pool and the steam concentration gradient are vector-superimposed to obtain the interactive characteristics of energy transfer and mass migration inside the molten pool. Based on the interaction features, the convection intensity distribution and vaporization suppression region boundary in the molten pool fluid are identified to obtain the molten pool state characteristics of the welding molten pool.

4. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 3, characterized in that, The identification of convection intensity distribution and vaporization suppression region boundaries in the molten pool fluid based on the interactive features, to obtain the molten pool state characteristics, includes: Collect plasma optical pattern morphology and ripple direction on the molten pool surface; The boundary of the vaporization suppression region is determined based on the contraction width of the plasma ripple morphology; The convection intensity distribution in the molten pool fluid is identified based on the convergence direction of the surface ripples of the molten pool, and the identified convection intensity distribution is used as the molten pool state characteristic of the welding molten pool.

5. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 3, characterized in that, The dynamic division of laser-arc energy distribution ratio based on the molten pool state characteristics includes: Based on the molten pool state characteristics, orthogonal physical parameter separation is performed to obtain the molten pool's flowability parameters and stability parameters; By superimposing the fluidity parameter and the stability parameter, the dynamic coupling effect of laser energy and electric arc energy is obtained; The laser-arc energy distribution ratio is obtained by adaptively allocating the energy based on the real-time changing trend of the dynamic coupling relationship.

6. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 5, characterized in that, The orthogonal physical parameter separation based on the molten pool state characteristics to obtain the molten pool's flowability and stability parameters includes: By mapping the heat flow vector in the molten pool state characteristics to the heat transport dominant axis of the three-dimensional temperature field of the molten pool, the first orthogonal component characterizing energy transfer is obtained. The melt flow vector in the molten pool state characteristics is mapped to the mass migration dominant axis of the three-dimensional temperature field of the molten pool to obtain the second orthogonal component characterizing mass transport. The modulus of the first orthogonal component is used as the melt flow parameter, and the directional consistency of the second orthogonal component is used as the melt stability parameter.

7. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 1, characterized in that, The adjustment of the laser focus offset position in the weld depth direction based on the laser-arc energy distribution ratio includes: The oscillation state of the weld pool is monitored in real time to obtain the oscillation frequency of the weld pool; Dynamic response analysis is performed on the deviation of the oscillation frequency from the preset reference frequency to obtain the melt depth control signal of the molten pool; Based on the melt depth control signal, the servo motor is transmitted a displacement command to obtain the depth direction displacement of the molten pool; The laser focus position is adjusted based on the depth direction displacement to obtain the offset position in the weld depth direction.

8. The method for controlling weld penetration in laser-arc hybrid welding as described in claim 7, characterized in that, The method of adjusting the arc voltage waveform via an inverter power supply includes: Extract the frequency change from the oscillation frequency to obtain the oscillation frequency change of the molten pool; The modulation response characteristics are mapped to the oscillation frequency change to obtain the arc voltage waveform parameters of the molten pool; The inverter power supply switching sequence is reconstructed based on the arc voltage waveform parameters to obtain the modulated arc voltage waveform.

9. A weld penetration control system for laser-arc hybrid welding, characterized in that, The system includes: The molten pool three-dimensional temperature field construction module is used to construct the molten pool three-dimensional temperature field based on the infrared thermal radiation signal of the welding molten pool; A molten pool metal vapor concentration calibration module is used to calibrate the molten pool metal vapor concentration based on the intensity of specific element spectral lines in the plasma spectral signal in the welding molten pool. The calibration of the molten pool metal vapor concentration based on the intensity of specific elemental spectral lines in the plasma spectral signal of the weld pool includes: The plasma spectral signal is acquired by dual-path beam splitting to obtain synchronous light intensity signals of characteristic spectral lines and reference spectral lines; The synchronous light intensity signal is differentially amplified to obtain the relative intensity ratio of the elemental spectral lines of the molten pool; The relative intensity ratio of the element spectral lines is linearly calibrated by photoelectric conversion to obtain the molten pool metal vapor concentration gradient value; Based on the steam concentration gradient value, the boundary of the concentration distribution field is reconstructed to obtain the molten pool metal steam concentration field; The integrated molten pool state characteristic module is used to couple and analyze the three-dimensional temperature field of the molten pool with the metal vapor concentration of the molten pool to obtain the molten pool state characteristics of the welding molten pool. A laser-arc energy distribution ratio module is used to dynamically divide the laser-arc energy distribution ratio based on the molten pool state characteristics. The arc voltage waveform adjustment module is used to adjust the offset position of the laser focus in the weld depth direction based on the laser-arc energy distribution ratio, and to adjust the arc voltage waveform through the inverter power supply.

Images