Vacuum electron beam welding quality monitoring method and system based on molten pool monitoring

Abstract

The invention discloses a vacuum electron beam welding quality monitoring method and system based on molten pool monitoring, and relates to the technical field of vacuum electron beam welding quality monitoring.The method comprises the steps that a molten pool panoramic image, gas flow, electron beam parameters and molten pool dynamic parameters are collected in real time and stored in an electron beam welding time sequence database; monitoring the pollution degree of a camera window to obtain a window pollution rate wic, setting a window pollution threshold Twi, and when wic is less than Twi, directly calculating the stability of a first molten pool; when wic is greater than or equal to Twi, dynamically changing the gas flow, cleaning a camera window, then running a molten pool stability correction strategy, and further compensating the molten pool stability according to the window pollution rate and the vacuum degree to obtain corrected molten pool stability; surface abnormity is monitored through a molten pool image, a workpiece joint surface abnormity coefficient is obtained, and an electron beam welding quality score is obtained by combining molten pool stability. According to the method, the influence of molten pool stability calculation errors caused by the window cleaning process is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of vacuum electron beam welding quality monitoring, and specifically to a method and system for vacuum electron beam welding quality monitoring based on molten pool monitoring. Background Art

[0002] Currently, the welding quality monitoring method based on molten pool monitoring has become a research hotspot, but there are still many deficiencies and challenges in practical applications. First of all, due to the metal vapor, spatter and deposits generated under the action of high temperature and high-energy beam, a pollution film will be formed on the window of the monitoring camera, resulting in blurred images and decreased contrast; this not only affects the edge extraction and feature detection of the molten pool area, but also makes it difficult for traditional image processing algorithms to accurately reflect the real molten pool state. After cleaning the window, in order to prevent metal deposition, measures such as argon purging are generally adopted; however, these measures will affect the dynamic behavior of the molten pool, and the traditional molten pool stability calculation method cannot effectively compensate for the measurement errors brought by measures such as argon purging, resulting in unstable monitoring data and affecting subsequent quality determination.

[0003] Therefore, the present invention provides a method and system for vacuum electron beam welding quality monitoring based on molten pool monitoring. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for vacuum electron beam welding quality monitoring based on molten pool monitoring to solve the existing problems mentioned in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A method for vacuum electron beam welding quality monitoring based on molten pool monitoring, including the following steps:

[0006] S1. Real-time collect the panoramic image of the molten pool, gas flow rate, electron beam parameters and molten pool dynamic parameters, and store them in the electron beam welding time series database;

[0007] S2. Monitor the pollution degree of the monitoring camera window to obtain the window pollution rate wic, set the window pollution threshold Twi. When wic < Twi, directly calculate the first molten pool stability and send it to S4; when wic ≥ Twi, dynamically change the gas flow rate, clean the camera window and then run S3;

[0008] S3. Set the molten pool stability correction strategy, further compensate the molten pool stability according to the window pollution rate and vacuum degree, and obtain the corrected molten pool stability and send it to S4;

[0009] S4. Monitor the surface abnormality through the molten pool image to obtain the surface abnormality coefficient of the workpiece joint, and combine the molten pool stability to obtain the electron beam welding quality score.

[0010] The further improvement of the present invention lies in that the specific calculation process of the window pollution rate includes:

[0011] S211. In the initial cleaning state, obtain the global contrast C0 of the reference image through a calibration experiment;

[0012] S212. Extract the current contrast mean value of each frame of image within the sampling period T through the global histogram mean value;

[0013] S213. Subsequently, calculate the window contrast decline rate within the sampling period T, denoted as the window pollution rate wic(t) = 1 - (aCt / C0), where aCt represents the current contrast mean value of each frame of image.

[0014] A further improvement of the present invention lies in that the specific calculation process of the first molten pool stability includes:

[0015] S221. Use threshold segmentation to obtain the ROI of the molten pool area in the panoramic image captured by the camera. Find the continuous boundary of the foreground contour features in the molten pool area, traverse the continuous boundary of the molten pool area, track and check the continuous boundary of the largest panoramic area, and count the number of pixels in the ROI of the molten pool area, denoted as the molten pool area A;

[0016] S222. Extract the time series data A(t) of the molten pool area, and perform Fourier transform on A(t) to obtain the time series spectrum of the molten pool area, and identify the frequency component with the largest amplitude in the spectrum as the molten pool oscillation frequency f pool ;

[0017] S223. Extract the molten pool temperature data, calculate the standard deviation of the temperature data within the time sampling period T, and obtain the molten pool temperature fluctuation index σ tem ;

[0018] S224. Calculate the first molten pool stability where f nom represents the molten pool oscillation frequency reference value, σ nom represents the molten pool temperature fluctuation index reference value, β = 0.05 represents the attenuation factor, and δ1 and δ2 represent the weight coefficients.

[0019] A further improvement of the present invention lies in that the dynamic change of the gas flow rate is achieved by setting a flow gain coefficient, and calculating the product of the flow gain coefficient and the window pollution rate and the maximum purge volume to obtain the current flow increment. When it is detected again that wic < Twi, the gas flow rate is gradually reduced to the initial gas flow rate according to the set step size.

[0020] A further improvement of the present invention lies in that the molten pool stability correction strategy includes a vacuum degree change calculation sub-strategy and an electron beam current correction sub-strategy;

[0021] The sub-strategy for calculating the change in vacuum degree extracts the initial vacuum degree, establishes a gas-pump dynamic equilibrium model, and quantitatively calculates the change in vacuum degree. Specifically, it includes extracting the current gas flow gaf cur and the initial gas flow gaf ini , and calculates the corresponding current steady-state pressure P cur and the initial steady-state pressure P ini through the ratio of the above to the pumping speed of the vacuum pump. Furthermore, the pressure change ΔP = P cur - P ini is obtained, and the pressure change is denoted as the vacuum degree change amv;

[0022] The sub-strategy for correcting the electron beam current extracts the change in vacuum degree, establishes a compensation relationship between the vacuum degree and the electron beam focusing current, and obtains the compensated dynamic correction coefficient If of the electron beam current cur , then If cur = If0·(1 + α(amv / P ini ))), where If0 represents the calibrated current, α represents the compensation coefficient, and is obtained by fitting after adjusting If cur to keep the molten pool depth constant during experimental calibration.

[0023] A further improvement of the present invention is that the molten pool stability correction strategy further includes calculating the effective energy input and molten pool forming ability during the welding process through the molten pool oscillation frequency f pool and the dynamic correction coefficient If of the electron beam current cur , and quantifying the negative impact on the molten pool quality through the change in vacuum degree and the molten pool temperature fluctuation index, and constructing the corrected molten pool stability CS pool , and the calculation formula is: where k1, k2, and k3 represent weight coefficients.

[0024] A further improvement of the present invention is that the electron beam welding quality score is obtained by calculating the weighted sum of the first molten pool stability obtained in step S2 or the corrected molten pool stability obtained in step S3 and the workpiece joint surface abnormality coefficient.

[0025] A further improvement of the present invention is that the specific calculation process of the workpiece joint surface abnormality coefficient includes: extracting the image edge in the molten pool area through the Canny operator, performing connected component analysis on the binary edge image, and screening independent contours with an area > 50 pixels as abnormal areas; extracting texture features, including calculating the contrast, energy, and homogeneity in 4 directions, and calculating the 8-neighborhood rotation-invariant LBP histogram to capture structural abnormalities Extract the texture features of all pixels to construct a feature vector, and perform unsupervised classification on all abnormal regions through the K-means++ algorithm, where K = 3, representing normal, splash, and crack, and assign weights to the defect types. Among them, the weight of normal defect w1 = 0.3, the weight of splash defect w2 = 0.5, and the weight of crack defect w3 = 0.9; at the same time, extract the area A1 of each abnormal region i , i represents the i-th abnormal region, calculate the proportion of all abnormal regions in the area of the molten pool region to obtain the surface abnormality coefficient, and perform weighted summation of the surface abnormality coefficients of all defect regions according to the defect type to obtain the surface abnormality coefficient of the workpiece joint

[0026] On the other hand, the present invention provides a quality monitoring system for vacuum electron beam welding based on molten pool monitoring, including:

[0027] A database establishment module, which collects panoramic images of the molten pool, gas flow rate, electron beam parameters, and dynamic parameters of the molten pool in real time and stores them in the electron beam welding time series database;

[0028] A camera window pollution degree monitoring module, which obtains the window pollution rate and sets a window pollution threshold. When the window pollution rate is less than the window pollution threshold, directly calculate the first molten pool stability and send it to the electron beam welding quality scoring module; when the window pollution rate is greater than or less than the window pollution threshold, dynamically change the gas flow rate, clean the camera window, and then run the molten pool stability correction module;

[0029] A molten pool stability correction module, which further compensates the molten pool stability according to the window pollution rate and the vacuum degree to obtain the corrected molten pool stability;

[0030] An electron beam welding quality scoring module, which monitors surface abnormalities through the molten pool image to obtain the surface abnormality coefficient of the workpiece joint, and combines the molten pool stability to obtain the electron beam welding quality score.

[0031] The further improvement of the present invention lies in that the molten pool stability correction module includes a vacuum degree change calculation unit and an electron beam current correction unit; the vacuum degree change calculation unit is used to establish a gas-pump dynamic balance model to quantitatively calculate the change in vacuum degree; the electron beam current correction unit is used to establish a compensation relationship between the vacuum degree and the electron beam focusing current to obtain a compensated electron beam current dynamic correction coefficient.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] 1. First, the present invention dynamically clears the deposits on the window by real-time monitoring the window pollution degree and adaptively adjusting the gas flow rate, keeps the image clear, ensures that the features of the molten pool region can be accurately extracted, and thus greatly improves the robustness and overall accuracy of the welding quality monitoring system;

[0034] 2. Secondly, by monitoring the window pollution degree in real time, dynamically adjusting the gas flow rate, and establishing the compensation relationship between the vacuum degree and the electron beam current, it is possible to effectively correct the fluctuations in the oscillation frequency, area, and temperature of the molten pool caused by the vacuum degree fluctuation after cleaning the window, thereby making the calculation of the molten pool stability more accurate and ensuring the stability of the welding process and the consistency of the weld quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a flowchart of a quality monitoring method for vacuum electron beam welding based on molten pool monitoring according to the present invention;

[0036] Figure 2 is a framework diagram of a quality monitoring system for vacuum electron beam welding based on molten pool monitoring according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] The technical solution of the present invention will be described in detail below with reference to the drawings and specific embodiments. It should be understood that the specific features in the embodiments of the present invention are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. Without conflict, the technical features in the embodiments of the present invention and the embodiments can be combined with each other.

[0038] The term "and / or" is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " generally represents an "or" relationship between the associated objects before and after.

[0039] Embodiment 1

[0040] Figure 1 shows a flowchart of a quality monitoring method for vacuum electron beam welding based on molten pool monitoring disclosed in this embodiment. The steps are as follows:

[0041] S1. Real-time collect the panoramic image of the molten pool, gas flow rate, electron beam parameters, and dynamic parameters of the molten pool, and store them in the time-series database of electron beam welding;

[0042] S2. Monitor the pollution degree of the camera window to obtain the window pollution rate wic, and set the window pollution threshold Twi. Since the instability of the molten pool easily leads to chaotic flow of the molten metal, it is difficult for the gas to escape smoothly, resulting in defects such as pores, cold laps, or root pores in the weld. These defects will weaken the tensile strength and fatigue performance of the welded joint. Therefore, when wic < Twi, directly calculate the first molten pool stability; when wic ≥ Twi, in a vacuum environment, metal vapor or spatter may deposit on the optical elements of the monitoring system, such as the lens or protective glass, resulting in a decrease in image quality and thus triggering false alarms. Therefore, at this time, dynamically change the gas flow rate to clean the camera window. The specific steps include:

[0043] S21: Calculate the window contamination rate. The specific calculation process includes:

[0044] S211. In the initial clean state, obtain the global contrast C0 of the reference image through a calibration experiment;

[0045] S212. Extract the current contrast mean of each frame of image within the sampling period T through the global histogram mean;

[0046] S213. Subsequently, calculate the window contrast decline rate within the sampling period T, denoted as the window contamination rate wic(t) = 1 - (aCt / C0), where aCt represents the current contrast mean of each frame of image.

[0047] S22: Calculate the stability of the first molten pool. The specific calculation process includes:

[0048] S221. Use threshold segmentation to obtain the ROI of the molten pool area in the panoramic image captured by the camera. Find the continuous boundary of the foreground contour features in the molten pool area, traverse the continuous boundary of the molten pool area, track and check the continuous boundary of the largest panoramic area, and count the number of pixels within the ROI of the molten pool area, denoted as the molten pool area A;

[0049] S222. Extract the time-series data A(t) of the molten pool area, and perform Fourier transform on A(t) to obtain the time-series spectrum of the molten pool area, and identify the frequency component with the largest amplitude in the spectrum as the molten pool oscillation frequency f pool , The molten pool oscillation frequency reflects the flow stability of the liquid metal. Abnormal fluctuations in frequency (such as a sudden drop) may indicate pores or lack of fusion defects;

[0050] S223. Extract the molten pool temperature data, calculate the standard deviation of the temperature data within the time sampling period T to obtain the molten pool temperature fluctuation index σ tem ;

[0051] S224. Calculate the stability of the first molten pool where f nom represents the molten pool oscillation frequency reference value, σ nom represents the molten pool temperature fluctuation index reference value, β = 0.05 represents the attenuation factor, calibrated through a hot crack sensitivity experiment, and δ1 and δ2 represent the weight coefficients.

[0052] S23. The dynamic gas flow rate is obtained by setting a flow gain coefficient and calculating the product of the flow gain coefficient, the window contamination rate, and the maximum purge volume to obtain the current flow increment. When wic < Twi is detected again, the gas flow rate is gradually reduced to the initial gas flow rate in accordance with the set step size.

[0053] Example 2

[0054] Based on the inventive concept of Embodiment 1, this embodiment proposes a molten pool stability correction strategy and evaluates the quality of electron beam welding; specifically including:

[0055] When wic ≥ Twi, dynamically change the gas flow rate. After cleaning the camera window, run the molten pool stability correction strategy, including:

[0056] S3. Since the gas flow rate was changed due to severe window contamination in S2, a new problem will occur at this time. No matter what gas enters the molten pool, the vacuum degree will change to a certain extent, and the vacuum degree will affect the stability of the molten pool. Therefore, further compensate the stability of the molten pool according to the window contamination rate and the vacuum degree to obtain the corrected molten pool stability; the molten pool stability correction strategy includes a vacuum degree change calculation sub-strategy and an electron beam current correction sub-strategy;

[0057] In vacuum electron beam welding, pressure and vacuum degree are essentially two expressions of the same physical quantity. Therefore, calculate the vacuum degree by calculating the influence of air flow velocity on pressure;

[0058] The vacuum degree change calculation sub-strategy extracts the initial vacuum degree, establishes a gas-pump dynamic equilibrium model, and quantitatively calculates the vacuum degree change. Specifically, it includes extracting the current gas flow rate gaf cur and the initial gas flow rate gaf ini , and calculates the corresponding current steady-state pressure P cur and the initial steady-state pressure P ini through their ratio to the pumping speed of the vacuum pump, and then obtains the pressure change amount ΔP = P cur -P ini , and records the pressure change amount as the vacuum degree change amount amv;

[0059] The electron beam current correction sub-strategy extracts the vacuum degree change amount, and establishes a compensation relationship between the vacuum degree and the electron beam focusing current to obtain the compensated electron beam current dynamic correction coefficient If cur , then If cur = If0·(1 + α(amv / P ini ), where, If0 represents the calibrated current, α represents the compensation coefficient, and is obtained by fitting after adjusting If cur to keep the molten pool depth constant during experimental calibration.

[0060] Introducing α(amv / P ini ), the compensation ratio of the vacuum degree fluctuation to the focusing current can be quantified, and the influence of the vacuum degree change on the molten pool stability can be dynamically compensated. If α = 0, the vacuum degree fluctuation will cause abnormal molten pool morphology (such as incomplete penetration or excessive burn-through), triggering an error alarm;

[0061] The molten pool stability correction strategy also includes through the molten pool oscillation frequency f poolAnd the dynamic correction coefficient If of the electron beam flow cur Calculate the effective energy input and the molten pool forming ability during the welding process, and quantify the negative impact on the molten pool quality through the change in vacuum degree and the molten pool temperature fluctuation index, and construct the corrected molten pool stability CS pool , and the calculation formula is: Where k1, k2, and k3 represent weight coefficients; the dynamic correction coefficient of the electron beam flow directly determines the energy density of the electron beam. When the dynamic correction coefficient of the electron beam flow increases, the penetration depth will also increase accordingly. The molten pool area can reflect the uniformity of the heat input distribution. If cur ·f pool reflects the synergistic effect between the electron beam current and the molten pool area; and the greater the vacuum degree, the stronger the electron scattering will be, the energy density will decrease, and the effects of vacuum fluctuation and temperature fluctuation are independent of each other, so they need to be superimposed and suppressed.

[0062] S4. Monitor the surface anomalies through the molten pool image to obtain the surface anomaly coefficient of the workpiece joint, and combine the molten pool stability to obtain the electron beam welding quality score.

[0063] The specific calculation process of the surface anomaly coefficient of the workpiece joint includes: extracting the image edge in the molten pool area through the Canny operator, performing connected domain analysis on the binary edge image, and screening the independent contour with an area > 50 pixels as the abnormal area; extracting texture features, including calculating the contrast, energy, and homogeneity in 4 directions (0°, 45, 90°, 135°), and calculating the 8-neighborhood rotation-invariant LBP histogram to capture structural anomalies Extract the texture features of all pixel points to construct a feature vector, and perform unsupervised classification on all abnormal areas through the K-means++ algorithm, where K = 3, representing normal, spatter, and crack, and assign weights to the defect types. Normal: w1 = 0.3 (less impact on airtightness), spatter: w2 = 0.5 (risk of stress concentration), crack: w3 = 0.9 (direct failure risk); at the same time, extract the area A1 of each abnormal area i , i represents the i-th abnormal area, calculate the proportion of all abnormal areas in the molten pool area to obtain the surface anomaly coefficient, and perform weighted summation of the surface anomaly coefficients of all defect areas according to the defect type to obtain the surface anomaly coefficient of the workpiece joint.

[0064] The electron beam welding quality score is obtained by extracting the first molten pool stability obtained in step S2 or the corrected molten pool stability obtained in step S3, and calculating the weighted summation of it and the surface anomaly coefficient of the workpiece joint.

[0065] The setting of the threshold and weight can be based on the default settings of the present invention, or can be set by the operator himself.

[0066] Example 3

[0067] Figure 2 The figure shows a framework diagram of a quality monitoring system for vacuum electron beam welding based on molten pool monitoring according to the present invention. Based on the same inventive concept as in Embodiment 1 and Embodiment 2, the present invention provides a quality monitoring system for vacuum electron beam welding based on molten pool monitoring, including:

[0068] A database establishment module that collects panoramic images of the molten pool, gas flow rate, electron beam parameters, and dynamic parameters of the molten pool in real time and stores them in the electron beam welding time series database;

[0069] A camera window pollution degree monitoring module that obtains the window pollution rate, sets a window pollution threshold. When the window pollution rate is less than the window pollution threshold, directly calculate the first molten pool stability and send it to the electron beam welding quality scoring module; when the window pollution rate is greater than or less than the window pollution threshold, dynamically change the gas flow rate, clean the camera window, and then run the molten pool stability correction module;

[0070] A molten pool stability correction module that further compensates the molten pool stability according to the window pollution rate and the vacuum degree to obtain the corrected molten pool stability; the molten pool stability correction module includes a vacuum degree change calculation unit and an electron beam current correction unit; the vacuum degree change calculation unit is used to establish a gas-pump dynamic balance model to quantitatively calculate the change in vacuum degree; the electron beam current correction unit is used to establish a compensation relationship between the vacuum degree and the electron beam focusing current to obtain a compensated dynamic correction coefficient of the electron beam current.

[0071] An electron beam welding quality scoring module that monitors surface anomalies through the molten pool image to obtain the surface anomaly coefficient of the workpiece joint seam, and combines the molten pool stability to obtain the electron beam welding quality score.

[0072] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0073] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing the processFigure 1 means for the functions specified in one or more processes and / or blocks Figure 1 or blocks.

[0074] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one Figure 1 or more processes and / or blocks Figure 1 or blocks.

[0075] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one Figure 1 or more processes and / or blocks Figure 1 or blocks.

[0076] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope protected by the claims of the present invention. These all fall within the protection scope of the present invention.

Claims

1. A vacuum electron beam welding quality monitoring method based on molten pool monitoring, characterized in that: The following steps are involved: S1, real-time acquisition of molten pool panoramic images, gas flow, electron beam parameters and molten pool dynamic parameters, and storage in the electron beam welding timing database; S2, monitor the camera window contamination degree, obtain the window contamination rate wic, set the window contamination threshold Twi, when wic<Twi, directly calculate the stability of the first molten pool and send it to S4; when wic≥Twi, dynamically change the gas flow rate, clean the camera window and run S3; S3, setting a molten pool stability correction strategy, further compensating the molten pool stability according to the window contamination rate and vacuum degree, and obtaining the corrected molten pool stability and sending it to S4; S4. Monitor surface anomalies through molten pool images to obtain the surface anomaly coefficient of the workpiece joint, and combine the molten pool stability to obtain the electron beam welding quality score.

2. A vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 1, characterized in that: The specific calculation process of the window contamination rate includes: S211, in an initial clean state, obtaining a global contrast C0 of a reference image through a calibration experiment; S212, extracting the current contrast mean of each frame image within the sampling period T through the global histogram mean; S213. Then, the window contrast decrease rate within the sampling period T is calculated and recorded as the window contamination rate wic(t)=1-(aCt / C0), where aCt represents the current contrast mean value of each frame of the image.

3. The vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 1 is characterized in that: The specific calculation process of the first molten pool stability includes: S221, using threshold segmentation to obtain a melt pool region ROI in the panoramic image taken by the camera, searching for a continuous boundary of a foreground contour feature in the melt pool region, traversing the continuous boundary of the melt pool region, tracking and checking the largest panoramic region continuous boundary, and counting the number of pixels in the melt pool region ROI as the melt pool area A; S222, extract the melt pool area time series data A(t), and perform Fourier transform on A(t) to obtain the melt pool area time series spectrum, and identify the frequency component with the largest amplitude in the spectrum as the melt pool oscillation frequency f pool ; S223, extract the molten pool temperature data, calculate the standard deviation of the temperature data within the time sampling period T, and obtain the molten pool temperature fluctuation index σ tem ; S224, calculate the stability of the first molten pool Among them, f nom Indicates the reference value of the molten pool oscillation frequency, σ nom represents the reference value of the molten pool temperature fluctuation index, β=0.05 represents the attenuation factor, and δ1 and δ2 represent the weight coefficients.

4. The vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 1 is characterized in that: The dynamic change of gas flow is achieved by setting a flow gain coefficient and calculating the product of the flow gain coefficient, the window contamination rate and the maximum purge volume to obtain the current flow increment. When wic<Twi is detected again, the gas flow is gradually reduced to the initial gas flow according to the set step size.

5. The vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 1, characterized in that: The molten pool stability correction strategy includes a vacuum degree change calculation sub-strategy and an electron beam current correction sub-strategy; The vacuum degree change calculation sub-strategy extracts the initial vacuum degree, establishes a gas-pump dynamic balance model, and quantitatively calculates the vacuum degree change, specifically including extracting the current gas flow gaf cur and initial gas flow gaf ini The current steady-state pressure P is calculated by the ratio of the vacuum pump speed to the vacuum pump speed. cur and the initial steady-state pressure P ini , and then the pressure change ΔP=P cur -P ini , the pressure change is recorded as the vacuum degree change amv; The electron beam current correction strategy extracts the vacuum degree change and establishes a compensation relationship between the vacuum degree and the electron beam focusing current to obtain the compensated electron beam current dynamic correction coefficient If cur , then cur =If0·(1+α(amv / P ini )), where If0 represents the calibration current, α represents the compensation coefficient, and If is adjusted in the experimental calibration cur The result is obtained by fitting while keeping the molten pool depth constant.

6. A vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 5, characterized in that: The molten pool stabilization correction strategy also includes adjusting the molten pool oscillation frequency f pool and electron beam current dynamic correction factor If cur Calculate the effective energy input and molten pool forming capacity of the welding process, quantify the negative impact of molten pool quality through vacuum change and molten pool temperature fluctuation indicators, and construct a corrected molten pool stability CS pool , the calculation formula is: Where k1, k2 and k3 represent weight coefficients.

7. The vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 1, characterized in that: The electron beam welding quality score is obtained by extracting the first molten pool stability obtained in step S2 or the modified molten pool stability obtained in step S3, and calculating the weighted sum of the first molten pool stability and the workpiece joint surface abnormality coefficient to obtain the electron beam welding quality score.

8. The vacuum electron beam welding quality monitoring method based on molten pool monitoring according to claim 7, characterized in that: The specific calculation process of the workpiece joint surface abnormality coefficient includes: extracting the image edge in the molten pool area through the Canny operator, performing connected domain analysis on the binary edge image, and screening independent contours with an area of ​​>50 pixels as abnormal areas; extracting texture features, including calculating the contrast, energy, and homogeneity in four directions, and calculating the 8-neighborhood rotation-invariant LBP histogram to capture structural abnormalities The texture features of all pixels are extracted to construct feature vectors. All abnormal areas are unsupervisedly classified using the K-means++ algorithm, where K = 3, representing normal, spatter, and crack. Weights are assigned to the defect types, where the normal defect weight w1 = 0.3, the spatter defect weight w2 = 0.5, and the crack defect weight w3 = 0.

9. At the same time, the area A1 of each abnormal area is extracted. i , i represents the i-th abnormal area. The proportion of all abnormal areas in the molten pool area is calculated to obtain the surface abnormality coefficient. The surface abnormality coefficients of all defective areas are weighted and summed by defect type to obtain the surface abnormality coefficient of the workpiece joint.

9. A vacuum electron beam welding quality monitoring system based on molten pool monitoring, used to execute a vacuum electron beam welding quality monitoring method based on molten pool monitoring as claimed in any one of claims 1 to 8, characterized in that: include: Database establishment module, real-time acquisition of molten pool panoramic images, gas flow, electron beam parameters and molten pool dynamic parameters, and storage in the electron beam welding timing database; The camera window contamination degree monitoring module obtains the window contamination rate and sets the window contamination threshold. When the window contamination rate is less than the window contamination threshold, the stability of the first molten pool is directly calculated and sent to the electron beam welding quality scoring module. When the window contamination rate is greater than or less than the window contamination threshold, the gas flow rate is dynamically changed, and the molten pool stability correction module is run after cleaning the camera window. The molten pool stability correction module further compensates the molten pool stability according to the window contamination rate and vacuum degree to obtain the corrected molten pool stability; The electron beam welding quality scoring module monitors surface anomalies through molten pool images, obtains the surface anomaly coefficient of the workpiece joint, and obtains the electron beam welding quality score based on the molten pool stability.

10. A vacuum electron beam welding quality monitoring system based on molten pool monitoring according to claim 9, characterized in that: The molten pool stability correction module includes a vacuum degree change calculation unit and an electron beam current correction unit; the vacuum degree change calculation unit is used to establish a gas-pump dynamic balance model and quantitatively calculate the vacuum degree change; the electron beam current correction unit is used to establish a compensation relationship between the vacuum degree and the electron beam focusing current to obtain the compensated electron beam current dynamic correction coefficient.

Images