Online detection method for all-position welding penetration based on the collaboration of electrical signals and weld pool images

Abstract

The present invention discloses an online detection method for all-position welding penetration that collaborates with electrical signals and molten pool images. The method employs dual robots to collaboratively weld pipelines in all positions. Initial segmentation is performed based on the welding position and further segmented at 5-10° angles. A laser scanner is used to scan the pipeline groove to obtain the groove dimensions of each section. Adaptive setting of welding process parameters is achieved based on a process knowledge base and a welding parameter adjustment model. The molten pool image and the time-domain eigenvalues ​​of the electrical signal are fused and discriminated using a neural network prediction model. Molten pool size characteristics are acquired based on the molten pool image, and a reference molten pool oscillation frequency is calculated based on a reference molten pool oscillation frequency model. Spectral analysis of the upper envelope signal of the voltage signal is performed to calculate the actual molten pool oscillation frequency. Penetration is then determined by comparing the actual and reference molten pool oscillation frequencies. This method implements all-position robot welding defect monitoring.

Description

Technical Field

[0001] The present invention relates to the field of robot automatic welding, in particular to an all-position welding penetration online detection method coordinated with an electrical signal and a molten pool image. Background Art

[0002] Currently, all-position welding of petrochemical process pipelines is primarily performed manually or with specialized automated welding machines. However, this method requires experienced technicians to adjust the welding process in real time based on the pipeline groove and position. This leads to high workload, high labor costs, low welding efficiency, and inconsistent welding quality. However, welding robots, with their long operating times, stable welding quality, and high efficiency, have become a major trend in replacing manual labor.

[0003] However, the application of welding machine technology currently faces many challenges. Currently, pipeline welding is mainly carried out through manual teaching. However, this method has high requirements for the workpiece itself and processing, and it does not respond quickly to problems that arise during the welding process and make timely adjustments. Therefore, in the actual on-site construction of petrochemical process pipeline robot full-position automatic welding, due to factors such as the external environment and assembly, it is easy to cause forming defects during the welding process. To solve the above problems, it is necessary to monitor and feedback the full-position automatic welding process of petrochemical process pipeline robots in real time, and take corresponding control measures to ultimately improve welding efficiency and reduce the waste of time and materials.

[0004] Nowadays, people generally use molten pool vision and computers to conduct intelligent real-time monitoring of the welding process. Generally, a convolutional neural network is used to directly establish a mapping model between welding defects and molten pool images, thereby realizing the monitoring of pipeline welding defects. However, this method is suitable for welding with unchanged welding position type and groove size. During the full-position welding process of pipelines with low processing accuracy, the welding position type, groove morphology and molten pool force are constantly changing, causing the same welding defect to exhibit different characteristics in different positions. If the molten pool image is simply classified according to the welding defect and a pipeline full-position welding defect prediction model is established through a convolutional neural network, it is easy to lead to a low prediction accuracy of the prediction model, making it impossible to effectively realize the monitoring of pipeline full-position welding defects. Summary of the Invention

[0005] The purpose of the present invention is to provide an online detection method for all-position welding penetration by coordinating electrical signals with molten pool images, which is used for real-time monitoring of the automatic all-position welding process of petrochemical pipeline robots.

[0006] The technical solution to achieve the purpose of the present invention is: a method for monitoring pipeline welding defects in all positions based on electrical signals and vision under the collaboration of two robots, comprising the following steps:

[0007] Step 1: Before welding, input the pipe material, wall thickness and diameter information, and start the welding robot equipped with a laser scanner to scan the pipe groove from bottom to top to obtain the original data of the groove profile at different positions.

[0008] Step 2: Based on the characteristics of overhead welding, flat welding and vertical welding, the pipeline groove is preliminarily segmented, that is, 0-30° is flat welding, 30°-150° is vertical welding and 150°-180° is overhead welding. Then, based on the preliminary segmentation, 5-10° segmentation is further performed. At the same time, the corresponding welding process parameters are matched from the process knowledge base according to the position and groove size of each segment.

[0009] Furthermore, corresponding welding process parameters are matched from a process knowledge base according to the position and groove size of each section. The process knowledge base includes material, wall thickness, pipe diameter, position type, groove gap, groove misalignment, welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time. The welding types include flat welding at 0°, vertical welding at 90°, and overhead welding at 180°.

[0010] The process of obtaining welding process parameters for each section is as follows:

[0011] First, the corresponding welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time are obtained based on the input material, wall thickness, pipe diameter, and welding position type, groove gap, and groove misalignment obtained by segmentation and laser scanner.

[0012] Secondly, calculate the position of each section of the slope. The specific formula is as follows:

[0013]

[0014] is the angle between the starting position of each segment and the vertical direction, is the angle between the end position of each segment and the vertical direction;

[0015] Finally, according to the position of each groove, some process parameters are adjusted;

[0016] When the section belongs to the flat welding position type, adjust the welding current and wire feeding speed. The specific formula is as follows:

[0017]

[0018] in, and To match the welding current and wire feeding speed in the process knowledge base, and is a constant;

[0019] When the section belongs to the vertical welding position type, adjust the welding current and welding speed. The specific formula is as follows:

[0020]

[0021]

[0022] in, To match the welding speed in the process knowledge base, 、 、 and is a constant;

[0023] When the section belongs to the overhead welding position type, adjust the wire feeding speed and welding speed. The specific formula is as follows:

[0024]

[0025] in, and is a constant.

[0026] Step 3: Synchronously start both welding robots, along with a welding parameter acquisition box (150-200 Hz) and an industrial camera (30-50 Hz) to collect electrical signals and weld pool visuals. Based on the symmetry of welding deformation and the complementary nature of arc start and arc end points, one robot welds clockwise from 180° on the left side to 3°-6° on the right side, while the other robot welds clockwise from 0° on the right side to 174°-177° on the right side.

[0027] Step 4: The voltage signal collected in each segment is framed using a rectangular window of 500-700 mm in length. Time domain analysis is performed on the voltage signal in each frame. The time domain analysis data, material, wall thickness, pipe diameter, groove position and size of the segment, welding process parameters, and the corresponding molten pool image in the rectangular window are input into the electrical signal-image collaborative perception burn-through defect prediction model. If the output result is burn-through, welding is stopped. If the output result is unknown, step 5 is executed.

[0028] Furthermore, the electrical signal-image collaborative perception burn-through defect prediction model includes:

[0029] The material, wall thickness, pipe diameter, groove position of the section, groove size, and welding process parameters are normalized. The material values ​​0.3, 0.6, and 0.9 represent carbon steel, stainless steel, and nickel-based alloy, respectively. Other parameters are normalized using the following calculation formula:

[0030]

[0031] in, and They are the maximum and minimum values ​​of the same type of data respectively;

[0032] The time domain analysis of voltage signals includes mean, variance, peak-to-peak value, and kurtosis. The calculation formulas are as follows:

[0033]

[0034]

[0035]

[0036]

[0037] in, is the length of the rectangular window, is the mean, is the variance, is the peak-to-peak value and is a steep value.

[0038] The electrical signal-image collaborative perception burn-through defect prediction model adopts a two-branch prediction model. One branch uses the Resnet convolutional neural network as the structural backbone and is used to extract the molten pool image features. The other branch uses the AlexNet convolutional neural network as the structural backbone and is used to extract data features consisting of variance, peak-to-peak value and kurtosis, material, wall thickness, pipe diameter, groove position, groove size, and welding process parameters. Finally, feature fusion is performed through the fully connected layer.

[0039] Step 5: Perform frequency domain analysis on the collected voltage signal and obtain the envelope spectrum data of the voltage signal, that is, the actual molten pool oscillation frequency. At the same time, obtain the molten pool size based on the molten pool image and calculate the reference oscillation frequency of the molten pool in the melt-through state. When the actual molten pool oscillation frequency is greater than the reference oscillation frequency of the molten pool in the melt-through state, take measures to immediately stop welding; when the actual molten pool oscillation frequency is less than or equal to the reference oscillation frequency of the molten pool in the melt-through state, continue welding.

[0040] Furthermore, obtaining the upper envelope spectrum data of the voltage signal includes:

[0041] First, the voltage signal in the time domain is converted into the spectrum of the frequency domain signal using Fourier transform;

[0042] Secondly, calculate the start and end frequencies of the intercepted frequency band according to the following formula:

[0043]

[0044] in, and They are the start and end frequencies of the intercepted frequency band when the pipe diameter is 150mm, the wall thickness is 7mm, the groove gap is 1.5mm, and the groove misalignment is 1mm; 、 、 and They are groove gap, groove misalignment, wall thickness and pipe diameter; 、 、 、 、 and is a constant.

[0045] Next, obtain the peak coordinates of the signal in the intercepted frequency band, sort the peak coordinates according to the size of the spectrum ordinate value, and save the two frequencies with the largest ordinate values. and ;

[0046] Then, and As the cutoff frequency of the band-pass filter, the collected voltage signal is filtered, and the upper envelope signal of the filtered voltage signal is obtained;

[0047] Finally, the upper envelope signal of the voltage signal is subjected to spectrum analysis to obtain the actual oscillation frequency of the molten pool.

[0048] Furthermore, obtaining the molten pool size based on the molten pool image and calculating the molten pool reference oscillation frequency in the melt-through state include:

[0049] First, the molten pool height and the length of the upper surface of the molten pool are obtained based on the molten pool image;

[0050] Secondly, based on the equivalent cylindrical film of molten metal in the melt-through state and the equal volume principle, the equivalent cylindrical diameter is obtained, and the following equation is obtained:

[0051]

[0052] in, is the height of the molten pool, is the length of the upper surface of the molten pool, is the equivalent melt pool radius.

[0053] At the same time, based on the fact that the molten pool is a convex frustum in the actual welding process, the above inequality is estimated to have a maximum value. Therefore, the above equation is simplified to:

[0054]

[0055] Thus, the equivalent melt pool diameter is obtained:

[0056]

[0057] Then, based on the oscillation model of the molten pool as a crescent-shaped molten pool, the wave number k is obtained, and the calculation formula is as follows:

[0058]

[0059] Finally, the oscillation frequency of the molten pool under the reference penetration state is calculated according to the following formula:

[0060]

[0061] in, is the density of molten metal, and is a constant.

[0062] Compared with the existing technology, the present invention has the following significant advantages: 1) The present invention selects welding process parameters based on the welding process adjustment model and process knowledge base, so that the welding process parameters are more adapted to the groove size and position, which is beneficial to improving the welding quality; 2) The present invention performs different treatments based on the different characteristics of burn-through defects and incomplete penetration defects in the molten pool visual and electrical signal manifestations, which is beneficial to eliminating interference factors and improving the accuracy of the prediction model; 3) The present invention takes into account the influence of the material and size of the pipeline itself, the welding position and the groove size on the welding defects, integrates these factors and fuses them with the visual and electrical signal characteristics, which is beneficial to enhancing the generalization ability and accuracy of the defect prediction model, thereby realizing automatic full-position welding defect monitoring of the pipeline robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a flow chart of the online detection method for all-position welding penetration by coordinating electrical signals with molten pool images of the present invention.

[0064] Figure 2 This is a schematic diagram of the pipeline position division in the all-position welding penetration online detection method based on the coordination of electrical signals and molten pool images of the present invention.

[0065] Figure 3 It is the electric signal and molten pool image without defects at the 40°-50° position in the online detection method for all-position welding penetration by coordinating the electric signal and the molten pool image of the present invention.

[0066] Figure 4 It is the electrical signal and molten pool image of the burn-through at the 40°-50° position in the online detection method for all-position welding penetration by coordinating the electrical signal and the molten pool image of the present invention.

[0067] Figure 5 It is the electric signal and molten pool image without defects at the 50°-60° position in the online detection method for all-position welding penetration by coordinating the electric signal and the molten pool image of the present invention.

[0068] Figure 6 It is the electrical signal and molten pool image of the burn-through at the 50°-60° position in the online detection method for all-position welding penetration by coordinating the electrical signal and the molten pool image of the present invention. DETAILED DESCRIPTION

[0069] The online detection method for all-position welding penetration by coordinating electrical signals with molten pool images of the present invention generally establishes a mapping model between welding formation defects and molten pool vision or electrical signals directly, thereby realizing the monitoring of pipeline welding formation defects. The actual quality of pipeline all-position welding formation is related to the groove size and welding position. Therefore, it is not possible to simply establish a mapping model between welding formation defects and molten pool images directly. Therefore, it is necessary to first establish a process knowledge base and a process parameter adjustment model according to the groove position and size to match and adjust the welding parameters; secondly, based on the fact that different welding defects exhibit different characteristics in different aspects of different information, different information source features are fused, and different welding defect prediction models are established through convolutional neural networks, thereby realizing accurate judgment and monitoring of pipeline all-position welding formation defects.

[0070] By monitoring the automated, all-position welding process of petrochemical process pipelines in real time, this approach addresses the challenge of identifying welding defects in real time for welding robots, thereby promoting the intelligent application of automated pipeline welding with robotic welding. Therefore, this paper proposes a mapping relationship between welding position, weld pool visual and electrical signals, and welding defects, establishing a model for predicting automated, all-position welding defects for robotic pipelines. This method implements a method for monitoring defects in robotic, all-position welding.

[0071] The present invention provides an online detection method for all-position welding penetration by coordinating electrical signals and molten pool images. Based on the symmetry of welding deformation and the complementarity of arc starting and extinction points, dual robots are used to coordinate all-position welding of pipelines. For the setting of all-position welding process parameters for pipelines, a preliminary segmentation is first performed based on the welding position and then further segmented at 5-10°. Then, a laser scanner is used to scan the pipeline groove to obtain the groove size of each section. Finally, the welding process parameters are adaptively set based on the process knowledge base and the welding parameter adjustment model. For the problem of pipeline burn-through defect identification, the molten pool image and the time domain eigenvalues ​​of the electrical signal are fused and judged through a neural network prediction model. For the problem of incomplete penetration identification, the molten pool size characteristics are first obtained based on the molten pool image and the reference molten pool oscillation frequency is calculated based on the reference molten pool oscillation frequency model. Then, the upper envelope signal of the voltage signal is subjected to spectrum analysis and the actual molten pool oscillation frequency is calculated. Finally, whether the weld is fully penetrated is determined by comparing the actual and reference molten pool oscillation frequencies.

[0072] The present invention will be further described below with reference to the accompanying drawings.

[0073] The specific process is as follows Figure 1 shown.

[0074] Please follow the steps below to implement it:

[0075] The present invention provides an online detection method for all-position welding penetration by coordinating electrical signals with molten pool images, comprising the following steps:

[0076] Step 1: Before welding, input the pipe material, wall thickness and diameter information, and start the welding robot equipped with a laser scanner to scan the pipe groove from bottom to top to obtain the original data of the groove profile at different positions.

[0077] Step 2: If Figure 2 As shown in the figure, the pipeline groove is initially segmented based on the characteristics of overhead welding, flat welding and vertical welding, that is, 0-30° is flat welding, 30°-150° is vertical welding and 150°-180° is overhead welding. Then, based on the preliminary segmentation, 5-10° segmentation is performed. At the same time, the corresponding welding process parameters are matched from the process knowledge base shown in Table 1 according to the position and groove size of each segment.

[0078] Table 1 Process knowledge base in the online detection method of all-position welding penetration coordinated by electrical signals and molten pool images

[0079]

[0080] Furthermore, the corresponding welding process parameters are matched from the process knowledge base according to the position and groove size of each section. The process knowledge base content includes material, wall thickness, pipe diameter, position type, groove gap, groove misalignment, welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time. The welding types include flat welding at 0°, vertical welding at 90°, and overhead welding at 180°.

[0081] The process of obtaining welding process parameters for each section is as follows:

[0082] First, the corresponding welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time are obtained based on the input material, wall thickness, pipe diameter, segmentation, and position type, groove gap, and groove misalignment obtained by the laser scanner.

[0083] Secondly, calculate the position of each section of the slope. The specific formula is as follows:

[0084]

[0085] is the angle between the starting position of each segment and the vertical direction, is the angle between the end position of each segment and the vertical direction;

[0086] Finally, according to the position of each groove, some process parameters are adjusted;

[0087] When the section belongs to the flat welding position type, adjust the welding current and wire feeding speed. The specific formula is as follows:

[0088]

[0089] in, and To match the welding current and wire feeding speed in the process knowledge base, and is a constant;

[0090] When the section belongs to the vertical welding position type, adjust the welding current and welding speed. The specific formula is as follows:

[0091]

[0092]

[0093] in, To match the welding speed in the process knowledge base, 、 、 and is a constant;

[0094] When the section belongs to the overhead welding position type, adjust the wire feeding speed and welding speed. The specific formula is as follows:

[0095]

[0096] in, and is a constant.

[0097] Step 3: Synchronously start both welding robots, along with a 160Hz welding parameter acquisition box and a 30Hz industrial camera to collect electrical signals and weld pool visuals. Based on the symmetry of welding deformation and the complementary nature of arc start and arc end points, one robot welds clockwise from 180° on the left side to 3° on the right side, while the other robot welds clockwise from 0° on the right side to 177° on the right side.

[0098] Step 4: If Figure 3 As shown, the welding is flawless and the welding position is between 40° and 50°. The arc voltage fluctuates slightly and is less than 12V. Figure 4As shown in the figure, there is a burn-through defect and the welding position is between 40° and 50°. The arc voltage fluctuates greatly and the maximum arc voltage value is higher than 30V. The molten pool image is incomplete. Therefore, for welding positions less than 50°, the burn-through defect can be analyzed by the time domain information of the arc voltage signal and the molten pool image. Figure 5 As shown, the welding is flawless and the welding position is between 50° and 60°. The arc voltage fluctuates slightly and is between 25 and 26 V. Figure 6 As shown, a burn-through defect exists and the welding position is between 50° and 60°. The arc voltage fluctuates slightly and is between 25 and 26V, but the weld pool image is incomplete. This shows that burn-through defects at different locations exhibit different characteristics in the electrical signal and weld pool image. Therefore, it is necessary to combine information such as the electrical signal, weld pool image, groove size, and position to predict burn-through defects.

[0099] The voltage signal collected within each segment is framed using a rectangular window of 500-700 mm in length. Time-domain analysis is then performed on the voltage signal within each frame. The time-domain analysis data, along with the material, wall thickness, pipe diameter, groove location and dimensions, welding process parameters, and the corresponding weld pool image within the rectangular window, are then fed into the electrical signal-image collaborative perception burn-through defect prediction model. If the output indicates a burn-through, welding is stopped.

[0100] Furthermore, the electrical signal-image collaborative perception burn-through defect prediction model includes:

[0101] The material, wall thickness, pipe diameter, groove position of the section, groove size, and welding process parameters are normalized. The material values ​​0.3, 0.6, and 0.9 represent carbon steel, stainless steel, and nickel-based alloy, respectively. Other parameters are normalized using the following calculation formula:

[0102]

[0103] in, and They are the maximum and minimum values ​​of the same type of data respectively.

[0104] The time domain analysis of voltage signals includes mean, variance, peak-to-peak value, and kurtosis. The calculation formulas are as follows:

[0105]

[0106]

[0107]

[0108]

[0109] in, is the length of the rectangular window, is the mean, is the variance, is the peak-to-peak value and is a steep value.

[0110] The electrical signal-image collaborative perception burn-through defect prediction model adopts a two-branch prediction model. One branch uses the Resnet convolutional neural network as the structural backbone and is used to extract the molten pool image features. The other branch uses the AlexNet convolutional neural network as the structural backbone and is used to extract data features consisting of variance, peak-to-peak value and kurtosis, material, wall thickness, pipe diameter, groove position, groove size, and welding process parameters. Finally, feature fusion is performed through the fully connected layer.

[0111] Step 5: If the output of the electrical signal-image collaborative perception burn-through defect prediction model is unknown, the collected voltage signal is subjected to frequency domain analysis to obtain the voltage signal's upper envelope spectrum data, i.e., the actual melt pool oscillation frequency. Simultaneously, the melt pool size is obtained based on the melt pool image, and the melt pool reference oscillation frequency under the melt-through state is calculated. If the actual melt pool oscillation frequency is greater than the reference oscillation frequency under the melt-through state, welding is immediately stopped. If the actual melt pool oscillation frequency is less than or equal to the reference oscillation frequency under the melt-through state, welding is continued.

[0112] Furthermore, obtaining the upper envelope spectrum data of the voltage signal includes:

[0113] First, the voltage signal in the time domain is converted into the spectrum of the frequency domain signal using Fourier transform;

[0114] Secondly, calculate the start and end frequencies of the intercepted frequency band according to the following formula:

[0115]

[0116] in, and They are the start and end frequencies of the intercepted frequency band when the pipe diameter is 150mm, the wall thickness is 7mm, the groove gap is 1.5mm, and the groove misalignment is 1mm; 、 、 and They are groove gap, groove misalignment, wall thickness and pipe diameter; 、 、 、 、 and is a constant.

[0117] Next, obtain the peak coordinates of the signal in the intercepted frequency band, sort the peak coordinates according to the size of the spectrum ordinate value, and save the two frequencies with the largest ordinate values. and ;

[0118] Then, and The voltage signal is filtered with the cutoff frequency of the bandpass filter, and the upper envelope signal of the filtered voltage signal is obtained.

[0119] Finally, the upper envelope signal of the voltage signal is subjected to spectrum analysis to obtain the actual oscillation frequency of the molten pool.

[0120] Furthermore, obtaining the molten pool size based on the molten pool image and calculating the molten pool reference oscillation frequency in the melt-through state include:

[0121] First, the molten pool height and the length of the upper surface of the molten pool are obtained based on the molten pool image;

[0122] Secondly, based on the equivalent cylindrical film of molten metal in the melt-through state and the equal volume principle, the equivalent cylindrical diameter is obtained, and the following equation is obtained:

[0123]

[0124] in, is the height of the molten pool, is the length of the upper surface of the molten pool, is the equivalent melt pool radius.

[0125] At the same time, based on the fact that the molten pool is a convex frustum in the actual welding process, the above inequality is estimated to have a maximum value. Therefore, the above equation is simplified to:

[0126]

[0127] Thus, the equivalent melt pool diameter is obtained:

[0128]

[0129] Then, based on the oscillation model of the molten pool as a crescent-shaped molten pool, the wave number k is obtained, and the calculation formula is as follows:

[0130]

[0131] Finally, the oscillation frequency of the molten pool under the reference penetration state is calculated according to the following formula:

[0132]

[0133] in, is the density of molten metal, and is a constant.

Claims

1. An online detection method for all-position welding penetration by coordinating electrical signals with molten pool images, characterized in that: The following steps are involved: Step 1: Before welding, input the pipe material, wall thickness, and diameter information. Simultaneously, start the welding robot equipped with a laser scanner to scan the pipe groove from bottom to top to obtain the original data of the groove profile at different locations. Step 2: Preliminary segmentation of the pipe groove is performed based on the characteristics of overhead welding, flat welding, and vertical welding. That is, 0-30° is flat welding, 30°-150° is vertical welding, and 150°-180° is overhead welding. Then, based on the preliminary segmentation, further 5-10° segments are performed. At the same time, the corresponding welding process parameters are matched from the process knowledge base according to the position and groove size of each segment. Step 3: Synchronously start the two welding robots, and simultaneously start the welding parameter acquisition box with a collection frequency of 150-200 Hz and the industrial camera with a collection frequency of 30-50 Hz to collect electrical signals and weld pool vision; one welding robot welds clockwise from 180° on the left half to 3°-6° on the right half, and the other welding robot welds clockwise from 0° on the right half to 174°-177° on the right half; Step 4: The voltage signal collected in each segment is framed using a rectangular window of 500-700 mm in length. Time domain analysis is performed on the voltage signal in each frame. The time domain analysis data, along with the material, wall thickness, pipe diameter, groove location and size of the segment, welding process parameters, and the corresponding molten pool image in the rectangular window, are input into the electrical signal-image collaborative perception burn-through defect prediction model. If the output result is a burn-through, welding is stopped. If the output result is unknown, step 5 is executed. The electrical signal-image collaborative perception burn-through defect prediction model includes: The material, wall thickness, pipe diameter, groove position of the section, groove size, and welding process parameters are normalized. The material values ​​0.3, 0.6, and 0.9 represent carbon steel, stainless steel, and nickel-based alloy, respectively. Other parameters are normalized using the following calculation formula: Among them, y max and y min They are the maximum and minimum values ​​of the same type of data respectively; The time domain analysis of voltage signals includes mean, variance, peak-to-peak value, and kurtosis. The calculation formulas are as follows: s p =s max -s min Where n is the length of the rectangular window, is the mean, F is the variance, s p is the peak-to-peak value and K is the steepness value; The electrical signal-image collaborative perception burn-through defect prediction model uses a two-branch prediction model. One branch uses a ResNet convolutional neural network as the backbone structure to extract weld pool image features, and the other branch uses an AlexNet convolutional neural network as the backbone structure to extract data features consisting of variance, peak-to-peak value and kurtosis, material, wall thickness, pipe diameter, groove position, groove size, and welding process parameters. Finally, the features are fused through a fully connected layer. Step 5: Perform frequency domain analysis on the collected voltage signal and obtain the envelope spectrum data of the voltage signal, i.e., the actual molten pool oscillation frequency. Simultaneously, the molten pool size is obtained based on the molten pool image, and the reference oscillation frequency of the molten pool in the melt-through state is calculated. When the actual molten pool oscillation frequency is greater than the reference oscillation frequency of the molten pool in the melt-through state, welding is immediately stopped. When the actual molten pool oscillation frequency is less than or equal to the reference oscillation frequency of the molten pool in the melt-through state, welding is continued. The step of obtaining the upper envelope spectrum data of the voltage signal includes: First, the voltage signal in the time domain is converted into the spectrum of the frequency domain signal using Fourier transform; Secondly, calculate the start and end frequencies of the intercepted frequency band according to the following formula: Among them, f s * and f e * are the start and end frequencies of the intercepted frequency band for a pipe diameter of 150 mm, a wall thickness of 7 mm, a groove gap of 1.5 mm, and a groove misalignment of 1 mm; σ, δ, w, and D are the groove gap, groove misalignment, wall thickness, and pipe diameter, respectively; a1, a2, b1, b2, c1, and c2 are constants; Next, obtain the peak coordinates of the signal in the intercepted frequency band, sort the peak coordinates according to the size of the spectrum ordinate value, and save the two frequencies f1 and f2 with the largest ordinate values; Then, the collected voltage signal is filtered using f1 and f2 as the cutoff frequencies of the bandpass filter, and the upper envelope signal of the filtered voltage signal is obtained; Finally, the upper envelope signal of the voltage signal is subjected to spectrum analysis to obtain the actual oscillation frequency of the molten pool; The method of obtaining the molten pool size based on the molten pool image and calculating the molten pool reference oscillation frequency in the melt-through state includes: First, the molten pool height and the length of the upper surface of the molten pool are obtained based on the molten pool image; Secondly, based on the equivalent cylindrical film of molten metal in the melt-through state and the equal volume principle, the equivalent cylindrical diameter is obtained, and the following equation is obtained: Where h is the height of the molten pool, r top is the length of the upper surface of the molten pool, r eq is the equivalent melt pool radius; At the same time, based on the fact that the molten pool is a convex frustum in the actual welding process, the above inequality is estimated to have a maximum value; therefore, the above equation is simplified to: Thus, the equivalent melt pool diameter is obtained: Then, based on the oscillation model of the molten pool as a crescent-shaped molten pool, the wave number k is obtained, and the calculation formula is as follows: Finally, the oscillation frequency of the molten pool under the reference penetration state is calculated according to the following calculation formula: Among them, ρ s is the density of molten metal, t cz and ε cz is a constant.

2. The method for online detection of all-position welding penetration by coordinating electrical signals and molten pool images according to claim 1 is characterized in that: The steps in step 2 to match the corresponding welding process parameters from the process knowledge base according to the position and groove size of each section include: The process knowledge base includes material, wall thickness, pipe diameter, welding position type, groove gap, groove misalignment, welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time. The welding position types include 0° flat welding, 90° vertical welding, and 180° overhead welding. The process of obtaining welding process parameters for each section is as follows: First, the corresponding welding current, wire feed speed, welding speed, swing length, swing amplitude, left dwell time, and right dwell time are obtained based on the input material, wall thickness, pipe diameter, and welding position type, groove gap, and groove misalignment obtained by segmentation and laser scanner. Secondly, calculate the position of each section of the slope. The specific formula is as follows: θ S is the angle between the starting position of each segment and the vertical direction, θ F is the angle between the end position of each segment and the vertical direction; Finally, according to the position of each groove, some process parameters are adjusted: When the section belongs to the flat welding position type, adjust the welding current and wire feeding speed. The specific formula is as follows: Among them, I * and V s * To match the welding current and wire feeding speed in the process knowledge base, η1 and η2 are constants; When the section belongs to the vertical welding position type, adjust the welding current and welding speed. The specific formula is as follows: Among them, V w * To match the welding speed in the process knowledge base, α1, α2, β1, and β2 are constants; When the section belongs to the overhead welding position type, adjust the wire feeding speed and welding speed. The specific formula is as follows: Where μ1 and μ2 are constants.

Images