An AI-based parameter management system for soldering stations

Abstract

The present invention discloses a parameter management system for a soldering station based on AI, which relates to the field of welding technology. By analyzing in detail the welding temperature state, appearance state, and morphological state of the soldering station, key characteristic parameters of welding quality are extracted, realizing comprehensive monitoring and analysis of the welding process, which helps to track and analyze process problems in the welding process and provides important data basis for subsequent quality control and improvement strategies. By comprehensively analyzing the characteristic parameters, a welding quality index is obtained, and whether the welding parameters need to be adjusted is intelligently judged according to the change trend of the welding quality index, improving the response speed. By comprehensively analyzing the average temperature, average defect, and average shape deviation, the welding speed is obtained, and dynamic adjustment is carried out according to the welding speed to ensure that the welding process is carried out in the optimal state, realizing effective control of welding quality, improving product quality and production efficiency, and reducing the quality risk of welded products.

Description

Technical Field

[0001] The present invention relates to the technical field of welding, and particularly to a parameter management system for a soldering station based on AI. Background Art

[0002] A soldering station is a general term for equipment or tools for welding operations, mainly used to control various parameters during the welding process, such as temperature, current, welding speed, etc.; the soldering station plays a very important role in the welding process, especially in large-scale production and applications that require high-precision welding; during the welding process, the welding speed is a crucial factor, which directly affects the quality, efficiency, and cost of the welding process;

[0003] Currently, the welding speed of a soldering station usually depends on manual adjustment by a human operator. This adjustment method is affected by the operator's skill level and experience, and there is a risk of introducing human errors, resulting in fluctuations in welding quality. Summary of the Invention

[0004] Based on this, in order to address the problems mentioned in the above background art, it is necessary to provide a parameter management system for a soldering station based on AI.

[0005] The object of the present invention can be achieved through the following technical solutions: A parameter management system for a soldering station based on AI, comprising: a database, a feature analysis module, a requirement analysis module, and a parameter management module;

[0006] The database communicates with a sensor (high-definition camera) to obtain welding information and stores it; the welding information includes a welding molten pool image, and several blocks are divided on the welding molten pool image, and each block is marked with a corresponding temperature;

[0007] The feature analysis module performs feature analysis on the welding temperature state, welding appearance state, and welding morphology state based on the welding information of the soldering station to extract feature parameters. Among them, the welding information includes a welding molten pool image; the feature parameters include a welding temperature deviation value, a welding defect value, and a morphology deviation value;

[0008] The requirement analysis module comprehensively analyzes the welding quality based on the feature parameters to determine whether welding adjustment is required and generates a welding control instruction;

[0009] The parameter management module regulates the welding speed based on the received welding control instruction. The specific steps are as follows:

[0010] Step 1: Retrieve the welding defect value and the morphology deviation value at each acquisition moment, and calculate their respective means to obtain a defect mean and a morphology deviation mean, which are respectively denoted as and ;

[0011] Step 2: Retrieve the welding molten pool images corresponding to each acquisition moment, extract the temperatures of each block in the welding molten pool images, calculate the temperatures of each block to obtain the welding temperatures at each acquisition moment; calculate the average value of the welding temperatures at each acquisition moment and denote it as ;

[0012] Step 3: Obtain the current welding speed denoted as V1, and substitute it together with the defect average value , the shape deviation average value , and the temperature average value into the set formula for calculation to obtain the welding speed V2, where n1, n2, and n3 are respectively set proportionality coefficients. When the standard temperature HT > the temperature average value , then m takes the value of -1; it indicates that the welding temperature is lower than the standard temperature and speed reduction is required, so the smaller V2 is; when the standard temperature HT < the temperature average value , then m takes the value of 1; it indicates that the welding temperature is higher than the standard temperature and speed increase is required, so the larger V2 is;

[0013] Step 4: Adjust the welding process according to the welding speed. Each time an adjustment is made, the feature analysis module, requirement analysis module, and parameter management module are repeated.

[0014] In some embodiments, the specific process of the feature analysis of the welding temperature state is as follows:

[0015] Set that each type of product welding corresponds to a standard process parameter respectively, where the standard process parameter includes the standard temperature denoted as HT and the standard shape of the welding molten pool;

[0016] Retrieve the welding molten pool images corresponding to each acquisition moment, and extract the temperature of each block in the welding molten pool images denoted as Tij; where i = 1, 2, 3...I, j = 1, 2, 3...J, I and J take positive integer values, I represents the total number of blocks in the welding molten pool image, i represents the serial number of any block in the welding molten pool image; J represents the total number of acquisition moments, and j represents the serial number of any one of the acquisition moments;

[0017] Compare the product type of the soldering station with all the set product types to match the corresponding standard process parameter, where the standard process parameter includes the standard temperature and the standard shape of the welding molten pool, and denote the standard temperature as HT; use the set formula for calculation to obtain the welding temperature deviation value Tbj, where b1 and b2 are respectively set proportionality coefficients.

[0018] In some embodiments, the specific process of the feature analysis of the welding appearance state is as follows:

[0019] Identify the pore parts, slag inclusion parts, and crack parts existing in the welding molten pool image through the neural network algorithm, and use the edge detection algorithm to calculate the pore area of each pore part, the slag inclusion area of each slag inclusion part, and the crack area of each crack part;

[0020] Count the number of pore parts, and sum up the pore areas of each pore part to obtain the total pore area, denoted as Z1;

[0021] Count the number of slag inclusion parts, and sum up the slag inclusion areas of each slag inclusion part to obtain the total slag inclusion area, denoted as Z2;

[0022] Compare and analyze the pore areas of each pore part and the slag inclusion areas of each slag inclusion part respectively to obtain the pore coefficient and the slag inclusion coefficient, and denote them as Qα and Qβ respectively;

[0023] Substitute the total pore area Z1, the pore defect coefficient Qα, the total slag inclusion area Z2, and the slag inclusion defect coefficient Qβ into the set formula Perform calculations to obtain the welding defect value Qb corresponding to the welding molten pool image, where b3 and b4 are the set proportionality coefficients respectively; thus, the welding defect values corresponding to each acquisition moment are denoted as Qbj.

[0024] In some embodiments, the specific process of comparing and analyzing the pore areas of each pore part and the slag inclusion areas of each slag inclusion part is as follows:

[0025] Compare and analyze the pore areas of each pore part with the set pore interval to divide the pore parts corresponding to the pore areas into high pore distribution, medium pore distribution, and low pore distribution, count the cumulative numbers of high pore distribution, medium pore distribution, and low pore distribution in the welding molten pool image respectively, and denote them as Q1, Q2, and Q3; substitute Q1, Q2, and Q3 into the set formula Perform calculations to obtain the pore defect coefficient Qα, where a1, a2, and a3 are the set proportionality coefficients respectively, and a1 > a2 > a3 > 0;

[0026] Similarly, compare and analyze the slag inclusion areas of each slag inclusion part with the set slag inclusion interval to obtain the slag inclusion defect coefficient.

[0027] In some embodiments, the specific process of the characteristic analysis of the welding morphology state is as follows:

[0028] Extract the edge of the welding molten pool image through the edge detection algorithm to obtain the welding molten pool morphology, and perform coincidence comparison with the standard welding molten pool morphology to obtain the union area and the intersection area, and denote them as G1 and G2 respectively. Use the set formula Perform calculations to obtain the morphology deviation value, where γ is the set correction factor respectively; thus, the morphology deviation values corresponding to each acquisition moment can be obtained.

[0029] In some embodiments, the steps for generating the welding control instruction are as follows:

[0030] Retrieve the characteristic parameters at each acquisition moment, where the characteristic parameters include the welding temperature deviation value Tbj, the welding defect value Qbj, and the shape deviation value Gbj, and calculate the welding quality index TQGj through a set formula where d1, d2, and d3 are respectively set proportionality coefficients;

[0031] Construct a two-dimensional rectangular coordinate system with time as the abscissa and the welding quality index as the ordinate. Input the welding quality index into the coordinate axis according to its corresponding acquisition moment, and record the position of the welding quality index in the coordinate axis as the control point. Connect the control points in sequence with a smooth curve to obtain a curve graph of the welding quality index changing with time;

[0032] Make a tangent to the curve at each control point, and calculate the slope of the tangent to obtain the control slope of each control point, denoted as Kj. Sum the control slopes greater than zero to obtain the control increase degree, denoted as D1, and sum the control slopes less than zero and take the absolute value to obtain the control decrease degree, denoted as D2;

[0033] Substitute the control slope Kj, the control increase degree D1, the control decrease degree D2, and the welding quality index TQGj into a set formula to calculate the demand value Dg, where g1, g2, and g3 are respectively set proportionality coefficients, is the average value of the control slopes of each control point; when the demand value is greater than the set demand threshold, a welding control instruction is generated.

[0034] Compared with the prior art, the beneficial effects of the present invention are:

[0035] 1. By analyzing the welding temperature state, appearance state, and shape state of the soldering station in detail, the key characteristic parameters of welding quality are extracted, including the welding temperature deviation value, the welding defect value, and the shape deviation value, realizing comprehensive monitoring and analysis of the welding process, which helps to track and analyze the process problems in the welding process and provides important data basis for subsequent quality control and improvement strategies;

[0036] 2. By comprehensively analyzing the characteristic parameters to obtain the welding quality index, it can be intelligently judged whether the welding parameters need to be adjusted according to the change trend of the welding quality index. The automatic judgment can timely detect potential welding quality problems and improve the response speed;

[0037] 3. The welding speed is obtained through comprehensive analysis of the average temperature, average defect, and average shape deviation, and dynamic adjustment is carried out based on the welding speed to ensure that the welding process is carried out in an optimal state, which can better control the heat input, thereby avoiding overheating or insufficient heating and saving energy consumption. At the same time, it reduces the need for manual adjustment, reduces the impact of human error, and makes the welding process more stable and consistent. It realizes effective control of welding quality, improves product quality and production efficiency, and reduces the quality risk of welded products.

[0038] In summary, the present invention can achieve comprehensive monitoring and intelligent optimization of the welding process, improve welding quality and process stability, and also enhance the automation level and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0040] Figure 1 It is a principle block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention in conjunction with the drawings. Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein. Those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0042] As Figure 1 shown, a parameter management system for a soldering station based on AI includes: a database, a feature analysis module, and a parameter management module;

[0043] The database communicates with a sensor (high-definition camera) to obtain welding information and stores it. The welding information includes a welding molten pool image, on which several blocks are divided, and each block is marked with a corresponding temperature. It should be noted that the welding molten pool image refers to the image of the welding molten pool captured by a vision system or imaging device (such as a camera, laser scanner, etc.) during the welding process. The molten pool is the molten metal area formed during the welding process, which is one of the key parts of the welding process and has an important impact on welding quality and joint performance. The temperature of each block is obtained through an infrared thermal imaging instrument and marked on the corresponding block of the welding molten pool image.

[0044] The feature analysis module analyzes based on the welding temperature state, appearance state, and morphological state of the soldering station to extract feature parameters, where the feature parameters include the welding temperature deviation value, welding defect value, and morphological deviation value; specifically:

[0045] It is set that each type of product welding corresponds to a standard process parameter, where the standard process parameter includes the standard temperature and the standard morphology of the welding molten pool; it should be noted that the standard temperature refers to the optimal temperature range to be maintained during the welding process, and different materials, welding methods, and product requirements have different standard temperatures to ensure the welding quality and performance; the standard morphology of the welding molten pool includes the shape, size, etc. of the molten pool, and the standard morphology is usually based on the design requirements of the welded joint to ensure the uniformity of the molten pool and the quality of the weld seam;

[0046] Temperature analysis:

[0047] Retrieve the welding molten pool images corresponding to each acquisition moment, and extract the temperature of each block in the welding molten pool image and record it as Tij; where i = 1, 2, 3...I, j = 1, 2, 3...J, I and J are positive integers, I represents the total number of blocks in the welding molten pool image, and i represents the serial number of any block in the welding molten pool image; J represents the total number of acquisition moments, and j represents the serial number of any one of the acquisition moments;

[0048] Compare the product type of the soldering station with all the set product types to match the corresponding standard process parameters, where the standard process parameters include the standard temperature and the standard morphology of the welding molten pool, and record the standard temperature as HT; use the set formula to calculate the welding temperature deviation value Tbj, where b1 and b2 are respectively the set proportionality coefficients; it can be seen from the formula that the more uneven the temperature of each block in the welding molten pool image, the larger the welding temperature deviation value; the farther the temperature of the block is from the standard temperature, the larger the welding temperature deviation value;

[0049] Appearance state:

[0050] The neural network algorithm is used to identify the porosity parts, slag inclusion parts, and crack parts existing in the welding molten pool image, and the edge detection algorithm is used to calculate the porosity area of each porosity part, the slag inclusion area of each slag inclusion part, and the crack area of each crack part; it should be noted that the porosity parts, slag inclusion parts, and crack parts are collectively referred to as defect parts. The edge detection algorithm is applied to identify the edges of the defect parts in the image and extract the contours of the defect parts. The area of each defect part is calculated using the contour information in the image, usually by using the pixel counting method, that is, calculating the number of pixel points inside the contour and then converting it into the actual area. Thus, the defect area of each defect part can be obtained; Porosity will cause a decrease in the metal strength inside the weld and affect the welding quality; improper control of the welding process, such as too fast welding speed, will lead to slag inclusion. Slag inclusion will cause a decrease in the weld strength because the metal in the slag inclusion area usually cannot be fully fused with the base material, affecting the appearance and final quality of the welded joint;

[0051] Count the number of porosity parts, and sum up the porosity areas of each porosity part to obtain the total porosity area, denoted as Z1;

[0052] Compare and analyze the porosity area of each porosity part with the set porosity interval. When the porosity area is greater than the upper limit of the set porosity interval, then accumulate a high porosity distribution; when the porosity area is within the set porosity interval, then accumulate a medium porosity distribution; when the porosity area is less than the lower limit of the set porosity interval, then accumulate a low porosity distribution; respectively count the cumulative numbers of the high porosity distribution, medium porosity distribution, and low porosity distribution in the welding molten pool image, and denote them as Q1, Q2, Q3 respectively; Use the set formula to calculate the porosity defect coefficient Qα, where a1, a2, a3 are respectively the set proportionality coefficients, and a1 > a2 > a3 > 0;

[0053] Count the number of slag inclusion parts, and sum up the slag inclusion areas of each slag inclusion part to obtain the total slag inclusion area, denoted as Z2;

[0054] Compare and analyze the slag inclusion area of each slag inclusion part with the set slag inclusion interval. When the slag inclusion area is greater than the upper limit of the set slag inclusion interval, then accumulate a high slag inclusion distribution; when the slag inclusion area is within the set slag inclusion interval, then accumulate a medium slag inclusion distribution; when the slag inclusion area is less than the lower limit of the set slag inclusion interval, then accumulate a low slag inclusion distribution; respectively count the cumulative numbers of the high slag inclusion distribution, medium slag inclusion distribution, and low slag inclusion distribution in the welding molten pool image, and denote them as Q4, Q5, Q6 respectively; Use the set formula to calculate the slag inclusion defect coefficient Qβ, where are respectively the set proportionality coefficients, and a4 > a5 > a6 > 0;

[0055] The total pore area Z1, the pore defect coefficient Qα, the total slag inclusion area Z2, and the slag inclusion defect coefficient Qβ are calculated through a set formula to obtain the welding defect value Qb corresponding to the welding molten pool image, where b3 and b4 are respectively set proportionality coefficients; thus, the welding defect values corresponding to each acquisition moment are denoted as Qbj;

[0056] Morphological analysis:

[0057] The edge of the welding molten pool image is extracted through an edge detection algorithm to obtain the welding molten pool morphology, and it is compared with the standard morphology of the welding molten pool to obtain the union area and the intersection area, which are respectively denoted as G1 and G2. The union area refers to the total area formed by the overlap of the actual welding molten pool morphology and the standard morphology of the welding molten pool, that is, the combined area of the two; the intersection area refers to the area of the overlapping part formed by the actual welding molten pool morphology and the standard morphology of the welding molten pool, that is, the area where the two coexist; it should be noted that when the intersection area and the union area are closer, it means that the extracted welding molten pool morphology is closer to the standard morphology of the welding molten pool and more in line with the process requirements of product welding; using the set formula to calculate the morphological deviation value, where γ is respectively the set correction factor; thus, the morphological deviation values corresponding to each acquisition moment are denoted as Gbj;

[0058] The welding temperature deviation value Tbj, the welding defect value Qbj, and the morphological deviation value Gbj corresponding to each acquisition moment are denoted as characteristic parameters and sent to the parameter management module;

[0059] By analyzing in detail the welding temperature state, the appearance state, and the morphological state of the soldering station, the key characteristic parameters of welding quality are extracted. The characteristic parameters include the welding temperature deviation value, the welding defect value, and the morphological deviation value, realizing the comprehensive monitoring and analysis of the welding process, which helps to track and analyze the process problems in the welding process and provides important data basis for subsequent quality control and improvement strategies.

[0060] The requirements analysis module analyzes and judges the regulation requirements for the welding work of the soldering station based on the received characteristic parameters to determine whether welding parameter regulation is required. If so, a welding regulation instruction is generated; specifically:

[0061] Retrieve the characteristic parameters at each acquisition moment, where the characteristic parameters include the welding temperature deviation value Tbj, the welding defect value Qbj, and the morphological deviation value Gbj, and use the set formula Calculate to obtain the welding quality index TQGj, where d1, d2, and d3 are respectively the set proportionality coefficients; It can be seen from the formula that when the welding temperature deviation value is larger, it indicates that the temperature deviation during the welding process is larger, which is more inconsistent with the process requirements and affects the welding quality of the product, then the welding quality index is larger; When the welding defect value is larger, it indicates that the impact on the appearance and final quality of the welding is greater, then the welding quality index is larger; When the morphological deviation value is larger, it indicates that the shape of the welded joint is more inconsistent with the design requirements, which will lead to insufficient joint strength, poor sealing, or other performance problems, then the welding quality index is larger;

[0062] Construct a two-dimensional rectangular coordinate system with time as the abscissa and the welding quality index as the ordinate. Input the welding quality index into the coordinate axis according to its corresponding acquisition time, and record the position of the welding quality index in the coordinate axis as the control point. Connect the control points in sequence with a smooth curve to obtain the curve of the welding quality index changing with time; At each control point, make a tangent to the curve and calculate the slope of the tangent to obtain the control slope of each control point, denoted as Kj; It should be noted that when the control slope is greater than zero, it indicates that the welding quality index shows an increasing trend at this control point; When the control slope is less than zero, it indicates that the welding quality index shows a decreasing trend at this control point; Sum the control slopes greater than zero to calculate the control increase degree, denoted as D1, and sum the control slopes less than zero and take the absolute value to calculate the control decrease degree, denoted as D2; Use the set formula Calculate to obtain the demand value Dg, where g1, g2, and g3 are respectively the set proportionality coefficients, is the average value of the control slopes of each control point; It can be seen from the formula that when the control slope fluctuates more, it indicates that the welding quality of the products on the soldering station fluctuates more, then the demand value is larger; When the overall increasing trend of the welding quality index is greater and the decreasing trend is smaller, then the demand value is larger;

[0063] Compare and analyze the demand value with the set demand threshold. When the demand value is greater than the set demand threshold, it indicates the risk that the welding quality of the products on the soldering station does not meet the process requirements, and it is necessary to adjust the welding parameters of the soldering station to improve the welding quality; Then generate a welding control instruction;

[0064] Obtain the welding quality index through comprehensive analysis of the characteristic parameters, and intelligently judge whether it is necessary to adjust the welding parameters based on the change trend of the welding quality index. The automatic judgment can timely detect potential welding quality problems and improve the response speed.

[0065] The parameter management module conducts in-depth welding analysis based on the received welding control instruction to obtain the welding adjustment parameters, and controls the welding speed accordingly to make it meet the working requirements of the welded product, so as to improve the quality and efficiency of the welded product; Specifically:

[0066] Step 1: Retrieve the welding defect value Qbj and the shape deviation value Gbj at each acquisition moment, and calculate their respective means to obtain the defect mean and the shape deviation mean, which are denoted as and ; It should be noted that when the welding defect value is high, it indicates that there are pores or excessive slag inclusions during the welding process. The welding speed can be reduced to increase the stability and filling time of the molten pool, and reduce the occurrence of defects such as pores and slag inclusions. When the shape deviation value is large, the welding speed can be reduced to improve the shape consistency of the weld;

[0067] Step 2: Retrieve the welding molten pool images corresponding to each acquisition moment, extract the temperature Tij of each block in the welding molten pool image, calculate the temperatures of each block to obtain the welding temperature at each acquisition moment; calculate the mean value of the welding temperatures at each acquisition moment to obtain the temperature mean, denoted as ;

[0068] Step 3: Obtain the current welding speed, denoted as V1, and calculate the welding speed V2 through the set formula where n1, n2, and n3 are respectively the set proportionality coefficients. When the standard temperature HT > the temperature mean , then m takes the value of -1; it indicates that the welding temperature is lower than the standard temperature and deceleration is required, so the smaller V2 is. When the standard temperature HT < the temperature mean , then m takes the value of 1; it indicates that the welding temperature is higher than the standard temperature and acceleration is required, so the larger V2 is. It can be seen from the formula that when the temperature is lower, the welding speed is lower to allow more heat to accumulate in the weld, thereby increasing the temperature. When the defect mean is larger, the welding speed is lower to increase the stability and filling time of the molten pool and reduce the occurrence of defects such as pores and slag inclusions. When the shape deviation mean is larger, the welding speed is lower;

[0069] Step 4: Adjust the welding process according to the welding speed V2. Each time an adjustment is made, the feature analysis module, the requirement analysis module, and the parameter management module are repeated;

[0070] The welding speed is obtained through comprehensive analysis of the temperature mean, the defect mean, and the shape deviation mean, and dynamic adjustment is carried out according to the welding speed to ensure that the welding process is carried out in an optimal state, which can better control the heat input, thereby avoiding overheating or insufficient heating and saving energy consumption; at the same time, it reduces the need for manual adjustment, reduces the influence of human error, and makes the welding process more stable and consistent; it realizes effective control of welding quality, improves product quality and production efficiency, and reduces the quality risk of welded products.

[0071] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0072] The above-described embodiments only represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. An AI-based parameter management system for a soldering station, characterized in that, It includes a feature analysis module, a requirement analysis module, and a parameter management module; The feature analysis module performs feature analysis on the welding temperature state, welding appearance state, and welding morphology state based on the welding information of the soldering station to extract feature parameters. Among them, the welding information includes the welding molten pool image; the feature parameters include the welding temperature deviation value, the welding defect value, and the morphology deviation value; The requirement analysis module comprehensively analyzes the welding quality based on the feature parameters to determine whether welding adjustment is required and generates a welding control instruction; The parameter management module adjusts the welding speed based on the received welding control instruction. The specific steps are as follows: Step 1: Retrieve the welding defect value and the morphology deviation value at each acquisition moment, and calculate their respective means to obtain the defect mean and the form deviation mean; Step 2: Retrieve the welding molten pool images corresponding to each acquisition moment, extract the temperature of each block in the welding molten pool image, calculate the welding temperature at each acquisition moment by calculating the temperatures of each block; calculate the mean value of the welding temperatures at each acquisition moment to obtain the temperature mean; Step 3: Obtain the current welding speed, denoted as V1, normalize it with the defect mean, the form deviation mean, and the temperature mean, take their numerical values, and analyze the numerical values to obtain the welding speed; Step 4: Adjust the welding process according to the welding speed. Each time an adjustment is made, the feature analysis module, the requirement analysis module, and the parameter management module are repeated; The specific process of the feature analysis of the welding temperature state is as follows: Set that each type of product welding corresponds to a standard process parameter, where the standard process parameter includes the standard temperature and the standard morphology of the welding molten pool; Retrieve the welding molten pool images corresponding to each acquisition moment, and extract the temperature of each block in the welding molten pool image, denoted as Tij; where i = 1, 2, 3...I, j = 1, 2, 3...J, I and J are positive integers, I represents the total number of blocks in the welding molten pool image, and i represents the serial number of any block in the welding molten pool image; J represents the total number of acquisition moments, and j represents the serial number of any one of the acquisition moments; Compare the product type of the soldering station with all the set product types to match the corresponding standard process parameters, where the standard process parameters include the standard temperature and the standard shape of the welding molten pool, and record the standard temperature as HT; use the set formula to calculate the soldering temperature deviation value Tbj, where b1 and b2 are the set proportionality coefficients respectively; The specific process of the feature analysis of the welding appearance state is as follows: Identify the pore sites, slag inclusion sites, and crack sites existing in the welding molten pool image through the neural network algorithm, and use the edge detection algorithm to calculate the pore area of each pore site, the slag inclusion area of each slag inclusion site, and the crack area of each crack site; Count the number of pore sites, and sum up the pore areas of each pore site to obtain the total pore area; Count the number of slag inclusion sites, and sum up the slag inclusion areas of each slag inclusion site to obtain the total slag inclusion area; Compare and analyze the pore areas of each pore site and the slag inclusion areas of each slag inclusion site respectively to obtain the pore defect coefficient and the slag inclusion defect coefficient; Normalize the total pore area, the pore defect coefficient, the total slag inclusion area, and the slag inclusion defect coefficient, take their numerical values, and analyze the numerical values to obtain the welding defect value corresponding to the welding molten pool image, and thus obtain the welding defect value corresponding to each acquisition moment; The specific process of comparing and analyzing the pore areas of each pore site and the slag inclusion areas of each slag inclusion site respectively is as follows: Compare and analyze the pore areas of each pore part with the set pore range to classify the pore parts corresponding to the pore areas into high pore distribution, medium pore distribution, and low pore distribution. Count the cumulative numbers of high pore distribution, medium pore distribution, and low pore distribution in the welding molten pool image respectively, and perform formula-based calculation and analysis on them to obtain the pore defect coefficient; Similarly, compare and analyze the slag inclusion areas of each slag inclusion part with the set slag inclusion range to obtain the slag inclusion defect coefficient; The specific process of the characteristic analysis of the welding morphology state is as follows: Extract the edges of the welding pool image through the edge detection algorithm to obtain the welding pool morphology, and compare it with the standard morphology of the welding pool to obtain the union area and the intersection area, which are respectively denoted as G1 and G2, and use the set formula to calculate the morphological deviation value, where γ is the set correction factor respectively; thus, the morphological deviation values corresponding to each acquisition moment can be obtained; The specific steps for generating the welding control instruction are as follows: Retrieve the characteristic parameters at each acquisition moment, where the characteristic parameters include the welding temperature offset value, the welding defect value, and the morphology deviation value, and perform formula-based calculation and analysis on them to obtain the welding quality index; Construct a two-dimensional rectangular coordinate system with time as the abscissa and the welding quality index as the ordinate. Input the welding quality index into the coordinate axis according to its corresponding acquisition moment, and record the position of the welding quality index in the coordinate axis as the control point. Connect the control points in sequence with a smooth curve to obtain the curve of the welding quality index changing with time; Make a tangent to the curve at each control point, and calculate the slope of the tangent to obtain the control slope of each control point. Sum up the control slopes greater than zero to obtain the control increase degree, and sum up the control slopes less than zero and take the absolute value to obtain the control decrease degree; Perform formula-based calculation and analysis on the control slope, the control increase degree, the control decrease degree, and the welding quality index to obtain the required value. When the required value is greater than the set required threshold, generate a welding control instruction.

Images