A welding process optimization method for single-sided heterogeneous composite steel plate

Abstract

The present invention discloses a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, and relates to the technical field of steel plate welding. The method comprises: interactively obtaining the usage scenario expectation and usage performance expectation of the composite steel plate to be produced to determine the surface steel plate and the base steel plate; interactively obtaining the design parameters of the composite steel plate to be produced, performing welding pretreatment analysis, and obtaining the surface welding interface and the base welding interface; obtaining the welding process associated parameters according to the design parameters of the composite steel plate to be produced, performing welding process optimization, and performing the single-sided heterogeneous composite steel plate welding based on the welding process optimization parameters after pretreatment of the surface welding interface and the base welding interface. The present invention solves the technical problems in the prior art that the welding quality fluctuates greatly, the residual stress after welding is large, and the reliability of the weldment is affected, and achieves the technical effect of improving the welding quality and ensuring the overall performance reliability of the composite steel plate.

Description

Technical Field

[0001] The invention relates to the technical field of steel plate welding, and in particular to a welding process optimization method for a single-sided heterogeneous composite steel plate. Background Art

[0002] In modern industrial production, single-sided heterogeneous composite steel plates are widely used in aerospace, automobile manufacturing, petrochemical and other fields due to their excellent properties, such as high strength, corrosion resistance, wear resistance, etc. However, welding of single-sided heterogeneous composite steel plates has always been a technical problem, because the two steel plates of different materials are prone to problems such as unstable welded joint performance, residual stress concentration, and large welding deformation during welding.

[0003] The existing technology has technical problems such as large fluctuations in welding quality and large residual stress after welding, which affects the reliability of weldments. Summary of the invention

[0004] The present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, which is used to solve the technical problems in the prior art such as large fluctuations in welding quality, large residual stress after welding, and influence on the reliability of weldments.

[0005] In view of the above problems, the present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate.

[0006] The present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, the method comprising: Interactively obtain the usage scenario expectations and usage performance expectations of the composite steel plate to be produced, and determine the surface steel plate and the base steel plate according to the usage scenario expectations and usage performance expectations; interactively obtain the design parameters of the composite steel plate to be produced, and refer to the design parameters, the surface steel plate and the base steel plate to perform steel plate call cutting to obtain the surface steel plate to be welded and the base steel plate to be welded; perform surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, and perform welding pretreatment analysis based on the acquisition results to obtain the surface welding joint surface and the base welding joint surface; according to the composite steel plate to be produced The design parameters are called to obtain welding process associated parameters, wherein the welding process associated parameters include surface thickness parameters, base thickness parameters, penetration welding penetration constraints and encapsulation welding penetration constraints; based on the welding process associated parameters, the surface steel plate and the base steel plate, welding process optimization is performed to obtain welding process optimization parameters; after pre-treating the surface welding joint surface and the base welding joint surface, the surface steel plate to be welded and the base steel plate to be welded are stacked based on the surface welding joint surface and the base welding joint surface, and single-sided heterogeneous composite steel plate welding is performed based on the welding process optimization parameters.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: Interactively obtain the usage scenario expectations and usage performance expectations of the composite steel plate to be produced, determine the surface steel plate and the base steel plate; interactively obtain the design parameters of the composite steel plate to be produced, and refer to the design parameters, the surface steel plate and the base steel plate to call and cut the steel plate to obtain the surface steel plate to be welded and the base steel plate to be welded; perform surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, perform welding pretreatment analysis, and obtain the surface welding joint surface and the base welding joint surface; obtain the welding process associated parameters according to the design parameters of the composite steel plate to be produced; perform welding process optimization based on the welding process associated parameters, the surface steel plate and the base steel plate to obtain the welding process optimization parameters; after pre-treating the surface welding joint surface and the base welding joint surface, stack the surface steel plate to be welded and the base steel plate to be welded based on the surface welding joint surface and the base welding joint surface, and perform single-sided heterogeneous composite steel plate welding based on the welding process optimization parameters. The technical effect of improving the welding quality and ensuring the overall performance reliability of the composite steel plate is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of a welding process optimization method for a single-sided heterogeneous composite steel plate provided in an embodiment of the present application; Figure 2 A schematic diagram of the process of obtaining the surface steel plate and the base steel plate in a welding process optimization method for a single-sided heterogeneous composite steel plate provided in an embodiment of the present application; Figure 3 A schematic diagram of the process of obtaining a surface welding joint surface and a base welding joint surface in a welding process optimization method for a single-sided heterogeneous composite steel plate provided in an embodiment of the present application. DETAILED DESCRIPTION

[0010] The present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, which is used to solve the technical problems in the prior art of large fluctuations in welding quality, large residual stress after welding, and impact on the reliability of weldments.

[0011] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application. Embodiment 1

[0012] like Figure 1 As shown, the present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, the method comprising: Step S100: interactively obtaining the usage scenario expectation and usage performance expectation of the composite steel plate to be produced, and determining the surface steel plate and the base steel plate according to the usage scenario expectation and usage performance expectation.

[0013] Specifically, by interacting with relevant parties, the expected usage scenarios and performance expectations of the composite steel plates to be produced are obtained, which involves communication with customers, designers, engineers or other stakeholders to understand the requirements for the composite steel plates in specific applications. The usage scenario expectations include the environment, working conditions and conditions in which the composite steel plates will be used. For example, the composite steel plates may be used in harsh environments such as high temperature, high pressure, corrosion or wear, or need to withstand mechanical loads such as heavy loads, impact or vibration. The performance expectations involve the performance characteristics that the composite steel plates are required to have, such as strength, hardness, toughness, corrosion resistance, wear resistance, thermal conductivity or electrical conductivity, etc.

[0014] According to the usage scenario expectations and performance expectations, select the steel plate material suitable for the surface layer. The surface steel plate usually needs to have good corrosion resistance, wear resistance or other specific properties to meet the requirements of the usage scenario. The base steel plate is the basic part of the composite steel plate and needs to provide strength and support. When determining the surface steel plate and the base steel plate, it is necessary to comprehensively consider the usage scenario expectations and performance expectations to ensure that the selected steel plate combination can meet the overall requirements of the composite steel plate. It is necessary to conduct material performance testing, simulation analysis or refer to previous experience and data to improve the applicability and performance of the composite steel plate to meet the needs of different fields and applications.

[0015] Step S200: interactively obtaining the design parameters of the composite steel plate to be produced, and referring to the design parameters, the surface steel plate and the base steel plate to perform steel plate call cutting to obtain the surface steel plate to be welded and the base steel plate to be welded.

[0016] Specifically, we interacted deeply with the personnel involved in all aspects of the production of composite steel plates in various ways to obtain the most accurate design parameters for the composite steel plates to be produced. The design parameters include length accurate to millimeters, specific values ​​of width, and clear requirements for thickness. They also include complex shape descriptions, such as certain specific angles of bends, special curve radians, etc. After fully obtaining these detailed design parameters, we accurately selected and called the existing steel plate material library, and compared the parameters of each steel plate with the design requirements to ensure that the most suitable steel plate is selected. In the cutting process, a high degree of accuracy and precision is required to ensure that the cut surface steel plates to be welded and the base steel plates to be welded fully meet the requirements of the design parameters in terms of size, shape, etc. Finally, the surface steel plates to be welded and the base steel plates to be welded specially prepared for subsequent welding work are obtained, which can perfectly meet the needs of subsequent welding processes and lay the foundation for the production of high-quality composite steel plates that fully meet design expectations.

[0017] Step S300: performing surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, and performing welding preprocessing analysis based on the acquisition results to obtain the surface welding joint surface and the base welding joint surface.

[0018] Specifically, a special image acquisition device, such as a high-definition industrial camera, is used to collect all-round and detailed surface images of the entire surface of the surface steel plate to be welded and the base steel plate to be welded. The acquisition process needs to ensure the clarity, resolution and accuracy of the image so that reliable data can be obtained for subsequent analysis. The large amount of image data collected is used for in-depth interpretation and analysis of these images using image processing algorithms and techniques. For example, the quality and feasibility of the welding joint surface can be evaluated by identifying features in the image, such as surface unevenness, scratches, rust, oil stains, etc. For the determination of the surface welding joint surface, it is necessary to comprehensively consider the flatness of the steel plate surface, whether there are obvious protrusions or depressions, because this will affect the uniformity and firmness of the welding; the cleanliness of the surface, whether there are impurities, oil stains, etc. that will hinder welding; and possible tiny cracks or defects. Through the analysis and evaluation of these factors, the most suitable surface area for welding, that is, the surface welding joint surface, is selected. Similarly, a similar analysis is also required for the base welding joint surface. Attention should be paid to the degree of matching between the surface of the base steel plate and the surface steel plate, such as the fit between the two, the size of the contact surface, etc. Only by ensuring a good base welding joint surface can a solid foundation be provided for subsequent welding, improving the welding quality and the overall performance reliability of the composite steel plate.

[0019] Step S400: obtaining welding process associated parameters according to the design parameters of the composite steel plate to be produced, wherein the welding process associated parameters include surface thickness parameters, substrate thickness parameters, penetration welding penetration constraints and encapsulation welding penetration constraints.

[0020] Specifically, the design parameters of the composite steel plate to be produced include various detailed requirements and specification information for this steel plate. Through the interpretation and analysis of the design parameters, the corresponding welding process-related parameters can be accurately called out, including surface thickness parameters, base thickness parameters, penetration welding penetration constraints and package welding penetration constraints, ensuring that the welding process is highly matched with the design of the composite steel plate to be produced, which helps to improve the welding quality and stability. Among them, the surface thickness parameter clarifies the thickness specification of the surface steel plate to be welded, which is crucial for the selection and adjustment of the welding process, because different thicknesses require different welding methods and parameter settings; the base thickness parameter specifies the thickness of the base steel plate to be welded, which also affects the heat input, welding speed and other aspects during the welding process; the penetration welding penetration constraint limits and requires the degree of penetration during penetration welding, such as stipulating the minimum or maximum penetration depth to ensure that the firmness and safety of the welding meet the design standards; the package welding penetration constraint constrains the penetration depth during package welding to meet specific packaging requirements and performance standards.

[0021] Step S500: Optimizing the welding process based on the welding process associated parameters, the surface steel plate and the base steel plate to obtain the welding process optimization parameters.

[0022] Specifically, the penetration welding optimization space and the package welding optimization space are pre-constructed, and their vector compositions include the surface thickness vector, the base thickness vector and the penetration welding penetration vector, and the surface thickness vector, the base thickness vector and the package welding penetration vector, respectively. According to the constraint condition of the material composition of the surface steel plate and the base steel plate, the welding process data is called to obtain the sample penetration welding control set and the sample package welding control set, which respectively include multiple groups of sample surface thickness, sample base thickness, sample penetration welding penetration and sample welding process parameters. The sample penetration welding control set and the sample package welding control set are respectively fitted with the penetration welding optimization space and the package welding optimization space to realize the spatial particle point positioning, thereby obtaining the penetration welding particle space and the package welding particle space. With the welding process associated parameters as constraints, in the penetration welding particle space and the package welding particle space, numerical optimization algorithms (such as gradient descent method, Newton method, etc.) and intelligent optimization algorithms (such as genetic algorithm, particle swarm optimization algorithm, etc.) are used to find the welding process parameters that make the objective function optimal through continuous iterative calculation and evaluation, that is, to obtain the welding process optimization parameters.

[0023] Step S600: After pre-treating the surface welding interface and the base welding interface, the surface steel plates to be welded and the base steel plates to be welded are stacked based on the surface welding interface and the base welding interface, and single-sided heterogeneous composite steel plate welding is performed based on the welding process optimization parameters.

[0024] Specifically, the surface welding joint surface and the base welding joint surface are pretreated. For the surface welding joint surface, use appropriate cleaning tools, such as sandblasting equipment or cleaning agents, to thoroughly remove impurities such as oil, rust, oxides, etc. on the surface to make it present a clean metal texture. Use tools such as grinders to finely grind it to remove small protrusions and uneven parts on the surface to improve the flatness of the welding joint. For the base welding joint surface, take the same cleaning and grinding steps to ensure that its surface condition is good. The pretreated surface steel plate to be welded and the base steel plate to be welded are stacked together. During the stacking process, professional fixtures or positioning devices should be used to ensure that the two steel plates are tightly fitted at the welding joint surface without gaps and misalignment. Check the stacking accuracy to ensure the accuracy of the welding position. Single-sided heterogeneous composite steel plate welding is performed according to the previously determined welding process optimization parameters. The welding process optimization parameters play a key guiding role and stipulate the specific current, voltage, welding speed and other parameters during welding to ensure that the welding can proceed smoothly and meet the expected quality standards. Through such orderly preparation and precise welding operations, the quality and reliability of welding can be improved to meet specific welding needs and requirements.

[0025] In a possible implementation, step S100 further includes: Step S110: Aggregate the usage scenario expectations and usage performance expectations to obtain multiple application performance expectations.

[0026] Step S120: interactively obtaining a plurality of sample measurement performance groups of a plurality of sample steel plate raw materials.

[0027] Step S130: Combining and enumerating the plurality of sample measurement performance groups based on the composite constraint, and traversing the plurality of application performance expectations based on the combination result, to screen and obtain a plurality of steel plate raw material combinations.

[0028] Step S140: performing welding adaptability analysis on the plurality of steel plate raw material combinations to obtain the surface steel plate and the base steel plate.

[0029] Specifically, the use scenario expectations comprehensively consider the specific conditions of various actual uses, including expectations such as the environment in which the steel plate will be used, whether it is a high temperature, low temperature, humid or dry environment; how much pressure, tension or impact will it withstand; whether it needs to have specific properties such as wear resistance and corrosion resistance. In-depth research is also conducted on the use performance expectations, including the requirements for the mechanical properties of the steel plate such as strength, hardness, toughness, ductility, and other special performance requirements involved, such as electromagnetic properties and thermal conductivity. Through specific algorithms (such as weighted calculations, weighted calculations are performed according to the importance of different factors), the use scenario expectations and use performance expectations are integrated, summarized and integrated to generate a variety of specific application performance expectations. These application performance expectations are refinements and specificizations of the use scenarios and performance requirements, such as specific application performance expectations such as having at least X strength and Y corrosion resistance under specific wind and climate conditions, and having a specific combination of hardness and toughness under high-load motion conditions. This provides a clear direction and standard for subsequent material selection and analysis, so that steel plate raw materials that meet the requirements can be found more accurately.

[0030] Interactive acquisition means obtaining relevant information by interacting with various data sources, including data exchange and communication with material databases, supplier systems, laboratory testing equipment, etc. By interacting with these data sources, detailed performance data of various sample steel plate raw materials are collected, and these data are organized into multiple sample measurement performance groups. Each sample measurement performance group contains a series of performance indicators of a specific sample steel plate raw material, such as chemical composition, mechanical properties (such as yield strength, tensile strength, elongation, etc.), physical properties (such as density, thermal conductivity, etc.), corrosion resistance, etc. The purpose of this is to build a rich material performance database so that it can be screened and compared in these sample measurement performance groups according to specific needs and constraints in the future to find the steel plate raw material that best meets the requirements.

[0031] Composite constraints are a multi-dimensional set of constraints, including economic constraints, such as budget constraints; technical constraints, such as specific processing technology requirements, size specification restrictions; and specific performance limitations, such as specific ranges of certain key performance indicators. Combining and enumerating multiple sample measurement performance groups is to comprehensively consider every possible sample combination. For example, if there are three sample measurement performance groups A, B, and C, then the combination enumeration will include various possible combinations such as AB, AC, BC, and ABC. Based on these combination results, multiple application performance expectations are traversed. This means that the specific performance of each combination is compared and matched with each requirement in the application performance expectation one by one. Only when a combination can meet or exceed the requirements in all important application performance expectations and meet the conditions in the composite constraints, it will be screened out as a qualified steel plate raw material combination. For example, assume that there are sample measurement performance groups A, B, C, and D. Through combination enumeration, multiple combinations such as AB, AC, AD, BC, BD, CD, etc. may be obtained. By comparing the application performance expectations, such as strength requirements, corrosion resistance requirements, etc., and considering the composite constraints such as the cost cannot exceed a certain value, qualified steel plate raw material combinations such as AC and BD are screened out. This is like accurately finding the "optimal solution" combination that meets both performance requirements and other constraints in a huge space of possibilities. Through this process, it can be ensured that the final selected steel plate raw material combination is the most suitable for the specific application scenario and needs.

[0032] It is crucial to conduct welding adaptability analysis, which mainly evaluates the various performances and feasibility of these steel plate raw material combinations during the welding process, including considering the difficulty of the welding process, such as some combinations may be more suitable for specific welding methods, with fewer defects during welding and higher quality of welded joints; it also includes the mechanical properties after welding, whether the original key performance indicators such as strength and toughness can be maintained; and the deformation, residual stress, etc. that may occur during welding. Through a comprehensive and in-depth welding adaptability analysis of multiple steel plate raw material combinations, the most suitable combination of surface steel plate and base steel plate for welding can be selected from these combinations. This combination must not only perform well in the welding process, ensure welding quality and reliability, but also meet the performance requirements of the entire structure or product. For example, after analysis, it is found that combination X has a simple process, high welded joint strength, and small deformation during welding, and can well meet the performance requirements of the surface and base, then combination X is determined as the final surface steel plate and base steel plate. This can provide clear guidance for subsequent actual welding production to ensure that the products produced have excellent quality and performance in welding.

[0033] In one possible implementation, Figure 2 , step S140 further includes: Step S141: based on the multiple sample measurement performance groups, obtain the first sample measurement performance group and the second sample measurement performance group of the first candidate steel plate and the second candidate steel plate in the first steel plate raw material combination, wherein the first steel plate raw material combination is any one of the multiple steel plate raw material combinations.

[0034] Step S142: preset welding related indicators, and based on the welding related indicators, obtain the first related parameter group and the second related parameter group from the first sample measurement performance group and the second sample measurement performance group, wherein the welding related indicators include melting point parameters, thermal expansion coefficient, and thermal conductivity parameters.

[0035] Step S143: interactively obtain the first raw material cost and the second raw material cost of the first candidate steel plate and the second candidate steel plate.

[0036] Step S144: introducing the first associated parameter group, the second associated parameter group, the first raw material cost and the second raw material cost into a pre-constructed welding adaptability evaluation function to obtain a first welding adaptability coefficient.

[0037] Step S145: Similarly, welding adaptability analysis is performed on the plurality of steel plate raw material combinations to obtain a plurality of welding adaptability coefficients.

[0038] Step S146: Serialize the plurality of welding adaptability coefficients, and obtain the surface steel plate and the base steel plate by combining the plurality of steel plate raw materials based on the sequence extreme values.

[0039] Specifically, the multiple sample measurement performance groups are a huge database that contains detailed performance information of various steel plates, and the first steel plate raw material combination is a specific combination randomly selected from the numerous steel plate raw material combinations. The performance data corresponding to the first candidate steel plate in the first steel plate raw material combination is extracted from this large database, which is the first sample measurement performance group. At the same time, the performance data corresponding to the second candidate steel plate in the first steel plate raw material combination is also found, which is the second sample measurement performance group.

[0040] Welding-related indicators are preset, including melting point parameters, thermal expansion coefficients, and thermal conductivity parameters. Based on these welding-related indicators, relevant parameters are called from the first sample measurement performance group and the second sample measurement performance group to obtain the first and second associated parameter groups, respectively. Among them, the melting point parameter represents the melting temperature of the steel plate, which plays an important role in the fusion and solidification during the welding process; the thermal expansion coefficient reflects the dimensional change of the steel plate during the heating and cooling process, which has an impact on the residual stress and deformation of the welded joint; the thermal conductivity parameter represents the thermal conductivity of the steel plate, which affects the heat distribution and cooling rate during the welding process. By calling these parameters associated with welding, the performance of the first candidate steel plate and the second candidate steel plate can be more comprehensively understood, providing a basis for the subsequent welding adaptability evaluation.

[0041] The raw material cost data of the first candidate steel plate and the second candidate steel plate are obtained from each other through various channels and methods. For example, the historical purchase prices of the two candidate steel plates can be obtained by querying the raw material procurement database of the enterprise, and used as their raw material costs, or the current market quotations can be obtained by communicating with suppliers as the raw material costs, or the cost records of the two steel plates in previous similar projects can be used to determine the raw material costs.

[0042] The first associated parameter group, the second associated parameter group, the first raw material cost, and the second raw material cost obtained previously are introduced into a pre-built welding adaptability evaluation function. This function will calculate a first welding adaptability coefficient based on the input parameters and cost data. This function will comprehensively consider factors such as welding associated parameters and raw material costs to evaluate the adaptability and cost-effectiveness of the first candidate steel plate and the second candidate steel plate in the welding process. By calculating the first welding adaptability coefficient, a value can be obtained to represent the welding adaptability of the first candidate steel plate and the second candidate steel plate. A higher welding adaptability coefficient may indicate that the two steel plates have a better overall performance in terms of welding performance and cost, while a lower welding adaptability coefficient may indicate that there are some problems or that other factors need to be further considered.

[0043] For each steel plate raw material combination, welding adaptability analysis is performed, that is, for each candidate steel plate in the combination, its welding associated parameter group, raw material cost and other information are obtained, and then introduced into the welding adaptability evaluation function to calculate the corresponding welding adaptability coefficient. In this way, multiple welding adaptability coefficients corresponding to multiple different combinations will be obtained. This step ensures that a comprehensive welding adaptability analysis is performed on all possible steel plate raw material combinations without omission.

[0044] Arrange the obtained multiple welding adaptability coefficients in a certain order to form a sequence. The sequence can be sorted by numerical value, etc. According to the extreme value (maximum value) in the sequence, the combination that best meets the requirements is called out from the numerous steel plate raw material combinations, and the steel plates are respectively determined as surface steel plates and base steel plates. In this way, the steel plate combination with the best welding adaptability can be selected from many possible combinations.

[0045] In a possible implementation, step S144 further includes: .

[0046] in, is the welding adaptation coefficient, is the first raw material cost, is the cost of the second raw material, is the first melting point, is the second melting point, is the difference in thermal expansion coefficient, is the first thermal conductivity, is the second thermal conductivity, is the thermal conductivity weight adjustment factor.

[0047] Specifically, The welding adaptability coefficient is the final result to be calculated. It is used to measure the adaptability of a certain steel plate combination in welding. The larger the calculated result, the higher the adaptability. The first raw material cost refers to the raw material cost of the first candidate steel plate, reflecting the impact of the cost factors of the steel plate on the welding adaptability; is the second raw material cost, and similarly, is the raw material cost of the second alternative steel plate; The first melting point is the melting point parameter of the first candidate steel plate, and the melting point will have an important influence on the welding process and welding quality; The second melting point is a melting point parameter of the second candidate steel plate; The difference in thermal expansion coefficients indicates the difference in thermal expansion coefficients between the two alternative steel plates. Inconsistent thermal expansion characteristics may cause stress and other problems after welding. The first thermal conductivity is the thermal conductivity of the first candidate steel plate, which will affect the heat conduction during welding; The second thermal conductivity is the thermal conductivity of the second candidate steel plate; The thermal conductivity weight adjustment factor is a factor used to adjust the weight of thermal conductivity in the welding suitability evaluation. Its specific value is determined according to actual needs and the characteristics of the welding process to more accurately reflect the influence of thermal conductivity on welding suitability. By inputting these specific parameters into the welding suitability evaluation function, the influence of various factors on welding suitability can be comprehensively considered, thereby obtaining a coefficient that can more comprehensively reflect welding suitability. When selecting a steel plate combination, decisions are made based on the size of the welding suitability coefficient to ensure that the selected combination has good performance in terms of welding performance and cost.

[0048] In one possible implementation, Figure 3 , step S300 further includes: Step S310: performing surface image acquisition on the surface steel plate to be welded to obtain a first surface steel plate image and a second surface steel plate image.

[0049] Step S320: performing surface image acquisition on the base steel plate to be welded to obtain a first base steel plate image and a second base steel plate image.

[0050] Step S330: pre-constructing a steel plate defect evaluation network.

[0051] Step S340: Synchronize the first surface steel plate image and the second surface steel plate image to the steel plate defect evaluation network to obtain the first surface defect complexity and the second surface defect complexity.

[0052] Step S350: comparing the first surface defect complexity and the second surface defect complexity to determine the surface welding interface.

[0053] Step S360: Synchronize the first base steel plate image and the second base steel plate image to the steel plate defect evaluation network to obtain the first base defect complexity and the second base defect complexity.

[0054] Step S370: Compare the first substrate defect complexity and the second substrate defect complexity to determine the substrate welding interface.

[0055] Specifically, two independent surface image acquisition operations are performed on the surface steel plate to be welded. The surface of the surface steel plate to be welded is photographed using a suitable image acquisition device (such as an industrial camera, etc.). The first surface steel plate image is obtained by the first shooting. This image records the surface characteristics of the surface steel plate at a specific time and angle, including surface texture, color, possible defects (such as scratches, dents, rust, etc.) and other surface details. Then, the second shooting is performed to obtain the second surface steel plate image. These two shots are performed at different time points or from different angles in order to obtain more comprehensive and multi-angle steel plate surface information, improve the accuracy and reliability of the steel plate surface feature evaluation, and make more reasonable welding decisions and operations.

[0056] The same image acquisition method is used to operate on the base steel plate to be welded. Through a specific image acquisition device, such as a high-resolution camera, the appearance of the base steel plate to be welded is photographed. The first shot results in the first base steel plate image, which intuitively presents the surface condition of the base steel plate at this time, covering information such as surface flatness, whether there are rust spots, wear marks or other abnormal features, and then a second shot is taken to obtain the second base steel plate image. These two image acquisitions of the base steel plate to be welded are carried out under different conditions, such as different lighting environments or different shooting angles. The purpose of this is to capture the surface characteristics of the base steel plate more comprehensively and from multiple angles. Subsequently, these images will be input into relevant analysis systems or programs for further evaluation of the status of the base steel plate, such as analyzing the degree and distribution of its surface defects, judging the area suitable for welding, etc.

[0057] A network specifically used for evaluating steel plate defects is pre-built. The network can be based on deep learning technology, such as convolutional neural networks, to collect a large number of steel plate image samples, including images of steel plates with and without defects, and annotate these images to indicate the type and location of defects, and select a neural network architecture suitable for evaluating steel plate defects, such as using multiple convolutional layers and pooling layers to extract image features, connecting fully connected layers for classification or regression, and using prepared data sets to train the network. The weights and biases of the network are adjusted through the back-propagation algorithm so that the network can accurately identify and evaluate steel plate defects.

[0058] The collected first surface steel plate images and second surface steel plate images are input into the pre-built steel plate defect evaluation network. The network will process and analyze these images and output corresponding defect complexity values, which are called the first surface defect complexity and the second surface defect complexity respectively. These complexity values ​​can reflect the degree and complexity of the surface defects of the steel plate.

[0059] Compare the complexity of the first surface defect and the complexity of the second surface defect. If the complexity of the first surface defect is lower, it means that its surface defects are relatively few or less complex, and it may be more suitable as the surface welding interface. On the contrary, if the complexity of the second surface defect is lower, the surface corresponding to the second surface steel plate image is selected as the surface welding interface.

[0060] The first substrate steel plate image and the second substrate steel plate image are input into the steel plate defect evaluation network. The network will process and analyze these images and output corresponding defect complexity values, which are respectively called the first substrate defect complexity and the second substrate defect complexity.

[0061] Compare the complexity of the first substrate defect and the complexity of the second substrate defect, and select the surface with lower complexity as the substrate welding joint surface to ensure the quality and reliability of the welding joint surface. Through the above steps, the steel plate defect evaluation network can be used to analyze the surface images of the surface and substrate steel plates to be welded, determine the joint surface that is relatively more suitable for welding, and thus improve the welding quality and efficiency.

[0062] In a possible implementation, step S330 further includes: Step S331: performing apparent defect type aggregation on the surface steel plate and the base steel plate to obtain a plurality of steel plate defect types, and constructing a plurality of steel plate defect recognition branches based on the plurality of steel plate defect types.

[0063] Step S332: performing defect image call based on the surface steel plate and the base steel plate to obtain multiple sample defect images.

[0064] Step S333: performing defect identification evaluation on the plurality of sample defect images based on the plurality of steel plate defect types to obtain a plurality of sample defect areas and a plurality of sample defect divergences.

[0065] Step S334: using the plurality of sample defect images, the plurality of sample defect areas and the plurality of sample defect divergences to train the plurality of steel plate defect recognition branches.

[0066] Step S335: The multiple steel plate defect recognition branches that have been trained are connected in parallel to obtain the steel plate defect evaluation network.

[0067] Specifically, the aggregation of apparent defect types for surface steel plates and base steel plates is to classify and organize all visible defect manifestations that may appear on these two types of steel plates, including but not limited to surface scratches, pits, protrusions, cracks, oxidation spots and other various appearance defects. By carefully observing and analyzing a large number of surface steel plate and base steel plate samples, defects with similar characteristics are classified into the same type, and a variety of steel plate defect types are obtained, such as scratches, dents, rust spots and other different specific defect types. Constructing multiple steel plate defect recognition branches based on multiple steel plate defect types means that each branch is designed and trained specifically for a specific type of steel plate defect. The advantage of this is that the network can handle each specific type of defect more focused and finely, thereby improving the accuracy and efficiency of recognition. For example, there is a branch specifically used to identify scratch defects and learn various features and patterns of scratches so that scratch defects can be accurately detected and distinguished in the input image. By constructing multiple such branches, the entire network can handle different types of steel plate defects comprehensively and specifically, laying the foundation for subsequent more accurate defect evaluation and analysis.

[0068] A large number of images containing various defect conditions are screened and extracted from an image database or storage area related to the surface steel plate and the base steel plate. These images are called multiple sample defect images. The sample defect images cover defect manifestations of different degrees and types.

[0069] According to the previously determined types of steel plate defects, such as scratches, dents, rust spots, etc., each sample defect image is carefully analyzed and judged based on these determined types. Among them, obtaining multiple sample defect areas means calculating the area occupied by each specific defect type in the image. For example, how large is the area of ​​a scratch defect in a certain image, or how much area is occupied by a dent defect, etc.; obtaining multiple sample defect divergences, where the divergence can be understood as the degree of discreteness or distribution range of the defect, for example, whether a certain defect is concentrated in a small area or distributed relatively dispersedly in a larger range, which helps to more comprehensively understand the characteristics and impact of the defect.

[0070] Multiple sample defect images serve as the basis of training data. They contain various forms of steel plate defects. These images will be input into each steel plate defect recognition branch. Multiple sample defect areas and multiple sample defect divergences, together with sample defect images, provide a more comprehensive and detailed description of defect features. Each steel plate defect recognition branch will process and analyze the input sample defect images and related area and divergence data. Through a large number of training iterations, these branches gradually learn the characteristics of different types of defects, the relationship between area and divergence, and how to accurately identify and evaluate them. For example, a branch is specifically responsible for identifying scratch defects. It will gradually master the shape, size, distribution characteristics of scratches, etc., as well as the correlation between these features and area and divergence by learning a large number of sample images with scratch defects and the corresponding area and divergence data. After sufficient training, each steel plate defect recognition branch can effectively identify and evaluate specific types of defects, thereby improving the accuracy and reliability of the entire steel plate defect evaluation network. When new steel plate images are input, these well-trained branches can quickly and accurately determine the defect type and related feature information.

[0071] After completing the parallel training of multiple steel plate defect recognition branches, each branch calculates the distribution divergence and total defect area of ​​a defect, and combines them to form a complete steel plate defect evaluation network. This network can comprehensively utilize the recognition capabilities of each branch to conduct a comprehensive defect evaluation on the input steel plate image. By fusing the results of multiple branches, the accuracy and reliability of the evaluation can be improved. The first surface steel plate image and the second surface steel plate image are synchronized to the steel plate defect evaluation network to obtain the first surface defect complexity and the second surface defect complexity. By comparing the defect complexity, the most suitable surface welding interface can be determined; the first base steel plate image and the second base steel plate image are synchronized to the steel plate defect evaluation network to obtain the first base defect complexity and the second base defect complexity. By comparing the defect complexity, the most suitable base welding interface can be determined.

[0072] In a possible implementation, step S500 further includes: Step S510: Pre-constructing a penetration welding optimization space and a package welding optimization space, wherein the vector composition of the penetration welding optimization space includes a surface thickness vector, a base thickness vector and a penetration welding penetration vector, and the vector composition of the package welding optimization space includes a surface thickness vector, a base thickness vector and a package welding penetration vector.

[0073] Step S520: Execute welding process data call with the material composition of the surface steel plate and the base steel plate as constraints to obtain a sample penetration welding control set and a sample encapsulation welding control set, wherein the sample penetration welding control set includes multiple groups of sample surface thickness-sample base thickness-sample penetration welding penetration-sample welding process parameters, and the sample encapsulation welding control set includes multiple groups of sample surface thickness-sample base thickness-sample encapsulation welding penetration-sample welding process parameters.

[0074] Step S530: fitting the sample penetration welding control set and the sample encapsulation welding control set to the penetration welding optimization space and the encapsulation welding optimization space respectively to perform spatial particle point positioning to obtain a penetration welding particle space and an encapsulation welding particle space.

[0075] Step S540: Taking the welding process associated parameters as constraints, performing welding process optimization in the penetration welding particle space and the encapsulation welding particle space to obtain the welding process optimization parameters.

[0076] Specifically, pre-constructing the penetration welding optimization space and the package welding optimization space means building these two specific space structures before formally conducting the relevant welding process analysis or optimization. For the penetration welding optimization space, the surface thickness vector reflects the thickness information of the surface layer of the welded part, which has an important influence on the effect of penetration welding. The base thickness vector reflects the thickness of the base part of the welding object, which is also related to the performance of welding. The penetration welding penetration vector is directly related to the key indicator of penetration welding, the penetration depth, which can measure the depth of welding. Similarly, the surface thickness vector and base thickness vector in the package welding optimization space have similar meanings to those in the penetration welding optimization space, while the package welding penetration vector focuses on describing the characteristics of the penetration depth in package welding. In this space, various welding parameters and conditions can be analyzed, simulated and optimized to find the best welding solution and improve welding quality and efficiency.

[0077] Taking the material composition of the surface steel plate and the base steel plate as constraints means that when calling the welding process data, the specific material composition and properties of the two steel plates must be taken into account. Different material compositions may have different effects on the welding process, such as differences in melting point, thermal conductivity, etc. Executing the welding process data call means obtaining data related to the welding process from the relevant database or storage. The sample penetration welding control set contains multiple groups of data, each of which consists of sample surface thickness-sample base thickness-sample penetration welding penetration-sample welding process parameters. The sample surface thickness and sample base thickness describe the thickness characteristics of the steel plate, the sample penetration welding penetration is related to the key effect of penetration welding, and the sample welding process parameters cover specific process setting information such as current, voltage, welding speed, etc. during the welding process. Similarly, the sample package welding control set also has a similar structure, but it is aimed at package welding. By obtaining such a detailed sample control set, it can provide a rich data basis for the subsequent analysis, research and optimization of different welding process conditions and parameters. Use these data to explore the relationship between different factors and find the optimal welding process configuration to achieve better welding quality and effect.

[0078] Fitting the sample penetration welding control set and the sample encapsulation welding control set to the penetration welding optimization space and the encapsulation welding optimization space respectively means that the previously obtained sample control set containing various specific parameters corresponds to the pre-constructed optimization space composed of specific vectors. Through this fitting, the sample data can be placed at a specific position in the optimization space. Positioning the spatial particle points is equivalent to determining the particle points corresponding to each sample data in the optimization space. These particle points represent different welding process states and parameter combinations, and obtain the penetration welding particle space and the encapsulation welding particle space. These two particle spaces are the sets composed of the particle points corresponding to all the fitted sample data. The significance of this is that a large amount of complex welding process data is converted into an intuitive representation in a specific optimization space, so that the relationship between different welding process parameters and their influence on the welding effect can be more clearly observed and analyzed.

[0079] Taking welding process associated parameters as constraints, the welding process associated parameters here can include a series of factors that have an important impact on welding quality and effect, such as welding current, voltage, welding speed, gas flow, etc. Taking these parameters as constraints means that in the process of welding process optimization, the range specified by these parameters cannot be exceeded to ensure that the optimization result is within the practical and reasonable process range. Performing welding process optimization in the penetration welding particle space and the encapsulation welding particle space is to perform optimization operations in the particle space that has been constructed and contains a large amount of welding process state information. This optimization can adopt various optimization algorithms or methods, such as genetic algorithms, simulated annealing algorithms, etc. Its purpose is to find the optimal welding process combination under the premise of meeting the constraints of welding process associated parameters, that is, the process parameters that can make the welding quality reach the best or meet specific requirements. Finally, the welding process optimization parameters are obtained. These optimization parameters are the specific values ​​of the best welding current, voltage, welding speed, etc. determined after the optimization process. These parameters will provide guidance for the actual welding operation. Welding according to these parameters can achieve better welding effects and quality. Through such a welding process optimization process, the efficiency and reliability of welding can be improved.

[0080] In a possible implementation, step S540 further includes: Step S541: the penetration welding particle space includes a plurality of sample penetration welding particle points, and each sample penetration welding particle point is identified by a sample welding process parameter.

[0081] Step S542: positioning virtual particle points in the penetration welding particle space according to the surface thickness parameter, base thickness parameter, and penetration welding penetration constraint in the welding process associated parameters to obtain penetration welding virtual points.

[0082] Step S543: preset a screening radius, and perform spatial particle point extraction based on the penetration welding virtual point and the screening radius to obtain H sample penetration welding particle points.

[0083] Step S544: performing weighted calculation of the H sample welding process parameters identified by the H sample penetration welding particle points according to the distances between the H sample penetration welding particle points and the penetration welding virtual point to obtain the penetration welding process parameters.

[0084] Step S545: Similarly, taking the welding process associated parameters as constraints, performing welding process optimization in the packaging welding particle space to obtain packaging welding process parameters, the penetration welding process parameters and the packaging welding process parameters constitute the welding process optimization parameters.

[0085] Specifically, within the specific spatial domain of the penetration welding particle space, there are a large number of sample penetration welding particle points. Each particle point does not exist in isolation, but carries the corresponding sample welding process parameters as its identification. The sample welding process parameters include welding current, voltage, welding speed, welding time and other key information. Through this identification method, each particle point represents a specific penetration welding process state.

[0086] According to the specific constraints of surface thickness parameters, base thickness parameters and penetration welding penetration contained in the welding process association parameters, the particle points in the penetration welding particle space are compared and analyzed to find a position that best meets these constraints, thereby determining the penetration welding virtual point. The purpose of this is to focus on a point in the particle space that is most likely to meet specific welding requirements.

[0087] A specific screening radius value is set in advance, and then, with the determined penetration welding virtual point as the center, in the penetration welding particle space, the sample penetration welding particle points within the radius are extracted according to this screening radius. The number of sample penetration welding particle points extracted in this way is H (the value range of H is a positive integer, such as 10, 50, 100). The significance of this is that by setting the screening radius, the particle points associated with the penetration welding virtual point within a certain spatial range are selected in a targeted manner. These extracted particle points may have similar characteristics or close welding process parameters to the virtual point. For example, if the screening radius is small, the extracted particle points will be more closely around the virtual point, which may represent very similar welding process conditions; if the screening radius is large, it will cover a wider range of particle points, which may include some welding process states that are slightly farther away but still have a certain association.

[0088] For the H extracted sample penetration welding particle points, the distances between them and the penetration welding virtual point are calculated respectively. Based on these distances, the H sample welding process parameters identified by the H sample penetration welding particle points are weightedly calculated. The closer the particle point is to the virtual point, the greater the weight of the corresponding sample welding process parameter in the weighted calculation. In this way, the correlation between each sample particle point and the virtual point can be comprehensively considered, so as to obtain a penetration welding process parameter that can better reflect the overall situation.

[0089] In the same way and logic, with welding process-related parameters (such as various specific conditions and requirements related to welding) as constraints, similar operations are performed in the specific space of the package welding particle space, that is, the optimal welding process parameters are found through continuous analysis, comparison and calculation in the package welding particle space. This process is called welding process optimization. The package welding process parameters finally obtained through this optimization operation can meet the requirements and constraints of package welding to the greatest extent. At the same time, the combination of the penetration welding process parameters obtained previously and the newly obtained package welding process parameters here constitutes a complete welding process optimization parameter. Together with the penetration welding process parameters, it provides comprehensive and optimized process guidance for the entire welding process to ensure that the welding quality and effect are in the best state.

[0090] In summary, the embodiments of the present application have at least the following technical effects: The present application provides a method for optimizing the welding process of a single-sided heterogeneous composite steel plate, which interactively obtains the usage scenario expectation and usage performance expectation of the composite steel plate to be produced, determines the surface steel plate and the base steel plate; interactively obtains the design parameters of the composite steel plate to be produced, and refers to the design parameters, the surface steel plate and the base steel plate to perform steel plate call cutting to obtain the surface steel plate to be welded and the base steel plate to be welded; performs surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, performs welding pretreatment analysis, and obtains the surface welding joint surface and the base welding joint surface; obtains welding process associated parameters according to the design parameter call of the composite steel plate to be produced; performs welding process optimization based on the welding process associated parameters, the surface steel plate and the base steel plate to obtain welding process optimization parameters; after pre-treating the surface welding joint surface and the base welding joint surface, the surface steel plate to be welded and the base steel plate to be welded are stacked based on the surface welding joint surface and the base welding joint surface, and the single-sided heterogeneous composite steel plate welding is performed based on the welding process optimization parameters. The technical effect of improving the welding quality and ensuring the overall performance reliability of the composite steel plate is achieved.

[0091] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0092] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0093] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A welding process optimization method for a single-sided heterogeneous composite steel plate, characterized in that: The method comprises: Interactively obtaining usage scenario expectations and usage performance expectations of the composite steel plate to be produced, and determining the surface steel plate and the base steel plate according to the usage scenario expectations and usage performance expectations; Interactively obtain the design parameters of the composite steel plate to be produced, and refer to the design parameters, the surface steel plate and the base steel plate to perform steel plate call cutting to obtain the surface steel plate to be welded and the base steel plate to be welded; Performing surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, and performing welding preprocessing analysis based on the acquisition results to obtain the surface welding joint surface and the base welding joint surface; Obtaining welding process associated parameters according to the design parameters of the composite steel plate to be produced, wherein the welding process associated parameters include a surface thickness parameter, a substrate thickness parameter, a penetration welding penetration constraint, and an encapsulation welding penetration constraint; Based on the welding process associated parameters, the surface steel plate and the base steel plate, the welding process is optimized to obtain the welding process optimization parameters; After pre-treating the surface welding joint surface and the base welding joint surface, the surface steel plates to be welded and the base steel plates to be welded are stacked based on the surface welding joint surface and the base welding joint surface, and single-sided heterogeneous composite steel plate welding is performed based on the welding process optimization parameters.

2. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 1, characterized in that: Interactively obtaining the usage scenario expectation and usage performance expectation of the composite steel plate to be produced, and determining the surface steel plate and the base steel plate according to the usage scenario expectation and usage performance expectation, the method further includes: Aggregate the usage scenario expectations and usage performance expectations to obtain multiple application performance expectations; Interactively obtain multiple sample measurement performance groups of multiple sample steel plate raw materials; Combining and enumerating the plurality of sample measurement performance groups based on the composite constraint, and traversing the plurality of application performance expectations based on the combination result, to screen and obtain a plurality of steel plate raw material combinations; A welding suitability analysis is performed on the plurality of steel plate raw material combinations to obtain the surface steel plate and the base steel plate.

3. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 2, characterized in that: Performing welding adaptability analysis on the plurality of steel plate raw material combinations to obtain the surface steel plate and the base steel plate, the method further comprising: Based on the multiple sample measurement performance groups, a first sample measurement performance group and a second sample measurement performance group of a first candidate steel plate and a second candidate steel plate in the first steel plate raw material combination are obtained, wherein the first steel plate raw material combination is any one of the multiple steel plate raw material combinations; Preset welding related indicators, and obtain a first related parameter group and a second related parameter group from the first sample measured performance group and the second sample measured performance group based on the welding related indicators, wherein the welding related indicators include a melting point parameter, a thermal expansion coefficient, and a thermal conductivity parameter; Interactively obtaining a first raw material cost and a second raw material cost of the first candidate steel plate and the second candidate steel plate; Introducing the first associated parameter group, the second associated parameter group, the first raw material cost and the second raw material cost into a pre-constructed welding adaptability evaluation function to obtain a first welding adaptability coefficient; Similarly, performing welding adaptability analysis on the plurality of steel plate raw material combinations to obtain a plurality of welding adaptability coefficients; The plurality of welding suitability coefficients are serialized, and the surface steel plate and the base steel plate are obtained by combining the plurality of steel plate raw materials based on the sequence extreme values.

4. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 3, characterized in that: The welding suitability evaluation function is as follows: ； in, is the welding adaptation coefficient, is the first raw material cost, is the cost of the second raw material, is the first melting point, is the second melting point, is the difference in thermal expansion coefficient, is the first thermal conductivity, is the second thermal conductivity, is the thermal conductivity weight adjustment factor.

5. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 1, characterized in that: Performing surface image acquisition on the surface steel plate to be welded and the base steel plate to be welded, and performing welding preprocessing analysis based on the acquisition results to obtain the surface welding joint surface and the base welding joint surface, the method also includes: Performing surface image acquisition on the surface steel plate to be welded to obtain a first surface steel plate image and a second surface steel plate image; Performing surface image acquisition on the base steel plate to be welded to obtain a first base steel plate image and a second base steel plate image; Pre-constructed steel plate defect evaluation network; Synchronizing the first surface steel plate image and the second surface steel plate image to the steel plate defect evaluation network to obtain the first surface defect complexity and the second surface defect complexity; Comparing the first surface defect complexity with the second surface defect complexity, determining the surface welding joint surface; Synchronizing the first substrate steel plate image and the second substrate steel plate image to the steel plate defect evaluation network to obtain the first substrate defect complexity and the second substrate defect complexity; The first substrate defect complexity and the second substrate defect complexity are compared to determine the substrate welding interface.

6. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 5, characterized in that: Pre-constructing a steel plate defect evaluation network, the method further comprises: Performing apparent defect type aggregation on the surface steel plate and the base steel plate to obtain multiple steel plate defect types, and constructing multiple steel plate defect recognition branches based on the multiple steel plate defect types; Perform defect image calling based on the surface steel plate and the base steel plate to obtain a plurality of sample defect images; Based on the multiple steel plate defect types, defect identification evaluation is performed on the multiple sample defect images to obtain multiple sample defect areas and multiple sample defect divergences; Using the plurality of sample defect images, the plurality of sample defect areas and the plurality of sample defect divergences to train the plurality of steel plate defect recognition branches; The multiple steel plate defect recognition branches are trained in parallel to obtain the steel plate defect evaluation network.

7. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 1, characterized in that: Based on the welding process associated parameters, the surface steel plate and the base steel plate, the welding process is optimized to obtain the welding process optimization parameters, and the method further includes: Pre-constructing a penetration welding optimization space and a package welding optimization space, wherein the vector composition of the penetration welding optimization space includes a surface thickness vector, a base thickness vector and a penetration welding penetration vector, and the vector composition of the package welding optimization space includes a surface thickness vector, a base thickness vector and a package welding penetration vector; The welding process data is called based on the material composition of the surface steel plate and the base steel plate to obtain a sample penetration welding control set and a sample encapsulation welding control set, wherein the sample penetration welding control set includes multiple groups of sample surface thickness-sample base thickness-sample penetration welding penetration-sample welding process parameters, and the sample encapsulation welding control set includes multiple groups of sample surface thickness-sample base thickness-sample encapsulation welding penetration-sample welding process parameters; The sample penetration welding control set and the sample encapsulation welding control set are respectively fitted to the penetration welding optimization space and the encapsulation welding optimization space to perform spatial particle point positioning, so as to obtain a penetration welding particle space and an encapsulation welding particle space; Taking the welding process associated parameters as constraints, welding process optimization is performed in the penetration welding particle space and the encapsulation welding particle space to obtain the welding process optimization parameters.

8. The method for optimizing the welding process of a single-sided heterogeneous composite steel plate according to claim 7, characterized in that: Taking the welding process associated parameters as constraints, performing welding process optimization in the penetration welding particle space and the encapsulation welding particle space to obtain the welding process optimization parameters, the method further includes: The penetration welding particle space includes a plurality of sample penetration welding particle points, and each sample penetration welding particle point is identified by a sample welding process parameter; Positioning virtual particle points in the penetration welding particle space according to the surface thickness parameter, the base thickness parameter, and the penetration welding penetration constraint in the welding process associated parameters to obtain a penetration welding virtual point; Preset a screening radius, and perform spatial particle point extraction based on the penetration welding virtual point and the screening radius to obtain H sample penetration welding particle points; Performing weighted calculation of H sample welding process parameters identified by the H sample penetration welding particle points according to the distances between the H sample penetration welding particle points and the penetration welding virtual point to obtain the penetration welding process parameters; By analogy, with the welding process associated parameters as constraints, welding process optimization is performed in the package welding particle space to obtain package welding process parameters, and the penetration welding process parameters and package welding process parameters constitute the welding process optimization parameters.

Images