Welding material welding manufacturability comprehensive quantitative evaluation method and system

Abstract

The invention relates to the field of electronic digital data processing, in particular to a comprehensive quantitative evaluation method and system for the welding manufacturability of welding materials, and the method comprises the steps: obtaining the quality data of a plurality of batches of historical welding materials, and calculating a Gini index by taking any item of quality data as a division point, an analytic hierarchy process is used for determining the influence weight of each data item on the welding quality, and then the welding quality index of each batch is calculated; the Gini indexes of the subsets on the two sides of the division point are compared, the segmentation optimization degree is evaluated and combined with the Gini indexes, and an optimized Gini index is formed and used for recursively constructing a decision tree; in the division process, whether division continues or not is determined by evaluating the division termination degree until construction of the decision tree is completed; and the welding quality is classified according to the division result, so that quantitative evaluation is carried out. According to the method, the Gini index is optimized, the advantages and disadvantages of each division point are evaluated, and the classification accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of electronic digital data processing, and more specifically, to a method and system for comprehensively quantifying the welding processability of welding materials. Background Art

[0002] In the field of modern manufacturing and engineering, welding technology is used in industries such as automobiles, aerospace, construction and shipbuilding. With the continuous advancement of industrial technology, the requirements for welding quality and efficiency are increasing. Different welding materials and process parameters not only affect the quality of welded joints, but are also directly related to the performance and safety of the final product. At present, although there are several welding quality evaluation standards and experimental methods, most of them still focus on a single performance indicator, such as mechanical properties or welding defects.

[0003] The existing Chinese patent application document with publication number CN115130892A discloses a comprehensive quantitative evaluation method and system for welding processability of welding materials, which relates to the field of welding. The evaluation method includes: obtaining welding data of target welding materials, calculating the measurement value of welding data corresponding to the target welding materials, the measurement value includes mean and variance; establishing a hierarchical structure model; establishing a priority relationship matrix of each level in the hierarchical structure model and calculating the weight set of each level; determining the corresponding membership function according to the underlying index data of welding processability of target welding materials; the underlying index data is the measurement value of welding data corresponding to the target welding materials.

[0004] In the application document, the target welding material processability quantitative evaluation index is determined by using a weight set and a comment set calculated by a membership function according to a hierarchical structure model. However, it is currently difficult to obtain accurate welding quality evaluation standards in the process of evaluating welding processes based on welding quality. Conventional decision tree algorithms directly classify data related to welding quality of welding materials according to the Gini index. However, in the data affecting welding quality of welding materials, each data has a different degree of influence on welding quality. Directly dividing the data according to the performance of the Gini index of the data is difficult to make a more reasonable distinction between data of different welding qualities, and may even divide data with completely different welding qualities into the same subset, resulting in inaccurate evaluation of welding material welding processes. Summary of the invention

[0005] In order to solve the problem that among the data affecting the welding quality of welding materials, each data has a different degree of influence on the welding quality, it is difficult to make a more reasonable distinction between the data of different welding qualities when dividing them through a conventional decision tree, and it is even possible to divide the data of completely different welding qualities into the same subset, thus causing inaccurate evaluation of the welding process of welding materials, the present invention provides solutions in the following multiple aspects.

[0006] In a first aspect, a comprehensive quantitative evaluation method for welding processability of welding materials includes: obtaining quality data of multiple batches of welding materials in history and preprocessing the data; calculating the Gini index with any quality data in the welding material as a dividing point, obtaining the influence weight of each quality data on the welding quality according to the hierarchical analysis method, and calculating the welding quality index of each batch of welding materials; obtaining the segmentation preference degree of the dividing point based on the Gini index of the data subsets on both sides of the dividing point, using the ratio between the segmentation preference degree of the dividing point and the Gini index as the adjusted optimized Gini index, and recursively constructing a decision tree to obtain the division termination degree of each subset after each division; judging whether to stop the division based on the division termination degree, and completing the decision tree division; classifying the welding quality according to the division result, and quantitatively evaluating the welding processability; the division termination degree Satisfies the following relationship: , where Indicates the number of data contained in the left subset or right subset after segmentation. Indicates the preset minimum value of the subset, The natural index The exponential function with base , Indicates the total number of quality data items, The first The weight of the quality data, Indicates the first The standard deviation of the quality data, Indicates all The standard deviation of the quality data.

[0007] The effect is as follows: by using the Gini index and the weights determined by the hierarchical analysis method, this method can more accurately evaluate and distinguish the quality of different welding batches, which is conducive to identifying quality differences. By introducing the segmentation preference degree and optimizing the Gini index, the decision tree is recursively constructed to ensure that each division point is based on the principle of maximizing and minimizing the distinction, thereby improving the prediction accuracy and generalization ability of the model. According to the degree of division termination, it automatically determines when to stop the division during the decision tree construction process, avoiding the problem of overfitting or underfitting. Through quantitative evaluation, the key factors affecting welding quality can be more clearly identified, providing targeted guidance for process improvement.

[0008] Preferably, the quality data includes: the degree of cracks on the weld surface, the degree of inclusions such as weld pores, the tensile strength of the weld joint, the yield strength of the weld joint, the hardness of the weld joint, the toughness of the weld joint and other data.

[0009] Preferably, the Gini index includes: According to the partition point, the quality data of the target item is divided into two parts: the left subset and the right subset, where the Gini index of the partition point is the weighted average of the Gini indexes of the left subset and the right subset, and the weight is the number of data points in the left subset and the right subset.

[0010] Preferably, the step of obtaining the influence weight of each quality data on the welding quality according to the hierarchical analysis method includes: Taking the welding quality as the evaluation target, a hierarchical structure is established, wherein the hierarchical structure includes: target, criterion and sub-criteria, wherein the target is at the top layer, and the criterion and sub-criteria are unfolded layer by layer until the bottom layer; constructing a judgment matrix according to expert scoring or voting; The judgment matrix is ​​normalized, that is, the elements of each column are divided by the column sum to obtain a normalized matrix, the sum of each row in the normalized matrix is ​​calculated, and normalization is performed so that the sum of each element in the vector is 1 to obtain the influence weight.

[0011] The effect is as follows: by establishing a hierarchical structure, the complex welding quality evaluation problem is decomposed into smaller and more manageable parts, making the decision-making process clearer; by constructing a judgment matrix and performing normalization, the qualitative expert judgment is converted into a quantitative weight value, so that the impact of different criteria and sub-criteria on welding quality can be quantitatively compared; through quantitative evaluation indicators, it is easy to compare and choose between different welding materials or processes, so as to make more reasonable decisions.

[0012] Preferably, the welding quality index includes: The quality data of any batch of welding materials is taken as a marked batch, and the average of the cumulative sum of the products between the target items and the influence weights of the target items in the marked batch is taken as the welding quality index of the marked batch.

[0013] The effect is that by multiplying and summing up the data of each target item with its corresponding influence weight, the overall quality of each batch of welding materials can be quantified to obtain a comprehensive welding quality index, which provides a unified evaluation standard for different batches of welding materials, facilitates comparison between batches, and identifies batches with higher or lower quality.

[0014] Preferably, the segmentation preference degree includes: The quality data of multiple batches of welding materials are obtained, and the quality data of any batch of welding materials is used as a marked batch, and the quality data of any item in the marked batch of welding materials is used as a target item. The difference between the target item and the minimum value of the same target item in multiple batches is divided by the range of the target item, and the absolute value after deducting one half is used as the absolute deviation, which is used to represent the position of the target item relative to the median, and the absolute deviation is attenuated using exponential decay; The difference between the average values ​​of the welding quality index of the two subsets on the left and right of the target item as the dividing point is calculated, and then divided by the range and squared to obtain the welding quality difference. The product of the attenuated absolute deviation and the welding quality difference is used as the segmentation optimization degree of the target item.

[0015] The effect is that by optimizing the selection of partition points, the model's ability to generalize new data can be improved and the risk of overfitting can be reduced. At the same time, multiple factors such as data distribution, deviation and quality differences can be comprehensively considered when constructing the decision tree, thereby evaluating the welding quality more comprehensively.

[0016] Preferably, judging whether to stop the division based on the division termination degree to complete the decision tree division includes: The decision tree is divided into two subsets according to the minimum value of the optimized Gini index, and the termination degree of the division of the two subsets is calculated respectively; In response to the termination degree of the division corresponding to the two subsets being greater than the termination threshold, the two subsets are marked as final subsets and no longer divided. Otherwise, other subsets are selected to continue the division until the termination degree of the division in all subsets is greater than the termination threshold, and the division of the decision tree is completed.

[0017] The effect is that the division is based on the minimum value of the optimized Gini index, which reduces human intervention. By setting the termination threshold, the depth and complexity of the decision tree can be controlled to avoid overfitting. By stopping the division after reaching a certain termination level, computing resources can be allocated more efficiently to avoid unnecessary calculations on subsets that are already pure enough.

[0018] Preferably, the quantitative evaluation of welding processability includes: The quality data is classified according to the decision tree to obtain multiple clusters. The average value of the welding quality index of each data in each cluster is taken as the welding quality of the corresponding cluster. The welding qualities of all clusters are sorted from large to small and numbered. The quality data of the welding material in the cluster with the largest welding quality is recorded as level one, and the other clusters are assigned welding quality levels according to their numbers to complete the quantitative scoring of the welding quality.

[0019] In a second aspect, a system for comprehensive quantitative evaluation of welding processability of welding materials includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned method for comprehensive quantitative evaluation of welding processability of welding materials is implemented.

[0020] The present invention has the following effects: 1. The present invention calculates the segmentation preference of the partition point and combines it with the Gini index to form an adjusted optimized Gini index, so that data can be partitioned more reasonably. It is beneficial to more accurately evaluate the pros and cons of each partition point in the process of recursively constructing the decision tree, thereby optimizing the structure of the decision tree and improving the classification effect.

[0021] 2. The present invention determines the influence weight of each welding quality data through the hierarchical analysis method, so that the welding quality index more accurately reflects the actual quality of the welding material, thereby distinguishing different welding quality data that are difficult to reasonably distinguish by conventional decision trees, avoiding erroneous classification of data with completely different qualities into the same subset, thereby improving the accuracy of welding process evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 It is a method flow chart of steps S1 to S4 in a method for comprehensive quantitative evaluation of welding processability of welding materials in an embodiment of the present invention.

[0023] Figure 2 This is a decision tree construction process in a comprehensive quantitative evaluation method for welding processability of welding materials in an embodiment of the present invention.

[0024] Figure 3 It is a structural block diagram of a comprehensive quantitative evaluation system for welding processability of welding materials according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0027] Reference Figure 1 A comprehensive quantitative evaluation method for welding processability of welding materials includes steps S1 to S4, which are as follows: S1: Obtain quality data of multiple batches of welding materials in history and pre-process the data.

[0028] Among them, quality data include but are not limited to: the degree of cracks on the weld surface, the degree of inclusions such as weld pores, the tensile strength of the weld joint, the yield strength of the weld joint, the hardness of the weld joint, the toughness of the weld joint and other data.

[0029] It should be noted that for welding quality data, it is necessary to evaluate the welding quality through various quality-related data, so it is necessary to collect various data on the welding joints.

[0030] Specifically, the degree of cracks on the surface of the welding weld is obtained through image processing; the degree of inclusions such as pores inside the weld is obtained through ultrasonic testing; the tensile strength and yield strength of the weld joint are obtained through tensile testing; the hardness of the weld joint is obtained through a Brinell hardness tester; and the toughness of the weld joint is obtained through a pendulum impact tester.

[0031] The acquired quality data is preprocessed, such as cleaning and normalizing the data. This method is well known to those skilled in the art and will not be described in detail.

[0032] Further analysis shows that for any batch of welding material quality data, in the process of classifying them through the decision tree, the Gini index of each division point is calculated, and the data is divided according to the Gini index. Therefore, the Gini index of each division point needs to be calculated first. The specific steps are as follows: S2: The Gini index is calculated by taking any quality data of the welding material as the dividing point, and the influence weight of each quality data on the welding quality is obtained according to the hierarchical analysis method, and the welding quality index of each batch of welding materials is calculated.

[0033] It should be noted that the decision tree is divided for the quality data of the collected welding materials. The classification prevents too many selected division points from causing excessive calculation. The data collected from each item of data can be selected as the division point for each division, which reduces the division calculation amount while ensuring the accuracy of the division.

[0034] Gini index, including: According to the partition point, the quality data of the target item is divided into two parts: the left subset and the right subset, where the Gini index of the partition point is the weighted average of the Gini indexes of the left subset and the right subset, and the weight is the number of data points in the left subset and the right subset.

[0035] Specifically, the Gini index satisfies the following relationship: ; In the formula, Indicates The Gini index when the quality data is used as the dividing point, Indicates The number of data points in the left subset when the item quality data is used as the partition point, Indicates The number of data points in the right subset when the item quality data is used as the dividing point, represents the total number of data points in the entire dataset, represents the Gini index of the left subset, represents the Gini index of the right subset.

[0036] For example, take the quality data of any batch of welding materials as an example, a data set containing 100 quality data, of which 50 quality data belong to the positive class and 50 quality data belong to the negative class. We choose the 50th quality data as the dividing point; There are 30 positive classes and 20 negative classes in the left subset, and 20 positive classes and 30 negative classes in the right subset. The Gini index of the left subset is ; Then the Gini index of the right subset is ; The Gini index of the dividing point is: ; Then iterate the above calculation method and select the partition point with the smallest Gini index as the best partition point, so as to select the best partition point to construct the decision tree.

[0037] Get the influence weight of welding quality, including: Taking welding quality as the evaluation target, a hierarchical structure is established, which includes: target, criteria and sub-criteria, where the target is at the top level, and the criteria and sub-criteria are expanded layer by layer until the bottom level; a judgment matrix is ​​constructed based on expert scoring or voting; In this embodiment, the analytic hierarchy process (AHP) is a well-known technology in the art and will not be described in detail. In the hierarchical structure, a 1-9 scale method is usually used, where 1 means that two elements are equally important, and 9 means that one element is extremely important than another element. The judgment matrix is ​​normalized, that is, the elements of each column are divided by the column sum to obtain a normalized matrix, the sum of each row in the normalized matrix is ​​calculated, and normalization is performed so that the sum of each element in the vector is 1 to obtain the influence weight.

[0038] Welding quality index, including: The quality data of any batch of welding materials is taken as a marked batch, and the average of the cumulative sum of the products between the target items and the influence weights of the target items in the marked batch is taken as the welding quality index of the marked batch.

[0039] Specifically, the welding quality index satisfies the following relationship: ; In the formula, Indicates Welding quality index of quality data of batch welding materials, Indicates the total number of quality data items, Indicates The weight of influence of welding quality of each quality data item, Indicates Quality data of batch welding materials Item quality data.

[0040] That is to say, Indicates The influence of each quality data of welding on the overall welding quality is comprehensively considered through weighted average, providing a quantitative index for evaluating the welding quality of welding materials.

[0041] In addition, another embodiment also includes: ; In the formula, Indicates Welding quality index of quality data of batch welding materials, Indicates the total number of quality data items, Indicates The weight of influence of welding quality of each quality data item, Indicates Quality data of batch welding materials Item quality data, No. The weighted average of all quality data for a batch of welding materials.

[0042] That is, the sum of the weighted data items is divided by the sum of the weights to get the weighted average, the weighted average Taking into account the weight of each data item, The comprehensive welding quality of the quality data of batch welding materials. It is suitable for scenarios where different data items have different degrees of influence on welding quality, and can more fairly reflect the overall level of welding quality.

[0043] Further analysis is performed to divide the data according to the Gini index of all division points, so as to complete the construction of the decision tree. However, in this process, since the division points are not divided into two subsets with large differences in welding quality for analysis and quantification, the accuracy of the division points cannot be guaranteed, resulting in the final decision tree being difficult to divide the data more accurately. Therefore, the division termination degree is set for the two subsets after each division point is divided. The specific steps are as follows: S3: Based on the Gini index of the data subsets on both sides of the partition point, the partition preference of the partition point is obtained, and the ratio between the partition preference of the partition point and the Gini index is used as the adjusted optimized Gini index, and a decision tree is recursively constructed to obtain the partition termination degree of each subset after each partition.

[0044] Reference Figure 2 , see the decision tree construction process for details.

[0045] The degree of segmentation preference, including: The quality data of multiple batches of welding materials are obtained, and the quality data of any batch of welding materials is used as a marked batch, and the quality data of any item in the marked batch of welding materials is used as a target item. The difference between the target item and the minimum value of the same target item in multiple batches is divided by the range of the target item, and the absolute value after deducting one half is used as the absolute deviation, which is used to represent the position of the target item relative to the median, and the absolute deviation is attenuated using exponential decay; The difference between the average values ​​of the welding quality index of the two subsets on the left and right of the target item as the dividing point is calculated, and then divided by the range and squared to obtain the welding quality difference. The product of the attenuated absolute deviation and the welding quality difference is used as the segmentation optimization degree of the target item.

[0046] Specifically, the segmentation preference degree satisfies the following relationship: ; In the formula, Indicates Batch welding materials The item quality data is used as the preferred degree of segmentation of the dividing points, Indicates Batch welding materials Item quality data, Indicates the first batch of welding materials among multiple batches. The minimum value of the quality data item, Indicates the first batch of welding materials The quality data of the items is extremely poor. Indicates Batch welding materials The quality data of each item is taken as the average value of the welding quality index in the left subset of the partition points. Indicates Batch welding materials The item quality data is taken as the average value of the welding quality index in the right subset of the partition points.

[0047] That is to say, Used to determine the division point in Whether the batch welding material is in the middle of all the data, the target item is used as the dividing point and the The difference between the minimum value of the item quality data and the range of the target item are compared. The closer the ratio is, the better. This means that the closer the partition point is to the middle of the data, the higher the partition preference is. It represents the difference in the average values ​​of the welding quality index of the two subsets. The larger the difference is, the more the division point divides the data into two parts with a large difference in quality, and the higher the segmentation preference is.

[0048] In addition, another embodiment also includes: ; In the formula, Indicates Batch welding materials The item quality data is used as the preferred degree of segmentation of the dividing points, Indicates Batch welding materials The variance of the welding quality index in the left subset with the item quality data as the partition point, Indicates Batch welding materials The variance of the welding quality index in the right subset of the item quality data as the partition point, Indicates Variance of the weld quality index of the quality data of a batch of welding materials.

[0049] In other words, the variance ratio measures the ratio of the variance of the subset after division to the variance of all welding quality indices. The smaller the value, the higher the consistency of the subset after division in welding quality and the better the segmentation preference.

[0050] S4: Determine whether to stop the division based on the division termination degree and complete the decision tree division; classify the welding quality according to the division results and quantitatively evaluate the welding processability.

[0051] The decision tree partitioning process includes: The decision tree is divided into two subsets according to the minimum value of the optimized Gini index, and the termination degree of the division of the two subsets is calculated respectively; Specifically, the division termination degree satisfies the following relationship: ; In the formula, Indicates the degree of termination of the division of the left or right subset of this segmentation. Indicates the number of data contained in the left subset or right subset after segmentation. Indicates the preset minimum value of the subset, The natural index The exponential function with base , Indicates the total number of quality data items, The first The weight of the quality data, Indicates the first The standard deviation of the quality data, Indicates all The standard deviation of the quality data.

[0052] That is, the experience value is ， Implementers can set the minimum value of data included in the subset according to the actual implementation situation. It represents the ratio between the minimum value and the amount of data contained in the subset after division. The larger the ratio is, the smaller the amount of data contained in the subset is. In order to ensure that the divided subset contains enough data to ensure that the segmentation is meaningful, the subset should not be further divided. It means that the standard deviation of each data item in the segmented subset is compared with the standard deviation of all the data items. The larger the ratio is, the higher the standard deviation of the two subsets is and the worse the segmentation effect is.

[0053] In response to the termination degree of the division corresponding to the two subsets being greater than the termination threshold, the two subsets are marked as final subsets and no longer divided. Otherwise, other subsets are selected to continue the division until the termination degree of the division in all subsets is greater than the termination threshold, and the division of the decision tree is completed.

[0054] It should be noted that the termination threshold is , implementers can set it according to the actual implementation situation.

[0055] The quality data is classified according to the decision tree to obtain multiple clusters. The average value of the welding quality index of each data in each cluster is taken as the welding quality of the corresponding cluster. The welding qualities of all clusters are sorted from large to small and numbered. The quality data of the welding material in the cluster with the largest welding quality is recorded as level one, and the other clusters are assigned welding quality levels according to their numbers to complete the quantitative scoring of the welding quality.

[0056] The present invention also provides a comprehensive quantitative evaluation system for welding processability of welding materials. Figure 3 As shown, the system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a comprehensive quantitative evaluation method for welding processability of welding materials according to the first aspect of the present invention is implemented.

[0057] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0058] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0059] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0060] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A comprehensive quantitative evaluation method for welding processability of welding materials, characterized in that: include: Obtain quality data of multiple batches of welding materials in history and pre-process the data; The Gini index is calculated by taking any quality data of welding materials as the dividing point, and the influence weight of each quality data on welding quality is obtained according to the hierarchical analysis method, and the welding quality index of each batch of welding materials is calculated; Based on the Gini index of the data subsets on both sides of the partition point, the partition preference degree of the partition point is obtained, and the ratio between the partition preference degree of the partition point and the Gini index is used as the adjusted optimized Gini index, and a decision tree is recursively constructed to obtain the partition termination degree of each subset after each partition; Based on the termination degree of the division, it is determined whether to stop the division, and the decision tree division is completed; the welding quality is classified according to the division result, and the welding processability is quantitatively evaluated; Among them, the division termination degree Satisfies the following relationship: , where Indicates the number of data contained in the left subset or right subset after segmentation. Indicates the preset minimum value of the subset, The natural index The exponential function with base , Indicates the total number of quality data items, The first The weight of the quality data, Indicates the first The standard deviation of the quality data, Indicates all The standard deviation of the quality data.

2. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: in, The quality data include: the degree of cracks on the weld surface, the degree of inclusions such as weld pores, the tensile strength of the weld joint, the yield strength of the weld joint, the hardness of the weld joint, the toughness of the weld joint and other data.

3. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The Gini index includes: According to the partition point, the quality data of the target item is divided into two parts: the left subset and the right subset, where the Gini index of the partition point is the weighted average of the Gini indexes of the left subset and the right subset, and the weight is the number of data points in the left subset and the right subset.

4. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The weight of influence of each quality data on welding quality is obtained according to the hierarchical analysis method, including: Taking the welding quality as the evaluation target, a hierarchical structure is established, wherein the hierarchical structure includes: target, criterion and sub-criteria, wherein the target is at the top layer, and the criterion and sub-criteria are unfolded layer by layer until the bottom layer; constructing a judgment matrix according to expert scoring or voting; The judgment matrix is ​​normalized, that is, the elements of each column are divided by the column sum to obtain a normalized matrix, the sum of each row in the normalized matrix is ​​calculated, and normalization is performed so that the sum of each element in the vector is 1 to obtain the influence weight.

5. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The welding quality index includes: The quality data of any batch of welding materials is taken as a marked batch, and the average of the cumulative sum of the products between the target items and the influence weights of the target items in the marked batch is taken as the welding quality index of the marked batch.

6. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The segmentation preference degree includes: The quality data of multiple batches of welding materials are obtained, and the quality data of any batch of welding materials is used as a marked batch, and the quality data of any item in the marked batch of welding materials is used as a target item. The difference between the target item and the minimum value of the same target item in multiple batches is divided by the range of the target item, and the absolute value after deducting one half is used as the absolute deviation, which is used to represent the position of the target item relative to the median, and the absolute deviation is attenuated using exponential decay; The difference between the average values ​​of the welding quality index of the two subsets on the left and right of the target item as the dividing point is calculated, and then divided by the range and squared to obtain the welding quality difference. The product of the attenuated absolute deviation and the welding quality difference is used as the segmentation optimization degree of the target item.

7. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The step of judging whether to stop the division based on the division termination degree to complete the decision tree division includes: The decision tree is divided into two subsets according to the minimum value of the optimized Gini index, and the termination degree of the division of the two subsets is calculated respectively; In response to the termination degree of the division corresponding to the two subsets being greater than the termination threshold, the two subsets are marked as final subsets and no longer divided. Otherwise, other subsets are selected to continue the division until the termination degree of the division in all subsets is greater than the termination threshold, and the division of the decision tree is completed.

8. A comprehensive quantitative evaluation method for welding processability of welding materials according to claim 1, characterized in that: The quantitative evaluation of welding processability includes: The quality data is classified according to the decision tree to obtain multiple clusters. The average value of the welding quality index of each data in each cluster is taken as the welding quality of the corresponding cluster. The welding qualities of all clusters are sorted from large to small and numbered. The quality data of the welding material in the cluster with the largest welding quality is recorded as level one, and the other clusters are assigned welding quality levels according to their numbers to complete the quantitative scoring of the welding quality.

9. A comprehensive quantitative evaluation system for welding processability of welding materials, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the comprehensive quantitative evaluation method for welding processability of welding materials according to any one of claims 1 to 8 is implemented.

Images