Method, apparatus, device, and computer program product for optimizing the quality of a stamped part

Abstract

A method, apparatus, device, and computer program product for optimizing the quality of a stamped part are disclosed. The method includes receiving one or more material parameters, one or more process parameters, and defect data; generating a correlation model based at least on the defect data, the one or more material parameters, and the one or more process parameters; and predicting one or more potential defects in an additional one or more stamped parts based at least on the correlation model.

Description

Methods, apparatus, equipment and computer programs for optimizing the quality of stamped parts Technical Field

[0001] This disclosure relates to the field of industrial production, and more specifically, to methods, apparatus, equipment, and computer program products for optimizing the quality of stamped parts. Background Technology

[0002] The production of modern industrial products requires a large number of parts. A portion of these parts are stamped components, formed through a stamping process. Stamping is a material forming process. For example, in automobile production, raw materials (e.g., steel, aluminum alloys) are stamped to form various automobile parts. These parts can then be assembled into the automobile. The selection of the material parameters of the raw materials and / or the process parameters of the stamping process is crucial to the quality of the stamped parts. Inappropriate combinations of material parameters and / or process parameters can result in defective stamped parts, thereby reducing production efficiency and increasing production costs. Summary of the Invention

[0003] This disclosure provides a method for optimizing the quality of stamped parts by a computing device, comprising: receiving a first set of parameters, the first set of parameters including one or more material parameters describing the properties of materials used to form one or more stamped parts; receiving a second set of parameters, the second set of parameters including one or more process parameters describing the properties of processes used to form the one or more stamped parts; receiving defect data indicating one or more defects identified from the one or more stamped parts; generating an association model based at least on the defect data, the first set of parameters, and the second set of parameters, the association model describing the relationship between the identified defects and at least one of the one or more material parameters and / or at least one of the one or more process parameters; and predicting one or more potential defects in the one or more stamped parts, at least based on the association model.

[0004] This disclosure also provides an apparatus for optimizing the quality of stamped parts, comprising components for performing the methods described herein.

[0005] This disclosure also provides an apparatus for optimizing the quality of stamped parts, comprising: at least one processor; and at least one storage device storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the methods described herein.

[0006] This disclosure also provides a computer program product including instructions that, when executed by a processor, cause to perform the methods described herein.

[0007] The methods and systems disclosed herein can automatically identify the correlation between defects in stamped parts and the materials and processes used to form the stamped parts, and predict potential defects in future stamped parts based on this correlation. If a potential defect is predicted, the systems and methods disclosed herein can automatically adjust the materials and processes used to form the stamped parts, thereby avoiding the formation of the potential defect. The methods and systems disclosed herein can handle combinations of multi-dimensional parameters associated with the stamping process, and make predictions and adjustments based on the overall combination of these parameters (not just a single parameter), avoiding the disadvantages that may occur when adjusting a single parameter. Attached Figure Description

[0008] The accompanying drawings, which form part of this specification, illustrate embodiments of this disclosure and, together with the specification, serve to explain the principles of this disclosure.

[0009] This disclosure will become clearer with reference to the accompanying drawings and the following detailed description, wherein:

[0010] Figure 1 shows a flowchart of a stamping process for forming a stamped part from raw materials according to an embodiment of the present disclosure.

[0011] Figure 2 shows a flowchart of a method for optimizing the quality of a stamped part according to an embodiment of the present disclosure.

[0012] Figure 3 shows a schematic diagram of an apparatus for optimizing the quality of a stamped part according to an embodiment of the present disclosure.

[0013] Figure 4 shows a schematic diagram of a computing device in which the present disclosure may be applied, according to an exemplary embodiment of the present disclosure. Detailed Implementation

[0014] The following description is provided to enable those skilled in the art to implement and use the embodiments, and is provided in the context of a particular system and its requirements. Various modifications will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and systems without departing from the spirit and scope of the embodiments. Therefore, the embodiments are not limited to those shown, but are to be given the widest scope consistent with the principles and features disclosed herein.

[0015] Figure 1 shows a flowchart of a stamping process 1000 for forming a stamped part from raw materials according to an embodiment of the present disclosure.

[0016] The stamping process 1000 can begin at step 1100. In step 1100, a destacking process can be performed. The destacking process may include breaking down the stacked raw materials into raw material sheets. For example, the raw materials can be steel, aluminum alloy, titanium alloy, or any other suitable raw material. These raw materials are typically produced as raw material sheets and rolled into stacks for easy transport and storage. The destacking process flattens the stacked raw materials into raw material sheets for subsequent processing.

[0017] The destacking process may involve one or more parameters, including material parameters related to the raw materials (referred to as “material parameters”) and parameters related to the equipment, processes, and auxiliary materials involved in performing the destacking process (referred to as “process parameters”). Examples of material parameters include, but are not limited to, the chemical composition of the raw materials (e.g., alloy ratios), mechanical properties (stiffness, ductility, density, etc.), etc. Examples of process parameters include, but are not limited to, the magnitude, location, direction, and duration of stress applied to the destacking material.

[0018] Then, the stamping process 1000 can continue to step 1200. In step 1200, a cleaning process can be performed. The cleaning process is used to clean the raw material boards obtained after the destacking process. Specifically, the raw material boards are brought into the cleaning machine by the inlet rollers. The cleaning machine pump station sprays oil from the oil tank onto the upper and lower surfaces of the raw material boards at a set pressure and flow rate, while cleaning brush rollers brush the raw material boards up and down. Particulate impurities on the upper and lower surfaces of the raw material boards are removed by the oil spraying and brushing. Then, the hydraulic station (or pneumatic in some cleaning machines) squeezes out the residual oil on the raw material boards by a set squeezing pressure, thereby leaving an oil film coating of a certain thickness on the surface of the raw material boards.

[0019] Oil film coatings play a crucial role in the forming process of stamped parts. For example, oil film coatings effectively lubricate the die and the workpiece. In addition, oil film coatings also provide cooling, anti-wear, extreme pressure properties, cleaning, and rust prevention. Oil film coatings significantly improve the surface quality of parts. Lubricity and extreme pressure properties are essential functions that effectively ensure the smooth progress of stamping and drawing processes, preventing surface cracking and wrinkling, ensuring good formability and demolding properties of the raw material sheet, and preventing scratches and sintering between the die surface and the raw material sheet. Improved lubrication can improve die performance, thereby reducing the requirements for raw materials, allowing the use of lower-grade raw materials, and significantly reducing scrap rates.

[0020] The thickness of the oil film coating can significantly affect the quality of stamped parts. For example, an excessively thin oil film will reduce the effectiveness of the coating, potentially leading to cracking or other defects on the stamped parts. Conversely, an excessively thick oil film may cause wrinkles, oil pockets, or other defects. The uniformity of the oil film coating thickness is also important. Inappropriate and / or uneven oil film thickness will reduce the yield rate of stamped parts.

[0021] The thickness of the oil film coating is influenced by multiple parameters. These parameters can include material parameters related to the raw materials (referred to as "material parameters") and parameters related to the equipment, processes, and auxiliary materials involved in the cleaning process (referred to as "process parameters"). For example, the material parameters of the raw material board may include, but are not limited to, the dimensions, thickness, surface roughness, surface cleanliness, and initial oil film thickness (the raw material may have been coated with a certain amount of oil at the time of manufacture). Process parameters may include parameters related to the oil used, including but not limited to the type of oil (composition, viscosity, density, acid value, flash point, moisture content), temperature, and purity (impurity content). Process parameters may also include parameters related to spraying, including but not limited to the spraying position, angle, rate, and duration. Process parameters may also include parameters related to the mechanical configuration of the cleaning machine, including but not limited to brush roller overlap, brush roller speed, and brush roller integrity. Additionally, the multiple parameters affecting the thickness of the oil film coating may also include multiple environmental parameters related to the environment, including but not limited to temperature, humidity, and air cleanliness.

[0022] Then, stamping process 1000 can continue to step 1300. In step 1300, the stamping forming process can be performed. The stamping forming process is used to form the raw material sheet that has undergone the cleaning process into the desired stamped part. The stamping forming process typically includes four steps: drawing, trimming, punching, and flanging. Each step can correspond to a set of dies. The drawing process is the main forming step. Through the drawing process, the main shape of the part is basically formed. In addition, the part after drawing can obtain a certain strength and rigidity through deformation strengthening. The trimming process mainly removes the excess auxiliary material from the drawing process, such as draw beads, process supplement surfaces, etc. The punching process punches holes in the part. The flanging process mainly turns up the edges around the part for use in subsequent edge-wrapping processes or subsequent assembly. Through the stamping forming process, a stamped part can be formed.

[0023] Multiple parameters associated with the stamping process can affect the quality of the resulting stamped parts. These parameters can include material parameters related to the raw materials (referred to as "material parameters") and parameters related to the equipment, processes, and auxiliary materials involved in performing the stamping process (referred to as "process parameters"). For example, material parameters can include the material parameters of the material sheet itself used in the stamping process, including but not limited to the chemical composition (alloy ratio), mechanical properties (stiffness, ductility, density, etc.), roughness, and thickness of the surface oil film. Process parameters can include parameters for each step of each of the drawing, trimming, punching, and flanging processes. As an example, process parameters for the drawing process can include, but are not limited to, the execution sequence of the various steps of the drawing process, the magnitude, location, direction, and duration of the drawing force used, etc. It should be noted that, although not listed individually, the configuration parameters of the equipment used to perform the stamping process, the timing relationship of the various steps of the stamping process, and the properties of other auxiliary materials used can all be considered as process parameters.

[0024] Stamping process 1000 produces stamped parts from raw materials, which can then be used as components in the production of industrial products (e.g., automobiles). To ensure the quality of these industrial products, the quality of the stamped parts must also be ensured. However, stamped parts produced by stamping process 1000 inevitably contain defects. These defects can be of various types, including cracking, necking, thinning, wrinkling, and so on. These defects are the result of the combined effect of the material and process parameters used in performing stamping process 1000.

[0025] To minimize defects in stamped parts, human experience can be relied upon to set the material and process parameters for the stamping process. However, human experience can be intuitive and superficial, as humans may not be able to discern the intrinsic connection between specific parameters and specific defects. They may also fail to notice the significant impact of subtle changes in specific parameters on defects. Furthermore, human experience is often one-dimensional. When considering dozens or even hundreds of parameters across multiple dimensions (e.g., material and process), humans struggle to provide a comprehensive overview. In addition, human experience lacks flexibility; it typically requires a considerable amount of time to develop, verify, and adjust.

[0026] Figure 2 shows a flowchart of a method 2000 for optimizing the quality of a stamped part according to an embodiment of the present disclosure. Method 2000 can be executed by a computing device. An example of a computing device is, for example, the apparatus 3000 described in Figure 3. Method 2000 can provide an improved method for optimizing the quality of a stamped part.

[0027] Method 2000 may begin at step 2100. In step 2100, a first set of parameters may be received. The first set of parameters may include one or more material parameters. The one or more material parameters may describe the properties of the material used to form one or more stamped parts. As an example, the properties of the material described by the one or more material parameters may include one or more of the following: chemical composition, roughness, thickness of the surface-coated oil film, mechanical properties, etc. For example, the one or more material parameters may include any part or all of the material parameters described with respect to the process of FIG1.

[0028] The first set of parameters may be received from one or more sources. For example, parameters related to the chemical composition, physical properties, and / or mechanical properties of raw materials may be received from the personnel or equipment of the raw material supplier, the raw material inventory manager, or the raw material inspector. Alternatively, one or more of the first set of parameters may be obtained from any other suitable source without limitation.

[0029] Then, method 2000 can proceed to step 2200. In step 2200, a second set of parameters may be received. The second set of parameters may include one or more process parameters. These one or more process parameters may describe the properties of the process used to form the one or more stamped parts. As an example, the properties of the process described by these one or more process parameters include process parameters associated with one or more of the material destacking process, cleaning process, and stamping process. For example, these one or more process parameters may include any part or all of the process parameters described with respect to the process of FIG1.

[0030] The second parameter set can be received from one or more sources. For example, the second parameter set can be received from personnel or equipment performing the stamping process 1000. Alternatively, one or more of the second parameter set can be obtained from any other suitable source without limitation.

[0031] Then, method 2000 can proceed to step 2300. In step 2300, defect data can be received. The defect data can indicate one or more defects identified from the aforementioned one or more stamped parts. The defect data can include any data associated with the one or more defects, such as the type of defect, the location of the defect, the severity level of the defect, etc. The type of defect can include, but is not limited to, cracking, necking, thinning, wrinkling, etc. In addition, the defect data can also include the type of the defective stamped part (e.g., part type).

[0032] Defect data can be received from one or more sources. In one embodiment, the stamped parts produced by the stamping process 1000 undergo manual inspection, where an inspection worker identifies defective stamped parts by visual inspection and records associated defect information (e.g., type of stamped part, type of defect, location of defect, severity level of defect, etc.). The inspection worker can send the recorded defect information as defect data to a computing device executing method 2000 via a terminal device. In an alternative embodiment, an inspection machine can be used instead of an inspection worker to obtain defect data. For example, the inspection machine can identify defects in the stamped parts by ultrasonic or infrared detection. Preferably, the inspection machine can also identify defects in the stamped parts by acquiring one or more images of the stamped parts and based on computer vision (e.g., computer image recognition). The inspection machine can send the identified defects as defect data to the computing device executing method 2000. Using computer vision to identify defects in stamped parts by an inspection machine can reduce the inspection worker's contact with the stamped parts and improve inspection efficiency.

[0033] Preferably, an identifier can be engraved for each stamped part during the stamping process 1000. An inspection worker or inspection machine can identify the engraved identifier and send it along with defect data to a computing device executing method 2000, thereby enabling the defect data to be associated with the corresponding stamped part (e.g., it can be associatedly stored in a database).

[0034] Then, method 2000 can proceed to step 2400. In step 2400, an association model can be generated based on the received defect data, the first parameter set, and the second parameter set. The generated association model can be stored and / or distributed.

[0035] In one embodiment, the generated correlation model can describe the relationship between an identified defect and at least one material parameter in a first set of parameters and / or at least one process parameter in a second set of parameters. This relationship can be causal. As an example, the correlation model can describe the relationship between cracking (or any other type of defect) and oil film thickness. This relationship can be represented by correlation parameters. In one example, the correlation parameter can have binary values, i.e., "related" or "unrelated". In another example, the correlation parameter can have numerical values ​​indicating the degree of correlation, such as normalized probability values. As another example, a defect can be represented as a defect vector (e.g., including numerical values ​​representing the type, location, and severity of the defect), and a set of material parameters or process parameters can be organized together into a parameter vector. In this case, the correlation parameters can include one or more matrices representing the correlation between the two. These one or more matrices can measure the influence of the set of material and process parameters formed as a whole (rather than each parameter individually) on the defect.

[0036] Preferably, the generated correlation model can also be described as a safety threshold estimated by one or more material parameters and / or one or more process parameters. Each safety threshold can be associated with a defect and one or a set of parameters. For example, a safety threshold for oil film thickness can be set for cracking. This safety threshold indicates one or more oil film thickness values ​​(or ranges of values) that will not cause cracking of the stamped part. Furthermore, a set of safety thresholds for a set of parameters can be set for cracking. This set of parameters can include a combination of a set of material parameters and process parameters. The set of safety thresholds can include one or more parameter vectors for this set of parameters, each parameter vector containing a set of values ​​that will not cause cracking of the stamped part.

[0037] In step 2400, machine learning can be applied to the defect data, the first parameter set, and the second parameter set to generate an association model. For example, the defect data, the first parameter set, and the second parameter set can be used as training data input to a machine learning model to train the model. The trained model can be used as the generated association model. Any suitable machine learning algorithm (such as neural networks, support vector machines, logistic regression, random forests, etc.) can be used to train the association model.

[0038] According to embodiments of this disclosure, when using a machine learning model, each of the defect data, the first parameter set, and the second parameter set can be quantified. For example, one or more defect scores can be generated based on one or more defects in a stamped part, and these defect scores can represent the type, location, severity level, etc., of the defect. Preferably, a single defect score can be used to comprehensively characterize the type, location, and severity level of the defect. The first parameter set and / or the second parameter set can be processed similarly to obtain material scores and process scores. Machine learning can be applied to the defect scores, material scores, and process scores to generate a correlation model.

[0039] Preferably, when generating the correlation model, the parameters involved in the stamping process 1000 can be divided into a dimensional hierarchy. This dimensional hierarchy can include multiple dimensional levels. A corresponding correlation model can be generated for each dimensional level.

[0040] For example, at the first-dimensional level, one or more parameters involved in the stamping process 1000 can be divided into two main dimensions: the material parameter dimension and the process parameter dimension. The material parameter dimension can be a synthesis of all considered material parameters, represented by a single material score. The process parameter dimension can be a synthesis of all considered process parameters, represented by a single process score. At the first-dimensional level, machine learning can be applied to the defect score, material score, and process score to generate a first correlation model. This first correlation model can indicate the relationship between defects in the stamped part and the single material score and the single process score. Since the material parameters and process parameters have been simplified to single material scores and single process scores respectively, the training and use of the first correlation model is relatively simple.

[0041] Furthermore, at the second dimension level, the material parameter dimension and process parameter dimension of the first dimension level can each be divided into more sub-dimensions. For example, the material parameter dimension can be divided into physical material parameter dimension and chemical material parameter dimension. The process parameter dimension can be divided according to the destacking process parameter dimension, cleaning process parameter dimension, and stamping process parameter dimension. Each sub-dimension can have a corresponding score, which can be obtained by synthesizing the parameters associated with that sub-dimension. For example, the corresponding score of the cleaning process parameter dimension can be obtained by synthesizing multiple process parameters associated with the cleaning process (e.g., parameters related to oil, parameters related to spraying, parameters related to the mechanical configuration of the equipment, etc.). At the second dimension level, machine learning can be applied to the defect score and the corresponding score of each sub-dimension to generate a second correlation model. The second correlation model can indicate the relationship between the defects of the stamped part and multiple sub-dimensions at the second dimension level (i.e., physical material parameter dimension, chemical material parameter dimension, destacking process parameter dimension, cleaning process parameter dimension, and stamping process parameter dimension). Compared with the first correlation model, the second correlation model involves parameters with finer granularity and more dimensions, thus having higher accuracy, but it is relatively more complex to train and use.

[0042] Furthermore, at the third-dimensional level, each of the sub-dimensions at the second-dimensional level can be further subdivided into more sub-dimensions. For example, the cleaning process parameter dimension can be divided into three sub-dimensions: oil-related parameters, spraying-related parameters, and equipment mechanical configuration-related parameters. Each sub-dimension can have a corresponding score, which can be derived by combining multiple parameters associated with that sub-dimension. For instance, the corresponding score for the spraying-related parameter dimension can be derived by combining multiple parameters such as spraying position, angle, rate, and duration. Each of the other sub-dimensions can also be subdivided into corresponding sub-dimensions. At the third-dimensional level, machine learning can be applied to the defect scores and the corresponding scores of each sub-dimension to generate a third correlation model. The third correlation model can indicate the relationship between defects in stamped parts and multiple sub-dimensions at the third-dimensional level. Compared to the second correlation model, the third correlation model involves finer-grained parameters and more dimensions, thus offering higher accuracy, but it is more complex to train and use.

[0043] Similarly, one or more parameters involved in the stamping process 1000 can be further divided into more dimensional levels, and a corresponding correlation model can be generated for each dimensional level. As the dimensional level increases, the parameters involved in the corresponding correlation model become more granular and have more dimensions, but training and use become more complex.

[0044] Preferably, at each dimension level, different weights can be applied to different parameter dimensions. For example, at the first dimension level, a first set of weights can be applied to one or more material parameters (material parameter dimensions) to generate a material score, and a second set of weights can be applied to one or more process parameters (process parameter dimensions) to generate a process score. Considering that the properties of the material itself may have a greater impact on defect generation than process parameters, the first set of weights can be greater than the second set of weights to highlight the influence of material parameters in the first correlation model. Similarly, when generating the second correlation model, different weights can be applied to each sub-dimension at the second dimension level. Furthermore, when generating the third correlation model, different weights can be applied to each grandchild dimension at the third dimension level.

[0045] Then, method 2000 can proceed to step 2500. In step 2500, one or more potential defects in one or more additional stampings can be predicted based on the correlation model generated in step 2400. The one or more additional stampings may refer to stampings put into production after the stampings used to generate the correlation model.

[0046] Specifically, in step 2500, an additional first set of parameters and an additional second set of parameters may be received. The additional first set of parameters includes one or more material parameters associated with the additional one or more stamped parts, and the additional second set of parameters includes one or more process parameters associated with the additional one or more stamped parts. The additional first set of parameters and the additional second set of parameters may be received from equipment or personnel currently performing the stamping process 1000 for the additional one or more stamped parts.

[0047] Then, based at least on the additional first set of parameters and the additional second set of parameters, the correlation model generated in step 2400 can be used to predict one or more potential defects in the additional one or more stampings.

[0048] In one example, data describing a potential defect can be inferred based on the values ​​of a specific parameter (or group of parameters) in the additional first set of parameters and / or the additional second set of parameters, and based on relational parameters that describe the relationship between the specific parameter (or group of parameters) and the defect. For example, data describing a potential defect can describe whether the potential defect will occur, its possible type, its possible location, and its possible severity level.

[0049] In another example, the values ​​of specific parameters (or groups of parameters) in an additional first set of parameters and / or an additional second set of parameters can be compared to the corresponding safety thresholds (or groups of thresholds) in the association model. If the value of a specific parameter (or group of parameters) does not fall within the corresponding safety threshold (or group of thresholds), a potential defect may occur.

[0050] In this way, the previously generated correlation model can be used to predict potential defects that may occur in the future.

[0051] Preferably, correlation models at different dimensional levels (e.g., the first correlation model, second correlation model, third correlation model, etc. described above) can be used for the additional first parameter set and / or the additional second parameter set. For each stamping (or a set of stampings using the same parameters), the first correlation model can be used first to make predictions in the main parameter dimensions. If the first correlation model does not predict a defect (or the confidence level of the prediction given is lower than a first predetermined threshold), it can be assumed that no defect will occur, and the prediction for that stamping is terminated. Otherwise, the second correlation model can be used for more refined predictions in the sub-dimensions. If the second correlation model does not predict a defect (or the confidence level of the prediction given is lower than a second predetermined threshold), it can be assumed that no defect will occur, and the prediction for that stamping is terminated. Otherwise, the third correlation model can be used for more refined predictions in the grandchild dimensions. If the third correlation model predicts that a defect will occur in the grandchild dimensions, it can be assumed that one or more potential defects have been predicted. This can also be extended to correlation models corresponding to fourth, fifth, or higher dimensional levels. In this way, we can avoid using high-dimensional association models (which have a large number of dimensions and fine granularity) all the time, thereby reducing the consumption of computation and resources while maintaining acceptable prediction accuracy.

[0052] Optionally, method 2000 may also include performing one or more actions when one or more potential defects are predicted. These one or more actions may be alarm actions and / or adjustment actions.

[0053] In one embodiment, when one or more potential defects are predicted, the one or more actions performed may include issuing an alert to an inspection worker who is inspecting the stampings. In response to the alert, the inspection worker may more carefully examine one or more additional stampings that are predicted to contain defects and reject them, thereby preventing defective stampings from spilling over into subsequent industrial products.

[0054] Additionally or alternatively, when one or more potential defects are predicted, the one or more actions performed may include adjusting one or more material parameters and / or one or more process parameters associated with one or more additional stamped parts based on the predicted one or more potential defects. The aim is to reduce the actual occurrence (or decrease the incidence) of the defects. For example, the values ​​of specific parameters (or sets of parameters) associated with the predicted potential defects may be adjusted to the corresponding safety thresholds (or sets of thresholds) indicated in the associated model. Note that since the material parameters are typically fixed after the stamping process for a given raw material sheet begins, the process parameters can be adjusted. For example, those safety thresholds (or sets of thresholds) that match the first set of parameters (material parameter set) may be found based on the additional first set of parameters (material parameter set), and the process parameters indicated in these safety threshold sets are then applied to the stamping process 1000. In this way, the material parameters and process parameters applied to the stamping process 1000 belong to the same safety threshold set.

[0055] Optionally, method 2000 may further include receiving additional defect data. This additional defect data indicates one or more actual defects identified from the additional one or more stampings. Further, the association model generated in step 2400 may be updated based on the one or more actual defects identified from the additional one or more stampings. Specifically, the additional defect data, an additional first set of parameters, and an additional second set of parameters may be applied as new training data to the machine learning model used in step 2400. If the additional first set of parameters and / or the additional second set of parameters are adjusted, the adjusted additional first set of parameters and / or the additional second set of parameters may be used as training data. Compared to the previous association model, the updated association model may have different relationship parameters and / or safety thresholds. In some embodiments, the updated association model may have different weight assignments (e.g., different first set of weights and different second set of weights).

[0056] Optionally, prior to performing step 2400, method 2000 may further include receiving a third set of parameters, which includes one or more environmental parameters that can describe the properties of the environment used to form one or more stamped parts (e.g., temperature, humidity, air cleanliness, etc.). Accordingly, in step 2400, the association model may also be additionally generated based on the third set of parameters. Furthermore, when one or more potential defects are predicted, the one or more actions performed may also include adjusting one or more environmental parameters associated with the defect, for example, adjusting temperature, humidity, and air cleanliness by controlling air conditioning and air purification equipment.

[0057] The method according to embodiments of this disclosure provides an improved approach for enhancing the quality of stamped parts. This method automatically identifies the correlation between defects in a stamped part and the materials and processes used to form the stamped part. The identified correlations become increasingly accurate with the accumulation of data. The method can also predict potential defects in future stamped parts based on the identified correlations. Furthermore, if a potential defect is predicted, the method can issue an alert or automatically adjust the materials and processes used to form the stamped part, thereby preventing the spillover of defective stamped parts and / or preventing the actual formation of potential defects. The method of this disclosure can handle combinations of multi-dimensional parameters associated with the stamping process and makes predictions and adjustments based on the overall combination of these parameters (not just a single parameter), avoiding the disadvantages that may occur when adjusting a single parameter. Simultaneously, the method of this disclosure can construct correlation models at multiple dimensional levels, which can provide a balance of flexibility, accuracy, and economy in its use.

[0058] Figure 3 shows a schematic diagram of an apparatus 3000 for optimizing the quality of stamped parts according to an embodiment of the present disclosure. The apparatus 3000 can be configured to perform the steps of the method described above with respect to Figure 2. As shown in Figure 3, the apparatus 3000 may include a data receiving unit 3100, a model generation unit 3200, and a defect prediction unit 3300.

[0059] The data receiving unit 3100 can be configured to receive a first set of parameters, a second set of parameters, and defect data. The first set of parameters includes one or more material parameters describing the properties of the material used to form one or more stamped parts. The second set of parameters includes one or more process parameters describing the properties of the process used to form the one or more stamped parts. The defect data indicates one or more defects identified from the one or more stamped parts.

[0060] The model generation component 3200 can be configured to generate an association model based at least on defect data, a first parameter set, and a second parameter set. This association model describes the relationship between the identified defects and at least one of one or more material parameters and / or at least one of one or more process parameters.

[0061] The defect prediction component 3300 can be configured to predict one or more potential defects in one or more additional stampings, based at least on the generated correlation model.

[0062] The apparatus 3000 may also include other components (not shown) for performing other steps of the method described above with respect to FIG2.

[0063] Figure 4 shows a schematic diagram of a computing device 4000 to which the present disclosure may be applied, according to an exemplary embodiment of the present disclosure.

[0064] Referring to Figure 4, a computing device 4000 will now be described as an example of a hardware device applicable to various aspects of this disclosure. The computing device 4000 can be any machine configured to perform processing and / or computation, and can be, but is not limited to, a workstation, server, desktop computer, laptop computer, tablet computer, personal digital assistant, smartphone, portable camera, or any combination thereof. The aforementioned apparatus 3000 can be implemented wholly or at least partially by the computing device 4000 or similar devices or systems.

[0065] The computing device 4000 may include elements capable of connecting to or communicating with the bus 4020 via one or more interfaces. For example, the computing device 4000 may include the bus 4020, one or more processors 4040, one or more input devices 4060, and one or more output devices 4080. The one or more processors 4040 may be any type of processor and may include, but are not limited to, one or more general-purpose processors and / or one or more dedicated processors (such as dedicated processing chips). The input device 4060 may be any type of device capable of inputting information to the computing device and may include, but is not limited to, a mouse, keyboard, touchscreen, microphone, and / or remote control. The output device 4080 may be any type of device capable of presenting information and may include, but is not limited to, a monitor, speaker, video / audio output terminal, and / or printer. The computing device 4000 may also include or be connected to a non-transient storage device 4100. The non-transient storage device 4100 may be any non-transient storage device capable of implementing a data repository, and may include, but is not limited to, disk drives, optical storage devices, solid-state storage devices, floppy disks, flexible disks, hard disks, magnetic tapes or any other magnetic media, compact disks or any other optical media, ROM (read-only memory), RAM (random access memory), cache memory and / or any other memory chips or cassettes, and / or any other media from which a computer can read data, instructions, and / or code. The non-transient storage device 4100 may be detachable from an interface. The non-transient storage device 4100 may have data / instructions / code for implementing the methods and steps described above. The computing device 4000 may also include a communication device 4120. The communication device 4120 can be any type of device or system capable of communicating with external devices and / or with a network, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication equipment and / or chipsets such as Bluetooth devices, 802.11 devices, Wi-Fi devices, Wi-Max devices, cellular communication facilities, etc.

[0066] Bus 4020 may include, but is not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0067] The computing device 4000 may also include working memory 4140, which may be any type of working memory that can store instructions and / or data useful for the operation of the processor 4040, and may include, but is not limited to, random access memory and / or read-only memory devices.

[0068] Software elements may reside in working memory 4140, including but not limited to operating system 4160, one or more application programs 4180, drivers, and / or other data and code. Instructions for performing the methods and steps described above may be included in one or more application programs 4180, and components of the aforementioned apparatus 3000 may be implemented by processor 4040 by reading and executing the instructions of one or more application programs 4180. More specifically, data receiving component 3100 may be implemented, for example, by processor 4040 when executing application program 4180 with instructions for executing steps 2100-2300. Model generation component 3200 may be implemented, for example, by processor 4040 when executing application program 4180 with instructions for executing step 2400. Defect prediction component may be implemented, for example, by processor 4040 when executing application program 4180 with instructions for executing step 2500. The executable code or source code of the software element's instructions may be stored in a non-transitory computer-readable storage medium (such as the storage device(s) 4100 described above) and may be read into the working memory 4140 if compilation and / or installation are possible. The executable code or source code of the software element's instructions may also be downloaded from a remote location.

[0069] This disclosure can be implemented as a system, method, and / or computer program product. The computer program product may include one or more computer-readable storage media having computer-readable program instructions thereon for causing a processor to perform aspects of this disclosure.

[0070] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example (but not limited to), electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of computer-readable storage media includes the following: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital universal disc (DVD), memory sticks, floppy disks, mechanical encoding devices (such as punch cards or recessed protrusions storing instructions thereon), and any suitable combination of the foregoing. As used herein, computer-readable storage media is not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0071] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network (e.g., the Internet, local area network, wide area network, and / or wireless network) to an external computer or external storage device. The network may include copper cables, fiber optic cables, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to computer-readable storage media within the respective computing / processing device.

[0072] Computer-readable program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​(such as Smalltalk, C++, etc.) and conventional procedural programming languages ​​(such as the "C" programming language or similar programming languages). The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network (including a local area network (LAN) or a wide area network (WAN)), or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, including, for example, programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), may be personalized by utilizing state information from the computer-readable program instructions to perform aspects of this disclosure.

[0073] This document describes aspects of the present disclosure with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0074] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine, such that, when executed by the processor of the computer or other programmable data processing apparatus, these instructions create means for implementing the functions / behaviors specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions may also be stored in a computer-readable storage medium that can instruct a computer, programmable data processing apparatus, and / or other device to operate in a particular manner, thereby including an article of manufacture comprising instructions for implementing aspects of the functions / behaviors specified in one or more blocks of the flowchart and / or block diagram.

[0075] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus, or other device to produce a computer-implemented process, thereby causing the instructions to be executed on the computer, other programmable apparatus, or other device to perform the functions / behaviors specified in one or more boxes of a flowchart and / or block diagram.

[0076] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a portion of a module, segment, or instruction containing one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the figures. For example, depending on the functions involved, two consecutive blocks may actually be executed substantially in parallel, or these blocks may sometimes be executed in reverse order. It will also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified function or behavior or executes a combination of dedicated hardware and computer instructions.

[0077] Those skilled in the art should also understand that the various operations illustrated in sequence in the embodiments of this disclosure do not necessarily have to be performed in the illustrated order. Those skilled in the art can adjust the order of operations as needed. They can also add more operations or omit some operations as needed.

[0078] Various embodiments of this disclosure have been described for illustrative purposes, but are not intended to be exhaustive or limiting of the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles of the embodiments, their practical application, or improvements to technology found in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for optimizing the quality of stamped parts, executed by a computing device, characterized in that, include: Receive a first set of parameters, the first set of parameters including one or more material parameters, the one or more material parameters describing the properties of the material used to form one or more stampings; Receive a second set of parameters, the second set of parameters including one or more process parameters, the one or more process parameters describing the properties of the process used to form the one or more stamped parts; Receive defect data, the defect data indicating one or more defects identified from the one or more stamped parts; generate an association model based at least on the defect data, the first parameter set, and the second parameter set, the association model describing the relationship between the identified defects and at least one material parameter and / or at least one process parameter among the one or more material parameters, wherein the association model includes: a first association model indicating the relationship between the identified defects and a single material dimension and a single process dimension, the single material dimension representing the sum of the one or more material parameters, and the single process dimension representing the sum of the one or more process parameters; a second association model indicating the relationship between the identified defects and multiple sub-material dimensions of the single material dimension and... The relationships between multiple sub-process dimensions of the single process dimension, the multiple sub-material dimensions including physical material parameter dimensions and chemical material parameter dimensions, the multiple sub-process dimensions including process parameter dimensions related to the destacking process, process parameter dimensions related to the cleaning process, and process parameter dimensions related to the stamping process; and, at least based on the association model, predicting one or more potential defects in one or more additional stamping parts, the prediction including: applying the first association model to one or more additional parameters associated with the one or more additional stamping parts; terminating the prediction in response to the first association model predicting no defects; and refining the prediction by applying the second association model to the one or more additional parameters associated with the one or more additional stamping parts in response to the first association model predicting one or more potential defects.

2. The method as described in claim 1, characterized in that, The correlation model is also described as a safety threshold estimated by the one or more material parameters and / or the one or more process parameters.

3. The method as described in claim 1, characterized in that, Predicting one or more potential defects in the additional one or more stamped parts includes: receiving an additional first set of parameters, the additional first set of parameters including the one or more material parameters associated with the additional one or more stamped parts; receiving an additional second set of parameters, the additional second set of parameters including the one or more process parameters associated with the additional one or more stamped parts; and using the association model to predict the one or more potential defects, at least based on the additional first set of parameters and the additional second set of parameters.

4. The method as described in claim 3, characterized in that, Also includes: Based on one or more predicted potential defects, adjust the one or more material parameters and / or one or more process parameters associated with the additional one or more stampings.

5. The method as described in claim 4, characterized in that, Also includes: Receive additional defect data, which indicates one or more actual defects identified from the additional one or more stampings; And the association model is updated based on one or more actual defects identified from the additional one or more stamped parts.

6. The method as described in claim 1, characterized in that, The properties of the material described by the one or more material parameters include one or more of the following: the chemical composition of the material; the roughness of the material; the thickness of the oil film coated on the surface of the material; and the mechanical properties of the material.

7. The method as described in claim 1, characterized in that, The attributes of the process described by the one or more process parameters include one or more of the following: process parameters associated with the destacking process of the material; process parameters associated with the cleaning process of the material; and process parameters associated with the stamping process of the material.

8. The method as described in claim 1, characterized in that, The defects in the stamped parts include one or more of the following: cracking, necking, thinning, and wrinkling.

9. The method as described in claim 8, characterized in that, The defect was obtained through computer image recognition.

10. The method as described in claim 1, characterized in that, Generating the association model includes applying machine learning to the defect data, the first parameter set, and the second parameter set to train the association model.

11. The method as described in claim 10, characterized in that, Generating the association model further includes: generating a defect score based on the one or more defects; applying a first set of weights to the one or more material parameters to generate a material score; applying a second set of weights to the one or more process parameters to generate a process score, wherein the second set of weights is less than the first set of weights; and applying machine learning to the defect score, the material score, and the process score to train the association model.

12. The method as described in claim 1, characterized in that, The method further includes: receiving a third parameter set, the third parameter set including one or more environmental parameters describing the properties of the environment used to form the one or more stampings; and wherein the association model is additionally generated based on the third parameter set.

13. A device for optimizing the quality of stamped parts, characterized in that, include: Components for performing the method as described in any one of claims 1-12.

14. A device for optimizing the quality of stamped parts, characterized in that, include: At least one processor; And at least one storage device, the at least one storage device storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the method as described in any one of claims 1-12.

15. A computer program product, characterized in that, The computer program product includes instructions that, when executed by a processor, cause the method described in any one of claims 1-12 to be performed.

Images