Trimming die shearing failure analysis and early warning method and device based on product analysis

Abstract

The application provides a trimming die shearing fault analysis early warning method and device based on product analysis, relates to the technical field of fault early warning, and comprises the following steps: continuously collecting shearing product features output by a trimming die, and generating product feature time sequences; performing product abnormal mode positioning; when the product abnormal mode is instantaneous abnormality or no abnormality, performing qualified feature fluctuation rule analysis, and generating qualified feature fluctuation rule information; performing product feature one-way trend verification, and generating one-way trend verification results; and generating first reminding information to perform fault early warning. Through the application, the technical problem that shearing fault early warning is not timely and production quality is affected due to the lack of qualified product quality fluctuation analysis in the prior art can be solved. In the case of instantaneous abnormality or no abnormality, potential fault early warning is performed through qualified feature fluctuation rule analysis and one-way trend verification, so that the timeliness of shearing fault early warning is ensured, and the technical effect of ensuring production quality is achieved.

Description

Technical Field

[0001] This application relates to the field of fault early warning technology, and in particular to a method and device for fault analysis and early warning of shearing of cutting edge die based on product analysis. Background Technology

[0002] Trimming dies are widely used in metal processing, plastic molding, and other industries. However, due to various reasons, such as wear and improper operation, trimming dies may experience shearing failures, affecting product quality and equipment lifespan.

[0003] Shearing failures can affect product quality, so quality analysis of manufactured products can be performed to provide early warnings of failures. Existing shearing failure analysis methods often set product quality thresholds for anomaly analysis. However, even if a product meets quality standards, potential failures may still exist, leading to continuous changes in product quality until anomalies occur.

[0004] In summary, existing technologies suffer from the technical problem of untimely warnings of shearing faults due to the lack of quality fluctuation analysis for qualified products, which is caused by setting multiple product quality thresholds for product quality anomaly analysis. This results in a lack of timely quality fault warnings and affects production quality. Summary of the Invention

[0005] The purpose of this application is to provide a method and device for analyzing and warning of shearing faults in cutting dies based on product analysis, in order to solve the technical problem in the prior art that due to the multiple product quality thresholds set for product quality anomaly analysis, there is a lack of quality fluctuation analysis for qualified products, resulting in untimely warning of shearing faults and affecting production quality.

[0006] In view of the above problems, this application provides a method and device for analysis and early warning of shearing failure of cutting die based on product analysis.

[0007] Firstly, this application provides a product analysis-based method for analyzing and warning of shearing faults in a cutting die. This method is implemented using a product analysis-based device and includes: continuously acquiring shearing product features output by the cutting die to generate a product feature time sequence, wherein the product feature time sequence includes multiple product features, and each product feature includes product geometric features and product surface features; performing product quality inspection based on preset product features and the multiple product features; locating abnormal product modes based on the product quality inspection results, wherein the abnormal product modes include regular abnormalities, instantaneous abnormalities, or no abnormalities; when the abnormal product mode is the instantaneous abnormality or the absence of abnormalities, performing a qualified feature fluctuation pattern analysis based on the multiple product features to generate qualified feature fluctuation pattern information; verifying the unidirectional trend of product features based on the feature fluctuation pattern information to generate a unidirectional trend verification result; and generating a first warning message for fault warning based on the unidirectional trend verification result.

[0008] Secondly, this application also provides a product analysis-based shearing die fault analysis and early warning device for performing the product analysis-based shearing die fault analysis and early warning method as described in the first aspect, comprising: a product feature acquisition module for continuously acquiring shearing product features output by the shearing die and generating a product feature time sequence, wherein the product feature time sequence includes multiple product features, and any product feature includes product geometric features and product surface features; a product quality inspection module for performing product quality inspection in combination with preset product features and the multiple product features, and locating product abnormal modes based on the product quality inspection results, wherein the product abnormal modes include regular abnormalities, instantaneous abnormalities, or no abnormalities; a fluctuation pattern analysis module for performing qualified feature fluctuation pattern analysis based on the multiple product features when the product abnormal mode is the instantaneous abnormality or the no abnormality, and generating qualified feature fluctuation pattern information; a one-way trend verification module for performing one-way trend verification of product features based on the feature fluctuation pattern information, and generating a one-way trend verification result; and a first reminder module for generating a first reminder message for fault early warning based on the one-way trend verification result.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The system continuously collects shearing product features output from the cutting die, generating a product feature time sequence. This time sequence includes multiple product features, each containing both geometric and surface features. Product quality inspection is performed using preset product features and the multiple product features. Based on the inspection results, abnormal product modes are located, including regular anomalies, transient anomalies, or no anomalies. When the abnormal mode is a transient anomaly or no anomaly, a qualified feature fluctuation pattern analysis is performed based on the multiple product features, generating qualified feature fluctuation pattern information. A one-way trend verification of product features is performed based on this fluctuation pattern information, generating a one-way trend verification result. A first warning message is generated based on the one-way trend verification result to provide a fault early warning. By continuously collecting shearing product features output from the cutting die and generating a product feature time sequence, combined with preset product features for quality inspection and abnormal mode location, the shearing process can be monitored in real time. In cases of transient or no anomalies, potential fault early warnings are provided through qualified feature fluctuation pattern analysis and one-way trend verification, ensuring timely shearing fault warnings and thus guaranteeing production quality.

[0011] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating the shearing fault analysis and early warning method for edge trimming dies based on product analysis, as described in this application.

[0014] Figure 2 This is a schematic diagram of the structure of the edge-cutting die shearing fault analysis and early warning device based on product analysis in this application.

[0015] Explanation of reference numerals in the attached figures:

[0016] Product feature acquisition module 11, product quality inspection module 12, fluctuation pattern analysis module 13, unidirectional trend verification module 14, first reminder module 15. Detailed Implementation

[0017] This application provides a method and apparatus for analyzing and warning of shearing faults in edge-cutting dies based on product analysis. This solves the technical problem in existing technologies where multiple product quality thresholds are set for anomaly analysis, but quality fluctuation analysis of qualified products is lacking, leading to untimely shearing fault warnings and impacting production quality. By continuously collecting the shearing product characteristics output by the edge-cutting die and generating a product characteristic time sequence, combined with preset product characteristics for quality inspection and anomaly mode localization, the shearing process can be monitored in real time. In cases of instantaneous anomalies or the absence of anomalies, potential fault warnings are provided through analysis of qualified characteristic fluctuation patterns and unidirectional trend verification, ensuring timely shearing fault warnings and thus guaranteeing production quality.

[0018] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0019] Example 1, please refer to the appendix. Figure 1 This application provides a product analysis-based method for analyzing and warning of shearing faults in cutting dies. The method is applied to a product analysis-based device for analyzing and warning of shearing faults in cutting dies, and specifically includes the following steps:

[0020] Step 1: Continuously collect the shearing product features output by the cutting die to generate a product feature time sequence. The product feature time sequence includes multiple product features, and each product feature includes product geometric features and product surface features.

[0021] Specifically, a trimming die is an existing industrial mold used to shear the edges of materials. It is typically used in manufacturing processes to remove excess material from product edges, ensuring dimensional accuracy and appearance quality. In this process, the shearing characteristics output by the trimming die are continuously collected. That is, during the die shearing process, various characteristic data of the product are monitored and recorded in real time. The collected product characteristic data is arranged in chronological order, forming a data sequence containing multiple time points. This time-series data reflects the changes in product characteristics over time. In the product characteristic time sequence, each product characteristic includes two parts: product geometric characteristics and product surface characteristics. Product geometric characteristics refer to the product's dimensions, shape, angles, and other geometric parameters, which directly relate to the product's function and performance. Product surface characteristics involve the product's surface quality, such as surface roughness, defects, and color, which have a significant impact on the product's appearance and durability. Continuously collecting product characteristics provides a foundation for subsequent fault analysis and early warning.

[0022] Step 2: Perform product quality inspection by combining preset product features and the multiple product features, and locate the abnormal product modes based on the product quality inspection results. The abnormal product modes include regular abnormalities, transient abnormalities, or no abnormalities.

[0023] Specifically, preset product characteristics refer to ideal product characteristic values ​​pre-set according to product design and quality standards, including key parameters such as product geometry and surface quality. These serve as the standard and basis for product quality inspection. A comparative analysis of the product characteristic sequence with the preset product characteristics is conducted to assess whether the product quality meets the standards. The comparison of any product characteristic in the product characteristic sequence with the preset product characteristics yields the product quality inspection result. Based on the product quality inspection result, abnormal modal localization is performed. Abnormal modalities are categorized into three types: regular abnormality, transient abnormality, or no abnormality. Regular abnormality refers to product characteristics continuously deviating from the preset product characteristics over a period of time; that is, a large number of product characteristics in the product characteristic sequence do not meet the preset product characteristics. This may be caused by mold wear, improper equipment adjustment, etc. Transient abnormality refers to product characteristics suddenly deviating from the preset values ​​within a short period of time; that is, a very small number of product characteristics in the product characteristic sequence do not meet the preset product characteristics and can quickly return to normal automatically. No abnormality indicates that the product characteristics meet the preset standards, and the product quality is qualified. No abnormality means that all product characteristics in the product characteristic sequence meet the preset product characteristics. This abnormal modal localization provides support for subsequent fault analysis.

[0024] Step 3: When the product abnormality mode is the instantaneous abnormality or the absence of abnormality, perform qualified feature fluctuation pattern analysis based on the multiple product features to generate qualified feature fluctuation pattern information.

[0025] Specifically, when a product's abnormal mode is transient or non-abnormal, it means that the product's quality generally meets the preset standards, but there may be variations within the normal range, which may evolve into a regular abnormality over a certain period of time. In this case, conducting a fluctuation pattern analysis of acceptable characteristics is to further understand the changing trends and patterns of product characteristics within the acceptable range, thereby identifying potential faults and providing timely alerts.

[0026] Acceptable feature fluctuation pattern analysis involves statistically analyzing multiple product features to identify patterns in their changes over time. For example, time series analysis and trend analysis can be used to identify trends, periodicity, and randomness in product feature changes. The generated acceptable feature fluctuation pattern information includes fluctuation patterns in both geometric and surface features. By analyzing these patterns, a better understanding of the dynamic changes in product quality can be achieved, leading to the prediction and alerting of potential faults. For instance, if continuous fluctuations in the same direction are found in a product feature, it may continue to change in that direction, potentially causing quality problems.

[0027] Step 4: Based on the aforementioned characteristic fluctuation pattern information, verify the unidirectional trend of the product characteristics and generate the unidirectional trend verification result.

[0028] Specifically, verifying the unidirectional trend of product characteristics based on characteristic fluctuation patterns aims to confirm whether product characteristics are continuously changing in a direction of non-compliance, thereby providing early warning of potential quality problems and preventing product quality degradation. Specifically, based on the characteristic fluctuation pattern information, the characteristic fluctuation trend direction over two consecutive time series is analyzed to obtain multiple continuous characteristic fluctuation trend directions. It is then determined whether these multiple continuous characteristic fluctuation trend directions are consistent and changing in a direction of non-compliance, obtaining the unidirectional trend verification result. This facilitates early warning of potential quality problems, reminding production management personnel to promptly inspect and maintain production equipment, preventing product quality degradation.

[0029] Step 5: Generate a first alert message based on the unidirectional trend verification result to provide a fault warning.

[0030] Specifically, if the unidirectional trend verification result passes, it indicates that multiple continuous feature fluctuation trends are consistent and moving towards the direction of non-compliance, meaning that a quality decline may occur subsequently. At this point, a first alert message is generated to issue a fault warning, promptly reminding production management personnel to take measures to prevent the fault from occurring, thereby ensuring product quality and production efficiency. The first alert message can be sent to production management personnel via email, SMS, system notifications, etc.

[0031] Furthermore, step two of this application includes:

[0032] Determine whether the multiple product features meet the preset product features. If yes, the product abnormality mode is no abnormality. If the multiple product features do not meet the preset product features, locate the abnormal product set. Perform abnormal clustering analysis based on the abnormal product set to generate abnormal clustering indicators. Use the abnormal clustering indicators to identify regular anomalies and instantaneous anomalies to generate the product abnormality mode.

[0033] Furthermore, this application also includes the following steps:

[0034] Configure a regular clustering threshold; construct an abnormal modality identifier based on the regular clustering threshold, judge the abnormal clustering index, and output the product abnormal modality.

[0035] Specifically, the steps for product quality inspection based on preset product characteristics and multiple product characteristics, and for locating product anomaly modes based on the inspection results, include: comparing the actually collected multiple product characteristics with preset product characteristics. If multiple product characteristics fall within the range of preset product characteristics, i.e., meeting the preset requirements, then the product anomaly mode is considered normal, indicating that the product quality is qualified. If the preset requirements are not met, further analysis is required. When product characteristics do not meet the preset product characteristics, it is necessary to locate the products that do not meet the preset product characteristics to form an anomaly product set. Anomaly cluster analysis is then performed on the located anomaly product set, i.e., analyzing the number of anomaly products. For example, it may be found that products generally have problems within a certain time period, then the anomalies tend to be regularized, and the anomaly cluster index is high.

[0036] Specifically, the anomaly clustering index can be calculated by statistically analyzing the distribution of abnormal products. For example, it can be calculated as the proportion of the number of abnormal products in an anomaly product cluster to the number of products corresponding to multiple product characteristics. This can be used as an anomaly clustering index to quantify the severity and distribution characteristics of the anomalies.

[0037] Anomaly clustering indicators are used to distinguish between regular anomalies and transient anomalies. Regular anomalies typically manifest as a large number of abnormalities, while transient anomalies are random anomalies that suddenly appear within a short period of time. The specific identification steps include: configuring a regular clustering threshold, which is used to distinguish between regular and transient anomalies. For example, a threshold can be set; if the anomaly clustering indicator exceeds this threshold, it is considered a regular anomaly. That is, transient anomalies are generally caused by random factors, such as 1 out of 100 products being defective, while regular anomalies refer to a large number of defective products appearing, such as more than 20 out of 100 products being defective, in which case the shearing fault is obvious. Based on the set regular clustering threshold, an anomaly modality recognizer is constructed. The anomaly modality recognizer is a rule-based logical judgment system used to judge the anomaly clustering indicator. The anomaly modality recognizer is used to judge the anomaly clustering indicator. If the anomaly clustering indicator exceeds the regular clustering threshold, it is identified as a regular anomaly; if the anomaly clustering indicator does not exceed the regular clustering threshold but anomalies are present, it is identified as a transient anomaly. Based on the judgment results of the abnormal mode recognizer, the abnormal mode of the product is output, which can realize the automatic identification and classification of product quality problems, assist in the timely analysis of potential quality problems, and ensure the stability and consistency of product quality.

[0038] Furthermore, step three of this application includes:

[0039] Based on the preset product features, configure geometric feature boundaries and surface feature boundaries; perform boundary distance analysis on the multiple product features based on the geometric feature boundaries to generate a geometric boundary distance distribution trend sequence; perform boundary distance analysis on the multiple product features based on the surface feature boundaries to generate a surface boundary distance distribution trend sequence; and generate the qualified feature fluctuation pattern information using the geometric boundary distance distribution trend sequence and the surface boundary distance distribution trend sequence.

[0040] Specifically, based on preset product characteristics, geometric feature boundaries and surface feature boundaries are defined. These boundaries define the acceptable range for product features; that is, the product feature data should fluctuate within these boundaries. In other words, they represent the critical conditions for quality compliance for geometric and surface features within the preset product characteristics, such as the acceptable threshold for surface roughness. Based on the defined geometric feature boundaries, boundary distance analysis is performed on the geometric data of multiple product features. This involves analyzing and calculating the distances between multiple product features and their geometric feature boundaries, generating a geometric boundary distance distribution trend sequence. This sequence demonstrates the trend of the relative positions of product features and geometric boundaries changing over time.

[0041] Similarly, based on the defined surface feature boundaries, boundary distance analysis is performed on the surface data of multiple product features to calculate the distance between the product features and the surface feature boundaries, generating a surface boundary distance distribution trend sequence. The surface boundary distance distribution trend sequence shows the trend of the relative position between the product features and the surface boundaries changing over time.

[0042] Finally, the geometric boundary distance distribution trend sequence and the surface boundary distance distribution trend sequence are combined to generate qualified characteristic fluctuation pattern information, which facilitates the analysis of dynamic changes in product quality, prediction of potential quality problems, and alerts to potential faults.

[0043] Furthermore, step four of this application includes:

[0044] Based on the geometric boundary distance distribution trend sequence, a geometric distance distribution trend consistency check is performed to generate a first trend consistency index; based on the surface boundary distance distribution trend sequence, a surface distance distribution trend consistency check is performed to generate a second trend consistency index; if the first trend consistency index is greater than a preset consistency index and / or the second trend consistency index is greater than a preset consistency index, the unidirectional trend verification result is a successful verification.

[0045] Furthermore, this application also includes the following steps:

[0046] Based on the geometric boundary distance distribution trend sequence, extract multiple sets of boundary distance distribution trend combinations under adjacent time series; perform trend direction analysis on the multiple sets of boundary distance distribution trend combinations to generate trend change vector time series; configure a predetermined unidirectional trend, wherein the predetermined unidirectional trend is a direction tending towards the geometric feature boundary; compare the similarity between the continuous change direction of the trend change vector time series and the predetermined unidirectional trend to generate the first trend consistency index.

[0047] Specifically, the steps for verifying the unidirectional trend of product features include: First, performing consistency verification on the geometric boundary distance distribution trend sequence, that is, analyzing the degree of consistency of the change direction of geometric features relative to the preset product features based on the geometric boundary distance distribution trend sequence, such as continuously changing in the direction of non-compliance, and generating a first trend consistency index, which is used to quantify the consistency of the trend.

[0048] Specifically, the step of verifying the consistency of geometric distance distribution trends based on the geometric boundary distance distribution trend sequence and generating a first trend consistency index includes: extracting multiple sets of boundary distance distribution trend combinations under adjacent time series based on the geometric boundary distance distribution trend sequence. Each set of boundary distance distribution trend combinations contains two geometric boundary distance distribution trends corresponding to two adjacent time series. The direction and magnitude of trend change within each set of boundary distance distribution trend combinations are calculated to generate a trend change vector. The trend change vector can be represented as the direction and magnitude of distance change from one time series to the next. All trend change vectors are arranged in chronological order to generate a trend change vector time series.

[0049] Configure a predetermined unidirectional trend, which refers to the direction in which product feature changes should avoid, typically tending towards the geometric feature boundary, as this may indicate that the product feature is approaching a non-conforming boundary. This configuration is done by professionals in the field based on practical considerations. The continuous change direction of the trend change vector time series is compared with the predetermined unidirectional trend to assess similarity. If the continuous change direction of the trend change vector time series is similar to the predetermined unidirectional trend, it indicates that the product feature may be moving towards non-conformity. Specifically, existing similarity analysis methods, such as cosine similarity, can be used for comparative analysis to generate a first trend consistency index. This index quantifies the similarity between the trend change vector time series and the predetermined unidirectional trend. Through this analysis and index generation process, the stability of product quality can be more accurately assessed, particularly whether product features are moving towards non-conformity, facilitating the timely detection and resolution of potential quality problems and ensuring the stability and consistency of product quality.

[0050] Next, the consistency of the surface boundary distance distribution trend sequence is verified using the same method as the geometric distance distribution trend consistency verification, which will not be elaborated here, generating a second trend consistency index. This index is also used to quantify trend consistency.

[0051] Finally, the first trend consistency index and the second trend consistency index are compared with the preset consistency index. If the first trend consistency index is greater than the preset consistency index and / or the second trend consistency index is greater than the preset consistency index, the unidirectional trend verification result is considered to be verified as passed, indicating that the product feature change trend is moving in the direction of non-compliance, and there may be potential faults. At this time, an early warning is issued to ensure the cutting quality of the trimming die and prevent the quality from continuing to decline.

[0052] Furthermore, this application also includes the following steps:

[0053] When the product's abnormal mode is a regular abnormality, a second reminder message is generated; a fault warning is issued based on the second reminder message, wherein the warning urgency of the second reminder message is greater than that of the first reminder message.

[0054] Specifically, when a product's abnormality pattern is a regular anomaly, it indicates that the product characteristics have exceeded the preset product characteristics, and this change is significant, foreshadowing a more serious quality problem. In this case, a second alert is generated. This second alert can take the form of an emergency alarm, SMS, email, etc., ensuring that relevant personnel receive and understand the urgency of the information quickly. The urgency level of the second alert should be higher than that of the first alert. This is because regular anomalies are usually more serious than momentary anomalies or no anomalies, requiring a faster response and more urgent handling. Upon receiving the second alert, production management personnel should immediately take appropriate measures to resolve the issue, which may include stopping production, inspecting equipment, adjusting process parameters, and replacing parts. This early warning mechanism ensures that production management personnel give sufficient attention to regular anomalies and take swift action to prevent further deterioration of quality problems. Simultaneously, it also helps improve the controllability and efficiency of the production process, reducing scrap and rework rates.

[0055] In summary, the product analysis-based shearing fault analysis and early warning method for edge trimming dies provided in this application has the following technical effects:

[0056] The system continuously collects shearing product features output from the cutting die, generating a product feature time sequence. This time sequence includes multiple product features, each containing both geometric and surface features. Product quality inspection is performed using preset product features and the multiple product features. Based on the inspection results, abnormal product modes are located, including regular anomalies, transient anomalies, or no anomalies. When the abnormal mode is a transient anomaly or no anomaly, a qualified feature fluctuation pattern analysis is performed based on the multiple product features, generating qualified feature fluctuation pattern information. A one-way trend verification of product features is performed based on this fluctuation pattern information, generating a one-way trend verification result. A first warning message is generated based on the one-way trend verification result to provide a fault early warning. By continuously collecting shearing product features output from the cutting die and generating a product feature time sequence, combined with preset product features for quality inspection and abnormal mode location, the shearing process can be monitored in real time. In cases of transient or no anomalies, potential fault early warnings are provided through qualified feature fluctuation pattern analysis and one-way trend verification, ensuring timely shearing fault warnings and thus guaranteeing production quality.

[0057] Example 2, based on the same inventive concept as the product analysis-based shearing fault analysis and early warning method for cutting dies in the previous examples, please refer to the appendix. Figure 2 This application also provides a product analysis-based shearing fault analysis and early warning device for edge trimming dies, including:

[0058] The product feature acquisition module 11 is used to continuously acquire the sheared product features output by the cutting die and generate a product feature time sequence. The product feature time sequence includes multiple product features, and each product feature includes product geometric features and product surface features.

[0059] Product quality inspection module 12 is used to perform product quality inspection by combining preset product features and the multiple product features, and to locate product abnormal modes based on the product quality inspection results, wherein the product abnormal modes include regular abnormalities, instantaneous abnormalities or no abnormalities.

[0060] The fluctuation pattern analysis module 13 is used to perform qualified feature fluctuation pattern analysis based on the multiple product features when the product abnormality mode is the instantaneous abnormality or the absence of abnormality, and generate qualified feature fluctuation pattern information.

[0061] The one-way trend verification module 14 is used to verify the one-way trend of product features based on the feature fluctuation pattern information and generate a one-way trend verification result.

[0062] The first reminder module 15 is used to generate a first reminder message for fault warning based on the unidirectional trend verification result.

[0063] Furthermore, the product quality inspection module 12 is also used for:

[0064] Determine whether the multiple product features meet the preset product features. If yes, the product abnormality mode is no abnormality. If the multiple product features do not meet the preset product features, locate the abnormal product set. Perform abnormal clustering analysis based on the abnormal product set to generate abnormal clustering indicators. Use the abnormal clustering indicators to identify regular anomalies and instantaneous anomalies to generate the product abnormality mode.

[0065] Furthermore, the product quality inspection module 12 is also used for:

[0066] Configure a regular clustering threshold; construct an abnormal modality identifier based on the regular clustering threshold, judge the abnormal clustering index, and output the product abnormal modality.

[0067] Furthermore, the fluctuation pattern analysis module 13 is also used for:

[0068] Configure geometric feature boundaries and surface feature boundaries based on the preset product features;

[0069] Based on the geometric feature boundaries, boundary distance analysis is performed on the multiple product features to generate a geometric boundary distance distribution trend sequence; based on the surface feature boundaries, boundary distance analysis is performed on the multiple product features to generate a surface boundary distance distribution trend sequence; the qualified feature fluctuation pattern information is generated using the geometric boundary distance distribution trend sequence and the surface boundary distance distribution trend sequence.

[0070] Furthermore, the unidirectional trend verification module 14 is also used for:

[0071] Based on the geometric boundary distance distribution trend sequence, a geometric distance distribution trend consistency check is performed to generate a first trend consistency index; based on the surface boundary distance distribution trend sequence, a surface distance distribution trend consistency check is performed to generate a second trend consistency index; if the first trend consistency index is greater than a preset consistency index and / or the second trend consistency index is greater than a preset consistency index, the unidirectional trend verification result is a successful verification.

[0072] Furthermore, the unidirectional trend verification module 14 is also used for:

[0073] Based on the geometric boundary distance distribution trend sequence, extract multiple sets of boundary distance distribution trend combinations under adjacent time series; perform trend direction analysis on the multiple sets of boundary distance distribution trend combinations to generate trend change vector time series; configure a predetermined unidirectional trend, wherein the predetermined unidirectional trend is a direction tending towards the geometric feature boundary; compare the similarity between the continuous change direction of the trend change vector time series and the predetermined unidirectional trend to generate the first trend consistency index.

[0074] Furthermore, the product analysis-based edge-cutting die shearing fault analysis and early warning device also includes a second reminder module, which is used for:

[0075] When the product's abnormal mode is a regular abnormality, a second reminder message is generated; a fault warning is issued based on the second reminder message, wherein the warning urgency of the second reminder message is greater than that of the first reminder message.

[0076] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1The product analysis-based shearing fault analysis and early warning method and specific examples in Example 1 are also applicable to the product analysis-based shearing fault analysis and early warning device for shearing dies in this embodiment. Through the foregoing detailed description of the product analysis-based shearing fault analysis and early warning method for shearing dies, those skilled in the art can clearly understand the product analysis-based shearing fault analysis and early warning device for shearing dies in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here. As for the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant details can be found in the method section.

[0077] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0078] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for analyzing and warning of shearing failures in edge trimming dies based on product analysis, characterized in that, include: The shearing product features output by the cutting die are continuously collected to generate a product feature time sequence, wherein the product feature time sequence includes multiple product features, and each product feature includes product geometric features and product surface features; Product quality inspection is performed by combining preset product features and the multiple product features. Based on the product quality inspection results, abnormal product modes are located, wherein the abnormal product modes include regular abnormalities, transient abnormalities, or no abnormalities. When the product abnormality mode is the instantaneous abnormality or the absence of abnormality, the qualified feature fluctuation pattern analysis is performed based on the multiple product characteristics to generate qualified feature fluctuation pattern information. Based on the aforementioned characteristic fluctuation pattern information, the product characteristic unidirectional trend is verified, and a unidirectional trend verification result is generated. Based on the unidirectional trend verification result, a first reminder message is generated to provide a fault warning. Product quality inspection is performed by combining preset product features and the multiple product features, and abnormal modal localization of the product is performed based on the product quality inspection results, including: Determine whether the multiple product features meet the preset product features; if so, the product abnormality mode is no abnormality. If the multiple product characteristics do not meet the preset product characteristics, locate the abnormal product set; Anomaly clustering analysis is performed on the aforementioned abnormal product set to generate anomaly clustering indicators. The abnormal clustering indicators are used to identify regular and transient anomalies, and the product anomaly modality is generated. The identification of regular and transient anomalies using the aforementioned anomaly aggregation indicators, and the generation of the product anomaly modality, includes: Configure a regular clustering threshold; An abnormal modality identifier is constructed based on the aforementioned regularity clustering threshold, which judges the abnormal clustering index and outputs the product abnormal modality. Based on the aforementioned multiple product characteristics, a fluctuation pattern analysis of qualification characteristics is performed to generate qualification characteristic fluctuation pattern information, including: Configure geometric feature boundaries and surface feature boundaries based on the preset product features; Based on the geometric feature boundaries, perform boundary distance analysis on the multiple product features to generate a geometric boundary distance distribution trend sequence; Based on the surface feature boundaries, boundary distance analysis is performed on the multiple product features to generate a surface boundary distance distribution trend sequence. The qualified feature fluctuation pattern information is generated using the geometric boundary distance distribution trend sequence and the surface boundary distance distribution trend sequence; Based on the aforementioned characteristic fluctuation pattern information, a one-way trend verification of product characteristics is performed, generating a one-way trend verification result, including: Based on the geometric boundary distance distribution trend sequence, a geometric distance distribution trend consistency check is performed to generate a first trend consistency index; Based on the surface boundary distance distribution trend sequence, a surface distance distribution trend consistency check is performed to generate a second trend consistency index; If the first trend consistency indicator is greater than the preset consistency indicator and / or the second trend consistency indicator is greater than the preset consistency indicator, the unidirectional trend verification result is verified as passed; Based on the geometric boundary distance distribution trend sequence, a geometric distance distribution trend consistency check is performed to generate a first trend consistency index, including: Extract multiple sets of boundary distance distribution trend combinations under adjacent time series based on the geometric boundary distance distribution trend sequence; The trend of the multiple sets of boundary distance distribution trend combinations is analyzed to generate a time series of trend change vectors; Configure a predetermined unidirectional trend, wherein the predetermined unidirectional trend is a direction that tends toward the geometric feature boundary; By comparing the continuous change direction of the trend change vector time series with the similarity of the predetermined unidirectional trend, the first trend consistency index is generated.

2. The method for analyzing and warning of shearing faults in edge-cutting dies based on product analysis as described in claim 1, characterized in that, Also includes: When the product's abnormal mode is a regular abnormality, a second reminder message is generated; Fault warnings are issued based on the second reminder information, wherein the urgency of the second reminder information is greater than that of the first reminder information.

3. A shearing fault analysis and early warning device for edge trimming dies based on product analysis, characterized in that, The steps for implementing the product analysis-based shearing fault analysis and early warning method for edge cutting dies according to any one of claims 1 to 2 include: The product feature acquisition module is used to continuously acquire the sheared product features output by the cutting die and generate a product feature time sequence. The product feature time sequence includes multiple product features, and each product feature includes product geometric features and product surface features. The product quality inspection module is used to perform product quality inspection by combining preset product features and the multiple product features, and to locate the abnormal mode of the product based on the product quality inspection results. The abnormal mode of the product includes regular abnormality, instantaneous abnormality or no abnormality. The fluctuation pattern analysis module is used to perform qualified feature fluctuation pattern analysis based on the multiple product characteristics when the product abnormality mode is the instantaneous abnormality or the absence of abnormality, and generate qualified feature fluctuation pattern information. The one-way trend verification module is used to verify the one-way trend of product features based on the characteristic fluctuation pattern information and generate the one-way trend verification result. The first reminder module is used to generate a first reminder message for fault warning based on the unidirectional trend verification result.

Images