Forging piece defect detection method and device based on digital analysis
The forging defect detection method constructed by multi-source data fusion and deep learning algorithms solves the problem of large defect identification error in the existing technology, realizes high-precision and efficient defect detection and process optimization, and meets the needs of automated production.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XUZHOU YIZHONG FORGING EQUIP
- Filing Date
- 2025-12-24
- Publication Date
- 2026-05-29
AI Technical Summary
Existing methods for detecting defects in forgings lack efficient fusion analysis of multi-source data and deep learning and intelligent recognition mechanisms based on digital models, resulting in significant errors in identification and judgment, and failing to meet the needs of high-precision, high-efficiency, and automated production.
By acquiring data from multiple sources and aligning it in time and space, a three-dimensional digital structure model is constructed. This model is then combined with deep learning algorithms for defect identification, generating digital defect archives. Process feedback optimization is then performed to achieve comprehensive, accurate identification and dynamic monitoring.
It achieves comprehensive and accurate identification and dynamic monitoring of defects in forgings, realizing multi-source data fusion, digital model-driven deep learning intelligent identification and real-time process optimization feedback, thereby improving production efficiency and quality control.
Smart Images
Figure CN122109482A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of defect detection technology, and in particular to a method and apparatus for detecting defects in forgings based on digital analysis. Background Technology
[0002] With the increasing demands for quality in modern manufacturing, traditional methods for detecting defects in forgings are no longer sufficient to meet the requirements of high precision, high efficiency, and intelligent operation.
[0003] Currently, most existing defect detection methods rely on human experience or are based on a single detection technology, such as ultrasonic testing or X-ray testing. These methods have weak ability to identify complex defects, especially when dealing with hidden or minute defects, where their accuracy and reliability are limited. Secondly, traditional detection methods are insufficient in the comprehensive analysis of multi-source data and in-depth exploration of defect formation mechanisms. They lack an effective intelligent analysis tool that integrates interdisciplinary and multi-dimensional data, making it impossible to comprehensively and accurately identify and judge defect types and their root causes.
[0004] In summary, existing technologies suffer from significant errors in identifying and judging defects in forging parts due to a lack of efficient fusion and analysis of multi-source data, a lack of deep learning and intelligent recognition mechanisms based on digital models, and a lack of adaptive process feedback and optimization control methods. These errors fail to accurately reflect the actual distribution and formation process of defects, further affecting the quality control and production efficiency of the forging process, and making it difficult to meet the technical requirements of high-precision, high-efficiency, and automated production. Summary of the Invention
[0005] The purpose of this application is to provide a method and apparatus for detecting defects in forging parts based on digital analysis, in order to solve the technical problems in the prior art. Due to the lack of efficient fusion analysis of multi-source data, the lack of deep learning and intelligent recognition mechanisms based on digital models, and the lack of adaptive process feedback and optimization control methods, the identification and judgment of defects in forging parts have large errors, which cannot accurately reflect the actual defect distribution and formation process, further affecting the quality control and production efficiency of the forging process, and making it difficult to meet the requirements of high precision, high efficiency and automated production.
[0006] In view of the above problems, this application provides a method and apparatus for detecting defects in forgings based on digital analysis.
[0007] In a first aspect, this application provides a method for detecting defects in forgings based on digital analysis, implemented by a device for detecting defects in forgings based on digital analysis. The method includes: acquiring multi-source data from the target forging; performing spatiotemporal alignment processing on the multi-source data to obtain standard data; performing anomaly analysis on a digital model obtained based on the standard data to generate candidate defect regions; analyzing the defect formation stage of the candidate defect regions and analyzing the defect evolution trend to generate a defect acceptance judgment; outputting a defect judgment result based on the defect acceptance judgment and generating a digital defect archive; and performing process feedback on the digital defect archive.
[0008] Preferably, the method for detecting defects in forgings based on digital analysis further includes: acquiring surface geometric morphology data through laser scanning; acquiring internal structure detection data through ultrasonic testing; synchronously collecting processing process data by the process monitoring units of the forging equipment and the heat treatment equipment; and acquiring the multi-source data based on the surface geometric morphology data, internal structure detection data, and processing process data.
[0009] Preferably, the method for detecting defects in forgings based on digital analysis further includes: constructing a three-dimensional digital structural model of the target forging based on the standard data; extracting structural feature parameters of the target forging to generate a multi-dimensional feature vector; and combining the three-dimensional digital structural model and the multi-dimensional feature vector to obtain the digital model.
[0010] Preferably, the method for detecting defects in forgings based on digital analysis further includes: inputting the digital model into the defect identification channel for feature anomaly comparison analysis to obtain abnormal forging structure features; and performing spatial clustering and boundary segmentation on the abnormal forging structure features to generate the defect candidate region.
[0011] Preferably, the method for detecting defects in forgings based on digital analysis further includes: collecting digital data of multiple batches of forgings that have been inspected and labeled; constructing a training sample set based on the digital data of the multiple forgings, and supervising the labeling of the training sample set according to a preset defect type; training the model of the training sample set using a deep learning algorithm to construct the defect recognition channel; evaluating the stability of the defect recognition channel through cross-validation, and obtaining the defect recognition channel when the evaluation result meets a preset stability threshold.
[0012] Preferably, the forging defect detection method based on digital analysis further includes: performing defect pattern matching analysis on the defect candidate region to obtain defect matching results; and performing correlation analysis on the defect matching results based on processing parameters to determine the defect formation stage.
[0013] Preferably, the method for detecting defects in forgings based on digital analysis further includes: constructing a defect influence factor based on the defect formation stage, and assessing the severity of the defect based on the defect influence factor; analyzing the defect evolution trend based on the defect severity, and generating a defect acceptance judgment based on the defect evolution trend.
[0014] Preferably, the forging defect detection method based on digital analysis further includes: outputting a defect judgment result based on the defect acceptance judgment, including defect location, defect type, defect level and risk assessment result; and associating and archiving the standard data, the defect candidate area and the defect judgment result to generate the defect digital file.
[0015] Preferably, the method for detecting defects in forgings based on digital analysis further includes: analyzing the defect formation process parameters in reverse according to the defect digital archive, generating optimization suggestions; and feeding the optimization suggestions back to the forging process control system for adaptive adjustment.
[0016] Secondly, this application also provides a forging defect detection device based on digital analysis, used to execute the forging defect detection method based on digital analysis as described in the first aspect, comprising: a standard data acquisition module, used to acquire multi-source data from the target forging and perform spatiotemporal alignment processing on the multi-source data to obtain standard data; a defect candidate region generation module, used to perform anomaly analysis based on the digital model obtained from the standard data to generate defect candidate regions; a defect acceptance judgment generation module, used to analyze the defect formation stage of the defect candidate regions and analyze the defect evolution trend to generate a defect acceptance judgment; a defect digital archive generation module, used to output defect judgment results based on the defect acceptance judgment and generate a defect digital archive; and a process feedback execution module, used to execute process feedback of the defect digital archive.
[0017] The technical solution provided in this application has at least the following technical effects or advantages: by realizing comprehensive and accurate identification and dynamic monitoring of defects in forging parts, it achieves the technical effects of deep learning intelligent identification based on multi-source data fusion and digital model-driven, as well as real-time process optimization feedback.
[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of 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
[0019] 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.
[0020] Figure 1 This is a flowchart illustrating the defect detection method for forging parts based on digital analysis proposed in this application.
[0021] Figure 2 This is a schematic diagram of the forging defect detection device based on digital analysis proposed in this application.
[0022] Figure labeling: Standard data acquisition module 1, defect candidate area generation module 2, defect acceptance judgment generation module 3, defect digital archive generation module 4, process feedback execution module 5. Detailed Implementation
[0023] This application provides a method and apparatus for detecting defects in forgings based on digital analysis. It addresses the technical problems in existing technologies where the lack of efficient fusion and analysis of multi-source data, the absence of deep learning and intelligent recognition mechanisms based on digital models, and the lack of adaptive process feedback and optimization control methods lead to significant errors in the identification and judgment of defects in forgings. These errors fail to accurately reflect the actual defect distribution and formation process, further impacting the quality control and production efficiency of the forging process and making it difficult to meet the demands of high-precision, high-efficiency, and automated production. The method achieves comprehensive and accurate identification and dynamic monitoring of defects in forgings, realizing the technical effects of multi-source data fusion, deep learning intelligent recognition driven by digital models, and real-time process optimization feedback.
[0024] 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.
[0025] Example 1, please refer to the appendix. Figure 1 This application provides a method for detecting defects in forgings based on digital analysis, which is applied to a device for detecting defects in forgings based on digital analysis. The method specifically includes the following steps: Multi-source data is collected from the target forging part, and the multi-source data is subjected to spatiotemporal alignment processing to obtain standard data.
[0026] Furthermore, this application also includes: acquiring surface geometric morphology data through laser scanning; acquiring internal structure detection data through ultrasonic testing; synchronously collecting processing process data by the process monitoring unit of the forging equipment and the heat treatment equipment; and acquiring the multi-source data based on the surface geometric morphology data, internal structure detection data, and processing process data.
[0027] Specifically, surface geometric topography data is acquired through laser scanning. A laser beam is emitted onto the surface of the target forging using a laser scanning device, and the echo signal formed after the laser beam is reflected or scattered on the surface of the forging is received. Based on the time difference, phase change, or displacement offset relationship between laser emission and reception, the three-dimensional contour information of the forging surface is reconstructed. The surface geometric topography data is used to characterize the size distribution, curvature change, surface undulation, and local discontinuity features of the forging surface, thereby providing fine surface structure basic data for subsequent defect identification.
[0028] Based on the obtained surface geometric morphology data, the internal structure detection data is obtained through ultrasonic testing. This refers to using an ultrasonic transducer to emit ultrasonic signals into the interior of the forging and receiving the reflection, refraction or attenuation signals generated by the ultrasonic waves during propagation within the material due to the interface of the structure, the boundary of the defect or the difference of the medium. According to the echo amplitude, propagation time and spectral characteristics, the density, structural continuity and the distribution of potential internal defects inside the forging are characterized, thereby forming internal structure detection data that can reflect the internal structural state of the forging.
[0029] Furthermore, the synchronous acquisition of processing data by the process monitoring units of the forging and heat treatment equipment refers to the real-time acquisition of process parameters such as forging pressure, deformation rate, forming temperature, holding time, and cooling rate by sensors or monitoring modules installed on the forging and heat treatment equipment during the forging and subsequent heat treatment processes. These parameters are then recorded synchronously according to a unified time reference to form processing data that reflects the entire process of forging and microstructure evolution, thereby providing process background information for defect formation mechanism analysis.
[0030] The acquisition of multi-source data based on surface geometry data, internal structure detection data, and processing data refers to the collection of data from different detection methods and acquisition channels to form a comprehensive data set containing surface structure information, internal organization information, and processing history information. Multi-source data is used to reflect the true state of the forging from multiple dimensions, avoiding information loss or judgment bias caused by a single data source.
[0031] After obtaining the multi-source data, spatiotemporal alignment processing is performed on the multi-source data to obtain standard data. This means that by establishing a unified time axis and spatial coordinate reference system, registration processing is performed on multi-source data from different sources, with different sampling frequencies and different spatial resolutions, so that various types of data have corresponding relationships at the same time node and the same spatial location. Furthermore, the differences between data are eliminated through scale normalization and format conversion, thereby generating standard data that can be directly used for subsequent digital modeling and anomaly analysis.
[0032] Anomaly analysis is performed on the digital model obtained based on the standard data to generate candidate defect regions.
[0033] Furthermore, this application also includes: constructing a three-dimensional digital structural model of the target forging based on the standard data; extracting structural feature parameters of the target forging to generate a multi-dimensional feature vector; and combining the three-dimensional digital structural model and the multi-dimensional feature vector to obtain the digital model.
[0034] Furthermore, this application also includes: inputting the digital model into the defect identification channel for feature anomaly comparison analysis to obtain abnormal forging structure features; performing spatial clustering and boundary segmentation on the abnormal forging structure features to generate the defect candidate region.
[0035] Furthermore, this application also includes: collecting digital data of multi-source forging parts that have completed multiple batches of inspection and labeling; constructing a training sample set based on the digital data of the multi-source forging parts, and supervising the labeling of the training sample set according to a preset defect type; training the model of the training sample set using a deep learning algorithm to construct the defect identification channel; evaluating the stability of the defect identification channel through cross-validation, and obtaining the defect identification channel when the evaluation result meets a preset stability threshold.
[0036] Specifically, constructing a three-dimensional digital structural model of the target forging based on standard data refers to using standard data as a unified data input basis after completing the spatiotemporal alignment and standardization of multi-source data. Then, through a three-dimensional reconstruction algorithm, the overall geometric shape, local structural morphology, and internal spatial relationships of the target forging are digitally expressed, thereby forming a three-dimensional digital structural model that can reflect the true structural state of the forging. The three-dimensional digital structural model is used to describe the structural integrity and morphological characteristics of the forging in the spatial dimension.
[0037] Based on the completion of the three-dimensional digital structural model, the structural feature parameters of the target forging are extracted and multi-dimensional feature vectors are generated. This means analyzing the three-dimensional digital structural model and its corresponding standard data to obtain various feature parameters used to characterize the structural state of the forging. The structural feature parameters include at least surface roughness parameters, geometric deviation parameters, internal signal response parameters, and microstructure continuity indicators. These various structural feature parameters are then combined according to preset feature dimensions to form a multi-dimensional feature vector that can comprehensively reflect the structural characteristics of the forging.
[0038] After obtaining the multidimensional feature vectors, the digital model is obtained by combining the three-dimensional digital structure model and the multidimensional feature vectors. This means that the three-dimensional digital structure model used to describe the spatial structural relationship of the forging is associated and fused with the multidimensional feature vectors used to describe the distribution state of structural features. This enables the digital model to have both spatial geometric expression capabilities and feature parameter expression capabilities, thereby constructing a unified digital analysis object that can be used for subsequent anomaly analysis, defect identification, and evolution assessment.
[0039] Furthermore, the digital model is input into the defect identification channel for feature anomaly comparison analysis to obtain abnormal forging structure characteristics. This means that after obtaining a digital model containing spatial structural information and structural feature distribution information of the forging part, the digital model is input into a pre-constructed defect identification channel as the analysis object. The defect identification channel compares and analyzes the multi-dimensional feature vectors in the digital model with the feature distribution patterns corresponding to the normal forging structure, thereby identifying feature information that deviates significantly from the normal forging structure in terms of geometric morphology, internal response, or structural continuity. The deviation features are defined as abnormal forging structure characteristics, which are used to characterize the initial manifestation of potential defects.
[0040] The method for constructing a defect identification channel includes: collecting multi-source digitized data of forging parts from multiple batches that have been inspected and labeled. During the actual production and inspection of forging parts, for different production batches of forging parts, corresponding digitized data is obtained through various inspection methods such as surface inspection, internal flaw detection, and processing monitoring. While acquiring the data, the defect type, location, or state is labeled. Multiple batches are used to characterize the diversity of samples in terms of process parameters, material state, and defect morphology. Multi-source digitized data of forging parts is used to characterize the comprehensive information of the same forging part under different inspection dimensions, thus providing a representative and scalable data foundation for subsequent model training. Furthermore, a training sample set is constructed based on the multi-source digitized data of forging parts, and supervised labeling is performed on the training sample set according to preset defect types. This involves organizing, filtering, and structuring the collected multi-source digitized data to form a sample set that can be directly used for model training. Based on a predefined defect classification system, each training sample is explicitly assigned a corresponding defect category label. The supervised labeling is used to indicate the correspondence between input data and target output during model learning, ensuring that the model can accurately distinguish different types of forging defect features.
[0041] Subsequently, a deep learning algorithm is used to train the model on the training sample set to construct a defect identification channel. This means that a deep neural network structure with multi-layer feature extraction capabilities is used to iteratively learn the training sample set that has been supervised and labeled. By continuously adjusting the internal parameters of the model, the model can automatically extract key defect discrimination features from the digital data of multi-source forging parts, thereby forming a defect identification channel that is used to input the digital model and output defect identification results.
[0042] Furthermore, the stability of the defect identification channel is evaluated through cross-validation. When the evaluation result meets the preset stability threshold, the defect identification channel is obtained. This means that after the model training is completed, the training sample set is divided into multiple subsets, and the model is validated multiple times in a rotating manner to evaluate the recognition consistency and generalization ability of the defect identification channel on different data subsets. When the results of multiple rounds of validation all reach the preset stability evaluation criteria, the defect identification channel is confirmed to have reliable recognition performance and is output as the final usable model.
[0043] Based on the obtained abnormal forging structure characteristics, spatial clustering and boundary segmentation are performed on the abnormal forging structure characteristics to generate defect candidate regions. This means that according to the spatial distribution relationship of abnormal forging structure characteristics in the three-dimensional digital structure model, abnormal feature points with similar abnormal features and spatial proximity are clustered, and the spatial boundary range of each abnormal feature cluster is further determined by the boundary segmentation algorithm, thereby forming several continuous regional units in the digital model. The regional units serve as defect candidate regions where defects may exist, and are used for subsequent defect type determination and defect formation mechanism analysis.
[0044] The defect formation stage of the defect candidate region is analyzed, and the defect evolution trend is analyzed to generate a defect acceptance judgment.
[0045] Furthermore, this application also includes: performing defect pattern matching analysis on the defect candidate region to obtain defect matching results; and performing correlation analysis on the defect matching results based on processing parameters to determine the defect formation stage.
[0046] Furthermore, this application also includes: constructing a defect impact factor based on the defect formation stage, and assessing the severity of the defect based on the defect impact factor; analyzing the defect evolution trend based on the defect severity, and generating the defect acceptance judgment based on the defect evolution trend.
[0047] Specifically, performing defect pattern matching analysis on defect candidate regions to obtain defect matching results refers to comparing and analyzing the defect candidate regions obtained through spatial clustering and boundary segmentation with a pre-established defect pattern library in terms of morphological features, scale distribution, gray-scale or echo response characteristics, and spatial structural relationships. The defect patterns are used to characterize the typical manifestations of different types of forging defects in the digital model. The similarity between the defect candidate regions and various defect patterns is determined through matching calculations, thereby outputting defect matching results that reflect the defect type and matching confidence.
[0048] Furthermore, based on the correlation analysis of the defect matching results with the processing parameters, the defect formation stage is determined. This means that the defect matching results are correlated with the processing parameters such as temperature, pressure, deformation rate and time nodes collected during forging, heat treatment or subsequent processing. By analyzing the correlation between defect characteristics and parameter changes in each process stage, the key process intervals or time stages in which the defect occurs are identified, thereby determining the formation stage of the defect in the entire processing flow.
[0049] Constructing defect impact factors based on the defect formation stage and assessing defect severity based on these impact factors involves, after determining the specific processing stage corresponding to the defect, summarizing and modeling factors such as process conditions, load states, thermo-mechanical coupling characteristics, and material microstructure response within that stage to form defect impact factors that characterize the sensitivity of defect generation and propagation. These impact factors quantify the degree to which defects are affected by process parameter fluctuations and material state changes at a specific formation stage. Through comprehensive calculation and weight analysis of each impact factor, the severity of defects in terms of size, morphological stability, and impact on the service performance of forgings is assessed.
[0050] Furthermore, analyzing the defect evolution trend based on the severity of the defect and generating a defect acceptance judgment based on the defect evolution trend means that after obtaining the defect severity evaluation result, combined with the repeated loads, thermal cycles or environmental effects that the defect may be subjected to during subsequent processing or use, predicting and analyzing the trend of defect expansion, merging or stabilization, and making an acceptance judgment on whether the defect meets the use or delivery requirements based on preset quality control standards or use safety thresholds, thereby generating the defect acceptance judgment result.
[0051] Based on the defect acceptance determination, the defect determination result is output, and a defect digital file is generated.
[0052] Furthermore, this application also includes: outputting a defect determination result based on the defect acceptance determination, including defect location, defect type, defect level, and risk assessment result; associating and archiving the standard data, the defect candidate area, and the defect determination result to generate the defect digital file.
[0053] Specifically, the defect acceptance determination output, including defect location, defect type, defect level, and risk assessment results, refers to the output of the determination conclusion in the form of structured information after the defect acceptability determination is completed. The defect location is used to characterize the specific coordinates or area of the defect in the three-dimensional space of the forging, the defect type is used to identify the corresponding organizational abnormality or structural defect category, the defect level is used to reflect the degree of impact of the defect on the mechanical properties and service safety of the forging, and the risk assessment results are used to comprehensively assess the probability and consequences of failure that the defect may cause in subsequent processing or service, thus forming a complete defect determination result.
[0054] Furthermore, the standard data, defect candidate regions, and defect judgment results are linked and archived to generate a defect digital archive. This means that the multi-source standard data used for analysis, the defect candidate region information obtained by spatial clustering and boundary segmentation, and the final output defect judgment results are linked according to a unified data indexing rule and stored in a digital storage system. This creates a corresponding defect digital archive for each forging part. The defect digital archive is used to support defect traceability, quality statistical analysis, and subsequent process optimization decisions.
[0055] The process feedback of the defect digital archive is executed.
[0056] Furthermore, this application also includes: analyzing the defect formation process parameters in reverse based on the defect digital archive, generating optimization suggestions; and feeding the optimization suggestions back to the forging process control system for adaptive adjustment.
[0057] Specifically, based on the defect digital archive, reverse analysis of defect formation process parameters generates optimization suggestions. This involves storing detailed defect data in the defect digital archive, including information such as defect location, type, severity, and formation stage. By analyzing the relationship between the causes of defects and relevant process parameters, key process parameters that may lead to defects are derived, such as forging temperature, deformation rate, and pressure distribution. Through reverse analysis of these process parameters, it is determined which process variables may need adjustment, thereby proposing optimization suggestions. The aim is to reduce or eliminate the occurrence of similar defects in future production by improving process parameters.
[0058] Furthermore, feeding optimization suggestions back to the forging process control system for adaptive adjustment refers to inputting the optimization suggestions generated through defect formation analysis into the forging process control system. This control system typically includes automation modules, sensors, and actuators for adjusting equipment parameters. Based on the optimization suggestions, process control parameters are automatically adjusted, such as the operating temperature, pressure, and deformation rate of the forging equipment, to ensure that process conditions align with defect suppression targets in subsequent production. This achieves adaptive adjustment of the production process, thereby improving the quality and production efficiency of forged parts.
[0059] In summary, the digital analysis-based defect detection method for forgings provided in this application has the following technical effects: by realizing comprehensive and accurate identification and dynamic monitoring of defects in forgings, it achieves the technical effects of deep learning intelligent identification based on multi-source data fusion and digital model-driven approach, as well as real-time process optimization feedback.
[0060] Example 2: Based on the same inventive concept as the forging defect detection method based on digital analysis in the previous examples, this application also provides a forging defect detection device based on digital analysis. Please refer to the appendix. Figure 2The system includes: a standard data acquisition module 1, used to acquire multi-source data from the target forging and perform spatiotemporal alignment processing on the multi-source data to obtain standard data; a defect candidate region generation module 2, used to perform anomaly analysis based on the digital model obtained from the standard data to generate defect candidate regions; a defect acceptance judgment generation module 3, used to analyze the defect formation stage of the defect candidate regions and analyze the defect evolution trend to generate a defect acceptance judgment; a defect digital archive generation module 4, used to output the defect judgment result based on the defect acceptance judgment and generate a defect digital archive; and a process feedback execution module 5, used to execute the process feedback of the defect digital archive.
[0061] Furthermore, the forging defect detection device based on digital analysis is also used to: acquire surface geometric morphology data through laser scanning; acquire internal structure detection data through ultrasonic testing; synchronously collect processing process data by the process monitoring unit of the forging equipment and the heat treatment equipment; and acquire the multi-source data based on the surface geometric morphology data, internal structure detection data, and processing process data.
[0062] Furthermore, the forging defect detection device based on digital analysis is also used to: construct a three-dimensional digital structural model of the target forging based on the standard data; extract the structural feature parameters of the target forging to generate a multi-dimensional feature vector; and combine the three-dimensional digital structural model and the multi-dimensional feature vector to obtain the digital model.
[0063] Furthermore, the forging defect detection device based on digital analysis is also used to: input the digital model into the defect identification channel for feature anomaly comparison analysis to obtain abnormal forging structure features; and perform spatial clustering and boundary segmentation on the abnormal forging structure features to generate the defect candidate region.
[0064] Furthermore, the forging defect detection device based on digital analysis is also used for: collecting digital data of multiple batches of forgings that have completed inspection and labeling; constructing a training sample set based on the digital data of the multiple forgings, and supervising the labeling of the training sample set according to a preset defect type; training the model of the training sample set using a deep learning algorithm to construct the defect recognition channel; evaluating the stability of the defect recognition channel through cross-validation, and obtaining the defect recognition channel when the evaluation result meets a preset stability threshold.
[0065] Furthermore, the forging defect detection device based on digital analysis is also used to: perform defect pattern matching analysis on the defect candidate area to obtain defect matching results; and perform correlation analysis on the defect matching results based on processing parameters to determine the defect formation stage.
[0066] Furthermore, the forging defect detection device based on digital analysis is also used to: construct a defect influence factor based on the defect formation stage, and assess the severity of the defect based on the defect influence factor; analyze the defect evolution trend according to the defect severity, and generate the defect acceptance judgment according to the defect evolution trend.
[0067] Furthermore, the forging defect detection device based on digital analysis is also used to: output defect judgment results based on the defect acceptance judgment, including defect location, defect type, defect level and risk assessment results; and associate and archive the standard data, the defect candidate area and the defect judgment results to generate the defect digital file.
[0068] Furthermore, the forging defect detection device based on digital analysis is also used to: reverse analyze the defect formation process parameters according to the defect digital archive, generate optimization suggestions, and feed the optimization suggestions back to the forging process control system for adaptive adjustment.
[0069] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The forging defect detection method and specific examples based on digital analysis in the aforementioned embodiment one are also applicable to the forging defect detection device based on digital analysis in this embodiment. Through the foregoing detailed description of the forging defect detection method based on digital analysis, those skilled in the art can clearly understand the forging defect detection device based on digital analysis in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.
[0070] 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.
[0071] 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 detecting defects in forgings based on digital analysis, characterized in that, include: Multi-source data is acquired from the target forging, and the multi-source data is then subjected to spatiotemporal alignment processing to obtain standard data. Anomaly analysis is performed on the digital model obtained from the standard data to generate candidate defect regions. The defect formation stage of the defect candidate region is analyzed, and the defect evolution trend is analyzed to generate a defect acceptance judgment. Based on the defect acceptance determination, the defect determination result is output, and a defect digital file is generated; The process feedback of the defect digital archive is executed.
2. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, Multi-source data acquisition was performed on the target forging, including: Surface geometric topography data are obtained by laser scanning; Internal structural detection data is obtained through ultrasonic testing; Process data is collected synchronously by the process monitoring units of the forging and heat treatment equipment; The multi-source data is obtained based on the surface geometry data, internal structure detection data, and processing data.
3. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, A digital model is obtained based on the aforementioned standard data, including: Based on the standard data, a three-dimensional digital structural model of the target forging is constructed; Extract the structural feature parameters of the target forging and generate a multi-dimensional feature vector; The digital model is obtained by combining the three-dimensional digital structure model and the multi-dimensional feature vector.
4. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, Anomaly analysis is performed on the digital model to generate candidate defect regions, including: The digital model is input into the defect identification channel for feature anomaly comparison analysis to obtain abnormal forging structure characteristics. Spatial clustering and boundary segmentation are performed on the abnormal forging structure features to generate the defect candidate regions.
5. The method for detecting defects in forgings based on digital analysis as described in claim 4, characterized in that, Construct a defect identification channel, including: Collect digital data of multi-source forgings that have completed multiple batches of testing and labeling; A training sample set is constructed based on the digital data of the multi-source forging parts, and the training sample set is supervised and labeled according to the preset defect types; The defect identification channel is constructed by training the model on the training sample set using a deep learning algorithm. The stability of the defect identification channel is evaluated by cross-validation. When the evaluation result meets the preset stability threshold, the defect identification channel is obtained.
6. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, The analysis of the defect formation stage of the defect candidate region includes: Perform defect pattern matching analysis on the defect candidate regions to obtain defect matching results; Based on the correlation analysis of the defect matching results using processing parameters, the defect formation stage is determined.
7. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, Analyze the defect evolution trend and generate defect acceptance criteria, including: Based on the defect formation stage, a defect impact factor is constructed, and the severity of the defect is assessed based on the defect impact factor. The defect evolution trend is analyzed based on the severity of the defect, and a defect acceptance judgment is generated based on the defect evolution trend.
8. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, Based on the defect acceptance determination, a defect determination result is output, and a defect digital file is generated, including: Based on the defect acceptance determination, a defect determination result is output, including defect location, defect type, defect level, and risk assessment result; The standard data, the defect candidate region, and the defect determination result are associated and archived to generate the defect digital file.
9. The method for detecting defects in forgings based on digital analysis as described in claim 1, characterized in that, The process feedback for executing the aforementioned defect digital profile includes: Based on the defect digital archive, reverse analysis is performed on the defect formation process parameters to generate optimization suggestions; The optimization suggestions are fed back to the forging process control system for adaptive adjustment.
10. A forging defect detection device based on digital analysis, characterized in that, The steps for implementing the digital analysis-based defect detection method for forgings according to any one of claims 1 to 9 include: The standard data acquisition module is used to acquire multi-source data from the target forging and perform spatiotemporal alignment processing on the multi-source data to obtain standard data. The defect candidate region generation module is used to perform anomaly analysis based on the digital model obtained from the standard data and generate defect candidate regions. The defect acceptance determination generation module is used to analyze the defect formation stage of the defect candidate region and analyze the defect evolution trend to generate a defect acceptance determination. The defect digital archive generation module is used to output a defect judgment result based on the defect acceptance judgment and generate a defect digital archive. The process feedback execution module is used to execute the process feedback of the defect digital archive.