Chamfered surface quality detection method and system based on material characteristics
By dynamically optimizing the chamfering surface quality detection method and system, the problems of low detection accuracy and efficiency caused by the discreteness of material composition and load are solved, and precise detection and efficient resource utilization are achieved.
Patent Information
- Application Number
- CN202510842172.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-23
- Publication Date
- 2025-09-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing technology has problems with poor detection accuracy and low efficiency in chamfered surface inspection. In particular, since the batch differences in material composition and the discreteness of bearing service loads are not effectively correlated, high-consistency batches are over-detected and low-consistency batches are missed, and there is a lack of dynamic adaptation to the defect distribution characteristics under material-load coupling.
By conducting sampling analysis on the material composition and load distribution of similar bearings, establishing material consistency and load consistency parameters, combining the prediction model of crack rate and roughness abnormality rate, dynamically optimizing the detection scale, and using an iterative optimization algorithm to generate the optimal detection scale, accurate detection is achieved.
It significantly improves the accuracy and resource utilization efficiency of chamfer surface quality inspection, automatically identifies high-variation risk batches, accurately anchors high-incidence areas of defects, avoids resource waste, and achieves a balance between inspection accuracy and cost.
Smart Images
Figure CN120703095A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of surface quality detection, and in particular to a method and system for detecting the surface quality of chamfering processing based on material properties. Background Art
[0002] Chamfering is a core machining method for mechanical systems, and its quality directly impacts assembly accuracy, stress distribution, and fatigue life. Existing technologies typically use fixed inspection scales to inspect chamfered surfaces for defects such as cracks and roughness. However, these methods have significant limitations. First, batch variations in material composition are not incorporated into inspection strategy design, resulting in over-inspection of highly consistent material batches and the under-inspection of low-consistency batches due to insufficient sampling. Second, the discrete nature of bearings' actual service loads is not effectively correlated, and microcrack risk areas under high-load conditions may not be captured due to the random nature of the fixed inspection scale. Third, chamfering defects exhibit a strong nonlinear relationship with material properties, but existing methods rely on manual experience to pre-determine inspection scales and are unable to dynamically adapt to the defect distribution characteristics under material-load coupling. In large-scale production, full inspection results in high costs and resources, while also reducing defect detection rates due to a lack of risk-based, targeted inspection. More critically, material composition fluctuations and service load variations combine to create spatial heterogeneity in defect occurrence, creating detection blind spots that lead to poor inspection accuracy and efficiency on chamfered surfaces. Summary of the Invention
[0003] The present application provides a method and system for detecting the quality of chamfered surface based on material properties, which is used to solve the technical problems of poor accuracy and low efficiency in the prior art of detecting the quality of chamfered surface.
[0004] In view of the above problems, the present application provides a method and system for detecting the surface quality of chamfering processing based on material characteristics.
[0005] In a first aspect, the present application provides a method for detecting surface quality of chamfering processing based on material properties, the method comprising: Conduct sampling material composition testing on similar bearings of the target bearing to obtain material composition distribution, analyze and obtain material consistency parameters, obtain the load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters; According to the material composition distribution and chamfer parameters, chamfer crack prediction and roughness anomaly prediction are performed to obtain crack rate distribution and roughness anomaly rate distribution; Randomly configure a detection scale for surface inspection of the chamfered bearing, and analyze and obtain the detection adaptability of the detection scale by combining the crack rate distribution, roughness abnormality rate distribution, material consistency parameter, and load consistency parameter; According to the detection adaptability, the detection scale is iteratively optimized to obtain the optimal detection scale. After the target bearing is chamfered, microscopic surface images are collected according to the optimal detection scale to perform surface quality detection.
[0006] In a second aspect, the present application provides a chamfering surface quality detection system based on material properties, comprising: The sampling analysis module is used to perform sampling material composition detection on similar bearings of the target bearing, obtain material composition distribution, analyze and obtain material consistency parameters, obtain the load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters; An abnormality prediction module, used to predict chamfer cracks and roughness abnormalities based on the material composition distribution and chamfer parameters, and obtain crack rate distribution and roughness abnormality rate distribution; A fitness analysis module is used to randomly configure a detection scale for surface inspection of the chamfered bearing, and analyze and obtain the detection fitness of the detection scale based on the crack rate distribution, roughness anomaly rate distribution, material consistency parameter, and load consistency parameter; The quality inspection module is used to iteratively optimize the inspection scale according to the inspection adaptability to obtain the optimal inspection scale. After the target bearing is chamfered, the microscopic surface image is collected according to the optimal inspection scale to perform surface quality inspection.
[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application proposes a chamfering surface quality detection method and system based on material properties. By dynamically integrating material composition fluctuations, service load discreteness, and defect distribution prediction results, it significantly improves the accuracy and resource utilization efficiency of bearing chamfer surface quality detection. Compared with traditional methods, the technical solution provided by this application significantly overcomes the rigid defects of fixed detection scales. First, based on the quantitative analysis of material consistency parameters and load consistency parameters, the system can automatically identify high-variation risk batches. In scenarios where the material composition is highly discrete or the service load is unevenly distributed, the detection density is intelligently improved, effectively capturing problems that are easily missed by traditional random inspections. Secondly, through a dual-branch prediction model of crack rate and roughness anomaly rate, the defect distribution law under different material and chamfer parameter combinations is pre-deduced, so that the detection scale optimization no longer relies on manual experience, but accurately anchors defect-prone areas, avoiding ineffective resource consumption in the full inspection mode. Finally, the detection fitness function incorporates defect coverage capability and resource consumption into a unified evaluation framework, ensuring that the iteratively generated optimal detection scale is both adapted to the individual risk of the current bearing and meets the common efficiency of the production line.
[0008] This application achieves the intelligent matching of detection intensity with material fluctuation, load conditions and defect probability while ensuring the defect detection rate, and provides a detection method that achieves both detection accuracy and detection cost for the quality control of chamfering processing. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0010] Figure 1 A schematic flow chart of a method for detecting surface quality of chamfering based on material properties provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a chamfering surface quality detection system based on material characteristics provided in an embodiment of the present application; In the accompanying drawings, the components represented by the reference numerals are described as follows: Sampling analysis module 100, abnormality prediction module 200, fitness analysis module 300, quality detection module 400. DETAILED DESCRIPTION
[0011] The present application provides a chamfering surface quality detection method and system based on material characteristics, which is used to solve the technical problems of poor accuracy and low efficiency in chamfering surface detection in the prior art.
[0012] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0013] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.
[0014] Example 1, as Figure 1 As shown, the present application provides a method for detecting surface quality of chamfering processing based on material characteristics, wherein the method includes: S10: Perform material composition testing on similar bearings of the target bearing to obtain material composition distribution, analyze and obtain material consistency parameters, obtain load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters.
[0015] Traditional testing methods fail to account for batch fluctuations in bearing material composition and the discrete nature of service loads, resulting in a lack of objective risk-based testing strategies. Existing technologies employ fixed testing scales and fail to identify hidden risks arising from material composition variations or uneven load distribution. Material composition dispersion can lead to localized performance degradation, while areas of high load are more susceptible to crack initiation.
[0016] Step S10 in the method provided in the embodiment of the present application includes: Collect the test results of sample material composition testing of similar bearings of the target bearing within a preset historical time range, obtain multiple test material compositions, and obtain material composition distribution; Calculating and obtaining material composition discrete parameters according to the material composition distribution; Calculating and obtaining material consistency parameters according to the material component discrete parameters; Within a preset historical time range, sample and collect the loads borne by similar bearings after installation to obtain the load distribution; According to the load distribution, load discrete parameters are calculated; The load consistency parameter is calculated based on the load discrete parameter.
[0017] In the embodiment of the present application, material composition test results within a preset historical time range are collected. For example, the historical time range is set to 1 year. From the historical test log, multiple test material compositions in the past year are collected to obtain the material composition distribution.
[0018] According to the material composition distribution, the material composition discrete parameters are calculated. The material composition discrete parameters are the standard deviation of each material component and are calculated using the standard deviation calculation formula.
[0019] The material consistency parameter is calculated based on the discrete parameters of the material components. The material consistency parameter = 1 ÷ the mean of the standard deviations of the material components. The larger the material consistency parameter, the smaller the change in material components and the better the uniformity.
[0020] The loads borne by similar bearings after installation within a preset historical time range are collected. For example, the historical time range is set to 1 year. The loads borne by multiple similar bearings after installation within the past year are collected from the historical operating data to obtain a load distribution.
[0021] According to the load distribution, the load discrete parameter is calculated. The load discrete parameter is the standard deviation of multiple similar bearing loads and is calculated using the standard deviation calculation formula.
[0022] Based on the load dispersion parameter, the load consistency parameter is calculated. The load consistency parameter = 1 ÷ the mean of the standard deviations of multiple similar bearing loads. The larger the load consistency parameter, the more concentrated the load distribution.
[0023] By quantitatively analyzing material composition and load distribution, a dynamic risk baseline is established. Material consistency parameters reflect the degree of compositional fluctuation, while load consistency parameters characterize the discreteness of service stresses. Together, these parameters identify high-risk batches and provide a basis for differentiated management in subsequent testing. Compared to traditional methods that ignore material and load variability, this application improves the targeted allocation of testing resources from the source.
[0024] S20: performing chamfer crack prediction and roughness anomaly prediction according to the material composition distribution and chamfer parameters to obtain crack rate distribution and roughness anomaly rate distribution.
[0025] Chamfering defects exhibit nonlinear characteristics, with their probability influenced by the coupling of fine-tuning of material composition and chamfering parameters. Existing methods rely on fixed thresholds or manual experience to predict defect distribution, failing to dynamically respond to material and process interactions. When material composition is widely distributed, the same chamfering parameters can induce differential defects in different material regions. Traditional spot checks, lacking the ability to predict the spatial distribution of defects, make it difficult to pinpoint high-incidence areas.
[0026] Step S20 in the method provided in the embodiment of the present application includes: Obtaining a chamfer quality predictor, wherein the chamfer quality predictor includes a crack prediction branch and a roughness prediction branch; Among them, obtaining the chamfer quality predictor includes: Based on the historical bearing chamfer data, we collected a set of sample chamfer parameters, a set of sample material compositions, and the percentage of areas with cracks and roughness anomalies on the bearing surface after chamfering. We then annotated the obtained sample crack rate set and roughness anomaly rate set. Using machine learning, with chamfer parameters and material composition as input features, and crack rate and roughness anomaly rate as output features, a crack prediction branch and a roughness prediction branch are constructed. Using the sample chamfer parameter set and the sample material composition set as input training data, and using the sample crack rate set and the roughness anomaly rate set as output training data, respectively, supervise the training of the crack prediction branch and the roughness prediction branch until convergence, and obtain a chamfer quality predictor; Get the chamfering parameters of the current target bearing; The chamfer parameters are respectively combined with a plurality of material components in the material component distribution, and input into the chamfer quality predictor, and the crack rate distribution and the roughness abnormality rate distribution are obtained by prediction output.
[0027] In an embodiment of the present application, based on the bearing chamfering data in historical time, a sample chamfering parameter set, a sample material composition set, and the proportion of areas where cracks appear on the bearing surface after chamfering and the proportion of areas where roughness anomalies appear are collected, and the sample crack rate set and the roughness anomaly rate set are marked. The sample chamfering parameter set and the sample material composition set are obtained by integrating the collected historical operation data and the historical detection data, and the proportion of areas where cracks appear and the proportion of areas where roughness anomalies appear are obtained by calculation and marking. The crack rate = the area of the cracked area ÷ the total surface area, and the roughness anomaly rate = the area of the roughness anomaly area ÷ the total surface area.
[0028] A chamfer quality predictor was constructed using machine learning. For example, a neural network was used to construct the chamfer quality predictor. First, a crack rate prediction branch was constructed. This branch employed a three-layer structure: an input layer with two nodes for receiving chamfer parameters and sample material composition; a hidden layer with 32 nodes activated using the ReLU function; and an output layer with one node activated using linear activation to output the predicted crack rate. A roughness prediction branch was constructed based on the same structure: an input layer with two nodes for receiving chamfer parameters and sample material composition; a hidden layer with 32 nodes activated using the ReLU function; and an output layer with one node activated using linear activation to output the predicted roughness.
[0029] A sample chamfer parameter set and a sample material composition set are used as input training data, and a sample crack rate set and a roughness anomaly rate set are used as output training data, respectively. The crack prediction branch and the roughness prediction branch are supervised and trained until convergence. That is, when the accuracy of the output predicted crack rate and roughness anomaly rate is above 95% for the input chamfer parameter and material composition set, the model training is completed, and a chamfer quality predictor is obtained. According to the chamfering parameter settings, collect the chamfering parameters of the current target bearing; The chamfering parameters are combined with multiple material components in the material composition distribution and input into the chamfering quality predictor, and the crack rate distribution and the roughness abnormality rate distribution are obtained as prediction output.
[0030] By generating a defect probability distribution through a dual-branch prediction model, we achieve precise defect prediction, transcending the traditional defect homogeneity assumption of testing. This accurately reveals the non-uniform distribution of risk resulting from the interaction between material properties and process parameters, providing targeted guidance for subsequent inspection scale optimization. Compared to the blindness of manually pre-set inspection areas, this step prioritizes inspection resources on high-risk coordinate areas where the predicted defect rate exceeds the threshold, significantly improving defect capture efficiency.
[0031] S30: randomly configuring a detection scale for surface detection of the chamfered bearing, and analyzing and obtaining the detection adaptability of the detection scale in combination with the crack rate distribution, roughness anomaly rate distribution, material consistency parameter, and load consistency parameter.
[0032] Choosing an inspection scale requires balancing defect capture capability and resource consumption. However, existing methods lack a unified evaluation standard, making it difficult to optimize both. Increasing inspection density based solely on defect rate distribution can waste resources in highly consistent batches. Reducing inspection intensity based solely on resource constraints can lead to missed detections in highly variable scenarios.
[0033] Step S30 in the method provided in the embodiment of the present application includes: Randomly configuring a detection scale for surface inspection of the chamfered bearing, wherein the detection scale includes a proportion of a collection area for collecting the chamfered image of the target bearing; Within the crack rate distribution and the roughness abnormality rate distribution, screening and calculating the crack rate proportion and the roughness abnormality rate proportion that are less than or equal to the detection scale, and obtaining the crack detection coefficient and the roughness detection coefficient; Analyzing and calculating the detection adaptability of the detection scale according to the crack detection coefficient, the roughness detection coefficient, the material consistency parameter, and the load consistency parameter; The detection adaptability of the detection scale is obtained by analyzing and calculating according to the crack detection coefficient, the roughness detection coefficient, the material consistency parameter, and the load consistency parameter, including: Get the preset detection scale; Calculating a ratio of the preset detection scale to the detection scale to obtain a detection resource utilization coefficient; The crack detection coefficient is calculated by correcting the material consistency parameter to obtain a corrected crack detection coefficient, and the roughness detection coefficient is calculated by correcting the load consistency parameter to obtain a corrected roughness detection coefficient; The detection adaptability is calculated based on the detection resource usage coefficient, the corrected crack detection coefficient, and the corrected roughness detection coefficient.
[0034] In this embodiment, a randomly configured inspection scale is used for surface inspection of a chamfered bearing. The inspection scale includes the percentage of the acquisition area used to capture the chamfered image of the target bearing. A random function is used to randomly generate a real number in the range of 0-1 as the inspection scale. For example, 0.6 is generated as the inspection scale.
[0035] Within the crack rate distribution and roughness anomaly rate distribution, filter and calculate the percentage of crack rates and roughness anomaly rates that are less than or equal to the test scale to obtain the crack detection coefficient and roughness detection coefficient. The crack detection coefficient = the percentage of crack rates less than or equal to the test scale / the total number of crack rates. The roughness detection coefficient = the percentage of roughness anomaly rates less than or equal to the test scale / the total number of roughness anomaly rates. For example, if the total number of crack rates is 100, the percentage of crack rates less than or equal to the test scale is 50, the total number of roughness anomaly rates is 100, and the roughness anomaly rates less than or equal to the test scale are 60, then the crack detection coefficient = 50 ÷ 100 = 0.5, and the roughness detection coefficient = 60 ÷ 100 = 0.6.
[0036] A preset detection scale is obtained. The preset detection scale is a preset proportion of the acquisition area of the target bearing chamfer image. For example, the preset detection scale is set to 0.5.
[0037] Calculate the ratio of the preset detection scale to the detection scale to obtain the detection resource utilization coefficient. Detection resource utilization coefficient = preset detection scale ÷ detection scale. For example, if the preset detection scale is 0.5 and the detection scale is 0.6, the detection resource utilization coefficient = 0.5 ÷ 0.6 = 0.83. A larger utilization coefficient indicates a smaller image acquisition area and less detection resource usage.
[0038] The material consistency parameter is used to correct the crack detection coefficient, resulting in a corrected crack detection coefficient. Corrected crack detection coefficient = material consistency parameter × crack detection coefficient. The load consistency parameter is used to correct the roughness detection coefficient, resulting in a corrected roughness detection coefficient = load consistency parameter × roughness detection coefficient. For example, if the material consistency parameter is 0.8, the crack detection coefficient is 0.5, the load consistency parameter is 0.6, and the roughness detection coefficient is 0.6, then the corrected crack detection coefficient = 0.8 × 0.5 = 0.4, and the corrected roughness detection coefficient = 0.6 × 0.6 = 0.36. The larger the material consistency parameter, the lower the probability that the crack rate of the current wheel hub bearing will change due to changes in material composition, the greater the credibility of the crack detection coefficient, and therefore, the larger the corrected roughness detection coefficient, and the greater the subsequent adaptability. The larger the load consistency parameter is, the more stable the load borne by the current wheel hub bearing is, the smaller the probability of roughness change caused by load change is, and the greater the credibility of the roughness detection coefficient is. Therefore, the larger the corrected roughness detection coefficient is, the greater the subsequent adaptability is.
[0039] The test fitness is calculated based on the test resource utilization factor, the corrected crack detection factor, and the corrected roughness detection factor. Test fitness = (corrected crack factor + corrected roughness factor + resource utilization factor) ÷ 3. For example, if the corrected crack factor is 0.4, the corrected roughness detection factor is 0.36, and the resource utilization factor is 0.83, then the test fitness = (0.4 + 0.36 + 0.83) ÷ 3 = 0.53. The greater the test fitness, the more accurate the test based on that test parameter, and the more suitable it is for testing complex conditions.
[0040] The crack detection coefficient and roughness detection coefficient are used to quantify the current inspection scale's ability to cover predicted defects. Material consistency parameters are introduced to modify the crack detection coefficient, while load consistency parameters are used to modify the roughness detection coefficient. Finally, a normalized evaluation is performed using the combined inspection resource utilization coefficient. This breaks through the traditional "defect-resource" two-dimensional optimization framework and incorporates material fluctuation risk and load condition risk into the inspection performance evaluation system. This allows high-variability batches to automatically receive weighted coverage coefficients, while low-variability batches receive relaxed inspection plans driven by the resource utilization coefficient, achieving intelligent decision-making that balances risk and inspection cost.
[0041] S40: Iteratively optimizing the detection scale according to the detection adaptability to obtain the optimal detection scale, collecting a microscopic surface image according to the optimal detection scale after the target bearing is chamfered, and performing surface quality detection.
[0042] Generating optimal inspection scales requires dynamic adaptation to real-time risks, but manual parameter adjustment is subject to lag and subjectivity. This is especially true when material composition is widely distributed or load distribution is complex. Fixed inspection scales cannot respond to sudden changes in local risk, while trial-and-error adjustments struggle to meet production line timelines.
[0043] Step S40 in the method provided in the embodiment of the present application includes: Randomly iteratively adjust the detection scale, calculate the detection fitness, and perform optimization; After reaching the preset number of optimization iterations, the optimal detection scale corresponding to the maximum value of the detection fitness is obtained, and after the target bearing is chamfered, a microscopic surface image is collected according to the optimal detection scale to perform surface quality detection.
[0044] The detection scale is randomly adjusted iteratively, the detection scale is randomly regenerated using a random function, the detection fitness is calculated according to the method of the embodiment of the present application, and iterative optimization is performed.
[0045] The preset number of optimization iterations is a pre-determined number of optimization iterations. Too few iterations may result in poor optimization results, while too many iterations may waste computing power. For example, 20 optimization iterations are set. After reaching the preset number of optimization iterations, the optimal detection scale corresponding to the maximum detection fitness is obtained. After chamfering the target bearing, microscopic surface images are collected according to the optimal detection scale for surface quality inspection.
[0046] Through the iterative optimization mechanism, a risk-responsive detection solution is output, achieving a dual breakthrough in accurate detection and dynamic adaptation. Based on the detection fitness function, a random optimization algorithm is used to dynamically adjust the detection scale parameters, and quickly converge to the maximum fitness within a preset number of iterations. This mechanism enables the optimal detection scale to automatically match the material composition fluctuation intensity, load distribution complexity, and defect prediction hotspot distribution. Compared with manual preset or fixed rules, the detection strategy provided by this application accurately allocates detection resources to the most risk-sensitive areas while ensuring the defect detection rate, ultimately achieving the global optimality of detection accuracy and cost-effectiveness.
[0047] Example 2, as Figure 2 As shown, based on the same inventive concept as the chamfering surface quality detection method based on material properties provided in the first embodiment, the embodiment of the present invention further provides a chamfering surface quality detection system based on material properties, comprising: The sampling and analysis module 100 is used to perform sampling material composition detection on similar bearings of the target bearing, obtain material composition distribution, analyze and obtain material consistency parameters, obtain load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters; An abnormality prediction module 200 is used to perform chamfer crack prediction and roughness abnormality prediction based on the material composition distribution and chamfer parameters, and obtain crack rate distribution and roughness abnormality rate distribution; The fitness analysis module 300 is used to randomly configure a detection scale for surface inspection of the chamfered bearing, and analyze and obtain the detection fitness of the detection scale based on the crack rate distribution, roughness anomaly rate distribution, material consistency parameter, and load consistency parameter. The quality inspection module 400 is used to iteratively optimize the inspection scale according to the inspection adaptability to obtain the optimal inspection scale, collect microscopic surface images according to the optimal inspection scale after the target bearing is chamfered, and perform surface quality inspection.
[0048] In one embodiment, the sampling analysis module 100 is further configured to: Collect the test results of sample material composition testing of similar bearings of the target bearing within a preset historical time range, obtain multiple test material compositions, and obtain material composition distribution; Calculating and obtaining material composition discrete parameters according to the material composition distribution; Calculating and obtaining material consistency parameters according to the material component discrete parameters; Within a preset historical time range, sample and collect the loads borne by similar bearings after installation to obtain the load distribution; According to the load distribution, load discrete parameters are calculated; The load consistency parameter is calculated based on the load discrete parameter.
[0049] In one embodiment, the anomaly prediction module 200 is further configured to: Obtaining a chamfer quality predictor, wherein the chamfer quality predictor includes a crack prediction branch and a roughness prediction branch; Among them, obtaining the chamfer quality predictor includes: Based on the historical bearing chamfer data, we collected a set of sample chamfer parameters, a set of sample material compositions, and the percentage of areas with cracks and roughness anomalies on the bearing surface after chamfering. We then annotated the obtained sample crack rate set and roughness anomaly rate set. Using machine learning, with chamfer parameters and material composition as input features, and crack rate and roughness anomaly rate as output features, a crack prediction branch and a roughness prediction branch are constructed. Using the sample chamfer parameter set and the sample material composition set as input training data, and using the sample crack rate set and the roughness anomaly rate set as output training data, respectively, supervise the training of the crack prediction branch and the roughness prediction branch until convergence, and obtain a chamfer quality predictor; Get the chamfering parameters of the current target bearing; The chamfer parameters are respectively combined with a plurality of material components in the material component distribution, and input into the chamfer quality predictor, and the crack rate distribution and the roughness abnormality rate distribution are obtained by prediction output.
[0050] In one embodiment, the fitness analysis module 300 is further configured to: Randomly configuring a detection scale for surface inspection of the chamfered bearing, wherein the detection scale includes a proportion of a collection area for collecting the chamfered image of the target bearing; Within the crack rate distribution and the roughness abnormality rate distribution, screening and calculating the crack rate proportion and the roughness abnormality rate proportion that are less than or equal to the detection scale, and obtaining the crack detection coefficient and the roughness detection coefficient; Analyzing and calculating the detection adaptability of the detection scale according to the crack detection coefficient, the roughness detection coefficient, the material consistency parameter, and the load consistency parameter; The detection adaptability of the detection scale is obtained by analyzing and calculating according to the crack detection coefficient, the roughness detection coefficient, the material consistency parameter, and the load consistency parameter, including: Get the preset detection scale; Calculating a ratio of the preset detection scale to the detection scale to obtain a detection resource utilization coefficient; The crack detection coefficient is calculated by correcting the material consistency parameter to obtain a corrected crack detection coefficient, and the roughness detection coefficient is calculated by correcting the load consistency parameter to obtain a corrected roughness detection coefficient; The detection adaptability is calculated based on the detection resource usage coefficient, the corrected crack detection coefficient, and the corrected roughness detection coefficient.
[0051] In one embodiment, the quality detection module 400 is further configured to: Randomly iteratively adjust the detection scale, calculate the detection fitness, and perform optimization; After reaching the preset number of optimization iterations, the optimal detection scale corresponding to the maximum value of the detection fitness is obtained, and after the target bearing is chamfered, a microscopic surface image is collected according to the optimal detection scale to perform surface quality detection.
[0052] In summary, the embodiments of the present application have at least the following technical effects: This application proposes a chamfering surface quality detection method and system based on material properties. By dynamically integrating material composition fluctuations, service load discreteness, and defect distribution prediction results, it significantly improves the accuracy and resource utilization efficiency of bearing chamfer surface quality detection. Compared with traditional methods, the technical solution provided by this application significantly overcomes the rigid defects of fixed detection scales. First, based on the quantitative analysis of material consistency parameters and load consistency parameters, the system can automatically identify high-variation risk batches. In scenarios where the material composition is highly discrete or the service load is unevenly distributed, the detection density is intelligently improved, effectively capturing problems that are easily missed by traditional random inspections. Secondly, through a dual-branch prediction model of crack rate and roughness anomaly rate, the defect distribution law under different material and chamfer parameter combinations is pre-deduced, so that the detection scale optimization no longer relies on manual experience, but accurately anchors defect-prone areas, avoiding ineffective resource consumption in the full inspection mode. Finally, the detection fitness function incorporates defect coverage capability and resource consumption into a unified evaluation framework, ensuring that the iteratively generated optimal detection scale is both adapted to the individual risk of the current bearing and meets the common efficiency of the production line.
[0053] This application achieves the intelligent matching of detection intensity with material fluctuation, load conditions and defect probability while ensuring the defect detection rate, and provides a detection method that achieves both detection accuracy and detection cost for the quality control of chamfering processing.
[0054] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0055] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.
[0056] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.
Claims
1. A chamfering surface quality detection method based on material properties, characterized in that: The method comprises: Conduct sampling material composition testing on similar bearings of the target bearing to obtain material composition distribution, analyze and obtain material consistency parameters, obtain the load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters; According to the material composition distribution and chamfer parameters, chamfer crack prediction and roughness anomaly prediction are performed to obtain crack rate distribution and roughness anomaly rate distribution; Randomly configure a detection scale for surface inspection of the chamfered bearing, and analyze and obtain the detection adaptability of the detection scale by combining the crack rate distribution, roughness abnormality rate distribution, material consistency parameter, and load consistency parameter; According to the detection adaptability, the detection scale is iteratively optimized to obtain the optimal detection scale. After the target bearing is chamfered, microscopic surface images are collected according to the optimal detection scale to perform surface quality detection.
2. The chamfering surface quality detection method based on material properties according to claim 1 is characterized in that: Conduct sampling material composition testing on similar bearings of the target bearing to obtain material composition distribution and analyze the material consistency parameters, including: Collect the test results of sample material composition testing of similar bearings of the target bearing within a preset historical time range, obtain multiple test material compositions, and obtain material composition distribution; Calculating and obtaining material composition discrete parameters according to the material composition distribution; The material consistency parameters are calculated based on the material component discrete parameters.
3. The chamfering surface quality detection method based on material properties according to claim 1 is characterized in that: Obtain the load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters, including: Within a preset historical time range, sample and collect the loads borne by similar bearings after installation to obtain the load distribution; According to the load distribution, load discrete parameters are calculated; The load consistency parameter is calculated based on the load discrete parameter.
4. The chamfering surface quality detection method based on material properties according to claim 1 is characterized in that: According to the material composition distribution and chamfer parameters, chamfer crack prediction and roughness anomaly prediction are performed to obtain crack rate distribution and roughness anomaly rate distribution, including: Obtaining a chamfer quality predictor, wherein the chamfer quality predictor includes a crack prediction branch and a roughness prediction branch; Get the chamfering parameters of the current target bearing; The chamfer parameters are respectively combined with a plurality of material components in the material component distribution, and input into the chamfer quality predictor, and the crack rate distribution and the roughness abnormality rate distribution are obtained by prediction output.
5. The chamfering surface quality detection method based on material properties according to claim 4 is characterized in that: Get the chamfer quality predictor, including: Based on the historical bearing chamfer data, we collected a set of sample chamfer parameters, a set of sample material compositions, and the percentage of areas with cracks and roughness anomalies on the bearing surface after chamfering. We then annotated the obtained sample crack rate set and roughness anomaly rate set. Using machine learning, with chamfer parameters and material composition as input features, and crack rate and roughness anomaly rate as output features, a crack prediction branch and a roughness prediction branch are constructed. The sample chamfer parameter set and the sample material composition set are used as input training data, and the sample crack rate set and the roughness anomaly rate set are used as output training data respectively. The crack prediction branch and the roughness prediction branch are supervised and trained until convergence to obtain a chamfer quality predictor.
6. The chamfering surface quality detection method based on material properties according to claim 1 is characterized in that: The detection scale for surface inspection of the chamfered bearing is randomly configured. Combined with the crack rate distribution, roughness abnormality rate distribution, material consistency parameters and load consistency parameters, the detection adaptability of the detection scale is analyzed and obtained, including: Randomly configuring a detection scale for surface inspection of the chamfered bearing, wherein the detection scale includes a proportion of a collection area for collecting the chamfered image of the target bearing; Within the crack rate distribution and the roughness abnormality rate distribution, screening and calculating the crack rate proportion and the roughness abnormality rate proportion that are less than or equal to the detection scale, and obtaining the crack detection coefficient and the roughness detection coefficient; The detection adaptability of the detection scale is obtained by analysis and calculation based on the crack detection coefficient, the roughness detection coefficient, the material consistency parameter and the load consistency parameter.
7. The chamfering surface quality detection method based on material properties according to claim 6 is characterized in that: The detection adaptability of the detection scale is obtained by analyzing and calculating according to the crack detection coefficient, the roughness detection coefficient, the material consistency parameter, and the load consistency parameter, including: Get the preset detection scale; Calculating a ratio of the preset detection scale to the detection scale to obtain a detection resource utilization coefficient; The crack detection coefficient is calculated by correcting the material consistency parameter to obtain a corrected crack detection coefficient, and the roughness detection coefficient is calculated by correcting the load consistency parameter to obtain a corrected roughness detection coefficient; The detection adaptability is calculated based on the detection resource usage coefficient, the corrected crack detection coefficient, and the corrected roughness detection coefficient.
8. The chamfering surface quality detection method based on material properties according to claim 1 is characterized in that: Based on the detection adaptability, the detection scale is iteratively optimized to obtain the optimal detection scale. After the target bearing is chamfered, microscopic surface images are collected according to the optimal detection scale to perform surface quality detection, including: Randomly iteratively adjust the detection scale, calculate the detection fitness, and perform optimization; After reaching the preset number of optimization iterations, the optimal detection scale corresponding to the maximum value of the detection fitness is obtained, and after the target bearing is chamfered, a microscopic surface image is collected according to the optimal detection scale to perform surface quality detection.
9. The chamfering surface quality detection system based on material properties is characterized by: A system for implementing the chamfering surface quality detection method based on material characteristics according to any one of claims 1 to 8, comprising: The sampling analysis module is used to perform sampling material composition detection on similar bearings of the target bearing, obtain material composition distribution, analyze and obtain material consistency parameters, obtain the load distribution of similar bearings during installation and operation, and analyze and obtain load consistency parameters; An abnormality prediction module, used to predict chamfer cracks and roughness abnormalities based on the material composition distribution and chamfer parameters, and obtain crack rate distribution and roughness abnormality rate distribution; A fitness analysis module is used to randomly configure a detection scale for surface inspection of the chamfered bearing, and analyze and obtain the detection fitness of the detection scale based on the crack rate distribution, roughness anomaly rate distribution, material consistency parameter, and load consistency parameter; The quality inspection module is used to iteratively optimize the inspection scale according to the inspection adaptability to obtain the optimal inspection scale. After the target bearing is chamfered, the microscopic surface image is collected according to the optimal inspection scale to perform surface quality inspection.