Intelligent detection method and system for metal shaft
By analyzing the surface image of the metal axis, identifying and comparing feature points, calculating feature deviation values, determining the theoretical correspondence time, and eliminating occasional feature points, the problem of inaccurate metal axis detection results is solved, and the accuracy of the remaining life of the metal axis is realized.
Patent Information
- Application Number
- CN202510430395.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-08
- Publication Date
- 2025-07-04
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, the residual service life detection of metal shafts depends on the experience of staff, resulting in inaccurate detection results and detection thresholds, making it difficult to achieve convenient and accurate life prediction.
By obtaining the circumferential surface image of the metal axis, identifying feature points and comparing them with feature points in the preset test database, calculating feature deviation values, determining the theoretical corresponding time, calculating the remaining usage time based on the installation and use time, and removing occasional feature points, updating the permissioned feature set to improve detection accuracy.
It realizes convenient and accurate detection of the remaining service life of the metal shaft, can eliminate occasional feature points, and improve the reliability and accuracy of the detection results.
Smart Images
Figure CN120259273A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of shaft part detection technologies, and particularly to an intelligent detection method and system for metal shafts. Background Art
[0002] Metal shafts are important components in mechanical parts. They are usually cylindrical. The basic function of a metal shaft is to support rotating parts and rotate with them to transmit motion, torque, or bending moment. Metal shafts can withstand various types of loads, including torque, bending moment, and axial load, so they are very crucial in mechanical design.
[0003] To ensure the quality and performance of metal shafts, tests need to be carried out when the metal shafts leave the factory. By simulating the usage conditions of the metal shafts, the theoretical service life of the metal shafts is determined to determine whether the metal shafts are qualified. Generally, however, the test environment of the metal shafts is different from the actual working environment. At this time, the determined theoretical service life of the metal shafts is not the service life in the actual situation. Therefore, in order to reduce the situation where the service life of the metal shafts is exhausted during use, it is necessary for the staff to regularly detect the metal shafts to know the wear situation. When determining the wear situation of the metal shafts, it is generally based on the experience of the staff. There are not only inaccurate life prediction results, but also a certain threshold for detection because experienced staff are required. Therefore, there is an urgent need to design a method that can more conveniently and accurately detect the remaining service life of metal shafts. Summary of the Invention
[0004] In order to more conveniently and accurately detect the remaining service life of metal shafts, this application provides an intelligent detection method and system for metal shafts.
[0005] In a first aspect, this application provides an intelligent detection method for metal shafts, adopting the following technical solution: An intelligent detection system for metal shafts, comprising: Obtain the circumferential surface image of the metal shaft and the installation and usage duration; Identify feature points in the circumferential surface image to determine the type and level of the identified features, and combine the type and level of the identified features to determine the identified feature points; Obtain the identified feature points at each test duration in a preset test database, and define the identified feature points as test feature points; Count according to the identified feature points to determine the number of identified features, and count according to the test feature points to determine the number of test features; Calculate the difference based on the number of recognition features and the number of test features to determine the increased demand quantity and the position of the increased demand, and generate virtual blank points according to the increased demand quantity and supplement them to the position of the increased demand; Determine the corresponding feature combinations based on the one-to-one correspondence of the recognition feature points, test feature points, and virtual blank points, and determine the overall feature deviation value according to the corresponding feature combinations; Determine the overall feature deviation value with the smallest value according to the preset sorting rule, and define this overall feature deviation value as the combined reasonable deviation value at this test duration; Determine the combined reasonable deviation value with the smallest value according to the sorting rule, and determine the theoretical corresponding duration corresponding to this combined reasonable deviation value; Calculate based on the installation and usage duration, the theoretical corresponding duration, and the preset theoretical usage duration to determine the remaining usage duration.
[0006] Optionally, after determining the recognition feature points, the intelligent detection method for the metal shaft further includes: Determine the set of license features corresponding to the installation and usage duration according to the preset license matching relationship; Judge whether the recognition feature points are within the set of license features; If the recognition feature points are within the set of license features, maintain the determined recognition feature points; If the recognition feature points are not within the set of license features, eliminate these recognition feature points and define them as accidental feature points; Determine the feature influence duration corresponding to the accidental feature points according to the preset influence matching relationship; Perform a summation calculation based on all the feature influence durations to determine the feature loss duration, and perform a difference calculation based on the remaining usage duration and the feature loss duration to update the remaining usage duration.
[0007] Optionally, after determining the accidental feature points, the intelligent detection method for the metal shaft further includes: Count all the recognition feature points before eliminating the recognition feature points to determine the number of features before elimination, and count the accidental feature points to determine the number of accidental features; Calculate based on the number of features before elimination and the number of accidental features to determine the overall accidental proportion; Judge whether the overall accidental proportion is less than the preset accidental demand proportion; If the overall accidental proportion is less than the accidental demand proportion, maintain the determined accidental feature points; If the overall accidental proportion is not less than the accidental demand proportion, restore the determined accidental feature points to recognition feature points.
[0008] Optionally, if the accidental overall ratio is not less than the accidental demand ratio, the intelligent metal shaft detection method further includes: Performing a summation calculation based on a preset supplementary duration and an installation and usage duration to update the installation and usage duration; Re-determining a license feature set according to the updated installation and usage duration, and restoring accidental feature points within the new license feature set to recognition feature points; After the recognition feature points are restored, re-determining the accidental overall ratio, and when the accidental overall ratio is not less than the accidental demand ratio, repeating the update of the installation and usage duration according to the supplementary duration until the accidental overall ratio is not less than the accidental demand ratio, and defining the installation and usage duration corresponding to the accidental overall ratio not less than the accidental demand ratio as the original analysis duration.
[0009] Optionally, after the theoretical corresponding duration is determined, the intelligent metal shaft detection method further includes: Performing a difference calculation based on the original analysis duration and the theoretical corresponding duration to determine an analysis interval duration; Judging whether the analysis interval duration is less than a preset limit interval duration; If the analysis interval duration is less than the limit interval duration, calculating a remaining usage duration according to the theoretical corresponding duration; If the analysis interval duration is not less than the limit interval duration, restoring all accidental feature points to recognition feature points to re-determine the theoretical corresponding duration.
[0010] Optionally, after the combined reasonable deviation value is determined, the intelligent metal shaft detection method further includes: Judging whether there are at least two test durations with the smallest and same combined reasonable deviation values; If there are no at least two test durations with the smallest and same combined reasonable deviation values, determining the test duration corresponding to the smallest combined reasonable deviation value as the theoretical corresponding duration; If there are at least two test durations with the smallest and same combined reasonable deviation values, defining the smallest combined reasonable deviation value as a reference reasonable deviation value, defining the corresponding test duration as an alternative duration, and defining the two test durations adjacent to the alternative duration as adjacent binding durations; Defining the combined reasonable deviation value determined by the adjacent binding durations as an adjacent reasonable deviation value; Performing a difference calculation based on the adjacent reasonable deviation value and the reference reasonable deviation value to determine an adjacent overall deviation value; Performing an average calculation based on two adjacent overall deviation values to determine an average adjacent deviation value, and calculating a comparison deviation value based on the average adjacent deviation value and a preset standard adjacent deviation value; Determine the comparison deviation value with the smallest numerical value according to the sorting rule, and determine the alternative duration corresponding to the comparison deviation value as the theoretical corresponding duration.
[0011] In a second aspect, the present application provides a metal shaft intelligent detection system, adopting the following technical solution: A metal shaft intelligent detection system includes: An acquisition module, configured to acquire the circumferential surface image of the metal shaft and the installation and usage duration. A processing module, connected to the acquisition module, for storing and processing information. The processing module performs feature point recognition in the circumferential surface image to determine the recognition feature type and the recognition feature level, and combines the recognition feature type and the recognition feature level to determine the recognition feature points. The acquisition module acquires the recognition feature points at each test duration from a preset test database, and enables the processing module to define the recognition feature points as test feature points. The processing module counts according to the recognition feature points to determine the recognition feature quantity, and counts according to the test feature points to determine the test feature quantity. The processing module calculates the difference according to the recognition feature quantity and the test feature quantity to determine the required increase quantity and the required increase position, and generates virtual blank points according to the required increase quantity to supplement them to the required increase position. The processing module determines the feature corresponding combination according to the one-to-one correspondence of the recognition feature points, the test feature points and the virtual blank points, and determines the overall feature deviation value according to the feature corresponding combination. The processing module determines the overall feature deviation value with the smallest numerical value according to the preset sorting rule, and defines the overall feature deviation value as the combined reasonable deviation value at this test duration. The processing module determines the combined reasonable deviation value with the smallest numerical value according to the sorting rule, and determines the test duration corresponding to the combined reasonable deviation value as the theoretical corresponding duration. The processing module calculates according to the installation and usage duration, the theoretical corresponding duration and the preset theoretical usage duration to determine the remaining usage duration.
[0012] In summary, the present application includes at least one of the following beneficial technical effects: When detecting the metal shaft, by analyzing the surface image of the metal shaft to determine each defect feature, and comparing the defect feature with the defect feature that appears during the test of the metal shaft to determine the actual usage situation of the current metal shaft, so that the remaining service life of the metal shaft can be detected more conveniently and accurately through comparison and analysis; during the process of data analysis, the defect features generated by unexpected situations can be determined and eliminated, so as to facilitate the comparison of the usage situation between the current metal shaft and the metal shaft during the test. Description of the Drawings
[0013] Figure 1 is a flowchart of the intelligent detection method for the metal shaft.
[0014] Figure 2 is a flowchart of the recognition feature point elimination method.
[0015] Figure 3 is a flowchart of the accidental feature point analysis method.
[0016] Figure 4 is a flowchart of the installation and usage duration update method.
[0017] Figure 5 is a flowchart of the interval duration analysis method.
[0018] Figure 6 is a flowchart of the test duration screening method.
[0019] Figure 7 is a module flowchart of the intelligent detection method for the metal shaft. Detailed Implementation Manner
[0020] In order to make the purpose, technical solutions and advantages of the present application clearer and more understandable, the following is a further detailed description of the present application in combination with Figures 1 - 7 and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0021] The following further describes the embodiments of the present application in detail with reference to the drawings of the specification.
[0022] The embodiments of the present application disclose an intelligent detection method for a metal shaft. When detecting the remaining service life of the metal shaft, by collecting and analyzing the image of the metal shaft, the usage progress of the metal shaft can be determined, so that the remaining service life of the metal shaft can be detected more conveniently and accurately.
[0023] Referring to Figure 1 , the method flow of the intelligent detection method for the metal shaft includes the following steps: Step S100: Obtain the circumferential surface image of the metal shaft and the installation and usage duration.
[0024] The circumferential surface image is the image of the circumferential surface of the metal shaft. The metal shaft to be detected can be placed at a preset detection position, and the image capture device on the periphery of the detection position can be controlled to capture the image of the surface of the metal shaft. At this time, the images obtained by each image capture device can be image-fused by point cloud fusion to obtain the circumferential surface image of the entire surface of the metal shaft; the installation and usage duration is the duration from the start of installation and use of the metal shaft to the current time point, that is, the overall duration of the metal shaft in use.
[0025] Step S101: Identify feature points in the circumferential surface image to determine the type and level of the identified features, and combine the type and level of the identified features to determine the identified feature points.
[0026] The type of the identified feature is the type of the features on the surface of the metal shaft that affect the service life of the metal shaft, such as dents, wear, rust, etc. The level of the identified feature is the level of each feature point. For example, wear can be divided into three levels: slight wear, moderate wear, and severe wear. The corresponding feature recognition method can be that the staff obtains samples of different defect features of each metal shaft in advance, records each defect feature through neural network learning to determine the corresponding feature recognition database, and then inputs the corresponding image into the corresponding feature recognition database to achieve the recognition of feature points; the identified feature points are to record the type and level of the identified features of each feature point for subsequent analysis.
[0027] Step S102: Obtain the identified feature points at each test duration in the preset test database, and define the identified feature points as test feature points.
[0028] The test database is a database composed of the images obtained by the metal shaft during the test. The surface conditions of the metal shaft are different at each test duration, and the corresponding identified feature points are also different. The specific test database can be determined through the one-to-one correspondence between the identified feature points and the test duration. This test database is gradually established by the staff during the metal shaft test. When the metal shaft test is completed, this test database is also established; at this time, the test feature points are defined to distinguish different identified feature points for subsequent analysis.
[0029] Step S103: Count according to the identified feature points to determine the number of identified features, and count according to the test feature points to determine the number of test features.
[0030] The number of identified features is the total number of the identified feature points determined by the currently detected metal shaft, and the number of test features is the total number of the test feature points determined by the tested metal shaft.
[0031] Step S104: Calculate the difference according to the number of identified features and the number of test features to determine the required increased quantity and the required increased position, and generate virtual blank points according to the required increased quantity to supplement them into the required increased position.
[0032] The required increased quantity is the difference between the number of recognition features and the number of test features, and this difference is an absolute value. The required increased position is the position where feature points need to be added, and this position includes the currently detected metal shaft and the test metal shaft. By generating virtual blank points, the number of recognition feature points on the currently detected metal shaft can be made consistent with the number of test feature points on the test metal shaft, facilitating subsequent pairwise comparison of feature points. Among them, when the number of recognition features is greater than the number of test features, the required increased position is the position of the test metal shaft; otherwise, it is the position of the currently detected metal shaft. Among them, virtual blank points are feature points without recognition feature types and recognition feature levels, and are only used for analysis of each feature point.
[0033] Step S105: Determine the corresponding feature combinations based on the one-to-one correspondence of recognition feature points, test feature points, and virtual blank points, and determine the overall feature deviation value based on the corresponding feature combinations.
[0034] The corresponding feature combination is the combination formed by the one-to-one correspondence of recognition feature points and test feature points. For example, if the recognition feature points include four points A, B, C, and D, and the test feature points are two points E and F, then two virtual blank points X and Y need to be generated and added to the test feature points. At this time, the test feature points are E, F, X, and Y. At this time, multiple corresponding feature combinations can be formed, such as AE, BF, CX, DY or AE, BF, CY, DX. It only needs to ensure that one recognition feature point corresponds to one test feature point. Therefore, in the case of the above example, a total of 24 corresponding feature combinations can be obtained. The overall feature deviation value is the corresponding deviation situation of all feature points under this corresponding feature combination. The larger the overall feature deviation value, the more different the usage conditions of the two metal shafts; on the contrary, it means that the usage conditions of the two metal shafts are more similar. Among them, the calculation method of the overall feature deviation value is as follows: When the recognition feature types of two feature points are different, a fixed calculation parameter I can be obtained by default. When the recognition feature types of two feature points are the same, the calculation parameter U can be obtained based on the difference between the recognition feature levels. The specific values of I and U can be determined in advance by the staff. Therefore, the calculation formula of the overall feature deviation value is , where is the overall feature deviation value, is the calculation parameter determined by the th group of feature points, is the calculation parameter determined by the th group of feature points. When the recognition feature types of two feature points are different, the corresponding is 0. When the recognition feature types of two feature points are the same, the corresponding is 0. Among them, is the calculation weight parameter for the same recognition feature type. Different defect features have different impacts on the life of the metal shaft, and the corresponding calculation weight parameters are also different. The specific values are determined in advance by the staff.
[0035] Step S106: Determine the overall feature deviation value with the smallest value according to a preset sorting rule, and define this overall feature deviation value as the combined reasonable deviation value at this test duration.
[0036] The sorting rule is a method set by the staff that can sort the numerical values, such as the bubble sort method. Through the sorting rule, the overall feature deviation value with the smallest value can be determined, that is, the feature corresponding combination under this overall feature deviation value can best reflect the similarity of the two metal shafts at the current experimental duration. At this time, it is defined as the combined reasonable deviation value for identification, which is convenient for subsequent analysis.
[0037] Step S107: Determine the combined reasonable deviation value with the smallest value according to the sorting rule, and determine the test duration corresponding to this combined reasonable deviation value as the theoretical corresponding duration.
[0038] Through the sorting rule, the combined reasonable deviation value with the smallest value can be determined, that is, the metal shaft at the test duration corresponding to this combined reasonable deviation value is closest to the defects of the currently detected metal shaft. Therefore, the situation of the currently detected metal shaft can be determined based on the situation of the metal shaft at this test duration. At this time, it is defined as the theoretical corresponding duration to distinguish different experimental durations, which is convenient for subsequent analysis.
[0039] Step S108: Calculate based on the installation and usage duration, the theoretical corresponding duration, and a preset theoretical usage duration to determine the remaining usage duration.
[0040] The remaining usage duration is the duration that the metal shaft can still be used theoretically in the future, that is, the service life of the metal shaft. The calculation formula is , where is the remaining usage duration, is the theoretical usage duration, is the theoretical corresponding duration, is the installation and usage duration, where the theoretical usage duration is the life duration that the test metal shaft adheres to during use.
[0041] Referring to Figure 2 , after the identification feature points are determined, the intelligent detection method for metal shafts further includes: Step S200: Determine the set of permission features corresponding to the installation and usage duration according to a preset permission matching relationship.
[0042] The set of permission features is a set composed of the feature points that will appear after the normal use and installation and usage duration of the metal shaft. The set of permission features corresponding to different installation and usage durations is different, and the permission matching relationship between the two is determined by the staff in advance.
[0043] Step S201: Determine whether the recognized feature points are within the permitted feature set.
[0044] The purpose of the determination is to find out whether the flaw feature points on the current metal shaft are the feature points that appear during the normal use of the metal shaft.
[0045] Step S2011: If the recognized feature points are within the permitted feature set, then maintain the determined recognized feature points.
[0046] When the recognized feature points are within the permitted feature set, it indicates that the recognized feature points are the feature points that appear during the normal use of the metal shaft. At this time, it is only necessary to normally maintain the determined recognized feature points.
[0047] Step S2012: If the recognized feature points are not within the permitted feature set, then eliminate the recognized feature points and define them as accidental feature points.
[0048] When the recognized feature points are not within the permitted feature set, it indicates that the recognized feature points are not the feature points that appear during the normal use of the metal shaft, that is, this feature point is probably a feature point caused by the interaction between an external foreign object and the metal shaft. At this time, eliminate the recognized feature points and define them as accidental feature points to distinguish different feature points, which is convenient for subsequent analysis.
[0049] Step S202: Determine the feature influence duration corresponding to the accidental feature points according to the preset influence matching relationship.
[0050] The feature influence duration is the duration by which the accidental feature points will shorten the overall life of the metal shaft. Different accidental feature points correspond to different feature influence durations. The influence matching relationship between the two is determined by the staff through multiple tests in advance and will not be elaborated here; this test can be achieved through the variable method. For example, through tests on two metal shafts, one metal shaft is a complete metal shaft, and the other metal shaft is a metal shaft with accidental feature points. By comparing the final life difference between the two metal shafts to determine the feature influence duration.
[0051] Step S203: Perform a summation calculation based on all the feature influence durations to determine the feature loss duration, and perform a difference calculation based on the remaining service duration and the feature loss duration to update the remaining service duration.
[0052] The feature loss duration is the sum of the feature influence durations brought by all accidental feature points. By excluding the accidental feature points, the determined remaining service duration can be made closer to the remaining life of the metal shaft in the actual situation. At this time, subtracting the feature loss duration from this remaining service duration can obtain a more accurate remaining service duration.
[0053] Refer to Figure 3, after the accidental feature points are determined, the intelligent detection method for the metal shaft further includes: Step S300: Count all the identified feature points before excluding the identified feature points to determine the number of features before exclusion, and count the accidental feature points to determine the number of accidental features.
[0054] The number of features before exclusion is the total number of all identified feature points before the accidental feature points are excluded, and the number of accidental features is the total number of all determined accidental feature points.
[0055] Step S301: Calculate based on the number of features before exclusion and the number of accidental features to determine the overall proportion of accidents.
[0056] The overall proportion of accidents is the ratio of the determined accidental feature points to all identified feature points, which is determined by dividing the number of accidental features by the number of features before exclusion.
[0057] Step S302: Determine whether the overall proportion of accidents is less than the preset required proportion of accidents.
[0058] The required proportion of accidents is the maximum overall proportion of accidents allowed by the staff to determine that the proportion of this accidental feature point is small and this accidental feature point is indeed an accidentally occurring feature point. The purpose of the judgment is to know whether the current accidental feature point is an accidentally occurring feature point in the actual situation.
[0059] Step S3021: If the overall proportion of accidents is less than the required proportion of accidents, maintain the determined accidental feature points.
[0060] When the overall proportion of accidents is less than the required proportion of accidents, it means that this accidental feature point is indeed an accidentally occurring feature point. At this time, it is only necessary to maintain this accidental feature point.
[0061] Step S3022: If the overall proportion of accidents is not less than the required proportion of accidents, restore the determined accidental feature points to the identified feature points.
[0062] When the overall proportion of accidents is not less than the required proportion of accidents, it means that the number of accidental feature points that appear is large, that is, there may be a situation where the entire metal shaft accelerates its service life due to external environmental influences. That is, the set of permitted features determined by the installation and use duration is inaccurate. At this time, the accidental feature points are restored to the identified feature points to better analyze the metal shaft.
[0063] Refer to Figure 4 , if the overall proportion of accidents is not less than the required proportion of accidents, the intelligent detection method for the metal shaft further includes: Step S400: Sum the preset supplementary duration and the installation and use duration to update the installation and use duration.
[0064] The supplementary duration is the fixed duration set for the staff. By adding the supplementary duration to the installation and usage duration, the update of the installation and usage duration can be achieved, which is convenient for re-determining the license feature set.
[0065] Step S401: Re-determine the license feature set according to the updated installation and usage duration, and restore the occasional feature points within the new license feature set to recognition feature points.
[0066] By updating the license feature set, further analysis of the accidental feature points can be achieved.
[0067] Step S402: Re-determine the overall proportion of accidents after the restoration of the recognition feature points. When the overall proportion of accidents is not less than the required proportion of accidents, update the installation and usage duration repeatedly according to the supplementary duration until the overall proportion of accidents is not less than the required proportion of accidents, and define the installation and usage duration corresponding to the situation where the overall proportion of accidents is not less than the required proportion of accidents as the original analysis duration.
[0068] Through the continuous update of the installation and usage duration by the supplementary duration, the actual duration of the current metal shaft can be determined. At this time, it is possible to determine some feature points that are actually accidental feature points, and at the same time, it is also possible to determine the feature points that are actually the change situation of the metal shaft, which is convenient for analyzing the overall usage situation of the metal shaft; defining the original analysis duration to distinguish different installation and usage durations is convenient for subsequent analysis.
[0069] Refer to Figure 5 , after the theoretical corresponding duration is determined, the intelligent detection method of the metal shaft further includes: Step S500: Calculate the difference between the original analysis duration and the theoretical corresponding duration to determine the analysis interval duration.
[0070] The analysis interval duration is the difference between the original analysis duration and the theoretical corresponding duration, and this difference is an absolute value.
[0071] Step S501: Judge whether the analysis interval duration is less than the preset limit interval duration.
[0072] The limit interval duration is the maximum analysis interval duration allowed when the staff determines that the original analysis duration and the theoretical corresponding duration are relatively close. The purpose of the judgment is to know whether the current determination of the accidental feature points is accurate.
[0073] Step S5011: If the analysis interval duration is less than the limit interval duration, calculate the remaining usage duration according to the theoretical corresponding duration.
[0074] When the analysis interval duration is less than the limit interval duration, it indicates that the actual usage duration of the metal shaft is relatively close to the installation and usage duration for analyzing the accidental feature points. That is, it is highly likely that the currently determined accidental feature points are indeed accidental occurrences. In this case, subsequent normal calculations can be performed.
[0075] Step S5012: If the analysis interval duration is not less than the limit interval duration, then all accidental feature points are restored to recognition feature points to re-determine the theoretical corresponding duration.
[0076] When the analysis interval duration is not less than the limit interval duration, it indicates that there is a significant difference between the installation and usage duration for determining the accidental feature points and the actual usage duration of the metal shaft. That is, the current determination of the accidental feature points is inaccurate. Therefore, the accidental feature points need to be restored to recognition feature points to re-analyze the theoretical corresponding duration, so as to accurately determine the actual usage situation of the metal shaft.
[0077] Refer to Figure 6 , after the combined reasonable deviation value is determined, the intelligent detection method for the metal shaft further includes: Step S600: Determine whether there are at least two test durations with the smallest and same combined reasonable deviation values.
[0078] The purpose of the determination is to find out whether there are multiple test durations that meet the requirements, so as to determine the unique theoretical corresponding duration subsequently.
[0079] Step S6001: If there are not at least two test durations with the smallest and same combined reasonable deviation values, then the test duration corresponding to the smallest combined reasonable deviation value is determined as the theoretical corresponding duration.
[0080] When there are not at least two test durations with the smallest and same combined reasonable deviation values, it indicates that there is only one test duration that meets the requirements. In this case, the theoretical corresponding duration can be determined normally according to this test duration.
[0081] Step S6002: If there are at least two test durations with the smallest and same combined reasonable deviation values, then the smallest combined reasonable deviation value is defined as the reference reasonable deviation value, and the corresponding test duration is defined as the alternative duration, and the two test durations adjacent to the alternative duration are defined as adjacent binding durations.
[0082] When there are at least two test durations with the smallest and same combined reasonable deviation values, it indicates that there are multiple test durations that meet the requirements. At this time, further analysis is needed to determine the unique and accurate theoretical corresponding duration; the reference reasonable deviation value, alternative duration, and adjacent binding durations are defined to distinguish different data for subsequent analysis.
[0083] Step S601: Define the combined reasonable deviation value determined by adjacent binding durations as the adjacent reasonable deviation value.
[0084] Define the adjacent reasonable deviation value to distinguish different combined reasonable deviation values, facilitating subsequent analysis.
[0085] Step S602: Calculate the difference between the adjacent reasonable deviation value and the reference reasonable deviation value to determine the adjacent overall deviation value.
[0086] The adjacent overall deviation value is the difference between the adjacent reasonable deviation value and the reference reasonable deviation value, and this difference is an absolute value.
[0087] Step S603: Calculate the average of two adjacent overall deviation values to determine the average adjacent deviation value, and calculate based on the average adjacent deviation value and the preset standard adjacent deviation value to determine the comparison deviation value.
[0088] The average adjacent deviation value is the average of two adjacent overall deviation values. The average adjacent deviation value is the deviation value that generally occurs between two adjacent test durations set by the staff. The comparison deviation value is the difference between the average adjacent deviation value and the standard adjacent deviation value, and this difference is an absolute value.
[0089] Step S604: Determine the comparison deviation value with the smallest value according to the sorting rule, and determine the theoretical corresponding duration as the alternative duration corresponding to this comparison deviation value.
[0090] Through the sorting rule, the comparison deviation value with the smallest value can be determined, that is, the specific situation of the metal shaft under the alternative duration corresponding to this comparison deviation value is closest to the metal shaft currently being detected. At this time, this alternative duration can be determined as the theoretical corresponding duration.
[0091] Refer to Figure 7 , based on the same inventive concept, an embodiment of the present invention provides a metal shaft intelligent detection system, including: An acquisition module, configured to acquire the circumferential surface image and the installation and use duration of the metal shaft; A processing module, connected to the acquisition module, for storing and processing information; The processing module performs feature point recognition in the circumferential surface image to determine the recognition feature type and the recognition feature level, and combines the recognition feature type and the recognition feature level to determine the recognition feature point; The acquisition module acquires the recognition feature points at each test duration from a preset test database, and enables the processing module to define this recognition feature point as the test feature point; The processing module counts according to the recognition feature points to determine the recognition feature quantity, and counts according to the test feature points to determine the test feature quantity; The processing module calculates the difference based on the number of recognition features and the number of test features to determine the increased demand quantity and the position of increased demand, and generates virtual blank points according to the increased demand quantity and supplements them to the position of increased demand; The processing module determines the corresponding feature combinations based on the one-to-one correspondence of the recognition feature points, test feature points, and virtual blank points, and determines the overall feature deviation value according to the corresponding feature combinations; The processing module determines the overall feature deviation value with the smallest value according to the preset sorting rule, and defines this overall feature deviation value as the combined reasonable deviation value at this test duration; The processing module determines the combined reasonable deviation value with the smallest value according to the sorting rule, and determines the test duration corresponding to this combined reasonable deviation value as the theoretical corresponding duration; The processing module calculates based on the installation and usage duration, the theoretical corresponding duration, and the preset theoretical usage duration to determine the remaining usage duration; The recognition feature point elimination module is used to determine and eliminate the accidentally appearing feature points, so as to facilitate the determination of the actual usage situation of the metal shaft; The occasional feature point analysis module analyzes the situation of the appearance of occasional feature points to determine whether to keep or remove the occasional feature points; The installation and usage duration update module is used to update the installation and usage duration to facilitate the determination of more accurate occasional feature points; The interval duration analysis module compares the installation and usage duration determined by the occasional feature points with the usage duration in the actual situation to determine the accuracy of the determination of the occasional feature points; The test duration screening module is used to screen multiple test durations that meet the requirements.
[0092] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the above division of each functional module is used as an example. In actual applications, the above functions can be assigned to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here.
Claims
1. An intelligent detection method for a metal shaft, characterized in that, Including: Obtain the circumferential surface image of the metal shaft and the installation and usage duration; Perform feature point recognition in the circumferential surface image to determine the recognition feature type and the recognition feature level, and combine the recognition feature type and the recognition feature level to determine the recognition feature points; Obtain the recognition feature points at each test duration in the preset test database, and define the recognition feature points as test feature points; Count according to the recognition feature points to determine the recognition feature quantity, and count according to the test feature points to determine the test feature quantity; Calculate the difference according to the recognition feature quantity and the test feature quantity to determine the required increase quantity and the required increase position, and generate virtual blank points according to the required increase quantity to supplement them into the required increase position; Determine the feature corresponding combination according to the one-to-one correspondence of the recognition feature points, the test feature points and the virtual blank points, and determine the overall feature deviation value according to the feature corresponding combination; Determine the overall feature deviation value with the smallest value according to the preset sorting rule, and define the overall feature deviation value as the combined reasonable deviation value at this test duration; Determine the combined reasonable deviation value with the smallest value according to the sorting rule, and determine the test duration corresponding to the combined reasonable deviation value as the theoretical corresponding duration; Calculate according to the installation and usage duration, the theoretical corresponding duration and the preset theoretical usage duration to determine the remaining usage duration.
2. The intelligent detection method for a metal shaft according to claim 1, wherein After the recognition feature points are determined, the intelligent detection method for the metal shaft further includes: Determine the permission feature set corresponding to the installation and usage duration according to the preset permission matching relationship; Judge whether the recognition feature points are within the permission feature set; If the recognition feature points are within the permission feature set, maintain the determined recognition feature points; If the recognition feature points are not within the permission feature set, eliminate the recognition feature points and define them as accidental feature points; Determine the feature influence duration corresponding to the accidental feature points according to the preset influence matching relationship; Perform a summation calculation according to all the feature influence durations to determine the feature loss duration, and perform a difference calculation according to the remaining usage duration and the feature loss duration to update the remaining usage duration.
3. The intelligent detection method for a metal shaft according to claim 2, characterized in that, After the accidental feature points are determined, the intelligent detection method for the metal shaft further includes: Count according to all the recognition feature points before the accidental feature points are eliminated to determine the feature quantity before elimination, and count according to the accidental feature points to determine the accidental feature quantity; Calculate according to the feature quantity before elimination and the accidental feature quantity to determine the overall accidental proportion; Judge whether the overall accidental proportion is less than the preset accidental requirement proportion; If the overall accidental proportion is less than the accidental requirement proportion, maintain the determined accidental feature points; If the overall accidental proportion is not less than the accidental requirement proportion, restore the determined accidental feature points to the recognition feature points.
4. The intelligent detection method for a metal shaft according to claim 3, wherein If the overall accidental proportion is not less than the accidental requirement proportion, the intelligent detection method for the metal shaft further includes: Perform a summation calculation according to the preset supplementary duration and the installation and usage duration to update the installation and usage duration; Redetermine the permission feature set according to the updated installation and usage duration, and restore the accidental feature points within the new permission feature set to the recognition feature points; After the accidental overall proportion is re - determined after the recognition feature points are restored, and when the accidental overall proportion is not less than the accidental demand proportion, the installation and usage duration is repeatedly updated according to the supplementary duration until the accidental overall proportion is not less than the accidental demand proportion, and the installation and usage duration corresponding to the situation where the accidental overall proportion is not less than the accidental demand proportion is defined as the original analysis duration.
5. The intelligent detection method for a metal shaft according to claim 4, characterized in that, After the theoretical corresponding duration is determined, the intelligent detection method for the metal shaft further includes: Calculating the difference between the original analysis duration and the theoretical corresponding duration to determine the analysis interval duration; Judging whether the analysis interval duration is less than a preset limit interval duration; If the analysis interval duration is less than the limit interval duration, calculate the remaining usage duration according to the theoretical corresponding duration; If the analysis interval duration is not less than the limit interval duration, restore all accidental feature points to the recognition feature points to re - determine the theoretical corresponding duration.
6. The intelligent detection method for metal shafts according to claim 1, wherein After the combined reasonable deviation value is determined, the intelligent detection method for the metal shaft further includes: Judging whether there are at least two test durations with the smallest and same combined reasonable deviation values; If there are not at least two test durations with the smallest and same combined reasonable deviation values, determine the test duration corresponding to the smallest combined reasonable deviation value as the theoretical corresponding duration; If there are at least two test durations with the smallest and same combined reasonable deviation values, define the smallest combined reasonable deviation value as the reference reasonable deviation value, define the corresponding test duration as the alternative duration, and define the two test durations adjacent to the alternative duration as the adjacent binding durations; Define the combined reasonable deviation value determined by the adjacent binding durations as the adjacent reasonable deviation value; Calculate the difference between the adjacent reasonable deviation value and the reference reasonable deviation value to determine the adjacent overall deviation value; Calculate the average value of the two adjacent overall deviation values to determine the average adjacent deviation value, and calculate according to the average adjacent deviation value and a preset standard adjacent deviation value to determine the comparison deviation value; Determine the comparison deviation value with the smallest value according to the sorting rule, and determine the alternative duration corresponding to the comparison deviation value as the theoretical corresponding duration.
7. An intelligent detection system for a metal shaft, characterized in that, Including: An acquisition module for acquiring the circumferential surface image of the metal shaft and the installation and usage duration; A processing module, connected to the acquisition module, for storing and processing information; The processing module performs feature point recognition in the circumferential surface image to determine the recognition feature type and the recognition feature level, and combines the recognition feature type and the recognition feature level to determine the recognition feature points; The acquisition module acquires the recognition feature points at each test duration from a preset test database, and enables the processing module to define the recognition feature points as test feature points; The processing module counts according to the recognition feature points to determine the recognition feature quantity, and counts according to the test feature points to determine the test feature quantity; The processing module calculates the difference between the recognition feature quantity and the test feature quantity to determine the demand increase quantity and the demand increase position, and generates virtual blank points according to the demand increase quantity to supplement them to the demand increase positions; The processing module determines the feature corresponding combinations according to the one - to - one correspondence of the recognition feature points, the test feature points, and the virtual blank points, and determines the feature overall deviation value according to the feature corresponding combinations; The processing module determines the overall feature deviation value with the smallest numerical value according to the preset sorting rule, and defines this overall feature deviation value as the combined reasonable deviation value at this test duration; The processing module determines the combined reasonable deviation value with the smallest numerical value according to the sorting rule, and determines the test duration corresponding to this combined reasonable deviation value as the theoretical corresponding duration; The processing module calculates based on the installation and usage duration, the theoretical corresponding duration, and the preset theoretical usage duration to determine the remaining usage duration.