Method for detecting quality defects of aluminum alloy profile for rail transit conductive rail

Abstract

The application discloses a kind of based on rail transit conductor rail aluminum alloy section quality defect detection method, belong to aluminum profile technical field.The application obtains the image data of target batch of aluminum alloy section by controlling preset camera, and according to the electrical conductivity data of the conductivity meter detection target batch of aluminum alloy section according to the image data, according to the electrical conductivity data and the image data determine the first quality characteristic data of the aluminum alloy section, according to the first quality characteristic data and the pre-set product standard determine the sampling information, according to the second quality characteristic data of the aluminum alloy section is collected according to the sampling information, and according to the first quality characteristic data and the second quality characteristic data as quality detection result, the type of second quality characteristic is destructive detection, the beneficial effect of improving the yield of high-quality aluminum alloy section is realized.

Description

Technical Field

[0001] This invention relates to the field of aluminum profile technology, and in particular to a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails. Background Technology

[0002] In rail transit, the power source for trains is generally provided by conductive rails. The conductive rails maintain continuous sliding contact with the train through current collectors, thereby providing a stable direct current to the train. Aluminum alloy conductive rails have significant advantages in conductivity, lightweight, and mechanical properties, making aluminum alloy the mainstream material for conductive rails. However, compared to ordinary aluminum alloys, aluminum alloy conductive rails have higher quality requirements. Current testing methods generally rely on random sampling to determine the quality of aluminum materials in the same batch. This can lead to the mixing of qualified and excellent products in the track, affecting the overall resistance and increasing power loss during later train operation.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails, aiming to improve the yield of high-quality aluminum alloy profiles. To achieve the above objective, this invention provides a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails, comprising the following steps:

[0005] The system controls a preset camera to acquire image data of a target batch of aluminum alloy profiles, and controls a conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data.

[0006] The first type of quality characteristic data of the aluminum alloy profile is determined based on the conductivity data and the image data;

[0007] The sampling inspection information is determined based on the first type of quality characteristic data and the preset product standards;

[0008] The second type of quality characteristic data of the aluminum alloy profile is collected based on the sampling inspection information, and the first type of quality characteristic data and the second type of quality characteristic data are used as the quality inspection results. The type of the second type of quality characteristic is destructive testing.

[0009] Optionally, the step of determining the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data includes:

[0010] The surface quality information of the target batch of aluminum alloy profiles is determined based on the image data and a preset image recognition algorithm.

[0011] The conductivity variation characteristics of the target batch of aluminum alloy profiles are determined based on the conductivity data and detection time.

[0012] Calculate the dimensional information of the aluminum alloy profile based on the image data;

[0013] The first type of quality characteristic data is determined based on the conductivity variation characteristics, surface quality information, and size information.

[0014] Optionally, the step of determining the surface quality information of the target batch of aluminum alloy profiles based on the image data and a preset image recognition algorithm includes:

[0015] The image data is divided into multiple recognition regions according to the image segmentation algorithm;

[0016] Surface defects in the identified area are detected using a deep learning classifier;

[0017] Calculate the defect feature data for each surface defect and count the number of defects in each identified area to obtain defect distribution information;

[0018] The surface quality information is determined based on the defect feature data and the defect distribution information.

[0019] Optionally, the step of determining the sampling inspection information based on the first type of quality characteristic data and the preset product standard includes:

[0020] A first quality scoring algorithm is generated based on the preset product standards;

[0021] Calculate the first quality score for each aluminum alloy profile based on the first type of quality characteristic data and the first quality scoring algorithm;

[0022] The sampling plan is generated based on the first quality score and used as the sampling information.

[0023] Optionally, the first type of quality feature data includes: first type of quality information corresponding to each aluminum alloy profile; the first quality scoring algorithm includes: a quality scoring formula; and the step of calculating the first quality score of each aluminum alloy profile based on the first type of quality feature data and the first quality scoring algorithm includes:

[0024] Determine the quality information of adjacent aluminum alloys produced in adjacent batches based on the inspection time of each aluminum alloy profile.

[0025] The confidence coefficient is determined by the first type of quality information and the quality information of the adjacent aluminum alloy;

[0026] The first quality score is calculated based on the confidence coefficient, the first type of quality information, and the quality scoring formula.

[0027] Optionally, the step of controlling the conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data includes:

[0028] Histogram statistics are performed on the grayscale information of the image data to obtain the image grayscale distribution map;

[0029] The detection location is determined based on the image grayscale distribution map;

[0030] The conductivity data of the target batch of aluminum alloy profiles is detected by controlling the conductivity meter according to the detection location.

[0031] Optionally, the step of collecting the second type of quality characteristic data of the aluminum alloy profile based on the sampling inspection information includes:

[0032] Based on the sampling information, the products to be destructively tested in the aluminum alloy profiles are determined.

[0033] The product to be tested is subjected to destructive testing to obtain the second type of quality data;

[0034] The second type of quality data is used as the second type of quality characteristic data of the aluminum alloy profile.

[0035] Furthermore, to achieve the above objectives, the present invention also provides a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails comprising:

[0036] The control module is used to control a preset camera to acquire image data of the target batch of aluminum alloy profiles, and to control a conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data.

[0037] The analysis module is used to determine the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data;

[0038] The selection module is used to determine sampling information based on the first type of quality characteristic data and preset product standards.

[0039] The detection module is used to collect the second type of quality characteristic data of the aluminum alloy profile according to the sampling information, and use the first type of quality characteristic data and the second type of quality characteristic data as the quality detection result. The type of the second type of quality characteristic is destructive testing.

[0040] Furthermore, to achieve the above objectives, the present invention also provides a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails. The quality defect detection device based on aluminum alloy profiles for rail transit conductive rails includes: a memory, a processor, and a quality defect detection program based on aluminum alloy profiles for rail transit conductive rails stored in the memory and executable on the processor. The quality defect detection program based on aluminum alloy profiles for rail transit conductive rails is configured to implement the steps of the quality defect detection method based on aluminum alloy profiles for rail transit conductive rails described in any of the above claims.

[0041] Furthermore, to achieve the above objectives, the present invention also provides a storage medium storing a quality defect detection program based on aluminum alloy profiles for rail transit conductive rails. When the quality defect detection program based on aluminum alloy profiles for rail transit conductive rails is executed by a processor, it implements the steps of the quality defect detection method based on aluminum alloy profiles for rail transit conductive rails described above.

[0042] This invention proposes a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails. The method involves controlling a pre-set camera to acquire image data of a target batch of aluminum alloy profiles, and then controlling a conductivity meter to detect the conductivity data of the target batch based on the image data. Based on the conductivity data and the image data, a first type of quality characteristic data for the aluminum alloy profiles is determined. Based on the first type of quality characteristic data and a pre-set product standard, sampling information is determined. This allows for the selection of aluminum alloy profiles requiring destructive testing based on non-destructive testing results. The first and second types of quality characteristic data are used as the quality inspection results, thereby improving the yield of high-quality aluminum alloy profiles and effectively reducing power consumption during later train operation. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the structure of a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, which is part of the hardware operating environment of the embodiment of the present invention.

[0044] Figure 2 This is a flowchart illustrating the first embodiment of the method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails according to the present invention.

[0045] Figure 3 This is a flowchart illustrating the second embodiment of the quality defect detection method for aluminum alloy profiles used in rail transit conductive rails according to the present invention.

[0046] Figure 4 This is a flowchart illustrating the third embodiment of the quality defect detection method for aluminum alloy profiles used in rail transit conductive rails according to the present invention.

[0047] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0048] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0049] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, which is part of the hardware operating environment of the embodiment of the present invention.

[0050] like Figure 1 As shown, the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, an interactive device 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The interactive device 1003 may include a display screen and an input unit such as a keyboard. Optionally, the interactive device 1003 may also connect to the communication bus via standard wired or wireless interfaces. The network interface 1004 may optionally include standard wired or wireless interfaces (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0051] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the quality defect detection equipment based on aluminum alloy profiles for rail transit conductive rails. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0052] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and a quality defect detection program based on aluminum alloy profiles for rail transit conductive rails.

[0053] exist Figure 1In the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails shown, the network interface 1004 is mainly used for data communication with other devices; the interactive device 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails of the present invention can be set in the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails. The quality defect detection device based on aluminum alloy profiles for rail transit conductive rails calls the quality defect detection program based on aluminum alloy profiles for rail transit conductive rails stored in the memory 1005 through the processor 1001, and executes the quality defect detection method based on aluminum alloy profiles for rail transit conductive rails provided in the embodiment of the present invention.

[0054] This invention provides a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of a method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails according to the present invention.

[0055] In this embodiment, the method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails includes:

[0056] Step S1: Control the preset camera to acquire image data of the target batch of aluminum alloy profiles, and control the conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data.

[0057] In this embodiment, the type and number of cameras are not limited. The camera obtains image data by photographing the aluminum alloy profile. Generally, before use, the physical display size represented by one pixel of the aluminum alloy profile at a preset position is determined. Therefore, collecting the image data is equivalent to obtaining the size information of the aluminum alloy profile, and subsequent analysis of the image data can yield information about the surface of the aluminum alloy profile. The conductivity meter here can be a standard portable eddy current conductivity meter, generally using a large-size surface probe, typically with a diameter of 8 mm. The probe measures the average conductivity within its coverage area. Since the conductivity measured at different points for normally extruded alloys will fluctuate relatively little, a default single-point detection or multi-point detection can be performed based on the image data. The average of the multiple data values ​​obtained after multi-point detection can be calculated as the conductivity data. Furthermore, since the target batch of aluminum alloy profiles can be multiple, image data and conductivity data for each aluminum alloy profile are generally collected in the order of production, and the corresponding collection timestamp is obtained during the collection process.

[0058] Step S2: Determine the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data;

[0059] In this embodiment, the conductivity data of each aluminum alloy profile is constructed into a corresponding conductivity data group according to the chronological order of the acquisition timestamps. Optionally, the fluctuation characteristics in the conductivity data group are extracted as the first type of quality feature data for the target batch of aluminum alloy profiles. Further, based on the image data, the data values ​​corresponding to multiple parameter types for each aluminum alloy profile in the target batch are determined. These parameter types include surface quality and aluminum alloy profile dimensions. Thus, the first type of quality feature data is determined based on the above data. It should be noted that the first type of quality feature here refers to the quality feature obtained through non-destructive testing. In addition, ultrasonic testing can also be used to determine whether there are internal defects.

[0060] Step S3: Determine the sampling inspection information based on the first type of quality characteristic data and the preset product standard;

[0061] In this embodiment, the first type of quality characteristic data is graded and quantified according to a preset product standard. This preset product standard refers to pre-defined parameter ranges for different grades of aluminum alloy profiles in different dimensions. Common preset product standards include: Excellent, Good, Qualified, and Unqualified. The parameter range for Excellent is smaller than that for Good, and the parameter range for Good is smaller than that for Qualified. When an aluminum alloy profile does not meet the parameter range for Qualified, it is marked as Unqualified. Specifically, based on the preset product standard, the grading of each aluminum alloy profile can be determined based on the first type of quality characteristic data, and aluminum alloy profiles for destructive testing can be selected according to the grading. Preferably, aluminum alloy profiles classified as Qualified are selected as products for destructive testing, thereby avoiding the selection of Unqualified aluminum alloy profiles for destructive testing, saving testing costs. Furthermore, selecting qualified aluminum alloy profiles can increase the ratio of Excellent to Good aluminum alloy profiles after destructive testing.

[0062] Step S4: Collect the second type of quality characteristic data of the aluminum alloy profile according to the sampling inspection information, and use the first type of quality characteristic data and the second type of quality characteristic data as the quality inspection result.

[0063] The second type of quality characteristic is destructive testing. In this embodiment, destructive testing may include mechanical property testing, wear resistance testing, and arc resistance testing. Optionally, grain structure can also be observed using a metallographic microscope. By combining the first type of quality characteristic data and the second type of quality characteristic data, the quality testing result can be obtained.

[0064] In this embodiment, image data of a target batch of aluminum alloy profiles is acquired by controlling a preset camera, and the conductivity data of the target batch of aluminum alloy profiles is detected by a conductivity meter based on the image data. The first type of quality characteristic data of the aluminum alloy profiles is determined based on the conductivity data and the image data, and the sampling information is determined based on the first type of quality characteristic data and the preset product standard. In this way, aluminum alloy profiles that need to be destructively tested can be selected based on the non-destructive test results, and the first type of quality characteristic data and the second type of quality characteristic data are used as the quality test results. This can improve the yield of high-quality aluminum alloy profiles and effectively reduce the power loss during the later operation of the train.

[0065] Furthermore, based on the first embodiment, a second embodiment of the present invention is proposed based on the method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails. In this embodiment, reference is made to... Figure 3 The step of determining the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data includes:

[0066] Step S21: Determine the surface quality information of the target batch of aluminum alloy profiles based on the image data and a preset image recognition algorithm;

[0067] In this embodiment, the preset image recognition algorithm may include edge extraction algorithm, deep learning algorithm, etc. The preset image recognition algorithm is used to identify the surface defects and corresponding number of defects in the target batch of aluminum alloy profiles, and the surface quality information is determined based on the surface defects and corresponding number of defects.

[0068] Step S22: Determine the conductivity variation characteristics of the target batch of aluminum alloy profiles based on the conductivity data and detection time;

[0069] Specifically, the detection time here refers to the detection time of conductivity data. It should be noted that the detection sequence is the same as the production sequence. By statistically analyzing the conductivity data over time, the data fluctuation of aluminum alloy profiles under the conductivity index can be determined.

[0070] Step S23: Calculate the dimensional information of the aluminum alloy profile based on the image data;

[0071] The number of pixels occupied by the aluminum alloy profile in the image is calculated, and the size information of the aluminum alloy profile is determined based on the correspondence between the number of occupied pixels and the actual size. This correspondence can be between pixels and actual dimensions.

[0072] Step S24: Determine the first type of quality characteristic data based on the conductivity change characteristics, surface quality information, and size information.

[0073] In this embodiment, the conductivity variation characteristics, surface quality information, and size information are used as the first type of quality feature data. Optionally, the above data can be used to construct a vector set as the first type of quality feature data.

[0074] In this embodiment, the surface quality information of the target batch of aluminum alloy profiles is determined by the image data and a preset image recognition algorithm. The conductivity change characteristics of the target batch of aluminum alloy profiles are determined by the conductivity data and the detection time. The size information of the aluminum alloy profiles is calculated by the image data. The first type of quality feature data is determined by the conductivity change characteristics, surface quality information and size information, thereby improving the accuracy of the detection.

[0075] Furthermore, the step of determining the surface quality information of the target batch of aluminum alloy profiles based on the image data and a preset image recognition algorithm includes:

[0076] The image data is divided into multiple recognition regions according to the image segmentation algorithm;

[0077] Surface defects in the identified area are detected using a deep learning classifier;

[0078] Calculate the defect feature data for each surface defect and count the number of defects in each identified area to obtain defect distribution information;

[0079] The surface quality information is determined based on the defect feature data and the defect distribution information.

[0080] In this embodiment, the image segmentation algorithm first divides the entire image into multiple 50mm×50mm regions along the track length direction; then, a classifier is used to identify scratches, inclusions, bubbles, and peeling in each region and generate corresponding defect labels. Here, the classifier can be the MobileNet-V3 classifier.

[0081] Optionally, when determining a surface quality score, the following steps can be performed: First, draw a minimum bounding rectangle for the detected defects, extracting multiple features such as aspect ratio, area, grayscale contrast, and distance to the nearest edge. Then, calculate the defect density and maximum contiguousness in each region. Finally, normalize the density, maximum contiguousness, and the largest region feature data among the minimum bounding rectangles corresponding to all defects in the region, and weight them to determine the surface quality score. The region feature data can be the area of ​​the minimum bounding rectangle. The density, maximum contiguousness, and largest region feature data can be normalized to the interval [0,1]. Since the defect density, maximum contiguousness, and largest region feature data are negatively correlated with the surface quality score, the higher the defect density, the smaller the normalized value; the higher the maximum contiguousness, the smaller the normalized value; and the larger the largest region feature data, the smaller the normalized value. This method eliminates the need for defect labeling and scores the surface quality based on the density, distribution, quantity, and size of the defects.

[0082] Preferably, the surface quality score corresponding to the region with the lowest surface quality score is selected as the surface quality score for the entire image data.

[0083] Alternatively, the average of the surface quality scores for each region can be calculated using a weighted average to obtain the surface quality score for the entire image data.

[0084] In this embodiment, the quality information of adjacent aluminum alloys produced in adjacent production is determined based on the inspection time of each aluminum alloy profile. The first type of quality information and the quality information of the adjacent aluminum alloys are used to determine a confidence coefficient. The first quality score is calculated based on the confidence coefficient, the first type of quality information and the quality scoring formula, thereby improving the accuracy of the first quality score and thus improving the accuracy of the subsequent sampling inspection information.

[0085] Furthermore, based on the first or second embodiment, a third embodiment of the present invention is proposed regarding the method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails, with reference to... Figure 4 The step of determining the sampling inspection information based on the first type of quality characteristic data and the preset product standard includes:

[0086] Step S31: Generate a first quality scoring algorithm based on the preset product standard;

[0087] The preset product standards here may include: conductivity tolerance, surface defect density limit, and dimensional tolerance. A scoring scheme or function is derived from the above data, for example: a 1% deviation in conductivity deducts 2 points, more than 3 scratches per square decimeter deducts 5 points, and a dimensional deviation of 0.1 mm deducts 3 points. This serves as the calculation method for each scoring dimension in the first scoring, and together with the quality scoring formula, constitutes the first quality scoring algorithm.

[0088] It should be noted that product standards may include multiple different dimensions, and different types of scores may be different. The scoring scheme described above is used to determine the scoring result of a specific scoring dimension, thereby obtaining specific conductivity scores, size scores, and surface quality scores, which can then be used to calculate the first quality score in subsequent formulas.

[0089] The above-mentioned deduction of 2 points for a 1% deviation in conductivity refers to a deduction of two points in the conductivity score.

[0090] More than 3 scratches per square decimeter deduct 5 points means deducting 5 points from the surface quality score;

[0091] A 3-point deduction for a 0.1mm deviation in size means a deduction of 3 points in the size assessment.

[0092] It should be noted that the conductivity score, surface quality score, and size score here will have corresponding default values ​​preset. Optionally, the value can be 100 points.

[0093] The target conductivity and target size of the aluminum alloy profile are determined according to the product standard. The deviation of conductivity is determined by the target conductivity and the conductivity data. The degree of dimensional deviation is determined by the target size and the size information.

[0094] It should be noted that this is only used to illustrate how to obtain the scoring results for each dimension, and is not intended to limit the scoring rules or deduction rules.

[0095] Step S32: Calculate the first quality score for each aluminum alloy profile based on the first type of quality feature data and the first quality scoring algorithm to obtain multiple first quality scores;

[0096] Based on the data of each aluminum alloy profile in the first type of quality characteristic data and the first quality scoring algorithm, the corresponding first quality score is output. The lower the score, the higher the defect risk.

[0097] Step S33: Generate a sampling plan based on multiple first quality scores as the sampling information.

[0098] In this embodiment, the sampling plan is dynamically generated according to the scoring range. It should be explained that products scoring below 60 points are considered unqualified and will not undergo destructive testing. Since destructive testing will damage the product, the higher the score, the smaller the sampling ratio. That is, in the example below, all products scoring above 90 points are considered excellent and are randomly sampled at a sampling ratio of 1%, etc., which ensures that a large number of excellent products are retained.

[0099] For example: Randomly sample all aluminum alloy profiles with scores below 70 but above or equal to 60, with a sampling rate of 10%; randomly sample all aluminum alloy profiles with scores between 70 and 90, with a sampling rate of 5%; and randomly sample all aluminum alloy profiles with scores above 90, with a sampling rate of 1%. Furthermore, random sampling can be replaced by Poisson sampling, with an average sampling rate of 10% for scores below 70 but above 60, 5% for scores between 70 and 90, and 1% for scores above 90. That is, the sampling rate differs for different score ranges, decreasing as the score increases. Therefore, compared to random sampling for destructive testing, this method can increase the proportion of excellent and good-quality aluminum alloy profiles.

[0100] Furthermore, the first type of quality characteristic data includes: first type of quality information corresponding to each aluminum alloy profile; the first quality scoring algorithm includes: a quality scoring formula; and the step of calculating the first quality score of each aluminum alloy profile based on the first type of quality characteristic data and the first quality scoring algorithm includes:

[0101] Determine the quality information of adjacent aluminum alloys produced in adjacent batches based on the inspection time of each aluminum alloy profile.

[0102] The confidence coefficient is determined based on the first type of quality information and the quality information of the adjacent aluminum alloys;

[0103] The first quality score is calculated based on the confidence coefficient, the first type of quality information, and the quality scoring formula.

[0104] The step of determining the confidence coefficient based on the first type of quality information and the adjacent aluminum alloy quality information includes:

[0105] Construct a data group by combining the first type of quality information and the first type of quality information corresponding to the adjacent aluminum alloy quality information;

[0106] Calculate the mean of the data set, and determine the standard deviation based on the mean;

[0107] The confidence coefficient is determined based on the standard deviation and a preset mapping table, and the confidence coefficient is negatively correlated with the standard deviation.

[0108] The preset mapping table includes a preset confidence coefficient corresponding to each standard deviation interval. The preset confidence coefficient corresponding to the standard deviation in the preset mapping table is used as the confidence coefficient.

[0109] Specifically, the first type of quality information includes: surface quality information, size information, and conductivity.

[0110] In this embodiment, based on the current detection time of the aluminum alloy profile, the first type of quality information of two adjacent profiles is taken from both the forward and backward timeframes to form three continuous samples. Optionally, the conductivity fluctuation data values ​​of the three samples are mapped to corresponding confidence coefficients, and the first quality score is calculated based on the confidence coefficients, the first type of quality information, and the quality scoring formula. Optionally, the quality scoring formula here is as follows:

[0111]

[0112] Among them, here For the first quality rating, here For surface quality scoring, here Rate the size here For conductivity scoring, here , , These are the scoring weights for each item, here... , , These are the confidence coefficients for each rating item. The rating weights here vary depending on the product. The confidence coefficients can range from [0,1], with a larger standard deviation resulting in a smaller confidence coefficient. , , This can be obtained based on step S31.

[0113] In this embodiment, by determining the quality information of adjacent aluminum alloys produced in adjacent production periods based on the inspection time of each aluminum alloy profile, a confidence coefficient is determined based on the first type of quality information and the quality information of the adjacent aluminum alloys, and the first quality score is calculated based on the confidence coefficient, the first type of quality information, and the quality scoring formula, thereby improving the reliability of the first quality score.

[0114] Furthermore, based on any of the above embodiments, a fourth embodiment of the present invention is proposed for the quality defect detection method of aluminum alloy profiles for rail transit conductive rails. The step of controlling the conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles according to the image data includes:

[0115] Histogram statistics are performed on the grayscale information of the image data to obtain the image grayscale distribution map;

[0116] The detection location is determined based on the image grayscale distribution map;

[0117] The conductivity data of the target batch of aluminum alloy profiles is detected by controlling the conductivity meter according to the detection location.

[0118] In this embodiment, the entire profile image is divided into blocks for grayscale histogram statistics. Specifically, the entire profile image is divided into multiple blocks, each of which may include multiple pixels. When the grayscale mean of a certain block is lower than the overall mean and the standard deviation is greater than a threshold, it is determined that the grains in that area are coarse or contain inclusions, and the conductivity may be abnormal. The area with the abnormal grayscale mean is then identified as the detection location. Preferably, the geometric center of the area with the abnormal grayscale mean can be directly used as the detection location. Furthermore, multi-point detection can be selected, and the user can choose between single-point and multi-point detection modes. In addition, when there are multiple areas with abnormal grayscale mean, multi-point detection is performed by default. During multi-point detection, additional detection points can be added based on the detection location. For example, five centimeters from the beginning and end can be used as mandatory test points, and control points can be added at equal intervals of five centimeters in the middle. In this embodiment, the conductivity meter probe is pressed point by point under the drive of a servo mechanism, collecting resistance values ​​multiple times per second, thereby ensuring that potential quality problems are completely covered and achieving efficient and accurate conductivity measurement guided by the image.

[0119] Furthermore, based on any of the above embodiments, a fifth embodiment of the present invention for the quality defect detection method of aluminum alloy profiles for rail transit conductive rails is proposed, wherein the step of collecting the second type of quality characteristic data of the aluminum alloy profiles according to the sampling inspection information includes:

[0120] Based on the sampling information, the products to be destructively tested in the aluminum alloy profiles are determined.

[0121] The product to be tested is subjected to destructive testing to obtain the second type of quality data;

[0122] The second type of quality data is used as the second type of quality characteristic data of the aluminum alloy profile.

[0123] In this embodiment, destructive testing can include tensile testing, hard indentation testing, intergranular corrosion testing, and fatigue impact testing. Information such as tensile strength, yield point, elongation, and Brinell hardness is recorded as the second type of quality characteristic data for the aluminum alloy profile.

[0124] Furthermore, this invention also proposes a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, the quality defect detection device based on aluminum alloy profiles for rail transit conductive rails comprising:

[0125] The control module is used to control a preset camera to acquire image data of the target batch of aluminum alloy profiles, and to control a conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data.

[0126] The analysis module is used to determine the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data;

[0127] The selection module is used to determine sampling information based on the first type of quality characteristic data and preset product standards.

[0128] The detection module is used to collect the second type of quality characteristic data of the aluminum alloy profile according to the sampling information, and use the first type of quality characteristic data and the second type of quality characteristic data as the quality detection result. The type of the second type of quality characteristic is destructive testing.

[0129] Furthermore, this invention also proposes a quality defect detection device based on aluminum alloy profiles for rail transit conductive rails. The quality defect detection device includes: a memory, a processor, and a quality defect detection program for aluminum alloy profiles for rail transit conductive rails stored in the memory and executable on the processor. The quality defect detection program for aluminum alloy profiles for rail transit conductive rails is configured to implement the steps of the quality defect detection method for aluminum alloy profiles for rail transit conductive rails described above.

[0130] Furthermore, this embodiment of the invention also proposes a storage medium storing a quality defect detection program based on aluminum alloy profiles for rail transit conductive rails. When the quality defect detection program based on aluminum alloy profiles for rail transit conductive rails is executed by a processor, it implements the steps of the quality defect detection method based on aluminum alloy profiles for rail transit conductive rails described above.

[0131] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0132] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0133] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0134] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails, characterized in that, The method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails includes the following steps: The system controls a preset camera to acquire image data of a target batch of aluminum alloy profiles, and controls a conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data. The first type of quality characteristic data of the aluminum alloy profile is determined based on the conductivity data and the image data; The sampling inspection information is determined based on the first type of quality characteristic data and the preset product standards; The second type of quality characteristic data of the aluminum alloy profile is collected based on the sampling inspection information, and the first type of quality characteristic data and the second type of quality characteristic data are used as the quality inspection results. The type of the second type of quality characteristic is destructive testing. The step of determining the sampling inspection information based on the first type of quality characteristic data and the preset product standard includes: A first quality scoring algorithm is generated based on the preset product standards; Based on the first type of quality characteristic data and the first quality scoring algorithm, a first quality score is calculated for each aluminum alloy profile, resulting in multiple first quality scores. A sampling plan is generated based on multiple first quality scores as the sampling information; The first type of quality feature data includes: first type of quality information corresponding to each aluminum alloy profile, the first type of quality information including surface quality information, dimensional information, and conductivity; the first quality scoring algorithm includes: a quality scoring formula; the step of calculating the first quality score of each aluminum alloy profile based on the first type of quality feature data and the first quality scoring algorithm includes: Based on the inspection time of each aluminum alloy profile, the quality information of adjacent aluminum alloys produced in adjacent production is determined. The quality information of adjacent aluminum alloys is the first type of quality information corresponding to the aluminum alloy profiles produced in adjacent production. The confidence coefficient is determined based on the first type of quality information and the quality information of the adjacent aluminum alloys; The first quality score is calculated based on the confidence coefficient, the first type of quality information, and the quality scoring formula. The step of determining the confidence coefficient based on the first type of quality information and the adjacent aluminum alloy quality information includes: Construct a data group from the first type of quality information and the adjacent aluminum alloy quality information; Calculate the mean of the data set, and determine the standard deviation based on the mean; The confidence coefficient is determined based on the standard deviation and a preset mapping table, and the confidence coefficient is negatively correlated with the standard deviation.

2. The method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails as described in claim 1, characterized in that, The step of determining the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data includes: The surface quality information of the target batch of aluminum alloy profiles is determined based on the image data and a preset image recognition algorithm. The conductivity variation characteristics of the target batch of aluminum alloy profiles are determined based on the conductivity data and detection time. Calculate the dimensional information of the aluminum alloy profile based on the image data; The first type of quality characteristic data is determined based on the conductivity variation characteristics, surface quality information, and size information.

3. The method for detecting quality defects in aluminum alloy profiles used in rail transit conductive rails as described in claim 2, characterized in that, The step of determining the surface quality information of the target batch of aluminum alloy profiles based on the image data and a preset image recognition algorithm includes: The image data is divided into multiple recognition regions according to the image segmentation algorithm; Surface defects in the identified area are detected using a deep learning classifier; Calculate the defect feature data for each surface defect and count the number of defects in each identified area to obtain defect distribution information; The surface quality information is determined based on the defect feature data and the defect distribution information.

4. The method for detecting quality defects in aluminum alloy profiles for rail transit conductive rails as described in any one of claims 1 to 3, characterized in that, The step of controlling the conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data includes: Histogram statistics are performed on the grayscale information of the image data to obtain the image grayscale distribution map; The detection location is determined based on the image grayscale distribution map; The conductivity data of the target batch of aluminum alloy profiles is detected by controlling the conductivity meter according to the detection location.

5. The method for detecting quality defects in aluminum alloy profiles for rail transit conductive rails as described in any one of claims 1 to 3, characterized in that, The step of collecting the second type of quality characteristic data of the aluminum alloy profile based on the sampling inspection information includes: Based on the sampling information, the products to be destructively tested in the aluminum alloy profiles are determined. The product to be tested is subjected to destructive testing to obtain the second type of quality data; The second type of quality data is used as the second type of quality characteristic data of the aluminum alloy profile.

6. A quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, characterized in that, The quality defect detection device based on aluminum alloy profiles for rail transit conductive rails includes: The control module is used to control a preset camera to acquire image data of the target batch of aluminum alloy profiles, and to control a conductivity meter to detect the conductivity data of the target batch of aluminum alloy profiles based on the image data. The analysis module is used to determine the first type of quality characteristic data of the aluminum alloy profile based on the conductivity data and the image data; The selection module is used to determine sampling inspection information based on the first type of quality characteristic data and the preset product standard; the step of determining the sampling inspection information based on the first type of quality characteristic data and the preset product standard includes: A first quality scoring algorithm is generated based on the preset product standards; Based on the first type of quality characteristic data and the first quality scoring algorithm, a first quality score is calculated for each aluminum alloy profile, resulting in multiple first quality scores. A sampling plan is generated based on multiple first quality scores as the sampling information; The first type of quality feature data includes: first type of quality information corresponding to each aluminum alloy profile, the first type of quality information including surface quality information, dimensional information, and conductivity; the first quality scoring algorithm includes: a quality scoring formula; the step of calculating the first quality score of each aluminum alloy profile based on the first type of quality feature data and the first quality scoring algorithm includes: Based on the inspection time of each aluminum alloy profile, the quality information of adjacent aluminum alloys produced in adjacent production is determined. The quality information of adjacent aluminum alloys is the first type of quality information corresponding to the aluminum alloy profiles produced in adjacent production. The confidence coefficient is determined based on the first type of quality information and the quality information of the adjacent aluminum alloys; The first quality score is calculated based on the confidence coefficient, the first type of quality information, and the quality scoring formula. The step of determining the confidence coefficient based on the first type of quality information and the adjacent aluminum alloy quality information includes: Construct a data group from the first type of quality information and the adjacent aluminum alloy quality information; Calculate the mean of the data set, and determine the standard deviation based on the mean; The confidence coefficient is determined based on the standard deviation and a preset mapping table, and the confidence coefficient is negatively correlated with the standard deviation; The detection module is used to collect the second type of quality characteristic data of the aluminum alloy profile according to the sampling information, and use the first type of quality characteristic data and the second type of quality characteristic data as the quality detection result. The type of the second type of quality characteristic is destructive testing.

7. A quality defect detection device based on aluminum alloy profiles for rail transit conductive rails, characterized in that, The quality defect detection device based on aluminum alloy profiles for rail transit conductive rails includes: a memory, a processor, and a quality defect detection program for aluminum alloy profiles for rail transit conductive rails stored in the memory and executable on the processor. The quality defect detection program for aluminum alloy profiles for rail transit conductive rails is configured to implement the steps of the quality defect detection method for aluminum alloy profiles for rail transit conductive rails as described in any one of claims 1 to 5.

8. A storage medium, characterized in that, The storage medium stores a quality defect detection program based on aluminum alloy profiles for rail transit conductive rails. When the processor executes the quality defect detection program based on aluminum alloy profiles for rail transit conductive rails, it implements the steps of the quality defect detection method based on aluminum alloy profiles for rail transit conductive rails as described in any one of claims 1 to 5.

Images