Maintenance system and method of test fixture and storage medium

Abstract

The invention provides a maintenance system and method of a test fixture and a storage medium, which can be applied to the technical field of maintenance of the test fixture. The maintenance system of the test fixture comprises a control device which is used for identifying a first image of the test fixture and determining the position information of a foreign matter when identifying that the foreign matter exists in an assembly of the test fixture from the first image, the first image of the test fixture is obtained by carrying out image acquisition on an assembly of the test fixture through an image acquisition device, the assembly comprises a movable to-be-tested plate and a bearing assembly for bearing the to-be-tested plate, and the to-be-tested plate is used for testing a tested object by the test fixture; the operation of the maintenance device is controlled according to the position information of the foreign matter and the bearing state of the bearing assembly, and the bearing state indicates whether the bearing assembly bears the to-be-tested plate or not; and the maintenance device is used for executing maintenance operation aiming at the test fixture under the control of the control device.

Description

Technical Field

[0001] The present invention relates to the technical field of test fixture maintenance, and in particular to a test fixture maintenance system, method and storage medium. Background Art

[0002] With the rapid development of cloud computing, big data, and artificial intelligence technologies, the scale of global data center construction continues to expand, and the requirements for server hardware performance are increasing exponentially. As the core component of a server, the stability and functionality of the control board directly determine the overall performance and reliability of the server. The control board in each server undergoes multiple tests and verifications before leaving the factory. Test fixtures are used during the control board testing and verification process, and they also require regular maintenance during use. Currently, the maintenance process for test fixtures suffers from low intelligence and incomplete cleaning, which not only reduces the maintenance efficiency of the test fixtures but also increases their failure rate. Summary of the Invention

[0003] In view of the above problems, the present invention provides a test fixture maintenance system, method and storage medium that improve the maintenance efficiency of the test fixture and reduce the failure rate of the test fixture.

[0004] One aspect of the present invention provides a maintenance system for a test jig, the system comprising: a control device for identifying a first image of the test jig and, if the first image identifies the presence of a foreign object in a component of the test jig, determining position information of the foreign object, wherein the first image of the test jig is obtained by image acquisition equipment performing image acquisition on the components of the test jig, the components comprising a movable test board and a supporting assembly for supporting the test board, the test board being used to enable the test jig to test an object under test; and controlling operation of a maintenance device based on the position information of the foreign object and a supporting state of the supporting assembly, the supporting state indicating whether the supporting assembly is supporting the test board; and a maintenance device for performing maintenance operations on the test jig under the control of the control device.

[0005] Another aspect of the present invention provides a maintenance method for a test jig, the method comprising: identifying a first image of the test jig; determining location information of the foreign object when a foreign object is identified in a component of the test jig from the first image; and controlling operation of a maintenance device based on the location information of the foreign object and a carrying state of the carrying component, wherein the carrying state indicates whether the carrying component is carrying a board to be tested.

[0006] Another aspect of the present invention further provides a computer-readable storage medium having a computer program or instructions stored thereon, which implements the steps of the above-mentioned test fixture maintenance method when the computer program or instructions are executed by a processor.

[0007] According to an embodiment of the present invention, a test jig is maintained by providing a control device and a maintenance device. The control device is used to identify a first image of the test jig and, when a foreign object is identified, determine the location information of the foreign object, and control the maintenance device to perform maintenance operations on the test jig based on the location information of the foreign object and the load status of the load-bearing component. During the maintenance process, the entire maintenance process, from scanning the test jig, identifying foreign objects, outputting the location of foreign objects to performing dual-mode foreign object cleaning based on the location and load status, can be fully automated. Moreover, by using machine equipment to identify foreign objects, the accuracy of foreign objects can be improved compared to manual identification of foreign objects with the naked eye, and thorough cleaning of foreign objects can be achieved, thereby achieving the technical effect of improving the maintenance efficiency of the test jig and reducing the failure efficiency of the test jig. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above contents and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0009] Figure 1 An application scenario diagram of a test fixture maintenance system according to an embodiment of the present invention is shown;

[0010] Figure 2 FIG2 shows an architecture diagram of a test fixture maintenance system according to an embodiment of the present invention;

[0011] Figure 3 FIG2 shows an architecture diagram of a test fixture maintenance system according to another embodiment of the present invention;

[0012] Figure 4 A flowchart of a test fixture maintenance method according to an embodiment of the present invention is shown;

[0013] Figure 5 A flow chart showing a foreign body recognition module identifying a foreign body and outputting the coordinates of the foreign body's position according to an embodiment of the present invention is shown;

[0014] Figure 6 shows a processing flow chart of a maintenance prediction model according to an embodiment of the present invention;

[0015] Figure 7 A flowchart of a test fixture maintenance method according to another embodiment of the present invention is shown;

[0016] Figure 8 The block diagram schematically shows an electronic device suitable for implementing a maintenance method for a test fixture according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concept of the present invention.

[0018] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the presence of the features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0019] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0020] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0021] As data center construction continues to expand globally, the demand for server hardware performance is increasing exponentially. Controller boards, such as motherboards, storage controller cards, and graphics processing unit (GPU) boards, are core server components. The stability and functionality of these components directly determine the overall performance and reliability of the server. Every server controller board undergoes multiple tests and verifications before shipment, including functional testing (FCT), in-circuit testing (ICT), and boundary scan testing (BSI). FCT provides a simulated operating environment for the target board, subjecting it to various design states to obtain parameters in each state and verifying its functionality. ICT is a testing technique that doesn't disconnect the circuit or remove component pins. BSI is a board-level circuit testing technology that non-invasively tests the solder joint quality, signal connectivity, and logic functionality of integrated circuits on computer boards by inserting boundary scan cells between the chip pins and the internal logic.

[0022] To improve test efficiency and ensure test quality during ICT, BSI, and FCT testing, customized test fixtures are developed for each test project. Test fixtures require regular maintenance during use. The fixture's exterior must be inspected, and the various modules, lower case, carrier plate, and upper mold must be checked for dirt and damage to ensure proper function. However, test fixture maintenance can be highly dependent on manual labor, incomplete inspections, low maintenance efficiency, interference with the test process, unpredictable cleaning results, and untimely maintenance.

[0023] Maintenance and cleaning efficiency is low, mainly reflected in manual maintenance, which requires operators to use a brush or cotton cloth with cleaning alcohol to remove debris (such as tin slag and plastic residue). In a printed circuit board assembly factory, the number of test fixtures normally tested on the production line is generally between 300 and 500. If all of them rely on operators for manual inspection and cleaning, maintenance efficiency is very low, affecting the normal use of the test fixtures.

[0024] This disruption to the test process primarily occurs when manual maintenance of the test fixture must be performed during shift changes or when the fixture is idle. Otherwise, testing must be stopped to allow for fixture maintenance. Checking all test fixture modules individually is time-consuming and requires dedicated personnel, disrupting the normal testing process on the production line.

[0025] The cleaning effect is uncontrollable, mainly because it is difficult to fully inspect and completely remove smaller foreign matter, such as debris smaller than 0.1mm³, through manual maintenance. Residual foreign matter can easily lead to misjudgment during board testing, affecting the first-time pass rate and test efficiency of the board test.

[0026] Although the industry currently uses semi-automatic cleaning devices to maintain and clean test fixtures, they use a combination of brush heads and slide rails. This method still has shortcomings such as incomplete inspection, untimely maintenance, and single functionality. The incomplete inspection and untimely maintenance are mainly reflected in the fact that the semi-automatic maintenance method requires manual inspection to detect dirt in the test fixture, and then clean it with the brush or vacuum cleaner in the cleaning device. If the dirt is not found manually, maintenance and cleaning cannot be carried out. The single functionality is mainly reflected in the inability to identify and clean smaller foreign objects through manual inspection, resulting in omissions during the maintenance process and the inability to fully maintain the test fixture.

[0027] In view of this, an embodiment of the present invention provides a maintenance system for a test fixture. After using a multispectral camera to capture an image of the test fixture and identify foreign objects in the image, a piezoelectric vibrator is used to loosen the foreign objects and a magnetically controlled suction nozzle is used to absorb the foreign objects. At the same time, combined with a maintenance prediction model, the time period when the test fixture requires maintenance in the future is predicted. The maintenance system can automatically adjust the fixture maintenance strategy based on the prediction results, thereby realizing self-maintenance of the fixture, improving fixture maintenance efficiency, reducing testing costs, and reducing the failure rate of the test fixture.

[0028] Figure 1 A diagram showing an application scenario of a test fixture maintenance system according to an embodiment of the present invention is shown.

[0029] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a test fixture 101 and a maintenance system 102. The test fixture 101 may include a movable test board 1011 and a supporting assembly 1012 for supporting the test board 1011. The supporting assembly 1012 may include a fixture middle frame 1012-1 and a test board carrier 1012-2. The maintenance system 102 may include a control device 1021 and a maintenance device 1022. Both the control device 1021 and the maintenance device 1022 are electrically connected to the test board 1011 and the supporting assembly 1012.

[0030] Control device 1021 can scan test board 1011 and support assembly 1012 of test fixture 101 to obtain a first image. If the first image identifies a foreign object in a component of the test fixture, it can determine the foreign object's location information and control the operation of maintenance device 1022 based on the foreign object's location information and the support status of the support assembly. Maintenance device 1022 is used to perform maintenance operations on test fixture 101 under the control of control device 1021.

[0031] It should be understood that Figure 1 The number of boards under test 1011, jig midframe 1012-1, board-under-test carrier 1012-2, control device 1021, and maintenance device 1022 is merely illustrative. Any number of boards under test 1011, jig midframe 1012-1, board-under-test carrier 1012-2, control device 1021, and maintenance device 1022 may be provided as needed.

[0032] The following will be based on Figure 1 The scene described by Figures 2 to 6 A maintenance system for a test fixture according to an embodiment of the present invention is described in detail.

[0033] Figure 2 The figure shows an architecture diagram of a test fixture maintenance system according to an embodiment of the present invention.

[0034] like Figure 2 As shown, the test fixture maintenance system 102 of this embodiment includes a control device 1021 and a maintenance device 1022 .

[0035] The control device 1021 is configured to identify a first image of the test jig and, if the first image identifies a foreign object in a component of the test jig, determine the location information of the foreign object. The first image of the test jig is obtained by capturing the components of the test jig using an image capture device. The components include a movable test board and a supporting assembly for supporting the test board. The test board is used to enable the test jig to test the object under test. The control device 1021 is configured to control the operation of the maintenance device 1022 based on the location information of the foreign object and the supporting status of the supporting assembly. The supporting status indicates whether the supporting assembly is supporting the test board.

[0036] The maintenance device 1022 is used to perform maintenance operations on the test fixture under the control of the control device 1021 .

[0037] Figure 3 FIG. 1 shows an architecture diagram of a test fixture maintenance system according to another embodiment of the present invention.

[0038] like Figure 3As shown, the control device 1021 may include an edge server 1021-1, a foreign object identification and positioning module 1021-2, and a visual recognition module 1021-3. The edge server 1021-1 can not only be used to control the operation of the foreign object identification and positioning module 1021-2, the visual recognition module 1021-3, and the maintenance device 1022, and connect the foreign object identification and positioning module 1021-2, the visual recognition module 1021-3, and the maintenance device 1022 together, but can also call a maintenance prediction model to predict the pre-maintenance period and store a foreign object distribution map. By integrating the foreign object identification and positioning module 1021-2, the visual recognition module 1021-3, and the maintenance device 1022 through the edge server, automated maintenance of the test fixture can be achieved.

[0039] Edge server 1021-1 can control the image acquisition device of visual recognition module 1021-3 to acquire a first image of a component of a test fixture. The components of the test fixture include a movable test board and a supporting assembly that supports the test board. The supporting assembly includes a fixture middle frame and a test board carrier. The test board is used to enable the test fixture to test the object under test.

[0040] The image acquisition device of visual recognition module 1021-3 may include a multispectral camera. A multispectral camera can detect foreign objects using visible light with a wavelength of 400nm to 700nm and shortwave infrared light with a wavelength of 900nm to 1700nm. Visible light can be used to detect organic residues, while shortwave infrared light can detect metal debris, thus achieving dual-band collaborative operation. Compared with traditional red-green-blue (RGB) cameras, multispectral cameras can more accurately detect foreign objects. Compared with RGB cameras, multispectral cameras have advantages such as higher accuracy and less restrictions on lighting environments. The main differences between multispectral cameras and RGB cameras are shown in Table 1.

[0041] Table 1

[0042]

[0043] When the multispectral camera obtains a first image, edge server 1021-1 can control foreign object identification and positioning module 1021-2 to identify the first image. If no foreign object is identified in the first image, the foreign object indicator light on the test fixture can turn green. If a foreign object is identified, the foreign object's three-dimensional coordinates can be generated based on the foreign object, and its location information can be obtained. Based on the foreign object's location information, a foreign object removal path can be automatically calculated. Based on the load status of the load-bearing component, the maintenance device can be controlled to perform maintenance on the foreign object.

[0044] The maintenance device may include a piezoelectric vibrator and a magnetic adsorption component. The carrying state may include the carrying component carrying the board to be tested, and the carrying component not carrying the board to be tested. When the carrying component carries the board to be tested, only the magnetic adsorption component may be controlled to adsorb foreign matter, thereby avoiding loosening of the component welding points on the board to be tested due to the vibration of the piezoelectric vibrator, thereby avoiding damage to the board to be tested caused by the piezoelectric vibrator. When the carrying component does not carry the board to be tested, the vibration of the piezoelectric vibrator may be controlled to loosen the foreign matter, and the magnetic adsorption component may be controlled to adsorb the loosened foreign matter, thereby achieving cleaning and maintenance of the foreign matter. The magnetic adsorption component may include a magnetic suction nozzle, and the caliber of the magnetic suction nozzle may be adapted to the size of the foreign matter.

[0045] According to an embodiment of the present invention, a test jig is maintained by providing a control device and a maintenance device. The control device is used to identify a first image of the test jig and, when a foreign object is identified, determine the location information of the foreign object, and control the maintenance device to perform maintenance operations on the test jig based on the location information of the foreign object and the load status of the load-bearing component. During the maintenance process, the entire maintenance process, from scanning the test jig, identifying foreign objects, outputting the location of foreign objects to performing dual-mode foreign object cleaning based on the location and load status, can be fully automated. Moreover, by using machine equipment to identify foreign objects, the accuracy of foreign objects can be improved compared to manual identification of foreign objects with the naked eye, and thorough cleaning of foreign objects can be achieved, thereby achieving the technical effect of improving the maintenance efficiency of the test jig and reducing the failure efficiency of the test jig.

[0046] Figure 4 A flow chart of a test fixture maintenance method according to an embodiment of the present invention is shown.

[0047] Based on the above-mentioned test fixture maintenance system, an embodiment of the present invention provides a maintenance method for the above-mentioned test fixture. The flowchart of the maintenance method is as follows: Figure 4 shown.

[0048] Figure 4 The test fixture maintenance method shown may include operations S401 to S426 . Figure 4The maintenance process of the test fixture shown starts after the fixture completes the test and is divided into two parts: self-maintenance before placing the board to be tested and self-maintenance after placing the board to be tested. Self-maintenance before placing the board to be tested includes operations S401 to S413, which refers to the advance maintenance of the board carrier and the middle frame of the test fixture before the board to be tested is placed in the fixture to ensure that there are no foreign objects inside the test fixture. Self-maintenance after placing the board to be tested includes operations S414 to S426, which mainly detects foreign objects on the board to be tested itself to ensure that there are no foreign objects on the board to be tested itself to avoid affecting the test results. By switching the process from cleaning by vibration and adsorption before loading the board to be tested to cleaning only by adsorption after loading the board to be tested, a dual-mode maintenance process can be achieved, thereby improving the maintenance diversity of the test fixture.

[0049] In operation S401 , the test fixture is opened after completing the test task.

[0050] In operation S402 , the board to be tested is removed.

[0051] After the test fixture completes the test, the lid will be opened and the operator or robotic arm will remove the board to be tested. After the test fixture detects that the board to be tested has been removed, the visual recognition module will start scanning and use the multispectral camera to capture images of the fixture frame and the board carrier to identify whether any foreign objects were generated during the previous round of testing.

[0052] In operation S403 , the visual recognition module starts scanning.

[0053] In operation S404 , a multispectral camera captures an image. The multispectral camera scans the components of the test fixture to obtain a first image.

[0054] In operation S405 , whether a foreign object is recognized, the foreign object recognition and positioning module performs operation S406 if a foreign object is recognized, and performs operation S314 if no foreign object is recognized.

[0055] In operation S406 , three-dimensional coordinates of the foreign object are generated, and position information of the foreign object is determined.

[0056] In operation S407 , a foreign matter removal path is generated.

[0057] In operation S408 , the piezoelectric vibrator is activated.

[0058] In operation S409 , the aperture of the magnetically controlled nozzle is adjusted.

[0059] In operation S410 , negative pressure adsorption is performed.

[0060] In operation S411, whether a foreign object is detected again can be determined by re-activating the visual recognition module and scanning the components of the test fixture to obtain a second image. If a foreign object is detected again, the process can be restarted from operation S308. If no foreign object is detected, the process can be continued to operation S412.

[0061] In operation S412 , coordinates of foreign objects are marked on the foreign object distribution map.

[0062] In operation S413 , the maintenance prediction model predicts a planned maintenance period for the next maintenance of the test fixture, and the operation is then executed starting from operation S414 .

[0063] In operation S414 , the prompt light turns green, indicating that the board to be tested can be placed.

[0064] In operation S415 , a board to be tested is placed.

[0065] In operation S416 , the visual recognition module starts scanning.

[0066] In operation S417 , the multispectral camera captures an image. The multispectral camera scans the components of the test fixture to obtain a first image.

[0067] In operation S418 , whether a foreign object is recognized, the foreign object recognition and positioning module performs operation S419 if a foreign object is recognized, and performs operation S426 if no foreign object is recognized.

[0068] In operation S419 , three-dimensional coordinates of the foreign matter are generated, and position information of the foreign matter is determined.

[0069] In operation S420 , a foreign matter removal path is generated.

[0070] In operation S421 , the aperture of the magnetically controlled nozzle is adjusted.

[0071] In operation S422 , negative pressure adsorption is performed.

[0072] In operation S423, whether a foreign object is scanned again can be determined by restarting the visual recognition module to scan again. If a foreign object is scanned again, the process can be started from operation S421. If no foreign object is scanned, the process can be started from operation S424.

[0073] In operation S424 , coordinates of foreign objects are marked on the foreign object distribution map.

[0074] In operation S425 , the maintenance prediction model predicts a planned maintenance period for the next maintenance of the test fixture.

[0075] In operation S426 , a test task of the test fixture is started.

[0076] In some embodiments, after image acquisition, the foreign object recognition module uses a recognition algorithm (such as the You Only Look Once (YOLO) algorithm) to identify foreign objects and outputs their location coordinates if a foreign object is detected. If no foreign object is detected, the fixture's foreign object indicator turns green, indicating that the board under test can be placed. The operator can then place the board under test into the fixture for testing.

[0077] In some embodiments, the above operation S405 can be determined by: extracting multi-scale features of the first image; weighting the features for the component based on the weight of the component to obtain a weighted feature map; determining multiple candidate boxes including features in the weighted feature map; when the confidence of the candidate box is greater than a predetermined confidence, determining that there is a foreign object in the component identified from the first image to the test fixture, and the confidence is used to describe the probability that the feature in the candidate box is a foreign object.

[0078] The above operation S406 can be determined in the following manner: when the feature is related to the candidate frame, the position information of the foreign object is obtained based on the position information of the candidate frame; when the feature is related to at least two candidate frames, the target frame is determined from at least two candidate frames based on the confidence of the candidate frame, and the position information of the foreign object is obtained based on the position information of the target frame.

[0079] Figure 5 A flow chart showing a foreign object recognition module identifying a foreign object and outputting the coordinates of the foreign object's position according to an embodiment of the present invention is shown.

[0080] like Figure 5 As shown, the process of the foreign object recognition module identifying foreign objects and outputting the foreign object position coordinates may include operations S501 to S509. The above-mentioned operation S405 can also be determined through operations S501 to S505, and operation S406 can also be determined through operations S506 to S509.

[0081] In operation S501 , a first image is input.

[0082] In operation S502 , the first image is pre-processed.

[0083] For the inspection components of test fixtures, the preprocessing step can enhance the feature retention of tiny foreign objects (such as 0.1mm debris), which is different from general image preprocessing and lays the foundation for subsequent image feature extraction and detection.

[0084] In operation S503 , multi-scale features of the first image are extracted.

[0085] An improved Cross Stage Partial Fusion (C2f) module extracts multi-scale features from the first image through cross-stage feature diversion and fusion. The scale is determined by the probability that a feature is a foreign object. For example, when the probability of a feature being a foreign object is greater than a first predetermined value, it is considered a key foreign object; when the probability of a feature being a foreign object is less than or equal to the first predetermined value, it is considered a secondary foreign object. This optimizes the feature extraction path to address the small-target, multi-material nature of test fixture inspection, addressing the problem of traditional networks missing features of small foreign objects in the complex environment of test fixtures.

[0086] In some embodiments, small targets can refer to foreign objects or defects (such as tiny scratches or welding slag) on ​​test fixtures, which may only occupy a few pixels in the image. Traditional networks easily lose these details when downsampling. Multi-materials can refer to the fact that test fixtures may be made of a variety of materials such as metal, plastic, and silicone. The differences in reflectivity and texture make traditional feature extractors incapable of generalization. Complex environments can refer to industrial scenes with noise such as uneven lighting and oil stains, which further complicate feature separation. By introducing the C2f module to extract multi-scale features, we can prevent the dilution of small target information in deep networks.

[0087] In operation S504 , multi-scale features are fused.

[0088] By combining spatial pyramid pooling with cross-stage convolution, the multi-scale features of the test fixture image (such as the local features of 0.1mm debris and the global features of the overall board structure) are fused to generate a feature map that contains details and semantics. Details can be fine-grained, high-resolution local features in the image, such as scratches (only 1-2 pixels wide), solder burrs, and the clear outlines of tiny foreign objects (dust / fibers). Semantics include the semantics of scratches (linear features) and stains (blocky features). This adapts to the scenarios in which tiny foreign objects coexist with complex backgrounds during test fixture inspection, ensuring that the small target features of the foreign objects are not overwhelmed by large structural features.

[0089] In operation S505, the multi-scale features are weighted.

[0090] Based on the weight of the component, the features for the component are weighted to obtain a weighted feature map. For example, a global attention mechanism is introduced to perform feature weighting on the features of high-risk components or areas (such as probe gaps, slots) for foreign objects in the first image, suppressing background noise, such as board textures and normal components, in order to simulate the focus of attention during manual inspection and improve the sensitivity of foreign object detection in key areas of the fixture. In some embodiments, multiple candidate boxes including features in the weighted feature map can be determined, and when the confidence of the candidate box is greater than a predetermined confidence level, it is determined that there is a foreign object in the test fixture. The confidence level can be the probability that the feature in the candidate box is a foreign object.

[0091] In operation S506 , the multi-scale features are output and dynamic label matching is performed.

[0092] In another embodiment, a task-aligned assigner can be used to dynamically assign labels based on the multi-scale features of the feature pyramid. For each candidate box in the weighted feature map, the real foreign body box (such as conductive debris, insulating impurities) is dynamically matched according to the classification confidence and positioning confidence of the candidate box, and labels are assigned to the candidate box, avoiding the mechanical nature of traditional anchor frame matching. The irregular shapes and diverse materials (such as metal and plastic debris) of the foreign bodies in the test fixture are adapted to improve the accuracy of label assignment and provide signals for prediction. The classification confidence can be the confidence of the specific foreign body type such as metal debris or conductive debris, and the positioning confidence can be the confidence of the foreign body at the current coordinate position.

[0093] In operation S507 , a bounding box and a category are predicted.

[0094] The detection head outputs the bounding box coordinates (accurate to ±0.1mm) and category probabilities (e.g., 80% for metal debris, 30% for dust) of foreign objects on the test fixture. It also supports multi-task output (such as foreign object outlines during instance segmentation). The output bounding box coordinate parameters facilitate cleaning operations performed by maintenance equipment.

[0095] In operation S508 , redundant frames are filtered.

[0096] In some embodiments, when a feature is associated with one candidate frame, the position information of the foreign object is obtained based on the position information of the candidate frame; when a feature is associated with at least two candidate frames, the target frame is determined from the at least two candidate frames based on the confidence of the candidate frame, and the position information of the foreign object is obtained based on the position information of the target frame. For example, non-maximum suppression can be used to filter out repeated prediction frames of the same foreign object (such as false detection frames caused by probe shadows) and retain the prediction results with the highest confidence. Ensure the uniqueness and accuracy of the fixture detection results and avoid repeated actions of the maintenance device. The confidence here can be the probability that the feature is a foreign object, or it can be the result of the weighted summation of the classification confidence and positioning confidence of the above-mentioned candidate frames. It can be adaptively adjusted according to actual needs.

[0097] In operation S509 , the position and category of the foreign matter are output.

[0098] After operations S501 to S509, the precise position of the foreign object in the first image (such as X: 113.3mm, Y: 50.5mm, corresponding to the fixture coordinate system), category (such as conductive debris, non-conductive dust), and confidence level (such as the probability of the feature being a foreign object is 99.2%, or metal debris is 99.2%, etc.) can be output, providing executable detection results for the test fixture self-maintenance system, directly driving the maintenance operations of subsequent maintenance devices, forming a closed loop of detection and execution.

[0099] According to an embodiment of the present invention, by extracting features of different scales, ensuring that tiny foreign objects retain valid information in the feature map, and using an attention focusing mechanism to perform weighted detection on high-risk areas inside the test fixture where foreign objects exist, suppressing background interference, and performing dynamic label allocation, it is possible to adapt to irregular foreign object shapes and achieve a foreign object classification accuracy rate greater than 99%. On the other hand, when a feature is associated with multiple candidate frames, the frame with the highest confidence is selected as the target candidate frame to avoid positioning ambiguity caused by overlapping candidate frames, thereby improving the accuracy of determining the foreign object location information.

[0100] In some embodiments, based on the foreign matter information obtained from the above operations, the control device can control the maintenance device to perform maintenance. For example, a removal path for removing foreign matter can be generated based on the position information; based on the carrying state of the carrying component, the maintenance device can be controlled to clean and maintain the foreign matter according to the removal path.

[0101] For single-point foreign matter, such as debris, welding slag or particles, a removal path can be generated based on a point-to-point straight line path. For linear foreign matter such as scratches or fibers, a segmented removal path can be generated.

[0102] The maintenance device may include a piezoelectric vibrator and a magnetic control adsorption component. When the carrying state indicates that the carrying component is not carrying the board to be tested: a first control instruction may be sent to the piezoelectric vibrator to make the piezoelectric vibrator vibrate at a predetermined frequency to loosen foreign matter on the carrying component; a second control instruction may be sent to the magnetic control adsorption component to make the magnetic control adsorption component adsorb the loose foreign matter on the carrying component according to the removal path, and the caliber of the magnetic control suction nozzle of the magnetic control adsorption component is adjusted to match the size of the foreign matter. When the carrying state indicates that the carrying component is carrying the board to be tested: a third control instruction may be sent to the magnetic control adsorption component to make the magnetic control adsorption component adsorb the foreign matter on the board to be tested according to the removal path, and the caliber of the magnetic control suction nozzle of the magnetic control adsorption component is adjusted to match the size of the foreign matter.

[0103] The piezoelectric vibrator can be made of lead zirconate titanate piezoelectric ceramic material, which generates high-frequency mechanical vibration when an alternating voltage is applied. The vibration frequency is adjustable within 10kHz~50kHz, and the foreign matter on the surface of the test fixture, especially micron-sized particles, is separated from the adsorption interface through vibration energy. The vibration frequency is intelligently matched to different types of foreign matter. For foreign matter such as metal debris, a high-frequency vibration of 30kHz~50kHz can be used to destroy electrostatic adsorption by the resonance effect. For fiber impurities, a low-frequency vibration of 10kHz~20kHz can be used, combined with sound wave scattering to loosen foreign matter. Structurally, the structure of a cantilever beam plus a vibrating stylus can be used to clean foreign matter deep inside the connector of the board to be tested.

[0104] Compared with air blowing cleaning methods, piezoelectric vibrators can solve the problem of poor air blowing cleaning effect on tiny particles (less than 0.1mm). Compared with traditional brush cleaning methods, this non-contact vibration cleaning method can effectively avoid the risk of mechanical damage to the fixture probe due to direct contact.

[0105] The magnetic control nozzle can include a magnetic control nozzle. The magnetic control nozzle uses nickel-titanium shape memory alloy with a phase transition temperature of 50°C to 70°C to make the nozzle frame. Utilizing the shape memory effect of nickel-titanium shape memory alloy, the alloy undergoes a phase transition through electrical heating, allowing the nozzle diameter to expand from the initial 0.1mm~2mm to 5mm to accommodate foreign objects of different sizes. It can also return to its original shape after power failure. This design allows the nozzle to dynamically adjust its diameter according to the size of the foreign object, improving cleaning efficiency and adaptability. At the same time, it simplifies the mechanical structure and avoids the problem of frequent replacement of traditional nozzles with multiple specifications.

[0106] The magnetic nozzle is embedded with a neodymium iron boron permanent magnet (surface magnetic field strength ≥ 300mT), providing additional magnetic attraction for ferromagnetic foreign matter (such as iron filings), synergistically acting with the vacuum suction force. The variable diameter design of the magnetic nozzle covers foreign matter ranging from 0.1mm to 5mm. A single magnetic nozzle replaces traditional nozzles with multiple sizes, simplifying the mechanical structure. The combined magnetic and vacuum modes increase the capture rate of conductive foreign matter from 90% to 99.5%, thereby reducing the risk of short circuits in the test fixture.

[0107] The magnetically controlled nozzle in the above operation S409 and operation S421 can be adjusted within a range of 0.3 mm to 2 mm. The adjusted nozzle can perform negative pressure adsorption under the conditions of three different levels of filtration ducts.

[0108] The maintenance method of the present invention differs from existing cleaning methods in the following ways: Regarding microscopic foreign matter removal, existing cleaning methods rely on airflow, resulting in poor cleaning efficiency for microscopic foreign matter. The maintenance method of the present invention combines vibration and magnetic attraction, resulting in a more efficient cleaning of microscopic foreign matter. Regarding nozzle adaptability, existing cleaning methods use a fixed nozzle diameter, requiring frequent nozzle replacement. The maintenance method of the present invention offers an adjustable nozzle diameter, eliminating the need for nozzle replacement. Regarding ferromagnetic foreign matter removal, existing cleaning methods rely solely on vacuum suction, which can easily lead to residual particles. The maintenance method of the present invention combines magnetic attraction and vacuum, achieving a near 100% capture rate. Regarding fixture damage risk, existing cleaning methods rely on hard contact with brushes, which can easily damage the probe. The maintenance method of the present invention uses non-contact vibration, minimizing the risk of damaging the test board. Overall, the use of piezoelectric vibration and magnetic attraction for non-contact foreign matter removal, while automatically adjusting the vibration frequency and nozzle diameter based on the type of foreign matter, effectively improves cleaning effectiveness.

[0109] In some embodiments, when maintaining foreign matter, an image acquisition device is used to capture an image of the position indicated by the position information to obtain a second image; when the second image identifies that there is still foreign matter at the position, the maintenance device is continued to be controlled to clean the foreign matter again; when the second image does not identify that there is foreign matter at the position, the position information is used to update a foreign matter distribution map, which is used to describe the distribution of foreign matter in the test fixture.

[0110] After the maintenance device has absorbed the foreign object, it can activate the visual recognition module to capture a second image of the location indicated by the position information, generating a second image. If the foreign object is still detected in the second image, meaning it has not been completely removed, the piezoelectric vibrator and magnetic nozzle are activated again to perform a second cleaning maintenance. If no foreign object is detected in the second image, the edge server can record the foreign object's location information and update the foreign object distribution map.

[0111] According to the embodiment of the present invention, by repeatedly detecting and cleaning foreign matter, the thoroughness of foreign matter cleaning can be ensured, and the failure rate of the test fixture can be reduced.

[0112] In some embodiments, a predetermined cleaning times threshold may be set. For example, if a foreign object is still present at the location identified by the second image, the number of cleaning times for the foreign object is determined. If the number of cleaning times is less than the predetermined cleaning times threshold, the maintenance device is controlled to clean the foreign object again. If the number of cleaning times is greater than or equal to the predetermined cleaning times threshold, an alarm message may be sent to a target object. The target object may include an operation and maintenance personnel or an intelligent robot.

[0113] According to an embodiment of the present invention, by setting a predetermined cleaning times threshold, and when the cleaning times are greater than or equal to the predetermined cleaning times threshold, an alarm message can be sent to the target object, and the target object can handle the foreign matter, thereby avoiding the adverse effects caused by multiple cleanings of the same area.

[0114] In some embodiments, the initial operating data of the test fixture can also be converted into a target format to obtain target operating data, where the target format is a format supported by the maintenance prediction model; the target operating data is processed using the maintenance prediction model to obtain a pre-cleaning period of the test fixture, so as to control the maintenance device to perform maintenance on the test fixture during the pre-cleaning period.

[0115] In some embodiments, a maintenance prediction model (such as a long short-term memory network) can be activated to predict the next cleaning period. As the number of foreign body identifications increases, more and more data will be included in the foreign body distribution map. Through continuous training of these foreign body data, the prediction results of the maintenance prediction model will become more and more accurate.

[0116] Figure 6 A processing flow chart of a maintenance prediction model according to an embodiment of the present invention is shown.

[0117] like Figure 6 As shown, the processing flow of maintaining the prediction model may include operations S601 to S606.

[0118] In operation S601 , initial operation data during a test process of a test fixture is collected.

[0119] In operation S602 , the initial operation data is preprocessed to obtain target operation data.

[0120] In operation S603 , a maintenance prediction model is trained using the target operation data.

[0121] In operation S604 , real-time operation data during the actual operation process is collected in real time into a maintenance prediction model.

[0122] In operation S605 , prediction is performed by combining the historical operation data and the real-time operation data.

[0123] In operation S606 , the maintenance strategy is adjusted according to the predicted result.

[0124] In some embodiments, initial operating data from the test fixture during operation, i.e., historical operating data, can be collected. This data includes test time, number of tests, amount of metal debris and dust detected, fixture temperature and vibration, etc. This data then needs to be preprocessed to convert the initial operating data into a target format recognizable by the maintenance prediction model. This generates target operating data. After preprocessing, the target operating data can be input into the maintenance prediction model, which can then be used to train the maintenance prediction model. Once training is complete, the maintenance prediction model can be imported into the edge server.

[0125] During the operation of the test fixture, real-time operation data can be collected and input into the maintenance prediction model. In this way, the maintenance prediction model can combine historical operation data and real-time operation data to predict the time point or time period when the test fixture may need cleaning and maintenance in the future and output the prediction results. The maintenance system can adjust the maintenance strategy based on the prediction results. For example, the fixture can be automatically maintained in advance before the time period when cleaning is predicted to be required.

[0126] According to an embodiment of the present invention, by adopting a maintenance prediction model in combination with historical data and real-time operation data to predict the time period in which the test fixture will require maintenance in the future, increasing the detection frequency of key areas where foreign objects appear, and performing focused scanning, the maintenance efficiency of the test fixture can be improved and the failure rate of the test fixture can be reduced.

[0127] In some embodiments, a performance degradation curve of the test fixture can also be constructed based on test data during the test of the board in the server by the test fixture, such as the positioning error, surface roughness, operating time, temperature and humidity of the test fixture, etc. The pre-cleaning period of the next maintenance of the test fixture can be predicted based on the real-time operating data of the test fixture and the performance degradation curve, and the test fixture can be predicted based on the predicted pre-cleaning period.

[0128] According to an embodiment of the present invention, by predicting the pre-cleaning period in combination with the performance attenuation curve of the test fixture itself, the health threshold of the fixture can be accurately determined, avoiding the blindness of fixed-cycle maintenance and extending the service life of the test fixture.

[0129] Figure 7 A flowchart of a maintenance method according to another embodiment of the present invention is shown.

[0130] Figure 7 The maintenance method shown can be applied to the above maintenance system. Figure 7 The maintenance method shown may include operations S710 to S730.

[0131] In operation S710 , a first image of a test jig is recognized.

[0132] In operation S720 , when a foreign object is identified as being present in a component of the test jig from the first image, position information of the foreign object is determined.

[0133] In operation S730, the operation of the maintenance device is controlled according to the position information of the foreign matter and the carrying state of the carrying assembly, where the carrying state indicates whether the carrying assembly carries the board to be tested.

[0134] The maintenance method for test fixtures provided by the embodiment of the present invention has a high degree of automation, and the entire maintenance process is unmanned. The high degree of automation avoids the influence of human factors. The maintenance efficiency is high, and there is no need to plan fixture maintenance time separately. Fixture maintenance can be completed during normal testing, thereby improving production testing efficiency. The maintenance effect is good, and foreign objects that are small in size and difficult to identify manually can be automatically identified and effectively cleaned. The test and fixture maintenance strategies are improved. Relying on the prediction results of the foreign object distribution map and the maintenance prediction model, the production line test operation mode and fixture maintenance strategy can be adjusted and optimized to improve test efficiency and extend the service life of the test fixture.

[0135] In some embodiments, the above solution can be applied to any business scenario that requires the use of test fixtures, such as related products in the fields of automotive electronics, consumer electronics, new energy, etc.

[0136] Figure 8 The block diagram schematically shows an electronic device suitable for implementing a maintenance method for a test fixture according to an embodiment of the present invention.

[0137] like Figure 8 As shown, an electronic device 800 according to an embodiment of the present invention includes a processor 801, which can perform various appropriate actions and processes based on programs stored in a read-only memory (ROM) 802 or programs loaded from a storage unit 808 into a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or related chipsets and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0138] Various programs and data required for the operation of the electronic device 800 are stored in the RAM 803. The processor 801, ROM 802, and RAM 803 are connected to each other via a bus 804. The processor 801 executes the programs in the ROM 802 and / or RAM 803 to perform various operations according to the method flow of the embodiment of the present invention. It should be noted that the programs may also be stored in one or more memories other than the ROM 802 and RAM 803. The processor 801 may also execute the programs stored in the one or more memories to perform various operations according to the method flow of the embodiment of the present invention.

[0139] According to an embodiment of the present invention, electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to bus 804. Electronic device 800 may also include one or more of the following components connected to I / O interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including devices such as a cathode ray tube (CRT), liquid crystal display (LCD), and speakers; a storage section 808 including a hard disk; and a communication section 809 including a network interface card such as a LAN card or modem. Communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to I / O interface 805 as needed. Removable media 811, such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed in drive 810 as needed, so that computer programs read from the removable media can be installed into storage section 808 as needed.

[0140] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the present invention.

[0141] According to an embodiment of the present invention, a computer-readable storage medium may be a non-volatile computer-readable storage medium, and may include, for example, but not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present invention, a computer-readable storage medium may include the ROM 802 and / or RAM 803 described above, and / or one or more memories other than ROM 802 and RAM 803.

[0142] The embodiments of the present invention further include a computer program product, which includes a computer program containing program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to cause the computer system to implement the method provided by the embodiments of the present invention.

[0143] The computer program executes the above functions defined in the system / device of the embodiment of the present invention when executed by the processor 801. According to the embodiment of the present invention, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0144] In one embodiment, the computer program may be stored on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal on a network medium, downloaded and installed via the communication portion 809, and / or installed from a removable medium 811. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0145] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 809 and / or installed from a removable medium 811. When the computer program is executed by the processor 801, the above-described functions defined in the system of the embodiment of the present invention are performed. According to the embodiment of the present invention, the systems, devices, means, modules, units, etc. described above can be implemented by computer program modules.

[0146] According to an embodiment of the present invention, the program code for executing the computer program provided by the embodiment of the present invention can be written in any combination of one or more programming languages. Specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C" or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect via the Internet).

[0147] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0148] It will be understood by those skilled in the art that the features described in the various embodiments of the present invention may be combined and / or coupled in various ways, even if such combinations or couplings are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention may be combined and / or coupled in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or couplings fall within the scope of the present invention.

[0149] The above describes embodiments of the present invention. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. Although each embodiment has been described separately above, this does not mean that the measures in each embodiment cannot be advantageously used in combination. Without departing from the scope of the present invention, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

Claims

1. A test fixture maintenance system, characterized in that: The system comprises: a control device for identifying a first image of a test jig and, if a foreign object is identified in the first image as being present in a component of the test jig, determining position information of the foreign object, wherein the first image of the test jig is obtained by image acquisition equipment performing image acquisition on a component of the test jig, the component comprising a movable test plate and a supporting assembly for supporting the test plate, the test plate being used to implement testing of an object under test by the test jig; and controlling operation of a maintenance device based on the position information of the foreign object and a supporting state of the supporting assembly, wherein the supporting state indicates whether the supporting assembly is supporting the test plate; The maintenance device is used to perform maintenance operations on the test fixture under the control of the control device.

2. The maintenance system according to claim 1, characterized in that: The control device is also used for: generating a removal path for removing the foreign matter according to the position information; According to the carrying state of the carrying assembly, the maintenance device is controlled to clean and maintain the foreign matter according to the removal path.

3. The maintenance system according to claim 2, characterized in that: The maintenance device includes a piezoelectric vibrator and a magnetically controlled adsorption component; The control device is further configured to: when the carrying state indicates that the carrying assembly is not carrying the board to be tested: sending a first control instruction to the piezoelectric vibrator to cause the piezoelectric vibrator to vibrate at a predetermined frequency to loosen foreign matter on the bearing assembly; A second control instruction is sent to the magnetic adsorption component, so that the magnetic adsorption component adsorbs loose foreign matter on the carrier component according to the removal path, and the diameter of the magnetic adsorption nozzle of the magnetic adsorption component is adjusted to match the size of the foreign matter.

4. The maintenance system according to claim 3, characterized in that: The control device is further configured to: when the carrying state indicates that the carrying assembly is carrying the board to be tested: A third control instruction is sent to the magnetic adsorption component, so that the magnetic adsorption component adsorbs the foreign matter on the test board according to the removal path, and the diameter of the magnetic adsorption nozzle of the magnetic adsorption component is adjusted to match the size of the foreign matter.

5. The maintenance system according to claim 1, characterized in that: The control device is also used for: extracting multi-scale features of the first image; Based on the weights of the components, weighting the features for the components to obtain a weighted feature map; Determining a plurality of candidate boxes including the features in the weighted feature map; When the confidence of the candidate frame is greater than a predetermined confidence, it is determined that a foreign object exists in the component of the test fixture identified from the first image, and the confidence is used to describe the probability that the feature in the candidate frame is the foreign object.

6. The maintenance system according to claim 5, characterized in that: The control device is also used for: When the feature is related to the candidate frame, obtaining the position information of the foreign object according to the position information of the candidate frame; In the case where the feature is related to at least two candidate frames, a target frame is determined from the at least two candidate frames according to the confidence of the candidate frames, and the position information of the foreign object is obtained according to the position information of the target frame.

7. The maintenance system according to claim 1, characterized in that: The control device is also used for: When maintaining the foreign object, using the image acquisition device to acquire an image of the position indicated by the position information to obtain a second image; If the foreign object is still present at the position identified from the second image, continue to control the maintenance device to clean the foreign object again; In the case that the foreign matter is not identified as existing at the position from the second image, a foreign matter distribution map is updated using the position information, where the foreign matter distribution map is used to describe the distribution of the foreign matter in the test fixture.

8. The maintenance system according to claim 1, wherein: The control device is also used for: Converting the initial operating data of the test fixture into a target format to obtain target operating data, wherein the target format is a format supported by the maintenance prediction model; The target operation data is processed using the maintenance prediction model to obtain a pre-cleaning period of the test fixture, so as to control the maintenance device to perform maintenance on the test fixture during the pre-cleaning period.

9. A test fixture maintenance method, characterized in that: The method is applied to the maintenance system according to any one of claims 1 to 8, and the method includes: recognizing the first image of the test fixture; In a case where a foreign object is identified in the component of the test jig from the first image, determining position information of the foreign object; The operation of the maintenance device is controlled according to the position information of the foreign matter and the carrying state of the carrying assembly, wherein the carrying state indicates whether the carrying assembly carries the board to be tested.

10. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instructions are executed by a processor, the steps of the method according to claim 9 are implemented.

Images