Test fixture maintenance system, method, and storage medium

Abstract

The application provides a test fixture maintenance system, method and storage medium, which can be applied to the technical field of test fixture maintenance. The test fixture maintenance system comprises: a control device configured to identify a first image of a test fixture, and determine position information of foreign matter in a case where it is identified from the first image that the test fixture has the foreign matter, the first image of the test fixture being obtained by image acquisition equipment collecting an image of a component of the test fixture, the component comprising a movable test board and a bearing component for bearing the test board, the test board being configured to implement test of a test object by the test fixture; and controlling operation of a maintenance device according to the position information of the foreign matter and a bearing state of the bearing component, the bearing state indicating whether the bearing component bears the test board; and the maintenance device is configured to execute a maintenance operation for the test fixture under control of the control device.

Description

TECHNICAL FIELD

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

[0002] With the rapid development of cloud computing, big data and artificial intelligence technology, the construction scale of global data centers continues to expand, and the requirement for server hardware performance is exponentially increasing. As a core component of the server, the stability and functionality of the control board card directly determine the overall performance and reliability of the server. The control board card in each server needs to undergo various tests and verifications before leaving the factory. During the testing and verification of the control board card, test fixtures are used, and the test fixtures also need to be regularly maintained during use. The current maintenance process of the test fixture has low intelligentization degree, incomplete cleaning effect and other problems, which not only reduces the maintenance efficiency of the test fixture, but also increases the failure rate of the test fixture. SUMMARY

[0003] In view of the above problems, the present application provides a test fixture maintenance system, method and storage medium for improving the maintenance efficiency of the test fixture and reducing the failure rate of the test fixture.

[0004] One aspect of the present application provides a test fixture maintenance system, which comprises: a control device configured to identify a first image of a test fixture, and determine position information of foreign matter in the case that the foreign matter exists in a component of the test fixture identified from the first image, the first image of the test fixture being obtained by image acquisition equipment performing image acquisition on the component of the test fixture, the component comprising a movable test board and a carrying component for carrying the test board, the test board being configured to realize testing of a measured object by the test fixture; and control operation of a maintenance device according to the position information of the foreign matter and a carrying state of the carrying component, the carrying state indicating whether the carrying component carries the test board; and the maintenance device configured to perform maintenance operation on the test fixture under control of the control device.

[0005] Another aspect of the present application also provides a test fixture maintenance method, which comprises: identifying a first image of a test fixture; determining position information of foreign matter in the case that the foreign matter exists in a component of the test fixture identified from the first image; and controlling operation of a maintenance device according to the position information of the foreign matter and a carrying state of the carrying component, the carrying state indicating whether the carrying component carries the test board.

[0006] Another aspect of the present application also provides a computer-readable storage medium having a computer program or instructions stored thereon, the computer program or instructions being executed by a processor to implement the steps of the above test fixture maintenance method.

[0007] According to the embodiment of the present application, the test fixture is maintained by setting a control device and a maintenance device, the control device is used to identify the first image of the test fixture, and in the case of identifying the foreign matter, the position information of the foreign matter is determined, and the maintenance device is controlled to perform the maintenance operation on the test fixture according to the position information of the foreign matter and the bearing state of the bearing assembly. Since in the maintenance process, from the scanning of the test fixture, the foreign matter identification, the foreign matter position output to the double-mode foreign matter cleaning according to the position and the bearing state, the whole maintenance process can be automatically executed, and by identifying the foreign matter by the machine equipment, compared with the manual naked eye identification of the foreign matter, the accuracy of the foreign matter can be improved, the thorough cleaning of the foreign matter is realized, and the technical effects of improving the maintenance efficiency of the test fixture and reducing the failure efficiency of the test fixture are achieved. BRIEF DESCRIPTION OF DRAWINGS

[0008] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application taken in conjunction with the accompanying drawings, in which:

[0009] Figure 1 A scene diagram of application of the maintenance system of the test fixture according to the embodiment of the present application is shown;

[0010] Figure 2 An architecture diagram of the maintenance system of the test fixture according to the embodiment of the present application is shown;

[0011] Figure 3 An architecture diagram of the maintenance system of the test fixture according to another embodiment of the present application is shown;

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

[0013] Figure 5 A flowchart of the foreign matter identification module identifying the foreign matter and outputting the foreign matter position coordinates according to the embodiment of the present application is shown;

[0014] Figure 6 A processing flowchart of the maintenance prediction model according to the embodiment of the present application is shown;

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

[0016] Figure 8 A block diagram of the electronic device suitable for implementing the maintenance method of the test fixture according to the embodiment of the present application is shown schematically. DETAILED DESCRIPTION

[0017] Embodiments of the present application will be described below with reference to the accompanying drawings. It should be understood, however, that the description which follows is illustrative only and is not intended to limit the scope of the present application. In the detailed description, procedures, apparatuses, and methods that are well known and commonly used in the art will not be described in detail in order to avoid obscuring the concept of the present application.

[0018] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used herein, the term "includes" and tautological equivalents thereof, means the inclusion of the stated features, steps, operations, and / or components but not to the exclusion of one or more other features, steps, operations, or components.

[0019] All terms used herein including technical and scientific terms have the same meanings as commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification, and not be interpreted in an idealized or overly formal way.

[0020] In the case where expressions similar to "at least one of A, B, and C, etc." are used, it should generally be interpreted to include any of the combinations of the items enumerated in the list (e.g., "a system having at least one of A, B, and C" should include but not be limited to a system having A alone, a system having B alone, a system having C alone, a system having both A and B, a system having both A and C, a system having both B and C, and / or a system having A, B, and C, etc.).

[0021] With the continuous expansion of the global data center construction scale, the requirements for server hardware performance are increasing exponentially. As the core components of servers, the stability and functionality of control board cards such as motherboards, storage control cards, and graphics processing unit (GPU) board cards directly determine the overall performance and reliability of the server. The control board cards in each server need to undergo various tests before leaving the factory, including functional test (FCT) and in-circuit test (ICT), boundary scan test (BSI), and other tests. FCT refers to providing a simulated operating environment for the test target board card, making the test target board card work in various design states, so as to obtain the parameters of each state to verify the test method of the test target board card function. ICT is a testing technology that does not open the circuit and does not remove the pins of the components. BSI is a board-level circuit testing technology that inserts a boundary scan unit between the chip pins and internal logic to achieve non-intrusive testing of the soldering quality, signal connectivity, and logic function of integrated circuits on the computer board card.

[0022] During ICT, BSI, and FCT testing, in order to improve testing efficiency and ensure testing quality, customized development of corresponding test fixtures is required for different test items. The test fixture needs to be maintained regularly during use, and the appearance of the fixture module needs to be checked to ensure that the test fixture is used normally. However, during the maintenance of the test fixture, there are problems such as high dependence on manual operation, incomplete troubleshooting, low maintenance efficiency, interference with the testing process, uncontrollable cleaning effect, and untimely maintenance.

[0023] The low maintenance and cleaning efficiency mainly reflects that manual maintenance requires operators to hold a brush or cotton cloth to clean debris (such as tin slag and plastic residue) with cleaning alcohol. In a printed circuit board assembly factory, the number of test fixtures for normal testing on the production line is generally between 300 and 500. If all of them rely on manual inspection and cleaning by operators, the maintenance efficiency is very low, which affects the normal use of the test fixture.

[0024] The interference with the testing process mainly reflects that when manual maintenance of the test fixture is adopted, it needs to be done during shift change or when the fixture is idle, otherwise the testing needs to be stopped for maintenance of the test fixture. If all modules of the test fixture are checked one by one, it will take a long time and require special organization of manpower, which will affect the normal testing process of the production line.

[0025] The cleaning effect is uncontrollable, mainly reflected in that it is difficult to conduct comprehensive inspection and thorough removal of small foreign matters, such as debris smaller than 0.1 mm3, by means of artificial maintenance, and the residual foreign matters are likely to cause misjudgment during the testing of the board card, thereby affecting the one-time pass rate and testing efficiency of the board card testing.

[0026] Although the semi-automatic cleaning device is currently used in the industry to maintain and clean the test fixture, the brush head and sliding rail combination is used, and this method still has the problems of incomplete troubleshooting, untimely maintenance and single function. The incomplete troubleshooting and untimely maintenance mainly reflect that the semi-automatic maintenance method needs to first find the dirt in the test fixture by artificial inspection, and then clean it by the brush or dust collector in the cleaning device. If the artificial inspection does not find the dirt, the maintenance and cleaning cannot be performed. The single function mainly reflects that the artificial inspection method cannot identify and clean small foreign matters, and the maintenance process will have the phenomenon of omission, and the test fixture cannot be maintained comprehensively.

[0027] Therefore, the embodiments of the present application provide a test fixture maintenance system, which collects the image of the test fixture by using a multispectral camera, identifies foreign matters in the image, loosens the foreign matters by using a piezoelectric vibrator, and adsorbs the foreign matters by using a magnetic control suction nozzle. Meanwhile, the maintenance prediction model is combined to predict the time period when the test fixture needs to be maintained in the future. The maintenance system can automatically adjust the fixture maintenance strategy according to the prediction result, so as to realize the self-maintenance of the fixture, improve the fixture maintenance efficiency, reduce the testing cost, and reduce the failure rate of the test fixture.

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

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

[0030] The control device 1021 can scan the to-be-tested board 1011 and the bearing assembly 1012 of the test fixture 101 to obtain a first image, and in a case where it is identified from the first image that there is a foreign matter on the assembly of the test fixture, determine position information of the foreign matter, and control operation of the maintenance device 1022 according to the position information of the foreign matter and a bearing state of the bearing assembly. The maintenance device 1022 is configured to perform a maintenance operation on the test fixture 101 under control of the control device 1021.

[0031] It should be understood that Figure 1 The number of the to-be-tested board 1011, the fixture middle frame 1012-1, the to-be-tested board carrier 1012-2, the control device 1021 and the maintenance device 1022 in the embodiment is only illustrative. Any number of the to-be-tested board 1011, the fixture middle frame 1012-1, the to-be-tested board carrier 1012-2, the control device 1021 and the maintenance device 1022 can be provided according to actual needs.

[0032] The following will describe the maintenance system of the test fixture of the embodiment of the application based on Figure 1 the scenario described in the embodiment. Figures 2-6 The maintenance system of the test fixture of the embodiment of the application will be described in detail.

[0033] Figure 2 An architectural diagram of the maintenance system of the test fixture according to the embodiment of the application is shown.

[0034] As shown in Figure 2 , the maintenance system 102 of the test fixture of the 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 fixture, and in a case where it is identified from the first image that there is a foreign matter on the assembly of the test fixture, determine position information of the foreign matter, the first image of the test fixture being obtained by image acquisition equipment performing image acquisition on the assembly of the test fixture, the assembly including a movable to-be-tested board and a bearing assembly for bearing the to-be-tested board, the to-be-tested board being configured to implement testing of a to-be-tested object by the test fixture; and control operation of the maintenance device 1022 according to the position information of the foreign matter and a bearing state of the bearing assembly, the bearing state indicating whether the bearing assembly bears the to-be-tested board.

[0036] The maintenance device 1022 is configured to perform a maintenance operation on the test fixture under control of the control device 1021.

[0037] Figure 3 An architectural diagram of the maintenance system of the test fixture according to another embodiment of the application is shown.

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

[0039] The edge server 1021-1 can control the image acquisition device of the visual identification module 1021-3 to acquire images of the components of the test fixture to obtain first images. The components of the test fixture include a movable test board and a bearing component that bears the test board, and the bearing component includes a fixture middle frame and a test board carrier plate, and the test board is used to realize the test of the test fixture on the measured object.

[0040] The image acquisition device of the visual identification module 1021-3 can include a multispectral camera. The multispectral camera can use visible light with a wavelength of 400nm~700nm and short-wave infrared light with a wavelength of 900nm~1700nm to detect foreign matter, the visible light can be used to detect organic residues, and the short-wave infrared light can detect metal debris, thereby realizing the cooperative work of double wave bands. Compared with a traditional Red-Green-Blue Camera (RGB camera), the multispectral camera can more accurately detect foreign matter, and has advantages such as high precision and less limitation by light environment. The main differences between the multispectral camera and the RGB camera are shown in Table 1.

[0041] Table 1

[0042]

[0043] In the case that the multispectral camera obtains the first images, the edge server 1021-1 can control the foreign matter identification and positioning module 1021-2 to identify the first images. In the case that no foreign matter is identified from the first images, the foreign matter indicator light of the test fixture can present green, and in the case that foreign matter is identified, the three-dimensional coordinates of the foreign matter can be generated according to the foreign matter to obtain the position information of the foreign matter, and the foreign matter removal path can be automatically calculated according to the position information of the foreign matter, and the maintenance device can be controlled to maintain the foreign matter according to the bearing state of the bearing component.

[0044] The maintenance device can include a piezoelectric vibrator and a magnetic control suction assembly. The carrying state can include that the carrying assembly carries the to-be-tested board and that the carrying assembly does not carry the to-be-tested board. In the case that the carrying assembly carries the to-be-tested board, only the magnetic control suction assembly can be controlled to suck the foreign matter, so as to avoid loosening of the electronic component welding point on the to-be-tested board due to vibration of the piezoelectric vibrator, and further avoid damage of the piezoelectric vibrator to the to-be-tested board. In the case that the carrying assembly does not carry the to-be-tested board, the piezoelectric vibrator can be controlled to vibrate, the foreign matter is loosened, and the magnetic control suction assembly is controlled to suck the loosened foreign matter, so as to realize cleaning and maintenance of the foreign matter. The magnetic control suction assembly can include a magnetic control suction nozzle, and a caliber of the magnetic control suction nozzle can be matched with a size of the foreign matter.

[0045] According to the embodiment of the present application, the test fixture is maintained by setting the control device and the maintenance device. The control device is used to identify the first image of the test fixture, and in the case that the foreign matter is identified, the position information of the foreign matter is determined, and the maintenance device is controlled to perform the maintenance operation on the test fixture according to the position information of the foreign matter and the carrying state of the carrying assembly. Since the whole maintenance process can be automatically executed from scanning of the test fixture, identification of the foreign matter, output of the position of the foreign matter, to double-mode cleaning of the foreign matter according to the position and the carrying state in the maintenance process, and the foreign matter is identified by the machine equipment, compared with manual naked-eye identification of the foreign matter, the accuracy of the foreign matter can be improved, complete cleaning of the foreign matter is realized, and the technical effects of improving the maintenance efficiency of the test fixture and reducing the failure efficiency of the test fixture are achieved.

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

[0047] Based on the above-mentioned maintenance system of the test fixture, an embodiment of the present application provides a maintenance method for the above-mentioned test fixture, and a flow chart of the maintenance method is shown in Figure 4

[0048] Figure 4 The maintenance method of the test fixture shown can include operations S401-S426. Figure 4 ​The maintenance process of the test fixture shown is divided into self-maintenance before the test board is placed and self-maintenance after the test board is placed, starting from when the test fixture completes testing. The self-maintenance before the test board is placed includes operations S401-S413, which means that the test board carrier plate and the fixture middle frame of the test fixture are maintained in advance before the test board is placed in the test fixture, to ensure that there is no foreign matter inside the test fixture. The self-maintenance after the test board is placed includes operations S414-S426, which mainly detects foreign matter on the test board itself to ensure that there is no foreign matter on the test board itself, so as to avoid affecting the test results. By switching the cleaning process from vibration and adsorption before the test board is loaded to adsorption only after the test board is loaded, a dual-mode maintenance process can be realized, improving the diversity of test fixture maintenance.

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

[0050] In operation S402, the test board is removed.

[0051] After the test fixture completes the test, it is opened and the test board is removed by an operator or a mechanical arm. After the test fixture detects that the test board is removed, the visual recognition module starts scanning, and the multispectral camera collects images of the fixture middle frame and the test board carrier plate, to identify whether foreign matter is generated in the previous test process.

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

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

[0054] In operation S405, whether foreign matter is identified. The foreign matter identification and positioning module executes operation S406 if foreign matter is identified, and executes operation S314 if no foreign matter is identified.

[0055] In operation S406, a three-dimensional coordinate of the foreign matter is generated to determine the position information of the foreign matter.

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

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

[0058] In operation S409, the magnetic control suction nozzle adjusts the caliber.

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

[0060] In operation S411, whether the foreign matter is scanned again. The second image obtained by scanning the components of the test fixture again through the visual recognition module can be determined. In the case of scanning the foreign matter again, operation S308 can be performed, and in the case of not scanning the foreign matter, operation S412 can be performed.

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

[0062] In operation S413, the maintenance prediction model predicts the pre-maintenance period of the next maintenance of the test fixture, and operation S414 is performed.

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

[0064] In operation S415, the board under test is placed.

[0065] In operation S416, the visual recognition module is started to scan.

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

[0067] In operation S418, whether the foreign matter is identified. The foreign matter identification and positioning module performs operation S419 in the case of identifying the foreign matter, and performs operation S426 in the case of not identifying the foreign matter.

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

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

[0070] In operation S421, the magnetic control suction nozzle adjusts the caliber.

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

[0072] In operation S423, whether the foreign matter is scanned again. The scanning can be determined by starting the visual recognition module again. In the case of scanning the foreign matter again, operation S421 can be performed, and in the case of not scanning the foreign matter, operation S424 can be performed.

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

[0074] In operation S425, the maintenance prediction model predicts the pre-maintenance period of the next maintenance of the test fixture.

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

[0076] In some embodiments, after the image acquisition is completed, the foreign object recognition module uses a recognition algorithm (such as a You Only Look Once (YOLO) algorithm) to recognize the foreign object, and outputs the position coordinates of the foreign object if the foreign object is recognized. If no foreign object is recognized, the jig foreign object indicator light turns green, indicating that the test board can be placed in the jig, and then the operator can place the test board in the jig for testing.

[0077] In some embodiments, the operation S405 can be determined by extracting the multi-scale features of the first image, weighting the features for the components based on the weights of the components to obtain a weighted feature map, determining a plurality of candidate boxes including the features in the weighted feature map, and determining that the component of the test jig is recognized to have a foreign object from the first image if the confidence of the candidate box is greater than a predetermined confidence, the confidence being used to describe the probability that the features in the candidate box are foreign objects.

[0078] The operation S406 can be determined by obtaining the position information of the foreign object according to the position information of the candidate box if the feature is related to the candidate box, and determining a target box from the at least two candidate boxes according to the confidence of the candidate box and obtaining the position information of the foreign object according to the position information of the target box if the feature is related to at least two candidate boxes.

[0079] Figure 5 A flowchart of the foreign object recognition module recognizing the foreign object and outputting the position coordinates of the foreign object according to an embodiment of the present application is shown.

[0080] As shown in Figure 5 The flow of the foreign object recognition module recognizing the foreign object and outputting the position coordinates of the foreign object can include operations S501-S509, and the operation S405 can also be determined by operations S501-S505, and the operation S406 can also be determined by operations S506-S509.

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

[0082] In operation S502, the first image is preprocessed.

[0083] For the detection components of the test jig, the preprocessing step can enhance the feature retention of small foreign objects (such as 0.1 mm debris) to distinguish from general image preprocessing, and lay a 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 module (Cross Stage Partial Fusion, C2f) is adopted to extract multi-scale features of the first image by cross-stage feature splitting and fusion. The scale can be determined according to the probability that the feature is a foreign object. For example, when the probability that the feature is a foreign object is greater than a first predetermined value, it is a key foreign object; and when the probability that the feature is a foreign object is less than or equal to the first predetermined value, it is a secondary key foreign object. In view of the characteristics of small targets and multiple materials detected by the test fixture, the feature extraction path is optimized to solve the problem of feature loss of small foreign objects in the traditional network under the complex environment of the test fixture.

[0086] In some embodiments, the small target can refer to the foreign object or defect (such as a tiny scratch, welding slag) on the test fixture, which can only occupy a few pixels of the image. When down-sampling in the traditional network, these details are easily lost. Multiple materials can refer to the test fixture, which can contain multiple materials such as metal, plastic, and silica gel. The reflectivity and texture difference of these materials leads to insufficient generalization ability of the traditional feature extractor. The complex environment can refer to the uneven light and oil stain interference in the industrial scene, which further increases the difficulty of feature separation. By introducing the C2f module, multi-scale features can be extracted to prevent small target information from being diluted in the deep network.

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

[0088] By combining spatial pyramid pooling and cross-stage convolution, the multi-scale features (such as local features of 0.1 mm debris and global features of the overall structure of the board) of the test fixture image are fused to generate a feature map containing details and semantics. The details can be local features with fine granularity and high resolution in the image, such as the clear outline of scratches (only 1-2 pixels wide), welding burrs, and tiny foreign objects (dust / fibers), and the semantics can be the semantics of scratches (linear features) and stains (block features), etc. Thus, the coexistence of tiny foreign objects and complex backgrounds in test fixture detection is adapted, and the small target features of the foreign objects are ensured not to be overwhelmed by large structure 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 weight the features of high-risk components or regions (such as probe gaps and slots) of foreign objects in the first image, and to suppress background noise such as board textures and normal elements, etc. In this way, the attention focus during manual detection is simulated, and the sensitivity of foreign object detection in the key area of the fixture is improved. In some embodiments, a plurality of candidate boxes including features in the weighted feature map can be determined, and in the case that the confidence of the candidate box is greater than a predetermined confidence, it is determined that there is a foreign object in the test fixture. The confidence can be the probability that the feature in the candidate box is a foreign object.

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

[0092] In another embodiment, based on the multi-scale features of the feature pyramid, a task-aligned assigner can be used to dynamically assign labels. For each candidate box in the weighted feature map, according to the classification confidence and the location confidence of the candidate box, a real foreign object box (such as a conductive debris or an insulating impurity) is dynamically matched, and a label is configured for the candidate box to avoid the mechanical matching of the traditional anchor box. The irregular shape and the various materials (such as metal and plastic debris) of the foreign object of the test fixture are adapted, and the label assignment accuracy is improved to provide a signal for prediction. The classification confidence can be the reliability of the foreign object being a specific foreign object type such as a metal debris or a conductive debris, and the location confidence can be the reliability of the foreign object at the current coordinate position.

[0093] In operation S507, the bounding box and the class are predicted.

[0094] The detection head outputs the bounding box coordinates (accurate to ±0.1 mm) and the class probability (such as 80% metal debris and 30% dust) of the foreign object of the test fixture, while supporting multi-task output (such as foreign object contour in instance segmentation). The output bounding box coordinate parameters can facilitate the maintenance device to perform a cleaning action.

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

[0096] In some embodiments, in the case where the feature is related to one candidate box, the location information of the foreign object is obtained according to the location information of the candidate box; in the case where the feature is related to at least two candidate boxes, the target box is determined from the at least two candidate boxes according to the confidence of the candidate box, and the location information of the foreign object is obtained according to the location information of the target box. For example, non-maximum suppression can be used to filter repeated prediction boxes of the same foreign object (such as false detection boxes caused by probe shadows), and the prediction result with the highest confidence is retained. The uniqueness and accuracy of the test fixture detection result are ensured, and repeated actions of the maintenance device are avoided. The confidence here can be the probability that the feature is a foreign object, or the weighted sum of the classification confidence and the location confidence of the candidate box, which can be adaptively adjusted according to actual needs.

[0097] In operation S509, the foreign object position and the class are outputted.

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

[0099] According to the embodiments of the present application, by extracting features of different scales, it is ensured that the micro foreign objects retain effective information in the feature map, the high-risk areas where foreign objects exist inside the test jig are weighted and detected by the attention focusing mechanism, background interference is suppressed, and dynamic label assignment is performed, which can adapt to irregular foreign object shapes and achieve an effect of a foreign object classification accuracy greater than 99%. On the other hand, in the case where the feature is related to multiple candidate boxes, the box with the highest confidence is selected as the target candidate box, avoiding positioning ambiguity caused by overlapping candidate boxes, and further improving the determination accuracy of the foreign object position information.

[0100] In some embodiments, according to the foreign object information obtained by the above operation, the control device can control the maintenance device to perform maintenance, for example, a removal path for removing the foreign object can be generated according to the position information; according to the bearing state of the bearing assembly, the maintenance device is controlled to clean and maintain the foreign object according to the removal path.

[0101] For single-point foreign objects such as debris, welding slag or particles, a removal path for removing the foreign object can be generated according to a straight-line path between points, and for linear foreign objects such as scratches or fibers, a segmented removal path can be generated.

[0102] The maintenance device can include a piezoelectric vibrator and a magnetic control adsorption assembly. In the case where the bearing state indicates that the bearing assembly is not bearing the test board: a first control instruction can be sent to the piezoelectric vibrator to make the piezoelectric vibrator vibrate at a predetermined frequency to loosen the foreign objects on the bearing assembly; a second control instruction is sent to the magnetic control adsorption assembly to make the magnetic control adsorption assembly adsorb the loosened foreign objects on the bearing assembly according to the removal path, and the caliber of the magnetic control suction nozzle of the magnetic control adsorption assembly is adjusted to be compatible with the size of the foreign object. In the case where the bearing state indicates that the bearing assembly is bearing the test board: a third control instruction can be sent to the magnetic control adsorption assembly to make the magnetic control adsorption assembly adsorb the foreign objects on the test board according to the removal path, and the caliber of the magnetic control suction nozzle of the magnetic control adsorption assembly is adjusted to be compatible with the size of the foreign object.

[0103] The piezoelectric vibrator can adopt a lead zirconate titanate piezoelectric ceramic material, generate high-frequency mechanical vibration when an alternating voltage is applied, and the vibration frequency can be adjusted within 10 kHz-50 kHz. The test fixture surface is separated from the adsorbed interface by vibration energy, especially for micrometer-level particles. For different types of foreign matter, the vibration frequency is intelligently matched. For metal debris type foreign matter, high-frequency vibration of 30 kHz-50 kHz can be used to destroy electrostatic adsorption by resonance effect. For fiber impurities, low-frequency vibration of 10 kHz-20 kHz can be used to loosen foreign matter by combining acoustic scattering. In terms of structure, the cantilever beam and vibrating stylus structure can be used to clean the connector inside the to-be-tested board.

[0104] Compared with the blowing cleaning method, the piezoelectric vibrator can solve the problem of poor effect of blowing cleaning on small particles (<0.1 mm). Compared with the traditional brush cleaning method, this non-contact vibration cleaning method can effectively avoid the risk of mechanical damage to the probe of the test fixture.

[0105] The magnetic control nozzle can include a magnetic control nozzle. The magnetic control nozzle is made of nickel-titanium memory alloy with a phase transition temperature of 50-70°C. The shape memory effect of the nickel-titanium memory alloy is used to expand the nozzle caliber from the initial 0.1-2 mm to 5 mm by heating and phase transition, so as to adapt to foreign matter of different sizes. After power-off, it can also restore to the original state. This design enables the nozzle to dynamically adjust the caliber according to the size of the foreign matter, improves the cleaning efficiency and adaptability, simplifies the mechanical structure, and avoids the problem of frequent replacement of traditional multi-specification nozzles.

[0106] The magnetic control nozzle can be embedded with a neodymium-iron-boron permanent magnet (surface magnetic field strength ≥300 mT), which can form an additional magnetic force to adsorb ferromagnetic foreign matter (such as iron filings) in cooperation with the vacuum suction force. The magnetic control nozzle variable diameter design covers a range of 0.1-5 mm of foreign matter. Single magnetic control nozzle replaces traditional multi-specification nozzles, simplifies the mechanical structure, and the adsorption of the magnetic suction and vacuum composite mode can improve the capture rate of conductive foreign matter from 90% to 99.5%, thereby reducing the short circuit risk of the test fixture.

[0107] The magnetic control nozzle in the above operation S409 and operation S421 can be adjusted within 0.3-2 mm. The adjusted nozzle can perform negative pressure adsorption under the condition of three different levels of filter air ducts.

[0108] Compared with the current cleaning method, the maintenance method of the embodiment of the application can have the following differences: in the dimension of micro-foreign matter removal rate, the current cleaning method relies on airflow blowing, and the cleaning effect on micro-foreign matter is poor, the maintenance method of the embodiment of the application combines vibration and magnetic attraction, and the cleaning efficiency on micro-foreign matter is higher. In the dimension of suction nozzle adaptation range, the current cleaning method is fixed in caliber, and the magnetic control suction nozzle needs to be frequently replaced, the caliber of the magnetic control suction nozzle of the maintenance method of the embodiment of the application is adjustable, and the suction nozzle can be replaced. In the dimension of ferromagnetic foreign matter processing, the current cleaning method only relies on vacuum suction, and residues are easy to be left, the maintenance method of the embodiment of the application combines magnetic attraction and vacuum, and the capture rate of foreign matter is close to 100%. In the dimension of jig damage risk, the current cleaning method cleans by hard contact with a brush, and probes are easy to be damaged, the maintenance method of the embodiment of the application is non-contact vibration, and the probability of damaging the test board is smaller. Overall, the piezoelectric vibration and magnetic control adsorption method is used to clean foreign matter non-contact, and the vibration frequency and the caliber of the suction nozzle are automatically adjusted according to the type of foreign matter, so that the cleaning effect is effectively improved.

[0109] In some embodiments, in the case of maintaining the foreign matter, the image acquisition device is used to acquire an image of the position indicated by the position information, to obtain a second image; in the case of identifying that the position still has foreign matter from the second image, the maintenance device is continued to be controlled to clean the foreign matter again; in the case of not identifying that the position has foreign matter from the second image, the position information is used to update the foreign matter distribution map, and the foreign matter distribution map is used to describe the distribution of the foreign matter on the test jig.

[0110] After the maintenance device completes the adsorption of the foreign matter, the vision recognition module can be started to acquire a second image of the position indicated by the position information. If the foreign matter is still identified in the second image, that is, the foreign matter is not completely removed, the piezoelectric vibrator and the magnetic control suction nozzle are continued to be started to clean the foreign matter again. If the foreign matter is not identified in the second image, the edge server can record the position information of the foreign matter and update the foreign matter distribution map.

[0111] According to the embodiment of the application, through multiple detection and cleaning of the foreign matter, the thoroughness of the foreign matter cleaning can be ensured, and the failure rate of the test jig is reduced.

[0112] In some embodiments, a predetermined cleaning frequency threshold can be set, for example, the cleaning frequency of the foreign matter is determined in the case of identifying that the position still has foreign matter from the second image; in the case of the cleaning frequency being less than the predetermined cleaning frequency threshold, the maintenance device is continued to be controlled to clean the foreign matter again, and in the case of the cleaning frequency being greater than or equal to the predetermined cleaning frequency threshold, the target object can be sent alarm information. The target object can include an operation and maintenance personnel or an intelligent robot.

[0113] According to the embodiment of the present application, by setting a predetermined cleaning frequency threshold, and in the case that the cleaning frequency is greater than or equal to the predetermined cleaning frequency threshold, an alarm information can be sent to the target object, and the target object can process the foreign matter, so that the adverse effects caused by multiple cleanings of the same area can be avoided.

[0114] In some embodiments, the initial running data of the test fixture can also be converted into a target format to obtain target running data, and the target format is a format supported by the maintenance prediction model; the target running data is processed by using the maintenance prediction model to obtain a pre-cleaning period of the test fixture, so that the maintenance device is controlled to maintain the test fixture in the pre-cleaning period.

[0115] In some embodiments, the maintenance prediction model (such as a long short-term memory network) can be started to predict the next cleaning period, and as the number of foreign matter identifications increases, the data in the foreign matter distribution diagram will be more and more, and through the continuous training of these foreign matter data, the prediction result of the maintenance prediction model will be more and more accurate.

[0116] Figure 6 A processing flowchart of the maintenance prediction model according to an embodiment of the present application is shown.

[0117] As shown in Figure 6 , the processing flow of the maintenance prediction model can include operation S601 to operation S606.

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

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

[0120] In operation S603, the target running data is used to train the maintenance prediction model.

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

[0122] In operation S605, the historical running data and the real-time running data are combined for prediction.

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

[0124] In some embodiments, initial running data during the running of the test fixture, i.e. historical running data, can be collected, including test time, test times, detected metal debris and dust quantity, temperature and vibration of the fixture, etc. Then the data needs to be preprocessed, and the format of the initial running data is processed into a target format that can be recognized by the maintenance prediction model to obtain target running data. After the preprocessing, the target running data can be input into the maintenance prediction model, and the target running data is used to train the maintenance prediction model. After the training, the maintenance prediction model can be imported into the edge server.

[0125] During the running of the test fixture, real-time running data can be collected and input into the maintenance prediction model. In this way, the maintenance prediction model can combine the historical running data and the real-time running 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 result. The maintenance system can adjust the maintenance strategy according to the prediction result, for example, the test fixture can be automatically maintained in advance before the time period predicted to need cleaning.

[0126] According to the embodiments of the present application, by using the maintenance prediction model to combine the historical data and real-time running data to predict the time period when the test fixture needs maintenance in the future, and increasing the detection frequency in the key area of foreign matter and performing key 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, the test data of the test fixture during the testing of the board in the server, such as the positioning error, surface roughness, running time, temperature and humidity of the test fixture, etc. can be used to construct the performance degradation curve of the test fixture. The real-time running data of the test fixture and the performance degradation curve are used to predict the pre-cleaning period of the test fixture for the next time, and the test fixture is predicted according to the predicted pre-cleaning period.

[0128] According to the embodiments of the present application, by combining the performance degradation curve of the test fixture itself to predict the pre-cleaning period, the health state threshold of the test fixture can be accurately judged, the blindness of fixed period maintenance can be avoided, and the service life of the test fixture can be prolonged.

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

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

[0131] In operation S710, a first image of the test fixture is identified.

[0132] In operation S720, in a case where it is identified from the first image that the component of the test fixture has a foreign matter, position information of the foreign matter is determined.

[0133] In operation S730, according to the position information of the foreign matter and a carrying state of the carrying component, an operation of the maintenance device is controlled, the carrying state indicating whether the carrying component carries the to-be-tested board.

[0134] The maintenance method of the test fixture provided by the embodiment of the present application has high automation degree, realizes unmanned operation in the whole maintenance process, has high automation degree, and avoids influence of human factors. The maintenance efficiency is high, the fixture maintenance time does not need to be planned separately, the fixture maintenance can be completed in the normal test process, the production test efficiency is improved. The maintenance effect is good, the foreign matter which is small in volume and difficult to be identified by manual work can be automatically identified and cleaned effectively. The test and fixture maintenance strategy is improved, the production line test operation mode and the fixture maintenance strategy can be adjusted and optimized according to the prediction result of the foreign matter distribution map and the maintenance prediction model, the test efficiency is improved, and the service life of the test fixture is prolonged.

[0135] In some embodiments, the above-mentioned scheme can be applied to any business scenario requiring the use of a test fixture, such as related products in the fields of automotive electronics, consumer electronics, new energy, etc.

[0136] Figure 8 A block diagram of an electronic device suitable for implementing the maintenance method of the test fixture according to the embodiment of the present application is schematically shown.

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

[0138] In the RAM 803, various programs and data required for the operation of the electronic device 800 are stored. The processor 801, the ROM 802, and the RAM 803 are connected to each other via the bus 804. The processor 801 performs various operations of the method flow according to the embodiments of the present application by executing the programs in the ROM 802 and / or the RAM 803. It should be noted that the programs can also be stored in one or more memories other than the ROM 802 and the RAM 803. The processor 801 can also perform various operations of the method flow according to the embodiments of the present application by executing the programs stored in the one or more memories.

[0139] According to the embodiments of the present application, the electronic device 800 can further include an input / output (I / O) interface 805, which is also connected to the bus 804. The electronic device 800 can further include one or more of the following components connected to the input / output (I / O) interface 805: an input part 806 including a keyboard, a mouse, etc.; an output part 807 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage part 808 including a hard disk, etc.; and a communication part 809 including a network interface card such as a LAN card, a modem, etc. The communication part 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the input / output (I / O) interface 805 as necessary. A removable recording medium 811 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 810 as necessary, so that a computer program read out therefrom is installed in the storage part 808 as necessary.

[0140] The present application also provides a computer readable storage medium, which can be included in the device / apparatus / system described in the above embodiments; or can exist separately without being assembled into the device / apparatus / system. The above computer readable storage medium carries one or more programs, when the one or more programs are executed, the method according to the embodiments of the present application is implemented.

[0141] According to an embodiment of the present application, the computer readable storage medium can be a non-transitory computer readable storage medium, for example, can include, but is not limited to, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in connection with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present application, the computer readable storage medium can include the ROM 802 and / or the RAM 803 described above and / or one or more memory other than the ROM 802 and the RAM 803.

[0142] Embodiments of the present application also include a computer program product that includes a computer program containing program code for executing the methods illustrated in the flowcharts. When the computer program product is run in a computer system, the program code is used to cause the computer system to implement the methods provided by the embodiments of the present application.

[0143] The above-described functions defined in the system / apparatus / module / unit of the embodiments of the present application are performed when the computer program is executed by the processor 801. According to an embodiment of the present application, the system, apparatus, module, unit, etc. described above can be implemented by computer program modules.

[0144] In one embodiment, the computer program can rely on a tangible storage medium, such as an optical storage device, a magnetic storage device, etc. In another embodiment, the computer program can also be transmitted in the form of a signal on a network medium and downloaded and installed by the communication part 809 and / or installed from the detachable medium 811. The program code contained in the computer program can be transmitted by any suitable network medium, including, but not limited to, wireless, wired, etc., or any suitable combination of the foregoing.

[0145] In such an embodiment, the computer program can be downloaded and installed from a network by the communication part 809 and / or installed from the detachable medium 811. When the computer program is executed by the processor 801, the above-described functions defined in the system of the embodiments of the present application are performed. According to an embodiment of the present application, the system, device, apparatus, module, unit, etc. described above can be implemented by computer program modules.

[0146] According to embodiments of the present application, program code for implementing the computer programs provided by embodiments of the present application can be written in any combination of one or more programming languages, and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. The programming language can include, but is not limited to, Java, C++, python, "C" language, or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the remote computing device, or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider.

[0147] The computer program instructions can also be loaded onto a computer or other programmable information processing apparatus to cause a series of operations to be performed on the computer or other programmable information processing apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable information processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0148] Those skilled in the art will appreciate that the features recited in the various embodiments of the present application can be combined and / or integrated in a variety of ways, even if such combinations or integrations are not expressly noted in the present application. In particular, the features recited in the various embodiments of the present application can be combined and / or integrated in a variety of ways without departing from the spirit and scope of the present application. All such combinations and / or integrations are within the scope of the present application.

[0149] The embodiments of the present application have been described above. However, these embodiments are merely for the purpose of illustration and are not intended to limit the scope of the present application. Although the embodiments are described separately above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Those skilled in the art can make various substitutions and modifications without departing from the scope of the present application, and these substitutions and modifications should fall within the scope of the present application.

Claims

1. A maintenance system for a test fixture, the maintenance system comprising: The system comprises: a control device configured to identify a first image of a test fixture, and determine position information of a foreign matter on a component of the test fixture in a case where the foreign matter is identified from the first image, the first image of the test fixture being obtained by an image acquisition device acquiring an image of the component of the test fixture, the component comprising a movable test board and a carrying component for carrying the test board, the test board being configured to implement testing of a test object by the test fixture; and control operation of a maintenance device according to the position information of the foreign matter and a carrying state of the carrying component, the carrying state indicating whether the carrying component carries the test board, and obtain a pre-cleaning period of the test fixture according to real-time running data and a performance attenuation curve of the test fixture, the performance attenuation curve being obtained according to test data of the test fixture; a maintenance device configured to perform a maintenance operation on the test fixture under control of the control device, the maintenance device comprising a piezoelectric vibrator and a magnetic control adsorption component; in a case where the carrying component carries the test board, the control device sends a first control instruction to the piezoelectric vibrator to vibrate the piezoelectric vibrator at a predetermined frequency to loosen the foreign matter on the carrying component, and sends a second control instruction to the magnetic control adsorption component to adsorb the loosened foreign matter on the carrying component according to a removal path, a caliber of a magnetic control suction nozzle of the magnetic control adsorption component being adjusted to be matched with a size of the foreign matter, and in a case where the carrying component does not carry the test board, the control device sends a third control instruction to the magnetic control adsorption component to adsorb the foreign matter on the test board according to the removal path, the caliber of the magnetic control suction nozzle of the magnetic control adsorption component being adjusted to be matched with the size of the foreign matter.

2. The maintenance system of claim 1, wherein, The control device is further configured to: generate a removal path for removing the foreign matter according to the position information; control the maintenance device to clean and maintain the foreign matter according to the removal path according to the carrying state of the carrying component.

3. The maintenance system of claim 1, wherein, The control device is further configured to: extract a multi-scale feature of the first image; weight the feature for the component based on a weight of the component to obtain a weighted feature map; determine a plurality of candidate boxes comprising the feature in the weighted feature map; determine that the foreign matter exists on the component of the test fixture in a case where a confidence of the candidate box is greater than a predetermined confidence, the confidence being configured to describe a probability that the feature in the candidate box is the foreign matter.

4. The maintenance system of claim 3, wherein, The control device is further configured to: obtain the position information of the foreign matter according to position information of the candidate box in a case where the feature is related to the candidate box; determine a target box from the at least two candidate boxes according to the confidence of the candidate box in a case where the feature is related to at least two candidate boxes, and obtain the position information of the foreign matter according to position information of the target box.

5. The maintenance system of claim 1, wherein, The control device is further configured to: In a case that the foreign matter is maintained, image acquisition is performed on the position indicated by the position information by using the image acquisition device, and a second image is obtained; In a case that the foreign matter is identified from the second image, the maintenance device is controlled to clean the foreign matter again; In a case that the foreign matter is not identified from the second image, the position information is used to update a foreign matter distribution map, which is used to describe the distribution of the foreign matter on the test fixture.

6. The maintenance system of claim 1, wherein, The control device is further configured to: convert initial operation data of the test fixture into a target format to obtain target operation data, the target format being a format supported by a maintenance prediction model; process the target operation data by using the maintenance prediction model to obtain a pre-cleaning period of the test fixture, so as to control the maintenance device to maintain the test fixture in the pre-cleaning period.

7. A method of maintaining a test fixture, the method comprising: The method is applied to the maintenance system according to any one of claims 1-6, and the method comprises: identifying the first image of the test fixture; in a case that a foreign matter is identified from the first image, determining position information of the foreign matter; controlling operation of the maintenance device according to the position information of the foreign matter and a bearing state of the bearing assembly, the bearing state indicating whether the bearing assembly bears the to-be-tested board.

8. A computer readable storage medium having stored thereon a computer program or instructions, characterized in that, The computer program or instructions are executed by the processor to realize the steps of the method according to claim 7.

Images