Method and apparatus for testing high power electronic fan printed circuit boards

Abstract

The application discloses a kind of high-power electronic fan printed circuit board assembly test method and equipment, the method includes: identifying the rigid contact relationship between the parts assembled in high-power electronic fan when vibrating;Part with rigid contact relationship is constructed with weight undirected graph;In the weight undirected graph, the multi-order neighborhood subgraph of target path and printed circuit board is extracted;Target path is the shortest path between printed circuit board and the edge of weight undirected graph;According to target path and multi-order neighborhood subgraph, the category to which high-power electronic fan belongs is divided;According to the sample vibration signal before sample electronic fan load fails under the same category, test case is generated;When high-power electronic fan is fixed on test table, test table is controlled according to test case to test high-power electronic fan, to generate test report.The embodiment improves the test efficiency of high-power electronic fan on the structure defect of assembly printed circuit board.

Description

Technical Field

[0001] This invention relates to the field of deep learning technology, and in particular to an assembly and testing method and equipment for a high-power electronic fan printed circuit board. Background Technology

[0002] For high-power motors in new energy vehicles, OEMs are gradually transitioning from configuring 3-4 low-power electric fans to configuring 1 high-power electric fan, which simplifies the vehicle wiring harness, reduces the overall size of the electric fan, and lowers the cost of the electric fan.

[0003] During the transition phase, OEMs typically develop high-power electric fans based on low-power electric fans. This results in many similarities between the high-power and low-power electric fans in their overall assembly processes. In particular, the PCBA (Printed Circuit Board Assembly) stage uses the assembly process of low-power electric fans, directly attaching the PCB (Printed Circuit Board) to the frame.

[0004] Because high-power electronic fans have a larger diameter and are heavier, the middle part of the printed circuit board will dent after the back cover is locked, so vibration (i.e. mechanical load) testing of high-power electronic fans is an important part of their performance testing.

[0005] Currently, vibration testing of printed circuit boards mainly involves resonance testing and simulated operating condition testing. Among them, simulated operating condition testing simulates the vibration of printed circuit boards based on actual operating scenarios, focusing on comprehensiveness, but has low testing efficiency for high-power electronic fans in the assembly of printed circuit board structural defects. Summary of the Invention

[0006] In view of this, the present invention provides an assembly testing method and equipment for high-power electronic fan printed circuit boards, so as to improve the testing efficiency of high-power electronic fans on structural defects in the assembled printed circuit boards.

[0007] The first aspect of the present invention provides an assembly and testing method for a high-power electronic fan printed circuit board, wherein the high-power electronic fan is an automotive engine cooling electronic fan with a power greater than a threshold, and the method includes:

[0008] Identify the rigid contact relationship between the components assembled in the high-power electronic fan during vibration; the components include printed circuit boards;

[0009] Construct a weighted undirected graph for the parts having the rigid contact relationship;

[0010] Extract the target path and the multi-order nearest neighbor subgraph of the printed circuit board from the weighted undirected graph; the target path is the shortest path between the printed circuit board and the edge of the weighted undirected graph;

[0011] The high-power electronic fan is classified into categories based on the target path and the multi-order nearest neighbor subgraph;

[0012] Test cases are generated based on the vibration signals of multiple sample electronic fans under the same category before the failure.

[0013] When the high-power electronic fan is fixed on the test bench, the test bench is controlled to test the high-power electronic fan according to the test case to generate a test report.

[0014] A second aspect of the present invention provides an assembly and testing apparatus for a high-power electronic fan printed circuit board, wherein the high-power electronic fan is an automotive engine cooling electronic fan with a power exceeding a threshold, and the apparatus comprises:

[0015] A rigid contact relationship identification module is used to identify the rigid contact relationship between the parts assembled in the high-power electronic fan during vibration; the parts include printed circuit boards;

[0016] A weighted undirected graph construction module is used to construct a weighted undirected graph for the parts having the rigid contact relationship;

[0017] The graph information extraction module is used to extract a multi-order nearest neighbor subgraph between the target path and the printed circuit board in the weighted undirected graph; the target path is the shortest path between the printed circuit board and the edge of the weighted undirected graph;

[0018] The category classification module is used to classify the category to which the high-power electronic fan belongs based on the target path and the multi-order nearest neighbor subgraph;

[0019] The test case generation module is used to generate test cases based on the vibration signals of multiple sample electronic fans under the same category before the failure.

[0020] The vibration testing module is used to control the test bench to test the high-power electronic fan according to the test cases when the high-power electronic fan is fixed on the test bench, so as to generate a test report.

[0021] A third aspect of the present invention provides an electronic device, the electronic device comprising:

[0022] At least one processor; and

[0023] A memory communicatively connected to the at least one processor; wherein,

[0024] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the assembly and testing method for the high-power electronic fan printed circuit board as described in the first aspect above.

[0025] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the assembly and testing method for a high-power electronic fan printed circuit board as described in the first aspect above.

[0026] The fifth aspect of the present invention provides a computer program product comprising a computer program that, when executed by a processor, implements the assembly and testing method for a high-power electronic fan printed circuit board as described in the first aspect above.

[0027] In this embodiment, the rigid contact relationship between components assembled in a high-power electronic fan during vibration is identified; the components include printed circuit boards (PCBs); a weighted undirected graph is constructed for the components with rigid contact relationships; a multi-order nearest neighbor subgraph between the target path and the PCB is extracted from the weighted undirected graph; the target path is the shortest path between the PCB and the edge of the weighted undirected graph; the high-power electronic fan is classified into categories based on the target path and the multi-order nearest neighbor subgraph; test cases are generated based on the sample vibration signals of multiple sample electronic fans in the same category before failure; when the high-power electronic fan is fixed on the test bench, the test bench is controlled to test the high-power electronic fan according to the test cases to generate a test report. This embodiment finds sample electronic fans with similar structures that transmit stress to the PCB, and uses their sample vibration signals before failure for specialized testing, improving the testing efficiency of high-power electronic fans for defects in PCB assembly structures, quickly uncovering defects in PCB assembly structures, and thus improving the testing efficiency of high-power electronic fans for defects in PCB assembly structures.

[0028] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of an assembly and testing method for a high-power electronic fan printed circuit board provided in Embodiment 1 of the present invention.

[0031] Figure 2 This is a front view of a high-power electronic fan provided in Embodiment 1 of the present invention.

[0032] Figure 3 This is a rear view of a high-power electronic fan provided in Embodiment 1 of the present invention.

[0033] Figure 4 This is an exploded view of a high-power electronic fan provided in Embodiment 1 of the present invention.

[0034] Figure 5 This is an architecture diagram of an embodiment of the present invention for identifying the probability of rigid contact between two adjacent parts.

[0035] Figure 6 This is a schematic diagram of the assembly and testing device for a high-power electronic fan printed circuit board provided in Embodiment 2 of the present invention.

[0036] Figure 7 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0037] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0038] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be used interchangeably where appropriate so that the embodiments of the invention described herein can cover implementations in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0039] Example 1

[0040] See Figure 1The diagram illustrates a flowchart of an assembly and testing method for a high-power electronic fan printed circuit board according to Embodiment 1 of the present invention. This method can be executed by an assembly and testing device for the high-power electronic fan printed circuit board, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0041] Step 101: Identify the rigid contact relationship between the parts assembled in the high-power electronic fan during vibration.

[0042] In this embodiment, the high-power electric fan is an automotive engine cooling electric fan, where high power means that its power is greater than a threshold (e.g., 1400W).

[0043] Specifically, high-power electric fans (refrigeration) refer to fan products used for cooling and heat dissipation of automobile engines, consisting of fan blades as the main body and components such as metal support plates.

[0044] A high-powered electric fan (cooling) prevents the car engine from overheating and maintains the optimal operating temperature of the air conditioning system by expelling hot air and promoting efficient heat exchange, ensuring the smooth and reliable operation of the car engine and air conditioning functions.

[0045] For example, the maximum airflow of a single 1500W high-power electric fan is typically greater than that of three 800W low-power electric fans (i.e., greater than 12000m³ at 100Pa). 3 The maximum airflow of a single 2000W high-power electric fan is typically greater than that of four 800W low-power electric fans (i.e., greater than 16000m³ / h at 100Pa). 3 / h).

[0046] In a heat dissipation design, such as Figure 2 , Figure 3 and Figure 4 As shown, the structure of a certain high-power electronic fan includes a fan cover 401, a back cover 402, a wiring harness 403, a printed circuit board 404, a rotor 405, a stator 406, a housing 407, fan blades 408, etc.

[0047] Furthermore, in addition to setting printed circuit boards as components, other structures (especially rotors, stators, and other assemblies) can be further broken down into finer-grained components.

[0048] In high-power electric fans, there may be rigid contact relationships between certain components. Rigid contact is a concept in mechanics that describes the contact state of two or more components. Its characteristics are that when the contacting parties are subjected to forces (such as vibration, impact, pressure), the relative displacement of the contact interface is extremely small (negligible), and the contact force can be efficiently transmitted through the contact interface.

[0049] In general, some parts of a high-power electric fan form a fixed rigid contact relationship through screws, clips, welding, etc. When vibrating, these parts will also form a rigid contact relationship. Some parts of a high-power electric fan form a temporary rigid contact relationship when vibrating due to their close proximity.

[0050] Printed circuit boards may fail due to vibration (such as component detachment, circuit breakage, or even breakage), mostly because the vibration stress exceeds the tolerance limit of the material and / or structure. Rigid contact largely determines the distribution pattern of vibration stress in high-power electric fans and the pattern of transmission to printed circuit boards.

[0051] In one embodiment of the present invention, a three-dimensional design drawing of a high-power electronic fan can be provided, the three-dimensional design drawing of the high-power electronic fan having a three-dimensional model of the parts assembled in the high-power electronic fan; step 101 may include the following steps:

[0052] Step 1011: Sample the surface of the 3D model to obtain the original point cloud data.

[0053] Generally, the 3D model of a part mainly describes the outline structure of the part and does not have an internal structure. Therefore, in order to reduce the interference of the internal structure and improve the utilization rate of the outline structure, multiple points can be sampled on the surface of the 3D model using a preset sampling frequency to obtain the original point cloud data. At this time, some projects developed for point clouds can be reused to reduce the development workload.

[0054] Step 1012: For the same part, map the material of the part to the original point cloud data of the part to obtain candidate point cloud data.

[0055] In this embodiment, the material of a part can be queried in the local design file based on the ID or other identifiers of the part's 3D model. These materials are classified in a granular manner, such as plastic or metal.

[0056] The material of the part is encoded using one-hot encoding and other methods. The encoded material of the part is then mapped to each point in the original point cloud data of the part, adding a feature dimension to each point in the original point cloud data, which is denoted as candidate point cloud data.

[0057] Step 1013: Classify the candidate point cloud data of two adjacent parts to obtain the probability of rigid contact between the two adjacent parts.

[0058] In this embodiment, two adjacent parts in three-dimensional space can be selected from the three-dimensional design drawings, and deep learning technology can be used to classify the candidate point cloud data of the two adjacent parts to obtain the probability of rigid contact between the two adjacent parts.

[0059] In one embodiment of the present invention, step 1013 may further include the following steps: Step 1031, Identifying the positional relationship between two parts based on candidate point cloud data.

[0060] In this embodiment, geometrical statistics can be used to identify the positional relationship between pairs of parts based on candidate point cloud data.

[0061] For example, a minimum bounding cube can be added to the candidate point cloud data of each part, and the minimum bounding cube can be expanded according to a preset magnification factor (such as 1.1). That is, the length, width and height of the minimum bounding cube are multiplied by the magnification factor to obtain new length, width and height, thereby completing the expansion.

[0062] If the two smallest bounding cubes after expansion intersect, then the positional relationship of the parts corresponding to the two smallest bounding cubes after expansion is determined to be adjacent.

[0063] This example relies on expansion and geometric intersection to identify adjacent parts. The calculation is simple, and expanding the smallest circumscribed cube can roughly simulate the scenario of displacement during vibration. This can initially screen out parts that may generate temporary rigid contact relationships during vibration, reducing the amount of subsequent calculations.

[0064] Of course, the above-described method for identifying the positional relationship between pairs of parts is merely an example. In implementing this embodiment, other methods for identifying the positional relationship between pairs of parts can be set according to actual circumstances. For example, when the distance between the candidate point cloud data of two parts is less than or equal to a certain threshold, the positional relationship between the two parts can be determined as adjacent, etc. This embodiment does not impose any limitations on this. Furthermore, in addition to the above-described method for identifying the positional relationship between pairs of parts, those skilled in the art can also employ other methods for identifying the positional relationship between pairs of parts according to actual needs, and this embodiment does not impose any limitations on this either.

[0065] Step 1032: Based on the vibration modes of the printed circuit board in the automobile, derive the candidate point cloud data of each part to obtain the target point cloud data.

[0066] In this embodiment, the vibration modes generated by the car during driving can be analyzed and statistically analyzed for the printed circuit board in the high-power electronic fan. Based on this, the candidate point cloud data of each part can be derived to generate new points and obtain target point cloud data to simulate the offset distribution of the part's contour during vibration. Furthermore, the density of the point cloud can be increased to facilitate subsequent deep learning-based calculations.

[0067] In practical implementation, the target line of two intersecting planes can be identified in the candidate point cloud data. The target line is usually the corner line in the part, which is a rigid structure and does not produce relative displacement when vibrating.

[0068] On the same plane, the distance L between the candidate point cloud data points and the target line is calculated. The greater the distance between a point and the target line (angle line), the greater the amplitude of the relative displacement during vibration. Conversely, the smaller the distance between a point and the target line (angle line), the smaller the amplitude of the relative displacement during vibration.

[0069] Furthermore, if there are multiple distances L for the same point in the candidate point cloud data, the minimum or average value of the multiple distances L can be taken as the final distance L.

[0070] Random vibration signal Vi is generated by taking multiple random values ​​within a preset amplitude range.

[0071] The target state transition matrix RVi is obtained by multiplying each random vibration signal Vi with the original state transition matrix R representing the vibration transmission law of the printed circuit board in the car. The target state transition matrix RVi can describe the vibration signal Vi after transmission correction.

[0072] The original state transition matrix R is usually an n×n matrix, representing the time domain dimension and the frequency domain dimension, and it has elements such as amplitude attenuation, relative offset, and frequency transfer coefficient.

[0073] Singular Value Decomposition (SVD) is performed on the target state transition matrix RVi to obtain the vibration constraint parameter SVD(RVi), thereby removing redundant information from the vibration signal and extracting the main transmission mode of the vibration.

[0074] Calculate the state transition parameter RViL, where the transition parameter RViL is the product between the target state transition matrix RVi and the distance L. It is the quantified value of the actual vibration state of the points in the candidate point cloud data after considering the position of the rigid reference (bent line). That is, the distance L is used as a weight to reflect the degree of influence of the rigid constraint on the vibration.

[0075] The second norm of the difference between the state transition parameter and the vibration constraint parameter SVD(RVi) is taken as the vibration offset parameter Z, i.e., Z=||RViL-SVD(RVi)||2. By combining the vibration state transmission with the position of the rigid structure (angle line), redundant information in the vibration signal is removed, and the effective vibration state deviation at a specific position (distance L from the angle line) is finally quantified to analyze the vibration transmission law near the rigid structure.

[0076] The target point cloud data is obtained by adding vibration offset parameters to the points in the candidate point cloud data in the manner of generating new points. The manner of generating new points refers to adding vibration offset parameters to the coordinates of the points in the candidate point cloud data while retaining other attributes (such as material) except coordinates.

[0077] Step 1033: Input the target point cloud data of each part into the attention network to generate the first point cloud feature.

[0078] In this embodiment, as Figure 5 As shown, the target point cloud data of each part can be traversed. The target point cloud data of the part is used as the Q (query) matrix, K (key) matrix and V (value) matrix to be input into the preset self-attention network to extract shallow features and obtain the first point cloud features.

[0079] On the one hand, the 3D model of the part is not a high-precision model and usually blurs some details, making the target point cloud data unstructured points. By calculating the attention weight between each point and other points through the Self-Attention network, it is possible to identify points related in the spatial structure, reduce the dependence on the grid or local neighborhood, and model the global spatial relationship.

[0080] On the other hand, the coordinates of a point (i.e., geometric information) are associated with the material (i.e., physical properties), thus realizing cross-modal correlation features.

[0081] By addressing these two aspects, noise interference in the target point cloud data can be reduced, and the focus can be placed on key local areas related to vibration in the target point cloud data (such as bend lines, material abrupt change areas, etc.), thereby reducing the cost of adapting to rigid contact tasks.

[0082] Step 1034: Combine the first point cloud features corresponding to two adjacent parts to form the second point cloud feature.

[0083] In this embodiment, as Figure 5 As shown, functions such as Concat can be used to stitch together the first point cloud features corresponding to two adjacent parts into a second point cloud feature.

[0084] Step 1035: Input the second point cloud feature into the backbone network to extract the third point cloud feature.

[0085] In this embodiment, as Figure 5 As shown, the second point cloud features can be input into the pre-set backbone network to further extract high-level features, thus obtaining the third point cloud features.

[0086] In practical implementation, the pre-trained neural network (such as PointNet, PointNet++, Point Transformer) can be fine-tuned based on samples from the rigid contact task to obtain the backbone network.

[0087] Step 1036: Input the third point cloud feature into the head network to generate the probability of rigid contact between two adjacent parts.

[0088] In this embodiment, as Figure 5 As shown, the third point cloud features can be input into the preset head network to perform a rigid contact task, generating the probability of rigid contact between two adjacent parts.

[0089] The head network includes fully connected layers (FC), activation function Sigmoid, and other components.

[0090] Step 1014: If the probability is greater than or equal to the preset threshold, then it is determined that there is a rigid contact relationship between two adjacent parts.

[0091] In this embodiment, the probability of rigid contact between two adjacent parts can be compared with a preset threshold.

[0092] If the probability is greater than or equal to the preset threshold, it indicates that the probability of rigid contact between two adjacent parts during vibration is high, and it can be determined that there is a rigid contact relationship between the two adjacent parts.

[0093] If the probability is less than the preset threshold, it means that the probability of rigid contact between two adjacent parts during vibration is low, and it can be determined that there is no rigid contact relationship between the two adjacent parts.

[0094] Step 102: Construct a weighted undirected graph for parts with rigid contact relationships.

[0095] In this embodiment, parts with rigid contact relationships can be modeled to obtain a weighted undirected graph.

[0096] In practical implementation, parts can be converted into nodes (also known as vertices).

[0097] If there is a rigid contact relationship between two parts, then add an edge between the corresponding nodes of the two parts. The edge has no direction, indicating that the rigid contact relationship between the nodes is bidirectional.

[0098] The cross-sectional area of ​​the intersection between the target point cloud data of two parts is statistically analyzed and used as the simplified area of ​​the rigid contact between the two parts. Weights are assigned to the edges according to the cross-sectional area to obtain a weighted undirected graph. The weights are positively correlated with the cross-sectional area, that is, the larger the cross-sectional area, the larger the weight, and vice versa. This indicates a positive correlation between the area of ​​rigid contact and the uniformity of stress transmission.

[0099] Step 103: Extract the multi-order nearest neighbor subgraph between the target path and the printed circuit board from the weighted undirected graph.

[0100] On the one hand, a target path can be extracted from a weighted undirected graph, where the target path is the shortest path between the printed circuit board and the edge of the weighted undirected graph (such as a node representing a shroud or blade).

[0101] On the other hand, a multi-order nearest neighbor subgraph of the printed circuit board is extracted from the weighted undirected graph. The multi-order nearest neighbor subgraph is the neighborhood structure of the nodes representing the printed circuit board. When the order is K, other nodes with paths less than or equal to K between them and the nodes representing the printed circuit board are extracted from the weighted undirected graph to obtain the K-order nearest neighbor subgraph.

[0102] Step 104: Classify the high-power electronic fan according to the target path and the multi-order nearest neighbor subgraph.

[0103] In this embodiment, sample electronic fans (especially low-power electronic fans) can be set. The sample electronic fans are automotive engine cooling electronic fans. The target path and multi-order nearest neighbor subgraph are constructed for them in the same way. The target path and multi-order nearest neighbor subgraph are used to cluster the sample electronic fans to obtain multiple categories.

[0104] The target path belongs to the vibration transmission topology, which can represent the vibration transmission capability to a certain extent. The multi-order nearest neighbor subgraph belongs to the surrounding structure topology. The characteristics of both can be combined to classify the category to which the high-power electric fan belongs.

[0105] Within the same category, electronic fans (including high-power electronic fans and sample electronic fans) exhibit similarities in their structure for transmitting stress to printed circuit boards.

[0106] On the one hand, the target path and multi-order nearest neighbor subgraphs are used in the graph dimension to identify the first probability of a high-power electronic fan belonging to each category.

[0107] In a practical implementation, the target path can be input into a pre-defined translation network (such as TransE, TransH, TransR, and TransD) to generate the first graph features;

[0108] Multi-order nearest neighbor subgraphs are input into translation networks (such as TransE, TransH, TransR, and TransD) to generate second graph features.

[0109] The features of the first image and the features of the second image are concatenated and input into a pre-set classifier (such as SVM (Support Vector Machine), K-means, etc.) to identify the first probability of a high-power electronic fan belonging to each category.

[0110] On the other hand, the semantics of high-powered electronic fans are added to the multi-order nearest neighbor subgraph, and the second probability of high-powered electronic fans belonging to each category is identified using the multi-order nearest neighbor subgraph in the dimensions of graph and semantics.

[0111] In the specific implementation, the descriptive information of the parts (such as information describing the structure, function, size and other attributes) is mapped to the nodes of the multi-order nearest neighbor subgraph. If the mapping is completed, the multi-order nearest neighbor subgraph is input into the graph neural network (GNN) to identify the second probability of the high-power electronic fan belonging to each category.

[0112] For the same category, the first probability and the second probability are merged into a third probability using methods such as weighted summation. The third probabilities of each category are then compared to determine which category the high-power electronic fan belongs to.

[0113] Step 105: Generate test cases based on the vibration signals of multiple sample electronic fans under the same category before the failure.

[0114] In general, each sample electronic fan is actually installed in the car while it is running. The operating status of the sample electronic fan is monitored. If the operating status is faulty, the vibration signal before the fault occurs can be collected and recorded as the sample vibration signal. This makes it easier to analyze the cause of the fault and optimize the sample electronic fan.

[0115] In this embodiment, test cases can be automatically generated based on the vibration signals of multiple sample electronic fans under the same category before the failure, using methods such as LLM (Large Language Model), and the test cases can be manually calibrated.

[0116] Step 106: When the high-power electronic fan is fixed on the test bench, control the test bench to test the high-power electronic fan according to the test cases to generate a test report.

[0117] In this embodiment, the tester fixes the high-power electronic fan on the test bench and controls the test bench to perform various tests on the high-power electronic fan according to the test cases, including vibration tests, thereby automatically generating a test report.

[0118] The test bench is an electromagnetic vibration table that can simulate the vibration (including sinusoidal vibration, random vibration, and impact vibration) transmitted from the car body to a high-power electric fan, providing controllable vibration excitation.

[0119] The test bench includes tools such as accelerometers and fixtures. The accelerometers are used to capture the vibration response of key parts of the high-power fan (especially the printed circuit board), and the fixtures are used to fix the high-power electric fan on the test bench to simulate the actual vehicle installation state (such as bolt connection, bracket fixation, etc.).

[0120] Of course, in addition to accelerometers, sensors such as RGB (red, green, and blue) cameras and infrared cameras can be deployed on printed circuit boards to more comprehensively capture the test responses of key parts of high-power fans (especially printed circuit boards).

[0121] The test bench is deployed in an environmental test chamber to simulate the temperature and humidity environment in which a high-power electronic fan actually operates.

[0122] In this embodiment, the rigid contact relationship between components assembled in a high-power electronic fan during vibration is identified; the components include printed circuit boards (PCBs); a weighted undirected graph is constructed for the components with rigid contact relationships; a multi-order nearest neighbor subgraph between the target path and the PCB is extracted from the weighted undirected graph; the target path is the shortest path between the PCB and the edge of the weighted undirected graph; the high-power electronic fan is classified into categories based on the target path and the multi-order nearest neighbor subgraph; test cases are generated based on the sample vibration signals of multiple sample electronic fans in the same category before failure; when the high-power electronic fan is fixed on the test bench, the test bench is controlled to test the high-power electronic fan according to the test cases to generate a test report. This embodiment finds sample electronic fans with similar structures that transmit stress to the PCB, and uses their sample vibration signals before failure for specialized testing, improving the testing efficiency of high-power electronic fans for defects in PCB assembly structures, quickly uncovering defects in PCB assembly structures, and thus improving the testing efficiency of high-power electronic fans for defects in PCB assembly structures.

[0123] Example 2

[0124] See Figure 6 This diagram illustrates the structure of an assembly and testing device for a high-power electronic fan printed circuit board according to Embodiment 2 of the present invention. The high-power electronic fan is an automotive engine cooling fan with a power exceeding a threshold value, such as... Figure 6As shown, the device includes:

[0125] Rigid contact relationship identification module 601 is used to identify the rigid contact relationship between the parts assembled in the high-power electronic fan during vibration; the parts include printed circuit boards;

[0126] Weighted undirected graph construction module 602 is used to construct a weighted undirected graph for the parts having the rigid contact relationship;

[0127] The graph information extraction module 603 is used to extract a multi-order nearest neighbor subgraph between the target path and the printed circuit board in the weighted undirected graph; the target path is the shortest path between the printed circuit board and the edge of the weighted undirected graph.

[0128] The category classification module 604 is used to classify the category to which the high-power electronic fan belongs based on the target path and the multi-order nearest neighbor subgraph;

[0129] The test case generation module 605 is used to generate test cases based on the vibration signals of multiple sample electronic fans under the same category before the failure.

[0130] The vibration testing module 606 is used to control the test bench to test the high-power electronic fan according to the test cases when the high-power electronic fan is fixed on the test bench, so as to generate a test report.

[0131] In one embodiment of the present invention, the three-dimensional design drawing of the high-power electronic fan includes a three-dimensional model of the parts assembled in the high-power electronic fan; the rigid contact relationship identification module 601 includes:

[0132] The raw point cloud data sampling module is used to sample the surface of the three-dimensional model to obtain raw point cloud data.

[0133] The candidate point cloud data generation module is used to map the material of the same part to the original point cloud data of the part to obtain candidate point cloud data.

[0134] The rigid contact probability calculation module is used to classify the candidate point cloud data of two adjacent parts to obtain the probability of rigid contact between the two adjacent parts.

[0135] A rigid contact relationship determination module is used to determine that there is a rigid contact relationship between two adjacent parts if the probability is greater than or equal to a preset threshold.

[0136] In one embodiment of the present invention, the rigid contact probability calculation module includes:

[0137] A positional relationship recognition module is used to identify the positional relationship between any two of the parts based on the candidate point cloud data;

[0138] The target point cloud data derivation module is used to derive the candidate point cloud data of each part based on the vibration mode of the printed circuit board in the automobile to obtain the target point cloud data.

[0139] The first point cloud feature generation module is used to input the target point cloud data of each of the parts into the attention network to generate the first point cloud features;

[0140] The second point cloud feature stitching module is used to stitch together the first point cloud features corresponding to two parts whose positional relationship is adjacent into a second point cloud feature;

[0141] The third point cloud feature extraction module is used to input the second point cloud features into the backbone network to extract the third point cloud features.

[0142] The adjacent probability generation module is used to input the third point cloud features into the head network to generate the probability of rigid contact between two adjacent parts.

[0143] In one embodiment of the present invention, the position relationship identification module includes:

[0144] The minimum bounding cube adding module is used to add a minimum bounding cube to the candidate point cloud data;

[0145] A minimum circumscribed cube expansion module is used to expand the minimum circumscribed cube according to a preset magnification factor;

[0146] The adjacent determination module is used to determine that the positional relationship of the parts corresponding to the two smallest outer cubes after expansion is adjacent if the two expanded cubes intersect.

[0147] In one embodiment of the present invention, the target point cloud data derivation module includes:

[0148] The target line identification module is used to identify target lines where two planes intersect in the candidate point cloud data;

[0149] The distance calculation module is used to calculate the distance between points in the candidate point cloud data and the target line on the same plane.

[0150] A random vibration signal generation module is used to generate random vibration signals within a preset amplitude range;

[0151] The target state transition matrix calculation module is used to calculate the product between the random vibration signal and the original state transition matrix representing the vibration transmission law of the printed circuit board in the car, so as to obtain the target state transition matrix.

[0152] The vibration constraint parameter generation module is used to perform singular value decomposition on the target state transition matrix to obtain vibration constraint parameters.

[0153] The vibration offset parameter calculation module is used to take the second norm of the difference between the state transition parameter and the vibration constraint parameter, and use it as the vibration offset parameter; the state transition parameter is the product of the target state transition matrix and the distance.

[0154] The vibration offset parameter adding module is used to add the vibration offset parameter to the points in the candidate point cloud data in a way that generates new points, so as to obtain the target point cloud data.

[0155] In one embodiment of the present invention, the weighted undirected graph construction module 602 includes:

[0156] A node conversion module is used to convert the part into a node;

[0157] An edge-adding module is used to add an edge between the nodes corresponding to the two parts if there is a rigid contact relationship between them.

[0158] The weight assignment module is used to calculate the cross-sectional area of ​​the intersection range between the target point cloud data of the two parts, and assign weights to the edges according to the cross-sectional area to obtain a weighted undirected graph; the weights are positively correlated with the cross-sectional area.

[0159] In one embodiment of the present invention, the category classification module 604 includes:

[0160] The first probability calculation module is used to identify the first probability of the high-power electronic fan belonging to each category using the target path and the multi-order nearest neighbor subgraph in the dimension of the graph;

[0161] The second probability calculation module is used to identify the second probability of the high-power electronic fan belonging to each category using the multi-order nearest neighbor subgraph in the dimensions of graph and semantics.

[0162] The third probability fusion module is used to fuse the first probability and the second probability into a third probability for the same category;

[0163] The category determination module is used to determine whether the high-power electronic fan belongs to the third category with the highest probability.

[0164] In one embodiment of the present invention, the first probability calculation module includes:

[0165] The first image feature generation module is used to input the target path into the translation network to generate the first image features;

[0166] The second graph feature generation module is used to input the multi-order nearest neighbor subgraph into the translation network to generate second graph features;

[0167] The first probability recognition module is used to concatenate the features of the first image and the features of the second image and input them into the classifier to identify the first probability of the high-power electronic fan belonging to each category.

[0168] The second probability calculation module includes:

[0169] The description information mapping module is used to map the description information of the part to the nodes of the multi-order nearest neighbor subgraph;

[0170] The second probability identification module is used to input the multi-order nearest neighbor subgraph into the graph neural network if the mapping is completed, and to identify the second probability of the high-power electronic fan belonging to each category.

[0171] The assembly and testing apparatus for high-power electronic fan printed circuit boards provided in this embodiment of the invention can execute the assembly and testing method for high-power electronic fan printed circuit boards provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the assembly and testing method for high-power electronic fan printed circuit boards.

[0172] Example 3

[0173] See Figure 7 This diagram illustrates a structural schematic of an electronic device according to an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, blade servers, mainframe computers, and other suitable computers. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0174] like Figure 7As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0175] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0176] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the assembly testing method for a high-power electronic fan printed circuit board.

[0177] In some embodiments, the assembly testing method for a high-power electric fan printed circuit board can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the assembly testing method for the high-power electric fan printed circuit board described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the assembly testing method for the high-power electric fan printed circuit board by any other suitable means (e.g., by means of firmware).

[0178] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0179] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0180] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0181] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0182] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0183] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0184] Example 4

[0185] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the assembly and testing method for a high-power electronic fan printed circuit board as provided in any embodiment of this invention.

[0186] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0187] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0188] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. An assembly and testing method for a high-power electronic fan printed circuit board, characterized in that, The high-power electric fan is a car engine cooling electric fan, and its power exceeds a threshold, which includes 1400W. The method includes: Identify the rigid contact relationship between components assembled in a high-power electric fan during vibration; components include printed circuit boards; Construct a weighted undirected graph for parts with rigid contact relationships; Extract the multi-order nearest neighbor subgraph between the target path and the printed circuit board in a weighted undirected graph; the target path is the shortest path between the printed circuit board and the edge of the weighted undirected graph. The high-power electronic fan is classified into categories based on the target path and the multi-order nearest neighbor subgraph; Test cases are generated based on the vibration signals of multiple sample electronic fans of the same category before the failure. When a high-power electronic fan is fixed on the test bench, the test bench is controlled to test the high-power electronic fan according to the test cases to generate a test report; Among them, the 3D design drawings of the high-power electric fan include 3D models of the parts assembled in the high-power electric fan; the rigid contact relationship between the parts assembled in the high-power electric fan during vibration is identified, including: Sampling is performed on the surface of the 3D model to obtain the raw point cloud data; For the same part, the material of the part is mapped to the original point cloud data of the part to obtain candidate point cloud data; Identify the positional relationships between pairs of parts based on candidate point cloud data; Based on the vibration modes of printed circuit boards in automobiles, candidate point cloud data of each component are derived to obtain target point cloud data; The target point cloud data of each part is input into the attention network to generate the first point cloud feature; The first point cloud features corresponding to two adjacent parts are stitched together to form the second point cloud features; The second point cloud feature is input into the backbone network to extract the third point cloud feature; The third point cloud feature is input into the head network to generate the probability of rigid contact between two adjacent parts; If the probability is greater than or equal to the preset threshold, then it is determined that there is a rigid contact relationship between two adjacent parts; Construct a weighted undirected graph for parts with rigid contact relationships, including: Convert parts into nodes; If two parts have a rigid contact relationship, then add an edge between the corresponding nodes of the two parts; The cross-sectional area of ​​the intersection between the target point cloud data of two parts is calculated, and the weights of the edges are assigned according to the cross-sectional area to obtain a weighted undirected graph; the weights are positively correlated with the cross-sectional area.

2. The method according to claim 1, characterized in that, The step of identifying the positional relationship between pairs of parts based on candidate point cloud data includes: Add a minimum bounding cube to the candidate point cloud data; The smallest circumscribed cube is expanded according to a preset magnification factor; If the two smallest circumscribed cubes after expansion intersect, then the positional relationship of the parts corresponding to the two smallest circumscribed cubes after expansion is determined to be adjacent.

3. The method according to claim 1, characterized in that, The process of deriving target point cloud data from the candidate point cloud data of each component based on the vibration modes of the printed circuit board in the automobile includes: Identify target lines where two planes intersect in the candidate point cloud data; On the same plane, calculate the distance between the points in the candidate point cloud data and the target line; Generate random vibration signals within a preset amplitude range; The target state transition matrix is ​​obtained by calculating the product between the random vibration signal and the original state transition matrix representing the vibration transmission law of the printed circuit board in the automobile. Singular value decomposition is performed on the target state transition matrix to obtain vibration constraint parameters; The second norm of the difference between the state transition parameter and the vibration constraint parameter is taken as the vibration offset parameter; the state transition parameter is the product of the target state transition matrix and the distance. The vibration offset parameter is added to the points in the candidate point cloud data in a way that generates new points, thereby obtaining the target point cloud data.

4. The method according to any one of claims 1-3, characterized in that, The classification of high-power electric fans based on the target path and multi-order nearest neighbor subgraphs includes: Using the target path and the multi-order nearest neighbor subgraph in the dimension of the graph, the first probability of the high-power electronic fan belonging to each category is identified; The second probability of classifying the high-power electronic fan into each category is identified using the multi-order nearest neighbor subgraph in both graph and semantic dimensions. For the same category, the first probability and the second probability are merged into a third probability; The high-power electronic fan was determined to belong to the third category with the highest probability.

5. The method according to claim 4, characterized in that, The first probability of identifying the high-power electronic fan belonging to each category using the target path and the multi-order nearest neighbor subgraph in the dimension of the graph includes: The target path is input into the translation network to generate the first graph features; The multi-order nearest neighbor subgraph is input into the translation network to generate the second graph feature; The first image feature and the second image feature are concatenated and input into the classifier to identify the first probability of the high-power electronic fan belonging to each category; The second probability of identifying the high-power electronic fan's classification into various categories using the multi-order nearest neighbor subgraph in both graph and semantic dimensions includes: The description information of the part is mapped to the nodes of the multi-order nearest neighbor subgraph; If the mapping is completed, the multi-order nearest neighbor subgraph is input into the graph neural network to identify the second probability of the high-power electronic fan belonging to each category.

6. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the assembly and testing method for a high-power electronic fan printed circuit board as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the assembly and testing method for a high-power electronic fan printed circuit board as described in any one of claims 1-5.

Images