A power distribution cabinet and a preparation method and system thereof

Abstract

The application provides a power distribution cabinet and a preparation method and system thereof, and relates to the technical field of power distribution cabinets.The method comprises the following steps: obtaining multiple photographed images of a power distribution cabinet to be detected under different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet; generating a three-dimensional simulation generated image of the power distribution cabinet by using a generative adversarial network based on the multiple photographed images of the power distribution cabinet to be detected under different ultrasonic frequencies; constructing a graph structure; determining the visual defect degree of the power distribution cabinet to be detected by processing the graph structure based on a graph autoencoder; obtaining a video of the power distribution cabinet to be detected being knocked; determining the structural deformation degree of the power distribution cabinet to be detected by using a knocking video processing model based on the video of the power distribution cabinet to be detected being knocked; and determining whether the power distribution cabinet to be detected is qualified in quality based on the visual defect degree of the power distribution cabinet to be detected and the structural deformation degree of the power distribution cabinet to be detected, so that the quality of the power distribution cabinet can be quickly and accurately evaluated.

Description

Technical Field

[0001] This invention relates to the field of power distribution cabinet technology, specifically to a power distribution cabinet and its manufacturing method and system. Background Technology

[0002] Distribution cabinets are indispensable and crucial equipment in power systems, used for distributing and controlling electrical energy. Their quality and performance directly affect the safety and reliability of the power grid. Therefore, ensuring the quality of distribution cabinets during the manufacturing process is of paramount importance. Traditional quality inspection methods mainly rely on manual inspection and limited mechanical testing. This approach is not only inefficient but also susceptible to subjective factors, making it difficult to guarantee accurate quality assessment of distribution cabinets every time.

[0003] Therefore, how to quickly and accurately assess the quality of distribution cabinets is an urgent problem to be solved. Summary of the Invention

[0004] The main technical problem this invention addresses is how to quickly and accurately assess the quality of power distribution cabinets.

[0005] According to a first aspect, the present invention provides a method for manufacturing a distribution cabinet, comprising: acquiring multiple images of a distribution cabinet to be tested at different ultrasonic frequencies and a standard three-dimensional image of a standard distribution cabinet; generating a three-dimensional simulated image of the distribution cabinet using a generative adversarial network based on the multiple images of the distribution cabinet to be tested at different ultrasonic frequencies; constructing a graph structure, the graph structure including two nodes and an edge between the two nodes, the two nodes including a detection node and a standard node, the node feature of the detection node being the three-dimensional simulated image of the distribution cabinet, the node feature of the standard node being the standard three-dimensional image of the standard distribution cabinet, and the edge between the two nodes being the similarity between the three-dimensional simulated image of the distribution cabinet and the standard three-dimensional image of the standard distribution cabinet; processing the graph structure based on a graph autoencoder to determine the visual defect degree of the distribution cabinet to be tested; acquiring a video of the distribution cabinet to be tested being struck; determining the structural deformation degree of the distribution cabinet to be tested using a striking video processing model based on the video of the distribution cabinet to be tested being struck; and determining whether the distribution cabinet to be tested is of acceptable quality based on the visual defect degree and the structural deformation degree of the distribution cabinet to be tested.

[0006] In one possible implementation, the tapping video processing model is a gated loop unit.

[0007] In one possible implementation, the graph autoencoder takes the graph structure as its input and outputs the visual defect rate of the distribution cabinet to be inspected.

[0008] In one possible implementation, determining whether the distribution cabinet under test is of acceptable quality based on the visual defects and structural deformation of the distribution cabinet under test includes:

[0009] The visual defects and structural deformation of the distribution cabinet under test are weighted and summed according to a predetermined weight to obtain the pass rate of the distribution cabinet under test; it is then determined whether the pass rate of the distribution cabinet under test is greater than the pass threshold; if the pass rate of the distribution cabinet under test is greater than the pass threshold, the distribution cabinet under test is determined to be of qualified quality; if the pass rate of the distribution cabinet under test is less than the pass threshold, the distribution cabinet under test is determined to be of unqualified quality.

[0010] According to a second aspect, the present invention provides a system for manufacturing a power distribution cabinet, comprising: a first acquisition module, used to acquire multiple images of the power distribution cabinet to be tested at different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet;

[0011] The generation module is used to generate a three-dimensional simulation image of the power distribution cabinet based on multiple images captured at different ultrasonic frequencies using a generative adversarial network.

[0012] A construction module is used to construct a graph structure, which includes two nodes and an edge between the two nodes. The two nodes include a detection node and a standard node. The node feature of the detection node is a 3D simulation image of the power distribution cabinet, and the node feature of the standard node is a standard 3D image of the standard power distribution cabinet. The edge between the two nodes is the similarity between the 3D simulation image of the power distribution cabinet and the standard 3D image of the standard power distribution cabinet.

[0013] The visual defect determination module is used to determine the visual defect degree of the power distribution cabinet to be inspected by processing the graph structure based on the graph autoencoder.

[0014] The second acquisition module is used to acquire video of the power distribution cabinet under test being struck.

[0015] The structural deformation determination module is used to determine the deformation degree of the distribution cabinet under test based on the video of the cabinet being struck using a striking video processing model.

[0016] The quality qualification determination module is used to determine whether the power distribution cabinet under test is qualified based on the visual defect degree and the structural deformation degree of the power distribution cabinet under test.

[0017] In one possible implementation, the tapping video processing model is a gated loop unit.

[0018] In one possible implementation, the graph autoencoder takes the graph structure as its input and outputs the visual defect rate of the distribution cabinet to be inspected.

[0019] In one possible implementation, the quality conformity determination module is further configured to:

[0020] The pass rate of the power distribution cabinet under test is obtained by weighting and summing the visual defects and structural deformation of the power distribution cabinet under test according to a predetermined weight.

[0021] Determine whether the pass rate of the power distribution cabinet to be tested is greater than the pass threshold;

[0022] If the pass rate of the distribution cabinet under test is greater than the pass threshold, the distribution cabinet under test is determined to be of qualified quality; if the pass rate of the distribution cabinet under test is less than the pass threshold, the distribution cabinet under test is determined to be of unqualified quality.

[0023] According to a third aspect, embodiments of the present invention provide an electronic device, comprising: a processor; a memory; and a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method described above, the method comprising: acquiring multiple captured images of a distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard distribution cabinet; generating a three-dimensional simulated image of the distribution cabinet using a generative adversarial network based on the multiple captured images of the distribution cabinet under test at different ultrasonic frequencies; and constructing a graph structure including two nodes and edges between the two nodes, the two nodes including a detection node and a standard node. The nodes are defined as follows: the node feature of the detection node is a 3D simulated image of the distribution cabinet; the node feature of the standard node is a standard 3D image of the standard distribution cabinet; the edge between two nodes represents the similarity between the 3D simulated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet; the visual defect degree of the distribution cabinet to be tested is determined by processing the graph structure based on a graph autoencoder; a video of the distribution cabinet to be tested being struck is acquired; the structural deformation degree of the distribution cabinet to be tested is determined using a striking video processing model based on the video of the distribution cabinet to be tested being struck; and the quality of the distribution cabinet to be tested is determined based on the visual defect degree and the structural deformation degree of the distribution cabinet to be tested.

[0024] According to the fourth aspect, this embodiment provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the program implements the aforementioned method for preparing a power distribution cabinet. The method includes: acquiring multiple images of a power distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet; generating a three-dimensional simulated image of the power distribution cabinet using a generative adversarial network based on the multiple images of the power distribution cabinet under test at different ultrasonic frequencies; and constructing a graph structure, the graph structure including two nodes and edges between the two nodes, the two nodes including a detection node and a standard node, and the detection node's nodes... The process involves several steps: first, generating a 3D simulated image of the distribution cabinet; second, identifying the visual defects of the distribution cabinet; third, acquiring a video of the distribution cabinet being struck; fourth, determining the structural deformation of the distribution cabinet using a striking video processing model; and finally, determining whether the distribution cabinet meets quality standards based on its visual defects and structural deformation.

[0025] This invention provides a power distribution cabinet and its fabrication method and system. The method includes acquiring multiple images of the power distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet; generating a three-dimensional simulated image of the power distribution cabinet using a generative adversarial network based on the multiple images of the power distribution cabinet under test at different ultrasonic frequencies; and constructing a graph structure, wherein the graph structure includes two nodes and edges between the two nodes, the two nodes being a detection node and a standard node, the node features of the detection node being the three-dimensional simulated image of the power distribution cabinet, and the node features of the standard node being the standard image of the standard power distribution cabinet. The method involves: generating a quasi-3D image, where the edge between two nodes represents the similarity between the 3D simulation image of the distribution cabinet and the standard 3D image of a standard distribution cabinet; processing the graph structure based on a graph autoencoder to determine the visual defects of the distribution cabinet under test; acquiring a video of the distribution cabinet under test being struck; using a striking video processing model based on the video of the distribution cabinet under test being struck to determine the structural deformation of the distribution cabinet under test; and determining whether the distribution cabinet under test is of acceptable quality based on the visual defects and the structural deformation of the distribution cabinet under test. This method enables a rapid and accurate assessment of the quality of distribution cabinets. Attached Figure Description

[0026] Figure 1 This is a schematic diagram illustrating an application scenario of a method for manufacturing a power distribution cabinet according to an embodiment of the present invention.

[0027] Figure 2 A schematic flowchart illustrating a method for manufacturing a power distribution cabinet according to an embodiment of the present invention;

[0028] Figure 3 A schematic diagram of a power distribution cabinet manufacturing system provided in an embodiment of the present invention;

[0029] Figure 4 A schematic diagram of an electronic device provided in an embodiment of the present invention;

[0030] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0031] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to the present invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0032] Figure 1 This is a schematic diagram illustrating an application scenario of a method for preparing a power distribution cabinet according to an embodiment of the present invention. Figure 1 The application scenarios of a method for manufacturing a power distribution cabinet can include a server 11, a network 12, a terminal 13, and a storage device 14.

[0033] In some embodiments, server 11 may be a single server or a group of servers. Server 11 can access information and / or data stored in terminal 13 or storage device 14 via network 12. In some embodiments, server 11 may be used to perform... Figure 2 The diagram shows a method for preparing a power distribution cabinet.

[0034] Network 12 can facilitate the exchange of information and / or data. In some embodiments, network 12 can be any form of wired or wireless network, or any combination thereof.

[0035] Terminal 13 may refer to one or more terminal devices used by a user. In some embodiments, terminal 13 may include one or more combinations of mobile devices, tablet computers, laptop computers, etc.

[0036] Storage device 14 can store data and / or instructions, for example, storage device 14 can store data instructions for a method of preparing a power distribution cabinet.

[0037] In this embodiment of the invention, the following are provided: Figure 2 The method for manufacturing a power distribution cabinet shown includes steps S1 to S7:

[0038] Step S1: Acquire multiple images of the distribution cabinet under test at different ultrasonic frequencies and a standard 3D image of a standard distribution cabinet.

[0039] Different ultrasonic frequencies can penetrate different depths of materials, thereby capturing more information about the internal structure.

[0040] The distribution cabinet generates a series of two-dimensional images at each ultrasonic frequency, which are multiple images captured at different ultrasonic frequencies.

[0041] The standard 3D image of a standard distribution cabinet is a pre-constructed, precise 3D image used for comparative analysis.

[0042] Step S2: Based on multiple images of the power distribution cabinet under different ultrasonic frequencies, a generative adversarial network is used to generate a three-dimensional simulation image of the power distribution cabinet.

[0043] A Generative Adversarial Network (GAN) consists of a generator and a discriminator. The generator attempts to create realistic data samples, while the discriminator tries to distinguish whether these generated data samples are real. They compete against each other, and through iterative training, the generator eventually becomes capable of generating highly realistic data. The input to the GAN is multiple images of the electrical distribution cabinet under different ultrasonic frequencies, and the output is a 3D simulated image of the cabinet.

[0044] The 3D simulation image is generated by a generative adversarial network based on multiple images of the distribution cabinet under different ultrasonic frequencies.

[0045] Different ultrasonic frequencies can penetrate materials to varying depths, thus capturing multi-layered information about the internal and external structure of a distribution cabinet. High-frequency ultrasound (such as 5MHz) can provide finer surface details, while low-frequency ultrasound (such as 20kHz) can penetrate deeper, revealing internal structures. By combining images from multiple frequencies, a more comprehensive picture can be obtained.

[0046] The 3D simulation image of the distribution cabinet provides a more intuitive and comprehensive visual reference, which helps to discover subtle defects or anomalies, such as cracks and uneven thickness, which may not be easily detected in 2D images.

[0047] Step S3: Construct a graph structure, which includes two nodes and the edges between the two nodes. The two nodes include a detection node and a standard node. The node feature of the detection node is the 3D simulation generated image of the distribution cabinet, and the node feature of the standard node is the standard 3D image of the standard distribution cabinet. The edge between the two nodes is the similarity between the 3D simulation generated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet.

[0048] A graph structure is a data structure consisting of vertices and edges. Each node represents an entity or state, while edges represent the relationships between these entities.

[0049] The graph structure contains two nodes—the detection node and the standard node—and the edges between them.

[0050] In some embodiments, a deep neural network model can be used to determine the similarity between a 3D simulated image of a distribution cabinet and a standard 3D image of a standard distribution cabinet. The input to the deep neural network model is the 3D simulated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet, and the output of the deep neural network model is the similarity between the 3D simulated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet.

[0051] Step S4: Process the graph structure based on the graph autoencoder to determine the visual defect degree of the distribution cabinet to be inspected.

[0052] A Graph Autoencoder (GAE) is a deep learning model specifically designed for processing graph-structured data. GAEs capture complex patterns of relationships between nodes and reconstruct the input graph structure. A graph structure consists of two types of nodes—detected nodes and standard nodes—and the edges between them. Each node represents a specific object or state, while the edges reflect the similarity between these two objects. Graph structures visually demonstrate the relationships between different images and their impact on quality, facilitating a more comprehensive and in-depth understanding of the interrelationships between images. By calculating the similarity score (e.g., 0.85 representing 85% similarity) between a 3D simulation image of a power distribution cabinet and a standard 3D image, and using this score as an edge, the difference between the two can be accurately measured. GAEs consider not only information from individual nodes but also the relationships between nodes (i.e., edge information). This allows GAEs to utilize both a global perspective (overall similarity) and local details (specific defect locations) for evaluation.

[0053] The graph autoencoder takes the graph structure as input and outputs the visual defect rate of the distribution cabinet under test. The visual defect rate quantifies the degree of defects in the appearance and structure of the distribution cabinet, and is a value between 0 and 1. The graph autoencoder can extract valuable features from the graph structure to quantify the visual defect rate.

[0054] Step S5: Obtain video of the power distribution cabinet under test being struck.

[0055] The video of the distribution cabinet under test being struck is a video recording the dynamic changes of the distribution cabinet under stress.

[0056] Step S6: Determine the structural deformation degree of the power distribution cabinet under test using a knocking video processing model based on the video of the cabinet being knocked.

[0057] The impact video processing model is a gated loop unit. The input to the impact video processing model is the video of the distribution cabinet being impacted, and the output is the structural deformation degree of the distribution cabinet. The impact video is essentially a continuous time-series dataset, where each frame is correlated with the previous frame. The gated loop unit can effectively capture this temporal correlation, thereby understanding the behavioral changes of the distribution cabinet throughout the impact process.

[0058] Gated Recurrent Units (GRUs) can effectively remember long-term information and selectively update or reset the state, making them ideal for processing time-series data such as video.

[0059] In some embodiments, the tapping video processing model includes a segmentation layer, a sound stability analysis layer, a visual distortion analysis layer, and a structural distortion judgment layer. The input to the segmentation layer is the tapping video processing model, and the output of the segmentation layer is a tapping audio sequence and a tapping image sequence. The input to the sound stability analysis layer is the tapping audio sequence, and the output of the sound stability analysis layer is the stability of the tapping sound in the tapping audio sequence. The input to the visual distortion analysis layer is the tapping image sequence, and the output of the visual distortion analysis layer is the deformation degree of the distribution cabinet under test during the tapping process. The input to the structural distortion judgment layer is the stability of the tapping sound in the tapping audio sequence and the deformation degree of the distribution cabinet under test during the tapping process, and the output of the structural distortion judgment layer is the structural deformation degree of the distribution cabinet under test.

[0060] Step S7: Determine whether the power distribution cabinet under test is of acceptable quality based on the visual defects and structural deformation of the power distribution cabinet under test.

[0061] In some embodiments, the visual defects and structural deformation of the distribution cabinet under test can be weighted and summed according to a predetermined weight to obtain the pass rate of the distribution cabinet under test; it is then determined whether the pass rate of the distribution cabinet under test is greater than the pass threshold; if the pass rate of the distribution cabinet under test is greater than the pass threshold, the distribution cabinet under test is determined to be of qualified quality; if the pass rate of the distribution cabinet under test is less than the pass threshold, the distribution cabinet under test is determined to be of unqualified quality.

[0062] Based on the same inventive concept Figure 3 This is a schematic diagram of a distribution cabinet manufacturing system provided in an embodiment of the present invention. The distribution cabinet manufacturing system includes:

[0063] The first acquisition module 31 is used to acquire multiple images of the power distribution cabinet under different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet.

[0064] The generation module 32 is used to generate a three-dimensional simulation image of the power distribution cabinet based on multiple captured images of the power distribution cabinet under different ultrasonic frequencies using a generative adversarial network.

[0065] The construction module 33 is used to construct a graph structure, which includes two nodes and an edge between the two nodes. The two nodes include a detection node and a standard node. The node feature of the detection node is a three-dimensional simulation image of the distribution cabinet, and the node feature of the standard node is a standard three-dimensional image of the standard distribution cabinet. The edge between the two nodes is the similarity between the three-dimensional simulation image of the distribution cabinet and the standard three-dimensional image of the standard distribution cabinet.

[0066] The visual defect determination module 34 is used to process the graph structure based on the graph autoencoder to determine the visual defect degree of the power distribution cabinet to be inspected.

[0067] The second acquisition module 35 is used to acquire video of the power distribution cabinet under test being struck.

[0068] The structural deformation determination module 36 is used to determine the deformation degree of the distribution cabinet under test based on the video of the distribution cabinet under test being hit using a hit video processing model.

[0069] The quality qualification determination module 37 is used to determine whether the power distribution cabinet under test is qualified based on the visual defect degree and the structural deformation degree of the power distribution cabinet under test.

[0070] Based on the same inventive concept, embodiments of the present invention provide an electronic device, such as... Figure 4 As shown, it includes:

[0071] The system includes: a processor 41; a memory 42; and a computer program; wherein the computer program is stored in the memory 42 and configured to be executed by the processor 41 to implement a method for manufacturing a power distribution cabinet as described above, the method comprising: acquiring multiple images of a power distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet; generating a three-dimensional simulated image of the power distribution cabinet using a generative adversarial network based on the multiple images of the power distribution cabinet under test at different ultrasonic frequencies; and constructing a graph structure, the graph structure including two nodes and edges between the two nodes, the two nodes including a detection node and a standard node. The node features of the detection node are the 3D simulated image of the distribution cabinet, and the node features of the standard node are the standard 3D image of the standard distribution cabinet. The edge between two nodes represents the similarity between the 3D simulated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet. The visual defect degree of the distribution cabinet to be tested is determined by processing the graph structure based on a graph autoencoder. A video of the distribution cabinet to be tested being struck is acquired. Based on the video of the distribution cabinet to be tested being struck, a striking video processing model is used to determine the structural deformation degree of the distribution cabinet to be tested. Based on the visual defect degree and the structural deformation degree of the distribution cabinet to be tested, it is determined whether the distribution cabinet to be tested is of acceptable quality.

[0072] Based on the same inventive concept, this embodiment provides a computer-readable storage medium storing a computer program that, when executed by processor 41, implements the aforementioned method for preparing a power distribution cabinet. The method includes: acquiring multiple images of a power distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard power distribution cabinet; generating a three-dimensional simulated image of the power distribution cabinet using a generative adversarial network based on the multiple images of the power distribution cabinet under test at different ultrasonic frequencies; and constructing a graph structure, the graph structure including two nodes and edges between the two nodes, the two nodes including a detection node and a standard node, wherein the detection node... The node features are generated from the 3D simulation image of the distribution cabinet, and the node features of the standard node are generated from the standard 3D image of the standard distribution cabinet. The edge between two nodes represents the similarity between the 3D simulation image of the distribution cabinet and the standard 3D image of the standard distribution cabinet. The graph structure is processed based on a graph autoencoder to determine the visual defect degree of the distribution cabinet to be tested. A video of the distribution cabinet to be tested being struck is acquired. Based on the video of the distribution cabinet to be tested being struck, a striking video processing model is used to determine the structural deformation degree of the distribution cabinet to be tested. Based on the visual defect degree and the structural deformation degree of the distribution cabinet to be tested, it is determined whether the distribution cabinet to be tested is of acceptable quality.

[0073] The method for manufacturing a power distribution cabinet provided in this application embodiment can be applied to terminal devices (such as mobile phones), tablet computers, laptops, ultra-mobile personal computers (UMPCs), handheld computers, netbooks, personal digital assistants (PDAs), wearable devices (such as smartwatches, smart glasses, or smart helmets), augmented reality (AR) / virtual reality (VR) devices, smart home devices, in-vehicle computers, and other electronic devices. This application embodiment does not impose any limitations on this.

[0074] Taking mobile phone 100 as an example of the aforementioned electronic devices, Figure 5 A structural schematic diagram of mobile phone 100 is shown.

[0075] like Figure 5 As shown, the mobile phone 100 may include a processing module 110, an external memory interface 120, an internal memory 121, a universal serial bus (USB) interface 130, a charging management module 140, a power management module 141, a battery 142, an antenna 1, an antenna 2, a mobile communication module 150, a wireless communication module 160, an audio module 170, a speaker 170A, a receiver 170B, a microphone 170C, a headphone jack 170D, a sensor module 180, buttons 190, a motor 191, an indicator 192, a camera 193, a display screen 194, and a subscriber identification module (SIM) card interface 195, etc.

[0076] The processing module 110 may include one or more processing units, such as an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). These different processing units may be independent devices or integrated into one or more processors.

[0077] The controller can be the nerve center and command center of the mobile phone 100, acting as the decision-maker that directs the various components of the mobile phone 100 to work in coordination according to instructions. The controller can generate operation control signals based on the instruction operation code and timing signals to complete the control of fetching and executing instructions.

[0078] The application processor can run the operating system of the mobile phone 100, which manages the hardware and software resources of the mobile phone 100. This includes managing and configuring memory, determining the priority of system resource allocation, managing the file system, and managing drivers. The operating system can also provide a user interface for interacting with the system. Various types of software can be installed within the operating system, such as drivers and applications (Apps). For example, the operating system of the mobile phone 100 could be Android, Linux, or another operating system.

[0079] The processing module 110 may also include a memory for storing instructions and data. In some embodiments, the memory in the processing module 110 is a cache memory. This memory can store instructions or data that the processing module 110 has just used or that are used repeatedly. If the processing module 110 needs to use the instruction or data again, it can directly retrieve it from the memory. This avoids repeated accesses, reduces the waiting time of the processing module 110, and thus improves the efficiency of the system.

[0080] In some embodiments, the processing module 110 may include one or more interfaces. Interfaces may include an inter-integrated circuit (I2C) interface, an inter-integrated circuit sound (I2S) interface, a pulse code modulation (PCM) interface, a universal asynchronous receiver / transmitter (UART) interface, a mobile industry processor interface (MIPI), a general-purpose input / output (GPIO) interface, a subscriber identity module (SIM) interface, and / or a universal serial bus (USB) interface, etc.

[0081] The processing module 110 can be used to: acquire multiple images of the distribution cabinet under test at different ultrasonic frequencies and a standard 3D image of a standard distribution cabinet; generate a 3D simulated image of the distribution cabinet using a generative adversarial network based on the multiple images of the distribution cabinet under test at different ultrasonic frequencies; construct a graph structure, the graph structure including two nodes and edges between the two nodes, the two nodes including a detection node and a standard node, the node feature of the detection node being the 3D simulated image of the distribution cabinet, the node feature of the standard node being the standard 3D image of the standard distribution cabinet, and the edge between the two nodes being the similarity between the 3D simulated image of the distribution cabinet and the standard 3D image of the standard distribution cabinet; process the graph structure based on a graph autoencoder to determine the visual defect degree of the distribution cabinet under test; acquire a video of the distribution cabinet under test being struck; determine the structural deformation degree of the distribution cabinet under test using a striking video processing model based on the video of the distribution cabinet under test being struck; and determine whether the distribution cabinet under test is of acceptable quality based on the visual defect degree and the structural deformation degree of the distribution cabinet under test.

[0082] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0083] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for manufacturing a power distribution cabinet, characterized in that, include: The system acquires multiple images of the distribution cabinet under test at different ultrasonic frequencies and a standard 3D image of a standard distribution cabinet. The standard 3D image of the standard distribution cabinet is a pre-constructed, precise 3D image. Based on multiple images of the power distribution cabinet under different ultrasonic frequencies, a three-dimensional simulation image of the power distribution cabinet is generated using a generative adversarial network. A graph structure is constructed, which includes two nodes and an edge between the two nodes. The two nodes include a detection node and a standard node. The node feature of the detection node is a 3D simulation image of the power distribution cabinet, and the node feature of the standard node is a standard 3D image of the standard power distribution cabinet. The edge between the two nodes is the similarity between the 3D simulation image of the power distribution cabinet and the standard 3D image of the standard power distribution cabinet. The visual defect level of the power distribution cabinet to be inspected is determined by processing the graph structure based on the graph autoencoder. Obtain video footage of the electrical distribution cabinet under test being struck; Based on the video of the power distribution cabinet being struck, a striking video processing model is used to determine the structural deformation of the power distribution cabinet. The striking video processing model includes a segmentation layer, a sound stability analysis layer, a visual deformation analysis layer, and a structural deformation judgment layer. The input of the segmentation layer is the striking video processing model, and the output of the segmentation layer is the striking audio sequence and the striking image sequence. The input of the sound stability analysis layer is the striking audio sequence, and the output of the sound stability analysis layer is the stability of the striking sound in the striking audio sequence. The input of the visual deformation analysis layer is the striking image sequence, and the output of the visual deformation analysis layer is the deformation of the power distribution cabinet under test during the striking process. The input of the structural deformation judgment layer is the stability of the striking sound in the striking audio sequence and the deformation of the power distribution cabinet under test during the striking process, and the output of the structural deformation judgment layer is the structural deformation of the power distribution cabinet under test. The quality of the power distribution cabinet under test is determined based on its visual defects and structural deformation.

2. The method for manufacturing a power distribution cabinet as described in claim 1, characterized in that, The tapping video processing model is a gated loop unit.

3. The method for manufacturing a distribution cabinet as described in claim 1, characterized in that, The input of the graph autoencoder is the graph structure, and the output of the graph autoencoder is the visual defect degree of the power distribution cabinet to be inspected.

4. The method for manufacturing a power distribution cabinet as described in claim 1, characterized in that, The process of determining whether the distribution cabinet under test is qualified based on the visual defects and structural deformation of the distribution cabinet under test includes: The pass rate of the power distribution cabinet under test is obtained by weighting and summing the visual defects and structural deformation of the power distribution cabinet under test according to a predetermined weight. Determine whether the pass rate of the power distribution cabinet to be tested is greater than the pass threshold; If the pass rate of the distribution cabinet under test is greater than the pass threshold, the distribution cabinet under test is determined to be of qualified quality; if the pass rate of the distribution cabinet under test is less than the pass threshold, the distribution cabinet under test is determined to be of unqualified quality.

5. A system for manufacturing a power distribution cabinet, characterized in that, include: The first acquisition module is used to acquire multiple images of the distribution cabinet under test at different ultrasonic frequencies and a standard three-dimensional image of a standard distribution cabinet. The standard three-dimensional image of the standard distribution cabinet is a precise three-dimensional image that has been artificially constructed in advance. The generation module is used to generate a three-dimensional simulation image of the power distribution cabinet based on multiple images captured at different ultrasonic frequencies using a generative adversarial network. A construction module is used to construct a graph structure, which includes two nodes and an edge between the two nodes. The two nodes include a detection node and a standard node. The node feature of the detection node is a 3D simulation image of the power distribution cabinet, and the node feature of the standard node is a standard 3D image of the standard power distribution cabinet. The edge between the two nodes is the similarity between the 3D simulation image of the power distribution cabinet and the standard 3D image of the standard power distribution cabinet. The visual defect determination module is used to determine the visual defect degree of the power distribution cabinet to be inspected by processing the graph structure based on the graph autoencoder. The second acquisition module is used to acquire video of the power distribution cabinet under test being struck. The structural deformation determination module is used to determine the deformation degree of the distribution cabinet under test based on the video of the distribution cabinet being struck, using a striking video processing model. The striking video processing model includes a segmentation layer, a sound stability analysis layer, a visual deformation analysis layer, and a structural deformation judgment layer. The input of the segmentation layer is the striking video processing model, and the output of the segmentation layer is a striking audio sequence and a striking image sequence. The input of the sound stability analysis layer is the striking audio sequence, and the output of the sound stability analysis layer is the stability of the striking sound in the striking audio sequence. The input of the visual deformation analysis layer is the striking image sequence, and the output of the visual deformation analysis layer is the deformation degree of the distribution cabinet under test during the striking process. The input of the structural deformation judgment layer is the stability of the striking sound in the striking audio sequence and the deformation degree of the distribution cabinet under test during the striking process, and the output of the structural deformation judgment layer is the structural deformation degree of the distribution cabinet under test. The quality qualification determination module is used to determine whether the power distribution cabinet under test is qualified based on the visual defect degree and the structural deformation degree of the power distribution cabinet under test.

6. The distribution cabinet manufacturing system as described in claim 5, characterized in that, The tapping video processing model is a gated loop unit.

7. The distribution cabinet manufacturing system as described in claim 5, characterized in that, The input of the graph autoencoder is the graph structure, and the output of the graph autoencoder is the visual defect degree of the power distribution cabinet to be inspected.

8. The distribution cabinet manufacturing system as described in claim 6, characterized in that, The quality conformity determination module is also used for: The pass rate of the power distribution cabinet under test is obtained by weighting and summing the visual defects and structural deformation of the power distribution cabinet under test according to a predetermined weight. Determine whether the pass rate of the power distribution cabinet to be tested is greater than the pass threshold; If the pass rate of the distribution cabinet under test is greater than the pass threshold, the distribution cabinet under test is determined to be of qualified quality; if the pass rate of the distribution cabinet under test is less than the pass threshold, the distribution cabinet under test is determined to be of unqualified quality.

9. An electronic device, characterized in that, include: processor; Memory; And a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor to implement a method for preparing a power distribution cabinet as described in any one of claims 1 to 4.

10. A power distribution cabinet prepared by the method of preparing a power distribution cabinet according to claim 1.

Images