Electric power material checking result verification and supplement method and system

Abstract

The invention provides an electric power material checking result verification and supplement method and system, and belongs to the field of electric power material warehouse management. The method comprises the following steps: acquiring image data of a to-be-checked area acquired by an unmanned aerial vehicle and real-time pose data of the unmanned aerial vehicle; inputting the image data and the pose data into a trained three-dimensional coupling verification model, processing the image data through a visual identification unit, and identifying and obtaining identification information of the electric power materials; processing the real-time pose data by using a digital twin space base, determining actual space coordinates of the electric power materials, and associating the actual space coordinates to corresponding goods locations; and calling business knowledge base data, performing coupling verification on the identification information, the actual space coordinates and the associated goods allocation information with the business knowledge base data, and generating and outputting a verification result. According to the invention, the problems of single identification dimension and lack of service context verification in the prior art are solved, and the accuracy of the inventory result and the intelligent level of warehouse management are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of power material storage management technology, and in particular relates to a method and system for verifying and supplementing the results of power material inventory. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Power equipment is a crucial asset for ensuring the stable operation of the power grid. The core objective of its warehousing management is to achieve complete consistency between the records in the management system, the physical goods in the warehouse, and the geographical location of the goods. Traditional methods of power equipment inventory mainly rely on manual labor, which is not only inefficient and costly, but also poses safety risks when dealing with materials at heights, heavy or hazardous materials. Furthermore, manual operation is prone to human errors such as miscounting and omissions, making it difficult to guarantee the real-time nature and accuracy of the data.

[0004] In recent years, with the development of drone technology and computer vision, drone-based automated inventory systems have begun to be applied to the management of power material warehouses. These systems typically use drones equipped with cameras to collect warehouse videos or images, and then utilize object detection models (such as YOLO and SSD) to identify the categories of materials in the images, thereby automatically generating an inventory list. This method, to some extent, frees up manpower and improves inventory efficiency.

[0005] However, existing technologies for inventory management using drones still have the following drawbacks. First, many electrical materials contain products that look similar but differ significantly in model and specifications. Current technology relies solely on the overall shape of the materials for identification, failing to distinguish between these similar-looking but substantively different materials, leading to unreliable identification results. Second, the system cannot link identified materials to their precise locations in the digital twin warehouse, making it impossible to determine whether materials are placed in the correct locations. Furthermore, the system does not correlate or verify with the warehouse management system (WMS)'s theoretical inventory list, material master data, storage rules (and material lifecycle status), thus failing to detect anomalies in material status. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, this invention provides a method and system for supplementing the verification of power material inventory results. By constructing a three-dimensional coupled verification model of vision, space, and business, and combining high-precision nameplate recognition and spatial pose information, intelligent verification of power material inventory results can be achieved.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for supplementing the verification of power material inventory results; A supplementary method for verifying the results of an inventory of electrical materials includes: Data acquisition steps: Acquire image data of the area to be inventoried by the drone and the real-time pose data of the drone; Visual recognition step: The image data and pose data are input into the trained three-dimensional coupled verification model. The three-dimensional coupled verification model processes the image data through a visual recognition unit to identify and obtain the identification information of the power materials. Spatial positioning and association steps: The real-time pose data is processed using a digital twin spatial base to determine the actual spatial coordinates of the identified electrical materials in the three-dimensional space of the warehouse and associate them with the corresponding storage locations; Knowledge base data retrieval steps: Retrieve the business knowledge base data corresponding to the current inventory task, wherein the business knowledge base data includes at least the theoretical inventory list of the target warehouse area; Coupled verification step: Couple and verify the identification information, the actual spatial coordinates and the associated cargo location information with the business knowledge base data, and generate and output the verification result.

[0008] As a further technical solution, the data acquisition step, which involves acquiring image data of the area to be inventoried by the drone and the real-time pose data of the drone, includes: Images of the area to be inventoried are acquired using the onboard visual sensor of the drone; The drone's real-time 3D coordinates and attitude angles are obtained through a pose estimation system integrated on the drone. The image is then timestamped along with the real-time 3D coordinates and attitude angles, and uploaded to the server via a wireless communication protocol.

[0009] As a further technical solution, in the visual recognition step, the image data is processed by a visual recognition unit to identify and obtain the identification information of the power materials, including: The nameplate area is located in the image data using an object detection model; The image of the located nameplate area is corrected by perspective transformation to obtain a standard front view; Optical character recognition is performed on the corrected nameplate image, and structured key information is extracted based on the fixed field structure of the power material nameplate. The extracted key information is combined to generate a unique material identifier.

[0010] As a further technical solution, the target detection model is an improved model based on the YOLOv5 architecture, which extracts image features through the backbone network and performs multi-scale feature fusion through the feature pyramid network and the path aggregation network. The nameplate area is located by using the bounding box coordinates and confidence level of the detection head.

[0011] As a further technical solution, in the spatial positioning and association step, the real-time pose data is processed using a digital twin spatial base to determine the actual spatial coordinates of the identified electrical materials in the three-dimensional space of the warehouse, and associate them with the corresponding storage locations, including: Align the real-time pose data with the coordinate system of the digital twin space base, and calculate the actual spatial coordinates of the identified electrical materials in the digital twin space base through coordinate transformation. The spatial inclusion of the actual spatial coordinates with the predefined bounding boxes of each cargo space in the digital twin spatial base is detected. When the actual spatial coordinates are detected to fall within the spatial bounding box of a certain storage location, the electrical material is associated and bound with the storage location ID of that location.

[0012] As a further technical solution, in the coupling verification step, the coupling verification is a three-dimensional coupling verification based on vision-space-business information, specifically including: Spatial verification: Determine whether the actual spatial coordinates of the electrical materials fall within the spatial bounding box of their expected storage location; Business verification: Determine whether the identity and / or quantity of the identified power materials match the expected material information of the corresponding location in the theoretical inventory list; Status verification: Based on the business rule base in the business knowledge base, determine whether the existence status of the identified power materials conforms to the predefined business rules.

[0013] As a further technical solution, the verification result includes the location ID, material ID, verification result, confidence level, timestamp, and anomaly type.

[0014] The second aspect of the present invention provides a supplementary system for verifying the results of an inventory of electrical materials.

[0015] A supplementary system for verifying the results of an inventory of electrical materials includes: The data acquisition module is configured to acquire image data of the area to be inventoried by the drone and the real-time pose data of the drone. The visual recognition module is configured to input the image data and pose data into a trained three-dimensional coupled verification model, wherein the three-dimensional coupled verification model processes the image data through a visual recognition unit to identify and obtain the identification information of power materials. The spatial positioning and association module is configured to: process the real-time pose data using a digital twin spatial base, determine the actual spatial coordinates of the identified power materials in the three-dimensional space of the warehouse, and associate them with the corresponding storage location; The data retrieval module is configured to retrieve business knowledge base data corresponding to the current inventory task, wherein the business knowledge base data includes at least the theoretical inventory list of the target warehouse area. The coupling verification module is configured to: couple and verify the identification information, the actual spatial coordinates and the associated cargo location information with the business knowledge base data, and generate and output the verification result.

[0016] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a method for supplementing the verification of inventory results of electrical materials as described in the first aspect of the present invention.

[0017] A fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the supplementary method for verifying the results of an inventory of electrical materials as described in the first aspect of the present invention.

[0018] The above one or more technical solutions have the following beneficial effects: (1) This invention constructs a three-dimensional coupled verification model of "vision-space-business" to deeply integrate the visual recognition results of UAVs, the precise spatial pose of materials, and the business rules of the warehouse management system. It not only identifies "what is there", but also intelligently judges "whether it is in the correct position" and "whether its status meets business expectations", thereby fundamentally ensuring 100% consistency among inventory records, physical objects, and storage locations, and greatly reducing management risks caused by misalignment, omission, or abnormal status.

[0019] Addressing the issue of similar appearances and diverse models of power equipment, this invention employs specially optimized target detection, perspective transformation correction, and structured information extraction techniques on nameplates to accurately extract key information such as "model," "capacity," and "factory serial number," forming a unique equipment identifier. This method effectively overcomes the shortcomings of general models that only recognize the form but not the substance, providing a highly reliable data foundation for all subsequent verification processes and significantly improving the overall system's recognition accuracy.

[0020] (2) This invention introduces a multi-layered verification process (spatial verification, business verification, and status verification) based on a rule engine, which can automatically detect a variety of complex problems such as misplaced materials, discrepancies in inventory, violations of storage rules, and abnormal lifecycle status. This makes the inventory process no longer a simple data collection, but an intelligent reasoning process with business understanding capabilities, which greatly enhances the intelligence level of the system.

[0021] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0022] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0023] Figure 1 This is a flowchart of the method in the first embodiment.

[0024] Figure 2 This is a system structure diagram of the second embodiment. Detailed Implementation

[0025] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0026] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0027] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0028] Example 1 This embodiment discloses a supplementary method for verifying the results of an inventory of power materials; like Figure 1 As shown, a supplementary method for verifying the results of an inventory of electrical materials includes: Step S1, Data Acquisition: Acquire image data of the area to be inventoried by the UAV and the real-time pose data of the UAV.

[0029] The flight path of the drone in the warehouse area to be inventoried for electrical materials is pre-planned. The flight path covers the material storage areas corresponding to all target locations within the warehouse, and preset parameters for flight altitude, speed, and shooting interval are set to ensure that the drone's onboard camera can completely capture image data of the materials in each location. The drone executes its flight mission according to the preset flight path, capturing continuous image frames of the area to be inventoried in real time through its onboard visual sensor. Simultaneously, the drone's pose estimation system acquires its own three-dimensional pose data in a preset coordinate system in real time, whereby the three-dimensional pose data includes position coordinates and attitude parameters. The UAV establishes a communication connection with the edge server or cloud via the MQTT or RTSP protocol, and synchronously associates the real-time acquired image frame data with the corresponding three-dimensional pose data. The association is achieved through timestamp matching to ensure that each frame of image data corresponds to a unique real-time pose data of the UAV. Then, the associated image data and three-dimensional pose data are continuously uploaded to the edge server or cloud according to a preset transmission strategy.

[0030] Step S2, visual recognition: The image data and pose data are input into the trained three-dimensional coupled verification model to generate and output the verification result; wherein, the three-dimensional coupled verification model processes the image data through a visual recognition unit to identify and obtain the identification information of the power materials.

[0031] In step S21, after receiving the image data to be processed uploaded by the UAV, the visual recognition unit first performs image enhancement operations, including adjusting the image brightness and contrast based on gamma correction, removing image noise through Gaussian filtering, and optimizing image details using histogram equalization to ensure that the image meets the quality requirements of subsequent processing.

[0032] Step S22: Subsequently, the model object detection model based on the improved YOLOv5 architecture is called to detect the preprocessed image, and the predicted bounding box coordinates, material type and detection confidence of the power material nameplate in the image are output. The effective prediction results are filtered by the preset confidence threshold, and redundant bounding boxes are eliminated by non-maximum suppression to determine the unique nameplate area.

[0033] In this embodiment, the CSPDarknet53 backbone structure of YOLOv5 is retained, but an attention mechanism is added to the shallow feature layers to enhance the feature capture capability for small targets such as nameplates. Based on the original feature pyramid network and path aggregation network of YOLOv5, a new small target detection branch, P2 layer, is added. By adjusting the downsampling rate, the detection accuracy of small nameplates is improved.

[0034] By constructing a "dedicated dataset for power equipment nameplates," covering nameplate images of common power equipment such as transformers, switchgear, and cables, and labeling the nameplate bounding boxes and their respective equipment types, the model's generalization ability is improved through Mosaic data augmentation and image fusion. Furthermore, considering the high accuracy requirements for bounding box localization in nameplate detection, the original YOLOv5 CIoU loss function is replaced with the DIoU (Distance-IoU) loss function, which focuses more on the distance and overlap of the bounding box center points, reducing localization errors.

[0035] Step S23: Extract image sub-blocks from the bounding box of the located nameplate area; correct tilted and distorted nameplate sub-blocks using a perspective transformation algorithm; determine the original image coordinates by selecting four non-collinear feature points (such as nameplate corners) in the nameplate sub-blocks, set the target coordinates of the standard frontal rectangle, construct the homography matrix H and solve it; perform perspective transformation on the nameplate sub-blocks based on the homography matrix to obtain a frontal, distortion-free standard nameplate image. Specifically, select four non-collinear feature points from the original nameplate sub-block image captured by the UAV, denoted as the original point set. .

[0036] Simultaneously, four corresponding vertices of the target standard image are defined, denoted as the target point set. .

[0037] Given the original point set and the target point set, the eight parameters of the homography matrix H are solved using the least squares method. For each corresponding point... Substituting the perspective transformation formula, we obtain two linear equations after expansion. The four corresponding points generate a total of eight linear equations. By solving the system of equations, we obtain the parameters of H.

[0038] The perspective transformation is performed, and the homography matrix H obtained by solving is used to map all pixels in the original nameplate sub-block image to obtain the pixel values ​​of the target standard image: for each pixel (x, y) in the original image, its mapped coordinates in the target image are calculated using H. After the final output standard front view transformation is completed, a standard image of the nameplate that is viewed from the front and without distortion is obtained, which meets the requirements of subsequent OCR recognition.

[0039] Step S24 involves processing the standard nameplate image using a multi-stage OCR pipeline. The first stage uses the CTPN algorithm to detect text line regions in the nameplate image and outputs text line bounding boxes. The second stage uses the TrOCR model to convert the text line regions into character sequences. The third stage performs text cleaning, combining a power industry dictionary for dictionary matching correction, rule correction based on power material naming rules (such as "Model: S11-M-500" format), and context correction using semantic association between adjacent texts to correct recognition errors. The fourth stage analyzes the text structure, dividing semantic blocks based on the spatial position relationship of text lines, and extracting key-value pairs of fixed fields such as "equipment name," "model," "specification," "capacity," and "factory number."

[0040] In step S25, the visual recognition unit verifies the validity of the extracted nameplate structured key-value pairs, removes invalid or conflicting fields, and supplements the material type, detection confidence level, and image ROI (region of interest) data of the original nameplate area obtained during the detection process to form complete power material identification information. The identification information includes material type, model, specifications, capacity, factory number, identification confidence level, and nameplate image ROI. Finally, the power material identification information is used as a unique material identifier.

[0041] Step S3, Spatial Positioning and Association: The real-time pose data is processed using a digital twin spatial base to determine the actual spatial coordinates of the identified power materials in the three-dimensional space of the warehouse and associate them with the corresponding storage locations.

[0042] The system retrieves preset global coordinate system parameters from the digital twin spatial base and simultaneously extracts the drone's own coordinate system parameters carried in the drone's real-time pose data. A coordinate transformation matrix is ​​then used to convert the drone's position coordinates and attitude parameters in its real-time pose data into pose data in the warehouse's global coordinate system, thus aligning the real-time pose data with the digital twin spatial base coordinate system.

[0043] The intrinsic parameter matrix of the UAV's onboard camera is obtained, and combined with the aligned global pose data of the UAV, a camera imaging model is established. From the pixel coordinates of the power material nameplate area output by the visual recognition unit, the center pixel of the nameplate is selected as the material feature point.

[0044] Based on the principle of perspective projection, the actual spatial coordinates of the identified power materials in the digital twin spatial base are calculated. The location space data interface of the digital twin spatial base is called to obtain the predefined bounding box parameters of all target locations within the warehouse. The bounding box of each location corresponds to the minimum and maximum three-dimensional coordinates in the global coordinate system of the warehouse. An inclusion check is performed on the actual spatial coordinates of the identified power materials and the bounding box parameters of each location. If there exists one and only one location i that satisfies the above spatial inclusion condition, the unique location D of that location is extracted and associated with the unique identifier of the identified power materials, generating material-location association data, which is synchronously stored in the material location mapping table of the digital twin spatial base and the inventory association table of the business knowledge base.

[0045] Step S4, Knowledge Base Data Retrieval: Retrieve the business knowledge base data corresponding to the current inventory task. The business knowledge base data includes at least the theoretical inventory list of the target warehouse area.

[0046] The system retrieves basic information about the current power material inventory task from the management platform. This basic information includes at least a unique task identifier, target warehouse area number, inventory time range, and task priority. The system then matches the corresponding task configuration record in the "Inventory Task Index Table" of the business knowledge base using the unique task identifier, and extracts the spatial range parameters of the target warehouse area and the associated inventory data query permissions.

[0047] Based on the parsed target warehouse area number and combined with the warehouse area-inventory association rules in the business knowledge base, the query dimensions for the theoretical inventory list are determined, and these dimensions are integrated into a structured query statement. By initiating a query request to the inventory database of the knowledge base, the business knowledge base extracts records that meet the query conditions from the WMS theoretical inventory master table. Each record contains at least the location ID, the expected material unique identifier, the expected material quantity, the most recent inventory count time, and the material lifecycle status, forming a structured theoretical inventory list for the target warehouse area. Finally, the standardized theoretical inventory list is associated and bound with the unique identifier of the current inventory count task, completing the entire retrieval process.

[0048] Step S5, Coupling Verification: The identification information, the actual spatial coordinates, and the associated cargo location information are coupled and verified with the business knowledge base data to generate and output the verification result.

[0049] The coupling verification is a three-dimensional coupling verification based on vision, space, and business information, specifically including: Spatial verification determines whether the actual spatial coordinates of the electrical materials fall within the bounding box of their expected storage location. Specifically, the expected bounding box parameters and maximum three-dimensional coordinates of the associated storage location are retrieved from the digital twin spatial base. The actual spatial coordinates of the identified electrical materials are then compared with the expected bounding box parameters to determine whether the spatial inclusion condition is met. If not, an anomaly is marked as "spatial coordinates deviate from expected storage location".

[0050] The business verification process determines whether the identity and / or quantity of the identified power materials match the expected material information for the corresponding storage location in the theoretical inventory list. Specifically, it extracts the expected material information corresponding to the associated storage location from the theoretical inventory list in the business knowledge base. This expected material information includes at least a unique identifier for the expected material and the expected quantity. The power material identification information output by the visual recognition unit is compared with the expected material information. If the unique identifiers are inconsistent, an "identity mismatch" anomaly is marked; if the unique identifiers are consistent but the quantities are mismatched, an "quantity mismatch" anomaly is marked.

[0051] Status verification, based on the business rule base in the business knowledge base, determines whether the existence status of the identified power materials conforms to predefined business rules. Specifically, it retrieves predefined business rules from the business rule base in the business knowledge base. These business rules include at least material lifecycle status rules and material storage association rules. Based on these business rules, the existence status of the identified power materials is verified. If the material lifecycle status is "scrapped" or "pending inspection" but appears within the normal inventory range, it is marked as "material status violation". If the material storage violates the association rules, it is marked as "storage association violation". Finally, the results of spatial verification, business verification, and status verification are summarized to generate a structured verification result containing the location ID, material ID, verification result, verification confidence level, timestamp, and anomaly type. The verification result is then packaged according to a preset format and synchronously pushed to the digital twin screen and verification result database of the management platform to complete the entire process of three-dimensional coupled verification.

[0052] Example 2 This embodiment discloses a supplementary system for verifying the results of power material inventory; like Figure 2 As shown, a supplementary system for verifying the results of an inventory of electrical materials includes: The data acquisition module is configured to acquire image data of the area to be inventoried by the drone and the real-time pose data of the drone. The visual recognition module is configured to input the image data and pose data into a trained three-dimensional coupled verification model, wherein the three-dimensional coupled verification model processes the image data through a visual recognition unit to identify and obtain the identification information of power materials. The spatial positioning and association module is configured to: process the real-time pose data using a digital twin spatial base, determine the actual spatial coordinates of the identified power materials in the three-dimensional space of the warehouse, and associate them with the corresponding storage location; The data retrieval module is configured to retrieve business knowledge base data corresponding to the current inventory task, wherein the business knowledge base data includes at least the theoretical inventory list of the target warehouse area. The coupling verification module is configured to: couple and verify the identification information, the actual spatial coordinates and the associated cargo location information with the business knowledge base data, and generate and output the verification result. Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0053] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the supplementary method for verifying the results of an inventory of electrical materials as described in Example 1.

[0054] Example 4 The purpose of this embodiment is to provide an electronic device.

[0055] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the supplementary method for verifying the results of an inventory of electrical materials as described in Example 1.

[0056] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0057] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0058] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for verifying and supplementing the result of inventory of electric power assets, characterized by, The method comprises the following steps: Data acquisition step: obtaining image data of the area to be checked collected by the unmanned aerial vehicle and real-time pose data of the unmanned aerial vehicle; Visual recognition step: inputting the image data and pose data into a trained three-dimensional coupling verification model, processing the image data by a visual recognition unit, identifying and obtaining identification information of the power materials; Spatial positioning and association step: processing the real-time pose data by a digital twin space base, determining the actual spatial coordinates of the identified power materials in the three-dimensional space of the warehouse, and associating them to the corresponding storage locations; Knowledge base data retrieval step: retrieving business knowledge base data corresponding to the current inventory task, which at least includes the theoretical inventory list of the target storage area; Coupling verification step: coupling and verifying the identification information, the actual spatial coordinates and the associated storage location information with the business knowledge base data to generate and output the verification result.

2. The method of claim 1, wherein the power asset inventory result verification supplementing method is characterized by, In the data acquisition step, the image data of the area to be checked collected by the unmanned aerial vehicle and the real-time pose data of the unmanned aerial vehicle are obtained, comprising: Collecting the image of the area to be checked by the on-board visual sensor of the unmanned aerial vehicle; Obtaining the real-time three-dimensional coordinates and attitude angle of the unmanned aerial vehicle itself by the pose estimation system integrated on the unmanned aerial vehicle; After the image and the real-time three-dimensional coordinates and attitude angle are stamped with a unified timestamp, they are uploaded to the server through a wireless communication protocol.

3. The method of claim 1, wherein the power asset inventory result verification supplementing method is characterized by, In the visual recognition step, the image data is processed by a visual recognition unit to identify and obtain the identification information of the power materials, comprising: Using a target detection model to locate the nameplate area in the image data; Performing perspective transformation correction on the located nameplate area image to obtain a standard front view; Performing optical character recognition on the corrected nameplate image, and based on the fixed field structure of the power material nameplate, extracting structured key information; Combining the extracted key information to generate a unique material identifier.

4. The method of claim 3, wherein the power asset inventory result verification supplementing method is characterized by, The target detection model is a model improved based on the YOLOv5 architecture, which extracts image features through a backbone network and performs multi-scale feature fusion through a feature pyramid network and a path aggregation network; Using the detection head to output the boundary box coordinates and confidence of the nameplate area to complete the location of the nameplate area.

5. The method of claim 1, wherein the power asset inventory result verification supplementing method is characterized by, In the spatial positioning and association step, the real-time pose data is processed by a digital twin space base to determine the actual spatial coordinates of the identified power materials in the three-dimensional space of the warehouse, and is associated to the corresponding storage location, comprising: Aligning the real-time pose data with the coordinate system of the digital twin space base, and calculating the actual spatial coordinates of the identified power materials in the digital twin space base through coordinate transformation; Performing spatial containment detection on the actual spatial coordinates and the pre-defined space bounding box of each storage location in the digital twin space base; When the actual spatial coordinates are detected to fall within the space bounding box of a storage location, the power material is associated and bound with the storage location ID of the storage location.

6. The method of claim 1, wherein the power asset inventory result verification supplementing method is characterized by, In the coupling verification step, the coupling verification is a three-dimensional coupling verification based on visual-spatial-business information, specifically comprising: Space verification: judging whether the actual space coordinates of the power materials fall within the space bounding box of the expected storage location; Business verification: judging whether the identity and / or quantity of the identified power materials match the expected material information of the corresponding storage location in the theoretical inventory list; State verification: judging whether the existence state of the identified power materials complies with the predefined business rules based on the business rule library in the business knowledge base.

7. The method of claim 1, wherein the power asset inventory result verification supplementing method is characterized by, The verification result includes storage location ID, material ID, verification result, confidence, timestamp, and exception type.

8. A system for verifying and supplementing the results of an inventory of electrical materials, characterized in that, It comprises: a data acquisition module configured to acquire image data of a to-be-inspected area collected by a UAV and real-time pose data of the UAV; a visual recognition module configured to input the image data and pose data into a trained three-dimensional coupling verification model, which processes the image data through a visual recognition unit to identify and obtain identification information of power materials; a space positioning and association module configured to process the real-time pose data using a digital twin space base to determine the actual space coordinates of the identified power materials in the three-dimensional space of the warehouse and associate them to the corresponding storage location; a knowledge base data retrieval module configured to retrieve business knowledge base data corresponding to the current inventory task, which at least includes the theoretical inventory list of the target storage area; a coupling verification module configured to perform coupling verification on the identification information, actual space coordinates, and associated storage location information with the business knowledge base data to generate and output verification results.

9. A computer-readable storage medium having stored thereon a program, characterized in that, The program is executed by the processor to implement the steps of the power material inventory result verification and supplement method according to any one of claims 1-7.

10. An electronic device comprising a memory, a processor, and a program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the power material inventory result verification and supplement method according to any one of claims 1-7.

Images