Method and system for identifying defects of overhead distribution line hardware fittings

Abstract

The invention discloses a method and a system for identifying defects of overhead distribution line hardware fittings, and belongs to the technical field of intelligent inspection of power distribution networks. The method comprises the following steps: receiving inspection data of an inspection terminal on overhead distribution line hardware, classifying the inspection data to obtain classified data, and preprocessing the classified data to obtain target data; identifying the target data by using a pre-constructed defect identification model to obtain an identification result; and determining the hardware defect of the overhead distribution line based on the identification result. The method can determine the defect category, solve the problem of poor identification effect of different types of defects in similar forms, realize large-scale expansion of features in a small-scale sample defect identification model training construction process, solve the problems of insufficient small-scale sample defect features and poor algorithm effect, and also solve the problem of poor algorithm effect caused by different equipment terminals. And therefore, the problem of poor model matching performance is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent inspection of distribution networks, and more specifically, to a method and system for identifying defects in fittings of overhead distribution lines. Background Art

[0002] China has a vast territory and a large base of overhead distribution lines. With the development of the economy and society, the scale of overhead distribution lines is constantly increasing, and the inspection tasks of distribution networks are becoming increasingly heavy, urgently requiring modern operation and maintenance methods. The application of automated inspection equipment such as unmanned aerial vehicles, vehicle-mounted devices, and handheld devices has greatly enhanced the efficiency of overhead distribution line maintenance. However, with the application of these automated equipment, the number of inspection images of distribution networks has increased explosively, and the relevant defects in the inspection images still need to be manually checked and verified to determine the defect locations, resulting in low efficiency. Currently, relevant object detection algorithms have begun to be applied to the analysis of inspection images of overhead distribution lines, greatly improving the efficiency of equipment defect detection and location in distribution network inspection images. However, there are still the following problems:

[0003] (1) For different types of defects with similar shapes, the recognition effect of the algorithm model is poor: There are a large number of different types of defects with similar appearances in the inspection images of overhead distribution lines. Commonly used template matching and object detection algorithms (such as the yolov1-v5 series) cannot effectively achieve accurate recognition of different defects under similar appearances, resulting in false detection and missed detection.

[0004] (2) Poor algorithm convergence under small-scale samples: Some equipment defects in overhead distribution lines will have a serious impact on the normal operation of the line, so they receive high attention, but the defect occurrence frequency is low, resulting in a small sample size. Currently, existing algorithms applicable to the analysis of distribution network inspection images need to be trained with large-scale samples, and have poor convergence for small-scale sample defects and poor recognition effects.

[0005] (3) Poor model matching under multiple devices and complex environments: The inspection environment of overhead distribution lines is complex and changeable, and there are various inspection devices. The robustness of the same algorithm varies in different environments, and there is a certain loss of accuracy due to model conversion and other problems. The recognition effects of the same algorithm in different inspection devices are also uneven. Summary of the Invention

[0006] In view of the above problems, the present invention proposes a method for identifying defects in fittings of overhead distribution lines, including:

[0007] Receiving the inspection data of the fittings of the overhead distribution line from the inspection terminal, classifying the inspection data to obtain classified data, and preprocessing the classified data to obtain target data;

[0008] Use a pre - constructed defect recognition model to identify the target data and obtain the recognition result;

[0009] Based on the recognition result, determine the fitting defects of the overhead distribution line.

[0010] Optionally, the inspection terminal includes at least one of the following: an unmanned aerial vehicle (UAV) terminal, a vehicle - mounted device terminal, and a handheld terminal.

[0011] Optionally, the inspection data includes at least one of the following: the inspection data of the UAV terminal, the inspection data of the vehicle - mounted device terminal, and the inspection data of the handheld terminal.

[0012] Optionally, pre - processing the classification data includes:

[0013] Clean the classification data and perform standardization processing after cleaning.

[0014] Optionally, the method further includes: after pre - processing the classification data, store the pre - processed target data in a standardized database.

[0015] Optionally, pre - constructing a defect recognition model includes:

[0016] Establish a prior knowledge relationship graph of the fittings of the overhead distribution line that combines the position relationship and the feature distribution, batch - simulate and generate small samples, fuse the small samples with the target data stored in the standardized database to generate fused data, and use a generator comparison network to train the fused data based on the prior knowledge relationship graph of the fittings of the overhead distribution line to generate a defect recognition model.

[0017] On the other hand, the present invention also proposes a system for identifying the defects of the fittings of the overhead distribution line, including:

[0018] A network communication and data management module, configured to receive the inspection data of the fittings of the overhead distribution line from the inspection terminal, classify the inspection data to obtain classification data, and pre - process the classification data to obtain target data;

[0019] A model construction and optimization module, configured to use a pre - constructed defect recognition model to identify the target data and obtain the recognition result;

[0020] An intelligent analysis module, based on the recognition result, determines the fitting defects of the overhead distribution line.

[0021] Optionally, the inspection terminal includes at least one of the following: an unmanned aerial vehicle (UAV) terminal, a vehicle - mounted device terminal, and a handheld terminal.

[0022] Optionally, the inspection data includes at least one of the following: inspection data of the UAV terminal, inspection data of the vehicle-mounted device terminal, and inspection data of the handheld terminal.

[0023] Optionally, preprocessing the classified data includes:

[0024] Cleaning the classified data and performing standardization processing after cleaning.

[0025] Optionally, the network communication and data management module is further configured to: after preprocessing the classified data, store the preprocessed target data in a standardized database.

[0026] Optionally, pre-constructing a defect recognition model includes:

[0027] Establishing a prior knowledge relationship graph of overhead distribution line fittings that combines position relationships and feature distributions, batch-simulating and generating small samples, fusing the small samples with the target data stored in the standardized database to generate fused data, and using a generator comparison network to train the fused data based on the prior knowledge relationship graph of overhead distribution line fittings to generate a defect recognition model.

[0028] Optionally, the model construction optimization module is further configured to optimize the defect recognition model.

[0029] Optionally, the system further includes: a model adaptation and transformation module for transforming and adapting the defect recognition model for different inspection terminal architectures and pushing it.

[0030] On the other hand, the present invention also provides a computing device, including: one or more processors;

[0031] The processor is configured to execute one or more programs;

[0032] When the one or more programs are executed by the one or more processors, the method as described above is implemented.

[0033] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, the method as described above is implemented.

[0034] Compared with the prior art, the beneficial effects of the present invention are:

[0035] The present invention provides a method for identifying defects in fittings for overhead distribution lines, including: receiving inspection data of fittings for overhead distribution lines from an inspection terminal, classifying the inspection data to obtain classified data, and preprocessing the classified data to obtain target data; using a pre-constructed defect identification model to identify the target data to obtain an identification result; and determining the defects in the fittings of the overhead distribution line based on the identification result. The present invention can determine the defect category, solve the problem of poor identification effect for different types of defects with similar shapes, achieve the large-scale expansion of features during the training and construction of a defect identification model for small-scale samples, solve the problem of insufficient defect features and poor algorithm effect for small-scale samples, and also solve the problem of poor model matching caused by different device terminals. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flowchart of the method of the present invention;

[0037] Figure 2 is a schematic diagram of the equipment position relationship map of the method of the present invention;

[0038] Figure 3 is a schematic diagram of the feature position relationship map of the method of the present invention;

[0039] Figure 4 is a schematic diagram of the model application process of the method of the present invention;

[0040] Figure 5 is a schematic diagram of the feature expansion network of the method of the present invention;

[0041] Figure 6 is a schematic diagram of the platform for identifying defects in fittings of overhead lines of the method of the present invention;

[0042] Figure 7 is a structural diagram of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] Now, exemplary embodiments of the present invention will be described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely, and to fully convey the scope of the present invention to those skilled in the art. The terms in the exemplary embodiments shown in the drawings are not intended to limit the present invention. In the drawings, the same unit / element is denoted by the same reference numeral.

[0044] Unless otherwise specified, the terms used herein (including technical terms) have the ordinary meaning understood by those skilled in the relevant technical field. Additionally, it can be understood that terms defined in commonly used dictionaries should be construed as having a meaning consistent with the context of their relevant fields, and should not be construed as having an idealized or overly formal meaning.

[0045] Embodiment 1:

[0046] The present invention proposes a method for identifying defects in fittings of overhead distribution lines, as Figure 1 shown, including:

[0047] Step 1: Receive the inspection data of the fittings of the overhead distribution line from the inspection terminal, classify the inspection data to obtain classified data, and preprocess the classified data to obtain target data;

[0048] Step 2: Use a pre-constructed defect identification model to identify the target data and obtain an identification result;

[0049] Step 3: Based on the identification result, determine the defects of the fittings of the overhead distribution line.

[0050] Among them, the inspection terminal includes at least one of the following: an unmanned aerial vehicle (UAV) terminal, a vehicle-mounted device terminal, and a handheld terminal.

[0051] Among them, the inspection data includes at least one of the following: the inspection data of the UAV terminal, the inspection data of the vehicle-mounted device terminal, and the inspection data of the handheld terminal.

[0052] Among them, preprocessing the classified data includes:

[0053] Clean the classified data, and perform standardization processing after cleaning.

[0054] Among them, the method further includes: after preprocessing the classified data, storing the preprocessed target data in a standardized database.

[0055] Among them, pre-constructing a defect identification model includes:

[0056] Establish a prior knowledge relationship graph of the fittings of the overhead distribution line that combines the position relationship and the feature distribution, batch-simulate and generate small samples, fuse the small samples with the target data stored in the standardized database to generate fused data, and use a generator comparison network to train the fused data based on the prior knowledge relationship graph of the fittings of the overhead distribution line to generate a defect identification model.

[0057] For the construction of the model, the present invention designs a method for discriminating similar form defects of a multi-object relationship graph, which combines the position relationship (such as Figure 2as shown) and the prior knowledge relationship graph of the overhead line fittings in the distribution network with the feature distribution (such as Figure 3 as shown), by comprehensively analyzing and judging various equipment features and relationships around the defects in the image, the defect category is determined, and the problem of poor recognition effect of different types of defects with similar forms is solved;

[0058] A small sample batch simulation generation and feature scale expansion method is also designed. By three-dimensional mapping to simulate the spatial features, color features and morphological features of equipment defects in the samples, relatively realistic equipment defects are generated, and then the equipment defects are fused with the existing image background to realize the batch simulation generation of small samples. A generator comparison network is added to scale the features during the training and construction process of the small-scale sample defect recognition model (the application process is as Figure 4 as shown), (such as Figure 5 as shown), and the problem of insufficient defect features of small-scale samples and poor algorithm effect is solved; A distribution network multi-source device data analysis platform is designed to connect the data of inspection devices such as drones, vehicle-mounted devices and handheld terminals to the data analysis platform to realize unified data management and analysis, and solve the problem of poor model matching caused by different device terminals.

[0059] A distribution network multi-source device data analysis platform (such as Figure 6 as shown) is also designed, including:

[0060] 1) Network communication module, which receives the inspection data transmitted by the inspection terminals (such as drones, vehicle-mounted devices, handheld terminals, etc.), collects and uploads the data to be analyzed of terminal devices such as drones, vehicle-mounted devices and handheld terminals. The relevant terminal devices can upload by themselves, or the data analysis platform can manage and upload various devices regularly and uniformly.

[0061] 2) Data management module, which classifies the received data according to the device type, such as drone inspection data, vehicle-mounted device inspection data, etc., and classifies, cleans and standardizes the data according to the model construction requirements to form a standardized database.

[0062] 3) Model construction and optimization module, which contains model construction and optimization algorithms, and can call the inspection data standard database to carry out the construction of the distribution network overhead line fitting defect recognition model and the independent optimization training of the existing model.

[0063] 4) Model adaptation and conversion module, which is mainly responsible for converting and adapting and pushing the existing models to different device (drone, vehicle-mounted device, handheld terminal, etc.) architectures.

[0064] 5) Intelligent analysis module, which is responsible for defect analysis of the received inspection data. By calling the existing models on the platform, as a cloud platform, it performs secondary analysis on the data transmitted back from the terminal device side, and generates an analysis report by integrating the analysis results on the terminal device side

[0065] 6) Task feedback module, which returns the analysis situation to the inspection terminal for guiding on-site inspection operations.

[0066] Embodiment 2:

[0067] The present invention also proposes a system 200 for identifying defects in fittings of overhead distribution lines, as Figure 7 shown, including:

[0068] Network communication and data management module 201, which is used to receive the inspection data of the fittings of the overhead distribution line from the inspection terminal, classify the inspection data, obtain classified data, and preprocess the classified data to obtain target data;

[0069] Model construction and optimization module 202, which is used to use a pre-constructed defect identification model to identify the target data and obtain an identification result;

[0070] Intelligent analysis module 203, which determines the defects of the fittings of the overhead distribution line based on the identification result.

[0071] Among them, the inspection terminal includes at least one of the following: unmanned aerial vehicle (UAV) terminal, vehicle-mounted equipment terminal, and handheld terminal.

[0072] Among them, the inspection data includes at least one of the following: inspection data of the UAV terminal, inspection data of the vehicle-mounted equipment terminal, and inspection data of the handheld terminal.

[0073] Among them, preprocessing the classified data includes:

[0074] Cleaning the classified data and performing standardization processing after cleaning.

[0075] Among them, the network communication and data management module is also used to: after preprocessing the classified data, store the preprocessed target data in a standardized database.

[0076] Among them, pre-constructing a defect identification model includes:

[0077] Establishing a prior knowledge relationship graph of fittings of overhead distribution lines that combines position relationships and feature distributions, batch-simulating and generating small samples, fusing the small samples with the target data stored in the standardized database to generate fused data, using a generator comparison network, and training the fused data based on the prior knowledge relationship graph of fittings of overhead distribution lines to generate a defect identification model.

[0078] Among them, the model construction and optimization module is also used to optimize the defect identification model.

[0079] Among them, the system further includes: a model adaptation and transformation module, which is used to transform and adapt and push the defect recognition model to different inspection terminal architectures.

[0080] The present invention can determine the defect category, solve the problem of poor recognition effect of different types of defects with similar shapes, realize the large-scale expansion of features during the training and construction of the defect recognition model with small-scale samples, solve the problem of insufficient defect features and poor algorithm effect with small-scale samples, and can also solve the problem of poor model matching caused by different device terminals.

[0081] Embodiment 3:

[0082] Based on the same inventive concept, the present invention further provides a computer device, which includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method process or corresponding function, so as to implement the steps of the method in the above embodiment.

[0083] Embodiment 4:

[0084] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device, used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. Moreover, in this storage space, one or more instructions suitable for being loaded and executed by the processor are also stored. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the steps of the method in the above embodiments.

[0085] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes. The solutions in the embodiments of the present invention can be implemented using various computer languages. For example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0086] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0087] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the procedures Figure 1 or blocks. Figure 1 The functions specified in one or more of the procedures

[0088] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the procedures Figure 1 or blocks. Figure 1 The functions specified in one or more of the blocks.

[0089] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A method for identifying defects in overhead distribution line hardware, characterized in that: include: Receiving inspection data of overhead power distribution line fittings from an inspection terminal, classifying the inspection data, obtaining classified data, and preprocessing the classified data to obtain target data; Using a pre-built defect recognition model, the target data is recognized to obtain a recognition result; Based on the identification result, hardware defects of the overhead distribution line are determined.

2. The method according to claim 1, characterized in that The inspection terminal includes at least one of the following: an unmanned aerial vehicle terminal, a vehicle-mounted device terminal and a handheld terminal.

3. The method according to claim 1, characterized in that The inspection data includes at least one of the following: inspection data of a drone terminal, inspection data of a vehicle-mounted device terminal, and inspection data of a handheld terminal.

4. The method according to claim 1, characterized in that: The preprocessing of the classified data comprises: The classified data is cleaned and then standardized.

5. The method according to claim 1, characterized in that: The method further comprises: after preprocessing the classified data, storing the preprocessed target data in a standardized database.

6. The method according to claim 1, characterized in that Pre-built defect identification models, including: A prior knowledge relationship map of overhead distribution line fittings that integrates positional relationships and feature distribution is established, small samples are generated by batch simulation, the small samples are fused with target data stored in a standardized database to generate fused data, and a generator comparison network is used to train the fused data based on the prior knowledge relationship map of overhead distribution line fittings to generate a defect recognition model.

7. A system for identifying defects in hardware of overhead power distribution lines, characterized in that: include: The network communication and data management module is used to receive the inspection data of the overhead distribution line fittings from the inspection terminal, classify the inspection data, obtain the classified data, and pre-process the classified data to obtain the target data; A model building optimization module, used to use a pre-built defect recognition model to identify the target data and obtain a recognition result; The intelligent analysis module determines the hardware defects of the overhead distribution line based on the identification result.

8. The system according to claim 7, characterized in that The inspection terminal includes at least one of the following: an unmanned aerial vehicle terminal, a vehicle-mounted device terminal and a handheld terminal.

9. The system according to claim 7, characterized in that The inspection data includes at least one of the following: inspection data of a drone terminal, inspection data of a vehicle-mounted device terminal, and inspection data of a handheld terminal.

10. The system according to claim 7, characterized in that The preprocessing of the classified data comprises: The classified data is cleaned and then standardized.

11. The system according to claim 7, characterized in that The network communication and data management module is also used to: after preprocessing the classified data, store the preprocessed target data in a standardized database.

12. The system according to claim 7, characterized in that Pre-built defect identification models, including: A prior knowledge relationship map of overhead distribution line fittings that integrates positional relationships and feature distribution is established, small samples are generated by batch simulation, the small samples are fused with target data stored in a standardized database to generate fused data, and a generator comparison network is used to train the fused data based on the prior knowledge relationship map of overhead distribution line fittings to generate a defect recognition model.

13. The system according to claim 7, characterized in that The model building optimization module is also used to optimize the defect recognition model.

14. The system according to claim 7, wherein the characteristic resource, The system also includes: a model adaptation and conversion module, which is used to convert and adapt the defect recognition model to different inspection terminal architectures.

15. A computer device, characterized in that: include: one or more processors; a processor for executing one or more programs; When the one or more programs are executed by the one or more processors, the method according to any one of claims 1 to 6 is implemented.

16. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed, the method according to any one of claims 1 to 6 is implemented.

Images