Power equipment defect detection method based on integrated confrontation training
Through the integrated adversarial training method, the training data is expanded by using data augmentation and adversarial samples, and combined with multiple deep learning models to extract features, the generalization ability and robustness of the power equipment defect detection model is improved, and the problems of insufficient data and poor generalization ability in the existing technology are solved.
Patent Information
- Application Number
- CN202411849102.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-16
- Publication Date
- 2025-04-08
AI Technical Summary
The existing power equipment defect detection methods face the problems of insufficient data scale, resulting in low model detection accuracy, changes in lighting conditions affect the stability of detection results, and poor generalization capabilities.
The integrated adversarial training method is adopted to generate adversarial samples through data augmentation and multiple adversarial attack methods, expand the training data set, and extract high-dimensional features using multiple deep learning models (convolutional neural network, deep residual network and ViT network), and combine multi-layer perceptron models for classification training to improve the generalization ability and robustness of the model.
It significantly improves the generalization ability of the power equipment defect detection model in multiple scenarios and the robustness of the real application scenarios, and improves the stability and accuracy of the detection results.
Smart Images
Figure CN119784700A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power equipment defect detection, and in particular to a power equipment defect detection method based on integrated adversarial training. Background Art
[0002] Deep learning technology has been widely used in power equipment defect detection, significantly improving the efficiency and accuracy of fault diagnosis. Through advanced models such as convolutional neural networks (CNN), researchers can extract useful features from complex power equipment images to achieve real-time monitoring of equipment status. For example, a deep learning-based defect detection system can automatically identify abnormal temperature of transformers, cracks in insulation materials, and other problems, greatly reducing the time and labor costs of manual inspections.
[0003] However, existing methods face many challenges: first, many models rely on a large amount of high-quality training data, while defect samples of power equipment are often scarce, which makes the model easily overfit to the limited data set; second, the lighting conditions in the power production environment vary greatly, which has an adverse effect on the image quality and feature extraction, and thus affects the stability of the detection results; in addition, the existing detection methods have poor generalization ability when dealing with different types of defects and often cannot effectively deal with unknown faults. Summary of the invention
[0004] The purpose of the present invention is to provide a method for power equipment defect detection based on integrated adversarial training, which can solve the problem of low model detection accuracy caused by insufficient data scale, and significantly improve the generalization ability of the model in multiple scenarios and its robustness in real application scenarios.
[0005] The technical solution of the present invention is:
[0006] In a first aspect, the present application provides a method for detecting defects in power equipment based on integrated adversarial training, which comprises the following steps:
[0007] S1. Obtain historical power equipment defect images to establish a data set, and perform data enhancement on the images in the data set;
[0008] S2, input the images of the data set after data enhancement into the convolutional neural network, deep residual network and ViT network for high-dimensional feature extraction to obtain three high-dimensional features;
[0009] S3, concatenating the three high-dimensional features to obtain integrated high-dimensional features, and inputting the integrated high-dimensional features into the multi-layer perceptron model to obtain the classification results, and using the cross entropy function as the objective function to train the model to obtain the classification model;
[0010] S4, selecting images from the data-enhanced dataset to generate adversarial samples using multiple adversarial attack methods, and expanding the original dataset through the adversarial samples to obtain an adversarially enhanced dataset;
[0011] S5. Traverse the adversarially enhanced data set to retrain the classification model to obtain a power equipment defect detection model, and input the power equipment defect image to be detected into the power equipment defect detection model to obtain a detection result.
[0012] Furthermore, in step S1, the process of performing data enhancement on the data set includes:
[0013] Select the images in the dataset and rotate them randomly;
[0014] The randomly rotated image and the original image are fused according to a preset ratio to obtain an enhanced image;
[0015] Traverse all the images in the dataset to get the enhanced dataset.
[0016] Furthermore, the calculation formula for the above data fusion is:
[0017] x eh =λ·x+(1-λ)·x'
[0018] In the formula, x eh is the image after data fusion, λ is the preset ratio, x is the defective image of power equipment in the data set, and x' is the new image after random rotation.
[0019] Furthermore, in step S2, the images of the data set after data augmentation are respectively input into the convolutional neural network, the deep residual network and the ViT network for high-dimensional feature extraction. The calculation formula includes:
[0020]
[0021] In the formula, is the high-dimensional feature extracted by the convolutional neural network, F CNN is the convolutional neural network feature extraction function, x i This is a picture of power equipment defects in the data set after data enhancement. is the high-dimensional feature extracted by the deep residual network, F ResNet It is a residual network composed of multiple convolutional units. F is the high-dimensional feature extracted by the ViT network. ViT It is the ViT deep network.
[0022] Furthermore, in step S4, the multiple adversarial attack methods used include PGD, MI-FGSM and DeepFool. Images are selected from the data set after data enhancement to generate adversarial sample expressions using multiple adversarial attack methods:
[0023]
[0024] In the formula, S e To combat the enhanced dataset, x eh The image after data fusion. It is the adversarial sample corresponding to the PGD adversarial attack method. is the adversarial sample corresponding to the MI-FGSM adversarial attack method. This is the adversarial sample corresponding to the DeepFool adversarial attack method.
[0025] In a second aspect, the present application provides an electronic device, including:
[0026] A memory for storing one or more programs;
[0027] processor;
[0028] When the one or more programs are executed by the processor, a method for detecting defects in electric power equipment based on integrated adversarial training as described in any one of the first aspects is implemented.
[0029] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for defect detection of power equipment based on integrated adversarial training as described in any one of the first aspects above.
[0030] Compared with the prior art, the present invention has at least the following advantages or beneficial effects:
[0031] (1) The present invention provides a method for detecting defects in electric power equipment based on ensemble adversarial training. On the one hand, the method uses image transformation and other methods to enhance data, thereby expanding the model training data set and alleviating the problem of low model detection accuracy due to insufficient data size. On the other hand, the method uses ensemble learning to integrate the advantages of multiple models from both the data and model levels, thereby improving the generalization ability of the method for detecting defects in electric power equipment in different scenarios.
[0032] (2) The present invention improves the robustness of the power equipment defect detection model in real applications through adversarial training, and improves the safe application capability of the power equipment defect detection model. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0034] Figure 1 A step diagram of a method for detecting defects in electric power equipment based on integrated adversarial training according to the present invention;
[0035] Figure 2 A flowchart of a method for detecting defects in electric power equipment based on integrated adversarial training according to the present invention;
[0036] Figure 3 The figure is a schematic structural block diagram of an electronic device according to an embodiment of the present invention.
[0037] Icon: 101, memory; 102, processor; 103, communication interface. DETAILED DESCRIPTION
[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.
[0039] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.
[0040] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.
[0041] It should be noted that, in this article, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprises..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0042] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.
[0043] Example 1
[0044] See also Figure 1 , Figure 1 Shown is a step diagram of a method for defect detection of electric power equipment based on integrated adversarial training provided in an embodiment of the present application.
[0045] In a first aspect, the present application provides a method for detecting defects in power equipment based on integrated adversarial training, which comprises the following steps:
[0046] S1. Obtain historical power equipment defect images to establish a data set, and perform data enhancement on the images in the data set;
[0047] S2, input the images of the data set after data enhancement into the convolutional neural network, deep residual network and ViT network for high-dimensional feature extraction to obtain three high-dimensional features;
[0048] S3, concatenating the three high-dimensional features to obtain integrated high-dimensional features, and inputting the integrated high-dimensional features into the multi-layer perceptron model to obtain the classification results, and using the cross entropy function as the objective function to train the model to obtain the classification model;
[0049] S4, selecting images from the data-enhanced dataset to generate adversarial samples using multiple adversarial attack methods, and expanding the original dataset through the adversarial samples to obtain an adversarially enhanced dataset;
[0050] S5. Traverse the adversarially enhanced data set to retrain the classification model to obtain a power equipment defect detection model, and input the power equipment defect image to be detected into the power equipment defect detection model to obtain a detection result.
[0051] As a preferred implementation, in step S1, the process of performing data enhancement on the data set includes:
[0052] Select the images in the dataset and rotate them randomly;
[0053] The randomly rotated image and the original image are fused according to a preset ratio to obtain an enhanced image;
[0054] Traverse all the images in the dataset to get the enhanced dataset.
[0055] As a preferred implementation, the calculation formula for data fusion is:
[0056] x eh =λ·x+(1-λ)·x'
[0057] In the formula, x eh is the image after data fusion, λ is the preset ratio, x is the defective image of power equipment in the data set, and x' is the new image after random rotation.
[0058] As a preferred implementation, in step S2, the pictures of the data set after data enhancement are respectively input into the convolutional neural network, the deep residual network and the ViT network for high-dimensional feature extraction. The calculation formula includes:
[0059]
[0060] In the formula, is the high-dimensional feature extracted by the convolutional neural network, F CNN is the convolutional neural network feature extraction function, x i This is a picture of power equipment defects in the data set after data enhancement. is the high-dimensional feature extracted by the deep residual network, F ResNet It is a residual network composed of multiple convolutional units. F is the high-dimensional feature extracted by the ViT network. ViT It is the ViT deep network.
[0061] As a preferred implementation, in step S4, the multiple adversarial attack methods used include PGD, MI-FGSM and DeepFool, and images are selected from the data set after data enhancement to generate adversarial sample expressions using multiple adversarial attack methods:
[0062]
[0063] In the formula, S e To combat the enhanced dataset, x eh The image after data fusion. It is the adversarial sample corresponding to the PGD adversarial attack method. is the adversarial sample corresponding to the MI-FGSM adversarial attack method. This is the adversarial sample corresponding to the DeepFool adversarial attack method.
[0064] like Figure 2 The flowchart of the method for detecting defects in electric equipment based on integrated adversarial training of the present invention is shown. The present invention establishes a data set of defective images of electric equipment, and then enhances the data set through feature transformation, and then expands the data set after adversarial enhancement through the data set of adversarial training, and inputs the multi-layer perceptron MLP model to output the abnormal type, that is, the abnormal result through high-dimensional feature extraction, and then trains the model using the cross entropy function as the objective function, and continuously improves the robustness of the model through adversarial training, until all images in the original data set are traversed to obtain the final electric equipment defect detection model, and then the image to be detected is input to output the detection result.
[0065] Example 2
[0066] See also Figure 3 , an electronic device includes a memory 101, a processor 102 and a communication interface 103, wherein the memory 101, the processor 102 and the communication interface 103 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines. The memory 101 can be used to store software programs and modules, and the processor 102 executes various functional applications and data processing by executing the software programs and modules stored in the memory 101. The communication interface 103 can be used to communicate signaling or data with other node devices.
[0067] Among them, the memory 101 can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.
[0068] The processor 102 may be an integrated circuit chip with signal processing capability. The processor 102 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0069] It is understood that the structure shown in the figure is only for illustration, and a method for detecting defects in power equipment based on integrated adversarial training may also include more or fewer components than those shown in the figure, or have a different configuration than those shown in the figure. Each component shown in the figure may be implemented by hardware, software, or a combination thereof.
[0070] In the embodiments provided in the present application, it should be understood that the disclosed method can also be implemented in other ways. The embodiments described above are merely schematic, for example, the flowchart or block diagram in the accompanying drawings shows the possible implementation architecture, functions and operations of the method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.
[0071] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
[0072] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.
[0073] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
[0074] It will be apparent to those skilled in the art that the present application is not limited to the details of the exemplary embodiments described above, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present application. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.
Claims
1. A method for detecting defects in power equipment based on integrated adversarial training, characterized in that: The following steps are involved: S1. Obtain historical power equipment defect images to establish a data set, and perform data enhancement on the images in the data set; S2, input the images of the data set after data enhancement into the convolutional neural network, deep residual network and ViT network for high-dimensional feature extraction to obtain three high-dimensional features; S3, concatenating the three high-dimensional features to obtain integrated high-dimensional features, and inputting the integrated high-dimensional features into the multi-layer perceptron model to obtain the classification results, and using the cross entropy function as the objective function to train the model to obtain the classification model; S4, selecting images from the data-enhanced dataset to generate adversarial samples using multiple adversarial attack methods, and expanding the original dataset through the adversarial samples to obtain an adversarially enhanced dataset; S5. Traverse the adversarially enhanced data set to retrain the classification model to obtain a power equipment defect detection model, and input the power equipment defect image to be detected into the power equipment defect detection model to obtain a detection result.
2. A method for detecting defects in electric power equipment based on integrated adversarial training according to claim 1, characterized in that: In step S1, the process of data enhancement on the data set includes: Select the images in the dataset and rotate them randomly; The randomly rotated image and the original image are fused according to a preset ratio to obtain an enhanced image; Traverse all the images in the dataset to get the enhanced dataset.
3. A method for detecting defects in electric power equipment based on integrated adversarial training as claimed in claim 2, characterized in that: The calculation formula of the data fusion is: x eh =λ x+(1-λ) x′ In the formula, x eh is the image after data fusion, λ is the preset ratio, x is the defective image of power equipment in the data set, and x′ is the new image after random rotation.
4. The method for detecting defects in electric power equipment based on integrated adversarial training according to claim 1, characterized in that: In step S2, the images of the data set after data augmentation are input into the convolutional neural network, deep residual network and ViT network respectively to calculate the formula for high-dimensional feature extraction include: In the formula, is the high-dimensional feature extracted by the convolutional neural network, F CNN is the convolutional neural network feature extraction function, x i This is a picture of power equipment defects in the data set after data enhancement. is the high-dimensional feature extracted by the deep residual network, F ResNet It is a residual network composed of multiple convolutional units. F is the high-dimensional feature extracted by the ViT network. ViT It is the ViT deep network.
5. The method for detecting defects in electric power equipment based on integrated adversarial training according to claim 1, characterized in that: In step S4, the multiple adversarial attack methods used include PGD, MI-FGSM and DeepFool. Images are selected from the data set after data enhancement to generate adversarial sample expressions using multiple adversarial attack methods: In the formula, S e To combat the enhanced dataset, x eh The image after data fusion. It is the adversarial sample corresponding to the PGD adversarial attack method. is the adversarial sample corresponding to the MI-FGSM adversarial attack method. This is the adversarial sample corresponding to the DeepFool adversarial attack method.
6. An electronic device, characterized in that: include: A memory for storing one or more programs; processor; When the one or more programs are executed by the processor, a method for detecting defects in electric power equipment based on integrated adversarial training as described in any one of claims 1 to 5 is implemented.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, an electric power equipment defect detection method based on integrated adversarial training as described in any one of claims 1 to 5 is implemented.