AI-based generator outlet circuit breaker contact wear detection method and system

Abstract

The invention discloses an AI-based generator outlet circuit breaker contact wear detection method and system. The method comprises the steps of obtaining and preprocessing a historical wear image of a circuit breaker contact to obtain a preprocessed image; extracting image features from the preprocessed image, analyzing the image features, determining key image features related to the wear of the circuit breaker contact, and constructing a feature recognition model based on the key image features; acquiring a real-time acquisition image of the circuit breaker contact and identifying key image features through the model; evaluating each key image feature to obtain a wear degree evaluation value; determining the weight of each key image feature, and determining a wear degree value based on the weight of each key image feature and the wear degree evaluation value; and sending out early warning information after judging that the wear degree value exceeds a preset threshold value. According to the method, the accuracy, efficiency and timeliness of wear detection of the contact of the generator outlet circuit breaker are improved, safe and stable operation of a power system is ensured, and meanwhile, a scientific basis and decision support are provided for maintenance and management.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breaker detection, and in particular to an AI-based generator outlet circuit breaker contact wear detection method and system. Background Art

[0002] The wear of the generator output circuit breaker contacts is an important fault problem in the power generation system. Due to long-term use, the circuit breaker contacts will be worn due to contact current, resulting in uneven contact surface, oxidation, ablation, etc., which in turn affects its normal working performance. Therefore, timely and accurate detection of the wear of the circuit breaker contacts is crucial to ensure the safe and stable operation of the power system.

[0003] However, traditional contact wear detection methods usually rely on manual inspections, which have problems such as low efficiency, strong subjectivity, and inability to monitor in real time. They often fail to promptly detect potential faults caused by contact wear. Therefore, automated contact wear detection methods based on computer vision and machine learning have become a solution that has attracted much attention. By using image processing technology and deep learning algorithms, the wear of circuit breaker contacts can be monitored and analyzed in real time, greatly improving the accuracy and efficiency of detection. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method and system for detecting contact wear of a generator output circuit breaker based on AI, comprising:

[0005] Acquire a historical wear image of a contact of a generator outlet circuit breaker, and preprocess the historical wear image to obtain a preprocessed image;

[0006] Extracting image features from the preprocessed image and analyzing the image features to determine key image features related to circuit breaker contact wear;

[0007] Building an initial feature recognition model based on key image features, and training and testing the initial feature recognition model to obtain a feature recognition model;

[0008] Acquire a real-time acquisition image of the generator outlet circuit breaker contact, and identify key image features in the real-time acquisition image through a feature recognition model;

[0009] Evaluate the wear degree of each key image feature to obtain a wear degree evaluation value of each key image feature;

[0010] Determine the weight of each key image feature, and determine the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value;

[0011] Determine the relationship between the wear degree value and the preset threshold, and issue a warning message when the wear degree value exceeds the preset threshold.

[0012] Furthermore, the preprocessing of the historical wear image to obtain the preprocessed image includes:

[0013] The historical wear image is converted into a grayscale image using a grayscale conversion method, and the noise in the grayscale image is removed using a denoising algorithm;

[0014] The contrast of the grayscale image is enhanced by histogram equalization, and the edge detection algorithm is used to identify and mark the edges and contours in the historical wear image to obtain the preprocessed image.

[0015] Furthermore, the image features are extracted from the preprocessed image, and the image features are analyzed to determine key image features related to the wear of the circuit breaker contacts in the image features, including:

[0016] A number of image features are extracted from the preprocessed image by using image processing technology, and the correlation between each image feature and the wear of the circuit breaker contact is calculated;

[0017] The average value of the correlation between all image features and circuit breaker contact wear is calculated, and image features with correlation exceeding the average value are screened out to be determined as key image features related to circuit breaker contact wear among the image features.

[0018] Furthermore, the initial feature recognition model is constructed based on the key image features, and the initial feature recognition model is trained and tested to obtain the feature recognition model, including:

[0019] Acquire image data corresponding to key image features to establish a data set, and build an initial feature recognition model based on the data set and a preset neural network model;

[0020] Divide the data set into training set, test set and validation set according to a certain ratio, and generate the convergence conditions of the initial model of feature recognition based on the data in the validation set;

[0021] The training set and the test set are input into the initial feature recognition model, and the initial feature recognition model is iteratively trained and tested until the convergence condition of the initial feature recognition model meets the preset convergence condition, thereby obtaining the feature recognition model.

[0022] Furthermore, the step of evaluating the degree of wear of each key image feature to obtain an evaluation value of the degree of wear of each key image feature includes:

[0023] Performing similarity analysis on the key image features and the corresponding standard image features, and calculating the similarity value between the key image features and the corresponding standard image features;

[0024] A corresponding relationship between the wear degree assessment value and the similarity value interval is preset, and the corresponding relationship between the wear degree assessment value and the similarity value interval is associated with a corresponding wear degree assessment value for each similarity value interval;

[0025] A similarity value is obtained, and based on a mapping relationship between a similarity value interval to which the similarity value belongs and a wear degree evaluation value corresponding to the similarity value interval, a wear degree evaluation value corresponding to the similarity value interval is selected as the wear degree evaluation value of the key image feature.

[0026] Furthermore, the calculating of the similarity value between the key image feature and the corresponding standard image feature includes:

[0027] The key image features and the corresponding standard image features are represented by vectors to obtain a key image feature vector and a standard image feature vector, and the key image feature vector and the standard image feature vector are subjected to dimensionality reduction processing to obtain a key image feature value and a standard image feature value;

[0028] The similarity value between the key image feature and the corresponding standard image feature is calculated according to the key image feature value and the standard image feature value, wherein the calculation formula of the similarity value is:

[0029] M=α*|D-D0|,

[0030] Among them, M is the similarity value, α is the similarity coefficient, D is the key image feature value, and D0 is the standard image feature value.

[0031] Further, the determining of the weight of each key image feature, and determining the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value, includes:

[0032] The relevance of each key image feature is obtained, and the weight of the key image feature is determined based on the relevance, wherein the calculation formula of the weight of the key image feature is:

[0033] R=Pi / (P1+P2+……+Pn),

[0034] Where R is the weight of the key image feature, Pi is the correlation of i key image features, and n is the number of all key image features;

[0035] The weight of each key image feature and the corresponding wear degree assessment value are weighted and added to obtain the wear degree value of the circuit breaker contact.

[0036] Furthermore, the determining the relationship between the wear degree value and a preset threshold value, and issuing a warning message after the wear degree value exceeds the preset threshold value, includes:

[0037] Determine the relationship between the wear degree value and the preset threshold value. If the wear degree value does not exceed the preset threshold value, it means that the wear degree of the circuit breaker contact is not serious and there is no need to issue an early warning message.

[0038] If the wear value exceeds the preset threshold, it means that the circuit breaker contacts are severely worn and an early warning message needs to be issued.

[0039] The present invention also provides an AI-based generator output circuit breaker contact wear detection system, comprising:

[0040] An acquisition module is used to acquire a historical wear image of a contact of a generator outlet circuit breaker, and preprocess the historical wear image to obtain a preprocessed image;

[0041] An extraction module, used for extracting image features from the preprocessed image, analyzing the image features, and determining key image features related to circuit breaker contact wear;

[0042] A modeling module is used to build an initial feature recognition model based on key image features, and to train and test the initial feature recognition model to obtain a feature recognition model;

[0043] A recognition module is used to obtain a real-time acquisition image of the generator outlet circuit breaker contacts, and to identify key image features in the real-time acquisition image through a feature recognition model;

[0044] An evaluation module is used to evaluate the degree of wear of each key image feature to obtain an evaluation value of the degree of wear of each key image feature;

[0045] A determination module, used to determine the weight of each key image feature, and determine the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value;

[0046] The judgment module is used to judge the relationship between the wear degree value and the preset threshold value, and issue a warning message after the wear degree value exceeds the preset threshold value.

[0047] Compared with the prior art, the method and system for detecting contact wear of a generator output circuit breaker based on AI in the embodiment of the present invention have the following beneficial effects:

[0048] The present invention can realize automatic detection of circuit breaker contact wear by constructing a feature recognition model, eliminating the subjectivity and low efficiency of traditional manual inspection;

[0049] The present invention can realize real-time monitoring of the wear of circuit breaker contacts based on feature recognition of real-time collected images, and timely discover potential fault hazards;

[0050] The present invention can improve the accurate assessment of the degree of wear through image feature analysis and model training, avoiding the uncertainty of subjective assessment in traditional methods;

[0051] Based on the evaluation value of the wear degree and the preset threshold, the present invention can establish an early warning mechanism to issue an alarm in time to avoid potential risks caused by severe contact wear;

[0052] The automated wear detection method of the present invention can reduce labor costs and time costs, improve maintenance efficiency, and reduce losses caused by severe wear that is not discovered in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the flow structure of a method for detecting contact wear of a generator output circuit breaker based on AI in an embodiment of the present invention;

[0054] Figure 2 It is a schematic diagram of the composition of the generator output circuit breaker contact wear detection system based on AI in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The specific implementation methods of the present application are further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0056] In the description of the present application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are the orientation or position relationship shown in the accompanying drawings, which are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the platform or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0057] The terms "second" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, a feature defined with "second" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "multiple" means two or more.

[0058] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0059] like Figure 1 As shown, in an embodiment of the present application, an AI-based generator outlet circuit breaker contact wear detection method is provided, including: S100: obtaining a historical wear image of the generator outlet circuit breaker contact, and preprocessing the historical wear image to obtain a preprocessed image; S200: extracting image features from the preprocessed image, analyzing the image features, and determining key image features related to the circuit breaker contact wear; S300: constructing a feature recognition initial model based on the key image features, and training and testing the feature recognition initial model to obtain a feature recognition model; S400: obtaining a real-time acquisition image of the generator outlet circuit breaker contact, and identifying the key image features in the real-time acquisition image through the feature recognition model; S500: evaluating the wear degree of each key image feature to obtain a wear degree evaluation value of each key image feature; S600: determining the weight of each key image feature, and determining the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value; S700: judging the relationship between the wear degree value and a preset threshold, and issuing a warning message after the wear degree value exceeds the preset threshold.

[0060] Furthermore, the present invention can realize automatic detection of circuit breaker contact wear by constructing a feature recognition model, eliminating the subjectivity and inefficiency of traditional manual inspections; the present invention can realize real-time monitoring of circuit breaker contact wear based on feature recognition of real-time acquired images, and timely discover potential fault hazards; the present invention can improve the accurate assessment of the degree of wear through image feature analysis and model training, avoiding the uncertainty of subjective assessment in traditional methods; the present invention can establish an early warning mechanism based on the assessment value of the degree of wear and the preset threshold, and issue an alarm in time to avoid potential risks caused by severe contact wear; the automated wear detection method of the present invention can reduce labor costs and time costs, improve maintenance efficiency, and at the same time reduce losses caused by severe wear not being discovered in time.

[0061] In an embodiment of the present application, a method for detecting contact wear of a generator output circuit breaker based on AI is provided, wherein the historical wear image is preprocessed to obtain a preprocessed image, including: converting the historical wear image into a grayscale image using a grayscale conversion method, and removing noise in the grayscale image using a denoising algorithm; enhancing the contrast of the grayscale image by a histogram equalization method, and using an edge detection algorithm to identify and mark edges and contours in the historical wear image to obtain a preprocessed image.

[0062] Specifically, the color historical wear image is converted into a grayscale image through the weighted averaging method to simplify image processing and reduce the amount of calculation; the denoising algorithm (such as Gaussian blur, median filtering, etc.) is used to remove the noise in the grayscale image, maintain the details of the image and reduce interference; the histogram equalization method is applied to enhance the contrast of the grayscale image to improve the clarity and visualization of the image. The histogram equalization can adjust the grayscale distribution of the image to make the contrast of the image more prominent; the edge detection algorithm (such as Sobel, Canny, etc.) is used to identify and mark the edges and contours in the historical wear image, highlight the contours of the worn part, and facilitate subsequent feature extraction and analysis.

[0063] In an embodiment of the present application, an AI-based method for detecting contact wear of a generator output circuit breaker is provided, wherein image features are extracted from a preprocessed image, and the image features are analyzed to determine key image features in the image features that are related to the wear of the circuit breaker contacts, including: extracting a number of image features from the preprocessed image by image processing technology, and calculating the correlation between each image feature and the wear of the circuit breaker contacts; calculating the average value of the correlation between all image features and the wear of the circuit breaker contacts, and screening out image features whose correlation exceeds the average value, and determining them as key image features in the image features that are related to the wear of the circuit breaker contacts.

[0064] Specifically, a number of image features are extracted from the preprocessed image through various image processing and computer vision techniques; for each extracted image feature, statistical methods (such as Pearson correlation coefficient and Spearman correlation coefficient) are used to calculate its correlation with the wear of the circuit breaker contacts, and determine which features have a strong correlation with the wear; the average value of the correlation between all image features and the wear of the circuit breaker contacts is calculated, and image features with correlations exceeding the average value are screened out to determine them as key image features among the image features that are related to the wear of the circuit breaker contacts.

[0065] In an embodiment of the present application, a method for detecting contact wear of a generator output circuit breaker based on AI is provided, wherein an initial feature recognition model is constructed based on key image features, and the initial feature recognition model is trained and tested to obtain the feature recognition model, including: acquiring image data corresponding to the key image features to establish a data set, and constructing the initial feature recognition model according to the data set and a preset neural network model; dividing the data set into a training set, a test set, and a validation set according to a certain ratio, and generating a convergence condition of the initial feature recognition model based on the data in the validation set; inputting the training set and the test set into the initial feature recognition model, iteratively training and testing the initial feature recognition model, and obtaining the feature recognition model after the convergence condition of the initial feature recognition model meets the preset convergence condition.

[0066] Specifically, the corresponding image data is obtained from the key image features, and a data set is established to ensure that each image feature can be correctly matched with its corresponding image data; based on the preset neural network model, a deep learning framework (such as TensorFlow, PyTorch) can be used to build a feature recognition initial model, and an appropriate network structure and parameter settings can be selected; the data set is divided into a training set, a test set, and a validation set according to a certain ratio, using a common division ratio, such as 70% of the data for training, 15% for testing, and 15% for validation; based on the data in the validation set, the convergence of the feature recognition initial model is monitored, and convergence conditions are defined, such as the accuracy on the validation set reaches a certain threshold or the training loss no longer decreases significantly; the training set and the test set are input into the feature recognition initial model, and the model is iteratively trained and tested. After each training cycle, the validation set is used to evaluate the performance of the model, and the model is monitored to see whether it meets the preset convergence conditions. When the feature recognition initial model meets the preset convergence conditions, the feature recognition model can be obtained.

[0067] In an embodiment of the present application, a method for detecting contact wear of a generator output circuit breaker based on AI is provided, wherein the wear degree of each key image feature is evaluated to obtain a wear degree evaluation value of each key image feature, including: performing similarity analysis on the key image feature and the corresponding standard image feature, and calculating the similarity value between the key image feature and the corresponding standard image feature; presetting a wear degree evaluation value-similarity value interval correspondence, wherein the wear degree evaluation value-similarity value interval correspondence is associated with a corresponding wear degree evaluation value for each similarity value interval; obtaining a similarity value, and based on a mapping relationship between the similarity value interval to which the similarity value belongs within the wear degree evaluation value-similarity value interval correspondence, selecting a wear degree evaluation value corresponding to the similarity value interval as the wear degree evaluation value of the key image feature.

[0068] Specifically, similarity analysis is performed on the key image features and the corresponding standard image features; the correspondence between the wear degree assessment value and the similarity value is preset. Based on actual conditions and experience, you can divide the similarity value into different intervals, assign a corresponding wear degree assessment value to each interval, obtain the similarity value of the key image features, and select the wear degree assessment value corresponding to the similarity value interval as the wear degree assessment value of the key image features based on the preset wear degree assessment value-similarity value interval correspondence.

[0069] In an embodiment of the present application, a generator outlet circuit breaker contact wear detection method based on AI is provided, wherein the similarity value between the key image feature and the corresponding standard image feature is calculated, including: representing the key image feature and the corresponding standard image feature with a vector to obtain a key image feature vector and a standard image feature vector, and performing dimensionality reduction processing on the key image feature vector and the standard image feature vector to obtain a key image feature value and a standard image feature value; calculating the similarity value between the key image feature and the corresponding standard image feature according to the key image feature value and the standard image feature value, wherein the calculation formula of the similarity value is:

[0070] M=α*|D-D0|,

[0071] Among them, M is the similarity value, α is the similarity coefficient, D is the key image feature value, and D0 is the standard image feature value.

[0072] Specifically, the key image features and the corresponding standard image features are represented by vectors, and various feature extraction methods (such as CNN, feature descriptors, etc.) are used to convert image features into vector representations; the key image feature vectors and the standard image feature vectors are subjected to dimensionality reduction processing to obtain key image feature values ​​and standard image feature values; and the similarity values ​​between them are calculated based on the key image feature values ​​and the standard image feature values.

[0073] In an embodiment of the present application, a generator outlet circuit breaker contact wear detection method based on AI is provided, wherein the weight of each key image feature is determined, and the wear degree value of the circuit breaker contact is determined based on the weight of each key image feature and the wear degree evaluation value, including: obtaining the correlation of each key image feature, and determining the weight of the key image feature based on the correlation, wherein the calculation formula of the weight of the key image feature is:

[0074] R=Pi / (P1+P2+……+Pn),

[0075] Among them, R is the weight of the key image feature, Pi is the correlation of i key image features, and n is the number of all key image features; the weight of each key image feature and the corresponding wear degree assessment value are weighted added to obtain the wear degree value of the circuit breaker contact.

[0076] In an embodiment of the present application, a generator output circuit breaker contact wear detection method based on AI is provided, wherein the relationship between the wear degree value and a preset threshold value is judged, and a warning message is issued after the wear degree value exceeds the preset threshold value, including: judging the relationship between the wear degree value and the preset threshold value, if the wear degree value does not exceed the preset threshold value, it indicates that the wear degree of the circuit breaker contact is not serious and there is no need to issue a warning message; if the wear degree value exceeds the preset threshold value, it indicates that the wear degree of the circuit breaker contact is serious and a warning message needs to be issued.

[0077] like Figure 2 As shown, in an embodiment of the present application, an AI-based generator outlet circuit breaker contact wear detection system is provided, including: an acquisition module, used to acquire historical wear images of generator outlet circuit breaker contacts, and preprocess the historical wear images to obtain preprocessed images; an extraction module, used to extract image features from the preprocessed images, analyze the image features, and determine key image features related to circuit breaker contact wear; a modeling module, used to build a feature recognition initial model based on key image features, and train and test the feature recognition initial model to obtain a feature recognition model; a recognition module, used to acquire real-time acquisition images of generator outlet circuit breaker contacts, and identify key image features in the real-time acquisition images through the feature recognition model; an evaluation module, used to evaluate the degree of wear of each key image feature, and obtain a wear degree evaluation value of each key image feature; a determination module, used to determine the weight of each key image feature, and determine the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value; a judgment module, used to judge the relationship between the wear degree value and a preset threshold, and issue a warning message after the wear degree value exceeds the preset threshold.

[0078] In summary, the embodiment of the present invention provides a method and system for detecting contact wear of a generator output circuit breaker based on AI, which includes: acquiring and preprocessing the historical wear image of the circuit breaker contact to obtain a preprocessed image; extracting image features from the preprocessed image and analyzing it, determining the key image features related to the wear of the circuit breaker contact, and building a feature recognition model based thereon; acquiring the real-time acquisition image of the circuit breaker contact and identifying the key image features through the model; evaluating each key image feature to obtain a wear degree assessment value; determining the weight of each key image feature, and determining the wear degree value based on the weight of each key image feature and the wear degree assessment value; issuing a warning message after judging that the wear degree value exceeds a preset threshold. The present invention improves the accuracy, efficiency and timeliness of contact wear detection of the generator output circuit breaker, helps to ensure the safe and stable operation of the power system, and provides a scientific basis and decision support for maintenance management.

[0079] Finally, it should be noted that it is obvious that 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 equivalents, the present invention is also intended to include these modifications and variations.

[0080] The above is only an example of implementation of the present invention, but it cannot be used to limit the scope of the present invention. Any structural changes made according to the present invention, as long as they do not lose the essence of the present invention, should be regarded as falling within the scope of protection of the present invention and being restricted. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process and related instructions of the platform described above can refer to the corresponding process in the aforementioned platform embodiment, and will not be repeated here.

[0081] The term "comprises" or any other similar term is intended to cover a non-exclusive inclusion such that a process, platform, article, or apparatus / platform that includes a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, platform, article, or apparatus / platform.

[0082] So far, the technical solutions of the present invention have been described in conjunction with the further embodiments shown in the accompanying drawings, but it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention.

Claims

1. A generator output circuit breaker contact wear detection method based on AI, characterized in that: include: Acquire a historical wear image of a contact of a generator outlet circuit breaker, and preprocess the historical wear image to obtain a preprocessed image; Extracting image features from the preprocessed image, analyzing the image features, and determining key image features related to circuit breaker contact wear; Building an initial feature recognition model based on key image features, and training and testing the initial feature recognition model to obtain a feature recognition model; Acquire a real-time acquisition image of the generator outlet circuit breaker contact, and identify key image features in the real-time acquisition image through a feature recognition model; Evaluate the wear degree of each key image feature to obtain a wear degree evaluation value of each key image feature; Determine the weight of each key image feature, and determine the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value; Determine the relationship between the wear degree value and the preset threshold, and issue a warning message when the wear degree value exceeds the preset threshold.

2. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 1, characterized in that: The preprocessing of the historical wear image to obtain the preprocessed image includes: The historical wear image is converted into a grayscale image using a grayscale conversion method, and the noise in the grayscale image is removed using a denoising algorithm; The contrast of the grayscale image is enhanced by histogram equalization, and the edge detection algorithm is used to identify and mark the edges and contours in the historical wear image to obtain the preprocessed image.

3. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 2, characterized in that: The extracting of image features from the preprocessed image and analyzing the image features to determine key image features related to circuit breaker contact wear in the image features include: A number of image features are extracted from the preprocessed image by using image processing technology, and the correlation between each image feature and the wear of the circuit breaker contact is calculated; The average value of the correlation between all image features and circuit breaker contact wear is calculated, and image features with correlation exceeding the average value are screened out to be determined as key image features related to circuit breaker contact wear among the image features.

4. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 3 is characterized in that: The initial feature recognition model is constructed based on the key image features, and the initial feature recognition model is trained and tested to obtain the feature recognition model, including: Acquire image data corresponding to key image features to establish a data set, and build an initial feature recognition model based on the data set and a preset neural network model; Divide the data set into training set, test set and validation set according to a certain ratio, and generate the convergence conditions of the initial model of feature recognition based on the data in the validation set; The training set and the test set are input into the initial feature recognition model, and the initial feature recognition model is iteratively trained and tested until the convergence condition of the initial feature recognition model meets the preset convergence condition, thereby obtaining the feature recognition model.

5. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 4, characterized in that: The step of evaluating the degree of wear of each key image feature to obtain an evaluation value of the degree of wear of each key image feature includes: Performing similarity analysis on the key image features and the corresponding standard image features, and calculating the similarity value between the key image features and the corresponding standard image features; A corresponding relationship between the wear degree assessment value and the similarity value interval is preset, and the corresponding relationship between the wear degree assessment value and the similarity value interval is associated with a corresponding wear degree assessment value for each similarity value interval; A similarity value is obtained, and based on a mapping relationship between a similarity value interval to which the similarity value belongs and a wear degree evaluation value corresponding to the similarity value interval, a wear degree evaluation value corresponding to the similarity value interval is selected as the wear degree evaluation value of the key image feature.

6. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 5, characterized in that: The calculating the similarity value between the key image feature and the corresponding standard image feature includes: The key image features and the corresponding standard image features are represented by vectors to obtain a key image feature vector and a standard image feature vector, and the key image feature vector and the standard image feature vector are subjected to dimensionality reduction processing to obtain a key image feature value and a standard image feature value; The similarity value between the key image feature and the corresponding standard image feature is calculated according to the key image feature value and the standard image feature value, wherein the calculation formula of the similarity value is: M=α*|D-D0|, Among them, M is the similarity value, α is the similarity coefficient, D is the key image feature value, and D0 is the standard image feature value.

7. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 5, characterized in that: The step of determining the weight of each key image feature and determining the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value includes: The relevance of each key image feature is obtained, and the weight of the key image feature is determined based on the relevance, wherein the calculation formula of the weight of the key image feature is: R=Pi / (P1+P2+……+Pn), Where R is the weight of the key image feature, Pi is the correlation of i key image features, and n is the number of all key image features; The weight of each key image feature and the corresponding wear degree assessment value are weighted and added to obtain the wear degree value of the circuit breaker contact.

8. The method for detecting contact wear of a generator output circuit breaker based on AI according to claim 7, characterized in that: The determining of the relationship between the wear degree value and the preset threshold value, and issuing a warning message after the wear degree value exceeds the preset threshold value, includes: Determine the relationship between the wear degree value and the preset threshold value. If the wear degree value does not exceed the preset threshold value, it means that the wear degree of the circuit breaker contact is not serious and there is no need to issue an early warning message. If the wear value exceeds the preset threshold, it means that the circuit breaker contacts are severely worn and an early warning message needs to be issued.

9. An AI-based generator output circuit breaker contact wear detection system, characterized in that: include: An acquisition module is used to acquire a historical wear image of a contact of a generator outlet circuit breaker, and preprocess the historical wear image to obtain a preprocessed image; An extraction module, used for extracting image features from the preprocessed image, analyzing the image features, and determining key image features related to circuit breaker contact wear; A modeling module is used to build an initial feature recognition model based on key image features, and to train and test the initial feature recognition model to obtain a feature recognition model; A recognition module is used to obtain a real-time acquisition image of the generator outlet circuit breaker contacts, and to identify key image features in the real-time acquisition image through a feature recognition model; An evaluation module is used to evaluate the degree of wear of each key image feature to obtain an evaluation value of the degree of wear of each key image feature; A determination module, used to determine the weight of each key image feature, and determine the wear degree value of the circuit breaker contact based on the weight of each key image feature and the wear degree evaluation value; The judgment module is used to judge the relationship between the wear degree value and the preset threshold value, and issue a warning message after the wear degree value exceeds the preset threshold value.

Images