Intelligent circuit breaker rapid tripping protection method and device based on illegal charging detection, electronic equipment and storage medium

Abstract

The invention discloses an intelligent circuit breaker rapid tripping protection method and device based on illegal charging detection, electronic equipment and a storage medium. The method comprises the steps of S1, monitoring and collecting related charging data of charging equipment through an intelligent circuit breaker; s2, carrying out preprocessing and feature extraction on the collected related charging data; s3, introducing a fault category tree to construct illegal charging symptom descriptions, defining a group of illegal charging symptom descriptions, and constructing a data attribute collaboration mechanism; s4, for the fault type of illegal charging, training an intelligent identification model on the labeled data set; s5, performing real-time monitoring and illegal charging detection, inputting features into the intelligent identification model, and detecting different illegal charging; and S6, if illegal charging is detected, performing an action according to the illegal charging type, and otherwise, returning to the step S5. The state of the charging pile can be monitored in real time, illegal charging behaviors can be accurately recognized, the charging safety of residents is guaranteed, and the utilization rate of the charging pile is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breakers, and in particular to a method, device, equipment and storage medium for rapid tripping protection of an intelligent circuit breaker based on illegal charging detection. Background Art

[0002] With the increasing popularity of electric vehicles, the number of charging stations in residential areas has gradually increased, providing convenient charging facilities for residents. However, due to the complex electricity usage environment in residential areas, the irregular use of charging stations, and the lack of safety awareness among some users, illegal charging practices such as overcharging, prolonged charging, and unauthorized wiring and connections are common. These illegal charging behaviors not only damage charging stations and circuit equipment but can also cause safety accidents such as fires and explosions, threatening the lives and property of residents.

[0003] Therefore, there is an urgent need for an intelligent circuit breaker that can monitor the status of charging piles in real time, accurately identify illegal charging behaviors, and perform rapid tripping protection to ensure residents' charging safety and improve the utilization rate of charging piles. Summary of the Invention

[0004] In response to the above technical problems, the present application provides, on the one hand, a smart circuit breaker fast tripping protection method based on illegal charging detection.

[0005] This application is implemented through the following scheme:

[0006] A method for rapid tripping protection of an intelligent circuit breaker based on illegal charging detection comprises the following steps:

[0007] S1. Monitor and collect relevant charging data of charging equipment through intelligent circuit breakers;

[0008] S2. Preprocessing the collected relevant charging data and performing feature extraction;

[0009] S3. Introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute coordination mechanism;

[0010] S4. Train an intelligent recognition model on a labeled dataset for the fault types of illegal charging.

[0011] S5. Perform real-time monitoring and illegal charging detection, input the features into the trained intelligent recognition model, and detect different illegal charging;

[0012] S6. If illegal charging is detected, take action according to the illegal charging type. If no illegal charging is detected, return to step S5.

[0013] Furthermore, in step S1, the relevant charging data is as follows:

[0014] X={I,V,T,t c ,SOC}

[0015] Among them, V represents the charging voltage, I represents the charging current, T represents the temperature, and t c represents the charging time, and SOC represents the state of charge of the charging device.

[0016] Furthermore, in step S2, SPCA is used to extract features from relevant charging data, specifically including the following steps:

[0017] S21, for the feature matrix X∈R n×p , where n is the number of samples, p is the number of features, and each feature is standardized as follows:

[0018]

[0019] Where x′ j represents the standardized data, x j represents the data before normalization, μ j represents the average value of the jth feature, σ j represents the standard deviation of the jth feature;

[0020] S22. Construct the covariance matrix C based on the standardized features. The formula is as follows:

[0021]

[0022] Among them, X T represents the transpose of matrix X;

[0023] S23. Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ k and the corresponding eigenvector v k , the formula is as follows:

[0024]

[0025] S24. Introduce sparsity constraints in PCA by solving the following optimization problem:

[0026]

[0027] Among them, u k represents the sparse principal component, v k represents the feature vector, α represents the regularization parameter that controls the sparsity level, and || ||1 represents the L1 norm;

[0028] S25. Use the obtained sparse principal components as input of the fault diagnosis model.

[0029] Furthermore, in step S3, a description of illegal charging symptoms is constructed, and a set of descriptions of illegal charging symptoms is defined, which specifically includes the following steps:

[0030] Define V = {V1, V2, ..., V n} is a set of defined faults, n = 4, including overcharge, overheating, mismatch between charger and battery type, and battery aging, U = {U1, U2, ..., U m} is a set of undefined fault types, m is the number of undefined fault types; the sample set of V is expressed as X V Represents the defined training fault samples, N V is the number of samples, D is the dimension of the feature, Y V Represents the defined training fault sample label, and the undefined fault type is represented by X U represents undefined training fault samples, N U is the number of samples, D is the dimension of the feature, Y U Represents the undefined training fault sample label, where Y U It is still an unknown quantity;

[0031] The fault symptom description matrix A∈R is constructed by V and U P×Q , P = n + m, Q represents the dimension of the symptom vector, and uses binary to indicate whether a symptom exists in a certain fault category. "1" indicates the presence of the symptom, and "0" indicates the absence of the symptom. Each fault should consist of k symptoms. The symptoms refer to the observable characteristics of the system when the fault occurs, including overvoltage and overcurrent.

[0032] Furthermore, in step S3, the data attribute coordination mechanism is constructed, specifically comprising the steps of:

[0033] S31. Define node types and corresponding fault categories: Introduce a category tree, which is a tree structure used to represent the complex relationship between fault categories. The category tree starts at the bottom of the first layer and goes up layer by layer, with the nth layer at the top. The first layer marks nodes as "known" or "unknown" based on whether the corresponding fault category belongs to a known fault category. The second to nth layers mark nodes as "known", "unknown" or "other" based on the child node type. In the second to nth layers, if all the child nodes of a node are "known", it is defined as "known". If all the child nodes of a node are "unknown", it is defined as "unknown". Otherwise, it is defined as "other".

[0034] S32. Construct data similarity matrix: Use SimSiam network as feature extractor to extract features from each sample, calculate the feature mean of all samples in each category, and obtain the centroid set C of the kth layerk , specifically expressed as follows:

[0035] Assume that the training set of the i-th fault category is recorded as The centroid set C of the kth layer k By averaging The features of each label in are obtained:

[0036]

[0037] Among them, f() is the feature extractor, || represents the number of samples in the set;

[0038] Calculate whether the distance between the centroids of two categories is less than the threshold Constructing data similarity matrix

[0039]

[0040] Among them, c i ,c j is the centroid set C k Points in ||c i ,c j || means c i ,c j The Euclidean distance between them is used to determine whether two fault categories are close enough to be classified as the same type through a threshold. Since only the centroid set of defined faults can be obtained, all undefined faults are taken as 0;

[0041] S33, construct attribute similarity matrix: by calculating whether the cosine distance between two category attribute vectors is less than the threshold And whether their centroid distance is within the tolerance range, construct the attribute similarity matrix

[0042]

[0043] Among them, c i ,c j is the centroid set C k The point in represents the attribute vector of the i-th fault in the k-th layer, express The cosine distance between is another threshold to exclude the case where the attributes are similar but the data are dissimilar. The attribute vector is calculated as follows:

[0044] Calculate the mean of the attribute vectors of all undefined and other fault categories under each parent node to obtain the attribute vector of the k+1th layer:

[0045]

[0046] in, is the attribute vector of the i-th fault category in the k+1th layer, γ i The number of fault types j that satisfy This formula can calculate the attribute vectors from the second layer to the top layer, and the labels of the fault types in the first layer have been defined, and the attribute vectors can be obtained directly. A is the fault symptom description matrix;

[0047] S34. Data similarity matrix and attribute similarity matrix Perform element-by-element logical “or” operation to obtain the fused similarity matrix:

[0048]

[0049] S35. Use the reverse Cuthill McKee algorithm to rearrange and partition the fused matrix so that the non-zero elements of the rearranged matrix are clustered around the diagonal to form a block structure. According to the non-zero elements on the diagonal of the rearranged matrix, the matrix is divided into multiple blocks, each block representing a group of sibling fault categories;

[0050] S36. Based on the matrix partitioning result, the labels of the defined fault categories and the undefined fault categories of the k+1th layer are reallocated according to the definition:

[0051]

[0052] in, Indicates the parent class of the k-th layer fault category "*", It represents the subclass of the fault category “*” at the kth layer. If a node is not classified into “defined” and “undefined”, it is defined as “other”, so that the nodes at each layer of the category tree are assigned labels.

[0053] Furthermore, the step S4 specifically includes the steps of:

[0054] S41. Use the relevant features extracted through SPCA as input to create a dataset for training a symptom classifier;

[0055] S42. Use the RF (random forest) classifier to train the symptom classifier of the intelligent recognition model. The RF formula is as follows:

[0056] {h1(x),h2(x),...,h T (x)}

[0057] Among them, h i (x) represents a category prediction of the corresponding output for a given input x, and the final set is determined by majority voting:

[0058]

[0059] Here, I() represents the indicator function, which returns 1 if the i-th tree predicts the category y of the input x, and 0 otherwise.

[0060] Furthermore, in step S5, after feature extraction of the real-time data, fault diagnosis is performed using the trained symptom classifier and the fault symptom description matrix. During fault diagnosis, the nearest neighbor search algorithm is used to calculate the Euclidean distance to identify the nearest point in the feature space:

[0061]

[0062] Among them, x and y are the feature vectors of two sample points, x i and y i is the value of these sample points on the i-th feature;

[0063] The k-nearest neighbor algorithm is used to determine the fault type:

[0064]

[0065] Among them, y i are the labels of the k nearest neighbors of x.

[0066] On the other hand, the present application provides an intelligent circuit breaker fast tripping protection device based on illegal charging detection, comprising:

[0067] A data acquisition module is used to monitor and collect relevant charging data of charging equipment through intelligent circuit breakers;

[0068] Feature extraction module, used to pre-process the collected relevant charging data and perform feature extraction;

[0069] A collaborative mechanism building module is used to introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute collaborative mechanism;

[0070] The model training module is used to train an intelligent recognition model based on labeled data sets for fault types related to illegal charging.

[0071] The illegal charging detection module is used for real-time monitoring and detection of illegal charging. It inputs features into a trained intelligent recognition model to detect different types of illegal charging;

[0072] The violation execution module is used to take action according to the type of illegal charging if illegal charging is detected. If no illegal charging is detected, it returns to the function of executing the illegal charging detection module.

[0073] On the other hand, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the intelligent circuit breaker rapid tripping protection method based on illegal charging detection are implemented.

[0074] On the other hand, the present application further provides a storage medium, which includes a stored program, and when the program is run, controls the device where the storage medium is located to execute the steps of the intelligent circuit breaker fast tripping protection method based on illegal charging detection.

[0075] Compared with the existing technology, this application has the following beneficial effects:

[0076] The present application provides a method, device, electronic device and storage medium for rapid tripping protection of an intelligent circuit breaker based on illegal charging detection. The method for rapid tripping protection of an intelligent circuit breaker based on illegal charging detection proposed in the present application reduces the amount of data required for training the fault identification model by introducing zero-sample fault diagnosis based on transfer of symptom descriptions. Only historically observed fault data is needed to build a diagnostic model. By establishing a cross-domain mapping relationship from fault features to fault labels, as well as a symptom sharing relationship between defined fault types and undefined fault types, the undefined fault types can be effectively identified, thereby enhancing the robustness of the protection method. The present application can help the model better understand the relationship between fault categories through a fault category tree, thereby improving diagnostic accuracy. At the same time, the data attribute collaborative mechanism can cluster similar fault categories together, so that the model can better identify similar unseen faults. At the same time, the mechanism can automatically construct a fault category tree, reduce manual intervention, and improve the practicality of the method.

[0077] In addition to the above-described purposes, features and advantages, the present application also has other purposes, features and advantages. The present application will be further described in detail below with reference to the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The drawings that constitute a part of this application are used to provide further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute improper limitations on this application.

[0079] Figure 1 This is a flow chart of a method for rapid tripping protection of an intelligent circuit breaker based on illegal charging detection in a preferred embodiment of the present application.

[0080] Figure 2 This is a schematic diagram of a module of an intelligent circuit breaker fast tripping protection device based on illegal charging detection in a preferred embodiment of the present application.

[0081] Figure 3This is a schematic block diagram of an electronic device entity according to a preferred embodiment of the present application.

[0082] Figure 4 It is a diagram of the internal structure of a computer device according to a preferred embodiment of the present application. DETAILED DESCRIPTION

[0083] The embodiments of the present application are described in detail below with reference to the accompanying drawings, but the present application can be implemented in a variety of different ways defined and covered below.

[0084] like Figure 1 As shown, the preferred embodiment of the present application provides a smart circuit breaker fast tripping protection method based on illegal charging detection, comprising the steps of:

[0085] S1. Monitor and collect relevant charging data of charging equipment through intelligent circuit breakers;

[0086] S2. Preprocessing the collected relevant charging data and performing feature extraction;

[0087] S3. Introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute coordination mechanism;

[0088] S4. Train an intelligent recognition model on a labeled dataset for the fault types of illegal charging.

[0089] S5. Perform real-time monitoring and illegal charging detection, input the features into the trained intelligent recognition model, and detect different illegal charging;

[0090] S6. If illegal charging is detected, take action based on the type of illegal charging, such as when short circuit, leakage, or overvoltage: immediately cut off the power supply; when overload or undervoltage: delay cutting off the power supply; when reverse connection or ground wire disconnection: issue an alarm and maintain power supply; if no illegal charging is detected, return to step S5.

[0091] This embodiment provides a method for rapid tripping protection of intelligent circuit breakers based on illegal charging detection. This method reduces the amount of data required to train the fault identification model by introducing zero-sample fault diagnosis based on symptom description transfer. Only historically observed fault data is needed to build the diagnosis model. By establishing a cross-domain mapping relationship from fault features to fault labels, as well as a symptom sharing relationship between defined fault types and undefined fault types, the undefined fault types can be effectively identified, thereby enhancing the robustness of the protection method. This embodiment can help the model better understand the relationship between fault categories through the fault category tree, thereby improving diagnostic accuracy. At the same time, the data attribute collaborative mechanism can cluster similar fault categories together, enabling the model to better identify similar unseen faults. At the same time, the mechanism can automatically construct a fault category tree, reducing manual intervention and improving the practicality of the method.

[0092] In a preferred embodiment of the present application, in step S1, the relevant charging data is as follows:

[0093] X={I,V,T,t c ,SOC}

[0094] Among them, V represents the charging voltage, I represents the charging current, T represents the temperature, and t c represents the charging time, and SOC represents the state of charge of the charging device.

[0095] The relevant charging data in this embodiment includes important data such as charging current, temperature, charging time, and charge status of the charging device, providing sufficient data preparation for subsequent fault diagnosis.

[0096] In a preferred embodiment of the present application, in step S2, SPCA is used to extract features from relevant charging data, specifically including the following steps:

[0097] S21, for the feature matrix X∈R n×p , where n is the number of samples, p is the number of features, and each feature is standardized using the following formula:

[0098]

[0099] Among them, x j ′ represents the standardized data, x j represents the data before normalization, μ j represents the average value of the jth feature, σ j represents the standard deviation of the jth feature;

[0100] S22. Construct the covariance matrix C based on the standardized features. The formula is as follows:

[0101]

[0102] Among them, X T represents the transpose of matrix X;

[0103] S23. Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ k and the corresponding eigenvector v k , the formula is as follows:

[0104]

[0105] S24. Introduce sparsity constraints in PCA by solving the following optimization problem:

[0106]

[0107] Among them, u k represents the sparse principal component, v k represents the feature vector, α represents the regularization parameter that controls the sparsity level, and || ||1 represents the L1 norm;

[0108] S25. Use the obtained sparse principal components as input of the fault diagnosis model.

[0109] The purpose of steps S21 to S25 of this embodiment is to effectively reduce the data dimension while preserving the data's primary information, thereby reducing storage space and computational complexity. Advantage: Compared to traditional dimensionality reduction methods, this approach introduces sparsity while maintaining the maximum variance of the data, resulting in a greater number of zero elements in the principal component loading vectors, thus achieving better data interpretability.

[0110] In a preferred embodiment of the present application, in step S3, constructing a description of illegal charging symptoms and defining a set of illegal charging symptom descriptions specifically includes the following steps:

[0111] Define V = {V1, V2, ..., V n} is a set of defined faults, n=4, including overcharge, overheating, mismatch between charger and battery type, and battery aging. See Table 1 for details:

[0112] Table 1

[0113] Fault type Label Overcharge #1 overheat #2 The charger does not match the battery type #3 Battery aging #4

[0114] U={U1,U2,...,U m} is a set of undefined fault types, m is the number of undefined fault types; the sample set of V is expressed as X V Represents the defined training fault samples, N V is the number of samples, D is the dimension of the feature, YV Represents the defined training fault sample label, and the undefined fault type is represented by X U represents undefined training fault samples, N U is the number of samples, D is the dimension of the feature, Y U Represents the undefined training fault sample label, where Y U It is still an unknown quantity;

[0115] The fault symptom description matrix A∈R is constructed by V and U P×Q , P = n + m, Q represents the dimension of the symptom vector, and uses binary to indicate whether a symptom exists in a certain fault category. "1" indicates the presence of the symptom, and "0" indicates the absence of the symptom. Each fault should consist of k symptoms. The symptoms refer to the observable characteristics of the system when the fault occurs, including overvoltage and overcurrent.

[0116] In this embodiment, the purpose of constructing and defining a set of symptom descriptions for illegal charging is to use symptom descriptions to link abnormal characteristics (such as overvoltage and overcurrent) during the charging process with specific illegal charging behaviors (such as overcharging and overheating), providing a basis for subsequent fault identification. Benefits include: Symptom descriptions can be used to extract common features across different fault types, thereby reducing the amount of data required to train intelligent recognition models. Symptom descriptions can also be used to identify fault types not covered by training data, improving the model's robustness and generalization capabilities.

[0117] In a preferred embodiment of the present application, in step S3, the construction of the data attribute collaboration mechanism specifically includes the following steps:

[0118] S31. Define node types and corresponding fault categories: Introduce a category tree, which is a tree structure used to represent the complex relationship between fault categories. The category tree starts at the bottom of the first layer and goes up layer by layer, with the nth layer at the top. The first layer marks nodes as "known" or "unknown" based on whether the corresponding fault category belongs to a known fault category. The second to nth layers mark nodes as "known", "unknown" or "other" based on the child node type. In the second to nth layers, if all the child nodes of a node are "known", it is defined as "known". If all the child nodes of a node are "unknown", it is defined as "unknown". Otherwise, it is defined as "other".

[0119] S32. Construct data similarity matrix: Use SimSiam network as feature extractor to extract features from each sample, calculate the feature mean of all samples in each category, and obtain the centroid set C of the kth layer k , specifically expressed as follows:

[0120] Assume that the training set of the i-th fault category is recorded as The centroid set C of the kth layer k By averaging The features of each label in are obtained:

[0121]

[0122] Among them, f() is the feature extractor, || represents the number of samples in the set;

[0123] Calculate whether the distance between the centroids of two categories is less than the threshold Constructing data similarity matrix

[0124]

[0125] Among them, c i ,c j is the centroid set C k Points in ||c i ,c j || means c i ,c j The Euclidean distance between them is used to determine whether two fault categories are close enough to be classified as the same type through a threshold. Since only the centroid set of defined faults can be obtained, all undefined faults are taken as 0;

[0126] S33, construct attribute similarity matrix: by calculating whether the cosine distance between two category attribute vectors is less than the threshold And whether their centroid distance is within the tolerance range, construct the attribute similarity matrix

[0127]

[0128] Among them, c i ,c j is the centroid set C k The point in represents the attribute vector of the i-th fault in the k-th layer, express The cosine distance between is another threshold to exclude the case where the attributes are similar but the data are dissimilar. The attribute vector is calculated as follows:

[0129] Calculate the mean of the attribute vectors of all undefined and other fault categories under each parent node to obtain the attribute vector of the k+1th layer:

[0130]

[0131] in, is the attribute vector of the i-th fault category in the k+1th layer, γ i The number of fault types j that satisfy This formula can calculate the attribute vectors from the second layer to the top layer, and the labels of the fault types in the first layer have been defined, and the attribute vectors can be obtained directly. A is the fault symptom description matrix;

[0132] S34. Data similarity matrix and attribute similarity matrix Perform element-by-element logical “or” operation to obtain the fused similarity matrix:

[0133]

[0134] S35. Use the reverse Cuthill McKee algorithm to rearrange and partition the fused matrix so that the non-zero elements of the rearranged matrix are clustered around the diagonal to form a block structure. According to the non-zero elements on the diagonal of the rearranged matrix, the matrix is divided into multiple blocks, each block representing a group of sibling fault categories;

[0135] S36. Based on the matrix partitioning result, the labels of the defined fault categories and the undefined fault categories of the k+1th layer are reallocated according to the definition:

[0136]

[0137] in, Indicates the parent class of the k-th layer fault category "*", It represents the subclass of the fault category “*” at the kth layer. If a node is not classified into “defined” and “undefined”, it is defined as “other”, so that the nodes at each layer of the category tree are assigned labels.

[0138] The purpose of steps S31 through S36 in this embodiment is to construct a fault category tree that clearly displays the hierarchical relationships and similarities between different fault types, helping the model better understand fault types and improving diagnostic accuracy. Benefits include automatically clustering similar fault categories and constructing a fault category tree, reducing manual intervention and improving the practicality of the method. The model can also identify unseen faults similar to defined fault types, further improving the model's generalization capabilities.

[0139] In a preferred embodiment of the present application, step S4 specifically includes the following steps:

[0140] S41. Use the relevant features extracted through SPCA as input to create a dataset for training a symptom classifier;

[0141] S42. Use the RF (random forest) classifier to train the symptom classifier of the intelligent recognition model. The RF formula is as follows:

[0142] {h1(x),h2(x),...,h T (x)}

[0143] Among them, h i (x) represents a category prediction of the corresponding output for a given input x, and the final set is determined by majority voting:

[0144]

[0145] Here, I() represents the indicator function, which returns 1 if the i-th tree predicts the category y of the input x, and 0 otherwise.

[0146] The purpose of steps S41 to S42 of this embodiment is to train a symptom classifier to associate symptom descriptions with specific fault types, thereby enabling automatic fault identification. Advantages include: the random forest algorithm has strong robustness and good generalization capabilities, and can effectively identify different types of illegal charging behaviors.

[0147] In a preferred embodiment of the present application, in step S5, after feature extraction of the real-time data, fault diagnosis is performed using a trained symptom classifier and a fault symptom description matrix. During fault diagnosis, a nearest neighbor search algorithm is used to calculate the Euclidean distance to identify the nearest point in the feature space:

[0148]

[0149] Among them, x and y are the feature vectors of two sample points, x i and y i is the value of these sample points on the i-th feature;

[0150] The k-nearest neighbor algorithm is used to determine the fault type:

[0151]

[0152] Among them, y i are the labels of the k nearest neighbors of x.

[0153] The purpose of the above-mentioned specific steps in step S5 of this embodiment is to associate symptom descriptions with specific fault types by training the symptom classifier, thereby achieving automatic fault identification. Advantages include: the random forest algorithm has advantages such as strong robustness and good generalization ability, and can effectively identify different types of illegal charging behaviors.

[0154] like Figure 2As shown, the preferred embodiment of the present application further provides an intelligent circuit breaker fast tripping protection device based on illegal charging detection, comprising:

[0155] A data acquisition module is used to monitor and collect relevant charging data of charging equipment through intelligent circuit breakers;

[0156] Feature extraction module, used to pre-process the collected relevant charging data and perform feature extraction;

[0157] A collaborative mechanism building module is used to introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute collaborative mechanism;

[0158] The model training module is used to train an intelligent recognition model based on labeled data sets for the types of faults associated with illegal charging.

[0159] The illegal charging detection module is used for real-time monitoring and detection of illegal charging. It inputs features into a trained intelligent recognition model to detect different types of illegal charging;

[0160] The violation execution module is used to take action according to the type of illegal charging if illegal charging is detected. If no illegal charging is detected, it returns to the function of executing the illegal charging detection module.

[0161] like Figure 3 As shown, a preferred embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the intelligent circuit breaker fast tripping protection method based on illegal charging detection in the above embodiment are implemented.

[0162] like Figure 4 As shown, the preferred embodiment of the present application further provides a computer device, which can be a terminal or a liveness detection server, and its internal structure diagram can be as shown in FIG. Figure 4 As shown. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with other external computer devices via a network connection. When the computer program is executed by the processor, the steps of the above-mentioned intelligent circuit breaker rapid tripping protection method based on illegal charging detection are implemented.

[0163] Those skilled in the art will understand that Figure 4The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0164] A preferred embodiment of the present application further provides a storage medium, which includes a stored program. When the program is executed, the device where the storage medium is located is controlled to execute the steps of the intelligent circuit breaker fast tripping protection method based on illegal charging detection in the above embodiment.

[0165] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0166] If the functions described in the method of this embodiment are implemented in the form of a software functional unit and sold or used as an independent product, they can be stored in a storage medium readable by one or more computing devices. Based on this understanding, the part of the embodiment of the present application that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for enabling a computing device (which can be a personal computer, server, mobile computing device or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0167] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal translation scripting language JavaScript, etc.

[0168] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0169] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0170] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0171] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0172] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A smart circuit breaker fast tripping protection method based on illegal charging detection, characterized in that: Including steps: S1. Monitor and collect relevant charging data of charging equipment through intelligent circuit breakers; S2. Preprocessing the collected relevant charging data and performing feature extraction; S3. Introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute coordination mechanism; S4. Train an intelligent recognition model on a labeled dataset for the fault types of illegal charging. S5. Perform real-time monitoring and illegal charging detection, input the features into the trained intelligent recognition model, and detect different illegal charging; S6. If illegal charging is detected, take action according to the illegal charging type. If no illegal charging is detected, return to step S5.

2. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 1 is characterized in that: In step S1, the relevant charging data is as follows: X={I,V,T,t c ,SOC} Among them, V represents the charging voltage, I represents the charging current, T represents the temperature, and t c represents the charging time, and SOC represents the state of charge of the charging device.

3. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 2 is characterized in that: In step S2, SPCA is used to extract features from relevant charging data, specifically including the following steps: S21, for the feature matrix X∈R n×p , where n is the number of samples, p is the number of features, and each feature is standardized using the following formula: Among them, x j ′ represents the standardized data, x j represents the data before normalization, μ j represents the average value of the jth feature, σ j represents the standard deviation of the jth feature; S22. Construct the covariance matrix C based on the standardized features. The formula is as follows: Among them, X T represents the transpose of matrix X; S23. Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalue λ k and the corresponding eigenvector v k , the formula is as follows: S24. Introduce sparsity constraints in PCA by solving the following optimization problem: Among them, u k represents the sparse principal component, v k represents the feature vector, α represents the regularization parameter that controls the sparsity level, and ||||1 represents the L1 norm; S25. Use the obtained sparse principal components as input of the fault diagnosis model.

4. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 2 is characterized in that: In step S3, a description of illegal charging symptoms is constructed, and a set of descriptions of illegal charging symptoms is defined, which specifically includes the following steps: Define V={V1,V2,...,V n } is a set of defined faults, n = 4, including overcharge, overheating, mismatch between charger and battery type, and battery aging, U = {U1, U2, ..., U m } is a set of undefined fault types, m is the number of undefined fault types; the sample set of V is expressed as X V Represents the defined training fault samples, N V is the number of samples, D is the dimension of the feature, Y V Represents the defined training fault sample label, and the undefined fault type is represented by X U represents undefined training fault samples, N U is the number of samples, D is the dimension of the feature, Y U Represents the undefined training fault sample label, where Y U It is still an unknown quantity; The fault symptom description matrix A∈R is constructed by V and U P×Q , P = n + m, Q represents the dimension of the symptom vector, and uses binary to indicate whether a symptom exists in a certain fault category. "1" indicates the presence of the symptom, and "0" indicates the absence of the symptom. Each fault should consist of k symptoms, which refer to the observable characteristics of the system when the fault occurs, including overvoltage and overcurrent.

5. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 1 is characterized in that: In step S3, the data attribute coordination mechanism is constructed, specifically including the following steps: S31. Define node types and corresponding fault categories: Introduce a category tree, a tree structure used to represent the complex relationships between fault categories. The category tree starts at the bottom of the first layer and goes up layer by layer, with the nth layer at the top. The first layer marks nodes as "known" or "unknown" based on whether the corresponding fault category belongs to a known fault category. The second to nth layers mark nodes as "known", "unknown", or "other" based on the child node type. In the second to nth layers, if all the child nodes of a node are "known", it is defined as "known". If all the child nodes of a node are "unknown", it is defined as "unknown". Otherwise, it is defined as "other". S32. Construct data similarity matrix: Use SimSiam network as feature extractor to extract features from each sample, calculate the feature mean of all samples in each category, and obtain the centroid set C of the kth layer k , specifically expressed as follows: Assume that the training set of the i-th fault category is recorded as The centroid set C of the kth layer k By averaging The features of each label in are obtained: Among them, f() is the feature extractor, || represents the number of samples in the set; Calculate whether the distance between the centroids of two categories is less than the threshold Constructing data similarity matrix Among them, c i ,c j is the centroid set C k Points in ||c i ,c j || means c i ,c j The Euclidean distance between them is used to determine whether two fault categories are close enough to be classified as the same type through a threshold. Since only the centroid set of defined faults can be obtained, all undefined faults are taken as 0; S33, construct attribute similarity matrix: by calculating whether the cosine distance between two category attribute vectors is less than the threshold And whether their centroid distance is within the tolerance range, construct the attribute similarity matrix Among them, c i ,c j is the centroid set C k The point in represents the attribute vector of the i-th fault in the k-th layer, express The cosine distance between is another threshold to exclude the case where the attributes are similar but the data are dissimilar. The attribute vector is calculated as follows: Calculate the mean of the attribute vectors of all undefined and other fault categories under each parent node to obtain the attribute vector of the k+1th layer: in, is the attribute vector of the i-th fault category in the k+1th layer, γ i The number of fault types j that satisfy This formula can calculate the attribute vectors from the second layer to the top layer, and the labels of the fault types in the first layer have been defined, and the attribute vectors can be obtained directly. A is the fault symptom description matrix; S34. Data similarity matrix and attribute similarity matrix Perform element-by-element logical "OR" operation to obtain the fused similarity matrix: S35. Use the reverse Cuthill McKee algorithm to rearrange and partition the fused matrix so that the non-zero elements of the rearranged matrix are clustered around the diagonal to form a block structure. According to the non-zero elements on the diagonal of the rearranged matrix, the matrix is divided into multiple blocks, each block representing a group of sibling fault categories; S36. Based on the matrix partitioning result, the labels of the defined fault categories and the undefined fault categories of the k+1th layer are reallocated according to the definition: in, Indicates the parent class of the k-th layer fault category "*", It represents the subclass of the fault category "*" at the kth level. If a node is not classified into "defined" and "undefined", it is defined as "other", so that the nodes at each level of the category tree are assigned labels.

6. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 1 is characterized in that: The step S4 specifically includes the following steps: S41. Use the relevant features extracted through SPCA as input to create a dataset for training a symptom classifier; S42. Use the RF classifier to train the symptom classifier of the intelligent recognition model. The RF formula is as follows: {h1(x),h2(x),...,h T (x)} Among them, h i (x) represents a category prediction of the corresponding output for a given input x, and the final set is determined by majority voting: Here, I() represents the indicator function, which returns 1 if the i-th tree predicts the category y of the input x, and 0 otherwise.

7. The intelligent circuit breaker fast tripping protection method based on illegal charging detection according to claim 1 is characterized in that In step S5, after feature extraction of the real-time data, fault diagnosis is performed using the trained symptom classifier and the fault symptom description matrix. During fault diagnosis, the nearest neighbor search algorithm is used to calculate the Euclidean distance to identify the nearest point in the feature space. Among them, x and y are the feature vectors of two sample points, x i and y i is the value of these sample points on the i-th feature; The k-nearest neighbor algorithm is used to determine the fault type: Among them, y i are the labels of the k nearest neighbors of x.

8. An intelligent circuit breaker fast tripping protection device based on illegal charging detection, characterized in that: include: A data acquisition module is used to monitor and collect relevant charging data of charging equipment through intelligent circuit breakers; Feature extraction module, used to pre-process the collected relevant charging data and perform feature extraction; A collaborative mechanism building module is used to introduce a fault category tree to construct a description of illegal charging symptoms, define a set of symptom descriptions of illegal charging, and build a data attribute collaborative mechanism; The model training module is used to train an intelligent recognition model based on labeled data sets for fault types related to illegal charging. The illegal charging detection module is used for real-time monitoring and detection of illegal charging. It inputs features into a trained intelligent recognition model to detect different types of illegal charging; The violation execution module is used to take action according to the type of illegal charging if illegal charging is detected. If no illegal charging is detected, it returns to the function of executing the illegal charging detection module.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the intelligent circuit breaker fast tripping protection method based on illegal charging detection as described in any one of claims 1 to 7 are implemented.

10. A storage medium comprising a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the steps of the intelligent circuit breaker fast tripping protection method based on illegal charging detection according to any one of claims 1 to 7.

Images