High-voltage circuit breaker mechanical fault diagnosis method, device and equipment and storage medium

Abstract

The invention relates to the technical field of fault diagnosis, and discloses a high-voltage circuit breaker mechanical fault diagnosis method, device and equipment and a storage medium, and the method comprises the steps: obtaining a fault vibration signal of a high-voltage circuit breaker; performing parallel decomposition on the fault vibration signal to obtain a plurality of intrinsic mode functions, and forming low-dimension parallel sub-data; extracting time domain and frequency domain energy entropy features of the intrinsic mode function to obtain feature vectors; inputting the feature vector into a hybrid classifier model, and performing fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model; wherein the hybrid classifier model comprises a single-class support vector machine and a probabilistic neural network, the single-class support vector machine is used for identifying whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used for identifying a fault category corresponding to an abnormal sample. The fault of the high-voltage circuit breaker can be accurately judged and detected, and the reliability and stability of a power system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a method, device, equipment and storage medium for diagnosing mechanical faults of high-voltage circuit breakers. Background Art

[0002] High-voltage circuit breakers play a vital role in power systems, and their reliable operation is essential to ensure the safety and stability of power systems. However, with the continuous development of power systems and the long-term operation of high-voltage circuit breakers, various faults may occur in circuit breakers. Traditional high-voltage circuit breaker fault detection methods often have problems such as low accuracy and insufficient ability to distinguish between minor fault types.

[0003] At present, the fault identification and detection of high-voltage circuit breakers mainly rely on manual inspections and some simple electrical parameter monitoring. These methods are difficult to accurately detect mechanical faults inside circuit breakers. Moreover, the existing fault detection technology is inefficient in processing large and complex vibration signal data, and cannot effectively extract key feature information from vibration signals, making it difficult to accurately distinguish different fault types.

[0004] Therefore, there is an urgent need for a more efficient and accurate high-voltage circuit breaker fault identification and detection method that can use the fault vibration signal generated by the circuit breaker to accurately identify the fault and improve the reliability and stability of the power system.

[0005] The above information is presented as background information only to assist with understanding the present disclosure and no determination or admission is made as to whether any of the above may be used as prior art with respect to the present disclosure. Summary of the invention

[0006] The present invention provides a high-voltage circuit breaker mechanical fault diagnosis method, device, equipment and storage medium to solve the problems existing in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for diagnosing mechanical faults of high-voltage circuit breakers, comprising:

[0009] Obtaining fault vibration signals of high voltage circuit breakers;

[0010] Decomposing the fault vibration signal in parallel to obtain a number of intrinsic mode functions, and forming low-dimensional parallel sub-data;

[0011] Extracting the time domain and frequency domain energy entropy characteristics of the intrinsic mode function to obtain a feature vector;

[0012] Inputting the feature vector into a hybrid classifier model, and performing fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model;

[0013] The hybrid classifier model includes a single-class support vector machine and a probabilistic neural network. The single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

[0014] Optionally, obtaining a fault vibration signal of a high-voltage circuit breaker includes:

[0015] The vibration signal generated by the high-voltage circuit breaker is collected by a signal acquisition circuit module, and the signal acquisition circuit module has a noise filtering function;

[0016] The collected vibration signal is converted into a digital signal through the analog-to-digital conversion module.

[0017] Optionally, the parallel decomposition of the fault vibration signal to obtain a plurality of intrinsic mode functions and compose low-dimensional parallel sub-data includes:

[0018] Determine to decompose the input signal into several modal components with specific center frequencies and limited bandwidths, decompose the original signal into several IMF components and construct a constrained variational problem with the goal of minimizing the bandwidth of the modal components;

[0019] The augmented Lagrangian function is introduced, and the alternating direction multiplier method is used to solve the saddle point of the variational problem. The modal components, center frequencies and Lagrangian multipliers are updated alternately in sequence, and the multiplier alternating direction method is used to solve and update alternately.

[0020] Repeat the iterations and when the convergence conditions are met, several intrinsic mode functions with different frequency characteristics are output to form low-dimensional parallel sub-data.

[0021] Optionally, extracting the time domain and frequency domain energy entropy features of the intrinsic mode function to obtain a feature vector includes:

[0022] The eigenmode functions obtained by the variational mode decomposition method are combined into a component matrix, where the number of rows is the number of modes and the number of columns is the signal length;

[0023] Divide the component matrix into multiple sub-matrices along the time axis. Each sub-matrix contains multiple time-frequency blocks. Calculate the energy of the time-frequency block and normalize it to the interval [0,1].

[0024] Calculate the time domain energy entropy of each submatrix, normalize the total energy of each mode, and then calculate the frequency domain energy entropy of the entire component matrix;

[0025] The time domain and frequency domain energy entropy features are concatenated into a vector as the feature vector.

[0026] Optionally, inputting the feature vector into a hybrid classifier model, and performing fault diagnosis on the high-voltage circuit breaker based on the feature vector by using the hybrid classifier model, comprises:

[0027] The feature vector is input into a single-class support vector machine. If the single-class support vector machine determines that it is a normal sample, the output is a normal state; if the single-class support vector machine determines that it is an abnormal sample, it is sent to a probabilistic neural network;

[0028] The probabilistic neural network is used to determine different fault types based on the feature vectors corresponding to the abnormal samples.

[0029] Optionally, the high-voltage circuit breaker mechanical fault diagnosis method further comprises constructing a hybrid classifier model by the following steps:

[0030] Acquire historical fault vibration signals, and process the historical fault vibration signals to obtain historical feature vectors;

[0031] The historical feature vectors are input into the training process of the hybrid classifier model to obtain the predicted labels, and the hybrid classifier model is constructed according to the predicted labels.

[0032] Optionally, the high-voltage circuit breaker mechanical fault diagnosis method further includes:

[0033] Visualize the fault diagnosis results.

[0034] The present invention further provides a high-voltage circuit breaker mechanical fault diagnosis device, which is used to implement the high-voltage circuit breaker mechanical fault diagnosis method as described in any one of the above items, comprising:

[0035] A signal acquisition module, used to acquire a fault vibration signal of a high-voltage circuit breaker;

[0036] A parallel decomposition module, used for parallel decomposing the fault vibration signal to obtain a plurality of intrinsic mode functions and compose low-dimensional parallel sub-data;

[0037] A feature extraction module is used to extract the time domain and frequency domain energy entropy features of the intrinsic mode function to obtain a feature vector;

[0038] A fault diagnosis module is used to input the feature vector into a hybrid classifier model, and perform fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model, wherein the hybrid classifier model includes a single-class support vector machine and a probabilistic neural network, the single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

[0039] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the high-voltage circuit breaker mechanical fault diagnosis method as described in any one of the above items when executing the computer program.

[0040] The present invention also provides a storage medium containing computer executable instructions, wherein the computer executable instructions are executed by a computer processor to implement the high-voltage circuit breaker mechanical fault diagnosis method as described in any one of the above items.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention provides a method, device, equipment and storage medium for diagnosing mechanical faults of high-voltage circuit breakers. First, a parallel decomposition method is used to decompose the vibration signal of the high-voltage circuit breaker into a number of intrinsic mode functions with physical significance, and a large and complex amount of serial data is converted into low-dimensional parallel sub-data, thereby improving data processing efficiency and making subsequent feature extraction and fault diagnosis more accurate.

[0043] Secondly, the improved time segment energy entropy method is used to extract the energy entropy characteristics of the signal in the time domain and frequency domain, which enhances the ability to distinguish minor fault types and can more accurately identify different fault types.

[0044] Finally, an optimized hybrid classifier model consisting of a single-class support vector machine and a probabilistic neural network is used to identify the vibration signal type of the high-voltage circuit breaker, which achieves accurate identification and detection of high-voltage circuit breaker faults and improves the reliability and stability of the power system.

[0045] The present invention has other features and advantages, which will be apparent from or will be described in detail in the accompanying drawings and the following detailed description incorporated herein, which together serve to explain the specific principles of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 It is a flow chart of a high-voltage circuit breaker mechanical fault diagnosis method provided by the present invention;

[0048] Figure 2It is a schematic diagram of a component matrix in a high-voltage circuit breaker mechanical fault diagnosis method provided by the present invention;

[0049] Figure 3 It is a structural block diagram of a high-voltage circuit breaker mechanical fault diagnosis device provided by the present invention;

[0050] Figure 4 The present invention provides a circuit schematic diagram of a high-voltage circuit breaker mechanical fault diagnosis device.

[0051] Reference numerals: 10, signal acquisition module; 20, parallel decomposition module; 30, feature extraction module; 40, fault diagnosis module. DETAILED DESCRIPTION

[0052] In order to explain in detail the possible application scenarios, technical principles, specific schemes that can be implemented, and the purposes and effects that can be achieved, the following is a detailed description of the specific embodiments listed in conjunction with the accompanying drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.

[0053] Reference to "embodiment" herein means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The term "embodiment" appearing in various places in the specification does not necessarily refer to the same embodiment, nor does it particularly limit its independence or association with other embodiments. In principle, in the present application, as long as there is no technical contradiction or conflict, the various technical features mentioned in the embodiments can be combined in any way to form a corresponding implementable technical solution.

[0054] Unless otherwise defined, the technical terms used in this document have the same meanings as those generally understood by those skilled in the art to which this application belongs; the use of relevant terms in this document is only for describing specific embodiments and is not intended to limit this application.

[0055] In the description of this application, the term "and / or" is an expression used to describe the logical relationship between objects, indicating that three relationships may exist, for example, A and / or B, which means: A exists, B exists, and A and B exist at the same time. In addition, the character " / " in this article generally indicates that the objects before and after are in an "or" logical relationship.

[0056] In the present application, terms such as “first” and “second” are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship of quantity, priority or sequence between these entities or operations.

[0057] Without further limitations, in this application, the words "include", "comprises", "has" or other similar expressions used in the sentences are intended to cover non-exclusive inclusion. These expressions do not exclude the presence of additional elements in the process, method or product including the elements, so that the process, method or product including a series of elements may include not only those limited elements, but also other elements not explicitly listed, or also include elements inherent to such process, method or product.

[0058] Similar to the understanding in the Examination Guidelines, in this application, expressions such as "greater than", "less than", "exceed" and the like are understood to exclude the number itself; expressions such as "above", "below", "within" and the like are understood to include the number itself. In addition, in the description of the embodiments of this application, "multiple" means more than two (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups", "multiple times", etc., unless otherwise clearly and specifically limited.

[0059] In the description of the embodiments of the present application, space-related expressions used, such as "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "vertical", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicate the orientation or position relationship based on the orientation or position relationship shown in the specific embodiments or drawings, and are only for the convenience of describing the specific embodiments of the present application or facilitating the reader's understanding, and do not indicate or imply that the referred device or component must have a specific position, a specific orientation, or be constructed or operated in a specific orientation, and therefore cannot be understood as a limitation on the embodiments of the present application.

[0060] Unless otherwise expressly specified or limited, in the description of the embodiments of the present application, the terms such as "install", "connect", "connect", "fix", "set", etc. used should be understood in a broad sense. For example, the "connection" can be a fixed connection, a detachable connection, or an integrated setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. For technicians in the technical field to which the present application belongs, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0061] High-voltage circuit breakers are an important type of switchgear that has the dual function of protecting power systems. However, high-voltage circuit breaker failures continue to occur, causing economic losses. A survey of high-voltage circuit breaker failures by the International Council on Large Electric Power Systems (CIGRE) showed that operating mechanism failures accounted for 61%. Specifically, 39% of minor failures and 44% of major failures were caused by mechanical degradation. Such as mechanism jamming, loose screws, insufficient lubrication leading to delays, and insufficient spring energy storage. Therefore, it is of great significance to study the fault diagnosis of high-voltage circuit breakers.

[0062] The present invention aims to provide a diagnostic solution for a high voltage circuit breaker to improve the safety and stability of a power system.

[0063] Embodiment 1

[0064] Please refer to Figure 1 The embodiment of the present invention provides a method for diagnosing mechanical faults of high-voltage circuit breakers, comprising:

[0065] S1. Obtaining a fault vibration signal of a high voltage circuit breaker;

[0066] S2, parallel decomposing the fault vibration signal to obtain several intrinsic mode functions and composing low-dimensional parallel sub-data;

[0067] S3, extracting the energy entropy characteristics of the intrinsic mode function in time domain and frequency domain to obtain a feature vector;

[0068] S4. Input the feature vector into the hybrid classifier model, and perform fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model.

[0069] In step S4, the hybrid classifier model includes a single-class support vector machine and a probabilistic neural network. The single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

[0070] In this embodiment, the fault vibration signal generated by the circuit breaker is used to perform fault identification and detection, the fault vibration signal of the high-voltage circuit breaker is used as input, and the high-voltage circuit breaker fault classification identification is output.

[0071] Specifically, the vibration signal of the high-voltage circuit breaker is decomposed into several intrinsic mode functions with physical meaning using the variational mode parallel decomposition method, which converts the originally large and complex serial data into low-dimensional parallel sub-data. Then, the energy entropy characteristics of the signal in the time domain and frequency domain are extracted using the improved time segment energy entropy method to enhance the ability to distinguish between minor fault types. Finally, an optimized hybrid classifier model consisting of a one-class support vector machine (OCSVM) and a probabilistic neural network (PNN) is used to identify the four types of vibration signals of the high-voltage circuit breaker.

[0072] The following is an introduction to each step in this embodiment.

[0073] In this embodiment, step S1: obtaining a fault vibration signal of a high-voltage circuit breaker includes:

[0074] The vibration signal generated by the high-voltage circuit breaker is collected by the signal acquisition circuit module, and the signal acquisition circuit module has the function of filtering noise;

[0075] The collected vibration signal is converted into a digital signal through the analog-to-digital conversion module.

[0076] It can be understood that, on the one hand, the vibration signal generated by the high-voltage circuit breaker is collected by the signal acquisition circuit module, and the module has the function of filtering noise to ensure that the acquired vibration signal reflects the actual operating status of the high-voltage circuit breaker as realistically as possible, reduces the interference of noise on subsequent analysis, and improves the accuracy of fault detection.

[0077] On the other hand, the collected vibration signals are converted into digital signals through the analog-to-digital conversion module, so that the vibration signals can be recognized and processed by the digital processing system, providing the prerequisite for the subsequent use of various digital signal processing methods for fault analysis and identification.

[0078] Furthermore, in step S2, the fault vibration signal is decomposed in parallel to obtain a number of intrinsic mode functions, and form low-dimensional parallel sub-data, including:

[0079] Determine to decompose the input signal into several modal components with specific center frequencies and limited bandwidths, decompose the original signal into several IMF components and construct a constrained variational problem with the goal of minimizing the bandwidth of the modal components;

[0080] The augmented Lagrangian function is introduced, and the alternating direction multiplier method is used to solve the saddle point of the variational problem. The modal components, center frequencies and Lagrangian multipliers are updated alternately in sequence, and the multiplier alternating direction method is used to solve and update alternately.

[0081] Repeat the iterations and when the convergence conditions are met, several intrinsic mode functions with different frequency characteristics are output to form low-dimensional parallel sub-data.

[0082] The purpose of step S2 is that since the pooling operation loses the position information of the feature, this affects the time series that is sensitive to the data series; this embodiment uses the attention mechanism to help the neural network learn the weight distribution and select the information that is more important to the current task goal. At the same time, in view of the weakening of the model convergence by the data characteristics of the vibration signal of the high-voltage circuit breaker, a variational mode parallel decomposition method is further proposed, combined with subsequent rectangular processing, to convert the original single channel into a parallel multi-channel mode, thereby improving the model iteration efficiency.

[0083] In this step, a one-dimensional vibration signal of a high-voltage circuit breaker is used as input, and a number of intrinsic mode functions (IMFs) are output, each of which represents a specific frequency component in the input signal.

[0084] Specifically, the principle of this step is as follows:

[0085] Assume that the input signal can be decomposed into K modal components with finite bandwidth and center frequency, each modal component is a center frequency of ω k The bandwidth of the original f t is decomposed into K IMF components. A constrained variational problem is constructed to minimize the bandwidth of each modal component.

[0086] (1)

[0087] Where * represents convolution, u k (t) is the modular function, ω k is the corresponding center frequency, K is the modulus, Indicates u k (analytic signal of t).

[0088] In order to obtain the optimal solution of equation (1), an augmented Lagrangian function is introduced, and the alternating direction multiplier method (ADMM) is used to solve the saddle point of the variational problem, and the modal components, central frequencies and Lagrangian multipliers are updated alternately.

[0089] (2)

[0090] where α is the quadratic multiplication factor and λ(t) is the Lagrange multiplier.

[0091] The saddle point of the above augmented Lagrangian function is solved by the alternating direction method of multipliers. Update alternately.

[0092]

[0093]

[0094]

[0095] The symbol ^ represents Fourier transform, Can be used as The real part of the inverse Fourier transform of .

[0096] Repeat the above iterations, and for a given e > 0, stop the iteration if the following conditions are met:

[0097]

[0098] If the condition is met, the iteration stops and the results of K narrowband IMF components are output, otherwise return to step (2). Repeat the iteration until the convergence condition is met and K intrinsic mode functions (IMFs) with different frequency characteristics are output.

[0099] Furthermore, in step S3, the time domain and frequency domain energy entropy features of the intrinsic mode function are extracted to obtain a feature vector, including:

[0100] The eigenmode functions obtained by the variational mode decomposition method are combined into a component matrix, where the number of rows is the number of modes and the number of columns is the signal length;

[0101] Divide the component matrix into multiple sub-matrices along the time axis. Each sub-matrix contains multiple time-frequency blocks. Calculate the energy of the time-frequency block and normalize it to the interval [0,1].

[0102] Calculate the time domain energy entropy of each submatrix, normalize the total energy of each mode, and then calculate the frequency domain energy entropy of the entire component matrix;

[0103] The time domain and frequency domain energy entropy features are concatenated into a vector as the feature vector.

[0104] The purpose of step S3 is to distinguish four working conditions, namely, normal working condition, closing spring fatigue fault, opening spring fatigue fault and iron core blocking fault, by effectively identifying faults on the field data.

[0105] In this step, K intrinsic mode functions (IMFs) obtained by the variational mode decomposition (VMD) method are first input, and then a vector containing energy entropy characteristics in the time domain and frequency domain is output to describe the energy distribution of the signal.

[0106] The specific principles are as follows:

[0107] like Figure 2 As shown, K intrinsic mode functions (IMFs) are organized into a component matrix, the number of rows in the matrix is ​​the number of modes, and the number of columns is the signal length (i.e., the number of sampling points).

[0108] The component matrix is ​​divided into 20 sub-matrices along the time axis. Each sub-matrix contains 7 time-frequency blocks. The energy of the time-frequency block represents the activity of the signal in the time-frequency region.

[0109] All time-frequency blocks Energy Normalize to the interval [0,1]. The normalization formula is:

[0110]

[0111] Calculate the time domain energy entropy of each sub-matrix to reflect the energy distribution characteristics of the signal in time. Where E is the energy of the entire component matrix. j As part of the signal characteristic energy distribution (time period energy entropy),

[0112]

[0113] The total energy of each mode is normalized, and then the frequency domain energy entropy of the entire component matrix is ​​calculated to reflect the energy distribution characteristics of the signal in frequency. is the total energy of each sub-event, then Q i As another part of the signal characteristic energy distribution (frequency band energy entropy):

[0114]

[0115] The time domain and frequency domain energy entropy features are concatenated into a vector as the feature vector of the signal for subsequent fault identification.

[0116]

[0117] I v Used as input vector of hybrid classifier to diagnose high voltage circuit breaker faults.

[0118] Furthermore, step S4 also includes constructing a hybrid classifier model by adopting the following steps:

[0119] Acquire historical fault vibration signals, and process the historical fault vibration signals to obtain historical feature vectors;

[0120] The historical feature vectors are input into the training process of the hybrid classifier model to obtain the predicted labels, and the hybrid classifier model is constructed according to the predicted labels.

[0121] The purpose of step S4 is to distinguish four working conditions, namely, normal working condition, closing spring fatigue fault, opening spring fatigue fault and iron core blocking fault, by effectively identifying faults on the field data.

[0122] Specifically, a set of eigenvectors is input, each of which corresponds to a vibration signal of a high voltage circuit breaker (HVCB). The eigenvector dimension is 27, the first 20 dimensions of the eigenvector reflect the energy distribution of the signal in the time domain, and the last 7 dimensions reflect the energy distribution of the signal in the frequency domain.

[0123] The output of the hybrid classifier model is then a set of prediction labels, each of which corresponds to a mechanical fault type of the HVCB. The prediction label has four possible values: normal state (Normal), loose base screws (Fault I), delayed action (Fault II), and buffer spring detachment (Fault III).

[0124] The specific principles are as follows:

[0125] First, the input feature vector is sent to OCSVM. If OCSVM judges it as a normal sample, the output is Normal; if OCSVM judges it as an abnormal sample, it is sent to PNN and determined as Fault I, Fault II or Fault III according to the output of PNN.

[0126] OCSVM is the training data set, then X contains the extracted n×m one-dimensional feature vectors. OCSVM requires a kernel function to transform the original low-dimensional feature space R n×m Mapping to high-dimensional feature space H: , then establish a high-dimensional space hyperplane It is necessary to solve the quadratic programming problem shown in equations (12) and (13):

[0127]

[0128]

[0129] Where w and b represent the normal vector and the intercept of the hyperplane respectively, ζ i is a slack variable, the error limit v ∈ (0, 1) is used to control the upper limit of the number of outliers in the training set, x i is an m-dimensional feature vector.

[0130] To solve the quadratic programming problem, the Lagrangian function is given as follows:

[0131]

[0132] where a i ≥ 0 and β i ≥ 0 is a Lagrange multiplier.

[0133] For variables w, ζ i Find the partial derivatives of and b and make them equal to 0, refer to the following equation:

[0134]

[0135] By solving and simplifying the above equations, we can obtain the following dual problem:

[0136]

[0137]

[0138] in, is the kernel function.

[0139] Finally, the classification decision function of OCSVM is calculated as:

[0140]

[0141] in, , which represents the symbolic function.

[0142] For a given test sample z, use the OCSVM function to make a decision. If f z If it is greater than 0, the test sample will be judged as a normal sample. Otherwise, it will be judged as an abnormal sample.

[0143] When determining an abnormal sample, enter the PNN network, which consists of an input layer, a pattern layer, a summation layer, and an output layer. Assume that the input vector to be classified is X = [x 1 , x 2 ,…, x n ] is the input layer. The number of neurons in the pattern layer is the same as the number of training samples. The summation layer simply adds the outputs of the training samples of the same category in the pattern layer. The decision layer is Y= [y 1 ,y 2 , · · · , y k ], and its output is binary output. That is, the corresponding position of the output layer will output 1, otherwise it will output 0.

[0144] Further, in step S4, the feature vector is input into a hybrid classifier model, and the hybrid classifier model is used to perform fault diagnosis on the high-voltage circuit breaker based on the feature vector, including:

[0145] The feature vector is input into the one-class support vector machine. If the one-class support vector machine determines that it is a normal sample, the output is a normal state; if the one-class support vector machine determines that it is an abnormal sample, it is sent to the probabilistic neural network;

[0146] The probabilistic neural network is used to determine different fault types based on the feature vectors corresponding to the abnormal samples.

[0147] The advantage of OCSVM is that it can effectively identify normal samples. By inputting the feature vector into OCSVM, high-voltage circuit breakers in normal state can be quickly distinguished. This helps to reduce the amount of subsequent complex classification calculations, because in power systems, the number of samples in normal state is often large.

[0148] When OCSVM determines that a sample is abnormal, it starts the probabilistic neural network (PNN) for further classification. This two-level classification structure can classify faults more accurately, because OCSVM and PNN each play their own advantages. OCSVM focuses on the coarse classification of normal / abnormal, while PNN focuses on the fine classification of abnormal samples, thereby determining different fault types.

[0149] In addition, by processing the feature vectors corresponding to abnormal samples through probabilistic neural networks, the ability of PNN in dealing with complex classification problems can be fully utilized. PNN can accurately identify different fault types of high-voltage circuit breakers, such as closing spring fatigue fault, opening spring fatigue fault, or iron core jamming fault, based on the input feature vectors and combined with its trained pattern recognition capabilities. This is crucial for taking targeted maintenance measures in a timely manner and ensuring the stable operation of the power system.

[0150] It can be understood that the use of this hybrid classifier model combines the advantages of two different classifiers. The combination of OCSVM's effective recognition of normal samples and PNN's detailed classification of abnormal samples can effectively reduce the misjudgment rate and improve the accuracy and reliability of the entire fault diagnosis system, thereby providing a stronger guarantee for the safe operation of the power system.

[0151] Furthermore, the high-voltage circuit breaker mechanical fault diagnosis method in this embodiment further includes:

[0152] The fault diagnosis results are displayed visually, including the fault type and location, so that the staff can quickly handle the fault.

[0153] Please refer to Figure 3 and Figure 4 Based on the above embodiments, an embodiment of the present invention provides a high-voltage circuit breaker mechanical fault diagnosis device, which is used to implement the high-voltage circuit breaker mechanical fault diagnosis method in the above embodiments.

[0154] In terms of hardware, the hardware part is mainly divided into signal acquisition circuit module, analog-to-digital conversion and data analysis module.

[0155] Among them, the signal acquisition module is used to collect the vibration signal generated by the high-voltage circuit breaker and has the function of filtering noise.

[0156] The analog-to-digital conversion module is used to convert the collected vibration signal into a digital signal.

[0157] The data analysis module is used to detect whether a series fault occurs.

[0158] According to the function, the high voltage circuit breaker mechanical fault diagnosis device can be divided into:

[0159] The signal acquisition module 10 is used to acquire the fault vibration signal of the high-voltage circuit breaker in order to collect the vibration signal generated by the high-voltage circuit breaker and has the function of filtering noise;

[0160] A parallel decomposition module 20 is used to perform parallel decomposition on the fault vibration signal to obtain a number of intrinsic mode functions and form low-dimensional parallel sub-data;

[0161] A feature extraction module 30 is used to extract the time domain and frequency domain energy entropy features of the intrinsic mode function to obtain a feature vector;

[0162] The fault diagnosis module 40 is used to input the feature vector into the hybrid classifier model, and perform fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model, wherein the hybrid classifier model includes a single-class support vector machine and a probabilistic neural network, the single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

[0163] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the high-voltage circuit breaker mechanical fault diagnosis method in the above embodiment is implemented.

[0164] The present invention also provides a storage medium containing computer executable instructions, and the computer executable instructions are executed by a computer processor to implement the high-voltage circuit breaker mechanical fault diagnosis method in the above embodiment.

[0165] Finally, it should be noted that although the above embodiments have been described in the specification and drawings of this application, this does not limit the scope of patent protection of this application. All technical solutions generated by replacing or modifying equivalent structures or equivalent processes based on the essential concept of this application using the contents recorded in the specification and drawings of this application, as well as directly or indirectly implementing the technical solutions of the above embodiments in other related technical fields, are included in the scope of patent protection of this application.

Claims

1. A method for diagnosing mechanical faults of high-voltage circuit breakers, characterized in that: include: Obtaining fault vibration signals of high voltage circuit breakers; Decomposing the fault vibration signal in parallel to obtain a number of intrinsic mode functions, and forming low-dimensional parallel sub-data; Extracting the time domain and frequency domain energy entropy characteristics of the intrinsic mode function to obtain a feature vector; Inputting the feature vector into a hybrid classifier model, and performing fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model; The hybrid classifier model includes a single-class support vector machine and a probabilistic neural network. The single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

2. The high-voltage circuit breaker mechanical fault diagnosis method according to claim 1, characterized in that: The method of obtaining a fault vibration signal of a high-voltage circuit breaker comprises: The vibration signal generated by the high-voltage circuit breaker is collected by a signal acquisition circuit module, and the signal acquisition circuit module has a noise filtering function; The collected vibration signal is converted into a digital signal through the analog-to-digital conversion module.

3. The high voltage circuit breaker mechanical fault diagnosis method according to claim 1, characterized in that: The parallel decomposition of the fault vibration signal obtains a plurality of intrinsic mode functions and forms low-dimensional parallel sub-data, including: Determine to decompose the input signal into several modal components with specific center frequencies and limited bandwidths, decompose the original signal into several IMF components and construct a constrained variational problem with the goal of minimizing the bandwidth of the modal components; The augmented Lagrangian function is introduced, and the alternating direction multiplier method is used to solve the saddle point of the variational problem. The modal components, center frequencies and Lagrangian multipliers are updated alternately in sequence, and the multiplier alternating direction method is used to solve and update alternately. Repeat the iterations and when the convergence conditions are met, several intrinsic mode functions with different frequency characteristics are output to form low-dimensional parallel sub-data.

4. The high-voltage circuit breaker mechanical fault diagnosis method according to claim 1, characterized in that: The step of extracting the time domain and frequency domain energy entropy features of the intrinsic mode function to obtain a feature vector includes: The eigenmode functions obtained by the variational mode decomposition method are combined into a component matrix, where the number of rows is the number of modes and the number of columns is the signal length; Divide the component matrix into multiple sub-matrices along the time axis. Each sub-matrix contains multiple time-frequency blocks. Calculate the energy of the time-frequency block and normalize it to the interval [0,1]. Calculate the time domain energy entropy of each submatrix, normalize the total energy of each mode, and then calculate the frequency domain energy entropy of the entire component matrix; The time domain and frequency domain energy entropy features are concatenated into a vector as the feature vector.

5. The high-voltage circuit breaker mechanical fault diagnosis method according to any one of claims 1 to 4, characterized in that: The step of inputting the feature vector into a hybrid classifier model and performing fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model comprises: The feature vector is input into a single-class support vector machine. If the single-class support vector machine determines that it is a normal sample, the output is a normal state; if the single-class support vector machine determines that it is an abnormal sample, it is sent to a probabilistic neural network; The probabilistic neural network is used to determine different fault types based on the feature vectors corresponding to the abnormal samples.

6. The high-voltage circuit breaker mechanical fault diagnosis method according to claim 1, characterized in that: It also includes building a hybrid classifier model using the following steps: Acquire historical fault vibration signals, and process the historical fault vibration signals to obtain historical feature vectors; The historical feature vectors are input into the training process of the hybrid classifier model to obtain the predicted labels, and the hybrid classifier model is constructed according to the predicted labels.

7. The high voltage circuit breaker mechanical fault diagnosis method according to claim 1, characterized in that: Also includes: Visualize the fault diagnosis results.

8. A high voltage circuit breaker mechanical fault diagnosis device, characterized in that: A method for diagnosing a mechanical fault of a high-voltage circuit breaker according to any one of claims 1 to 7, comprising: A signal acquisition module, used to acquire a fault vibration signal of a high-voltage circuit breaker; A parallel decomposition module, used for parallel decomposing the fault vibration signal to obtain a plurality of intrinsic mode functions and compose low-dimensional parallel sub-data; A feature extraction module is used to extract the time domain and frequency domain energy entropy features of the intrinsic mode function to obtain a feature vector; A fault diagnosis module is used to input the feature vector into a hybrid classifier model, and perform fault diagnosis on the high-voltage circuit breaker based on the feature vector through the hybrid classifier model, wherein the hybrid classifier model includes a single-class support vector machine and a probabilistic neural network, the single-class support vector machine is used to identify whether the high-voltage circuit breaker is a normal sample, and the probabilistic neural network is used to identify the fault category corresponding to the abnormal sample.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the high-voltage circuit breaker mechanical fault diagnosis method according to any one of claims 1 to 7 is implemented.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are executed by a computer processor to implement the high-voltage circuit breaker mechanical fault diagnosis method according to any one of claims 1 to 7.

Images