A series arc fault detection method and system based on variational modal decomposition

Abstract

The application belongs to the technical field of power systems, and discloses a series arc fault detection method based on variational mode decomposition, which comprises the following steps: decomposing and extracting a current high-frequency signal to obtain a characteristic extraction signal sequence, and obtaining the variance, peak-peak value and pulse factor of the characteristic extraction signal according to the characteristic extraction signal sequence; recording the calculated variance, peak-peak value and pulse factor as time-frequency domain characteristic value data; obtaining the clustering center of a fault state and the clustering center of a normal state by using the variance, peak-peak value and pulse factor of the current high-frequency signal in the fault state and the current high-frequency signal in the normal state; and judging whether a series arc fault occurs in the DC line according to the distance between the time-frequency domain characteristic value data and the clustering center of the fault state and the clustering center of the normal state. The application has better anti-noise capability and the advantages of not needing to rely on experience to select a fault characteristic frequency band and set a detection threshold.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method and system for detecting series arc faults based on variational mode decomposition. Background Technology

[0002] With the rapid development of power electronics technology, DC power supply systems have been further promoted and applied. However, at the same time, problems such as aging insulation of power supply lines, loose joints, or broken connecting wires causing poor contact can lead to series DC arcing faults, endangering the safety of the power supply system.

[0003] To address the problem of DC series arc fault detection, scholars both domestically and internationally have proposed various methods from different perspectives. For example, a combined fault detection method using Ensemble Empirical Mode Decomposition (EEMD) and Fuzzy C-means Clustering (FCM) has been proposed. This method studies the influence of different operating conditions and external factors on arc fault detection in photovoltaic DC systems, but it does not solve the mode aliasing problem inherent in EEMD. Another example is the use of wavelet transform to extract characteristic frequency bands from the series arc fault signal and the use of a moving time window method to statistically analyze the energy values ​​of the high-frequency coefficients after wavelet decomposition, using these values ​​to characterize the disorder and chaos of the arc fault signal, thereby diagnosing series arc faults. However, this method does not consider the selection of the wavelet function. Yet another example is the use of time-frequency domain combination methods such as the stage frequency domain ratio judgment method and the frequency domain maximum value discrimination method for the discrimination of series DC arc faults, but these methods do not consider the impact of noise on the detection algorithm.

[0004] In other words, traditional detection methods based on time-domain quantities are easily affected by noise interference, while time-frequency domain detection methods such as wavelet transform and empirical mode decomposition have problems such as selecting fault frequency bands based on experience and setting detection thresholds.

[0005] Therefore, how to provide a series arc fault detection method with better noise resistance and without relying on experience to select wavelet functions and fault characteristic frequency bands is an urgent problem to be solved. Summary of the Invention

[0006] This invention provides a series arc fault detection method, system, device, and medium based on variational mode decomposition, to solve the problems of existing technologies being easily affected by noise interference and relying on experience to select wavelet functions and fault characteristic frequency bands.

[0007] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or to describe the scope of protection of these embodiments. Its sole purpose is to present some concepts in a simple form as a prelude to the detailed description that follows.

[0008] According to a first aspect of the present invention, a method for detecting series arc faults based on variational mode decomposition is provided.

[0009] In one embodiment, the series arc fault detection method based on variational mode decomposition includes:

[0010] The loop current of the monitored DC line is sampled to obtain the loop current signal, and the loop current signal is bandpass filtered to obtain a high-frequency current signal.

[0011] The high-frequency current signal is decomposed and extracted to obtain a feature extraction signal sequence. Based on the feature extraction signal sequence, the variance, peak-to-peak value, and impulse factor of the feature extraction signal are obtained. The calculated variance, peak-to-peak value, and impulse factor are recorded as time-frequency domain feature value data.

[0012] The variance, peak-to-peak value, and impulse factor of the high-frequency current signal under fault conditions and the high-frequency current signal under normal conditions are used to obtain the cluster centers of the fault conditions and the cluster centers of the normal conditions.

[0013] Based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state, it is determined whether a series arc fault has occurred in the DC line.

[0014] In one embodiment, the sampling rate of the loop current sampling is greater than or equal to 200kHz, and the frequency of the bandpass filtering is 10kHz to 100kHz.

[0015] In one embodiment, decomposing and extracting the high-frequency current signal to obtain a feature extraction signal sequence includes:

[0016] By using a constrained variational model, variational mode decomposition is performed on the high-frequency current signal to obtain a set of mode components with different numbers of modes;

[0017] The alternating direction multiplier algorithm is used to calculate each group of modal components to obtain the center frequency and bandwidth of each modal component;

[0018] Based on the center frequency and bandwidth of each modal component, the component coefficients of each modal component are obtained, and the component coefficients are compared to obtain the modal component with the largest component coefficient in each component.

[0019] Based on the modal component with the largest component coefficient for each component, determine the kurtosis of the modal component with the largest component coefficient for each component; and compare the kurtosis to obtain the modal component with the largest kurtosis.

[0020] The signal sequence of the mode component with the largest component coefficient and kurtosis is used as the feature extraction signal sequence.

[0021] In one embodiment, the formulas for calculating the variance, peak-to-peak value, and impulse factor of the feature-extracted signal, based on the feature-extracted signal sequence, are as follows:

[0022]

[0023]

[0024] I pp =max(|s t |)-min(|s t |)

[0025]

[0026] In the formula, s(t) is the extracted signal sequence, and s(t) = s1, s2, ..., s N ; is the average value of the feature extraction signal s(t); t is the sampling time, and t=1,2,......,N.

[0027] In one embodiment, obtaining the cluster centers for the fault state and the cluster centers for the normal state from the pre-obtained variance, peak-to-peak value, and impulse factor of the high-frequency current signal under fault conditions and the high-frequency current signal under normal conditions includes:

[0028] The variance, peak-to-peak value, and pulse factor of the high-frequency current signal under fault conditions in the M groups and the variance, peak-to-peak value, and pulse factor of the high-frequency current signal under normal conditions in the M groups are averaged to obtain the average variance, average peak-to-peak value, and average pulse factor of the corresponding M groups.

[0029] The average variance, average peak-to-peak value, and average impulse factor of the M groups under fault conditions are denoted as the cluster centers of the fault conditions.

[0030] The average variance, average peak-to-peak value, and average impulse factor of group M under normal conditions are denoted as the cluster centers under normal conditions.

[0031] In one embodiment, M is greater than or equal to 10.

[0032] In one embodiment, determining whether a series arc fault has occurred in the DC line based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state includes:

[0033] The distance between the time-frequency domain feature value data and the cluster centers under the fault state and the cluster centers under the normal state is calculated using the following formula:

[0034]

[0035]

[0036] In the formula, x3 is the variance of the signal, y3 is the peak-to-peak value of the signal, z3 is the impulse factor of the signal, and D A D is the distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center A(x1,y1,z1) in the normal state. B The distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center B(x2,y2,z2) under the fault state;

[0037] When D A <D B When D is in a certain condition, it is determined that no DC arc fault has occurred in the DC line; when D A >D B When this occurs, it is determined that a DC arc fault has occurred in the DC line.

[0038] According to a second aspect of the present invention, a series arc fault detection system based on variational mode decomposition is provided.

[0039] In one embodiment, the series arc fault detection system based on variational mode decomposition includes:

[0040] The sampling and processing module is used to sample the loop current of the monitored DC line to obtain the loop current signal, and to perform bandpass filtering on the loop current signal to obtain a high-frequency current signal.

[0041] The data calculation module is used to decompose and extract the high-frequency current signal to obtain a feature extraction signal sequence, and to obtain the variance, peak-to-peak value and impulse factor of the feature extraction signal based on the feature extraction signal sequence. The calculated variance, peak-to-peak value and impulse factor are recorded as time-frequency domain feature value data.

[0042] The clustering processing module is used to obtain the cluster centers of the fault state and the cluster centers of the normal state from the variance, peak-to-peak value and impulse factor of the high-frequency current signal under the fault state and the high-frequency current signal under the normal state obtained in advance.

[0043] The fault determination module is used to determine whether a series arc fault has occurred in the DC line based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the cluster centers of the normal state.

[0044] In one embodiment, the sampling rate of the loop current sampling is greater than or equal to 200kHz, and the frequency of the bandpass filtering is 10kHz to 100kHz.

[0045] In one embodiment, the data calculation module includes: a signal extraction module, used to decompose and extract features from the high-frequency current signal to obtain a feature extraction signal sequence, and the signal extraction module includes a mode decomposition submodule, an algorithm calculation submodule, a coefficient determination submodule, a kurtosis determination submodule, and a signal determination submodule, wherein...

[0046] The mode decomposition submodule is used to perform variational mode decomposition on the high-frequency current signal using a constrained variational model to obtain mode component groups with different numbers of modes.

[0047] The algorithm calculation submodule is used to calculate each group of modal components using the alternating direction multiplier algorithm to obtain the center frequency and bandwidth of each modal component;

[0048] The coefficient determination submodule is used to obtain the component coefficients of each modal component based on the center frequency and bandwidth of each modal component, and compare the component coefficients to obtain the modal component with the largest component coefficient in each component.

[0049] The kurtosis determination submodule is used to determine the kurtosis of the modal component with the largest component coefficient for each component based on the obtained modal component with the largest component coefficient; and to compare the kurtosis to obtain the modal component with the largest kurtosis.

[0050] The signal determination submodule is used to extract the feature signal sequence from the signal sequence of the mode component with the largest component coefficient and kurtosis.

[0051] In one embodiment, the data calculation module includes: a signal calculation module, used to obtain the variance, peak-to-peak value, and impulse factor of the feature-extracted signal based on the feature-extracted signal sequence; and the calculation formula is:

[0052]

[0053]

[0054] I pp =max(|s t |)-min(|s t |)

[0055]

[0056] In the formula, s(t) is the extracted signal sequence, and s(t) = s1, s2, ..., s N ; is the average value of the feature extraction signal s(t); t is the sampling time, and t=1,2,......,N.

[0057] In one embodiment, when the clustering processing module obtains the cluster centers for the fault state and the normal state from the variance, peak-to-peak value, and pulse factor of the high-frequency current signals under the fault state and the high-frequency current signals under the normal state obtained in advance, it calculates the average values ​​of the variance, peak-to-peak value, and pulse factor of the M groups of high-frequency current signals under the fault state and the M groups of high-frequency current signals under the normal state obtained in advance, and obtains the average variance, average peak-to-peak value, and average pulse factor of the corresponding M groups; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the fault state are recorded as the cluster centers for the fault state; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the normal state are recorded as the cluster centers for the normal state.

[0058] In one embodiment, M is greater than or equal to 10.

[0059] In one embodiment, the fault determination module calculates the distance between the time-frequency domain feature value data and the cluster centers in the fault state and the cluster centers in the normal state:

[0060]

[0061]

[0062] In the formula, x3 is the variance of the signal, y3 is the peak-to-peak value of the signal, z3 is the impulse factor of the signal, and D A D is the distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center A(x1,y1,z1) in the normal state. B The distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center B(x2,y2,z2) under the fault state;

[0063] And, when D A <D B When the fault judgment module determines that no DC arc fault has occurred in the DC line; when D A >D B When the fault determination module determines that a DC arc fault has occurred in the DC line, the fault determination module determines that a DC arc fault has occurred in the DC line.

[0064] According to a third aspect of the present invention, a computer device is provided.

[0065] In some embodiments, the computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method described above.

[0066] According to a fourth aspect of the present invention, a computer-readable storage medium is provided.

[0067] In one embodiment, a computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the above method.

[0068] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0069] This invention focuses on the impact of noise under actual working conditions. It performs multiple variational mode decompositions on the DC arc current signal to obtain multiple mode components. Then, it uses the correlation coefficient and kurtosis of the decomposed signal as screening criteria to determine the optimal number of decomposition layers and the component with the highest correlation to calculate variance features, peak-to-peak features, and impulse features, constructing a three-dimensional feature vector. Finally, it uses an algorithm of Euclidean distance comparison to effectively identify series arc faults.

[0070] Compared to detection methods based solely on time-domain quantities, this method offers better noise immunity. Compared to traditional time-frequency domain methods, it eliminates the need for empirical selection of fault characteristic frequency bands and setting detection thresholds. It can effectively improve the detection success rate of DC arc faults and reduce the hazards caused by them.

[0071] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0072] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0073] Figure 1 This is a flowchart illustrating a series arc fault detection method based on variational mode decomposition according to an exemplary embodiment.

[0074] Figure 2 This is a schematic diagram of a series arc fault detection system based on variational mode decomposition, according to an exemplary embodiment.

[0075] Figure 3 It is an image of current signal sample values ​​shown according to an exemplary embodiment;

[0076] Figure 4 This is a filtered current signal image shown according to an exemplary embodiment;

[0077] Figure 5 This is an image showing the result of a two-layer VMD decomposition of a current signal according to an exemplary embodiment;

[0078] Figure 6 This is an image showing the result of a 3-layer VMD decomposition of a current signal according to an exemplary embodiment;

[0079] Figure 7 This is an image showing the result of a 4-layer VMD decomposition of a current signal according to an exemplary embodiment;

[0080] Figure 8 This is an image showing the 5-layer VMD decomposition result of a current signal according to an exemplary embodiment;

[0081] Figure 9 This is a feature extraction signal image illustrated according to an exemplary embodiment;

[0082] Figure 10 This is a schematic diagram of the structure of a computer device according to an exemplary embodiment. Detailed Implementation

[0083] The following description and accompanying drawings fully illustrate specific embodiments described herein to enable those skilled in the art to practice them. Some embodiments may include or substitute parts and features of other embodiments. The scope of the embodiments herein encompasses the entire scope of the claims and all available equivalents thereof. Throughout this document, the terms “first,” “second,” etc., are used only to distinguish one element from another without requiring or implying any actual relationship or order between the elements. Indeed, a first element can also be referred to as a second element, and vice versa. Furthermore, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a structure, apparatus, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a structure, apparatus, or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the structure, apparatus, or device that includes said element. The various embodiments described herein are presented in a progressive manner, with each embodiment focusing on its differences from other embodiments; similar or identical parts between embodiments can be referred to interchangeably.

[0084] The terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer" used in this document to indicate orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings. They are used solely for the convenience of describing the document and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In the description herein, unless otherwise specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two elements; they can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0085] In this document, unless otherwise stated, the term "multiple" means two or more.

[0086] In this article, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0087] In this article, the term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0088] It should be understood that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order constraint on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the diagram may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0089] The modules in the apparatus or system of this application can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0090] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0091] Figure 1An embodiment of a series arc fault detection method based on variational mode decomposition according to the present invention is shown.

[0092] In this optional embodiment, the series arc fault detection method based on variational mode decomposition includes:

[0093] Step S101: Sample the loop current of the monitored DC line to obtain the loop current signal, and perform bandpass filtering on the loop current signal to obtain a high-frequency current signal.

[0094] Step S103: Decompose and extract the high-frequency current signal to obtain a feature extraction signal sequence, and obtain the variance, peak-to-peak value and impulse factor of the feature extraction signal based on the feature extraction signal sequence. Record the calculated variance, peak-to-peak value and impulse factor as time-frequency domain feature value data.

[0095] Step S105: Obtain the cluster centers of the fault state and the normal state from the variance, peak-to-peak value and impulse factor of the high-frequency current signal under the fault state and the high-frequency current signal under the normal state obtained in advance.

[0096] Step S107: Determine whether a series arc fault has occurred in the DC line based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state.

[0097] Figure 2 An embodiment of a series arc fault detection system based on variational mode decomposition according to the present invention is shown.

[0098] In this optional embodiment, the series arc fault detection system based on variational mode decomposition includes:

[0099] The sampling and processing module 201 is used to sample the loop current of the monitored DC line to obtain the loop current signal, and to perform bandpass filtering on the loop current signal to obtain a high-frequency current signal.

[0100] The data calculation module 203 is used to decompose and extract the high-frequency current signal to obtain a feature extraction signal sequence, and to obtain the variance, peak-to-peak value and pulse factor of the feature extraction signal based on the feature extraction signal sequence. The calculated variance, peak-to-peak value and pulse factor are recorded as time-frequency domain feature value data.

[0101] The clustering processing module 205 is used to obtain the cluster centers of the fault state and the cluster centers of the normal state from the variance, peak-to-peak value and impulse factor of the high-frequency current signal under the fault state and the high-frequency current signal under the normal state obtained in advance.

[0102] The fault judgment module 207 is used to judge whether a series arc fault has occurred in the DC line based on the distance between the time-frequency domain feature value data and the cluster center of the fault state and the cluster center of the normal state.

[0103] In one embodiment, the sampling rate of the loop current sampling is greater than or equal to 200kHz, and the frequency of the bandpass filtering is 10kHz to 100kHz.

[0104] In one embodiment, when decomposing and extracting the high-frequency current signal to obtain a feature extraction signal sequence, a constrained variational model is used to perform variational mode decomposition on the high-frequency current signal to obtain mode component groups with different numbers of modes. An alternating direction multiplier algorithm is used to calculate the center frequency and bandwidth of each mode component. Based on the center frequency and bandwidth of each mode component, the component coefficients of each mode component are obtained, and the component coefficients are compared to obtain the mode component with the largest component coefficient in each group. Based on the obtained mode component with the largest component coefficient in each group, the kurtosis of the mode component with the largest component coefficient in each group is determined. The kurtosis is then compared to obtain the mode component with the largest kurtosis. The signal sequence of the mode component with both the largest component coefficient and the largest kurtosis is used as the feature extraction signal sequence.

[0105] Specifically, when using the VMD (Variational Mode Decomposition) algorithm to extract and decompose a signal, it is necessary to set an important parameter K. The value of K represents the number of modes to be decomposed. Perform L VMD decompositions on the original signal f(t), and it is recommended that L≥4. The number of decomposition layers K can be K=2,……,L+1 respectively.

[0106] When decomposing a signal using VMD, the center frequencies and bandwidths of the intrinsic mode components (IMFs) are continuously updated through the solution process of the constrained variational model, resulting in a series of bandwidth-constrained IMFs. The corresponding constrained variational model is:

[0107]

[0108] In the formula, δ(t) is the unit impulse function; j is the imaginary unit; π is pi, with a value of 3.1415926; t is the sampling time; and * is the convolution operator. To find the partial derivative of the function; {u k} represents the K eigenmode components obtained from the decomposition, {u k}={u1,u2,u3,......,u k};{ω k} represents the center frequency of each component, {ω k}={ω1,ω2,ω3,......,ω k}; f(t) is the objective function, representing the total bandwidth.

[0109] To solve the above variational model, the constrained variational problem is transformed into an unconstrained variational problem, and the augmented Lagrangian expression is obtained as follows:

[0110]

[0111] In the formula, L(·) is the Lagrange function, α is the penalty factor, λ(t) is the Lagrange multiplier with respect to time t, and k = 1, 2, ..., K. It is the L2 norm of the gradient function, mainly used to evaluate bandwidth. This is a secondary penalty item. These are Lagrange multipliers, and these two terms are mainly used to ensure strict refactoring constraints.

[0112] The alternating direction multiplier algorithm is used to solve for the saddle points of the equation, resulting in a series of u. k and ω k This refers to the center frequency and bandwidth of the K modal components. The specific steps are as follows:

[0113] a) Initialize {u k 1},{ω k 1},λ 1 and n = 0.

[0114] b) Iterative update u k ω k And λ, that is:

[0115]

[0116]

[0117]

[0118] In the formula, τ is the step size of the Lagrange multiplier term. For the Fourier transform of f(t), For u k (t) is the value of the (n+1)th iteration in the Fourier domain. For ω k The value of the (n+1)th iteration, where ω is the frequency parameter. Let λ(t) be the (n+1)th iteration value in the Fourier domain.

[0119] c) For a given discrimination precision ε > 0, if the iteration stopping condition is satisfied... Then stop iterating; otherwise, return to step b).

[0120] d) Output K modal components and their corresponding center frequency values.

[0121] After iteration, the decomposition result of VMD {u k}={u1,u2,u3,......,u k Calculate their correlation coefficient α respectively. k ={α1,α2,α3,……,α k} and kurtosis H k ={H1,H2,H3,……,H k}, for signal u i (t), i = 1, 2, ..., k, assuming sampled values ​​are x1, x2, ..., xk. N N is the number of sampling points, and the average value is denoted as N. Its corresponding correlation coefficient α i The calculation method is as follows:

[0122]

[0123]

[0124] Where f(t) = f1, f2, ..., f N It is the original signal before decomposition. It is the average value of the original signal before decomposition, u i (t)=x t It is a decomposed signal. It is the average value of the decomposed signals. The decomposed signal with the largest correlation coefficient is selected, and its kurtosis H is... i The calculation method is as follows:

[0125]

[0126]

[0127] Where, σ 2 It is the variance of the decomposed signal, σ 4 It is the standard deviation of the decomposed signal. The original signal f(t) is decomposed L times using VMD, and it is recommended that L≥4. The number of decomposition layers K is taken as K=2,......,L+1 respectively. In each decomposition, the decomposed signal with the largest correlation coefficient is selected, its kurtosis is calculated, and the signal with the largest kurtosis is selected as the feature extraction signal sequence s(t). The optimal number of decomposition layers K is determined.

[0128] In one embodiment, the formulas for calculating the variance, peak-to-peak value, and impulse factor of the feature extraction signal based on the feature extraction signal sequence are as follows:

[0129]

[0130]

[0131] I pp =max(|s t |)-min(|s t |)

[0132]

[0133] In the formula, s(t) is the extracted signal sequence, and s(t) = s1, s2, ..., s N ; is the average value of the feature extraction signal s(t); t is the sampling time, and t=1,2,......,N.

[0134] In one embodiment, when obtaining cluster centers for the fault state and the normal state based on the variance, peak-to-peak value, and pulse factor of the high-frequency current signal under the fault state and the high-frequency current signal under the normal state obtained in advance, the average values ​​of the variance, peak-to-peak value, and pulse factor of the high-frequency current signal under the fault state of M groups and the variance, peak-to-peak value, and pulse factor of the high-frequency current signal under the normal state of M groups (M is greater than or equal to 10) are calculated to obtain the average variance, average peak-to-peak value, and average pulse factor of the corresponding M groups; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the fault state are recorded as the cluster centers of the fault state; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the normal state are recorded as the cluster centers of the normal state.

[0135] Based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state, it is determined whether a series arc fault has occurred in the DC line, as follows:

[0136] The distance between the time-frequency domain feature value data and the cluster centers under the fault state and the cluster centers under the normal state is calculated using the following formula:

[0137]

[0138]

[0139] In the formula, x3 is the variance of the signal, y3 is the peak-to-peak value of the signal, z3 is the impulse factor of the signal, and D A D is the distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center A(x1,y1,z1) in the normal state. B The distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center B(x2,y2,z2) under the fault state;

[0140] When D A <DB When D is in a certain condition, it is determined that no DC arc fault has occurred in the DC line; when D A >D B When this occurs, it is determined that a DC arc fault has occurred in the DC line.

[0141] For ease of understanding, the above-mentioned technical solution of the present invention will be described in detail below through specific examples, as follows:

[0142] 1) Sample the current of the monitored DC power supply line at a sampling frequency of 250kHz to obtain the DC line current signal w, such as... Figure 3 As shown, a low-pass filter is applied to the current, and the filtered high-frequency current signal f is as follows. Figure 4 As shown.

[0143] 2) Perform L-fold VMD decomposition on the filtered high-frequency current signal f, where L is set to 4 in this case. Figures 5-8 These are images obtained by decomposing the high-frequency current signal f into 2 to 5 layers.

[0144] 3) Calculate the correlation coefficient α of the decomposed signals from level 2 to level 5. i The correlation coefficients of the decomposed signals under different K values ​​are shown in Table 1.

[0145] Table 1. Correlation coefficients of the decomposed signals under different K values.

[0146]

[0147] Table 1 shows that when K=2, the signal with the highest correlation is IMF2; when K=3, the signal with the highest correlation is IMF4; when K=4, the signal with the highest correlation is IMF4; and when K=5, the signal with the highest correlation is IMF4. Analyzing these signals, their respective kurtosis H is calculated. i Table 2 shows the kurtosis values ​​of the decomposed signals with the highest correlation coefficients under different K values.

[0148] Table 2. Kurtosis values ​​of the highest correlation coefficient decomposed signals under different K values.

[0149] K value Modal components Correlation coefficient cliff 2 IMF2 0.81 3.39 3 IMF2 0.67 2.88 4 IMF4 0.66 3.85 5 IMF4 0.62 2.95

[0150] 4) As shown in Table 1, when K=4, the kurtosis of the decomposed signal IMF4 is 3.85, which is the highest value. Using it as the feature extraction signal s(t), the image is as follows: Figure 9 As shown, calculate its variance σ. 2 Peak-to-peak value I pp and pulse factor I f The results are shown in Table 3.

[0151] Table 3. Variance, peak-to-peak value, and impulse factor of the feature-extracted signal.

[0152] Signal <![CDATA[Variance / A 2 > Peak-to-peak value / A cliff s(t) <![CDATA[1.53×10 -5 ]]> 0.0314 9.603

[0153] 5) Gather the 10 sets of data each for the fault state and the non-fault state obtained in advance into a training sample set X, and calculate the cluster center A(x1, y1, z1) for the normal state and the cluster center B(x2, y2, z2) for the fault state, as shown in Table 4:

[0154] Table 4. Numerical values ​​of the training sample set

[0155]

[0156] By calculating the average value of the dataset, two cluster centers were obtained: the normal state cluster center A (2.99024 × 10⁻⁶). -6 The values ​​are 0.01105 and 3.6671, and the cluster center B of the fault state is 1.85513 × 10⁻⁶. -5 (0.0290919, 9.79475), such as Figure 7 As shown.

[0157] The time-frequency domain eigenvalue data C (1.53 × 10⁻⁶) is obtained from step 4). -5 The distances from (0.0314, 9.603) to the cluster center AB are respectively:

[0158]

[0159]

[0160] It can be seen that D A >D B A series arc fault occurred in the line.

[0161] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores static and dynamic information data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the steps in the above method embodiments.

[0162] Those skilled in the art will understand that Figure 10The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0163] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0164] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the method embodiments described above.

[0165] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0166] This invention is not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this invention is limited only by the appended claims.

Claims

1. A method for detecting series arc faults based on variational mode decomposition, characterized in that, include: The loop current of the monitored DC line is sampled to obtain the loop current signal, and the loop current signal is bandpass filtered to obtain a high-frequency current signal. The high-frequency current signal is decomposed and extracted to obtain a feature extraction signal sequence. Based on the feature extraction signal sequence, the variance, peak-to-peak value, and impulse factor of the feature extraction signal are obtained. The calculated variance, peak-to-peak value, and impulse factor are recorded as time-frequency domain feature value data. The variance, peak-to-peak value, and impulse factor of the high-frequency current signal under fault conditions and the high-frequency current signal under normal conditions are used to obtain the cluster centers of the fault conditions and the cluster centers of the normal conditions. Based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state, it is determined whether a series arc fault has occurred in the DC line.

2. The series arc fault detection method based on variational mode decomposition according to claim 1, characterized in that, The sampling rate of the loop current sampling is greater than or equal to 200kHz, and the frequency of the bandpass filtering is 10kHz to 100kHz.

3. The series arc fault detection method based on variational mode decomposition according to claim 1, characterized in that, The high-frequency current signal is decomposed and extracted to obtain a feature extraction signal sequence including: By using a constrained variational model, variational mode decomposition is performed on the high-frequency current signal to obtain a set of mode components with different numbers of modes; The alternating direction multiplier algorithm is used to calculate each group of modal components to obtain the center frequency and bandwidth of each modal component; Based on the center frequency and bandwidth of each modal component, the component coefficients of each modal component are obtained, and the component coefficients are compared to obtain the modal component with the largest component coefficient in each component. Based on the modal component with the largest component coefficient for each component, determine the kurtosis of the modal component with the largest component coefficient for each component; and compare the kurtosis to obtain the modal component with the largest kurtosis. The signal sequence of the mode component with the largest component coefficient and kurtosis is used as the feature extraction signal sequence.

4. The series arc fault detection method based on variational mode decomposition according to claim 3, characterized in that, Based on the feature extraction signal sequence, the formulas for calculating the variance, peak-to-peak value, and impulse factor of the feature extraction signal are as follows: In the formula, s(t) is the extracted signal sequence, and s(t) = s1, s2, ..., s N ; is the average value of the feature extraction signal s(t); t is the sampling time, and t=1,2,......,N.

5. The series arc fault detection method based on variational mode decomposition according to claim 4, characterized in that, Based on the variance, peak-to-peak value, and impulse factor of the high-frequency current signal under fault conditions and under normal conditions obtained in advance, the cluster centers for the fault conditions and the cluster centers for the normal conditions are obtained as follows: The variance, peak-to-peak value, and pulse factor of the high-frequency current signal under fault conditions in the M groups and the variance, peak-to-peak value, and pulse factor of the high-frequency current signal under normal conditions in the M groups are averaged to obtain the average variance, average peak-to-peak value, and average pulse factor of the corresponding M groups. The average variance, average peak-to-peak value, and average impulse factor of the M groups under fault conditions are denoted as the cluster centers of the fault conditions. The average variance, average peak-to-peak value, and average impulse factor of group M under normal conditions are denoted as the cluster centers under normal conditions.

6. The series arc fault detection method based on variational mode decomposition according to claim 5, characterized in that, M is greater than or equal to 10.

7. The series arc fault detection method based on variational mode decomposition according to claim 6, characterized in that, The determination of whether a series arc fault has occurred in the DC line is based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the normal state, including: The distance between the time-frequency domain feature value data and the cluster centers under the fault state and the cluster centers under the normal state is calculated using the following formula: In the formula, x3 is the variance of the signal, y3 is the peak-to-peak value of the signal, z3 is the impulse factor of the signal, and D A D is the distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center A(x1,y1,z1) in the normal state. B The distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center B(x2,y2,z2) under the fault state; When D A <D B When D is in a certain condition, it is determined that no DC arc fault has occurred in the DC line; when D A >D B When this occurs, it is determined that a DC arc fault has occurred in the DC line.

8. A series arc fault detection system based on variational mode decomposition, characterized in that, include: The sampling and processing module is used to sample the loop current of the monitored DC line to obtain the loop current signal, and to perform bandpass filtering on the loop current signal to obtain a high-frequency current signal. The data calculation module is used to decompose and extract the high-frequency current signal to obtain a feature extraction signal sequence, and to obtain the variance, peak-to-peak value and impulse factor of the feature extraction signal based on the feature extraction signal sequence. The calculated variance, peak-to-peak value and impulse factor are recorded as time-frequency domain feature value data. The clustering processing module is used to obtain the cluster centers of the fault state and the cluster centers of the normal state from the variance, peak-to-peak value and impulse factor of the high-frequency current signal under the fault state and the high-frequency current signal under the normal state obtained in advance. The fault determination module is used to determine whether a series arc fault has occurred in the DC line based on the distance between the time-frequency domain feature value data and the cluster centers of the fault state and the cluster centers of the normal state.

9. The series arc fault detection system based on variational mode decomposition according to claim 8, characterized in that, The sampling rate of the loop current sampling is greater than or equal to 200kHz, and the frequency of the bandpass filtering is 10kHz to 100kHz.

10. The series arc fault detection system based on variational mode decomposition according to claim 8, characterized in that, The data calculation module includes: a signal extraction module, used to decompose and extract the high-frequency current signal to obtain a feature extraction signal sequence, and the signal extraction module includes a mode decomposition submodule, an algorithm calculation submodule, a coefficient determination submodule, a kurtosis determination submodule, and a signal determination submodule, wherein... The mode decomposition submodule is used to perform variational mode decomposition on the high-frequency current signal using a constrained variational model to obtain mode component groups with different numbers of modes. The algorithm calculation submodule is used to calculate each group of modal components using the alternating direction multiplier algorithm to obtain the center frequency and bandwidth of each modal component; The coefficient determination submodule is used to obtain the component coefficients of each modal component based on the center frequency and bandwidth of each modal component, and compare the component coefficients to obtain the modal component with the largest component coefficient in each component. The kurtosis determination submodule is used to determine the kurtosis of the modal component with the largest component coefficient for each component based on the obtained modal component with the largest component coefficient; and to compare the kurtosis to obtain the modal component with the largest kurtosis. The signal determination submodule is used to extract the feature signal sequence from the signal sequence of the mode component with the largest component coefficient and kurtosis.

11. The series arc fault detection system based on variational mode decomposition according to claim 10, characterized in that, The data calculation module includes: a signal calculation module, used to obtain the variance, peak-to-peak value, and impulse factor of the feature-extracted signal based on the feature-extracted signal sequence; and the calculation formula is: I pp =max(|s t |)-min(|s t |) In the formula, s(t) is the extracted signal sequence, and s(t) = s1, s2, ..., s N ; is the average value of the feature extraction signal s(t); t is the sampling time, and t=1,2,......,N.

12. The series arc fault detection system based on variational mode decomposition according to claim 11, characterized in that, When the clustering processing module obtains the cluster centers for the fault state and the normal state from the variance, peak-to-peak value, and pulse factor of the high-frequency current signals under the fault state and the high-frequency current signals under the normal state obtained in advance, it calculates the average values ​​of the variance, peak-to-peak value, and pulse factor of the high-frequency current signals under the fault state and the high-frequency current signals under the normal state obtained in advance, and obtains the average variance, average peak-to-peak value, and average pulse factor of the corresponding M groups; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the fault state are recorded as the cluster centers for the fault state; the average variance, average peak-to-peak value, and average pulse factor of the M groups under the normal state are recorded as the cluster centers for the normal state.

13. The series arc fault detection system based on variational mode decomposition according to claim 12, characterized in that, M is greater than or equal to 10.

14. The series arc fault detection system based on variational mode decomposition according to claim 13, characterized in that, The fault determination module calculates the distance between the time-frequency domain feature value data and the cluster centers in the fault state and the cluster centers in the normal state: In the formula, x3 is the variance of the signal, y3 is the peak-to-peak value of the signal, z3 is the impulse factor of the signal, and D A D is the distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center A(x1,y1,z1) in the normal state. B The distance between the time-frequency domain feature point C(x3,y3,z3) represented by the current signal and the cluster center B(x2,y2,z2) under the fault state; And, when D A <D B When the fault judgment module determines that no DC arc fault has occurred in the DC line; when D A >D B When the fault determination module determines that a DC arc fault has occurred in the DC line, the fault determination module determines that a DC arc fault has occurred in the DC line.

15. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method of any one of claims 1 to 7.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of any one of claims 1 to 7.

Images