Fault detection method, device, equipment, medium and product for low-voltage distribution network

Abstract

The application relates to a fault detection method, device, equipment, medium and product of a low-voltage distribution network. The method comprises the following steps: constructing a Hankel matrix according to a current sequence of a target measuring point in the low-voltage distribution network; determining a delay characteristic sequence according to the Hankel matrix; constructing a HAVOK linear forced model according to the delay characteristic sequence; performing fitting processing on the HAVOK linear forced model, and determining a fault detection result of the target measuring point according to the amplitude of a forced operator in the fitting processing process; wherein the forced operator is a delay signal located at the end of the delay characteristic sequence. The method can improve the accuracy of fault detection.

Description

Technical Field

[0001] This application relates to the field of power distribution network technology, and in particular to a fault detection method, device, equipment, medium and product for low-voltage power distribution networks. Background Technology

[0002] Arc faults are a common and highly dangerous type of fault in low-voltage power distribution systems, especially at distribution cabinets, branch circuits, and end load sides. They are easily masked by load fluctuations, switch operations, harmonics, and measurement noise, leading to missed detections or false alarms. Existing fault detection methods for low-voltage distribution networks generally use preset thresholds to detect the collected current signal and obtain the fault detection result.

[0003] However, because electric arcs in low-voltage distribution networks may exhibit characteristics such as high impedance, intermittency, and strong randomness, resulting in insignificant fault current amplitude and unstable waveform distortion, using threshold detection methods will reduce the accuracy of fault detection. Summary of the Invention

[0004] Therefore, it is necessary to provide a fault detection method, device, equipment, medium, and product for low-voltage distribution networks to address the aforementioned technical problems and improve the accuracy of fault detection.

[0005] Firstly, this application provides a fault detection method for low-voltage distribution networks, including:

[0006] Construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network;

[0007] Determine the delayed feature sequence based on the Hankel matrix;

[0008] Based on the delayed feature sequence, construct the Koopman HAVOK linear forced model;

[0009] The HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process; where the forced operator is the delayed signal at the end of the delayed feature sequence.

[0010] In one embodiment, a Hankel matrix is ​​constructed based on the current current sequence of the target measurement point in the low-voltage distribution network, including:

[0011] For each current signal in the current current sequence of the target measurement point in the low-voltage distribution network, the current signal is processed according to the delay embedding dimension and delay step size to obtain the delay vector corresponding to the current signal; based on the delay vector corresponding to each current signal, a Hankel matrix is ​​constructed.

[0012] In one embodiment, determining the delayed feature sequence based on the Hankel matrix includes:

[0013] Singular value decomposition is performed on the Hankel matrix to obtain the decomposition result, which includes the right singular vector matrix. Based on the effective rank parameter, the delayed feature sequence is determined from the right singular vector matrix.

[0014] In one embodiment, a Koopman-Havok linear forced model is constructed based on the delayed feature sequence, including:

[0015] For each delayed signal in the delayed feature sequence, the delayed signal is differentially processed to obtain the corresponding time derivative value; based on the corresponding time derivative values ​​of each delayed signal, a Koopman HAVOK linear forced model is constructed.

[0016] In one embodiment, the fault detection result of the target measurement point is determined based on the amplitude of the forcing operator during the fitting process, including:

[0017] The amplitude of the forced operator during the fitting process is compared with at least one fault threshold; based on the comparison result, the fault detection result of the target measurement point is determined.

[0018] In one embodiment, the method further includes:

[0019] Based on the historical current sequence of candidate measurement points under current abnormality scenarios in low-voltage distribution networks, at least one fault threshold is determined; wherein, current abnormality scenarios include at least one of arc fault, non-arc fault, and non-arc disturbance.

[0020] Secondly, this application also provides a fault detection device for a low-voltage distribution network, comprising:

[0021] The matrix construction module is used to construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network.

[0022] The sequence determination module is used to determine the delayed feature sequence based on the Hankel matrix;

[0023] The model building module is used to construct the Koopman HAVOK linear forced model based on the delayed feature sequence.

[0024] The fault detection module is used to fit the HAVOK linear forced model and determine the fault detection result of the target measurement point based on the amplitude of the forced operator during the fitting process; wherein, the forced operator is the delayed signal located at the end of the delayed feature sequence.

[0025] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0026] Construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network;

[0027] Determine the delayed feature sequence based on the Hankel matrix;

[0028] Based on the delayed feature sequence, construct the Koopman HAVOK linear forced model;

[0029] The HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process; where the forced operator is the delayed signal at the end of the delayed feature sequence.

[0030] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0031] Construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network;

[0032] Determine the delayed feature sequence based on the Hankel matrix;

[0033] Based on the delayed feature sequence, construct the Koopman HAVOK linear forced model;

[0034] The HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process; where the forced operator is the delayed signal at the end of the delayed feature sequence.

[0035] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0036] Construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network;

[0037] Determine the delayed feature sequence based on the Hankel matrix;

[0038] Based on the delayed feature sequence, construct the Koopman HAVOK linear forced model;

[0039] The HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process; where the forced operator is the delayed signal at the end of the delayed feature sequence.

[0040] The aforementioned fault detection methods, devices, equipment, media, and products for low-voltage distribution networks construct a Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network. Based on the Hankel matrix, a delay characteristic sequence is determined, and a HAVOK linear forced model is constructed based on the delay characteristic sequence. The HAVOK linear forced model is then fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process. Using this method, the delayed signal at the end of the delay characteristic sequence is used as the forced operator. By fitting the HAVOK linear forced model, the amplitude of the forced operator characterizing the fault condition of the target measurement point can be obtained. Subsequently, the fault detection result is determined based on the amplitude of the forced operator, which effectively improves the accuracy of fault detection result determination. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating a fault detection method for a low-voltage distribution network in one embodiment;

[0043] Figure 2 This is a flowchart illustrating the process of determining the HAVOK linear forced model in one embodiment;

[0044] Figure 3 This is a flowchart illustrating the process of determining fault detection results in one embodiment;

[0045] Figure 4 This is a flowchart illustrating a fault detection method for a low-voltage distribution network in another embodiment;

[0046] Figure 5 This is a structural block diagram of a fault detection device for a low-voltage distribution network in one embodiment;

[0047] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0049] Arc faults are a common and highly dangerous type of fault in low-voltage power distribution systems, especially at distribution cabinets, branch circuits, and end load sides. They are easily masked by load fluctuations, switch operations, harmonics, and measurement noise, leading to missed detections or false alarms. Existing fault detection methods for low-voltage distribution networks generally use preset thresholds to detect the collected current signal and obtain the fault detection result.

[0050] However, because electric arcs in low-voltage distribution networks may exhibit characteristics such as high impedance, intermittency, and strong randomness, resulting in insignificant fault current amplitude and unstable waveform distortion, using threshold detection methods will reduce the accuracy of fault detection.

[0051] Based on this, in an exemplary embodiment, a fault detection method for a low-voltage distribution network is provided. This embodiment uses the application of this method to a server as an example for illustration. It can be understood that this method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and implemented through the interaction between the terminal and the server. Figure 1 As shown, it includes the following steps:

[0052] S101, construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network.

[0053] The target measurement point is any location in the low-voltage distribution network where a fault may exist, including but not limited to the main incoming line, branch line, and switchgear outgoing line. The current sequence can be understood as the sequence of current signals flowing through the target measurement point during the current time period. The Hankel matrix is ​​a special matrix in linear algebra whose core characteristic is that all elements on each subdiagonal (antidiagonal) are equal.

[0054] Optionally, current signals flowing through the target measurement point within the current time period can be collected using a current sensor according to a preset signal acquisition interval, and the current signals can be time-sequentially arranged to obtain the current sequence. Current sensors, such as current transformers (CTs), Hall effect current sensors, and Rogowski coils, can be installed on the main incoming line or key branches of the low-voltage distribution network.

[0055] For example, the current current sequence can be , Where t0 is the time when signal acquisition begins; t k This refers to the k-th data acquisition time. i(t) is the preset sampling interval; k ) represents the current signal at the k-th acquisition time.

[0056] Furthermore, the matrix features of the Hankel matrix can be combined with the current current sequence to construct the corresponding Hankel matrix. For example, the current current sequence can be input into a trained first construction model, which processes the current current sequence to obtain the corresponding Hankel matrix. The first construction model can convert the current sequence into a Hankel matrix.

[0057] It is worth noting that, in order to improve the stability of the values ​​during subsequent processing, after acquiring the current signals, the current signals can be time-sequentially arranged to obtain an initial current sequence. Then, a sliding window is applied to the initial current sequence, and the truncated result undergoes light preprocessing to obtain the current current sequence. This light preprocessing can include, but is not limited to, DC component removal, amplitude limiting, normalization, and simple filtering.

[0058] S102, Determine the delayed feature sequence based on the Hankel matrix.

[0059] The delayed feature sequence can be understood as the feature sequence obtained after performing singular value decomposition on the Hankel matrix.

[0060] Optionally, the Hankel matrix can be subjected to singular value decomposition, and the delayed feature sequence can be obtained from the decomposition result. The decomposition result may include a left singular vector matrix, a singular value diagonal matrix, and a right singular vector matrix. For example, the delayed feature sequence can be obtained from the right singular vector matrix.

[0061] S103. Based on the delayed feature sequence, construct the Koopman HAVOK linear forced model.

[0062] Among them, the HAVOK linear forced model, also known as the Hankel Alternative View of Koopman linear forced model, is a data-driven dynamic system identification method. It can decompose complex nonlinear dynamic systems into linear dominant systems and intermittent forcing terms. It does not rely on prior knowledge of the system and can identify and reconstruct the dynamic characteristics of the system simply by observing the current sequence. It is a key model connecting the Hankel matrix and dynamic system analysis.

[0063] Optionally, the time derivative values ​​corresponding to each delayed signal in the delayed feature sequence can be calculated to construct a HAVOK linear forced model. Alternatively, the delayed feature sequence can be input into a trained second model, which processes the delayed feature sequence to obtain the corresponding HAVOK linear forced model. The second model can convert the delayed feature sequence into a HAVOK linear forced model.

[0064] S104. Fit the HAVOK linear forced model and determine the fault detection result of the target measurement point based on the amplitude of the forced operator during the fitting process.

[0065] The fault detection result is the result obtained after fault detection at the target measurement point. The amplitude can characterize the change in the value of the forcing operator. In this embodiment, the forcing operator is the delayed signal at the end of the delayed feature sequence, used to characterize the unexplainable part in the process of converting the nonlinear model into a linear model, so as to make the obtained low-dimensional linear model effective. In practical applications, if there is no fault at the target measurement point, the forcing operator tends to 0; if the target measurement point encounters nonlinearity, abrupt change, or a fault, the forcing operator is activated, injecting the effect of the fault into the linear model, so that the low-dimensional linear model can still accurately reproduce the dynamics of the real system. Therefore, the change in the forcing operator can effectively reflect the fault status of the target measurement point.

[0066] Optionally, the HAVOK linear forced model can be fitted first. For example, the least squares method can be used to fit the HAVOK linear forced model. Then, the fault detection result of the target measuring point can be determined based on the magnitude of the forcing operator during the fitting process. For example, the magnitude can be compared with a preset magnitude range. If the magnitude is within the range, it is determined that there is an arc fault at the target measuring point; if the magnitude is outside the range, it is determined that there is no arc fault at the target measuring point. The magnitude range can be [0.06, 0.18].

[0067] After determining the fault detection results, the fault type (e.g., arc fault / non-arc disturbance / other fault / uncertain) and alarm trigger timestamp can be output to the operation and maintenance terminal. Additionally, the peak value of the forced operator, burst duration, and window number can be output for subsequent log and traceability analysis. Simultaneously, local audible and visual alarms, communication transmission, circuit breaker tripping, or fault recording operations can also be performed.

[0068] It is understood that in this embodiment, fault detection can be achieved by relying solely on a single-channel current signal, requiring minimal hardware modifications and making it suitable for locations such as low-voltage distribution network switchgear, distribution boxes, and terminal protection devices.

[0069] In the aforementioned fault detection method for low-voltage distribution networks, a Hankel matrix is ​​constructed based on the current current sequence of the target measurement point in the low-voltage distribution network. Based on the Hankel matrix, a delay feature sequence is determined, and a HAVOK linear forced model is constructed based on the delay feature sequence. Then, the HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process. Using this method, the delayed signal at the end of the delay feature sequence is used as the forced operator. By fitting the HAVOK linear forced model, the amplitude of the forced operator characterizing the fault condition of the target measurement point can be obtained. Subsequently, the fault detection result is determined based on the amplitude of the forced operator, which effectively improves the accuracy of fault detection result determination.

[0070] Based on the above embodiments, this application provides an optional method for constructing a Hankel matrix. Specifically, for each current signal in the current current sequence of the target measurement point in the low-voltage distribution network, the current signal is processed according to the delay embedding dimension and delay step size to obtain the delay vector corresponding to the current signal; and a Hankel matrix is ​​constructed according to the delay vector corresponding to each current signal.

[0071] Delayed embedding is a method that maps a one-dimensional time series to a high-dimensional space. Its core principle is to construct a high-dimensional vector using past sequence values, thereby reconstructing the dynamic characteristics of the original system. The dimension of the delayed embedding can be understood as the length of the high-dimensional vector, i.e., how many past sample points are used to construct the vector; the delay step size can be understood as the time interval between two adjacent signals. The delayed vector can be understood as the vector obtained after delay embedding.

[0072] Optionally, the autocorrelation function method or the mutual information method can be used to set the delay embedding dimension and delay step size; or the delay embedding dimension and delay step size can be set according to the historical experience of those skilled in the art. This embodiment does not impose any restrictions on this.

[0073] For each current signal in the current current sequence, starting from that current signal, take several current signals with a delay embedding dimension according to the delay step size to form the delay vector corresponding to that current signal. For example, the delay vector can be referred to the following formula (1). Where q is the delay embedding dimension; To delay the step size, ;T is the transpose symbol.

[0074] (1)

[0075] Furthermore, the delay vectors can be stacked side by side to construct the Hankel matrix. For example, the Hankel matrix H can be obtained by referring to the following formula (2). Where p = N - (q - 1)m is the number of available columns.

[0076] (2)

[0077] In this embodiment of the application, by performing delay embedding processing on each current signal, a one-dimensional current sequence can be converted into a Hankel matrix, laying a reasonable foundation for subsequent fault detection processing.

[0078] Based on the above embodiments, this application provides an optional method for determining the delayed feature sequence, specifically, performing singular value decomposition on the Hankel matrix to obtain the decomposition result; and determining the delayed feature sequence from the right singular vector matrix according to the effective rank parameter.

[0079] The decomposition result includes the right singular vector matrix. The effective rank parameter can be understood as the number of singular values ​​that can represent the true signal.

[0080] Optionally, the Hankel matrix can be subjected to singular value decomposition to obtain a decomposition result containing a left singular vector matrix, a singular value diagonal matrix, and a right singular vector matrix. For example, refer to the formula. The Hankel matrix is ​​subjected to singular value decomposition. Here, U is the left singular vector matrix. ; V is a diagonal matrix of singular values; V is a right singular vector matrix. This corresponds to the feature sequence in the time dimension.

[0081] Furthermore, an optimal hard threshold strategy or a strategy based on the singular value energy ratio can be adopted to select the effective rank parameter r; then, the first r feature sequences in the right singular vector matrix are used as delayed feature sequences.

[0082] For example, the delayed feature sequence can be represented as , where v r (t k ) represents the r-th delayed signal in the delayed feature sequence; the length of each delayed signal in the delayed feature sequence is p, that is, each delayed signal represents a column of data in the right singular vector matrix.

[0083] In this embodiment, by performing singular value decomposition on the Hankel matrix and determining the delayed feature sequence from the right singular vector matrix representing the feature sequence in the time dimension, the correlation between the delayed feature sequence and the time dimension can be guaranteed, thus laying a reliable foundation for subsequent fault detection.

[0084] Based on the above embodiments, this application provides an optional method for determining the HAVOK linear forced model, such as... Figure 2 As shown, the specific steps include:

[0085] S201, for each delayed signal in the delayed feature sequence, perform differential processing on the delayed signal to obtain the time derivative value corresponding to the delayed signal.

[0086] The time derivative can be understood as the instantaneous rate of change of the delayed signal.

[0087] Optionally, for each delayed signal in the delayed feature sequence, a fourth-order central difference processing can be performed on the delayed signal to obtain the corresponding time derivative value. For example, the fourth-order central difference processing can refer to the following formula (3). Where, v j (t) represents the j-th delayed signal in the delayed feature sequence; t k This represents the k-th time point where the derivative is to be calculated; v j (t k ) represents the value of the j-th delayed signal at the k-th time sampling point, v j (t k+2 ) represents the value of the j-th delayed signal at the (k+2)-th time sampling point, v j (t k+1 ) represents the value of the j-th delayed signal at the (k+1)-th time sampling point, v j (t k-1 ) represents the value of the j-th delayed signal at the (k-1)-th time sampling point, v j (t k-2 () represents the value of the j-th delayed signal at the (k-2)-th time sampling point; For v j (t) at t k The first time derivative at time t is used to characterize the j-th delayed signal t. k The instantaneous rate of change at any given moment; For time step.

[0088] (3)

[0089] S202. Based on the corresponding time derivative values ​​of each delayed signal, construct the Koopman HAVOK linear forced model.

[0090] Optionally, the delayed signals and their corresponding time derivatives can be aligned, and differential boundary points can be removed to obtain a regression dataset. Then, using this regression dataset, a HAVOK linear forced model can be constructed. For example, the first r delayed signals can be used as a linearly interpretable state vector, i.e. The r-th delayed signal v r (t k ) is defined as the forcing operator, i.e., u(t) k )=v r (t k Based on this, the HAVOK linear forced model can be referred to the following formula (4).

[0091] (4)

[0092] in, A is the time derivative of the master state vector; A is the linear dynamic matrix (system matrix), used to describe the linear evolution of the master state vector z(t) itself, and determines the main linear dynamic characteristics of the system. B is the forced coupling matrix (input matrix), used to determine how the forced operator u(t) acts on each component of the main state, controlling the intensity and direction of the nonlinear effect. .

[0093] Furthermore, the least squares method can be used to solve for A and B in the HAVOK linear forced model. Specifically, the state matrix can be set as Z = [z(t1), z(t2), ..., z(t... M ], derivative matrix , forcing vector U=[u(t1),u(t2),...,u(t M Then, the augmented regression matrix shown in formula (5) can be constructed. Processing formula (5) yields the solutions for A and B shown in formula (6). Wherein, This indicates a pseudo-inverse. Furthermore, during the solution process for A and B, methods such as ridge regression (L2 regularization) can be incorporated to improve stability under noisy conditions.

[0094] (5)

[0095] (6)

[0096] Based on the above process, it can be seen that the magnitude of the forcing operator is s(t) k )=|u(t k )|=|v r (t k )|.

[0097] In the embodiments of this application, by performing differential processing on each delayed signal and constructing a HAVOK linear forced model based on the differential processing results, the rationality of the construction of the HAVOK linear forced model can be guaranteed.

[0098] Based on the above embodiments, this application provides an optional method for determining fault detection results, such as... Figure 3 As shown, the specific steps include:

[0099] S301, compare the magnitude of the forcing operator during the fitting process with at least one fault threshold.

[0100] The fault threshold can be understood as a numerical value that measures the magnitude of the forcing operator.

[0101] In one optional implementation, at least one fault threshold can be determined based on the historical current sequence of candidate measuring points under current anomaly scenarios in a low-voltage distribution network. The current anomaly scenarios include at least one of arc faults, non-arc faults, and non-arc disturbances. Non-arc faults can be short-circuit faults, etc.; non-arc disturbances can include load switching, etc. Candidate measuring points can be other measuring points of the same type as the target measuring point, or they can be the target measuring point itself; the historical current sequence can be understood as the current sequence flowing through the candidate measuring points under current anomaly scenarios.

[0102] Optionally, for each current anomaly scenario, historical current sequences of candidate measurement points under that scenario can be obtained. Then, a corresponding Hankel matrix is ​​constructed based on the historical current sequence; a delay feature sequence is determined based on the Hankel matrix; and a HAVOK linear forced model is constructed based on the delay feature sequence. The HAVOK linear forced model is then fitted to obtain the amplitude of the forcing operator for the candidate measurement points under that current anomaly scenario. Furthermore, the amplitude range of the forcing operator corresponding to multiple candidate measurement points under that current anomaly scenario is statistically analyzed to determine the fault threshold corresponding to that current anomaly scenario.

[0103] Optionally, the magnitude of the forcing operator can be compared with at least one fault threshold to obtain a comparison result.

[0104] S302. Based on the comparison results, determine the fault detection result of the target measuring point.

[0105] Optionally, the fault detection result of the target measuring point can be determined based on the current anomaly characterized by the comparison results.

[0106] For example, the fault threshold corresponding to non-arc disturbance can be 0.045, that is, when the amplitude of the forcing operator is less than 0.045, the fault detection result of the target measuring point is that there is a non-arc disturbance at the target measuring point; the fault threshold corresponding to arc fault can be 0.06 and 0.08, that is, when the amplitude of the forcing operator is in the range of [0.06, 0.08], the fault detection result of the target measuring point is that there is an arc fault at the target measuring point; the fault threshold corresponding to non-arc fault can be 0.2, that is, when the amplitude of the forcing operator is greater than 0.2, the fault detection result of the target measuring point is that there is a non-arc fault at the target measuring point.

[0107] It is worth noting that when the amplitude of the forcing operator is in the range of [0.045, 0.06] or (0.18, 0.2], an uncertain state can be output, triggering a secondary criterion to further reduce false alarms. This secondary criterion can include duration, phase consistency, and the number of repeated bursts. Specifically, duration determines how long the amplitude remains within the interval, excluding transient spikes; phase consistency determines whether the forcing operators of all three phases simultaneously malfunction, avoiding false alarms due to single-phase noise; and the number of repetitions determines whether multiple triggers occur within a short period, distinguishing between single disturbances and persistent faults.

[0108] In this embodiment of the application, by comparing the amplitude of the forced operator with the fault threshold, the fault detection result of the target measurement point can be obtained, which can improve the accuracy of the fault detection result determination.

[0109] Figure 4 This is a flowchart illustrating a fault detection method for a low-voltage distribution network in another embodiment. Based on the above embodiments, this embodiment provides an optional example of a fault detection method for a low-voltage distribution network. (Combined with...) Figure 4 The specific implementation process is as follows:

[0110] S401 processes each current signal in the current current sequence of the target measurement point in the low-voltage distribution network according to the delay embedding dimension and delay step size to obtain the delay vector corresponding to the current signal.

[0111] S402, construct the Hankel matrix based on the delay vectors corresponding to each current signal, and perform singular value decomposition on the Hankel matrix to obtain the right singular vector matrix.

[0112] S403, determine the delayed feature sequence from the right singular vector matrix based on the effective rank parameter.

[0113] S404 performs differential processing on each delayed signal in the delayed feature sequence to obtain the time derivative value corresponding to the delayed signal.

[0114] S405. Based on the corresponding time derivative values ​​of each delayed signal, construct the HAVOK linear forced model and perform fitting processing on the HAVOK linear forced model.

[0115] S406, compare the amplitude of the forced operator during the fitting process with at least one fault threshold, and determine the fault detection result of the target measurement point based on the comparison result.

[0116] The forcing operator is the delayed signal located at the end of the delayed feature sequence.

[0117] Optionally, at least one fault threshold is determined based on the historical current sequence of candidate measurement points under current anomaly scenarios in the low-voltage distribution network; wherein, the current anomaly scenarios include at least one of arc fault, non-arc fault, and non-arc disturbance.

[0118] The specific processes of S401-S406 described above can be found in the description of the above method embodiments. Their implementation principles and technical effects are similar, and will not be repeated here.

[0119] It should be understood that although the steps in the flowcharts of the embodiments described above 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 restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0120] Based on the same inventive concept, this application also provides a fault detection device for a low-voltage distribution network to implement the fault detection method for the low-voltage distribution network described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the fault detection device for low-voltage distribution networks provided below can be found in the limitations of the fault detection method for low-voltage distribution networks described above, and will not be repeated here.

[0121] In one exemplary embodiment, such as Figure 5 As shown, a fault detection device 1 for a low-voltage distribution network is provided, comprising: a matrix construction module 10, a sequence determination module 20, a model construction module 30, and a fault detection module 40, wherein:

[0122] Matrix construction module 10 is used to construct a Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network;

[0123] Sequence determination module 20 is used to determine the delayed feature sequence based on the Hankel matrix;

[0124] Model building module 30 is used to build a Koopman HAVOK linear forced model based on the delayed feature sequence;

[0125] The fault detection module 40 is used to fit the HAVOK linear forced model and determine the fault detection result of the target measurement point based on the amplitude of the forced operator during the fitting process; wherein, the forced operator is the delayed signal located at the end of the delayed feature sequence.

[0126] In one exemplary embodiment, the matrix construction module 10 is specifically used for:

[0127] For each current signal in the current current sequence of the target measurement point in the low-voltage distribution network, the current signal is processed according to the delay embedding dimension and delay step size to obtain the delay vector corresponding to the current signal; based on the delay vector corresponding to each current signal, a Hankel matrix is ​​constructed.

[0128] In one exemplary embodiment, the sequence determination module 20 is specifically used for:

[0129] Singular value decomposition is performed on the Hankel matrix to obtain the decomposition result, which includes the right singular vector matrix. Based on the effective rank parameter, the delayed feature sequence is determined from the right singular vector matrix.

[0130] In one exemplary embodiment, the model building module 30 is specifically used for:

[0131] For each delayed signal in the delayed feature sequence, the delayed signal is differentially processed to obtain the corresponding time derivative value; based on the corresponding time derivative values ​​of each delayed signal, a Koopman HAVOK linear forced model is constructed.

[0132] In one exemplary embodiment, the fault detection module 40 is specifically used for:

[0133] The amplitude of the forced operator during the fitting process is compared with at least one fault threshold; based on the comparison result, the fault detection result of the target measurement point is determined.

[0134] In an exemplary embodiment, the fault detection device 1 for a low-voltage distribution network further includes a threshold determination module, wherein the threshold determination module is specifically used for:

[0135] Based on the historical current sequence of candidate measurement points under current abnormality scenarios in low-voltage distribution networks, at least one fault threshold is determined; wherein, current abnormality scenarios include at least one of arc fault, non-arc fault, and non-arc disturbance.

[0136] Each module in the aforementioned low-voltage distribution network fault detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0137] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores current operating data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a fault detection method for a low-voltage power distribution network.

[0138] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0139] 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.

[0140] 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 above method embodiments.

[0141] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0142] It should be noted that the data involved in this application (including but not limited to current operation data) are all data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0143] 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 above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0144] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0145] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fault detection method for a low-voltage distribution network, characterized in that, The method includes: Construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network; Determine the delayed feature sequence based on the Hankel matrix; Based on the aforementioned delay feature sequence, a Koopman HAVOK linear forced model is constructed; The HAVOK linear forced model is fitted, and the fault detection result of the target measurement point is determined based on the amplitude of the forced operator during the fitting process; wherein, the forced operator is the delayed signal located at the end of the delayed feature sequence.

2. The method according to claim 1, characterized in that, The construction of the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network includes: For each current signal in the current current sequence of the target measurement point in the low-voltage distribution network, the current signal is processed according to the delay embedding dimension and the delay step size to obtain the delay vector corresponding to the current signal. Based on the delay vectors corresponding to each current signal, construct the Hankel matrix.

3. The method according to claim 1, characterized in that, The step of determining the delayed feature sequence based on the Hankel matrix includes: Singular value decomposition is performed on the Hankel matrix to obtain the decomposition result; wherein, the decomposition result includes the right singular vector matrix; The delayed feature sequence is determined from the right singular vector matrix based on the effective rank parameter.

4. The method according to claim 1, characterized in that, The step of constructing the Koopman-Havok linear forced model based on the delayed feature sequence includes: For each delayed signal in the delayed feature sequence, the delayed signal is differentially processed to obtain the time derivative value corresponding to the delayed signal; Based on each delayed signal and its corresponding time derivative, a Koopman-HAVOK linear forced model is constructed.

5. The method according to claim 1, characterized in that, The step of determining the fault detection result of the target measurement point based on the amplitude of the forcing operator during the fitting process includes: The magnitude of the forcing operator during the fitting process is compared with at least one fault threshold. Based on the comparison results, the fault detection result of the target measuring point is determined.

6. The method according to claim 1, characterized in that, The method further includes: Based on the historical current sequence of candidate measurement points under the current anomaly scenario in the low-voltage distribution network, at least one fault threshold is determined; wherein, the current anomaly scenario includes at least one of arc fault, non-arc fault and non-arc disturbance.

7. A fault detection device for a low-voltage distribution network, characterized in that, The device includes: The matrix construction module is used to construct the Hankel matrix based on the current current sequence of the target measurement point in the low-voltage distribution network. The sequence determination module is used to determine the delayed feature sequence based on the Hankel matrix; The model building module is used to construct a Koopman HAVOK linear forced model based on the delay feature sequence. The fault detection module is used to fit the HAVOK linear forced model and determine the fault detection result of the target measurement point based on the amplitude of the forced operator during the fitting process; wherein the forced operator is the delayed signal located at the end of the delayed feature sequence.

8. 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 according to any one of claims 1 to 6.

9. 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 according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images