Differential protection abnormal graph signal diagnosis method and device based on substation recording wave

Abstract

The application discloses a differential protection abnormal graph signal diagnosis method and device based on a power station recording wave, and the method comprises the following steps: obtaining power station recording wave data and related parameters, and calculating differential currents according to the power station recording wave data and related parameters; constructing a current graph signal model based on power station differential protection according to the differential currents; performing graph Fourier transform calculation on the graph signal according to the current graph signal model to obtain characteristic values and graph Fourier transform coefficients of the graph signal; analyzing the differential current change condition based on the power station differential protection according to the graph Fourier transform coefficients, and calculating the differential current change importance of each node in the current graph signal model; and judging the abnormal node of the differential protection according to the differential current change importance of each node. The method is sensitive to the change of the GFT coefficient of the graph signal caused by the abnormal differential current during the fault, and therefore, more abnormal conditions can be diagnosed and identified under the condition that the differential current is very small.

Description

Technical Field

[0001] This invention relates to the field of differential protection technology for power plants, and specifically to a method and apparatus for diagnosing differential protection anomaly signals based on power plant waveform recording. Background Technology

[0002] Differential protection is a crucial primary protection system in power grids, playing a vital role in the power system. While its principle is simple and its sensitivity to faults within the protection zone is high, it is susceptible to various factors, and maloperation is common during faults outside the protection zone. For example, inrush currents and asynchronous closing of circuits outside the protection zone have caused maloperation of generator differential protection. Several cases of busbar protection maloperation due to faults outside the protection zone have also been reported. Transformer differential protection, due to its larger unbalanced current and more influencing factors, is even more prone to maloperation during faults outside the protection zone. To prevent maloperation during faults outside the differential protection zone, in-depth research was conducted on preventing maloperation of generator differential protection and transformer differential protection, and several preventative measures were proposed, achieving good results. In addition to preventative measures, improving safety during faults outside the differential protection zone and preventing maloperation requires strengthening the detection of differential currents in various differential protection systems to diagnose potentially abnormal differential protection and implement targeted measures.

[0003] Load testing of differential protection is a crucial method for detecting differential circuits, revealing numerous issues. However, due to load current limitations, it struggles to identify certain anomalies, such as incorrect TV parameters or discrepancies between transient and steady-state current transformer (CT) characteristics. During external faults, the increased fault current leads to a corresponding increase in unbalanced currents caused by differential protection component parameters or CT characteristics. Utilizing fault waveform data from external faults to diagnose the differential current of the differential protection system can uncover more anomalies undetectable under load current conditions. Because the unbalanced current during external faults remains relatively small, and due to safety-enhancing safety features, differential protection systems often fail to activate and record differential current conditions during external faults. While the differential current used in various component differential protection systems may differ in reference selection and balance coefficients, fault waveform recorders can calculate the differential current of each component within the plant, indirectly diagnosing any abnormalities in the differential protection. When there is a fault outside the zone, the current flowing through the differential protection of generators and other devices inside the plant is basically the same, so their differential currents are also not much different. The simplest way is to directly compare and analyze the differential currents. However, since the unbalanced current is still very small, the ability to detect abnormalities is limited. Abnormal differential protection can only be diagnosed and identified when the fault outside the zone is very serious and the differential current is large.

[0004] The main methods of existing differential protection in power plants have the following disadvantages: (1) Although load testing of differential protection is a very important means of differential circuit detection and can find many problems of differential protection, it is difficult to identify some anomalies due to the limitation of load current, such as TV parameter errors, and the difference between transient and steady-state errors of TA characteristics. (2) Graph signal processing (GSP) technology has developed rapidly in recent years. Its key idea is to extend the concept of digital signal processing (DSP) to the data connected on the graph. GSP can define classic signal processing concepts such as filtering, sampling and modulation for signals related to the underlying graph structure. GSP is increasingly widely used in graph signal mutation detection, graph localization, graph clustering and graph denoising. GSP is less used in power systems and is currently mainly used in non-intrusive load detection, power grid anomaly attack data detection, power grid PMU data anomaly detection, etc.

[0005] In particular, existing differential protection methods for power plants cannot accurately and effectively analyze situations where faults occur outside the power plant area, and where the differential current is small but the differential protection malfunctions, potentially leading to false tripping of the differential protection. Summary of the Invention

[0006] The technical problem to be solved by this invention is that existing differential protection methods for power plants cannot accurately and effectively analyze situations where faults occur outside the power plant area, the differential current is small, and the differential protection malfunctions, leading to false tripping of the differential protection.

[0007] The purpose of this invention is to provide a method and device for diagnosing differential protection anomalies based on substation waveform recordings. This method involves modeling the differential current waveform signal using substation waveform recording data and then employing graphical Fourier transform (GFT) for differential protection anomaly diagnosis. First, a current waveform signal model based on substation differential protection is constructed using fault waveform recording data and related parameters. Second, the graphical Fourier transform (GFT) coefficients of this current waveform signal model are calculated. Finally, based on the variation characteristics of the GFT coefficients, a diagnostic method is proposed that considers the importance of differential current changes in the GFT coefficients to diagnose potentially abnormal differential protection. Actual field waveform recording data verifies the correctness and effectiveness of this method. This method utilizes the characteristic that abnormal differential current during a fault causes changes in the mid-to-high frequency GFT coefficients of the waveform signal, thus enabling the identification of more anomalies even with small differential currents. This method is more effective than directly analyzing the differential current in diagnosing and identifying whether differential protection is abnormal and identifying abnormal components.

[0008] This invention is achieved through the following technical solution:

[0009] In a first aspect, the present invention provides a differential protection anomaly signal diagnosis method based on substation waveform recording, the method comprising:

[0010] Acquire power plant waveform data and related parameters, and calculate differential current based on power plant waveform data and related parameters;

[0011] Based on the differential current, a current diagram signal model based on substation differential protection is constructed;

[0012] Based on the current graph signal model, the graph signal is subjected to graph Fourier transform calculation to obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal;

[0013] Based on the Fourier transform coefficients, the differential current variation of the substation differential protection is analyzed, and the importance of the differential current variation of each node in the current graph signal model is calculated.

[0014] Based on the importance of the differential current changes at each node, the abnormal nodes of the differential protection are identified.

[0015] The power plant and substation waveform data and related parameters include: fault waveform information of power plants and substations when a fault occurs, actual wiring diagram of the power plant and substation at the time of the fault, and configuration parameters of differential protection for generators, transformers and busbars.

[0016] Furthermore, based on the differential current, a current diagram signal model for substation differential protection is constructed, including:

[0017] Using the effective value amplitude of the differential current as the node signal of the graph and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; whereby the differential current is the differential current of bus / transformer / generator differential protection, etc.

[0018] Furthermore, the formula for calculating the node signal f(i) of the graph is:

[0019] f(i) = I di

[0020] In the formula, I di The effective value amplitude of the differential current for busbars / transformers / generators, etc.

[0021] The weight w of the graph ij The calculation formula is:

[0022] w ij =I ij

[0023] In the formula, I ij Let be the per-unit value of the effective amplitude of the branch current connecting nodes i and j in the graphical signal model.

[0024] Furthermore, based on the Fourier transform coefficients, the variation of differential current in the substation differential protection is analyzed, including:

[0025] Under normal operating conditions, the current in each branch is the load current, and the differential current of each component in the substation is also very small. The weight of the node differential current in each branch changes very little. At this time, the differential current fluctuation of the graphical signal based on the substation differential protection is small, and the graphical signal is smooth, i.e., the characteristic value λ n The Fourier transform coefficients f(λ) are relatively small. n It's also very small;

[0026] When a fault occurs outside the differential protection zone of the substation, the current in each branch is a relatively large fault current, and the differential current at the nodes also increases slightly. At this time, the characteristic value λ of the graphical signal based on the substation differential protection... n The Fourier transform coefficients f(λ) are relatively large. n The difference also increases slightly when the overall differential flow difference is not significant. n It is also relatively small;

[0027] When the differential current of some nodes in the current graph signal model changes more significantly than that of other nodes, an anomaly will occur, resulting in larger mid-to-high frequency components in the graph Fourier transform coefficients f(λ). n Furthermore, nodes with large differential current variations contribute more to the Fourier transform coefficients of the mid-to-high frequency band diagram than other nodes with smaller differential currents.

[0028] By utilizing the signal variation characteristics of differential current as shown in the figure above, and by analyzing the variation characteristics of the Fourier transform coefficients in the mid-to-high frequency band, the differential protection nodes of the substation with abnormal differential currents during external faults can be diagnosed.

[0029] Furthermore, the formula for calculating the importance I(i) of the differential current change at each node is as follows:

[0030] I(i)=k(i) max +k(i) max-1

[0031] k(i) n =f(i)u n (i) / f(λ n )

[0032] In the formula, k(i) max and k(i) max-1 The contribution ratio of the graph Fourier transform coefficients corresponding to the two nodes with the largest graph Fourier transform coefficients in the mid-to-high frequency band; k(i) max For the mid-to-high frequency band f(λ) n Maximum value f(λ) max The corresponding k(i) n The value, i.e., k(i). max =f(i)u max (i) / f(λ max ), k(i)max-1 For the mid-to-high frequency band f(λ) n The second largest value f(λ) max-1 The corresponding k(i) n The value, i.e., k(i). max-1 =f(i)u max-1 (i) / f(λ max-1 );

[0033] f(i) is the signal value of node i; u n (i) is the feature vector u n The i-component; f(λ) n () is based on the Fourier transform of the graph with respect to the eigenvalue λ n The coefficients, i.e., the graphical Fourier transform coefficients; k(i) n The feature value contributed by node i is λ. n The proportion of the Fourier transform coefficients in the graph.

[0034] Furthermore, based on the importance of the differential current changes at each node, the abnormal nodes of the differential protection are identified, including:

[0035] Based on the importance of the differential current change at each node, the node with the largest deviation in importance from other nodes is diagnosed as the differential current abnormal node of the differential protection.

[0036] Secondly, the present invention provides a differential protection anomaly signal diagnostic device based on substation waveform recording, the device comprising:

[0037] The acquisition unit is used to acquire power plant waveform recording data and related parameters.

[0038] The differential current calculation unit is used to calculate the differential current based on the power plant waveform data and related parameters.

[0039] The current diagram signal construction unit is used to construct a current diagram signal model based on the differential current of the substation differential protection.

[0040] The graph Fourier transform unit is used to perform graph Fourier transform calculations on the graph signal based on the current graph signal model, and obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal.

[0041] The differential current change importance calculation unit is used to analyze the differential current change based on the differential protection of the substation according to the Fourier transform coefficients of the graph, and to calculate the importance of the differential current change of each node in the current graph signal model.

[0042] The abnormal node judgment unit is used to determine the abnormal nodes of the differential protection based on the importance of the differential current change of each node.

[0043] Furthermore, the execution process of the graph signal construction unit is as follows:

[0044] Using the effective value amplitude of the differential current as the node signal of the graph and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; whereby the differential current is the differential current of bus / transformer / generator differential protection, etc.

[0045] Thirdly, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the differential protection anomaly signal diagnosis method based on substation waveform recording.

[0046] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the differential protection anomaly signal diagnosis method based on substation waveform recording.

[0047] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0048] This invention relates to a method and apparatus for diagnosing differential protection anomalies based on substation waveform recordings. First, a current waveform signal model of the substation's differential protection is constructed based on fault waveform recording data and related parameters. Second, the waveform signal model's waveform Fourier transform is performed to calculate its coefficients. Finally, based on the variation characteristics of the waveform Fourier transform coefficients, a diagnostic method is proposed that considers the importance of differential current changes in the waveform Fourier transform coefficients to diagnose potentially abnormal differential protection. Actual field waveform recording data verifies the correctness and effectiveness of this method. This invention utilizes the characteristic that abnormal differential current during a fault causes changes in the mid-to-high frequency GFT coefficients of the waveform signal, making it highly sensitive to changes. Therefore, it can diagnose and identify more anomalies even with very small differential currents; compared to directly analyzing the differential current, it is more effective in diagnosing and identifying whether the differential protection is abnormal and identifying abnormal components. Attached Figure Description

[0049] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0050] Figure 1 This is a flowchart of the differential protection anomaly signal diagnosis method based on power plant waveform recording according to the present invention;

[0051] Figure 2 This is a schematic diagram of the differential protection signal for a 220kV substation with a double busbar connection, as shown in Embodiment 1 of the present invention.

[0052] Figure 3 This is a flowchart of Embodiment 1 of the present invention;

[0053] Figure 4 This is a topology diagram for abnormal diagnosis of power plant differential protection according to Embodiment 1 of the present invention;

[0054] Figure 5 The waveform diagrams for the line, bus tie, and main transformer branches in Embodiment 1 of the present invention are shown below.

[0055] Figure 6 This is a diagram showing the diagnostic analysis results of phase B differential protection under normal conditions in Embodiment 1 of the present invention;

[0056] Figure 7 This is a diagram showing the diagnostic results of incorrect main transformer (CT) parameter settings for bus differential protection in Embodiment 1 of the present invention.

[0057] Figure 8 This is a diagram showing the diagnostic results of incorrect bus coupler (TA) parameter settings for the bus differential protection in Embodiment 1 of the present invention.

[0058] Figure 9 This is a diagram showing the diagnostic results of incorrect TA parameter settings for generator differential protection in Embodiment 1 of the present invention.

[0059] Figure 10 This is a block diagram of the differential protection anomaly signal diagnostic device based on power plant waveform recording according to the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0061] Existing differential protection methods for power plants cannot accurately and effectively analyze situations where faults occur outside the power plant area, and the differential current is small, but the differential protection malfunctions, leading to false tripping of the differential protection.

[0062] Therefore, this invention designs a differential protection abnormality diagram signal diagnosis method and device based on substation waveform recording. It utilizes the characteristic that the GFT coefficient of the diagram signal is more sensitive to changes in the mid-to-high frequency band due to abnormal differential current during a fault. Therefore, it can diagnose and identify more abnormal situations when the differential current is very small. Compared with direct analysis using differential current, it can more effectively diagnose and identify whether the differential protection is abnormal and abnormal components.

[0063] (1) Current Diagram Signal Modeling Based on Substation Differential Protection: Current diagram signal modeling based on substation differential protection uses the amplitude of the effective value of the differential current of the busbar, transformer, generator, etc., as the node signal, and the amplitude of the effective value of the branch fault current as the weight of the diagram. This modeling method essentially analyzes and diagnoses the change of the differential current of the differential protection in the weight of the fault branch current. Since the differential current is mainly unbalanced current when there is a fault outside the differential protection zone, this method actually diagnoses and analyzes the change of unbalanced current, that is, the change of the safety margin of the differential protection when there is a fault outside the zone. Since the secondary windings used by the fault recorder and the component differential protection are different, and their own sampling and calculation links are also different, directly using the waveform recording and differential current of the differential protection device for diagnostic analysis can more accurately diagnose the abnormal differential current of the differential protection device.

[0064] (2) Differential Protection Anomaly Diagnosis and Analysis Method Based on Graph Fourier Transform: Anomaly diagnosis and analysis of the constructed differential current graph signal is performed using the graph Fourier transform method, which has a strong anomaly detection capability in graph signal processing technology. The graph Fourier transform uses the eigenvectors of the graph's Laplace matrix as the transform basis to decompose the graph signal into graph signals with different smoothness levels, just as the traditional Fourier transform decomposes a function into functions of different frequencies.

[0065] Example 1

[0066] like Figure 1 As shown, the present invention provides a differential protection anomaly signal diagnosis method based on substation waveform recording. This method includes:

[0067] Step 1: Obtain the power plant waveform data and related parameters, and calculate the differential current based on the power plant waveform data and related parameters; based on the differential current, construct a current diagram signal model based on the power plant differential protection.

[0068] The power plant and substation waveform recording data and related parameters include: fault waveform recording information of power plants and substations when a fault occurs, actual wiring diagram of the power plant and substation at the time of the fault, and configuration parameters of differential protection for generators, transformers and busbars.

[0069] Specifically, based on the differential current, a current diagram signal model for substation differential protection is constructed, including:

[0070] Using the effective value amplitude of the differential current as the node signal of the graph and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; whereby the differential current is the differential current of bus / transformer / generator differential protection, etc.

[0071] Specifically, the graph signal in the current graph signal model is a signal defined at the nodes. Using the differential current of the differential protection as the node signal, the calculation formula for the node signal f(i) of the graph is:

[0072] f(i) = I di (1)

[0073] In the formula, I di This refers to the effective value amplitude of the differential current for the busbar / transformer / generator, etc. For ease of comparison, the differential current is calculated using a per-unit value of 100MVA on this side. When calculating the differential current based on the waveform information, a general calculation method or the manufacturer's technical specifications can be used, followed by a unified calculation. If the differential current of the protection device is available, using the differential current calculated by the device will be more accurate.

[0074] The weights of the graph (i.e., the weights of the graph signals) w ij The calculation formula is:

[0075] w ij =I ij (2)

[0076] In the formula, I ij Let be the per-unit value of the effective amplitude of the branch current connecting nodes i and j in the graphical signal model.

[0077] Since the current on the opposite side of the line cannot be obtained at this station, the differential unbalanced current of the fiber optic line cannot be obtained. Therefore, the branch currents of the line are not considered in the modeling. However, the unbalanced current calculation of the bus differential node needs to consider all connected branches. The low-voltage side bus of the main transformer is not equipped with a bus differential, so its corresponding nodes and branches are not considered. In some renewable energy stations that are equipped with bus differentials, they can be included in the graphical signal model.

[0078] Figure 2 This is a schematic diagram of the differential protection signal for a 220kV substation with a double busbar connection. Both the 220kV and 110kV sides are equipped with bus couplers and bus differential protection. Main transformer 1 and main transformer 2 are connected to bus I and bus II, respectively.

[0079] Step 2: Based on the current graph signal model, perform a graph Fourier transform on the graph signal to obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal; specifically including:

[0080] 2.1 Graph Laplacian Matrix and Eigenvectors

[0081] The Laplace matrix L of a graph is defined as:

[0082] L = DW (3)

[0083] Where D is the degree matrix of graph G, which is a diagonal matrix, D = diag[d1, d2, ..., d]. N ],d i Let be the degree of the i-th node, and let be the sum of the weights of the edges associated with node i. Let W be the weight matrix of the graph.

[0084] The set of eigenvalues ​​of the graph's Laplacian matrix L is called the graph's Laplacian spectrum. The spectrum of a graph G with N nodes is:

[0085] λ(g)={λ0,λ1,...,λ N-1} (4)

[0086] Where, 0 = λ0 ≤ λ1 ≤ λ2 ... ≤ λ N-1 For each eigenvalue, the corresponding eigenvector is:

[0087] U = [u0, u1, ..., u N-1 (5)

[0088] For an undirected graph with positive edge weights, the eigenvalues ​​and eigenvectors of L are all real numbers, and it has a complete set of orthonormal eigenvectors.

[0089] 2.2 Graph Fourier Transform (GFT)

[0090] The graphical Fourier transform uses the eigenvector U of formula (5) as the transform basis. The graphical Fourier transform (GFT) of the graphical signal f is defined as follows:

[0091] f(λ n )=U T f (6)

[0092] Where f(λ) n ) is the GFT with respect to the eigenvalue λ n The coefficients are called graphical Fourier transform coefficients (GFT coefficients). U is the eigenvector of formula (5), and f is the graphical signal vector. The graphical frequency is the eigenvalue λ(g) of the graphical Laplace matrix, similar to the harmonics of the classical signal Fourier transform, while the corresponding GFT coefficients are similar to the magnitude of the harmonic components. Smaller eigenvalues ​​correspond to low frequencies, and larger eigenvalues ​​correspond to high frequencies. λ0 = 0 corresponds to zero frequency, where all values ​​of the graphical signal are the same and do not change, equivalent to the DC component in classical signal processing.

[0093] When the GFT coefficients corresponding to low-frequency eigenvalues ​​are large and the GFT coefficients corresponding to high-frequency eigenvalues ​​are small, it means that the graph signal changes relatively slowly and the graph signal is relatively smooth; while when the GFT coefficients corresponding to high-frequency eigenvalues ​​are large, it means that the graph signal changes relatively large and the graph signal oscillates and fluctuates more.

[0094] Step 3: Analyze the differential current variation based on the graph Fourier transform coefficients, calculate the importance of differential current variation at each node in the current graph signal model, and determine the abnormal nodes of the differential protection based on the importance of differential current variation at each node.

[0095] Specifically, based on the Fourier transform coefficients, the differential current variation of the substation differential protection is analyzed, including:

[0096] Under normal operating conditions, the current in each branch is the load current, and the differential current of each component in the substation is also very small. The weight of the node differential current in each branch changes very little. At this time, the differential current fluctuation of the graphical signal based on the substation differential protection is small, and the graphical signal is smooth, i.e., the characteristic value λ n The Fourier transform coefficients f(λ) are relatively small. n It's also very small;

[0097] When a fault occurs outside the differential protection zone of the substation, the current in each branch is a relatively large fault current, and the differential current at the nodes also increases slightly. At this time, the characteristic value λ of the graphical signal based on the substation differential protection... n The Fourier transform coefficients f(λ) are relatively large. n The difference also increases slightly when the overall differential flow difference is not significant. n It is also relatively small;

[0098] When the differential current of some nodes in the current graph signal model changes more significantly than that of other nodes, an anomaly will occur, resulting in larger mid-to-high frequency components in the graph Fourier transform coefficients f(λ). n Furthermore, nodes with large differential current variations contribute more to the Fourier transform coefficients of the mid-to-high frequency band diagram than other nodes with smaller differential currents.

[0099] By utilizing the signal variation characteristics of differential current as shown in the figure above, and by analyzing the variation characteristics of the Fourier transform coefficients in the mid-to-high frequency band, the differential protection nodes of the substation with abnormal differential currents during external faults can be diagnosed.

[0100] Specifically, in the graphical Fourier transform coefficients (GFT coefficients) corresponding to the mid-to-high frequency eigenvalues, nodes with large signal changes near their nodes account for a larger proportion of the GFT coefficients, while nodes with small changes near their nodes account for a smaller proportion of the GFT coefficients. The eigenvalue contributed by node i is λ. n The proportion k(i) in the GFT coefficients n for:

[0101] k(i) n =f(i)u n (i) / f(λ n (7)

[0102] In the formula, f(i) is the signal value of node i; un (i) is the feature vector u n The i-component; f(λ) n () is based on the Fourier transform of the graph with respect to the eigenvalue λ n The coefficients, i.e., the graphical Fourier transform coefficients; k(i) n The feature value contributed by node i is λ n The proportion of the Fourier transform coefficients in the graph;

[0103] Since the GFT coefficient changes significantly when the differential current changes during an external fault, these changes are reflected in two components in the mid-to-high frequency band. Therefore, k(i) of the two nodes with the largest GFT coefficients in the mid-to-high frequency band is used. n The sum is used as a criterion to diagnose abnormal nodes in differential current.

[0104] The importance I(i) of the differential current change of each node signal in the differential current diagram signal of differential protection is defined as:

[0105] I(i)=k(i) max +k(i) max-1 (8)

[0106] In the formula, k(i) max and k(i) max-1 The contribution ratio of the graph Fourier transform coefficients corresponding to the two nodes with the largest graph Fourier transform coefficients in the mid-to-high frequency band; k(i) max For the mid-to-high frequency band f(λ) n Maximum value f(λ) max The corresponding k(i) n The value, i.e., k(i). max =f(i)u max (i) / f(λ max ), k(i) max-1 For the mid-to-high frequency band f(λ) n The second largest value f(λ) max-1 The corresponding k(i) n The value, i.e., k(i). max-1 =f(i)u max-1 (i) / f(λ max-1 ).

[0107] Since the actual differential current changes are very complex, there may be a situation where only one high-frequency component changes during the fault period, or many high-frequency components change simultaneously. In this case, depending on the GFT component situation, only one or three k(i) can be used. n The sum is used to calculate the value of I(i).

[0108] Specifically, based on the importance of the differential current changes at each node, abnormal nodes of the differential protection are identified, including:

[0109] Based on the importance of differential current changes at each node, the node with the largest deviation in importance from other nodes is diagnosed as the node with abnormal differential current in the differential protection. Since the recorded waveform data is dynamic data during the fault period, I(i) can reflect any changes in differential current at different times during the fault period. The implementation flowchart is as follows: Figure 3 As shown.

[0110] In practice, when a phase-to-phase (BC) fault occurred on an outgoing line of a 220kV power plant, the abnormal situation of the differential protection within the substation was analyzed and verified. The network topology diagram is as follows: Figure 4 As shown, there are 12 nodes in total, including 220kV I and II bus differential protection, 5 main transformer protection differential protection, and 5 generator protection differential protection. There are 11 branches in total: 1 220kV bus tie branch, 5 transformer branches, and 5 generator branches. This substation is a double busbar connection with a total of 6 generator-transformer units. During a fault, Unit 2 is out of service; generator-transformer units 1, 5, and 6 operate on bus I, and generator-transformer units 3 and 4 operate on bus II. Because the lines are not included in the signal modeling, two lines are not shown. The waveforms of the lines, bus tie, and main transformer branches during a fault, as well as the relevant parameters of the main transformer protection and busbar protection, are as follows: Figure 5 As shown in Tables A2 and A3, the full-wave Fourier transform algorithm is used to calculate the phase B current amplitude of each branch in the waveform recording, and the phase B differential current is calculated separately.

[0111] Table A2 Busbar Protection Related Parameters

[0112]

[0113] Table A3 Transformer Protection Related Parameters

[0114]

[0115] Normally, the abnormal diagnosis results of power plant differential protection are as follows: Figure 6 As shown. Figure 6 The results of phase B analysis are as follows. Figure 6 (a) is a graph showing the variation of the effective value of the differential current amplitude at each node. Due to the influence of the non-periodic component, the differential current is larger in the first cycle from 40 to 60 ms, and the differential current of each branch of the transformer is the largest. However, the maximum amplitude of the differential current is only 0.06 Ie. During the fault period from 50 to 100 ms, the differential current actually decreases. Figure 6 (b) shows the variation of the fundamental frequency and GFT coefficients of the differential current in each frequency band, f(λ) 10 The differential current values ​​of f(λ5) and f(λ5) are relatively large, while the other high-frequency and low-frequency components are very small. Furthermore, there are no significant mid-to-high frequency changes during the 50-100ms fault period, and the values ​​even decrease slightly. The importance of the differential current change I(i) of each node is as follows: Figure 6As shown in (c), the differential pressure of main transformer No. 4 and bus No. 1 was slightly larger during the fault, but the difference with other nodes was not significant and there was no obvious abnormality.

[0116] The evaluation of changes in busbar protection parameters is as follows;

[0117] When the current transformer ratio of the No. 1 main transformer branch of the busbar protection is set from 1600 to 1650, the busbar differential protection diagnosis is as follows: Figure 7 As shown. Because there are many bus differential branches and the current transformer (CT) changes little in non-fault branches, the differential current is small during a fault, only 0.03Ie. Figure 7 As shown in (a). Figure 7 (b) The GFT coefficients f(λ) in the high-frequency band of the figure. 11 The coefficients f(λ5) and f(λ5) increase significantly during the fault period, becoming the two largest GFT coefficient components during the fault period. Figure 7 (c) The significance value I(i) of the differential current change of the bus differential protection differs significantly from other differential protections during a fault, and is superior to diagnostic methods that rely solely on differential current. From Figure 7 (d) In terms of deviation, during the fault period, the deviation of differential current diagnosis is 1 to 1.3, while the importance of differential current change is 2 to 3.5, which is significantly higher than the ability to identify abnormalities by relying solely on differential current.

[0118] When the bus coupler current transformer ratio is set from 3000 to 3050, the differential current diagnosis of the bus differential protection is as follows: Figure 8 As shown. Figure 8 As shown in (a), due to the large bus tie current during a fault, the differential currents of bus I and bus II change significantly during the fault, but are not much different from other differential protections. Figure 8 (b) The GFT coefficient f(λ5) in the high-frequency band of the figure increases significantly. Since only this component changes significantly during the fault, only this component is used when calculating I(i). The importance of the differential current changes of bus I and bus II, I(i), increases more significantly during the fault than other differential protection currents, such as... Figure 8 As shown in (c). Due to the incorrect setting of the bus tie current transformer (TA), the differential current of both bus I and bus II increased. Therefore, the differential current deviation was compared between the node with the largest deviation and the third largest deviation. Figure 8 As shown in (d), the difference in importance of differential current change I(i) of bus differential protection during faults is better than that of other differential protection methods that rely solely on differential current for diagnosis.

[0119] The generator differential protection parameter error diagnosis test is as follows:

[0120] When the differential protection of generator No. 6 has a 2% deviation in the current transformer ratio on one side, the differential current diagnosis result is as follows: Figure 9As shown. At this time, although the change in TA error was very small, the differential current of generator No. 6 increased significantly during the fault, but the magnitude was not very large. Figure 9 (b) In the figure, the GFT coefficient f(λ4) of the frequency band increases significantly during the fault, becoming the largest GFT coefficient component, and f(λ7) increases significantly compared to before the fault, becoming the second largest GFT coefficient component. The differential current importance I(i) value of generator protection No. 6 is significantly higher than that of other differential protections during the fault, such as... Figure 9 As shown in (c), this indicates an abnormality in the differential current of the differential protection for generator No. 6. Figure 9 As shown in (d), the difference in importance of differential current changes during the fault period is 2 to 3 times, which is better than 1 to 1.6 times that of directly using differential current diagnosis method.

[0121] Because the fault is on the line and far from the generator, the unevenness of the overall signal is not as great as that of the bus and transformer protection, and the increase in high-frequency components is not as obvious. However, the change in intermediate frequency components is obvious, so it is also very sensitive to abnormal detection of generator differential protection.

[0122] Based on the above on-site test and verification analysis of actual waveform recordings, it can be seen that the method proposed in this invention, which uses substation fault waveform recording to model the current diagram signal based on substation differential protection, calculates GFT coefficients through graphical Fourier transform, and diagnoses abnormal nodes of differential protection based on the importance of differential current changes constructed according to the characteristics of GFT coefficient changes, can more accurately diagnose and identify abnormal differential currents in generator, transformer, and bus protection when the differential current is slight. It can also effectively diagnose multiple differential protection anomalies when the differential current is small.

[0123] Because the magnitude of the unbalanced current during external faults is often closely related to the fault current, especially when the fault current is large, differences in TA characteristics, saturation, and parameter errors cause the unbalanced current of the faulty component to increase, and the differential current gradient on the connected fault branch current to increase, thus generating larger mid-to-high frequency components. The Fourier transform spectrum analysis method is highly capable of capturing frequency changes during such faults. Using waveform recording data, dynamic diagnosis of frequency characteristic changes during the fault process can be performed. Therefore, this graphical signal processing method can quantitatively and more accurately diagnose and analyze differential protection anomalies during external faults with relatively small differential currents, providing a basis for further diagnosis and analysis of the causes of differential current anomalies, enabling preventative measures to be taken, and avoiding maloperation of differential protection in more serious situations.

[0124] Example 2

[0125] like Figure 10As shown, the difference between this embodiment and Embodiment 1 is that this embodiment provides a differential protection anomaly diagram signal diagnosis device based on substation waveform recording. This device is used to implement the differential protection anomaly diagram signal diagnosis method based on substation waveform recording described in Embodiment 1. The device includes:

[0126] The acquisition unit is used to acquire power plant waveform recording data and related parameters.

[0127] The differential current calculation unit is used to calculate the differential current based on the power plant waveform data and related parameters.

[0128] The current diagram signal construction unit is used to construct a current diagram signal model based on the differential current of the substation differential protection.

[0129] The graph Fourier transform unit is used to perform graph Fourier transform calculations on the graph signal based on the current graph signal model, and obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal.

[0130] The differential current change importance calculation unit is used to analyze the differential current change based on the differential protection of the substation according to the Fourier transform coefficients of the graph, and to calculate the importance of the differential current change of each node in the current graph signal model.

[0131] The abnormal node judgment unit is used to determine the abnormal nodes of the differential protection based on the importance of the differential current change of each node.

[0132] As a further implementation, the execution process of the graph signal construction unit is as follows:

[0133] Using the effective value amplitude of the differential current as the node signal of the graph and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; whereby the differential current is the differential current of bus / transformer / generator differential protection, etc.

[0134] The execution process of each unit can be carried out according to the steps of the differential protection anomaly signal diagnosis method based on substation waveform recording described in Example 1, and will not be described in detail in this example.

[0135] Meanwhile, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the differential protection anomaly signal diagnosis method based on substation waveform recording.

[0136] Meanwhile, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the differential protection anomaly signal diagnosis method based on substation waveform recording.

[0137] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0139] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0140] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0141] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A differential protection anomaly signal diagnosis method based on substation waveform recording, characterized in that, The method includes: Acquire power plant waveform data and related parameters, and calculate differential current based on the power plant waveform data and related parameters; Based on the differential current, a current diagram signal model based on the differential protection of the substation is constructed; Based on the current graph signal model, a graph Fourier transform is performed on the graph signal to obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal. Based on the Fourier transform coefficients of the graph, the differential current variation of the differential protection of the plant is analyzed, and the importance of the differential current variation of each node in the current graph signal model is calculated. Based on the importance of the differential current changes at each node, identify the abnormal nodes of the differential protection. Based on the differential current, a current diagram signal model based on substation differential protection is constructed, including: Using the effective value amplitude of the differential current as the node signal of the graph, and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; wherein, the differential current is the differential current of bus / transformer / generator differential protection; Based on the Fourier transform coefficients of the aforementioned figure, the differential current variation of the substation differential protection is analyzed, including: When the substation is in normal operation, the current in each branch is the load current, the differential current of each component in the substation is small, and the weight of the node differential current in each branch changes little. At this time, the differential current fluctuation of the graph signal based on the substation differential protection is small, the graph signal is smooth, that is, the eigenvalue is small, and the graph Fourier transform coefficient is small. When the fault occurs outside the differential protection zone of the substation, the current in each branch is a large fault current, and the node differential current increases. At this time, the characteristic value of the graphical signal based on the substation differential protection is large, and the graphical Fourier transform coefficient also increases. When the differential current of some nodes in the current graph signal model changes more than that of other nodes, an anomaly will occur, resulting in larger graph Fourier transform coefficients of mid-to-high frequency components. Furthermore, the contribution rate of nodes with large differential current changes to the graph Fourier transform coefficients of mid-to-high frequency bands is greater than that of other nodes with small differential currents. By utilizing the signal variation characteristics of the differential current as shown in the above figure, and by analyzing the variation characteristics of the Fourier transform coefficients in the mid-to-high frequency band, the differential protection nodes of the substation with abnormal differential currents during faults outside the protection zone can be diagnosed.

2. The differential protection anomaly signal diagnosis method based on substation waveform recording according to claim 1, characterized in that, The node signals of the diagram The calculation formula is: ； In the formula, The effective value amplitude of the differential current of the busbar / transformer / generator; The weights of the graph The calculation formula is: ； In the formula, Let be the per-unit value of the effective amplitude of the branch current connecting nodes i and j in the graphical signal model.

3. The differential protection anomaly signal diagnosis method based on substation waveform recording according to claim 1, characterized in that, Importance of differential current changes at each node The calculation formula is: ； ； In the formula, and The contribution ratio of the graph Fourier transform coefficients to the two nodes with the largest graph Fourier transform coefficients in the mid-to-high frequency band; Mid-to-high frequency band Maximum value corresponding Value, that is , Mid-to-high frequency band Second largest value corresponding Value, that is ; The signal value of node i; For feature vectors The i-component; It is based on the graph Fourier transform relative to the eigenvalues The coefficients, i.e., the graph Fourier transform coefficients; The feature value contributed by node i is The proportion of the Fourier transform coefficients in the graph.

4. The differential protection anomaly signal diagnosis method based on substation waveform recording according to claim 1, characterized in that, Based on the importance of the differential current changes at each node, identify the abnormal nodes of the differential protection, including: Based on the importance of the differential current change at each node, the node with the largest deviation in importance from other nodes is diagnosed as the differential current abnormal node of the differential protection.

5. A differential protection anomaly signal diagnostic device based on substation waveform recording, characterized in that, The device includes: The acquisition unit is used to acquire power plant waveform recording data and related parameters. The differential current calculation unit is used to calculate the differential current based on the power plant waveform data and related parameters. The diagram signal construction unit is used to construct a current diagram signal model based on the differential current of the substation differential protection. The graph Fourier transform unit is used to perform graph Fourier transform calculation on the graph signal according to the current graph signal model to obtain the eigenvalues ​​and graph Fourier transform coefficients of the graph signal. The differential current change importance calculation unit is used to analyze the differential current change based on the differential protection of the substation according to the Fourier transform coefficients of the graph, and to calculate the importance of the differential current change of each node in the current graph signal model. The abnormal node judgment unit is used to judge the abnormal nodes of the differential protection based on the importance of the differential current change of each node; The execution process of the graph signal construction unit is as follows: Using the effective value amplitude of the differential current as the node signal of the graph, and the per-unit value of the effective value amplitude of the branch fault current as the weight of the graph, a current graph signal model based on substation differential protection is constructed; wherein, the differential current is the differential current of bus / transformer / generator differential protection; Based on the Fourier transform coefficients of the aforementioned figure, the differential current variation of the substation differential protection is analyzed, including: When the substation is in normal operation, the current in each branch is the load current, the differential current of each component in the substation is small, and the weight of the node differential current in each branch changes little. At this time, the differential current fluctuation of the graph signal based on the substation differential protection is small, the graph signal is smooth, that is, the eigenvalue is small, and the graph Fourier transform coefficient is small. When the fault occurs outside the differential protection zone of the substation, the current in each branch is a large fault current, and the node differential current increases. At this time, the characteristic value of the graphical signal based on the substation differential protection is large, and the graphical Fourier transform coefficient also increases. When the differential current of some nodes in the current graph signal model changes more than that of other nodes, an anomaly will occur, resulting in larger graph Fourier transform coefficients of mid-to-high frequency components. Furthermore, the contribution rate of nodes with large differential current changes to the graph Fourier transform coefficients of mid-to-high frequency bands is greater than that of other nodes with small differential currents. By utilizing the signal variation characteristics of the differential current as shown in the above figure, and by analyzing the variation characteristics of the Fourier transform coefficients in the mid-to-high frequency band, the differential protection nodes of the substation with abnormal differential currents during faults outside the protection zone can be diagnosed.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the differential protection anomaly signal diagnosis method based on substation waveform recording as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the differential protection anomaly signal diagnosis method based on substation waveform recording as described in any one of claims 1 to 4.

Images