A transient detection method and apparatus

Abstract

The application discloses a transient detection method and device, the method comprises the following steps: obtaining power data of a to-be-detected power distribution network in a preset period, and calculating an operation index according to the power data; when it is determined that the to-be-detected power distribution network is in a transient state according to the change of the operation index, power data of the to-be-detected power distribution network is divided into data in different sections by a variational mode decomposition method; wherein the data in different sections comprises medium-frequency section frequency data and high-frequency section frequency data; when it is determined that the ratio between the signal amplitude of the medium-frequency section frequency data and the original signal amplitude is greater than a preset value, it is determined that the transient state of the to-be-detected power distribution network is caused by distributed photovoltaic fluctuation; when it is determined that the ratio between the signal amplitude of the high-frequency section frequency data and the original signal amplitude is greater than a preset value, it is determined that the transient state of the to-be-detected power distribution network is caused by system failure of the to-be-detected power distribution network. The application embodiment effectively improves the precision of transient detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power grid transient detection technology, and in particular to a transient detection method and device. BACKGROUND

[0002] In order to improve the renewable energy consumption capacity, based on the distribution network directly accessing renewable energy has become a major trend of current power grid development, that is, in the form of distributed microgrid interacting with the existing distribution network.

[0003] The distributed renewable energy access to the distribution network will inevitably lead to new control and safety problems. Specifically, the load, the power grid, and the distributed renewable power source interact with each other, and once any part is abnormal, it will cause the power grid to enter a certain transient state from the existing steady state. How to quickly (millisecond level) identify the transient starting time and identify the cause of the power grid entering the transient state becomes the key. However, the existing detection method mainly detects the transient state based on real-time data with a 3s interval, such as whether the voltage calculated at 3s exceeds a certain fixed limit value, which has low sensitivity; when judging that a transient state occurs, only the waveform and the event are recorded, and usually only the power grid fault is considered to cause the transient state to occur, and the fluctuation of the distributed power source is rarely considered to cause the transient state to change, and the considered factors are not comprehensive enough, thereby resulting in low transient detection accuracy.

[0004] From the above, the existing transient detection method has the problem of low detection accuracy. SUMMARY

[0005] The embodiments of the present application provide a transient detection method and device, which improves the accuracy of transient detection.

[0006] The first aspect of the embodiments of the present application provides a transient detection method, comprising:

[0007] Obtaining power data of a to-be-detected distribution network in a preset period, and calculating an operation index according to the power data;

[0008] When it is determined that the to-be-detected distribution network is in a transient state according to the change of the operation index, the power data of the to-be-detected distribution network is divided into data of different sections by a variational mode decomposition method; wherein the data of different sections include: medium-frequency section frequency data and high-frequency section frequency data;

[0009] When it is determined that the ratio between the signal amplitude of the medium-frequency section frequency data and the original signal amplitude is greater than a preset value, it is determined that the transient state of the to-be-detected distribution network is caused by the fluctuation of the distributed photovoltaic; when it is determined that the ratio between the signal amplitude of the high-frequency section frequency data and the original signal amplitude is greater than a preset value, it is determined that the transient state of the to-be-detected distribution network is caused by the system fault of the to-be-detected distribution network.

[0010] In one possible implementation of the first aspect, the operating indicators are calculated based on power data, specifically as follows:

[0011] Power data includes: current data, power data, and voltage data;

[0012] The operational indicators include: the primary operational indicator and the secondary operational indicator;

[0013] The calculation process for the first operational indicator is as follows:

[0014]

[0015] The calculation process for the second operational indicator is as follows:

[0016]

[0017] Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents the 2-cycle interval, i.e. the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

[0018] In one possible implementation of the first aspect, the distribution network under test is determined to be in a transient state based on changes in operating indicators, specifically as follows:

[0019] When the values ​​of both the first and second operating indicators increase within a preset time period, the distribution network under test is determined to be in a transient state.

[0020] One possible implementation of the first aspect further includes: calculating the transient start time of the distribution network to be detected, specifically:

[0021] The voltage change formula of the distribution network under test during the transient process is obtained by fitting the function, and the fitting coefficient is obtained by solving the voltage change formula by the least squares method.

[0022] The transient start time of the distribution network to be detected is calculated based on the fitting coefficients and the first preset time interval.

[0023] In one possible implementation of the first aspect, the transient start time of the distribution network to be detected is calculated based on the fitting coefficients and a first preset time interval, specifically as follows:

[0024] Based on the fitting coefficients and a first preset time interval, a first formula is constructed. When the first formula satisfies a certain condition, the distribution network under test is in a transient state; when the first formula does not satisfy a certain condition, the distribution network under test is not in a transient state. Specifically, when the first formula satisfies a certain condition:

[0025]

[0026] in, Take 0.1; a n (t) is the fitting coefficient at time t; a n (tT) is the fitting coefficient at time tT; T is the first preset time interval; m is the total number of the first preset time intervals.

[0027] In one possible implementation of the first aspect, the power data of the distribution network to be tested is divided into different segments using variational mode decomposition, specifically:

[0028] The power data of the distribution network to be tested is decomposed by variational mode decomposition according to a first preset time interval to obtain K discrete sub-signals; where K is a positive integer greater than 1.

[0029] Based on K discrete sub-signals, a preset objective function, and preset constraints, data for different segments are calculated.

[0030] In one possible implementation of the first aspect, after determining that the transient state of the power grid under test is caused by a system fault in the distribution network under test, the method further includes:

[0031] The system fault type of the distribution network to be tested is determined as follows:

[0032] Based on the Monte Carlo sampling method, N operating scenarios and their power and voltage data of the power grid to be tested are obtained; where N is a positive integer.

[0033] Cluster the N running scenarios according to the clustering algorithm and generate clustering results;

[0034] The voltage fitting coefficients of the clustering results are calculated using the polynomial fitting method.

[0035] Based on the polynomial fitting method, the actual fault voltage trend is fitted, and the first difference between the voltage fitting coefficient and the library function is obtained. The system fault type of the distribution network to be tested is determined based on the first difference.

[0036] In one possible implementation of the first aspect, the data for different segments also includes: low-frequency segment frequency data.

[0037] A second aspect of this application provides a transient detection device, including: a calculation module, a decomposition module, and a detection module;

[0038] The calculation module is used to acquire power data of the distribution network to be tested for a preset time period and calculate the operating indicators based on the power data.

[0039] The decomposition module is used to divide the power data of the distribution network under test into different segments by using variational mode decomposition method when the network is determined to be in a transient state based on changes in operating indicators. The different segments include: mid-frequency segment data and high-frequency segment data.

[0040] The detection module is used to determine that the transient of the distribution network under test is caused by distributed photovoltaic fluctuations when the ratio between the signal amplitude of the intermediate frequency band data and the original signal amplitude is greater than a preset value; and to determine that the transient of the distribution network under test is caused by a system fault in the distribution network under test when the ratio between the signal amplitude of the high frequency band data and the original signal amplitude is greater than a preset value.

[0041] In one possible implementation of the second aspect, operating indicators are calculated based on power data, specifically as follows:

[0042] Power data includes: current data, power data, and voltage data;

[0043] The operational indicators include: the primary operational indicator and the secondary operational indicator;

[0044] The calculation process for the first operational indicator is as follows:

[0045]

[0046] The calculation process for the second operational indicator is as follows:

[0047]

[0048] Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents the 2-cycle interval, i.e. the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

[0049] Compared to existing technologies, this invention provides a transient detection method and apparatus. The method includes: acquiring power data of a distribution network under test for a preset time period; calculating operating indicators based on the power data; when the distribution network under test is determined to be in a transient state based on changes in the operating indicators, dividing the power data of the distribution network under test into different segments using variational mode decomposition; wherein the different segments include: mid-frequency segment frequency data and high-frequency segment frequency data; when the ratio between the signal amplitude of the mid-frequency segment frequency data and the original signal amplitude is determined to be greater than a preset value, determining that the transient state of the distribution network under test is caused by distributed photovoltaic fluctuations; when the ratio between the signal amplitude of the high-frequency segment frequency data and the original signal amplitude is determined to be greater than a preset value, determining that the transient state of the distribution network under test is caused by a system fault in the distribution network under test.

[0050] The beneficial effects are as follows: In this embodiment of the invention, when the distribution network under test is determined to be in a transient state based on changes in its operating indicators, the power data of the distribution network under test is divided into mid-frequency band data and high-frequency band data using variational mode decomposition. When the ratio between the signal amplitude of the mid-frequency band data and the original signal amplitude is greater than a preset value, the transient state of the distribution network under test is determined to be caused by distributed photovoltaic fluctuations. When the ratio between the signal amplitude of the high-frequency band data and the original signal amplitude is greater than a preset value, the transient state of the distribution network under test is determined to be caused by a system fault in the distribution network under test. In the transient detection process of the power grid, this embodiment of the invention divides the power data of the distribution network under test into mid-frequency band data and high-frequency band data, and further determines the cause of the transient state as distributed photovoltaic fluctuations or a system fault in the distribution network under test based on the ratio between the signal amplitude of the mid-frequency band data and the high-frequency band data and the original signal amplitude. This comprehensively considers the factors contributing to the transient state and effectively improves the accuracy of transient detection. Attached Figure Description

[0051] Figure 1 This is a schematic flowchart of a transient detection method provided in an embodiment of the present invention;

[0052] Figure 2 This is a power grid transient voltage change trend diagram provided in an embodiment of the present invention;

[0053] Figure 3 This is a mid-frequency power trend diagram provided in an embodiment of the present invention;

[0054] Figure 4 This is a schematic diagram of a transient detection device provided in an embodiment of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Reference Figure 1 , Figure 1 This is a flowchart illustrating a transient detection method according to an embodiment of the present invention, including steps S101-S103:

[0057] S101: Obtain the power data of the distribution network to be tested for a preset time period, and calculate the operating indicators based on the power data.

[0058] In this embodiment, the calculation of operating indicators based on the power data specifically includes:

[0059] The power data includes: current data, power data, and voltage data;

[0060] The operational indicators include: a first operational indicator and a second operational indicator;

[0061] The calculation process for the first operational indicator is as follows:

[0062]

[0063] The calculation process for the second operating indicator is as follows:

[0064]

[0065] Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents a 2-cycle interval, i.e., the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

[0066] S102: When the distribution network under test is determined to be in a transient state based on the changes in operating indicators, the power data of the distribution network under test is divided into data of different segments by variational mode decomposition method.

[0067] The data for different segments include: low-frequency segment frequency data, mid-frequency segment frequency data, and high-frequency segment frequency data.

[0068] In this embodiment, the step of dividing the power data of the distribution network under test into different segments using variational mode decomposition specifically involves:

[0069] The power data of the distribution network under test is decomposed by variational mode decomposition according to the first preset time interval to obtain K discrete sub-signals; where K is a positive integer greater than 1.

[0070] Based on the K discrete sub-signals, the preset objective function, and the preset constraints, the data for the different segments are calculated.

[0071] In this embodiment, determining that the distribution network under test is in a transient state based on changes in the operating indicators specifically involves:

[0072] When the values ​​of both the first operating indicator and the second operating indicator increase within the preset time period, the distribution network under test is determined to be in a transient state.

[0073] In one specific embodiment, the method further includes: calculating the transient start time of the distribution network to be detected, specifically:

[0074] The voltage change formula of the distribution network under test during the transient process is obtained by fitting the function, and the fitting coefficient is obtained by solving the voltage change formula by the least squares method.

[0075] The transient start time of the distribution network to be detected is calculated based on the fitting coefficient and the first preset time interval.

[0076] In one specific embodiment, the step of calculating the transient start time of the distribution network to be detected based on the fitting coefficient and the first preset time interval specifically involves:

[0077] Based on the fitting coefficients and the first preset time interval, a first formula is constructed. When the first formula satisfies a condition, the distribution network under test is in a transient state; when the first formula does not satisfy a condition, the distribution network under test is not in a transient state. Specifically, the condition that the first formula satisfies is:

[0078]

[0079] in, Take 0.1; a n (t) is the fitting coefficient at time t; a n (tT) is the fitting coefficient at time tT; T is the first preset time interval; m is the total number of the first preset time intervals.

[0080] S103: When the ratio between the signal amplitude of the intermediate frequency band data and the original signal amplitude is greater than a preset value, the transient state of the distribution network under test is determined to be caused by distributed photovoltaic fluctuations; when the ratio between the signal amplitude of the high frequency band data and the original signal amplitude is greater than a preset value, the transient state of the distribution network under test is determined to be caused by a system fault in the distribution network under test.

[0081] In this embodiment, after determining that the transient state of the power grid under test is caused by a system fault in the distribution network under test, the method further includes:

[0082] The system fault type of the distribution network to be tested is determined as follows:

[0083] Based on the Monte Carlo sampling method, N operating scenarios and their power and voltage data of the power grid to be tested are obtained; where N is a positive integer;

[0084] The N running scenarios are clustered according to the clustering algorithm to generate clustering results;

[0085] The voltage fitting coefficients of the clustering results are calculated using the polynomial fitting method.

[0086] According to the polynomial fitting method, the actual fault voltage trend is fitted, and the first difference between the voltage fitting coefficient and the library function is obtained. The system fault type of the distribution network to be detected is confirmed based on the first difference.

[0087] In a preferred embodiment, the specific process of the transient detection method is as follows:

[0088] Step 1: When the distribution network is in steady state, the various indicators of the power grid are in a smooth and slow state of change, so the correlation between the indicators is not obvious (e.g., voltage U(t), current I(t), active power P(t)). When the distribution is subjected to shocks such as distributed power sources, loads or power grid faults, the power grid operation indicators change drastically, so the correlation between these indicators will be more obvious at this time.

[0089] In this embodiment of the invention, the current and power indicators (i.e., the first operating indicator) corr1(P,I,t) and the voltage and power indicators (i.e., the second operating indicator) corr2(P,U,t) are calculated and generated with a 2-cycle interval (when the power grid frequency is 50Hz, one cycle time is 0.02s, and two cycles are 0.04s). This is because the time interval of 2 cycles is shorter than that of the prior art, which can achieve higher recognition accuracy.

[0090] Specifically, corr1(P,I,t) is shown in equation (1), and corr2(P,U,t) is shown in equation (2). t is time, t0 is the start time within the 0.04s interval. It should be noted that other indicators can also be selected to calculate the operational correlation, such as frequency, harmonics, etc., and k is a positive integer.

[0091]

[0092]

[0093] When corr1(P,I,t) and corr2(P,U,t) both increase compared to corr1(P,I,t-2k△t) and corr2(P,U,t-2k△t) in the previous interval 2k△t (i.e., when the values ​​of the first and second operating indicators both increase within the preset time period 2k△t), the correlation increases, satisfying equations (3) and (4), where τ1 and τ2 are both positive numbers. To avoid the influence of steady-state disturbances, τ1 and τ2 are both taken as 0.05, indicating that the power grid state has changed and is in a transient state.

[0094] Corr1(P,I,t)-Corr1(P,I,t-2kΔt)>τ1; (3)

[0095] Corr2(P,U,t)-Corr2(P,U,t-2kΔt)>τ2; (4)

[0096] Step 2: Once the conditions of Step 1 are met, a more refined assessment of transient change characteristics is needed. From the perspective of a single indicator, when the power grid transitions from a steady state to a transient state, its operating indicators show significant discontinuity, i.e., a change in trend. Due to the existence of system damping, multiple trend changes may occur during the transient process, such as... Figure 2 As shown, Figure 2 This is a power grid transient voltage change trend diagram provided in an embodiment of the present invention. Figure 2 We can obtain, Figure 2 This is an example of a power grid recovering to a steady state after a transient change. When the power grid changes from a steady state to a transient state, the voltage index shows obvious fluctuations. After the transient change, the voltage first drops and then rises. However, due to the existence of power grid damping, multiple trend changes also occur during the transient process.

[0097] To further confirm that the system has indeed entered a transient state (start time), taking voltage as an example, the voltage change ΔU(t) between adjacent first preset time intervals is calculated using equation (5):

[0098] ΔU(t)=U(t)-U(tT); (5)

[0099] According to power quality standards, T is 10 cycles, or 0.2 seconds.

[0100] To effectively determine the transient trend changes in the power grid, a fitting function (which can be a polynomial fitting, trigonometric function, exponential function, etc.) is used. Equation (6) is the formula for obtaining the transient process voltage change trend through polynomial fitting, where a n Here are the polynomial coefficients, and n is the fitting order. The solution method is the least squares method.

[0101]

[0102] Obtain the fitting coefficient a for each first preset time interval T n (t)={a0(t), a1(t)...a m (t)}, compare with a n (t) and a n The variation range of (tT) further divides the transient process into multiple sub-segments, as shown in Equation (7) (i.e., the first formula). That is, the cumulative change of the fitting coefficient exceeds in Take 0.1. If the first formula (i.e., formula 7) satisfies the condition, the distribution network under test is in a transient state; if the first formula does not satisfy the condition, the transient state of the distribution network under test ends.

[0103]

[0104] Step 3: Mode decomposition, classifying the causes of changes in the power grid state into distributed output fluctuations, load fluctuations, and grid faults. To effectively identify the cause, variational mode decomposition (VMD) is used at power intervals of T. VMD decomposes any x(t) into K discrete sub-signals u. k By using the preset objective function shown in equation (8) and the preset constraints shown in equation (9), the stability of each discrete sub-signal u is guaranteed. k Compactly distributed at the corresponding center frequency w k Nearby. In equation (8) δ(t) represents the gradient operation, δ(t) represents the Dirac distribution, and * represents the convolution operation.

[0105]

[0106]

[0107] To distinguish between power grid state changes caused by three factors—distributed output fluctuations, load fluctuations, and grid faults—K is set to 3, corresponding to w respectively. k The signals are categorized into low-frequency, mid-frequency, and high-frequency. Low-frequency refers to signals below 40Hz, mid-frequency to 160Hz-1280Hz, and high-frequency to above 1280Hz. VMD decomposes the signals into three segments: low-frequency, low-to-mid-frequency, and mid-to-high-frequency. Low-frequency signals are determined based on grid load safety, such as a rated power change rate not exceeding 5%. Mid-frequency signals are determined according to distributed energy output fluctuation requirements, such as photovoltaic output fluctuations. High-frequency signals correspond to fluctuations when a short-circuit fault occurs in the grid. The mid-frequency signal obtained from VMD decomposition is as follows: Figure 3 As shown, Figure 3 This is a mid-frequency power trend diagram provided in an embodiment of the present invention. Figure 3 In the diagram, the vertical axis represents power data, and the horizontal axis represents time data.

[0108] Furthermore, when the ratio between the signal amplitude of the intermediate frequency band data and the original signal amplitude is greater than a preset value, it is considered that the intermediate frequency signal is significant while the high frequency signal is close to 0. This indicates that distributed photovoltaic (PV) fluctuations have caused a change in the grid state (i.e., the transient state of the distribution network under test is determined to be caused by PV fluctuations), and battery energy storage is considered to mitigate the power fluctuations caused by PV. In other words, the power used by the intermediate frequency signal decomposed by VMD should be compensated for by battery energy storage for PBS.

[0109] Similarly, when the ratio between the signal amplitude of the high-frequency band data and the original signal amplitude is greater than a preset value, it is considered that the high-frequency signal is significant while the intermediate-frequency signal is close to 0. This indicates that a fault in the distribution network system has caused a change in the power grid state (i.e., the transient state of the power grid under test is determined to be caused by a system fault in the distribution network under test). To further determine the type of power grid fault, fault location needs to be achieved based on the following methods:

[0110] (1) Based on the Monte Carlo sampling method, power and voltage data under N operating scenarios in the distributed distribution network (distribution network under test) are generated, as shown in Equation (10). The power data generated by single-phase grounding fault, two-phase short-circuit fault and three-phase short-circuit fault in the distributed distribution network under N operating scenarios are P W The voltage data generated by single-phase grounding faults, two-phase short-circuit faults, and three-phase short-circuit faults in a distributed distribution network under N operating scenarios is U. W , corresponding to w being 0, 1, and 2 respectively.

[0111]

[0112]

[0113] In the above formula Expanded into a time series Similarly, Expanded into a time series

[0114] (2) For scenarios where step (1) is reduced based on clustering algorithms, the kNN algorithm is specifically adopted, with P W As a clustering object, a running scenario is randomly selected. As a cluster center.

[0115] First, the Euclidean distance between other scenarios and the cluster center is calculated using equation (11).

[0116]

[0117] Secondly, when equation (12) is satisfied, They are considered to be of the same type. У can be set according to the rated power ratio, such as 0.1, 0.15... times the rated power.

[0118]

[0119] (3) Based on the polynomial fitting method in equation (6), the voltage fitting coefficients a under different scenarios are obtained respectively. lib n (t), and store it.

[0120] (4) Based on the polynomial fitting method in equation (6), fit the actual fault voltage trend, and obtain the first difference between the voltage fitting coefficient and the library function. The first difference Δcoffect i The calculation process is shown in equation (13):

[0121]

[0122] Choose the smallest first difference Δcoffect i As the corresponding fault type. Specifically: the actual obtained a real n (t) are respectively compared with the voltage fitting coefficients a under different scenarios. lib n (t) performs difference calculations, and then takes the minimum value among these differences, that is, the smallest difference corresponds to the corresponding scenario.

[0123] To further explain the transient detection device, please refer to... Figure 4 , Figure 4 This is a schematic diagram of a transient detection device according to an embodiment of the present invention, including: a calculation module 401, a decomposition module 402 and a detection module 403;

[0124] The calculation module 401 is used to acquire power data of the distribution network to be tested during a preset time period, and calculate the operating indicators based on the power data.

[0125] The decomposition module 402 is used to divide the power data of the distribution network under test into different segments by using variational mode decomposition when it is determined that the distribution network under test is in a transient state based on the changes in the operating indicators; wherein, the data of the different segments include: mid-frequency segment frequency data and high-frequency segment frequency data;

[0126] The detection module 403 is used to determine that the transient state of the distribution network under test is caused by distributed photovoltaic fluctuations when the ratio between the signal amplitude of the mid-frequency band data and the original signal amplitude is greater than a preset value; and to determine that the transient state of the distribution network under test is caused by a system fault of the distribution network under test when the ratio between the signal amplitude of the high-frequency band data and the original signal amplitude is greater than the preset value.

[0127] In this embodiment, the calculation of operating indicators based on the power data specifically includes:

[0128] The power data includes: current data, power data, and voltage data;

[0129] The operational indicators include: a first operational indicator and a second operational indicator;

[0130] The calculation process for the first operational indicator is as follows:

[0131]

[0132] The calculation process for the second operating indicator is as follows:

[0133]

[0134] Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents a 2-cycle interval, i.e., the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

[0135] In this embodiment, determining that the distribution network under test is in a transient state based on changes in the operating indicators specifically involves:

[0136] When the values ​​of both the first operating indicator and the second operating indicator increase within the preset time period, the distribution network under test is determined to be in a transient state.

[0137] In one specific embodiment, the method further includes: calculating the transient start time of the distribution network to be detected, specifically:

[0138] The voltage change formula of the distribution network under test during the transient process is obtained by fitting the function, and the fitting coefficient is obtained by solving the voltage change formula by the least squares method.

[0139] The transient start time of the distribution network to be detected is calculated based on the fitting coefficient and the first preset time interval.

[0140] In one specific embodiment, the step of calculating the transient start time of the distribution network to be detected based on the fitting coefficient and the first preset time interval specifically involves:

[0141] Based on the fitting coefficients and the first preset time interval, a first formula is constructed. When the first formula satisfies a condition, the distribution network under test is in a transient state; when the first formula does not satisfy a condition, the distribution network under test is not in a transient state. Specifically, the condition that the first formula satisfies is:

[0142]

[0143] in, Take 0.1; a n (t) is the fitting coefficient at time t; a n (tT) is the fitting coefficient at time tT; T is the first preset time interval; m is the total number of the first preset time intervals.

[0144] In one specific embodiment, the step of dividing the power data of the distribution network under test into different segments using variational mode decomposition specifically involves:

[0145] The power data of the distribution network under test is decomposed by variational mode decomposition according to the first preset time interval to obtain K discrete sub-signals; where K is a positive integer greater than 1.

[0146] Based on the K discrete sub-signals, the preset objective function, and the preset constraints, the data for the different segments are calculated.

[0147] In this embodiment, after determining that the transient state of the power grid under test is caused by a system fault in the distribution network under test, the method further includes:

[0148] The system fault type of the distribution network to be tested is determined as follows:

[0149] Based on the Monte Carlo sampling method, N operating scenarios and their power and voltage data of the power grid to be tested are obtained; where N is a positive integer;

[0150] The N running scenarios are clustered according to the clustering algorithm to generate clustering results;

[0151] The voltage fitting coefficients of the clustering results are calculated using the polynomial fitting method.

[0152] According to the polynomial fitting method, the actual fault voltage trend is fitted, and the first difference between the voltage fitting coefficient and the library function is obtained. The system fault type of the distribution network to be detected is confirmed based on the first difference.

[0153] In one specific embodiment, the data for the different segments further includes: low-frequency segment frequency data.

[0154] In this embodiment of the invention, a calculation module acquires power data of the distribution network under test for a preset time period, and calculates operating indicators based on the power data. A decomposition module, when determining that the distribution network under test is in a transient state based on changes in the operating indicators, divides the power data of the distribution network under test into different segments using variational mode decomposition. These different segments include: mid-frequency segment data and high-frequency segment data. A detection module, when determining that the ratio between the signal amplitude of the mid-frequency segment data and the original signal amplitude is greater than a preset value, determines that the transient state of the distribution network under test is caused by distributed photovoltaic fluctuations. Similarly, when determining that the ratio between the signal amplitude of the high-frequency segment data and the original signal amplitude is greater than a preset value, determines that the transient state of the distribution network under test is caused by a system fault in the distribution network under test.

[0155] This invention, in its embodiments, determines that the distribution network under test is in a transient state based on changes in its operating indicators. It uses variational mode decomposition to divide the power data of the distribution network under test into mid-frequency and high-frequency data. When the ratio between the signal amplitude of the mid-frequency data and the original signal amplitude is greater than a preset value, the transient state is determined to be caused by distributed photovoltaic fluctuations. When the ratio between the signal amplitude of the high-frequency data and the original signal amplitude is greater than a preset value, the transient state is determined to be caused by a system fault in the distribution network. In the transient state detection process, this invention divides the power data of the distribution network under test into mid-frequency and high-frequency data. Based on the ratio between the signal amplitude of the mid-frequency / high-frequency data and the original signal amplitude, it further determines whether the transient state is caused by distributed photovoltaic fluctuations or a system fault in the distribution network under test. This comprehensive consideration of transient factors effectively improves the accuracy of transient state detection.

[0156] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A transient detection method, characterized in that, include: Obtain power data of the distribution network to be tested within a preset time period, and calculate operating indicators based on the power data; When the distribution network under test is determined to be in a transient state based on the changes in the operating indicators, the power data of the distribution network under test is divided into different segments using the variational mode decomposition method; wherein, the different segments include: mid-frequency segment frequency data and high-frequency segment frequency data; When the ratio between the signal amplitude of the mid-frequency band data and the original signal amplitude is determined to be greater than a preset value, the transient state of the distribution network under test is determined to be caused by distributed photovoltaic fluctuations; when the ratio between the signal amplitude of the high-frequency band data and the original signal amplitude is determined to be greater than the preset value, the transient state of the distribution network under test is determined to be caused by a system fault in the distribution network under test.

2. The transient detection method according to claim 1, characterized in that, The operational indicators calculated based on the power data are specifically as follows: The power data includes: current data, power data, and voltage data; The operational indicators include: a first operational indicator and a second operational indicator; The calculation process for the first operational indicator is as follows: ； The calculation process for the second operating indicator is as follows: ； Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents a 2-cycle interval, i.e., the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

3. The transient detection method according to claim 2, characterized in that, The step of determining that the distribution network under test is in a transient state based on changes in the operating indicators specifically involves: When the values ​​of both the first operating indicator and the second operating indicator increase within the preset time period, the distribution network under test is determined to be in a transient state.

4. The transient detection method according to claim 3, characterized in that, Also includes: The transient start time of the distribution network under test is calculated as follows: The voltage change formula of the distribution network under test during the transient process is obtained by fitting the function, and the fitting coefficient is obtained by solving the voltage change formula by the least squares method. The transient start time of the distribution network to be detected is calculated based on the fitting coefficient and the first preset time interval.

5. The transient detection method according to claim 4, characterized in that, The transient start time of the distribution network to be detected is calculated based on the fitting coefficient and the first preset time interval, specifically as follows: Based on the fitting coefficients and the first preset time interval, a first formula is constructed. When the first formula satisfies a condition, the distribution network under test is in a transient state; when the first formula does not satisfy a condition, the distribution network under test is not in a transient state. Specifically, the condition that the first formula satisfies is: ； in, Take 0.1; a n (t) is the fitting coefficient at time t; a n (tT) is the fitting coefficient at time tT; T is the first preset time interval; m is the total number of the first preset time intervals.

6. The transient detection method according to claim 4, characterized in that, The process of dividing the power data of the distribution network under test into different segments using variational mode decomposition is as follows: The power data of the distribution network under test is decomposed by variational mode decomposition according to the first preset time interval to obtain K discrete sub-signals; where K is a positive integer greater than 1. Based on the K discrete sub-signals, the preset objective function, and the preset constraints, the data for the different segments are calculated.

7. The transient detection method according to claim 1, characterized in that, After determining that the transient state of the power grid under test is caused by a system fault in the distribution network under test, the method further includes: The system fault type of the distribution network to be tested is determined as follows: Based on the Monte Carlo sampling method, N operating scenarios and their power and voltage data of the power grid to be tested are obtained; where N is a positive integer; The N running scenarios are clustered according to the clustering algorithm to generate clustering results; The voltage fitting coefficients of the clustering results are calculated using the polynomial fitting method. According to the polynomial fitting method, the actual fault voltage trend is fitted, and the first difference between the voltage fitting coefficient and the library function is obtained. The system fault type of the distribution network to be detected is confirmed based on the first difference.

8. The transient detection method according to claim 1, characterized in that, The data for the different segments also includes: frequency data for the low-frequency segment.

9. A transient detection device, characterized in that, include: Calculation module, decomposition module, and detection module; The calculation module is used to acquire power data of the distribution network to be tested for a preset time period, and calculate the operating indicators based on the power data. The decomposition module is used to divide the power data of the distribution network under test into different segments by using variational mode decomposition when the distribution network under test is determined to be in a transient state based on the changes in the operating indicators; wherein, the data of the different segments include: mid-frequency segment frequency data and high-frequency segment frequency data; The detection module is used to determine that the transient state of the distribution network under test is caused by distributed photovoltaic fluctuations when the ratio between the signal amplitude of the mid-frequency band data and the original signal amplitude is greater than a preset value; and to determine that the transient state of the distribution network under test is caused by a system fault of the distribution network under test when the ratio between the signal amplitude of the high-frequency band data and the original signal amplitude is greater than the preset value.

10. A transient detection device according to claim 9, characterized in that, The operational indicators calculated based on the power data are specifically as follows: The power data includes: current data, power data, and voltage data; The operational indicators include: a first operational indicator and a second operational indicator; The calculation process for the first operational indicator is as follows: ； The calculation process for the second operating indicator is as follows: ； Wherein, corr1(P,I,t) is the first operating indicator; corr2(P,U,t) is the second operating indicator; 2kΔt represents a 2-cycle interval, i.e., the preset time period; P(t) is the power data; I(t) is the current data; and U(t) is the voltage data.

Images