Method and system for detecting phenomena in a power network

Abstract

Method for detecting a phenomenon in an electric power network, the electric power network comprising at least one protection zone having at least two sub-zones. The method comprises: measuring in a sub-zone of the protection zone at least one current signal actually in the respective sub-zone; determining a linear dependency between the at least one measured current signal actually in at least one sub-zone of the protection zone and the corresponding at least one actual measured current signal actually in at least one other sub-zone of the protection zone; and detecting a phenomenon in the protection zone of the electric power network based on the at least one determined linear dependency. Furthermore, a system for detecting a phenomenon in an electric power network.

Description

Technical Field

[0001] This invention relates to methods and systems for detecting phenomena in power networks. Background Technology

[0002] A power grid is an interconnected system of currents, in which a phenomenon or event (such as a fault) occurring at a point in the grid is at least partially visible at all measurement points in the interconnected grid, and thus also affects the main operating conditions, i.e., the state, of the power grid.

[0003] Typically, power networks have two main operating conditions or states: a normal operating state, in which power is supplied to loads connected to the network and there are no specific anomalies in the operation of the network; and an abnormal operating state, in which there is at least one phenomenon or event that deviates from the normal operation of the network, such as a fault occurring at some point in the current interconnection network, and this phenomenon or event may have some impact on the operation of the network and / or on the power supply to at least one load connected to the network.

[0004] Currently, fault detection and management are based on single-point measurement, where a specific feeder electrical quantity that dominates in different feeders or racks of the power network is measured at a specific point in each feeder. Accordingly, fault detection and management are performed by feeder-specific equipment, often referred to as protection and control relays.

[0005] Feeder-specific methods for fault detection (especially in large networks with several input and output feeders) result in complex systems with numerous devices whose operation must be precisely tuned, particularly for fault management, to respond only to faults occurring in specific protected areas or zones designated for that particular device. Different relays create different events, and these events often require time-consuming analysis to determine the actual original event or the root cause. Furthermore, events for which no corresponding dedicated detection application is available or used are missed. Therefore, it is often difficult even to identify some events that have occurred, or at least to quickly determine what has happened. Summary of the Invention

[0006] The purpose of this invention is to provide a novel method and system for detecting phenomena in power networks.

[0007] The invention is characterized by the features of the independent claims.

[0008] The present invention is based on the concept of determining the linear dependence between current signals in different sub-regions of at least one protection zone in a power network, and detecting phenomena in the power network based on the determined linear dependence.

[0009] This invention provides a simple application for detecting events or phenomena occurring at a point in a power network.

[0010] Some embodiments of the present invention are disclosed in the dependent claims. Attached Figure Description

[0011] The present invention will now be described in more detail with reference to the accompanying drawings and preferred embodiments, in which...

[0012] Figure 1 This is a schematic diagram of a substation in a power grid;

[0013] Figure 2a , Figure 2b , Figure 2c and Figure 2d The schematic disclosure of the connection to Figure 1 The correlation coefficient is determined by the actual current signal in the phase of the feeder in the substation.

[0014] Figure 3 An implementation of a solution for detecting phenomena in a power grid is schematically disclosed.

[0015] Figure 4a and Figure 4b Exemplary sample sequences for detecting phenomena in power networks are schematically disclosed;

[0016] Figure 5 This is a schematic diagram illustrating the correlation coefficients determined from actual measurements of residual current in different feeders within a substation in the power network during a ground fault in one feeder; and

[0017] Figure 6 This is a schematic diagram showing the correlation coefficient determined by actual measured phase currents in a phase of a different feeder in a substation within the power network during a short-circuit fault in one feeder.

[0018] For clarity, the accompanying drawings illustrate some embodiments of the invention in a simplified manner. In the drawings, the same reference numerals denote the same elements. Detailed Implementation

[0019] Figure 1 This is a schematic diagram of a power station ES or substation ES, shown schematically with a dashed box. The power station ES forms part of the power network EPN, which is shown schematically with a solid box.

[0020] The power station ES includes or is equipped with an input feeder F1, namely the first feeder F1, which supplies power to the power station ES. Figure 1Three output feeders, F2, F3, and F4, also were disclosed, namely the second feeder F2, the third feeder F3, and the fourth feeder F4, which supply power forward from the power station ES. However, the number of input and output feeders is not limited to this. Figure 1 The number shown. Depending on the power network configuration and the possible distributed generation within the network, some feeders can be configured to operate as input or output feeders at different times under different power network configurations. A power plant ES with at least one input feeder and at least one output feeder provides one type of node in the power network.

[0021] For clarity, in Figure 1 In this diagram, only one line (i.e., the line indicated by the reference numerals F1, F2, F3, or F4) is used to indicate all phase lines and possible neutral conductors of the corresponding feeder. The actual number of phase lines and the availability of the neutral conductor depend on local power supply practices. In the following, it is assumed that each feeder F1, F2, F3, F4 includes three phase lines, namely the first phase L1, the second phase L2, and the third phase L3, and the neutral conductor N.

[0022] In the following text, for example, it may be assumed that input feeder F1 represents a high-voltage transmission line, such as a 110kV transmission line, and output feeders F2, F3, and F4 represent medium-voltage distribution lines, such as 20kV distribution lines, and substation ES represents a main substation. However, the applicability of the solution disclosed herein is not limited to this power network configuration, but can be used in any power network configuration including appropriate nodes with an appropriate number of input and output feeders.

[0023] Figure 1 A Central Intelligent Electronic Device (CIED) is also disclosed, which can be configured to receive measurement or operational information obtained from devices connected to a power grid (EPN), process the received information by running various software-based applications using the received information as input, and / or further transmit the received information. The CIED can also be configured to control devices connected to the EPN based on the processed information or information received from higher-level control applications. For this purpose, the CIED may include a control output (CO) for controlling the operation of other devices connected to the EPN. Figure 1 In the example, the Central Intelligent Electronic Device (CIED) is located at the power station ES, and the CIED can be, for example, a computer, such as a substation computer equipped with high computing power. However, other locations and implementations of the CIED are also possible.

[0024] Each of feeders F1 to F4 includes a corresponding measuring instrument for measuring the actual electrical quantity in the respective feeder. Therefore, the first feeder F1 includes a first measuring instrument MI1 for measuring the actual electrical quantity in the first feeder F1, the second feeder F2 includes a second measuring instrument MI2 for measuring the actual electrical quantity in the second feeder F2, the third feeder F3 includes a third measuring instrument MI3 for measuring the actual electrical quantity in the third feeder F3, and the fourth feeder F4 includes a fourth measuring instrument MI4 for measuring the actual electrical quantity in the fourth feeder F4.

[0025] Typically, the electrical quantities of the power network EPN of interest can be zero voltage (residual voltage), zero current (residual current), at least one phase voltage and / or at least one phase current. Therefore, measuring instruments MI1 to MI4 can include means for measuring both the required voltage and current information. In the solution presented herein, as disclosed in more detail below, only the phase current and zero current (i.e., residual current) are the electrical quantities of interest. Therefore, measuring instruments MI1 to MI4 can include only instrument transformers for measuring only the necessary current information. Thus, the first measuring instrument MI1 can be configured to measure the phase currents IL1(F1), IL2(F1), and IL3(F1) in phases L1, L2, and L3 of the first feeder F1, and the residual current IN(F1) in the neutral conductor N. The second measuring instrument MI2 can be configured to measure the phase currents IL1(F2), IL2(F2), and IL3(F2) in phases L1, L2, and L3 of the second feeder F2. The third measuring instrument MI3 can be configured to measure the phase currents IL1(F3), IL2(F3), and IL3(F3) in phases L1, L2, and L3 of the third feeder F3, as well as the residual current IN(F3) in the neutral conductor N. The fourth measuring instrument MI4 can be configured to measure the phase currents IL1(F4), IL2(F4), and IL3(F4) in phases L1, L2, and L3 of the fourth feeder F4, as well as the residual current IN(F4) in the neutral conductor N.

[0026] Each measuring instrument MI1 to MI4 is connected to the central intelligent electronic device (CIED) via corresponding data transmission lines DL1 to DL4 to transmit the measurement information obtained by the respective measuring instruments MI1 to MI4 to the CIED. More precisely, the measurement information obtained by the first measuring instrument MI1 is transmitted to the CIED via the first data transmission line DL1, the measurement information obtained by the second measuring instrument MI2 is transmitted to the CIED via the second data transmission line DL2, the measurement information obtained by the third measuring instrument MI3 is transmitted to the CIED via the third data transmission line DL3, and the measurement information obtained by the fourth measuring instrument MI4 is transmitted to the CIED via the fourth data transmission line DL4. The actual implementation of data transmission lines DL1 to DL4 can be wired and / or wireless.

[0027] The measurement information obtained by measuring instruments MI1 to MI4 is typically represented by one or more sample sequences of the corresponding electrical quantity to be measured, which, as mentioned above, are the actual phase current and residual current in the corresponding feeder. To provide samples of the electrical quantity of interest, measuring instruments MI1 to MI4 are equipped with corresponding sampling circuits configured to acquire samples of the electrical quantity of interest. The general construction and operation of different sampling circuits are well known to those skilled in the art and therefore will not be discussed in more detail here. To transmit the acquired samples or sample sequences to the central intelligent electronic device (CIED), measuring instruments MI1 to MI4 are equipped with corresponding transmitters. The general construction and operation of different transmitters are well known to those skilled in the art and therefore will not be discussed in more detail here. In cases where the measuring instruments do not include sampling circuits or transmitters, the measuring instruments can be connected to feeder-specific intelligent electronic devices, such as data collectors, that provide the corresponding functionality.

[0028] Feeders F1 to F4 are also equipped with corresponding circuit breakers. By operating the circuit breaker, the corresponding feeder can be disconnected from the current connection with other feeders, or reconnected to the current connection with other feeders. More precisely, the first circuit breaker CB1 can be operated to disconnect the current connection of the first feeder F1 from the other feeders F2, F3, and F4, or to reconnect the first feeder F1 back to the other feeders F2, F3, and F4; the second circuit breaker CB2 can be operated to disconnect the current connection of the second feeder F2 from the other feeders F1, F3, and F4, or to reconnect the second feeder F2 back to the other feeders F1, F3, and F4; the third circuit breaker CB3 can be operated to disconnect the current connection of the third feeder F3 from the other feeders F1, F2, and F4, or to reconnect the third feeder F3 back to the other feeders F1, F2, and F4; and the fourth circuit breaker CB4 can be operated to disconnect the current connection of the fourth feeder F4 from the other feeders F1, F2, and F3, or to reconnect the fourth feeder F4 back to the other feeders F1, F2, and F3. The circuit breaker CB1 in the input feeder F1 can be located on the low-voltage side or the high-voltage side of the transformer (not shown) at the power station ES.

[0029] By disconnecting any one of circuit breakers CB1, CB2, CB3, and CB4, power supply from or to the corresponding feeder can be blocked. By closing any one of circuit breakers CB1, CB2, CB3, and CB4, power supply from or to the corresponding feeder can be restored. Each of circuit breakers CB1, CB2, CB3, and CB4 therefore provides a coupling device configured in its closed state to connect power to / from the corresponding feeder and configured in its open state to disconnect power to / from the corresponding feeder. Circuit breakers CB1, CB2, CB3, and CB4 may include a single coupling device common to all phase lines L1, L2, and L3, or phase-specific coupling devices that allow phase-specific disconnection and closure. Circuit breakers CB1, CB2, CB3, and CB4 include respective control connections CO-CB1, CO-CB2, CO-CB3, and CO-CB4 for receiving control signals that control the operation of circuit breakers CB1, CB2, CB3, and CB4. For example, due to a fault in the corresponding feeder, there may be a need to disconnect the power supply to / from the corresponding feeder. This, for example, can protect equipment or components in both the faulty feeder and other parts of the interconnected power network from harmful overvoltages or overcurrents. After the fault has been cleared or eliminated, power can be restored by closing the corresponding circuit breaker.

[0030] exist Figure 1In the example, the power station ES provides or forms a protection zone in the power network EPN, and each feeder F1, F2, F3, F4 provides or forms a sub-zone of the protection zone. Figure 1 The protection zone in the example therefore comprises a total of four sub-zones interconnected by current. The number of sub-zones within the protection zone can change if at least one feeder is removed from the service, or if at least one feeder is disconnected from the substation ES due to a change in network configuration, or if at least one additional feeder is connected to the substation ES. In addition to the sub-zones formed by the feeders, the protection zone may include additional sub-zones, such as sub-zones for one or more transformers or buses at the substation ES.

[0031] Figure 1 The power plant ES and the above disclosure are merely examples where the solutions for detecting phenomena or events in power networks, disclosed below, can be utilized. In reality, countless different network configurations or topologies exist where the solutions disclosed herein are applicable. Detailed examples of the solutions are explained below, particularly considering phase line L1; however, similar considerations apply to other phase lines L2, L3, and the neutral conductor N.

[0032] Typically, a power grid interconnection network (EPN) has two main operating conditions or states. The first main operating condition or state is the normal operating state, where power is supplied to loads connected to the network without any specific anomalies in network operation. The second main operating condition or state is the abnormal operating state, where at least one phenomenon or event deviates from the normal operation of the network, such as a fault occurring at some point in the current interconnection network, and this phenomenon or event can have some impact on the operation of the network and / or on the power supply to at least one load connected to the network. In particular, phenomena or events that require certain control operations to be performed in the power network can be considered to constitute phenomena or events of concern in the network. Besides faults occurring in the network, such phenomena or events of concern can also be the connection / disconnection of high-power loads to / from the network.

[0033] In the solution disclosed herein for detecting phenomena or events in the protection zone of a power network EPN, at least one actual current signal in a corresponding sub-region of the current-interconnected protection zone is measured by a corresponding measuring instrument; a linear dependence between at least one measured current signal in at least one sub-region of the protection zone and a corresponding at least one measured current signal in at least another sub-region of the protection zone is determined by at least one intelligent electronic device; and the phenomenon in the protection zone of the power network is detected by at least one intelligent electronic device based on at least one determined linear dependence.

[0034] In other words, refer to Figure 1 For example, in the feeders forming the protection zone, one or more corresponding current signals, i.e., the corresponding phase currents and / or residual currents, are measured in the feeder, and the linear dependence between the actual corresponding measured current signals, i.e., the corresponding phase currents and / or residual currents, in at least one feeder and in at least one other feeder is determined. During normal network operation, only a weak (if any) actual linear dependence exists between the phase currents and / or residual currents in different feeders, but during abnormal network operation, a fairly strong actual linear dependence typically exists between the phase currents and / or residual currents in at least some feeders.

[0035] According to the implementation method, a linear dependence is determined between at least one measured current signal actually in each sub-region of the protection zone and at least one corresponding measured current signal actually in each other sub-region of the protection zone, and phenomena in the protection zone of the power network are detected based on the determined linear dependence. (Refer to...) Figure 1 For example, a linear dependency is determined between a measured current signal in each sub-region of a protection zone and a corresponding measured current signal in each other sub-region of the protection zone, and phenomena in the protection zone of the power network are detected based on the determined linear dependency. In this embodiment, a determined linear dependency exists between each sub-region of the protection zone. At its minimum, a phenomenon or event occurring in the protection zone can be detected by monitoring the linear dependency between only two sub-regions, but preferably, there are at least three sub-regions, among which linear dependencies are monitored. The higher the number of sub-regions where linear dependencies are determined, the higher the sensitivity for detecting phenomena or events occurring in the protection zone.

[0036] According to the implementation, the linear dependence between actual current signals in one sub-region of a protected area and their corresponding actual current signals in another sub-region can be determined by at least one intelligent electronic device determining the correlation between actual measured current signals in one sub-region and their corresponding actual measured current signals in another sub-region. In other words, the linear dependence between current signals in a sub-region can be determined by determining the correlation between one or more corresponding measured current signals (i.e., corresponding phase currents and / or residual currents) in the sub-region. This correlation provides an efficient way to determine the linear dependence between one or more actual current signals in different sub-regions forming interconnected protected areas in a power network. It should be noted that correlation calculation is only one possible way to determine the linear dependence between two time series. For example, a simple summation or difference of two time series provides some examples of other ways to determine the linear dependence between two time series.

[0037] According to the implementation, the linear dependence between a measured current signal in one sub-region and its corresponding measured current signal in another sub-region can be determined by calculating a correlation coefficient between the measured current signal in one sub-region and its corresponding measured current signal in another sub-region. In other words, the linear dependence between at least one measured current signal in one sub-region and its corresponding at least one measured current signal in at least one other sub-region can be determined by calculating a correlation coefficient between one or more corresponding measured current signals (i.e., corresponding phase currents and / or residual currents) in one sub-region and at least one other sub-region, by at least one intelligent electronic device. Calculating the correlation coefficient is a simple and effective way to provide information about the linear dependence between at least one measured current signal in one sub-region and at least one measured current signal in at least one other sub-region, because if the relationship between two variables is suspected to be linear, the calculation of the correlation coefficient provides a numerical measurement of linear dependence or linear relationship, i.e., a numerical measurement of the strength or degree of linear dependence between the two variables.

[0038] The Pearson product-moment correlation coefficient is one of the most commonly used correlation coefficients, and it is defined as the product of the covariance of the variables divided by the standard deviation of the variables.

[0039]

[0040] Where r is the Pearson product-moment correlation coefficient, and Cov(x, y) is the covariance of x and y. σ x and σ yLet r be the standard deviation of x and y. Here, r is a dimensionless quantity ranging from -1 to +1, i.e., -1 ≤ r ≤ +1.

[0041] When considering Figure 1 When determining the linear dependence between the actual phase currents IL1(F1), IL1(F2), IL1(F3), and IL1(F4) in the first phase L1 of feeders F1, F2, F3, and F4 in the protection zone of a power network EPN, this linear dependence can be determined, for example, by calculating the correlation coefficient between the actual phase currents IL1(F1), IL1(F2), IL1(F3), and IL1(F4) in the first phase L1 of each feeder F1, F2, F3, and F4 and the corresponding actual phase currents IL1(F1), IL1(F2), IL1(F3), and IL1(F4) in the first phase L1 of each other feeder F1, F2, F3, and F4. The correlation coefficient is calculated as follows: Figure 1 A set of sixteen correlation coefficients for the phase current of the first phase L1 in the protection zone of the power network EPN: r(IL(F1), IL1(F1)), r(IL1(F1), IL1(F2)), r(IL1(F1), IL1(F3)), r(IL1(F1), IL1(F4)), r(IL1(F2), IL1(F1)), r(IL1(F2), IL1(F2)), r(IL1(F2), IL1(F3)), r( IL1(F2), IL1(F4)), r(IL1(F3), IL1(F1)), r(IL1(F3), IL1(F2)), r(IL1(F3), IL1(F3)), r(IL1(F3), IL1(F4)), r(IL1(F4), IL1(F1)), r(IL1(F4)), IL1(F2)), r(IL1(F4)), IL1(F3)), r(IL1(F4)), IL1(F4)). For a more intuitive illustration, the calculated correlation coefficients can be presented in matrix form, such as... Figure 2a The diagram illustrates this. Because the correlation between a variable and itself is always one, the correlation coefficients r(IL1(F1), IL1(F1)), r(IL1(F2), IL1(F2)), r(IL1(F3), IL1(F3)) and r(IL1(F4), IL1(F4)) are all one, and therefore do not actually need to be calculated at all.

[0042] The linear dependence between the phase current in phase L2 or L3, or the zero current or residual current in the neutral conductor N between feeders F1, F2, F3, and F4, can be determined accordingly. This provides a total of four sets of correlation coefficients with sixteen coefficients in each set: one set of correlation coefficients for phase current IL2, one set for phase current IL3, and one set for residual current IN. Figure 2b , Figure 2c and Figure 2d As shown schematically in the diagram.

[0043] Below, refer to Figure 1 , Figure 3 Figure 4 and a more detailed description are given of an implementation method for detecting phenomena or events in a protected area of ​​a power grid, wherein... Figure 3 Some procedural steps for detecting phenomena or events in a power grid are schematically disclosed, and Figure 4a and Figure 4b An exemplary sample sequence used in determining the correlation coefficient is schematically disclosed.

[0044] As Figure 3 The first step involves collecting samples of at least one actual current signal in each feeder, substantially continuously, using appropriate measuring instruments. Preferably, samples are collected in all phase conductors within the protection zone of the power network and in the neutral conductor (if available) within all feeders within the protection zone of the network.

[0045] Therefore, refer to Figure 1 The first measuring instrument MI1 is configured to collect samples of the phase currents IL1, IL2, IL3 and the residual current IN in the first feeder F1 substantially continuously. Correspondingly, other measuring instruments MI2, MI3, and MI4 are configured to collect samples of the phase currents IL1, IL2, IL3 and the residual current IN in their respective feeders F2, F3, and F4 substantially continuously. The samples are then transmitted to the central intelligent electronic device (CIED) for further processing.

[0046] exist Figure 3 In the next step, at least one sample sequence of a predetermined number of samples of the corresponding current signals is created. More precisely, at least one sample sequence is created for each current signal measured in each feeder. Thus, referring to the example in Figure 4, at least sixteen sample sequences are created in total, namely, one sample sequence for measuring the phase currents IL1, IL2, IL3 and the residual current IN in each feeder FI, F2, F3, F4. Unless provided by the measuring instrument, the creation of the sample sequences can be performed in the Central Intelligent Electronic Device (CIED).

[0047] Figure 4aFour sample sequences are schematically disclosed, each consisting of ten samples T1 to T10 acquired at different times Tk, k = 1, ... 10. These sample sequences are measured in the first phase L1 of each feeder F1, F2, F3, and F4. Therefore, Figure 4a A sample sequence of current signals in the first phase L1 of each feeder F1, F2, F3, F4 is disclosed.

[0048] The number of sample sequences for each current signal in the power network can also be higher; for example, each current signal in the power network can have two sample sequences with different numbers of samples. In this case, the first sample sequence can have a lower number of samples, while the second sample sequence can have a higher number of samples. The first sample sequence can thus be intended for protection purposes, i.e., to protect the power network from unwanted phenomena or events, such as faults, occurring in the protection zone of the power network, whereby the number of samples in the sample sequence can correspond to a time period of, for example, 5 ms to 20 ms for a 50 Hz baseband power network. Because the number of samples in the sample sequence is multiplied by the number of sample sequences of corresponding current signals from the available different feeders, the very low number of samples in a single sample sequence may be sufficient to detect faults occurring in the protection zone of the power network. The second sample sequence can then be used for non-real-time post-fault analysis of phenomena or events before and after a fault, whereby the number of samples in the sample sequence can correspond to a time period of, for example, 200 ms to 400 ms for a 50 Hz baseband power network, such that at least some samples are obtained before the phenomenon or event occurs, and at least some samples are obtained after the phenomenon or event occurs.

[0049] As Figure 3 The next step in the implementation is to calculate the correlation coefficient between at least one sample sequence of the actual at least one current signal in each feeder and at least one corresponding sample sequence of the actual at least one current signal in each other feeder. (Refer to...) Figure 4a For example, calculate the correlation coefficient between the sample sequence IL1(F1) of the first phase current IL1 in the first feeder F1 and the sample sequence IL1(F2) of the first phase current IL1 in the second feeder F2, the correlation coefficient between the sample sequence IL1(F1) of the first phase current IL1 in the first feeder F1 and the sample sequence IL1(F3) of the first phase current IL1 in the third feeder F3, and the correlation coefficient between the sample sequence IL1(F1) of the first phase current IL1 in the first feeder F1 and the sample sequence IL1(F4) of the first phase current IL1 in the fourth feeder F4, until completion. Figure 2a The calculation ends at the first row of the matrix. Continue calculating the correlation coefficients until all available correlation coefficients have been calculated, that is, until the correlation coefficients have been determined. Figure 2a , Figure 2b , Figure 2c and Figure 2d This refers to all the correlation coefficients that have been schematically disclosed. The calculation of correlation coefficients can be performed within the Central Intelligent Electronic Device (CIED).

[0050] Typically, the Pearson correlation coefficient r between measurement sequences x and y is written as (x1, y1), (x2, y2), ..., (x... n y n The following equation can be used to calculate:

[0051]

[0052] Where n is the length of the dataset, and the correlation coefficient is calculated based on this dataset, therefore taking into account... Figure 4a In the example, n is ten. k It is a single sample of measurement x at time k, and y k It is a single sample of the measurement y at time k. It measures the average value of x. It measures the average value of y.

[0053] Because new samples are acquired substantially continuously at sampling frequencies typically in the kilohertz range, the amount of data samples would be enormous unless older samples were ignored. Therefore, according to one implementation, in response to the collection of at least one new sample of the current signal, at least one new sample is included in the corresponding sample sequence to become at least one last sample in the sample sequence, and at least one new sample is discarded from the corresponding sample sequence, the corresponding number of samples being the first in the sample sequence, and thereafter a correlation coefficient is calculated between at least one sample sequence of the actual at least one current signal in each feeder and the corresponding at least one sample sequence of the actual at least one current signal in each other feeder. Figure 4b The example sample sequences disclosed schematically are shown in the diagram, wherein at time T13, the three oldest samples T1, T2, and T3 in each sample sequence have been discarded, and each sample sequence already includes the three newest samples T11, T12, and T13. According to this embodiment, the correlation coefficient of the sample sequences is calculated by utilizing a sliding time window, thereby considering new samples in the calculation and discarding a corresponding number of the oldest samples based on the calculation.

[0054] According to the implementation method, in order to detect possible phenomena or events, such as faults, in the protection zone of a power grid EPN, at least one intelligent electronic device compares a calculated correlation coefficient with at least one (i.e., one or more) settable or fixed limit values, which indicate the significance, i.e., the degree of linear dependence, between the actual current signal in each feeder and the corresponding actual current signal in each other feeder. Figure 3 The example illustrates this step schematically. For instance, a limit value can be set in a central intelligent electronic device (CIED). For example, a limit value indicating the significance of linear dependence can be set based on the following principle:

[0055] • 0 indicates that there is no linear dependence between the corresponding phase currents or residual currents in the two sub-regions to be checked each time;

[0056] +1 indicates a perfectly linear positive dependence, meaning that the corresponding phase current or residual current in the two sub-regions being checked will increase or decrease simultaneously each time.

[0057] -1 indicates a completely linear negative dependence, meaning that when the phase current or residual current in the other sub-regions to be checked decreases, the corresponding phase current or residual current in one of the two sub-regions to be checked will increase each time.

[0058] • 0 < r ≤ 0.3 and -0.3 ≤ r < 0 indicate a weak positive linear dependence or a weak negative linear dependence between the corresponding phase currents or residual currents in the two sub-regions to be checked each time;

[0059] • 0.3 < r ≤ 0.7 and -0.7 ≤ r < -0.3 indicate a moderately positive or moderately negative linear dependence between the corresponding phase currents or residual currents in the two sub-regions to be checked each time; and

[0060] • 0.7 < r < +1 and -1 < r < -0.7 indicate a strong positive or strong negative linear dependence between the corresponding phase currents or residual currents in the two sub-regions to be checked each time.

[0061] According to one embodiment, the absolute value of the limit value used to indicate the significance of the linear dependence between an actual measured current signal in one sub-region and a corresponding measured current signal in another sub-region can be set to, for example, 0.5. Thus, a correlation coefficient with a value less than or equal to -0.5 is considered to indicate the significance of a negative dependence between the actual current signals in the two sub-regions in question, while a correlation coefficient with a value greater than or equal to 0.5 is considered to indicate the significance of a positive dependence between the actual current signals in the two sub-regions in question. Of course, other options can be chosen for the absolute value of the limit value used to indicate the significance of the linear dependence between an actual measured current signal in one sub-region and a corresponding measured current signal in another sub-region.

[0062] According to an implementation, at least one intelligent electronic device indicates an anomaly or event that has occurred in a protected area of ​​the power network, in response to at least one calculated correlation coefficient indicating the significance of the linear dependence between corresponding current signals in at least two different sub-regions. Figure 3 The example illustrates this step schematically. An anomaly or event in the protection zone of a power network indicates at least one phenomenon or event that deviates from the normal operation of the network, such as a fault occurring at a point in the current interconnection network within the protection zone of the power network.

[0063] Without considering any significant degree of correlation coefficient between corresponding current signals in at least two different sub-regions, it is assumed that the power network operates under normal operating conditions, i.e., it operates under conditions where power is supplied to loads connected to the network and there are no specific anomalies in the operation of the network.

[0064] According to the implementation, in response to an abnormal state of an indicated protected area of ​​the power network, at least one intelligent electronic device compares correlation coefficients indicating a significant linear dependence between corresponding current signals in at least two different sub-regions to determine the sub-regions where undesired phenomena or events have occurred and / or the type of phenomena or events such as faults. The type of fault can be, for example, a short-circuit fault involving two or more phase lines or a ground fault involving one or more phase lines, i.e., a ground fault.

[0065] According to the implementation, in response to an indication of an anomaly or event, particularly a fault, in at least one sub-region, a protection operation is initiated by at least one intelligent electronic device to clear or eliminate the fault. In other words, at least one intelligent electronic device initiates the process of clearing the fault by providing a control signal via its communication channel to a corresponding circuit breaker to disconnect power from / to the faulty sub-region / feeder in some or all phases of the faulty sub-region / feeder, typically for a very short time, e.g., only a few milliseconds. Therefore, the solution disclosed herein can be applied to protect a power network EPN from faults therein, and the length of the sample sequence used in the protection application can be the same as that just disclosed above.

[0066] According to an implementation, in response to at least one calculated correlation coefficient indicating the significance of the linear dependence between corresponding current signals in at least two different sub-regions, at least one sample sequence of the corresponding current signal in each sub-region is stored for post-analysis. The current signals can be stored, for example, in non-volatile memory within a central intelligent electronic device (CIED), or in a more advanced system within the control of a power grid (EPN). The length of the sample sequence to be stored can be the same as that just disclosed above.

[0067] The solutions disclosed herein have several advantages in monitoring the operation of power networks and detecting phenomena or events such as faults that occur therein.

[0068] This solution is easy to implement because it primarily uses statistical analysis of measurements, meaning it does not require extensive power system theory; rather, it is based solely on the fundamental properties of interconnected systems.

[0069] The correlation coefficient is calculated directly using measured samples of the corresponding current signal. That is, it does not require separating the fundamental network frequency components and higher-order harmonic components in the current signal, or calculating based on the phasor values ​​of these frequency components. In other words, the disclosed solution utilizes the entire spectrum of the phenomenon or event.

[0070] In addition, only a small number of parameters, namely one or more limit values, are needed to detect anomalies or events occurring in the network.

[0071] Moreover, compared to currently used applications, the disclosed solution can detect potential faults in the network much faster. This application is based on phasor calculations, which require a sample sequence length of at least one fundamental frequency cycle at their minimum, while preliminary offline tests show that the disclosed solution can detect similar faults even within a quarter of a fundamental frequency cycle. As mentioned above, this is based on a large amount of available data, i.e., all current measurements from all sub-regions / feeders forming the protection zone in the interconnected network, although a single current sample sequence does not include a very high number of samples.

[0072] Furthermore, utilizing all available current data from all sub-regions of the protection zone / feeders in the interconnected power network can also increase the sensitivity of fault detection, especially high-impedance ground faults, which typically do not cause any easily detectable increase in the magnitude of the residual current in a single feeder.

[0073] Figure 5 and Figure 6 The diagram schematically illustrates the correlation coefficients calculated for specific current measurements collected at 110kV / 20kV substations in a power network during an actual fault event. The power network has a total of eleven measurement points. Figure 5 and Figure 6 The figures are indicated by reference numerals J01 to J11. Figure 5 and Figure 6 In the attached diagram, reference numeral J01 refers to a measurement point in the 110kV input feeder supplying power to the substation, and reference numerals J05, J06, J07, J08, J10, and J11 refer to measurement points in the corresponding 20kV output feeders. Reference numerals J02, J03, J04, and J09 refer to measurement points located inside the substation or at unused feeder locations.

[0074] Figure 5 This is a schematic diagram of the correlation coefficient in matrix form. This correlation coefficient is determined for the actual measured zero current or residual current at different measurement points in the substation relative to measurement point J05 during a ground fault in the feeder. From Figure 5It can be seen that there is a strong negative correlation between measurement point J05 in one output feeder and measurement points J06, J07, J08, J10, and J11 in the corresponding other output feeders. This strong negative correlation is caused by the phenomenon of residual current flowing from healthy output feeders corresponding to measurement points J06, J07, J08, J10, and J11 to the faulty feeder corresponding to measurement point J05. Since there is at least no perceptible residual current in the other measurement points J01, J02, J03, J04, and J09, there is no perceptible correlation between the faulty feeder corresponding to measurement point J05 and the other healthy parts of the substation corresponding to measurement points J01, J02, J03, J04, and J09.

[0075] exist Figure 5 Among the correlation coefficients, a strong negative correlation in the residual current between the faulty output feeder corresponding to measurement point J05 and the healthy output feeders corresponding to measurement points J06, J07, J08, J10, and J11 indicates a ground fault in the output feeder corresponding to measurement point J05. This indication can be further verified by the absence of any perceptible correlation between the faulty feeder corresponding to measurement point J05 and other healthy parts of the substation corresponding to measurement points J01, J02, J03, J04, and J09. Therefore, according to the embodiment, a ground fault in an output feeder can be detected by a strong negative correlation in the residual current between the output feeder and at least one other output feeder.

[0076] Because the correlation between a variable and itself is always one, Figure 5 The diagonal of the matrix and subsequently Figure 6 The diagonal in the matrix indicates a perfectly linear positive dependency.

[0077] Figure 6 This is a schematic diagram of the correlation coefficient in matrix form, which is determined for the actual measured phase current in phase L1 at different measurement points in the substation relative to measurement point J05 during a short-circuit fault in the feeder. Figure 6 It can be seen that there is a strong positive correlation between measurement point J05 in each output feeder and measurement point J01 in each healthy input feeder. This strong positive correlation stems from the phenomenon that, in phase L1 of the network, the phase current mainly flows from the healthy input feeder corresponding to measurement point J01 to the faulty feeder corresponding to measurement point J05, which has a high current. The healthy output feeders corresponding to measurement points J06, J07, J08, J10, and J11 have normal low-level phase currents. Figure 6This indicates a low positive correlation between the healthy output feeders corresponding to measurement points J06, J07, J08, J10, and J11 and the healthy input feeder corresponding to measurement point J01, as well as the fault output feeder corresponding to measurement point J05. The other healthy components of the substation corresponding to measurement points J02, J03, J04, and J09 have no perceptible correlation with either the healthy input feeder corresponding to measurement point J01 or the fault output feeder corresponding to measurement point J05.

[0078] exist Figure 6 Among the correlation coefficients, a strong positive correlation in the phase current of phase L1 between the faulty output feeder corresponding to measurement point J05 and the healthy input feeder corresponding to measurement point J01, and a low positive correlation in the phase current of phase L1 between the faulty output feeder corresponding to measurement point J05 and other healthy output feeders corresponding to measurement points J06, J07, J08, J10, and J11, indicates a short-circuit fault in phase L1 of the output feeder corresponding to measurement point J05. This indication can also be verified by the absence of any perceptible correlation between the faulty feeder corresponding to measurement point J05 and other healthy parts of the substation corresponding to measurement points J02, J03, J04, and J09. Therefore, according to the embodiment, a short-circuit fault in the output feeder can be detected by the strong positive correlation of the phase current between the input feeder and the output feeder and by the low positive correlation of the phase current between the output feeder and at least one other output feeder.

[0079] Because the behavior of short-circuit faults between two-phase and three-phase provides a basis for... Figure 6 The correlation coefficient matrices shown are essentially similar, so a matrix can also be applied as a combination of the correlation coefficients of the phase currents of all phases L1, L2, L3 available in the power network in use.

[0080] Combinations of correlation matrices can be used to minimize the computational cost of calculating correlation coefficients and the storage space required to store them, even though the computational power and storage space available to central intelligent electronic devices are usually more than sufficient.

[0081] like Figure 5 and Figure 6 As can be seen, different types of phenomena or events occurring in current interconnected networks provide essentially unique patterns, such as fingerprints of phenomena or events in the network. When represented by correlation coefficients arranged in matrix form, this allows for simple yet highly informative visualization of network phenomena or events.

[0082] Although in the above example all the necessary analysis, determination, and calculation are provided by the Central Intelligent Electronic Device (CIED), the disclosed solution can also be applied to conventional power networks where feeder-specific intelligent electronic devices, such as protection and control relays, exist, provided that at least one of these feeder-specific intelligent electronic devices can perform the functions of the CIED disclosed herein. Alternatively, lower-level and higher-level intelligent electronic devices can exist, allowing the tasks to be performed to be assigned to different intelligent electronic devices. For example, refer to... Figure 1 For example, the central intelligent electronic device (CIED) at the power plant's ES (Electronic Storage Equipment) can provide lower-level intelligent electronic devices that simply collect measurement data and send it to another intelligent electronic device that provides a higher-level intelligent electronic device for further processing. Such higher-level intelligent electronic devices can also be implemented, for example, as cloud services.

[0083] It will be apparent to those skilled in the art that the concept of this invention can be implemented in various ways as technology advances. The invention and its embodiments are not limited to the examples described above, but can be varied within the scope of the claims.

Claims

1. A method for detecting phenomena in a power grid, said power grid comprising at least one protection zone having at least two sub-regions, the method comprising: Measure at least one actual current signal in the corresponding sub-region of the protected area. Determine the linear dependence between at least one measured current signal in each sub-region of the protected area and the corresponding at least one measured current signal in each other sub-region of the protected area, and The phenomenon in the protection zone of the power network is detected based on the determined linear dependence. in, The linear dependence is determined by calculating at least one correlation coefficient between at least one actual measured current signal in each sub-region of the protection zone and at least one corresponding actual measured current signal in each other sub-region of the protection zone. Specifically, at least one calculated correlation coefficient is compared with at least one limit value indicating the degree of linear dependence to detect phenomena within the protection zone of the power network. Specifically, in response to at least one calculated correlation coefficient indicating the significance of the linear dependence between the actual corresponding current signals in at least two sub-regions of the protection zone, a phenomenon that has occurred in the protection zone of the power network is detected. Specifically, correlation coefficients, which indicate the significance of the linear dependence between actual corresponding current signals in at least two different sub-regions, are compared to determine the sub-regions where the phenomenon has occurred and / or the type of the phenomenon. The phenomenon described refers to a fault occurring in an interconnected power grid.

2. The method according to claim 1, characterized in that... Continuously collect samples of at least one actual current signal in the sub-region. At least one sample sequence that generates a predetermined number of samples of the corresponding current signal, and Calculate the correlation coefficient between the sample sequence of the actual current signal in the sub-region and the corresponding sample sequence of the actual current signal in another sub-region.

3. The method according to claim 2, characterized in that... In response to the collection of at least one new sample of the current signal, at least one newly collected sample is added to the corresponding sample sequence as at least one last sample in the sample sequence, and the corresponding number of samples that are first in the sample sequence are discarded from the corresponding sample sequence. Calculate the correlation coefficient between the sample sequence of the actual current signal in the sub-region and the corresponding sample sequence of the actual current signal in another sub-region.

4. A system for detecting phenomena in a power grid, said power grid comprising at least one protection zone having at least two sub-regions, said system comprising At least one measuring instrument for each sub-region of the protected area, for measuring at least one actual current signal in the corresponding sub-region. At least one intelligent electronic device is configured to determine the linear dependence between at least one measured current signal actually in each sub-region of the protected area and a corresponding at least one measured current signal actually in each other sub-region of the protected area. At least one intelligent electronic device is configured to detect phenomena in the protection zone of the power network based on a determined linear dependency. At least one intelligent electronic device is configured to determine the linear dependence by calculating at least one correlation coefficient between at least one actual measured current signal in each sub-region of the protected area and a corresponding at least one actual measured current signal in another sub-region of the protected area. At least one intelligent electronic device is configured to compare at least one calculated correlation coefficient with at least one limit value indicating the degree of linear dependence to detect phenomena in the protection zone of the power network. At least one intelligent electronic device is configured to detect phenomena occurring within the protection zone of the power network in response to at least one calculated correlation coefficient, which indicates the significance of the linear dependence between actual corresponding current signals in at least two sub-regions of the protection zone. At least one intelligent electronic device is configured to compare correlation coefficients that indicate the significance of the linear dependence between actual corresponding current signals in at least two different sub-regions to determine the sub-region where the phenomenon has occurred and / or the type of the phenomenon. in, The phenomenon described is a fault that occurs in an interconnected power grid.

5. The system according to claim 4, characterized in that, At least one measuring instrument in the sub-region is configured to continuously collect samples of at least one actual current signal in the sub-region, and The system also includes: At least one intelligent electronic device configured to generate at least one sample sequence of a predetermined number of samples of a corresponding current signal, and At least one intelligent electronic device is configured to calculate a correlation coefficient between a sample sequence of actual current signals in the said sub-region and a corresponding sample sequence of actual current signals in another sub-region.

6. The system according to claim 5, characterized in that, Also includes At least one intelligent electronic device is configured to, in response to the collection of at least one new sample of the current signal, add at least one newly collected sample to a corresponding sample sequence as at least one last sample in the sample sequence, and discard the corresponding number of samples that are first in the sample sequence from the corresponding sample sequence. At least one intelligent electronic device is configured to calculate a correlation coefficient between a sample sequence of actual current signals in the said sub-region and a corresponding sample sequence of actual current signals in another sub-region.

Images