Percent conversion transformer interface power system coherent analysis method based on transfer coefficients

Abstract

The present application discloses belongs to the technical field of power system stability control, and particularly relates to a 100% converter interface power system homology analysis method based on transfer coefficient, comprising: generating a system node admittance matrix Y according to network topology structure; generating a node impedance matrix Z corresponding to the system node admittance matrix Y based on the system node admittance matrix Y; generating a transfer coefficient matrix K based on diagonal elements and non-diagonal elements of the node impedance matrix Z; using K-Mean clustering analysis to obtain homology clustering results for the transfer coefficient matrix K of all power supply node rows, and realizing 100% converter interface power system homology analysis. The present application discloses a converter homology clustering mechanism after a large disturbance occurs in a 100% converter interface power system, proposes a converter homology clustering identification method based on transfer coefficient, and realizes 100% converter interface power system homology analysis.

Description

Technical Field

[0001] This invention belongs to the field of power system stability control technology, and specifically relates to a method for analyzing the coherence of a 100% converter interface power system based on the transfer coefficient. Background Technology

[0002] With the continuous increase in installed capacity of new energy sources, the proportion of converter-interfaced generation (CIG) power sources has risen sharply, posing a significant challenge to analyzing the stability of new power systems. In the near future, there is a possibility that traditional synchronous machine power sources in power systems will be completely replaced by converter-interfaced generation, thus forming a 100% converter-interfaced generation power system, fundamentally changing the dynamic characteristics of the power system.

[0003] The analysis of the coherence characteristics of power systems studies the oscillation and clustering problems among different generating units after large disturbances. This is of great significance for understanding the interaction modes of different power sources through the power grid after large disturbances and is the foundation for studies such as the Extended Equal Area Rule (EEAC) and dissociation. In traditional power system coherence analysis, the dynamic process of synchronous generators is generally described using a second-order rotor motion equation model. The smaller the electrical distance between any two generators, the greater the probability of coherence, thus transforming coherence analysis into judging the size of the electrical distance between different generators. In existing technology, synchronous generator nodes are equivalent to voltage source series impedances. Then, Kron simplification is used to equivalently shrink all non-generator nodes in the power grid to obtain a simplified diagram containing only generator nodes. The electrical distance between any two generators can be characterized by the impedance between the potential nodes within the two generators, also known as transfer impedance.

[0004] Analyzing the compatibility characteristics of a 100% converter-interface power system is crucial for revealing the interaction between the converter-interface power source and the power grid, as it is influenced by various factors such as converter topology, control strategy, and the connection method between the converter and the grid. Whether a 100% converter-interface power system possesses a relatively fixed compatibility pattern similar to traditional power systems remains inconclusive. Differences in control methods between grid-forming and grid-following converters affect the compatibility characteristics of the converter system. Grid-forming (GFM) and grid-following (GFL) converters exhibit voltage source and current source characteristics respectively in their external characteristics, and their steady-state internal impedances are finite and infinite, respectively. Therefore, traditional compatibility analysis methods based on equivalent condensation and transfer impedance in power systems are no longer applicable. Thus, an effective method for analyzing the compatibility characteristics of a 100% converter-interface power system is urgently needed. Summary of the Invention

[0005] The purpose of this invention is to provide a method for homogeneity analysis of a 100% converter interface power system based on the transfer coefficient, characterized by the following steps:

[0006] Step S1: Generate the system node admittance matrix Y based on the network topology;

[0007] Step S2: Based on the system node admittance matrix Y, generate the node impedance matrix Z corresponding to the system node admittance matrix Y;

[0008] Step S3: Generate the transfer coefficient matrix K based on the diagonal and off-diagonal elements of the nodal impedance matrix Z;

[0009] Step S4: For the blocks of all power node rows, based on the transition coefficient matrix K, use K-Mean clustering analysis to obtain the homology grouping results, and realize the homology analysis of the 100% converter interface power system.

[0010] The system node admittance matrix Y in step S1 is:

[0011]

[0012] In the formula, Indicates the voltage at node i. Represents the voltage at node j. Y represents the injected current at node j. ij This represents the ratio of the injected current at node j to the voltage at node i when all other nodes except node i are grounded.

[0013] The step of generating the system node admittance matrix Y in step S1 is as follows:

[0014] Based on the network topology, the equivalent impedance of the load power and the internal impedance of the power source are included in the power flow calculation system node admittance matrix in the form of admittance, and then corrected according to the fault type to obtain the system node admittance matrix Y. In the system node admittance matrix Y, the internal impedance of the grid-type converter is considered to be infinite, and the internal impedance of the network-type converter is considered to be a non-zero finite value according to the control parameters and circuit parameters.

[0015] The node impedance matrix Z in step S2 is:

[0016]

[0017] In the formula, Indicates the voltage at node i. Represents the voltage at node j. This indicates the current injected into node i. Z represents the injected current at node j. ijZ represents the ratio of the voltage at node j to the current at node i when all other nodes except node i have zero injected current. It can also be described as the voltage caused by a unit current at node j. Its value is non-zero, meaning the node impedance matrix Z is full-scale. When i equals j, Z... ij This represents the equivalent impedance of the network as seen from node i, where n is the total number of nodes in the node impedance matrix Z.

[0018] In step S3, the transfer coefficient matrix K is generated, and the transfer impedance z is defined. ij for:

[0019]

[0020] In the formula, z ij Let be the transfer impedance of the power supply connected from node j to node i. Let be the voltage value of the potential source connected to node i. Z is the transfer current from node j to node i. jj Z represents the self-impedance of node j in the nodal impedance matrix. ij Let k be the mutual impedance between node j and node i in the nodal impedance matrix. ij Z is the transfer coefficient of the power supply from node j to node i. ig Let be the internal impedance of the power supply connected to node i.

[0021] The transition coefficient matrix K in step S3 is:

[0022]

[0023] In the formula, k ij This indicates the power supply connected to node i from node j. The transfer coefficient.

[0024] The step S4, which uses K-Mean clustering analysis to obtain homology clustering results, is as follows:

[0025] Assuming there are n nodes in the network and m nodes connected to the power source, the magnitude of the m×n dimensional transfer coefficients in the row corresponding to the power source in the obtained transfer coefficient matrix K is used for cluster analysis using the K-Means algorithm to obtain the cluster analysis results of the system after the fault, and then the coherence grouping results of the power source when a three-phase short circuit occurs at that node are obtained.

[0026] The beneficial effects of this invention are as follows:

[0027] This invention discloses a method for homology analysis of 100% converter interface power systems based on transfer coefficients. The transfer coefficient is proposed based on the concept of transfer impedance and serves as a representation of electrical distance. The transfer coefficient effectively solves the problem of the inability to measure electrical distance due to infinite internal impedance of current sources, and formally unifies the index for evaluating electrical distance. Based on the transfer coefficient matrix K, homology clustering results of the 100% converter interface power system are obtained. Furthermore, the transfer coefficient can also measure system stability to a certain extent. Simulation results verify the theoretical analysis. This method has certain reference value for the transient stability analysis of power systems with a high proportion of converters. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the flow chart of the 100% converter interface power system homology analysis method based on the transfer coefficient of the present invention.

[0029] Figure 2 A simplified diagram of a multi-port system architecture;

[0030] Figure 3 This is a schematic diagram of the impedance transformation at node j of the system.

[0031] Figure 4 This is a schematic diagram illustrating the physical meaning of the elements in the j-th column of the transfer coefficient matrix K;

[0032] Figure 5 Schematic diagram of a 100% converter interface power system for an IEEE 39-node network.

[0033] Figure 6 Schematic diagram of the IEEE 39-node hybrid 100% converter interface power system;

[0034] Figure 7 The simulation diagram of a three-phase short circuit at node 11 of the grid-type 100% converter interface power system is shown, where: (a) is the relative angle difference of the converter, and (b) is the relative frequency difference of the converter.

[0035] Figure 8 A schematic diagram of the clustering results of the three-phase short-circuit transfer coefficient of node 11 of the grid-type 100% converter interface power system is shown, where: (a) is a schematic diagram of the clustering analysis results when k=3, and (b) is a schematic diagram of the clustering analysis results when k=2;

[0036] Figure 9 The simulation diagram of a three-phase short circuit at node 22 of the interface power system of a hybrid 100% converter is shown, where: (a) is the relative angle difference of the converter, and (b) is the relative frequency difference of the converter.

[0037] Figure 10The diagram shows the clustering results of the three-phase short-circuit transfer coefficient at node 22 of the interface power system of the hybrid 100% converter. Among them, (a) is the clustering analysis result when k=2, (b) is the clustering analysis result when k=3, (c) is the clustering analysis result when k=4, (d) is the clustering analysis result when k=5, and (e) is the clustering analysis result when k=6. Detailed Implementation

[0038] This invention provides a method for homogeneity analysis of a 100% converter interface power system based on the transfer coefficient. The invention will be further described in detail below with reference to the accompanying drawings.

[0039] like Figure 1 The embodiment of the present invention shown provides a method for homogeneity analysis of a 100% converter interface power system based on the transfer coefficient, including the following steps:

[0040] Step S1: Generate the system node admittance matrix Y based on the network topology;

[0041] Based on the network topology, the equivalent impedance of the load power and the internal impedance of the power source are included in the system node admittance matrix for power flow calculation in the form of admittance, and then corrected according to the fault type to obtain the system node admittance matrix Y:

[0042]

[0043] In the formula, Indicates the voltage at node i. Represents the voltage at node j. Y represents the injected current at node j. ij This represents the ratio of the injected current at node j to the voltage at node i when all other nodes except node i are grounded.

[0044] In this embodiment, the equivalent impedance of the load power and the internal impedance of the power source are included in the system node admittance matrix Y for power flow calculation in the form of admittance. In the system node admittance matrix Y, the internal impedance of the grid-type converter is considered to be infinite, that is, the admittance is 0; the internal impedance of the grid-type converter is considered to be a non-zero finite value based on the control parameters and circuit parameters, that is, it is not included in the form of an ideal voltage source.

[0045] Step S2: Based on the system node admittance matrix Y, generate the node impedance matrix Z corresponding to the system node admittance matrix Y;

[0046] The node impedance matrix Z corresponding to Y in step S2 is obtained by using the inverse relationship between the system node admittance matrix Y and the node impedance matrix Z:

[0047]

[0048] In the formula, Indicates the voltage at node i. Represents the voltage at node j. This indicates the current injected into node i. Z represents the injected current at node j. ij Z represents the ratio of the voltage at node j to the current at node i when all other nodes except node i have zero injected current. It can also be described as the voltage caused by a unit current at node j. Its value is non-zero, meaning the node impedance matrix Z is full-scale. When i equals j, Z... ij This represents the equivalent impedance of the network as seen from node i, where n is the total number of nodes in the node impedance matrix Z.

[0049] Step S3: Generate the transfer coefficient matrix K based on the diagonal and off-diagonal elements of the nodal impedance matrix Z;

[0050] For a multi-node network system, when the potential source When the voltage value of the potential source connected to node i exists alone, it is equivalent to injecting current into node i. Z ig This is the internal impedance of the power source connected to node i. From the node impedance matrix Z, we know that a voltage will be generated at node j. If the three phases at node j are short-circuited, there will be current.

[0051] In step S3, the transfer coefficient matrix K is generated, and the transfer impedance z is defined. ij for:

[0052]

[0053] In the formula, z ij Let be the transfer impedance of the power supply connected from node j to node i. Let be the voltage value of the potential source connected to node i. Z is the transfer current from node j to node i. jj Z represents the self-impedance of node j in the nodal impedance matrix. ij Let k be the mutual impedance between node j and node i in the nodal impedance matrix. ij Z is the transfer coefficient of the power supply from node j to node i. ig Let be the internal impedance of the power supply connected to node i.

[0054] The transfer impedance z ij This is equivalent to creating a virtual impedance at node j that is directly connected to the power supply connected to node i, mapping node i in the network to j. The impedance only needs to be multiplied by the transfer factor k. ij Then the potential source from node j can be obtained. The transfer impedance.

[0055] Note the transfer coefficient k ij It is composed of the corresponding elements of the node impedance matrix Z, and therefore contains information on the power supply external characteristics, network parameters and load levels of the entire system.

[0056] From the transfer coefficient k ij The transition coefficient matrix K can be constructed as follows:

[0057]

[0058] In the formula k ij This indicates the power supply connected to node i from node j. The transfer coefficient.

[0059] In this embodiment, a transfer coefficient is proposed based on the concept of transfer impedance, which serves as a representation of electrical distance. The transfer coefficient effectively solves the problem of the inability to measure electrical distance due to the infinite internal impedance of the current source, and formally unifies the index for evaluating electrical distance. Based on the transfer coefficient matrix K, the homology grouping results of the power system of the 100% converter interface power supply are obtained. In addition, the transfer coefficient can also measure system stability to a certain extent.

[0060] Step S4: For the blocks of all power node rows, based on the transition coefficient matrix K, use K-Mean clustering analysis to obtain the homology grouping results, and realize the homology analysis of the 100% converter interface power system.

[0061] Assuming there are n nodes in the network and m nodes connected to power sources, the magnitudes of the m×n dimensional transfer coefficients in the row corresponding to the power source in the obtained transfer coefficient matrix K are used for cluster analysis using the K-Means algorithm. This yields the cluster analysis results of the system after a fault, and further, the homogeneity grouping results of the power sources experiencing a three-phase short circuit at that node. Specifically, as the value of parameter k increases, the clustering becomes more rigorous, resulting in more clusters. This indicates that as the observation time window length increases, the dynamic behavior of different converters is more finely divided, leading to more homogeneous results. It should be noted that the larger the system scale and the stronger the coupling between power sources, the more pronounced the homogeneity becomes.

[0062] In summary, this invention discloses a method for coherence analysis of 100% converter interface power systems based on transfer coefficients. The equivalent impedance of the load power and the internal impedance of the power source are included in the system node admittance matrix Y for power flow calculation in the form of admittance. The node impedance matrix Z is obtained by inverting the system node admittance matrix Y. The transfer coefficient matrix K is generated from the diagonal and off-diagonal elements of the node impedance matrix Z. The electrical distance from a certain point in the system to the power source is obtained based on the magnitude of the transfer coefficient matrix K. Based on the electrical distance, K-Mean clustering analysis is used to obtain the system fault power source coherence clustering results, thus realizing 100% converter interface power system coherence analysis.

[0063] The proposed method for harmonic analysis of 100% converter interface power systems based on transfer coefficients is based on the converter harmonic mechanism of large disturbances causing faults in 100% converter interface power systems. It implements a method for identifying converter harmonic clusters based on transfer coefficients. This method unifies grid-connected converters (with external characteristics as current sources) and grid-connected converters (with external characteristics as voltage sources) into a system based on harmonic analysis determined by electrical distance, enabling rapid and convenient analysis of the harmonic characteristics of 100% converter systems.

[0064] To verify the effectiveness of the 100% converter interface power system coherence analysis method based on transfer coefficient disclosed in this invention, a multi-port system was constructed, and the following simulation experiments were conducted for verification.

[0065] The multi-port system architecture is as follows: Figure 2 As shown in the diagram, the transformation of the power supply from node j to node i after eliminating intermediate nodes is illustrated in the figure below. Figure 3 As shown, the definition of the transfer coefficient is obtained, and the physical meaning of the j-th column element of the transfer coefficient is as follows: Figure 4 As shown. In this embodiment, the IEEE 39-node network-type 100% converter interface power system is defined as follows: Figure 5 As shown, G1-G10 are all grid-type converters; the IEE E39 node hybrid 100% converter interface power system is defined as follows: Figure 6 As shown, the power supplies G1, G2 and G3 connected to nodes 30, 31 and 32 are grid-connected converters, while the rest (G4-G10) are grid-connected converters.

[0066] In both systems described above, the grid-type converter controls the voltage of the LCL capacitor to a given value. Therefore, the inductor branch of the grid-type converter furthest from the inverter side is considered its internal impedance, while the internal impedance of the grid-type converter is considered infinite. The angle θ of the synchronous machine or converter connected to bus 1 is also considered. 1* and frequency ω 1* Set as the baseline value, δ i =θ i* -θ 1* and ω i =ωi* -ω 1* (i = 30, 31, ..., 38) represents the angle and frequency changes of the synchronous machine or converter connected to node i relative to the synchronous machine and converter connected to node 1. It should be noted that since the simulation results are relative curves, the row containing the baseline point needs to be removed during cluster analysis, and cluster analysis should be performed using the data from the (m-1)×n dimensional transfer coefficient matrix. In this embodiment, a short circuit is set, which requires adding a large number to the corresponding self-admittance of the node admittance matrix to indicate that the node is short-circuited to ground. For example... Figure 4 As shown, from the physical meaning of the elements in the j-th column of the transfer coefficient matrix K, it can be concluded that the values ​​of the corresponding columns of the transfer coefficients obtained from the network matrix after the short-circuit fault repair are consistent with the values ​​obtained from the steady state before the fault.

[0067] In such Figure 5 In the IEEE 39-node network type 100% converter interface power system shown, a three-phase short-circuit fault occurs at node 11. The simulation time is 2s, the fault occurs at 0.1s, the system frequency is 50Hz, the fault lasts for 10 cycles, and the fault is cleared at 0.3s.

[0068] In this embodiment, the simulation results of the three-phase short circuit at node 11 of the grid-type 100% converter interface power system are as follows: Figure 7 As shown, (a) represents the relative angle difference of the converters, and (b) represents the relative frequency difference of the converters; the horizontal axis of (a) and (b) is the simulation time.

[0069] Clustering results of the three-phase short-circuit transfer coefficient at node 11 of the grid-type 100% converter interface power system are as follows: Figure 8 As shown, (a) is a schematic diagram of the cluster analysis results when k=3, and (b) is a schematic diagram of the cluster analysis results when k=2; Figure 8 (a) and Figure 8 (b) presents the clustering analysis results for K-Means with parameter k set to 3 and 2, respectively, in the homology analysis method. Figure 8 As shown, its clustering results are similar to Figure 7 The simulation results shown are consistent.

[0070] In such Figure 6 In the IEEE 39-node hybrid 100% converter interface power system shown, a three-phase short-circuit fault is considered at node 22. The simulation time is 1.5s, the fault occurs at 0.1s, the system frequency is 50Hz, the fault lasts for 30 cycles, and the fault is cleared at 0.7s.

[0071] In this embodiment, the simulation results of the three-phase short circuit at node 22 of the hybrid 100% converter interface power system are as follows: Figure 9As shown, (a) represents the relative angle difference of the converters, and (b) represents the relative frequency difference of the converters; the horizontal axis of (a) and (b) is the simulation time.

[0072] Clustering results of the three-phase short-circuit transfer coefficient at node 22 of the hybrid 100% converter interface power system are as follows: Figure 10 As shown, (a) is a schematic diagram of the cluster analysis results when k=2, (b) is a schematic diagram of the cluster analysis results when k=3, (c) is a schematic diagram of the cluster analysis results when k=4, (d) is a schematic diagram of the cluster analysis results when k=5, and (e) is a schematic diagram of the cluster analysis results when k=6. Figure 10 (a)- Figure 10 (e) presents the clustering results for K-Means with parameter k set to 2, 3, 4, 5, and 6 in the homology analysis method, such as... Figure 10 As shown, its clustering results are similar to Figure 9 The simulation results shown are consistent.

[0073] In summary, this invention discloses a method for harmonic analysis of 100% converter interface power systems based on transfer coefficients. This method addresses the problem that impedance-based electrical distance analysis methods are no longer applicable because grid-connected converters exhibit current source characteristics and have infinite internal impedance. It unifies grid-connected converters (with current source external characteristics) and grid-connected converters (with voltage source external characteristics) into a harmonic analysis system based on electrical distance. This allows for rapid and convenient analysis of the harmonic characteristics of 100% converter systems and solves the problem of uniformly measuring the electrical distance from a point in the system to the converter interface power supply.

Claims

1. A method for homogeneity analysis of a 100% converter interface power system based on transfer coefficients, characterized in that, Includes the following steps: Step S1: Generate the system node admittance matrix Y based on the network topology; Step S2: Based on the system node admittance matrix Y, generate the node impedance matrix Z corresponding to the system node admittance matrix Y; Step S3: Generate the transfer coefficient matrix K based on the diagonal and off-diagonal elements of the nodal impedance matrix Z; Step S4: For the blocks of all power node rows, based on the transition coefficient matrix K, use K-Mean clustering analysis to obtain the homology clustering results, and realize the homology analysis of the 100% converter interface power system. In step S3, the transfer coefficient matrix K is generated, and the transfer impedance z is defined. ij for: (3) In the formula, z ij Let be the transfer impedance of the power supply connected from node j to node i. Let be the voltage value of the potential source connected to node i. Z is the transfer current from node j to node i. jj Let j be the self-impedance of node j in the nodal impedance matrix. Let be the mutual impedance between node j and node i in the nodal impedance matrix. Let be the power transfer coefficient from node j to node i. Let be the internal impedance of the power supply connected to node i.

2. The method for homogeneity analysis of 100% converter interface power systems based on transfer coefficients according to claim 1, characterized in that, The system node admittance matrix Y in step S1 is: (1) In the formula, , Indicates the voltage at node i. Represents the voltage at node j. This indicates the injected current at node j. This represents the ratio of the injected current at node j to the voltage at node i when all other nodes except node i are grounded.

3. The method for homogeneity analysis of 100% converter interface power systems based on transfer coefficients according to claim 1, characterized in that, The step of generating the system node admittance matrix Y in step S1 is as follows: Based on the network topology, the equivalent impedance of the load power and the internal impedance of the power source are included in the power flow calculation system node admittance matrix in the form of admittance, and then corrected according to the fault type to obtain the system node admittance matrix Y. In the system node admittance matrix Y, the internal impedance of the grid-type converter is considered to be infinite, and the internal impedance of the network-type converter is considered to be a non-zero finite value according to the control parameters and circuit parameters.

4. The method for homogeneity analysis of 100% converter interface power systems based on transfer coefficients according to claim 1, characterized in that, The node impedance matrix Z in step S2 is: (2) In the formula, , Indicates the voltage at node i. Represents the voltage at node j. This indicates the current injected into node i. This indicates the injected current at node j. This represents the ratio of the voltage at node j to the current at node i when all other nodes except node i have zero injected current. It can also be called the voltage value caused by a unit current at node j. Its value is a non-zero value, that is, the node impedance matrix Z is a full matrix. When i equals j This represents the equivalent impedance of the network as seen from node i, where n is the total number of nodes in the node impedance matrix Z.

5. The method for homogeneity analysis of 100% converter interface power systems based on transfer coefficients according to claim 1, characterized in that, The transition coefficient matrix K in step S3 is: (4) In the formula, k ij This indicates the power supply connected to node i from node j. The transfer coefficient.

6. The method for homogeneity analysis of 100% converter interface power systems based on transfer coefficients according to claim 1, characterized in that, The step S4, which uses K-Mean clustering analysis to obtain homology grouping results, is as follows: Assuming there are n nodes in the network and m nodes connected to the power source, the magnitude of the m×n dimensional transfer coefficients in the row corresponding to the power source in the obtained transfer coefficient matrix K is used for cluster analysis using the K-Means algorithm to obtain the cluster analysis results of the system after the fault, and then the coherence grouping results of the power source when a three-phase short circuit occurs at that node are obtained.

Images