A method for identifying a driving node of a power system and application thereof

Abstract

The application discloses a kind of identification method and application of power system driving node, belong to power system analysis technical field;The present application considers practical application scene, according to controllable condition, candidate driving node is iteratively optimized, remaining set is regarded as part driving node, and unmarked node is controlled under the part driving node set and is marked;Part driving node and its connection edge are deleted, and then for each connected subgraph in remaining structure, find part driving node that can control unmarked node within limited steps;Such cycle, constantly expand optimization driving node set, until network is completely controllable, obtain driving node set under given limited step.The application can realize the identification of driving node of network control under given step length, for the complex network of high spatial dimension such as power system, accurate identification of driving node can be realized with lower complexity.

Description

Technical Field

[0001] This invention belongs to the field of power system analysis technology, and more specifically, relates to a method and application for identifying driving nodes in a power system. Background Technology

[0002] In practical power system control, failures in certain components can lead to large-scale cascading failures, resulting in widespread regional blackouts, severe social impacts, and significant revenue losses. Therefore, identifying and controlling the driving components of the power system to restore normal system operation within the expected timeframe is of great practical significance.

[0003] A power system can be viewed as a large, complex network with a set of nodes and edges, which is a complex dynamic network system. Identification of driving nodes in a power system is an NP-hard combinatorial optimization problem. Most existing methods directly use the maximum matching algorithm for solution. However, the theoretical analysis of this method does not consider the time required for the control signal to fully propagate in the network. Due to the high spatial dimension of such large and complex networks as power systems, in practical applications of power systems, controlling the driving nodes identified by the maximum matching algorithm and thus controlling the entire network often requires a long time and cannot achieve accurate identification of driving nodes within a fixed time step. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the prior art, the present invention provides a method and application for identifying drive nodes in a power system, so as to solve the technical problem that the prior art cannot achieve accurate identification of drive nodes with low complexity.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for identifying drive nodes in a power system, comprising the following steps:

[0006] S1. Initialize the candidate driver node set K'; determine matrix B based on set K'; where matrix B is used to represent the connection relationship between the driver node and all nodes; the power system network is a network established with generators, loads and buses in the power system as nodes and transmission lines connecting each node as edges;

[0007] S2. Substitute the current matrix B into the expression F(Δt,n,A,B) of the coefficient matrix to obtain the current coefficient matrix; where the expression F(Δt,n,A,B) of the coefficient matrix is ​​obtained by discretizing the dynamic formula of the power system network; Δt is the discrete step length; n is the number of evolution steps in the discretization process; matrix A is the adjacency matrix of the power system network.

[0008] S3. If the rank R of the current coefficient matrix is ​​full rank, then all nodes in the current set K' are taken as the identified target driving nodes; otherwise, go to step S4.

[0009] S4. Calculate the rank R' of the coefficient matrix corresponding to each node removed from set K'; when R' corresponding to node P is equal to R, remove node P from set K', obtain the coefficient matrix corresponding to the current set K', perform linear independence analysis on the coefficient matrix, and mark the nodes with linear independence.

[0010] S5. Delete the nodes and their edges in set K' from the backup network of the power system network to obtain a new power system network. From the new power system network, obtain the nodes that can control the unmarked nodes to form a set D.

[0011] S6. Merge set D into set K' to update set K', update the coefficient matrix accordingly, and go to step S3.

[0012] More preferably, the dynamic formula of the above-mentioned power system network is:

[0013]

[0014] Where X is the state vector of each node in the power system; X is the derivative of X; U is the vector formed by the control signals of each drive node in the power system.

[0015] More preferably, the discretized expression of the dynamic formula of the power system network is as follows:

[0016]

[0017] Among them, X n Let X0 be the state vector of the power system at time nΔt; I is the identity matrix; X0 is the state vector at time 0; U q It is a vector composed of the control signals of each driving node at time qΔt in the power system.

[0018] More preferably, the dynamic formula of the power system network is discretized and represented in matrix form as follows:

[0019]

[0020] in, For U q The vector formed is q = 0, 1, ..., n-1.

[0021] More preferably, the coefficient matrix F(Δt,n,A,B)=[B,(I+AΔt) 1B,...,(I+AΔt) n-1 B].

[0022] More preferably, the method for initializing the candidate driver node set K' includes: adding the driver nodes obtained by solving the dynamic formula to the set K' to initialize the set K'.

[0023] More preferably, the method for obtaining the above set D includes:

[0024] A1. Form a set of nodes to be matched from all unmarked nodes;

[0025] A2. Based on the maximum matching algorithm, all nodes in the set of nodes to be matched in the new power system network are matched to obtain the set of matched nodes. Then, all nodes in the new power system network that failed to match are merged with the set of matched nodes to reconstruct the set of nodes to be matched.

[0026] A3. Repeat step A2 for iteration until the number of iterations is n. At this point, the set of matching nodes is set D.

[0027] In a second aspect, the present invention provides a power system drive node identification system, comprising: a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the power system drive node identification method provided in the first aspect of the present invention.

[0028] Thirdly, the present invention provides a control method for a power system, comprising: identifying a drive node in a power system using the power system drive node identification method provided in the first aspect of the present invention, and controlling the identified drive node.

[0029] Fourthly, the present invention provides a power system, including: a drive node identification module, used to execute the identification method provided in the first aspect of the present invention to obtain the power system drive nodes.

[0030] Fifthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed by a processor, it controls the device where the storage medium is located to execute the power system drive node identification method provided in the first aspect of the present invention and / or the power system control method provided in the third aspect of the present invention.

[0031] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects:

[0032] 1. This invention provides a method for identifying driving nodes in a power system. It transforms an NP-hard problem into a deterministic iterative optimization. Considering practical application scenarios, it iteratively optimizes candidate driving nodes based on controllability conditions (rank of the coefficient matrix), using the remaining set as a subset of driving nodes and marking the nodes controllable within this subset. The subset of driving nodes and their edges are then removed. For each connected subgraph in the remaining structure, the method searches for driving nodes capable of controlling unmarked nodes within a finite number of steps. This process is repeated, continuously expanding and optimizing the set of driving nodes until the network is fully controllable, obtaining the set of driving nodes for a given finite number of steps. This invention enables the identification of driving nodes for network control within a given step size. For large and complex networks like power systems with high spatial dimensions, it achieves accurate identification of driving nodes with relatively low complexity.

[0033] 2. Furthermore, the power system drive node identification method provided by the present invention transforms the problem of selecting drive nodes to reach any target state within a given step energy control network into maximum matching of complex networks and finite step target control, thereby effectively solving the problem of large spatial dimensions and high computational complexity of large complex networks.

[0034] 3. Furthermore, the power system drive node identification method provided by the present invention can accurately identify the set of drive nodes that can control the network within a given step, which can effectively improve the determinism of network controllability and increase the controllability and planning in practical applications. Attached Figure Description

[0035] Figure 1 This is a schematic diagram illustrating the use of an n-step greedy algorithm to determine set D according to Embodiment 1 of the present invention;

[0036] Figure 2 This is an unweighted directed graph with 7 nodes provided in Embodiment 1 of the present invention;

[0037] Figure 3 This is a schematic diagram illustrating the process of implementing the maximum matching algorithm on a power system network according to Embodiment 1 of the present invention;

[0038] Figure 4 This is a schematic diagram of the identification process of the driving node of the power system network when the given step size Δt = 0.1 and the number of steps n = 2, as provided in Embodiment 1 of the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0040] Example 1

[0041] A method for identifying driving nodes in a power system models the power system as a complex network system. The proposed problem of selecting driving nodes to reach an arbitrary target state in a finite-step internal energy control network is transformed into maximum matching of the complex network and finite-step target control. Specifically, the method includes the following steps:

[0042] S1. Initialize the candidate driver node set K'; determine matrix B based on set K'; where matrix B... This is used to represent the connection relationship between the driving node and all nodes; in this embodiment, N is the total number of nodes in the power system; M is the number of driving nodes.

[0043] Specifically, in this embodiment, a power system network is established using generators, loads, and buses in the power system as nodes, and transmission lines connecting each node as edges.

[0044] The dynamic formula for the power system network is as follows:

[0045]

[0046] Among them, matrix For the adjacency matrix of the power system network, Represents a set of real number matrices of size N×N; Let each node in the power system be a state vector. Represents a set of real number vectors of dimension N; The derivative of X is the rate of change of the node states in the power system; the matrix As part of matrix A, it is used to represent the connection relationship between the driving node and all nodes. The element in the i-th row and j-th column indicates whether the i-th driving node and the j-th node are connected, with 1 indicating connection and 0 indicating no connection. It is a vector composed of the control signals of each drive node in the power system.

[0047] Preferably, the method for initializing the candidate driver node set K' includes: adding the driver nodes obtained by solving the dynamic formula to the set K' to initialize the set K'. Specifically, the matrix B in the dynamic formula can be solved using algorithms such as maximum matching, heuristics, or logn approximation to obtain the driver nodes. It should be noted that, according to the characteristics of network control, if the network is controllable under a given step size, then the network is definitely controllable. However, the fact that the network is controllable does not directly lead to the conclusion that the network is controllable under a given step size. Therefore, some nodes in the driver node set found at this time must be a subset considering the step size requirement of the control problem. Thus, it is feasible to use the candidate driver node set K' at this time as the seed nodes for subsequent operations.

[0048] S2. Substitute the current matrix B into the expression F(Δt,n,A,B) of the coefficient matrix to obtain the current coefficient matrix; where the expression F(Δt,n,A,B) of the coefficient matrix is ​​obtained by discretizing the dynamic formula of the power system network; Δt is the discrete step length; n is the number of evolution steps in the discretization process; the matrix... is the adjacency matrix of the power system network; N is the total number of nodes in the power system; M is the number of driver nodes to be identified;

[0049] Specifically, the process of discretizing the above dynamic formula is as follows:

[0050]

[0051]

[0052]

[0053]

[0054] After simplification, we can obtain equation (3):

[0055]

[0056] Where Δt is the discrete step length; n is the number of evolution steps in the discrete process; X0 is the state vector at time 0; X n Let I be the state vector of the power system at time nΔt; I is an N×N identity matrix; U q It is a vector composed of the control signals of each driving node at time qΔt in the power system.

[0057] Observations show that when n and Δt are given, the first term of equation (3) is a constant and does not affect the controllability of the network. Next, equation (3) is written in matrix form as shown in equation (4):

[0058] Where const is For U q The vector formed is q = 0, 1, ..., n-1.

[0059] The expression obtained after representing the discretized formula in matrix form is as follows:

[0060]

[0061] Therefore, the condition for the power system network to be controllable is that the coefficient matrix F(Δt,n,A,B) has full row rank, which means that the controllability condition related to the step size Δt, the number of steps n, the adjacency matrix A, and the control input matrix B has full row rank. The controllability index is the controllability rank, which is the rank of the coefficient matrix R = Rank(F(Δt,n,A,B)). The larger R is, the more nodes the currently selected set of driving nodes can control.

[0062] Preferably, the coefficient matrix is: F(Δt,n,A,B)=[B,(I+AΔt) 1 B,...,(I+AΔt) n-1 B].

[0063] It should be noted that the coefficient matrix is ​​not limited to the above form; it can also be F(Δt,n,A,B)=[BΔt,(I+AΔt)]. 1 BΔt,...,(I+AΔt) n-1 BΔt], the effect is the same as the above formula.

[0064] This invention first discretizes the system dynamics, and based on the given step size Δt and number of steps n, it deduces the relationship between the system state after n steps and the initial state, and determines the controllable conditions related to the step size, number of steps and network structure, that is, the rank of the coefficient matrix, denoted as the controllable rank.

[0065] S3. If the rank R of the current coefficient matrix is ​​full rank, then all nodes in set K' are taken as the identified target driving nodes; otherwise, go to step S4.

[0066] S4. Calculate the rank R' of the coefficient matrix corresponding to each node removed from set K'; when R' corresponding to node P is equal to R, delete node P from set K', obtain the coefficient matrix corresponding to the current set K', perform linear independence analysis on the coefficient matrix, and mark the nodes with linear independence; specifically, in this embodiment, the nodes corresponding to the linearly independent rows of the coefficient matrix are marked (it should be noted that when the columns of the coefficient matrix B correspond to all nodes, the nodes corresponding to the linearly independent columns of the coefficient matrix are marked);

[0067] It should be noted that since the columns in matrix B correspond one-to-one with the elements in set K', when the nodes in set K' change, matrix B also changes accordingly to ensure that the columns in matrix B are the same as the columns of the nodes in set K'.

[0068] Specifically, the above process is as follows: For each node P in set K', calculate the controllable rank R' that makes the columns in matrix B retain only the columns corresponding to nodes other than node P, thus obtaining a set of controllable ranks R'; for the controllable ranks R' in the obtained controllable rank set, obtain the set S consisting of all nodes P corresponding to controllable ranks R' that are equal to R; remove all nodes in set S from set K' to obtain set K; obtain the coefficient matrix after removing the columns corresponding to all nodes in set S from matrix B, and mark the nodes corresponding to the linearly independent rows in the current coefficient matrix, that is, mark the nodes that can be controlled under set K:

[0069] In this process, it is verified whether the controllable rank changes after removing each node from set K'. Nodes whose controllable rank changes are retained in the driving set K', while nodes whose indices do not change are removed from set K'. At the same time, the columns corresponding to all nodes whose indices do not change are removed from matrix B. Then, based on the situation of linearly independent rows in the current coefficient matrix, the controllable nodes are marked, that is, the nodes corresponding to the largest linearly independent row set.

[0070] S5. Delete the nodes and their edges in set K' from the backup network of the power system network to obtain a new power system network. From the new power system network, obtain the nodes that can control the unmarked nodes to form a set D.

[0071] Specifically, the nodes and their edges in set K' are removed from the backup network of the power system network. This operation may disrupt the connectivity of the network. Therefore, for each connected subgraph in the remaining structure, the set D consisting of nodes that can control the unlabeled nodes is then searched.

[0072] Specifically, in one alternative implementation, an n-step greedy algorithm is used to determine the set D for each connected subgraph, as shown in the specific algorithm below. Figure 1 As shown (where solid nodes represent control targets, i.e., nodes in the set of nodes to be matched):

[0073] A1. Form a set of nodes to be matched from all unmarked nodes;

[0074] A2. Based on the maximum matching algorithm, all nodes in the set of nodes to be matched in the new power system network are matched to obtain the set of matched nodes. Then, all nodes in the new power system network that failed to match are merged with the set of matched nodes to reconstruct the set of nodes to be matched.

[0075] A3. Repeat step A2 for iteration until the number of iterations is n. At this point, the set of matching nodes is set D.

[0076] S6. Merge set D into set K' to update set K', update the coefficient matrix accordingly, and go to step S3.

[0077] Specifically, this invention iteratively optimizes the candidate driver node set K' based on controllable conditions. Through the aforementioned iterative process, a set of driver nodes capable of network control within a given number of steps is ultimately obtained. The driver node set is continuously expanded and optimized during the iteration process, and finally, the optimized set of nodes that meet the control conditions is output. This set is the set of driver nodes capable of controlling the network within a given step size.

[0078] It should be noted that a complex dynamic network system is considered completely controllable if, by applying control signals to a finite number of nodes (driving nodes), the network state can be driven from any initial state to any desired state within a finite time. As a ubiquitous form of connection in both natural and artificial environments, the system parameters of many complex networks cannot be precisely known; we only know whether there are links between individuals. This has given rise to the concept of structural controllability. By definition, if it is possible to choose non-zero weights to make the system controllable, then the system is structurally controllable. Therefore, structural controllability provides a method to overcome inherently incomplete weight information.

[0079] A power system can be viewed as a large, complex network with a set of nodes and edges. For a power system, its generators, loads, and buses can be modeled as nodes, and the connected transmission lines as lines or edges. Depending on the current in the grid, the power system can be modeled as a directed or undirected network. Edge weights are typically defined as reactance to convert the shortest path into the shortest electrical distance. In addition, admittance, which is proportional to the amount of electricity flowing through any line in the transmission network, can be used. However, these values ​​are not readily available; therefore, the aforementioned controllability can be well applied to power systems. Furthermore, based on Kirchhoff's laws, the nodal voltage method is a good approach for analyzing large power systems and can be used to calculate the current in the power system. The direction of the current is fixed under any steady state. Therefore, suppose G = (V, E) is a power system with n nodes and k lines, where V is the set of nodes and E is the adjacency matrix defining network connectivity. If there is current from node i to node j, then eji is 1; otherwise, eji is 0. The resulting adjacency matrix can then be used to calculate the network's controllability.

[0080] To achieve specific control at a defined time step, this invention proposes a driver node identification problem for complex networks with fixed-step controllability, namely, how to find a set of driver nodes that can achieve the control objective of the network at a fixed time step. For large-scale complex networks, this problem is an NP-hard combinatorial optimization problem, and exhaustive search is often impractical. Therefore, this invention proposes a driver node identification method based on maximum matching and finite-step greedy search. This method can effectively handle large complex networks, overcoming the difficulties of excessively high problem space dimensionality and unacceptable computational complexity.

[0081] To further illustrate the power system drive node identification method provided by the present invention, a specific embodiment is described in detail below:

[0082] This specific implementation method models a sparse directed graph with 7 nodes for the power system network structure, such as... Figure 2 As shown.

[0083] Step 1: Construct a linear time-invariant system for the directed graph as shown in formula (1), where A is the adjacency matrix of the network and B is the matrix to be determined. According to the actual graph structure, we know that:

[0084]

[0085] Step 2: Given a step size Δt = 0.1 and a number of steps n = 2, calculate the explicit expression for the controllable rank R = Rank(F(Δt,n,A,B)), which is used to verify the node recognition results. The result is:

[0086]

[0087] Step 3: As Figure 3 As shown, the maximum matching algorithm is used to process the system, and the set of driving nodes C = {1, 3, 6} of the network without considering the step requirement is obtained. The set of candidate driving nodes K' = C = {1, 3, 6} of the original problem is obtained, and the current controllable rank R = 5 is calculated as the reference controllable rank.

[0088] Step 4: Optimize the candidate driver node set K' by reducing its size based on the controllability conditions (step S5 above): Specifically, the controllability rank R' when node 1 is removed from set K' is 3; the controllability rank R' when node 3 is removed from set K' is 3; the controllability rank R' when node 6 is removed from set K' is 4; compared with the reference controllability rank, all of the above controllability ranks have changed, that is, the controllability indices have changed; the nodes with changed indices are retained in the driver set K', so the optimized set K' is {1,3,6}; remove the columns corresponding to all nodes whose indices have not changed from matrix B. Based on the current linearly independent rows of the coefficient matrix, the current set of uncontrolled nodes is {5,7}.

[0089] Step 5: In the backup network of the original power system network, delete the nodes and their edges in set K'. The network structure then becomes two parts, as follows: Figure 4 As shown in the process, the uncontrolled nodes are distributed in two subgraphs. A greedy search of n steps (2 steps) is performed on each graph to locate the candidate driver node set D = {4,7}.

[0090] Step 6: Determine the candidate driver node set K' = D∪K' = {1,3,4,6,7} of the original problem, and substitute it into the control index expression. At this time, the controllable rank R = 7, which is used as the new reference controllable rank.

[0091] Step 7: Optimize the candidate driver node set K' based on the controllable rank. When node 4 in set C is removed from set K', its corresponding controllable rank R' = 6, which changes compared to the new reference controllable rank, but it remains in set K'. When node 7 in set C is removed from set K', its corresponding controllable rank R' = 7, which remains unchanged compared to the new reference controllable rank, and it cannot be used as a driver node, so it is removed from set K'. The final optimized K' = {1,3,4,6}. According to the controllable condition, applying control to the selected node set at this point can achieve control of the entire network under given conditions, i.e., the currently uncontrolled node set D = {}. At this point, the candidate driver node set K' = D∪K' = {1,3,4,6}. Since the rank is now full, the solution process ends, and the candidate driver node set K' is output, i.e., the identified driver nodes are nodes 1, 3, 4, and 6. Apply control signals to these 4 nodes. According to equation (2), when n=1, the control signal can control nodes {1,3,4,6}. When n=2, the control signal can control node 2 through node 1, node 7 through node 3, and node 5 through node 4. At this time, the controlled set is {1,2,3,4,5,6,7}, realizing two-step control of the network.

[0092] In summary, this invention relates to a method for identifying drive nodes in power systems based on the controllability of complex networks, belonging to the technical field of complex network controllability. This method considers practical application scenarios and can accurately find the set of drive nodes under given conditions, effectively solving large-scale complex NP-hard networks and overcoming the difficulties of excessively high problem space dimensionality and computational complexity.

[0093] Example 2

[0094] A power system drive node identification system includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the power system drive node identification method provided in Embodiment 1 of the present invention.

[0095] The relevant technical solutions are the same as in Embodiment 1, and will not be repeated here.

[0096] Example 3

[0097] A power system control method includes: identifying the driving nodes in the power system using the power system driving node identification method provided in Embodiment 1 of the present invention, and controlling the identified driving nodes.

[0098] Specifically, a suitable control law can be designed based on the required control objective (e.g., to ensure that the power of all node states reaches a certain level or that the node voltage phases are consistent) – based on the state X after the i-th step evolution. i Generate the required control input signal U i , will U i The effects are applied to the drive nodes and transmitted through the network structure, ultimately enabling the control of all nodes. For example, when a power system fault occurs, the current, voltage, and other parameters of the identified drive nodes can be controlled specifically, allowing the system to quickly return to normal operation.

[0099] The relevant technical solutions are the same as in Embodiment 1, and will not be repeated here.

[0100] Example 4

[0101] A power system includes: a drive node identification module, used to execute the identification method provided in Embodiment 1 of the present invention to obtain the drive nodes of the power system.

[0102] The relevant technical solutions are the same as in Embodiment 1, and will not be repeated here.

[0103] Example 5

[0104] A computer-readable storage medium includes a stored computer program, wherein when the computer program is run by a processor, it controls the device where the storage medium is located to execute the power system drive node identification method provided in Embodiment 1 of the present invention.

[0105] The relevant technical solutions are the same as in Embodiment 1, and will not be repeated here.

[0106] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying drive nodes in a power system, characterized in that, Includes the following steps: S1. Initialize the candidate driver node set K'; determine the matrix based on set K'. The matrix Used to represent the connection relationship between the driving node and all nodes; S2, the matrix The expression for substituting the coefficient matrix In the expression, obtain the current coefficient matrix; where, It is obtained by discretizing the dynamic formula of the power system network; For a long walk; The number of evolution steps in the discrete process; matrix This is the adjacency matrix of the power system network; S3. If the rank R of the current coefficient matrix is ​​full rank, then all nodes in set K' are taken as the identified target driving nodes; otherwise, go to step S4. S4. Calculate the rank R' of the coefficient matrix when each node is removed from set K'; when the node... P When the corresponding R' is equal to R, delete the node from set K'. P Obtain the coefficient matrix corresponding to the current set K', perform linear independence analysis on the coefficient matrix, and mark the nodes that have linear independence relationships; S5. Delete the nodes and their edges in set K' from the backup network of the power system network to obtain a new power system network. Obtain the nodes that can control the unmarked nodes from the new power system network to form a set D. S6. Merge the set D into the set K' to update the set K', update the coefficient matrix accordingly, and go to step S3; The dynamic formula of the power system network is: ； in, For each node in the power system, there is a state vector. for The derivative; It is a vector composed of the control signals of each drive node in the power system; The discretized expression of the dynamic formula is: in, For the power system n The state vector at time t; It is the identity matrix; The state vector at time 0; For the first power system q The vector formed by the control signals of each driving node at a given time. After discretizing the dynamic formula, it is represented in matrix form as follows: for The vector formed q =0,1,..., n -1.

2. The identification method according to claim 1, characterized in that, The coefficient matrix = .

3. The identification method according to claim 1, characterized in that, The method for initializing the candidate driver node set K' includes: adding the driver nodes obtained by solving the dynamic formula to the set K' to initialize the set K'.

4. The identification method according to any one of claims 1-3, characterized in that, The methods for obtaining the set D include: A1. Form a set of nodes to be matched from all unmarked nodes; A2. Based on the maximum matching algorithm, all nodes in the set of nodes to be matched in the new power system network are matched to obtain the matching node set. All nodes in the new power system network that failed to match are merged with the matching node set to reconstruct the set of nodes to be matched. A3. Repeat step A2 iteratively until the number of iterations is reached. n The current set of matching nodes is the set D.

5. A power system drive node identification system, characterized in that, include: A memory and a processor, wherein the memory stores a computer program, and the processor executes the identification method according to any one of claims 1-4 when executing the computer program.

6. A control method for a power system, characterized in that, include: The identification method described in any one of claims 1-4 is used to identify the drive nodes in the power system and to control the identified drive nodes.

7. An electric power system, characterized in that, include: A drive node identification module is used to execute the identification method according to any one of claims 1-4 to obtain the drive nodes of the power system.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed by a processor, it controls the device in which the storage medium is located to perform the identification method according to any one of claims 1-4 and / or the control method according to claim 6.

Images