Equivalent model generation method and system based on direct-current double-end station coupling

Abstract

The invention provides an equivalent model generation method and system based on direct-current double-end station coupling. The method comprises the following steps: acquiring a target direct-current line and a corresponding direct-current double-end station; constructing a whole network unweighted graph model and a main network weighted graph model according to the original power grid data file; searching the shortest path, except the target direct current line, between the direct current double-end stations from the main network weighted graph model, so as to construct a first node set; performing transient simulation on the power grid, and screening and determining a plurality of main power grid nodes to construct a second node set according to the voltage change of each main power grid node at the fault moment and a preset condition; dividing each power grid node into an external region node set and an internal region node set according to the first node set and the second node set; the power grid characteristics of the external region are fitted at the boundary nodes between the external region node set and the internal region node set, the corresponding equivalent model is generated, and the simulation response speed and the simulation accuracy of the equivalent model are improved.

Description

Technical Field

[0001] The present application relates to the technical field of power system simulation, and in particular to a method and system for generating an equivalent model based on DC double-terminal station coupling. Background Art

[0002] High-voltage direct current (HVDC) transmission technology has become a highly efficient transmission technology and is widely used in my country's power grid. Together with the AC transmission system, it forms a hybrid AC / DC transmission system. Because the two-terminal stations of the HVDC system are coupled through the AC system in this hybrid AC / DC transmission mode, the DC system stability mechanism exhibits different characteristics from those of a pure DC transmission channel. Therefore, specific research is needed on the HVDC system in this hybrid AC / DC system. However, when conducting stability analysis of an HVDC system under a specific operating mode based on grid data files, simulating and modeling the entire network is time-consuming and laborious, as the grid data files contain the main stations of the entire network. Furthermore, most stations have little impact on the HVDC system. To achieve efficient stability analysis of HVDC systems, nodes with minimal impact on the transient characteristics of the HVDC system can be simplified and equivalently represented, forming an equivalent DC system model for the hybrid AC / DC system.

[0003] Currently, dynamic equivalent modeling techniques for DC systems in AC / DC hybrid systems can be broadly divided into two main approaches. The first is the hierarchical network simplification approach, which retains the DC converter station, 500 kV and above backbone network nodes, and adjacent high-capacity power generation nodes as the core. This approach expands outward through a tree-like divergent structure to form an internal network, while simplifying the remaining nodes. This approach simplifies the entire network topology using Ward equivalents or coherent unit aggregation algorithms (e.g., using residual voltage thresholds to demarcate boundaries). While this approach maintains the interactive characteristics of the AC / DC system, the number of retained nodes is still large, resulting in high computational complexity. The second is the infinite power source equivalent approach, which simplifies the AC system on both sides of the DC link into an infinite power source with controllable voltage amplitude and phase, and then combines parameter adjustments to fit the transient response of the original system. While this approach significantly improves simulation efficiency (e.g., the subsystem partitioning strategy in RTDS real-time simulation), it breaks the dynamic coupling between the converter station and the AC system (e.g., the correlation between reactive power support capacity and voltage fluctuations), leading to biased voltage stability assessments in scenarios with high renewable energy penetration.

[0004] In summary, although the hierarchical network simplification method reduces network complexity through topological screening and coherent aggregation, its equivalent model still retains too many nodes, resulting in excessive computing power requirements for simulation in electromagnetic transient environments, making it difficult to meet the real-time simulation requirements of AC / DC transmission systems; and although the infinite power equivalent method improves computational efficiency through virtual power parameter fitting, it dynamically decouples the converter station from the AC system, making the equivalent system unable to accurately represent the complex transient characteristics in the AC / DC coupling scenario. Summary of the Invention

[0005] In response to the above technical problems, the present application provides a method and system for generating an equivalent model based on DC two-terminal site coupling to improve the simulation response speed and simulation accuracy of the equivalent model of the AC / DC system.

[0006] In a first aspect, an embodiment of the present application provides a method for generating an equivalent model based on DC two-terminal site coupling, comprising:

[0007] Acquire a target DC line and a DC double-terminal station of the target DC line;

[0008] Constructing an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes;

[0009] Searching for the shortest path between the DC double-terminal stations excluding the target DC line from the main grid weighted graph model, extracting the main grid nodes on the shortest path to construct a first node set;

[0010] performing a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determining a number of main power grid nodes from each main power grid node to form a second node set based on voltage changes of each main power grid node at the moment of the fault and preset conditions;

[0011] Merging the first node set and the second node set into a core node set, and dividing each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, wherein the internal area node set includes all main power grid nodes in the core node set;

[0012] The grid characteristics of the external area are fitted at the boundary nodes between the external area node set and the internal area node set, and the original grid data file is modified accordingly, and then an equivalent model of the target DC line is generated according to the modified original grid data file.

[0013] The embodiment of the present application provides an equivalent model generation method based on DC double-terminal site coupling, which uses the shortest path algorithm to extract the first node set from the main network weighted graph model, uses transient simulation to screen out the second node set from each main grid node, and then constructs a core node set. Finally, based on the core node set and the unweighted graph model of the entire network, internal and external regions are divided and external equivalents are performed to generate the corresponding equivalent model. Among them, the extraction of nodes by the shortest path algorithm can effectively retain the DC site through the AC system coupling channel, ensuring that the subsequent equivalent model accurately reflects the dynamic interaction characteristics of the AC / DC system. At the same time, since the main network weighted graph model used has already screened each grid node according to the preset power threshold, it effectively reduces the node scale in the subsequent generated equivalent model and improves the computational efficiency of the equivalent model. By screening the voltage fluctuation sensitive nodes through transient simulation, the node scale to be simulated can be significantly reduced, the computational complexity can be reduced, and the decoupling error of the traditional infinite power equivalent method can be avoided. The core node set constructed by combining the two makes the equivalent AC / DC system greatly reduce the scale while retaining the coupling relationship between the large power grid and the DC system, improving the simulation response speed and simulation accuracy of the AC / DC system equivalent model.

[0014] Furthermore, the construction of the network-wide unweighted graph model and the main network weighted graph model based on the original grid data file and the preset power threshold includes:

[0015] Constructing the unweighted graph model of the entire network according to the connection relationship between each grid node in the original grid data file;

[0016] Determine a number of main grid nodes whose power is greater than the preset power threshold from the original grid data file, and construct an initial main grid weighted graph model based on the connection relationship between each of the main grid nodes and each of the main grid nodes;

[0017] Traversing each connection edge in the initial main network weighted graph model, determining the weight value of the connection edge according to the power grid line type corresponding to the connection edge, and then obtaining the main network weighted graph model;

[0018] Among them, if the grid line type corresponding to the connection edge is an AC line, the weight value of the connection edge is the inverse of the equivalent impedance of the AC line; if the grid line type corresponding to the connection edge is a DC line, the weight value of the connection edge is 1.

[0019] The embodiment of the present application provides a method for constructing an unweighted graph model of the entire network and a weighted graph model of the main network. A number of main grid nodes are screened out according to a preset power threshold, and the edge weights of the main grid weighted graph are dynamically set according to the line type, thereby generating a weighted graph model of the main network. This helps to retain the characteristics of the DC site coupled through the AC system in the subsequent equivalent model, while minimizing the number of retained nodes, achieving dual optimization of computational efficiency and simulation accuracy. In addition, the embodiment of the present application sets the AC line weight to the inverse of the equivalent impedance and the DC line weight to 1, which is more in line with the physical characteristics of the actual power grid, improves the simulation accuracy of the main grid weighted graph model for the actual power distribution, and further improves the simulation accuracy of the AC / DC system equivalent model.

[0020] In one possible implementation, searching the main grid weighted graph model for the shortest path between the DC double-terminal stations excluding the target DC line, and extracting the main grid nodes on the shortest path to construct a first node set includes:

[0021] Removing the connection edges corresponding to the target DC line from the main network weighted graph model to obtain a temporary weighted graph;

[0022] Based on the Dijkstra algorithm and the weight values ​​of each connection edge in the temporary weighted graph, searching and determining the shortest path between the DC double-terminal stations in the temporary weighted graph;

[0023] Main grid nodes on the shortest path are extracted to construct a first node set.

[0024] An embodiment of the present application provides a method for constructing a first node set. After removing the target DC line, the Dijkstra algorithm is used to search for the shortest path between two-terminal DC stations. The grid nodes on the shortest path are extracted to construct the first node set. This method effectively determines the main channel representing the coupling of the two DC stations through the AC system and ignores other irrelevant main grid nodes. This ensures that the subsequently established equivalent model accurately reflects the dynamic interaction characteristics of the AC / DC system, while reducing the computing power required for simulating the equivalent model of the AC / DC system. In addition, the Dijkstra algorithm can efficiently ensure the rapid determination of the shortest path in complex networks and increase the speed of generating equivalent models.

[0025] In one possible implementation, performing transient simulation on the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determining a plurality of main grid nodes from each main grid node to construct a second node set based on voltage changes of each main grid node at the moment of a fault and preset conditions, includes:

[0026] Construct a blank initial second node set;

[0027] According to the original power grid data file, in a power grid electromechanical transient simulation environment, a transient simulation of the power grid is performed by setting a single-ended three-phase short-circuit fault at the DC double-terminal station, and obtaining a voltage change value of each main power grid node at the moment of the fault;

[0028] Each of the voltage change values ​​is compared with a preset threshold value. If the voltage change value is greater than or equal to the preset threshold value, the main grid node corresponding to the voltage change value is added to the initial second node set. After completing all threshold comparisons, the second node set is obtained.

[0029] An embodiment of the present application provides a method for constructing a second node set. By using transient simulation, nodes with significant voltage changes at the moment of fault are screened from all main power grid nodes to construct the second node set. By screening voltage-sensitive nodes, dynamic response distortion caused by static topology simplification can be effectively avoided. While reasonably simplifying the model, the accuracy of the equivalent model in the fault crossing scenario is improved.

[0030] In one possible implementation, dividing each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model includes:

[0031] Constructing a blank initial outer region node set and an initial inner region node set containing the core node set;

[0032] Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph;

[0033] Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory;

[0034] Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set;

[0035] After the traversal is completed, the external area node set and the internal area node set are obtained.

[0036] The embodiment of the present application provides a method for dividing the system into an external area node set and an internal area node set. After determining the core node set, it is also necessary to screen some internal nodes from the remaining power grid nodes to further improve the robustness of the equivalent model. The embodiment of the present application divides the temporary unweighted graph into several connected components based on the connected component theory, and divides the internal and external area nodes by traversing the connected components, so that the system can efficiently process large-scale power grid topology during the area division process, reducing the computational complexity of the area division. In the specific division process, the embodiment of the present application divides the external area by judging whether there is a main power grid node, ensuring that the equivalence of the external network does not affect the internal dynamic characteristics, and further improving the simulation accuracy of the equivalent model of the AC / DC system.

[0037] In one possible implementation, fitting the grid characteristics of the external region at the boundary nodes between the external region node set and the internal region node set and modifying the original grid data file accordingly, and then generating the equivalent model of the target DC link according to the modified original grid data file, includes:

[0038] Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set;

[0039] Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model;

[0040] Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram;

[0041] Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file;

[0042] Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model;

[0043] The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link.

[0044] The embodiment of the present application provides a method for generating an equivalent model, which determines a number of boundary nodes, a number of internal and external connection edges, an internal area node set, and a number of internal connection edges based on an unweighted graph model of the entire network, an external area node set, and an internal area node set, constructs a target area power grid node line diagram, and modifies the original power grid data file based on the target area power grid node line diagram to obtain a target power grid data file that only retains the internal area power grid data. Finally, a preset initial generator-load model is connected at each boundary node and the system curve of the original power grid is fitted to achieve effective segmentation of the target area power grid and the external power grid, and the external power grid characteristics are fitted at the boundary nodes to generate the corresponding equivalent model. The embodiment of the present application accurately simulates the power interaction of the external area with the core network by dynamically adjusting the generator-load model parameters, avoids the static deviation of the traditional Ward equivalent, and effectively improves the simulation accuracy of the AC / DC system equivalent model.

[0045] Furthermore, the equivalent model generation method further includes simplifying the internal power grid nodes of the equivalent model to obtain a simplified equivalent model, specifically:

[0046] Get the pre-reserved node entered by the user;

[0047] Classifying each internal power grid node of the equivalent model as a reserved node or a node to be deleted according to a preset screening condition and the pre-reserved node;

[0048] Sorting each of the internal power grid nodes according to their types, and establishing a block node admittance matrix according to the sorting results;

[0049] The block node admittance matrix is ​​simplified into an equivalent admittance matrix using the Kron simplification method;

[0050] The equivalent model is simplified according to the equivalent admittance matrix and the type of each internal power grid node to obtain the simplified equivalent model.

[0051] This embodiment of the present application provides an internal simplification method for equivalent models. Based on preset screening criteria and user-specified pre-reserved nodes, each internal power grid node in the equivalent model is classified as either retained or to be deleted. This method, combined with block admittance matrix sorting, prioritizes simplification of non-critical areas, further optimizing computational resource allocation. The Kron simplification method then compresses the block structure of the large-scale admittance matrix into a low-dimensional equivalent admittance matrix through mathematical transformation. This method directly reduces the computational complexity of the equivalent model in electromagnetic transient simulations and improves the simulation efficiency of the equivalent model.

[0052] In a second aspect, an embodiment of the present application provides an equivalent model generation system based on DC two-terminal site coupling, comprising an acquisition module, a graph construction module, a first node set construction module, a second node set construction module, a region division module, and an equivalent model generation module;

[0053] The acquisition module is used to acquire a target DC line and a DC double-terminal station of the target DC line;

[0054] The graph construction module is used to construct an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes;

[0055] The first node set construction module is used to search the shortest path between the DC double-terminal stations excluding the target DC line from the main grid weighted graph model, and extract the main grid nodes on the shortest path to construct a first node set;

[0056] The second node set construction module is configured to perform a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determine a number of main power grid nodes from each main power grid node to construct a second node set based on the voltage change of each main power grid node at the moment of the fault and preset conditions;

[0057] The area division module is used to merge the first node set and the second node set into a core node set, and divide each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, wherein the internal area node set includes all main power grid nodes in the core node set;

[0058] The equivalent model generation module is used to fit the power grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modify the original power grid data file accordingly, and then generate the equivalent model of the target DC line according to the modified original power grid data file.

[0059] Furthermore, the area division module divides each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, including:

[0060] Constructing a blank initial outer region node set and an initial inner region node set containing the core node set;

[0061] Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph;

[0062] Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory;

[0063] Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set;

[0064] After the traversal is completed, the external area node set and the internal area node set are obtained.

[0065] Furthermore, the equivalent model generation module fits the grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modifies the original grid data file accordingly, and then generates the equivalent model of the target DC link according to the modified original grid data file, including:

[0066] Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set;

[0067] Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model;

[0068] Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram;

[0069] Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file;

[0070] Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model;

[0071] The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 A flow chart of a method for generating an equivalent model based on DC two-terminal site coupling provided in an embodiment of the present application;

[0073] Figure 2 A schematic diagram of a process for constructing a first node set in a method for generating an equivalent model based on DC two-terminal site coupling provided in an embodiment of the present application;

[0074] Figure 3 A schematic structural diagram of an equivalent generator-comprehensive load model connected to an equivalent model generation method based on DC two-terminal station coupling provided in an embodiment of the present application;

[0075] Figure 4 A flowchart of another method for generating an equivalent model based on DC two-terminal site coupling provided in an embodiment of the present application;

[0076] Figure 5 A schematic diagram showing a simulation comparison between an equivalent model generated by the equivalent model generation method provided by an embodiment of the present application and an original system under a fault condition;

[0077] Figure 6 A schematic diagram showing a simulation comparison between an equivalent model generated by the equivalent model generation method provided in an embodiment of the present application and an original system under another fault condition;

[0078] Figure 7 A structural diagram of an equivalent model generation system based on DC two-terminal site coupling provided in an embodiment of the present application. DETAILED DESCRIPTION

[0079] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0080] It should be noted that the step numbers herein are for convenience of explanation of the specific embodiments and do not serve to define the order in which the steps are to be performed. In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature designated "first" or "second" may explicitly or implicitly include one or more of such features.

[0081] Example 1:

[0082] like Figure 1As shown, embodiment 1 provides a method for generating an equivalent model based on DC two-terminal site coupling, including steps S1-S6:

[0083] Step S1: obtaining a target DC line and a DC double-terminal station of the target DC line;

[0084] Step S2: constructing an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes;

[0085] Step S3: searching the main grid weighted graph model for the shortest path between the DC double-terminal stations excluding the target DC line, and extracting the main grid nodes on the shortest path to construct a first node set;

[0086] Step S4: performing a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determining a number of main grid nodes from each main grid node to form a second node set based on voltage changes of each main grid node at the moment of the fault and preset conditions;

[0087] Step S5: merging the first node set and the second node set into a core node set, and dividing each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, wherein the internal area node set includes all main power grid nodes in the core node set;

[0088] Step S6: Fitting the grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modifying the original grid data file accordingly, and then generating an equivalent model of the target DC link according to the modified original grid data file.

[0089] The embodiment of the present application provides an equivalent model generation method based on DC double-terminal site coupling, which uses the shortest path algorithm to extract the first node set from the main network weighted graph model, uses transient simulation to screen out the second node set from each main grid node, and then constructs a core node set. Finally, based on the core node set and the unweighted graph model of the entire network, internal and external regions are divided and external equivalents are performed to generate the corresponding equivalent model. Among them, the extraction of nodes by the shortest path algorithm can effectively retain the DC site through the AC system coupling channel, ensuring that the subsequent equivalent model accurately reflects the dynamic interaction characteristics of the AC / DC system. At the same time, since the main network weighted graph model used has already screened each grid node according to the preset power threshold, it effectively reduces the node scale in the subsequent generated equivalent model and improves the computational efficiency of the equivalent model. By screening the voltage fluctuation sensitive nodes through transient simulation, the node scale to be simulated can be significantly reduced, the computational complexity can be reduced, and the decoupling error of the traditional infinite power equivalent method can be avoided. The core node set constructed by combining the two makes the equivalent AC / DC system greatly reduce the scale while retaining the coupling relationship between the large power grid and the DC system, improving the simulation response speed and simulation accuracy of the AC / DC system equivalent model.

[0090] Furthermore, in step S2, the whole network unweighted graph model and the main network weighted graph model are constructed according to the original grid data file and the preset power threshold, including:

[0091] Constructing the unweighted graph model of the entire network according to the connection relationship between each grid node in the original grid data file;

[0092] Determine a number of main grid nodes whose power is greater than the preset power threshold from the original grid data file, and construct an initial main grid weighted graph model based on the connection relationship between each of the main grid nodes and each of the main grid nodes;

[0093] Traversing each connection edge in the initial main network weighted graph model, determining the weight value of the connection edge according to the power grid line type corresponding to the connection edge, and then obtaining the main network weighted graph model;

[0094] Among them, if the grid line type corresponding to the connection edge is an AC line, the weight value of the connection edge is the inverse of the equivalent impedance of the AC line; if the grid line type corresponding to the connection edge is a DC line, the weight value of the connection edge is 1.

[0095] The embodiment of the present application provides a method for constructing an unweighted graph model of the entire network and a weighted graph model of the main network. A number of main grid nodes are screened out according to a preset power threshold, and the edge weights of the main grid weighted graph are dynamically set according to the line type, thereby generating a weighted graph model of the main network. This helps to retain the characteristics of the DC site coupled through the AC system in the subsequent equivalent model, while minimizing the number of retained nodes, achieving dual optimization of computational efficiency and simulation accuracy. In addition, the embodiment of the present application sets the AC line weight to the inverse of the equivalent impedance and the DC line weight to 1, which is more in line with the physical characteristics of the actual power grid, improves the simulation accuracy of the main grid weighted graph model for the actual power distribution, and further improves the simulation accuracy of the AC / DC system equivalent model.

[0096] In a preferred embodiment, step S2 constructs a weighted graph model of the main power grid to be analyzed and an unweighted graph model of the entire network, specifically:

[0097] Read the original power grid data file and define the topology model of the entire network as an unweighted undirected graph G = (V, E), where the vertex set V = {v i} represents the power grid site (including AC bus, converter station, substation and power generation station), edge set E = {e ij} represents node v i 、v j Electrical connection relationship (including lines, transformers, DC lines, etc.). Construct the backbone network of AC stations and DC system stations with voltage levels of 500kV and above. main =(V main ,E main ), and the weights of each edge are set as follows:

[0098]

[0099] In one possible implementation, in step S3, searching the main grid weighted graph model for the shortest path between the DC double-terminal stations excluding the target DC line, and extracting the main grid nodes on the shortest path to construct a first node set includes:

[0100] Removing the connection edges corresponding to the target DC line from the main network weighted graph model to obtain a temporary weighted graph;

[0101] Based on the Dijkstra algorithm and the weight values ​​of each connection edge in the temporary weighted graph, searching and determining the shortest path between the DC double-terminal stations in the temporary weighted graph;

[0102] Main grid nodes on the shortest path are extracted to construct a first node set.

[0103] An embodiment of the present application provides a method for constructing a first node set. After removing the target DC line, the Dijkstra algorithm is used to search for the shortest path between two-terminal DC stations. The grid nodes on the shortest path are extracted to construct the first node set. This method effectively determines the main channel representing the coupling of the two DC stations through the AC system and ignores other irrelevant main grid nodes. This ensures that the subsequently established equivalent model accurately reflects the dynamic interaction characteristics of the AC / DC system, while reducing the computing power required for simulating the equivalent model of the AC / DC system. In addition, the Dijkstra algorithm can efficiently ensure the rapid determination of the shortest path in complex networks and increase the speed of generating equivalent models.

[0104] In a preferred embodiment, Figure 2 As shown, the DC station is retained through the AC system coupling path node through step S3, specifically:

[0105] Determine the double-ended converter station node set of the target DC line Build source-target node pairs (v s ,v t ), where v s =v dc1 、vt=v dc2 Remove the mainnet entitled graph G main The edge e corresponding to the DC line to be studied dc ∈E main , forming the main network graph G′ main ={V main ,E main \{e dc},W}. Then search G′ based on Dijkstra algorithm main (v s ,v t ), its cost function is:

[0106]

[0107] Where P s→i Indicates that from v s to v i The cumulative cost of the node is updated iteratively through the priority queue Q:

[0108] 1. Initialization:

[0109] 2. Loop: Select the minimum cost u in Q

[0110] 3. Update adjacent nodes:

[0111] Finally, the optimal path P is obtained opt ={v s ,v1,v2,...,vt}, which satisfies:

[0112]

[0113] The path node set V path ={vk|vk∈Popt} is included in the first node set B1, representing the main channel for coupling the two DC stations through the AC system.

[0114] In one possible implementation, in step S4, performing transient simulation on the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determining a plurality of main grid nodes from each main grid node to construct a second node set based on voltage changes of each main grid node at the moment of a fault and preset conditions, including:

[0115] Construct a blank initial second node set;

[0116] According to the original power grid data file, in a power grid electromechanical transient simulation environment, a transient simulation of the power grid is performed by setting a single-ended three-phase short-circuit fault at the DC double-terminal station, and obtaining a voltage change value of each main power grid node at the moment of the fault;

[0117] Each of the voltage change values ​​is compared with a preset threshold value. If the voltage change value is greater than or equal to the preset threshold value, the main grid node corresponding to the voltage change value is added to the initial second node set. After completing all threshold comparisons, the second node set is obtained.

[0118] An embodiment of the present application provides a method for constructing a second node set. By using transient simulation, nodes with significant voltage changes at the moment of fault are screened from all main power grid nodes to construct the second node set. By screening voltage-sensitive nodes, dynamic response distortion caused by static topology simplification can be effectively avoided. While reasonably simplifying the model, the accuracy of the equivalent model in the fault crossing scenario is improved.

[0119] In a preferred embodiment, the nodes with the largest electrical coupling strength with the DC station are retained in step S4, specifically:

[0120] In the grid data file under the grid electromechanical transient simulation environment, set a single-ended three-phase short circuit fault at time t0 at both ends of the DC, and perform transient simulation, recording each fault instant t=t 0+ Voltage of each 500kV node at the moment The dynamic response sensitivity is calculated based on the voltage change value of each node, and the dynamic response sensitivity is compared with the preset sensitivity threshold, and the grid nodes are screened based on the comparison results.

[0121] The dynamic response sensitivity criterion is defined here:

[0122]

[0123] in is the normal operating voltage of node i, U b The sensitivity threshold η is set to 0.3, and the nodes that meet the criteria are added to the second node set B2.

[0124] Merge B1 and B2 into core node set B.

[0125] In one possible implementation, in step S5, dividing each grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model includes:

[0126] Constructing a blank initial outer region node set and an initial inner region node set containing the core node set;

[0127] Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph;

[0128] Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory;

[0129] Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set;

[0130] After the traversal is completed, the external area node set and the internal area node set are obtained.

[0131] The embodiment of the present application provides a method for dividing the system into an external area node set and an internal area node set. After determining the core node set, it is also necessary to screen some internal nodes from the remaining power grid nodes to further improve the robustness of the equivalent model. The embodiment of the present application divides the temporary unweighted graph into several connected components based on the connected component theory, and divides the internal and external area nodes by traversing the connected components, so that the system can efficiently process large-scale power grid topology during the area division process, reducing the computational complexity of the area division. In the specific division process, the embodiment of the present application divides the external area by judging whether there is a main power grid node, ensuring that the equivalence of the external network does not affect the internal dynamic characteristics, and further improving the simulation accuracy of the equivalent model of the AC / DC system.

[0132] In one possible implementation, fitting the grid characteristics of the external region at the boundary nodes between the external region node set and the internal region node set and modifying the original grid data file accordingly, and then generating the equivalent model of the target DC link according to the modified original grid data file, includes:

[0133] Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set;

[0134] Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model;

[0135] Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram;

[0136] Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file;

[0137] Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model;

[0138] The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link.

[0139] The embodiment of the present application provides a method for generating an equivalent model, which determines a number of boundary nodes, a number of internal and external connection edges, an internal area node set, and a number of internal connection edges based on an unweighted graph model of the entire network, an external area node set, and an internal area node set, constructs a target area power grid node line diagram, and modifies the original power grid data file based on the target area power grid node line diagram to obtain a target power grid data file that only retains the internal area power grid data. Finally, a preset initial generator-load model is connected at each boundary node and the system curve of the original power grid is fitted to achieve effective segmentation of the target area power grid and the external power grid, and the external power grid characteristics are fitted at the boundary nodes to generate the corresponding equivalent model. The embodiment of the present application accurately simulates the power interaction of the external area with the core network by dynamically adjusting the generator-load model parameters, avoids the static deviation of the traditional Ward equivalent, and effectively improves the simulation accuracy of the AC / DC system equivalent model.

[0140] In a preferred embodiment, the external network dynamic equivalence is performed based on the selected core node set B, and then the equivalence model is generated, specifically:

[0141] 1. First, remove the nodes in node set B from graph G and segment the main network topology using graph connected component theory:

[0142] G′=(V\B,E′)where E′={e ij =(v i ,v j )|v i ,v j ∈V\B}

[0143] Divide the graph into multiple connected components according to its connectivity

[0144] 2. Traverse each subnet G k , determine each subnet G k Is there a 500kV node? If so, the subnet node is divided into the external area A1, otherwise it is divided into A2 together with the nodes in B. And the communication channel E between the two areas is extracted. edge and boundary nodes V edge :

[0145] E edge ={e ij |v i ∈A1,v j ∈A2}

[0146]

[0147] The node line set of the regional power grid system is constructed based on this

[0148] G reserch =(V A2 ∪V edge ,E A2 ∪E edge )

[0149] 3. Traverse the power grid data file and remove the reserch There are no corresponding nodes and edges for electrical components such as lines and transformers. And the power exchange between the external power grid and the power grid in the study area at the boundary nodes is counted:

[0150]

[0151] in represents the total injected power at the boundary node i, represents the total outflow power of boundary node i, represents the power injected from node j to node i, represents the power flowing from node i to node j.

[0152] 4. Fit the external grid dynamic characteristics at the boundary nodes. For example, in the V edge Each node is connected to an equivalent generator-comprehensive load model, and its model structure is as follows Figure 3 As shown in Figure 1, it consists of three parts: a sixth-order generator G, a dynamic load M, and a ZIP load. The generator power is set to The maximum power generation is The total load power is The dynamic load power is The ZIP load power is

[0153] The mathematical model of the generator is represented as follows

[0154]

[0155] The generator grid connection point voltage u in the dq plane d 、u q , sub-transient electromotive force E″ d , E″ q , transient electromotive force E′ d , E′ q , current i d 、i d , generator angular velocity ω and generator power angle δ are generator state variables, which are numerically calculated according to the above differential equations based on the generator operating state and grid working conditions during the simulation process; rated angular velocity ω s , reference power S b is the grid reference parameter; excitation voltage E f0 , mechanical torque T M0 is the external input variable; armature resistance r a , subtransient reactance X″ in dq plane d 、X″ q , transient reactance X′ d , X′ q , synchronous reactance X d 、X q , subtransient time constant T″ d0 , T″ q0 , transient time constant T′ d0 , T′ q0 , generator kinetic energy E a are the generator parameters.

[0156] The power injected into the system by the generator can be expressed as:

[0157] P e =u d i d +u q iq

[0158] The dynamic load is mainly composed of induction motors, and its mathematical model is as follows:

[0159]

[0160] Among them, P i 、U i is the real-time active power and node voltage of node i; R R is the rotor resistance, R S is the stator resistance, X R is the rotor reactance, X S is the stator reactance, X M Magnetizing reactance

[0161] The mathematical expression of the ZIP load mathematical model is as follows

[0162]

[0163] Among them, P i0 , Q i0 、U i0 is the active power, reactive power and node voltage of node i in steady state; P i , Q i 、U i is the real-time active power, reactive power and node voltage of node i; k Z 、k I 、k P is the coefficient of the load constant impedance component, constant current component, and constant power component, satisfying k Z +k I +k P =1.

[0164] Among the above parameters, the excitation voltage E f0 , mechanical torque T M0 , armature resistance r a , subtransient reactance X″ in dq plane d 、X″ q , transient reactance X′ d , X′ q , synchronous reactance X d 、X q , subtransient time constant T″ d0 , T″ q0 , transient time constant T′ d0 , T′ q0 , generator kinetic energy E a , rotor resistance R R , stator resistance R S , rotor reactance X R , stator reactance X S, magnetizing reactance X M , constant impedance component coefficient k of load Z , constant current component coefficient k I , constant power component coefficient k P The load type proportional coefficient α is an adjustable parameter of the equivalent generator-load model. By adjusting these parameters, the dynamic characteristics of the equivalent generator-load model can be changed.

[0165] With the goal of fitting the equivalent generator-load model curve with the original system curve, the above model parameters are adjusted to fit the curve of the boundary node under fault conditions to achieve external equivalence. Combined with the target power grid data file, an equivalent model of the target DC line is generated.

[0166] Furthermore, the equivalent model generation method further includes simplifying the internal power grid nodes of the equivalent model to obtain a simplified equivalent model, specifically:

[0167] Get the pre-reserved node entered by the user;

[0168] Classifying each internal power grid node of the equivalent model as a reserved node or a node to be deleted according to a preset screening condition and the pre-reserved node;

[0169] Sorting each of the internal power grid nodes according to their types, and establishing a block node admittance matrix according to the sorting results;

[0170] The block node admittance matrix is ​​simplified into an equivalent admittance matrix using the Kron simplification method;

[0171] The equivalent model is simplified according to the equivalent admittance matrix and the type of each internal power grid node to obtain the simplified equivalent model.

[0172] This embodiment of the present application provides an internal simplification method for equivalent models. Based on preset screening criteria and user-specified pre-reserved nodes, each internal power grid node in the equivalent model is classified as either retained or to be deleted. This method, combined with block admittance matrix sorting, prioritizes simplification of non-critical areas, further optimizing computational resource allocation. The Kron simplification method then compresses the block structure of the large-scale admittance matrix into a low-dimensional equivalent admittance matrix through mathematical transformation. This method directly reduces the computational complexity of the equivalent model in electromagnetic transient simulations and improves the simulation efficiency of the equivalent model.

[0173] In a preferred embodiment, the internal grid nodes of the equivalent model are simplified to obtain a simplified equivalent model, specifically:

[0174] 1. Based on the regional power grid equivalent work purpose with DC as the research object, determine the nodes that need to be retained for internal network simplification: DC nodes, nodes above 500KV, regional synchronous power sources that may be coupled with the DC system, and new energy station nodes V s1 , and add some nodes V that need to be retained manually s2 , establish an internal reserved node set V s =V s1 ∩V s2 , the remaining nodes are nodes to be eliminated V r =V research \V s ;

[0175] 2. Sort the nodes by eliminating nodes and retaining nodes, and generate the all-zero block admittance matrix Y

[0176]

[0177] Taking the outflow power of each node as positive, the admittance corresponding to the element self-admittance is generated for the power element of each node.

[0178]

[0179] For lines and transformers, the impedance values ​​they represent are directly converted into corresponding self-admittance values ​​and mutual-admittance values. The final calculation results are filled into the block admittance matrix, and the following formula is applied to simplify the network and establish the equivalent system admittance matrix.

[0180] Y eq =Y ss -Y sr (Y rr ) -1 Y rs

[0181] 3. Equally generated Y eq The off-diagonal elements of the matrix (Y eq ) ij Converted to line / transformer impedance between boundary nodes i and j.

[0182]

[0183] Then apply the following formula to generate the new reactive and active powers of the retained nodes after equalizing the earth admittance.

[0184]

[0185] The above power is converted into active power and reactive power of the node load, and filled into the target power grid data file corresponding to the equivalent model, and then the corresponding simplified equivalent model is generated according to the modified target power grid data file.

[0186] In a preferred embodiment, in order to further verify the simulation effect of the equivalent model provided by the embodiment of the present application, a comparative experiment was conducted in the embodiment of the present application. Taking the actual DC line to be studied in a certain area as an example, Figure 4 The steps shown in the flow separate the regional grid data containing the DC link to be studied from the original system data, perform equivalent and simplified operations, and generate an equivalent model. Electromechanical transient simulations are then performed on the original system and the equivalent model.

[0187] Comparison of simulation time: Using the electromechanical transient simulation software DSP as the simulation tool, the time for performing a fault simulation in the original system is 5.585 seconds, while the time for performing a fault simulation using the equivalent model established in the embodiment of the present application is 0.936 seconds, which is only 16.8% of the original system.

[0188] Verification of the fitting degree of the simulation curve: Here, a short circuit fault is set at the AC station at the DC sending end and in the middle of the DC line, and some simulation curves are obtained as follows Figure 5 、 Figure 6 As shown, Figure 5 The simulation diagram of some power devices of the equivalent model and the original system when a three-phase grounding fault occurs at the DC sending end is shown in the figure. Figure 6 Simulation diagram of some power components of the equivalent model and the original system when a DC midpoint grounding fault occurs.

[0189] In summary, applying the equivalent model generation method provided in the embodiment of the present application to perform equivalent modeling of the DC line can effectively preserve the electrical characteristics of the DC line and its coupling with the AC system, while significantly reducing the scale of the power grid and thus reducing the simulation time, thereby improving the simulation response speed and simulation accuracy of the equivalent model of the AC / DC system.

[0190] Example 2:

[0191] like Figure 7 As shown, embodiment 2 provides an equivalent model generation system based on DC two-terminal site coupling, including an acquisition module 10, a graph construction module 20, a first node set construction module 30, a second node set construction module 40, a region division module 50 and an equivalent model generation module 60;

[0192] The acquisition module 10 is used to acquire a target DC line and a DC double-terminal station of the target DC line;

[0193] The graph construction module 20 is used to construct an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes;

[0194] The first node set construction module 30 is used to search the shortest path between the DC double-terminal stations excluding the target DC line from the main grid weighted graph model, and extract the main grid nodes on the shortest path to construct a first node set;

[0195] The second node set construction module 40 is configured to perform a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determine a number of main power grid nodes from each main power grid node to construct a second node set based on the voltage change of each main power grid node at the moment of the fault and preset conditions;

[0196] The region division module 50 is configured to merge the first node set and the second node set into a core node set, and divide each power grid node into an external region node set and an internal region node set according to the core node set and the whole network unweighted graph model, wherein the internal region node set includes all main power grid nodes in the core node set;

[0197] The equivalent model generation module 60 is used to fit the power grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modify the original power grid data file accordingly, and then generate the equivalent model of the target DC line according to the modified original power grid data file.

[0198] Furthermore, the graph construction module 20 constructs an unweighted graph model of the entire network and a weighted graph model of the main network according to the original grid data file and the preset power threshold, including:

[0199] Constructing the unweighted graph model of the entire network according to the connection relationship between each grid node in the original grid data file;

[0200] Determine a number of main grid nodes whose power is greater than the preset power threshold from the original grid data file, and construct an initial main grid weighted graph model based on the connection relationship between each of the main grid nodes and each of the main grid nodes;

[0201] Traversing each connection edge in the initial main network weighted graph model, determining the weight value of the connection edge according to the power grid line type corresponding to the connection edge, and then obtaining the main network weighted graph model;

[0202] Among them, if the grid line type corresponding to the connection edge is an AC line, the weight value of the connection edge is the inverse of the equivalent impedance of the AC line; if the grid line type corresponding to the connection edge is a DC line, the weight value of the connection edge is 1.

[0203] In one possible implementation, the first node set construction module 30 searches the main grid weighted graph model for the shortest path between the DC double-terminal stations excluding the target DC line, extracts the main grid nodes on the shortest path to construct the first node set, including:

[0204] Removing the connection edges corresponding to the target DC line from the main network weighted graph model to obtain a temporary weighted graph;

[0205] Based on the Dijkstra algorithm and the weight values ​​of each connection edge in the temporary weighted graph, searching and determining the shortest path between the DC double-terminal stations in the temporary weighted graph;

[0206] Main grid nodes on the shortest path are extracted to construct a first node set.

[0207] In one possible implementation, the second node set construction module 40 performs transient simulation on the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determines a number of main power grid nodes from each main power grid node to construct a second node set based on voltage changes of each main power grid node at the moment of a fault and preset conditions, including:

[0208] Construct a blank initial second node set;

[0209] According to the original power grid data file, in a power grid electromechanical transient simulation environment, a transient simulation of the power grid is performed by setting a single-ended three-phase short-circuit fault at the DC double-terminal station, and obtaining a voltage change value of each main power grid node at the moment of the fault;

[0210] Each of the voltage change values ​​is compared with a preset threshold value. If the voltage change value is greater than or equal to the preset threshold value, the main grid node corresponding to the voltage change value is added to the initial second node set. After completing all threshold comparisons, the second node set is obtained.

[0211] Furthermore, the area division module 50 divides each grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, including:

[0212] Constructing a blank initial outer region node set and an initial inner region node set containing the core node set;

[0213] Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph;

[0214] Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory;

[0215] Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set;

[0216] After the traversal is completed, the external area node set and the internal area node set are obtained.

[0217] Furthermore, the equivalent model generation module 60 fits the grid characteristics of the external region at the boundary nodes between the external region node set and the internal region node set and modifies the original grid data file accordingly, and then generates the equivalent model of the target DC link according to the modified original grid data file, including:

[0218] Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set;

[0219] Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model;

[0220] Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram;

[0221] Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file;

[0222] Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model;

[0223] The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link.

[0224] In one possible implementation, the equivalent model generation system further includes a simplification module, which is configured to simplify the internal power grid nodes of the equivalent model to obtain a simplified equivalent model, specifically:

[0225] Get the pre-reserved node entered by the user;

[0226] Classifying each internal power grid node of the equivalent model as a reserved node or a node to be deleted according to a preset screening condition and the pre-reserved node;

[0227] Sorting each of the internal power grid nodes according to their types, and establishing a block node admittance matrix according to the sorting results;

[0228] The block node admittance matrix is ​​simplified into an equivalent admittance matrix using the Kron simplification method;

[0229] The equivalent model is simplified according to the equivalent admittance matrix and the type of each internal power grid node to obtain the simplified equivalent model.

[0230] The embodiment of the present application provides an equivalent model generation system based on DC double-terminal site coupling, which uses the shortest path algorithm to extract the first node set from the main grid weighted graph model, uses transient simulation to screen out the second node set from each main grid node, and then constructs a core node set. Finally, based on the core node set and the unweighted graph model of the entire network, internal and external regions are divided and external equivalents are performed to generate the corresponding equivalent model. Among them, the extraction of nodes by the shortest path algorithm can effectively retain the DC site through the AC system coupling channel, ensuring that the subsequent equivalent model accurately reflects the dynamic interaction characteristics of the AC / DC system. At the same time, since the main grid weighted graph model used has already screened each grid node according to the preset power threshold, it effectively reduces the node scale in the subsequent generated equivalent model and improves the computational efficiency of the equivalent model. By screening the voltage fluctuation sensitive nodes through transient simulation, it can significantly reduce the node scale to be simulated, reduce the computational complexity, and avoid the decoupling error of the traditional infinite power equivalent method. The core node set constructed by combining the two makes the equivalent AC / DC system greatly reduce the scale while retaining the coupling relationship between the large power grid and the DC system, improving the simulation response speed and simulation accuracy of the AC / DC system equivalent model.

[0231] The more detailed working principle and process flow of this embodiment can be referred to, but not limited to, the relevant records of the first embodiment.

[0232] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application by those skilled in the art should be included within the scope of protection of this application.

Claims

1. A method for generating an equivalent model based on DC two-terminal site coupling, characterized in that: include: Acquire a target DC line and a DC double-terminal station of the target DC line; Constructing an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes; Searching for the shortest path between the DC double-terminal stations excluding the target DC line from the main grid weighted graph model, extracting the main grid nodes on the shortest path to construct a first node set; performing a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determining a number of main power grid nodes from each main power grid node to form a second node set based on voltage changes of each main power grid node at the moment of the fault and preset conditions; Merging the first node set and the second node set into a core node set, and dividing each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, wherein the internal area node set includes all main power grid nodes in the core node set; The grid characteristics of the external area are fitted at the boundary nodes between the external area node set and the internal area node set, and the original grid data file is modified accordingly, and then an equivalent model of the target DC line is generated according to the modified original grid data file.

2. The method for generating an equivalent model based on DC two-terminal site coupling according to claim 1, characterized in that: The method of constructing a network-wide unweighted graph model and a main network weighted graph model based on the original grid data file and the preset power threshold includes: Constructing the unweighted graph model of the entire network according to the connection relationship between each grid node in the original grid data file; Determine a number of main grid nodes whose power is greater than the preset power threshold from the original grid data file, and construct an initial main grid weighted graph model based on the connection relationship between each of the main grid nodes and each of the main grid nodes; Traversing each connection edge in the initial main network weighted graph model, determining the weight value of the connection edge according to the power grid line type corresponding to the connection edge, and then obtaining the main network weighted graph model; Among them, if the grid line type corresponding to the connection edge is an AC line, the weight value of the connection edge is the inverse of the equivalent impedance of the AC line; if the grid line type corresponding to the connection edge is a DC line, the weight value of the connection edge is 1.

3. The method for generating an equivalent model based on DC two-terminal site coupling according to claim 1, characterized in that: The step of searching the main grid weighted graph model for the shortest path between the DC double-terminal stations excluding the target DC line, and extracting the main grid nodes on the shortest path to construct a first node set includes: Removing the connection edges corresponding to the target DC line from the main network weighted graph model to obtain a temporary weighted graph; Based on the Dijkstra algorithm and the weight values ​​of each connection edge in the temporary weighted graph, searching and determining the shortest path between the DC double-terminal stations in the temporary weighted graph; Main grid nodes on the shortest path are extracted to construct a first node set.

4. The method for generating an equivalent model based on DC two-terminal site coupling according to claim 1, wherein: The method further comprises: performing transient simulation on the power grid in a power grid electromechanical transient simulation environment according to the original power grid data file, and then determining a plurality of main power grid nodes from each main power grid node to form a second node set according to voltage changes of each main power grid node at the moment of the fault and preset conditions, including: Construct a blank initial second node set; According to the original power grid data file, in a power grid electromechanical transient simulation environment, a transient simulation of the power grid is performed by setting a single-ended three-phase short-circuit fault at the DC double-terminal station, and obtaining a voltage change value of each main power grid node at the moment of the fault; Each of the voltage change values ​​is compared with a preset threshold value. If the voltage change value is greater than or equal to the preset threshold value, the main grid node corresponding to the voltage change value is added to the initial second node set. After completing all threshold comparisons, the second node set is obtained.

5. The method for generating an equivalent model based on DC two-terminal site coupling according to claim 1, characterized in that: The dividing of each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model includes: Constructing a blank initial outer region node set and an initial inner region node set containing the core node set; Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph; Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory; Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set; After the traversal is completed, the external area node set and the internal area node set are obtained.

6. The method for generating an equivalent model based on DC two-terminal site coupling according to claim 1, characterized in that: The step of fitting the grid characteristics of the external region at the boundary nodes between the external region node set and the internal region node set and modifying the original grid data file accordingly, and then generating the equivalent model of the target DC line according to the modified original grid data file, includes: Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set; Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model; Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram; Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file; Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model; The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link.

7. The method for generating an equivalent model based on DC two-terminal site coupling according to any one of claims 1 to 6, characterized in that: The equivalent model generation method further includes simplifying the internal power grid nodes of the equivalent model to obtain a simplified equivalent model, specifically: Get the pre-reserved node entered by the user; Classifying each internal power grid node of the equivalent model as a reserved node or a node to be deleted according to a preset screening condition and the pre-reserved node; Sorting each of the internal power grid nodes according to their types, and establishing a block node admittance matrix according to the sorting results; The block node admittance matrix is ​​simplified into an equivalent admittance matrix using the Kron simplification method; The equivalent model is simplified according to the equivalent admittance matrix and the type of each internal power grid node to obtain the simplified equivalent model.

8. An equivalent model generation system based on DC two-terminal site coupling, characterized in that: It includes an acquisition module, a graph construction module, a first node set construction module, a second node set construction module, a region partitioning module and an equivalent model generation module; The acquisition module is used to acquire a target DC line and a DC double-terminal station of the target DC line; The graph construction module is used to construct an unweighted graph model of the entire network and a weighted graph model of the main network according to the original power grid data file and the preset power threshold, wherein the unweighted graph model of the entire network includes the connection relationship between each power grid node and each power grid node, and the weighted graph model of the main network includes the connection relationship between a number of main grid nodes having a power greater than the preset power threshold and each of the main grid nodes; The first node set construction module is used to search the shortest path between the DC double-terminal stations excluding the target DC line from the main grid weighted graph model, and extract the main grid nodes on the shortest path to construct a first node set; The second node set construction module is configured to perform a transient simulation of the power grid in a power grid electromechanical transient simulation environment based on the original power grid data file, and then determine a number of main power grid nodes from each main power grid node to construct a second node set based on the voltage change of each main power grid node at the moment of the fault and preset conditions; The area division module is used to merge the first node set and the second node set into a core node set, and divide each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, wherein the internal area node set includes all main power grid nodes in the core node set; The equivalent model generation module is used to fit the power grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modify the original power grid data file accordingly, and then generate the equivalent model of the target DC line according to the modified original power grid data file.

9. The equivalent model generation system based on DC two-terminal site coupling according to claim 8, characterized in that: The area division module divides each power grid node into an external area node set and an internal area node set according to the core node set and the whole network unweighted graph model, including: Constructing a blank initial outer region node set and an initial inner region node set containing the core node set; Removing a number of power grid nodes corresponding to the core node set from the full-network unweighted graph model to obtain a temporary unweighted graph; Splitting the temporary unweighted graph into a plurality of connected components based on graph theory connected component theory; Traversing each of the connected components, if a main power grid node exists in the connected component, classifying the power grid node in the connected component into an initial external region node set; otherwise, classifying the power grid node in the connected component into an initial internal region node set; After the traversal is completed, the external area node set and the internal area node set are obtained.

10. The equivalent model generation system based on DC two-terminal site coupling according to claim 8, characterized in that: The equivalent model generation module fits the grid characteristics of the external area at the boundary nodes between the external area node set and the internal area node set and modifies the original grid data file accordingly, and then generates the equivalent model of the target DC line according to the modified original grid data file, including: Determining a plurality of boundary nodes and a plurality of internal and external connection edges based on the full-network unweighted graph model, the external area node set, and the internal area node set, wherein the boundary node is a power grid node in the external area node set that has a connection edge with any power grid node in the internal area node set, and the internal and external connection edge is a connection edge between the boundary node and any power grid node in the internal area node set; Determining internal connection edges between each power grid node in the internal regional node set according to the full network unweighted graph model; Merging the plurality of boundary nodes, the plurality of internal and external connection edges, the internal area node set, and the plurality of internal connection edges to construct a target area power grid node line diagram; Deleting and simplifying the content of the original power grid data file according to the target area power grid node diagram, retaining the connection edges and power grid nodes corresponding to the target area power grid node diagram, and obtaining a target power grid data file; Inserting a preset initial generator-load model at a boundary node in the target power grid data file, and adjusting parameters of the initial generator-load model based on a system curve corresponding to the original power grid data file to obtain an equivalent generator-load model; The equivalent generator-load model is combined with the target power grid data file to generate an equivalent model of the target DC link.

Images