Main grid topology gridding construction method based on OP interconnection
Through the main grid topology grid-based grid network topology grid construction method, the complexity and real-time problems of large-scale topology modeling and simulation of power grid are solved, and efficient grid management and analysis are achieved.
Patent Information
- Application Number
- CN202411436742.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-15
- Publication Date
- 2025-02-11
AI Technical Summary
Existing power grid technologies are difficult to effectively handle the topology of large-scale networks, which increases the complexity of modeling and simulation, cannot meet real-time requirements, and lacks intuitive visual interfaces and analysis tools.
Using the main network frame topology grid construction method based on OP interconnection, collect distribution network data through monitoring equipment, network modeling and topology design based on OP interconnection results, generate distribution network main network frame topology model, and maintain it with the same source The model is mapped, combined with the hybrid clustering method, and the power supply grid is divided into a mapping relationship to form a current grid analysis model.
It realizes effective modeling and simulation of large-scale power grids, meets real-time requirements, provides intuitive visual interfaces and analysis tools, improves the operating efficiency and management level of the power grid, and enhances the ability to adapt to faults and load changes.
Smart Images
Figure CN119419745A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power grids, and in particular to a main grid topology grid construction method based on OP interconnection. Background Art
[0002] With the continuous expansion of network scale and the increase of complexity, higher requirements are put forward for the modeling and simulation of distribution network topology. How to effectively handle the topology of large-scale networks and accurately simulate network performance is still a challenge at the current stage. Modern networks usually involve multiple protocols and technologies, such as Ethernet, wireless LAN, IPv6, etc. When modeling and simulating these protocols and technologies, special attention should be paid to the interaction and compatibility between them, which usually increases the complexity of modeling and simulation. Some network applications also have high real-time requirements, such as VoIP and video surveillance. In the modeling and simulation of topology, it is necessary to consider how to meet the real-time requirements and dynamically reflect the changes in network topology. How to provide intuitive and easy-to-understand visualization interfaces and analysis tools to effectively display the results of topology modeling and simulation is also a challenge that needs to be solved urgently.
[0003] In order to solve the above problems, the present invention proposes a main grid topology grid construction method based on OP interconnection. Summary of the invention
[0004] The purpose of the present invention is to propose a main grid topology grid construction method based on OP interconnection to solve the problems raised in the background technology: The existing distribution network is large in scale and highly complex; the interaction and compatibility between networks involving multiple protocols and technologies increase the complexity of modeling and simulation; the real-time requirements in the topology modeling process cannot be met; and the existing technology cannot provide an intuitive display interface.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions: The main grid topology grid construction method based on OP interconnection includes the following steps: S1: Collect distribution network data based on monitoring equipment; S2: Based on the collected distribution network data and OP interconnection results, network modeling and topology design of the distribution network are performed to generate a distribution network main grid topology model, and the generated distribution network main grid topology model is mapped and associated with the homologous maintenance model; S3: Divide the power supply grid based on the loads in different areas of the power supply map combined with the hybrid clustering method; S4: Establish a mapping relationship between the distribution network main grid topology model and the power supply grid to form a current grid analysis model within the power supply grid.
[0006] Preferably, in S2, a topological model of the power grid is established based on the collected distribution network data using the OP interconnection results, and the node and connection information is defined and stored in accordance with the OPC standard format, the OPC technical standard is selected, and the server and client are configured; then the established topological model is verified and optimized to obtain the distribution network main grid topological model.
[0007] Preferably, the topological model of the power grid determines the nodes and edges of the main grid based on the OP interconnection results, establishes the logical and physical connection relationship between the nodes and the connection lines, and constructs the topological structure based on the minimum spanning tree algorithm, as follows: S2.1.1: Define the states, actions, and rewards of the network; S2.1.2: Based on the improved DQN algorithm, learn the behavior pattern of the network and learn the optimal strategy through interaction with the environment; S2.1.3: Dynamically adjust edge weights based on learning results; S2.1.4: Sort all the edges in the network determined based on the OP interconnection results in ascending order of weight; S2.1.5: Create an empty minimum spanning tree T and prepare a union-find data structure to manage the connections between nodes; S2.1.6: Consider each edge in order: S2.1.7: For the edge (u,v) currently under consideration, use union-find to determine whether u and v are already in the same connected component; if not, add ((u,v)) to the minimum spanning tree T and merge u and v into one connected component; S2.1.8: If u and v are already in the same connected component, skip this edge to avoid forming a loop; S2.1.9: When the weight of an edge changes, update the order of the edge list based on the change; S2.1.10: Repeat the above steps until the minimum spanning tree T contains n-1 edges, where n is the number of nodes in the network.
[0008] Preferably, the improved DQN algorithm in S2.1.2 uses a stochastic gradient descent method to update the value function of the neural network, as follows:
[0009] in, , are the value functions of the neural network before and after the update respectively; is the learning rate; For immediate rewards; is the reward discount factor; is the optimal strategy; is the policy gradient; For the current strategy; Optimize network parameters based on particle swarm dynamic optimization algorithm; The particle swarm dynamic optimization algorithm is specifically as follows:
[0010]
[0011]
[0012] in, , For the Particle No. Vidi The iteration and The speed of iterations; For the The weight of the iteration; , is the learning factor; , for A random number between is the current optimal value; is the global optimal value; , For the Particle No. Vidi The iteration and The position of the iteration; , is the maximum and minimum weight; is the maximum number of iterations.
[0013] Preferably, in S2.1.9, a heap or a balanced binary tree is used to update and query edge weight changes.
[0014] Preferably, in S2.1.9, when the weight of an edge decreases and the edge is already in the minimum spanning tree, reconsider whether the minimum spanning tree condition is still met; If the weight of an edge increases and the edge is in the minimum spanning tree, it is removed from the minimum spanning tree and the minimum spanning tree is rebuilt.
[0015] Preferably, the construction of the homology maintenance model in S2 includes the following steps: S2.2.1: Determine the homologous devices in the distribution network based on the distribution network data; S2.2.2: Analyze the operating characteristics and impact of each device and use this to divide the grid into a number of homogeneous groups; S2.2.3: Determine the interactions and communication between homologous groups; S2.2.4: Develop a maintenance plan and strategy for the homologous maintenance model.
[0016] Preferably, the hybrid clustering algorithm in S3 performs hybrid clustering based on a hierarchical clustering algorithm and a K-means algorithm, as follows: The distribution network area is divided into several sub-clusters at different levels based on the regional load. In each level of the hierarchical clustering, the initial cluster center point is selected according to the center point. The initial center point selected in the hierarchical clustering stage is used as the initial center point of the k-means algorithm. The k-means algorithm is iteratively optimized based on the initial center point to further refine the clustering results.
[0017] Preferably, in S4, the mapping relationship between the main grid topology model of the distribution network and the power supply grid is established as follows: S4.1: Find nodes related to the power supply grid boundary in the constructed distribution network main grid topology model, and mark the found nodes in the model to indicate the power supply grid to which the nodes belong; S4.2: Create a corresponding relationship table to record each power grid and the associated topology model node in a one-to-one correspondence; S4.3: Associate the lines extending from these key nodes with the corresponding power supply grids to clarify the grid range served by the lines; S4.4: Dynamically update the mapping relationship according to changes in the distribution network and adjustments to the power supply grid.
[0018] Compared with the prior art, the present invention provides a main grid topology grid construction method based on OP interconnection, which has the following beneficial effects: The present invention realizes the interconnection and intercommunication of various devices and systems based on OP interconnection including standardized communication protocols, data formats and interfaces, and at the same time combines the improved DQN algorithm to generate the optimal dynamic distribution network main grid topology model to meet the real-time requirements in the modeling process, and replaces the distribution network main grid topology model with a homologous maintenance model, which can improve the operation efficiency and flexibility of the distribution network. The power supply area is divided, which can effectively improve the operation efficiency and management level of the distribution network, and at the same time enhance the adaptability of the power grid to faults and load changes; a mapping relationship is established between the distribution network main grid topology model and the power supply grid to form a current grid analysis model in the power supply grid, which can support the distribution network grid analysis and grid topology display functions, and improve the efficiency, reliability and manageability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 This is a flow chart of the method mentioned in Example 1 of the present invention; Figure 2 This is a schematic diagram of the power supply grid unit division mentioned in Example 1 of the present invention. DETAILED DESCRIPTION
[0020] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.
[0021] The present invention realizes the interconnection and intercommunication of various devices and systems based on OP interconnection including standardized communication protocols, data formats and interfaces, and at the same time combines the improved DQN algorithm to generate the optimal dynamic distribution network main grid topology model to meet the real-time requirements in the modeling process, and replaces the distribution network main grid topology model with a homologous maintenance model, which can improve the operation efficiency and flexibility of the distribution network, and then divide the power supply area, which can effectively improve the operation efficiency and management level of the distribution network, and at the same time enhance the adaptability of the power grid to faults and load changes; establish a mapping relationship between the distribution network main grid topology model and the power supply grid, and form a current grid analysis model in the power supply grid, which can support the distribution network grid analysis and grid topology display functions, and improve the efficiency, reliability and manageability of the power system. Specifically including the following contents.
[0022] Embodiment 1: See also Figure 1-2 The present invention provides a main grid topology grid construction method based on OP interconnection, comprising: S1: Collect distribution network data based on monitoring equipment; the details are as follows: First, various data of the power system are collected and analyzed, including the geographical location, electrical parameters, operating status and other information of power stations, substations, distribution equipment, transmission lines, etc. These data can be obtained through sensors, SCADA systems and other monitoring equipment.
[0023] S2: Based on the collected distribution network data and OP interconnection results, network modeling and topology design of the distribution network are performed to generate a distribution network main grid topology model, and the generated distribution network main grid topology model is mapped and associated with the homologous maintenance model; the details are as follows: The grid modeling and topology design of the power system are carried out based on the collected data and OP interconnection results. This includes determining the nodes (such as power stations, substations) and connection lines (transmission lines, distribution lines) of the main grid, and establishing the logical and physical connection relationships between them. OP interconnection results usually refer to an open interconnection solution that can include standardized communication protocols, data formats and interfaces. In the grid construction of the main grid topology, OP interconnection results can be used to achieve the interconnection of various devices and systems to achieve data sharing and collaborative control.
[0024] During the construction process, network security and communication requirements need to be considered to ensure that the application of OP interconnection results will not affect the stability and security of the power system. Therefore, appropriate security measures and communication strategies need to be taken to prevent network attacks and data leakage.
[0025] Based on the collected distribution network data, the OP interconnection results are used to establish the topological model of the power grid. The topological model of the power grid determines the nodes and edges of the main grid based on the OP interconnection results, and establishes the logical and physical connection relationship between the nodes and the connection lines, and builds the topological structure based on the minimum spanning tree algorithm, as follows: S2.1.1: Define the states, actions, and rewards of the network; specifically: Status: defines the current state of the network, which can include node load, connection quality, data transfer rate, etc.
[0026] Actions: Defines actions that can be performed in the network, such as increasing bandwidth, adjusting latency, etc.
[0027] Reward: Set a reward function, defined based on the improvement of network performance or the degree of task completion. For example, reducing transmission delay can give positive rewards, and increasing bandwidth can give negative rewards.
[0028] S2.1.2: Based on the improved DQN algorithm, the behavior pattern of the learning network is learned, and the optimal strategy is learned through interaction with the environment; the improved DQN algorithm uses stochastic gradient descent to update the value function of the neural network, as follows:
[0029] in, , are the value functions of the neural network before and after the update respectively; is the learning rate; For immediate rewards; is the reward discount factor; is the optimal strategy; is the policy gradient; For the current strategy; Optimize network parameters based on particle swarm dynamic optimization algorithm; The particle swarm dynamic optimization algorithm is specifically as follows:
[0030]
[0031]
[0032] in, , For the Particle No. Vidi The iteration and The speed of iterations; For the The weight of the iteration; , is the learning factor; , for A random number between is the current optimal value; is the global optimal value; , For the Particle No. Vidi The iteration and The position of the iteration; , is the maximum and minimum weight; is the maximum number of iterations; The improved DQN algorithm can learn strategies in complex environments and adapt to changes in dynamic environments.
[0033] S2.1.3: Dynamically adjust edge weights based on learning results; details are as follows: Once the reinforcement learning algorithm is trained, the weights of the edges in the network can be dynamically adjusted based on the learned strategy. The specific steps include: Edge weight adjustment during learning: During training, edge weights can be adjusted dynamically based on the decisions made by the reinforcement learning algorithm, for example, to optimize bandwidth utilization or reduce latency.
[0034] Application to real networks: In real networks, edge weights are dynamically adjusted by monitoring the network state and using the policy recommended by the reinforcement learning model. This can be achieved through a network controller or an automated system.
[0035] S2.1.4: Sort all the edges in the network determined based on the OP interconnection results in ascending order of weight; S2.1.5: Create an empty minimum spanning tree T and prepare a union-find data structure to manage the connections between nodes; S2.1.6: Consider each edge in order: S2.1.7: For the edge (u,v) currently under consideration, use the union-find function to determine whether u and v are already in the same connected component. If they are not in the same connected component, add ((u,v)) to the minimum spanning tree T and merge u and v into one connected component. Among them, optimization measures such as path compression and rank merging can be used to ensure the efficiency of the union-find operation, so that the time complexity of each operation is close to a constant.
[0036] S2.1.8: If u and v are already in the same connected component, skip this edge to avoid forming a loop; S2.1.9: When the weight of an edge changes, update the order of the edge list based on the change; use a heap or balanced binary tree to update and query edge weight changes, and when the weight of an edge changes, update the order of the edge list based on the change. If the weight of an edge decreases and the edge is already in the minimum spanning tree, reconsider whether the minimum spanning tree conditions are still met (you may need to rebuild the minimum spanning tree). If the weight of an edge increases and the edge is in the minimum spanning tree, you may need to remove it from the minimum spanning tree and rebuild the minimum spanning tree.
[0037] S2.1.10: Repeat the above steps until the minimum spanning tree T contains n-1 edges, where n is the number of nodes in the network; when the algorithm is completed, the minimum spanning tree T is a minimum spanning tree of the network that connects all nodes and has the smallest sum of edge weights; This greedy Kruskal algorithm ensures the local optimal solution at each step, and the final global solution is the overall optimal solution, that is, the minimum spanning tree of the network. This algorithm is applicable to various types of weighted graphs, such as undirected graphs or directed graphs, and is an efficient method for solving the minimum spanning tree problem.
[0038] Define and store the node and connection information in the OPC standard format, and select OPC technical standards, such as OPC UA, to ensure that the system can support existing power grid equipment and communication protocols. Configure the server and client to ensure that data can be collected from field devices stably and efficiently, and transmitted to the central system or database through the network; then verify whether the established topology model accurately reflects the structure and connection relationship of the actual power grid. This may involve comparison and verification with actual operating data. Based on actual operating experience and data feedback, optimize and adjust the topology model to improve its accuracy and practicality. Use the established topology model and OPC technology to achieve real-time monitoring and management of the power grid operation status. This includes the collection, analysis and visualization of real-time data to support power grid operation and fault management decisions.
[0039] S3: The power grid is divided based on the loads in different areas of the power supply map combined with the hybrid clustering method; the details are as follows: The clustering results based on the hybrid clustering method are divided into corresponding power supply units. For details, please refer to Table 1: Table 1 Load range of power supply unit divided based on hybrid clustering method
[0040] First, the load intervals of the power supply unit are divided into [0,5), [5,10), [10,15) and [15,25) based on the hierarchical clustering algorithm. Then, the load interval [15,25) is further divided into [20,21), [21,22), [22,23), [23,24) and [24,25) based on the K-means algorithm. Figure 2 The division result.
[0041] S4: Establish a mapping relationship between the main distribution network topology model and the power supply grid to form a current grid analysis model within the power supply grid. The details are as follows: Determine the association points: Find the nodes related to the power grid boundary in the topology model, such as substation outgoing line nodes, etc. These nodes will serve as the key association points for establishing the mapping.
[0042] Mark the corresponding nodes of the grid: clearly mark these key nodes in the topology model to indicate the power supply grid to which they belong.
[0043] Establish a correspondence table: Create a correspondence table to record the one-to-one correspondence between each power supply grid and the associated topology model node.
[0044] Associated lines: Associate the lines extending from these key nodes with the corresponding power supply grids to clarify the grid range served by the lines.
[0045] Dynamic update: As the distribution network changes and the power supply grid is adjusted, the mapping relationship is updated in a timely manner to ensure accuracy.
[0046] The final current grid analysis model within the power supply grid can support distribution network analysis and grid topology display functions.
[0047] For example, for a power grid that has been divided and has a topology model, you can first find the correspondence between each grid and the substation outgoing line nodes in the topology model, and mark these nodes as belonging to a specific grid in the model. Then record each grid and the corresponding node and related lines in detail in the relationship table. When a new substation is put into operation or the line is modified, this mapping relationship is updated in a timely manner to ensure that the corresponding power grid and topology structure can be accurately associated when conducting analysis and decision-making. Please refer to the following example for details: According to the nine power supply grids of the urban distribution network in S3: DY-01, DY-02, DY-03, DY-04, DY-05, DY-06, DY-07, DY-08, DY-09; In the distribution network main grid topology model, we identified nine substation outgoing line nodes as association points: node 1, node 2, node 3, node 5, node 7, node 10, node 12, node 16 and node 18.
[0048] Step 1: Determine the connection points After analysis, it was found that node 1 is connected to the main area of the DY-01 grid, node 2 is connected to the core area of the DY-02 grid, node 3 is responsible for the power supply of the DY-03 grid, node 5 is connected to the important load area of the DY-04 grid, node 10 is connected to the emerging commercial area of the DY-05 grid, node 12 is connected to the industrial concentration area of the DY-06 grid, node 7 is connected to the densely populated area of the DY-07 grid, node 18 is connected to the science and technology park of the DY-08 grid, and node 16 is connected to the schools and surrounding areas of the DY-09 grid.
[0049] Step 2: Mark the corresponding nodes of the grid In the topology model, we explicitly mark node 1 as “DY-01 grid-associated node”, node 2 as “DY-02 grid-associated node”, node 3 as “DY-03 grid-associated node”, node 5 as “DY-04 grid-associated node”, node 10 as “DY-05 grid-associated node”, node 12 as “DY-06 grid-associated node”, node 7 as “DY-07 grid-associated node”, node 18 as “DY-08 grid-associated node”, and node 16 as “DY-09 grid-associated node”.
[0050] Step 3: Create a corresponding relationship table Table 2 Correspondence table
[0051] Step 4: Associate the lines Lines 1-1, 1-2 and 1-3 extending from node 1 are associated with the DY-01 grid and are responsible for supplying power to users within the grid.
[0052] Lines 2-1, 2-2 extending from node 2 serve the DY-02 grid.
[0053] Lines 3-1 and 3-2 extending from node 3 correspond to the DY-03 grid.
[0054] Lines 5-1, 5-2, 5-3, 5-4 extending from node 5 serve the DY-04 grid.
[0055] Lines 10 - 1 , 10 - 2 extending from the node 10 serve the DY-08 grid.
[0056] Lines 12 - 1 , 12 - 2 extending from node 12 serve the DY-06 grid.
[0057] Lines 7-1, 7-2, 7-3 extending from node 7 serve the DY-07 grid.
[0058] Lines 18 - 1 , 18 - 2 extending from node 18 serve the DY-08 grid.
[0059] Lines 16-1, 16-2, 16-3 extending from node 16 serve the DY-09 grid.
[0060] Step 5: Dynamic Update After a period of time, due to urban development, a large factory was added to the DY-01 grid, resulting in an increase in electricity demand. To this end, a new substation was built and a new outgoing node 4 was added to supply power to the factory. At this time, the updated mapping relationship can be referred to in Table 3: Table 3 Updated correspondence table
[0061] Finally, based on the updated results in the above table, a current grid analysis model of the corresponding power supply grid is generated.
[0062] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A main grid topology grid construction method based on OP interconnection, characterized in that: The steps include: S1: Collect distribution network data based on monitoring equipment; S2: Based on the collected distribution network data and OP interconnection results, network modeling and topology design of the distribution network are performed to generate a distribution network main grid topology model, and the generated distribution network main grid topology model is mapped and associated with the homologous maintenance model; S3: Divide the power supply grid based on the loads in different areas of the power supply map combined with the hybrid clustering method; S4: Establish a mapping relationship between the distribution network main grid topology model and the power supply grid to form a current grid analysis model within the power supply grid.
2. The main grid topology grid construction method based on OP interconnection according to claim 1 is characterized in that: In S2, a topological model of the power grid is established based on the collected distribution network data using the OP interconnection results, and the node and connection information is defined and stored in accordance with the OPC standard format, the OPC technical standard is selected, and the server and client are configured; then the established topological model is verified and optimized to obtain the distribution network main grid topological model.
3. The main grid topology grid construction method based on OP interconnection according to claim 2 is characterized in that: The topological model of the power grid determines the nodes and edges of the main grid based on the OP interconnection results, establishes the logical and physical connection relationship between the nodes and the connection lines, and builds the topological structure based on the minimum spanning tree algorithm, as follows: S2.1.1: Define the states, actions, and rewards of the network; S2.1.2: Based on the improved DQN algorithm, learn the behavior pattern of the network and learn the optimal strategy through interaction with the environment; S2.1.3: Dynamically adjust edge weights based on learning results; S2.1.4: Sort all the edges in the network determined based on the OP interconnection results in ascending order of weight; S2.1.5: Create an empty minimum spanning tree T and prepare a union-find data structure to manage the connections between nodes; S2.1.6: Consider each edge in order: S2.1.7: For the edge (u,v) currently under consideration, use union-find to determine whether u and v are already in the same connected component; if not, add ((u,v)) to the minimum spanning tree T and merge u and v into one connected component; S2.1.8: If u and v are already in the same connected component, skip this edge to avoid forming a loop; S2.1.9: When the weight of an edge changes, update the order of the edge list based on the change; S2.1.10: Repeat the above steps until the minimum spanning tree T contains n-1 edges, where n is the number of nodes in the network.
4. The main grid topology grid construction method based on OP interconnection according to claim 3 is characterized in that: The improved DQN algorithm in S2.1.2 uses stochastic gradient descent to update the value function of the neural network, as follows: in, , are the value functions of the neural network before and after the update respectively; is the learning rate; For immediate rewards; is the reward discount factor; is the optimal strategy; is the policy gradient; For the current strategy; Optimize network parameters based on particle swarm dynamic optimization algorithm; The particle swarm dynamic optimization algorithm is specifically as follows: in, , For the Particle No. Vidi The iteration and The speed of iterations; For the The weight of the iteration; , is the learning factor; , for A random number between is the current optimal value; is the global optimal value; , For the Particle No. Vidi The iteration and The position of the iteration; , is the maximum and minimum weight; is the maximum number of iterations.
5. The main grid topology grid construction method based on OP interconnection according to claim 4 is characterized in that: In S2.1.9, a heap or a balanced binary tree is used to update and query the weight changes of edges.
6. The main grid topology grid construction method based on OP interconnection according to claim 5 is characterized in that: In S2.1.9, when the weight of an edge decreases and the edge is already in the minimum spanning tree, reconsider whether the minimum spanning tree condition is still met; If the weight of an edge increases and the edge is in the minimum spanning tree, it is removed from the minimum spanning tree and the minimum spanning tree is rebuilt.
7. The main grid topology grid construction method based on OP interconnection according to claim 6 is characterized in that: The construction of the homology maintenance model in S2 includes the following steps: S2.2.1: Determine the homologous devices in the distribution network based on the distribution network data; S2.2.2: Analyze the operating characteristics and impact of each device and use this to divide the grid into a number of homogeneous groups; S2.2.3: Determine the interactions and communication between homologous groups; S2.2.4: Develop a maintenance plan and strategy for the homologous maintenance model.
8. The main grid topology grid construction method based on OP interconnection according to claim 1 is characterized in that: The hybrid clustering algorithm in S3 is based on the hierarchical clustering algorithm and the K-means algorithm for hybrid clustering, as follows: The distribution network area is divided into several sub-clusters at different levels based on the regional load. In each level of the hierarchical clustering, the initial cluster center point is selected according to the center point. The initial center point selected in the hierarchical clustering stage is used as the initial center point of the k-means algorithm. The k-means algorithm is iteratively optimized based on the initial center point to further refine the clustering results.
9. The main grid topology grid construction method based on OP interconnection according to claim 1 is characterized in that: In S4, the mapping relationship between the main grid topology model of the distribution network and the power supply grid is established as follows: S4.1: Find nodes related to the power grid boundary in the constructed distribution network main grid topology model, and mark the found nodes in the model to indicate the power grid to which the nodes belong; S4.2: Create a corresponding relationship table to record each power grid and the associated topology model node in a one-to-one correspondence; S4.3: Associate the lines extending from these key nodes with the corresponding power supply grids to clarify the grid range served by the lines; S4.4: Dynamically update the mapping relationship according to changes in the distribution network and adjustments to the power supply grid.