Islanded active distribution network power supply restoration method and system based on dynamic mapping of topology graph

Abstract

This invention addresses active distribution networks operating in islanded environments, providing a power restoration method and system based on dynamic mapping of topology graphs. This method enables rapid and accurate fault location, efficient isolation, and graded load restoration, thereby comprehensively improving the speed of power grid fault handling, restoration accuracy, and system reliability. It relates to the field of distribution network automation control technology and includes: each distributed intelligent terminal constructing and updating in real-time a local distribution network topology graph model centered on its own node and showing its connections to adjacent nodes; identifying fault edges and locking their two endpoints in the local distribution network topology graph model; forming a fault isolation region; determining whether the fault isolation region is within a planned island; relaying all connected available tie switch nodes and load nodes to be restored in the local distribution network topology graph models of downstream adjacent nodes; and restoring power supply to the loads along the path.

Description

Technical Field

[0001] This invention relates to the field of power distribution network automation control technology, specifically a method and system for power supply restoration in an active power distribution network with islanded areas based on dynamic mapping of topology maps. Background Technology

[0002] With the large-scale integration of distributed generation (DG), distribution networks have evolved from traditional single-source, unidirectional power flow networks to complex networks with multiple sources and bidirectional power flow. Islanded operation has become one of the important operating states of active distribution networks. Existing distribution network fault recovery solutions have two major drawbacks: First, centralized control relies on master station decisions, and communication delays and computational loads lead to slow fault recovery responses, making it difficult to adapt to the dynamic changes of islanded operation. Second, topology information management uses static table storage, which suffers from problems such as unintuitive information association, delayed updates, and low efficiency in fault location and recovery path planning, failing to meet the multi-scenario, high-reliability power supply requirements of active distribution networks.

[0003] While current distributed feeder automation solutions reduce reliance on the master station, topology information exchange still relies on text-based data transmission, lacking visualization and dynamic mapping capabilities. This leads to redundant calculations in node identification and path searching during fault handling, and insufficient accuracy in island boundary delineation and load restoration priority determination. Therefore, there is an urgent need to construct a power restoration system based on dynamic topology mapping. Through graphical dynamic modeling and distributed collaboration, this system can address the problems of inefficient topology information management, delayed fault handling response, and poor adaptability of power restoration in existing technologies. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a power supply restoration method and system for active distribution networks with islanded operation based on dynamic mapping of topology maps, so as to achieve rapid and accurate fault location, efficient isolation and load hierarchical restoration, thereby comprehensively improving the speed of power grid fault handling, restoration accuracy and system reliability.

[0005] The technical solution adopted by this invention to solve its technical problem is: a power supply restoration method for active distribution networks with islands based on dynamic mapping of topology graphs, comprising the following steps: S1. Based on distributed intelligent terminals deployed at various nodes of the distribution network and forming a peer-to-peer communication network through high-speed optical fibers, each distributed intelligent terminal constructs and updates in real time a local distribution network topology graph model centered on its own node and showing the connection relationships with adjacent nodes; the local distribution network topology graph model uses vertices to represent the node and its adjacent nodes, and edges to represent the connection lines between the node and its adjacent nodes, and the vertices and edges are associated with a set of information that dynamically maps the real-time state and decision attributes of the entities they represent; wherein, the vertices are also associated with switch identifiers; S2. Each distributed intelligent terminal collects real-time current data of each connecting line between this node and adjacent nodes in real time, and associates the current data with the corresponding edge in the local distribution network topology graph model of this node; performs graph matching between the associated current data and the fault criteria preset in the fault feature library, and identifies the fault edge and locks its two vertices in the local distribution network topology graph model of this node according to the matching result. S3. Based on the fault edge and its two vertices, according to the switch identifier associated with the vertex, the distributed intelligent terminal corresponding to the upstream vertex of the fault edge controls the upstream switch to open, and the distributed intelligent terminal corresponding to the downstream vertex of the fault edge controls the downstream switch to open. The state attributes of the corresponding edge are updated in their respective local distribution network topology graph models to form a fault isolation area. S4. The downstream vertex of the fault isolation region determines whether the fault isolation region is within the planned island based on the predefined island boundary identifier and the node identifier within the island. If the downstream vertex of the fault isolation region is not within the planned island, then steps S5 and S6 are executed. If the downstream vertex of the fault isolation region is within the planned island, then the operation ends. S5. Starting from the downstream vertex of the formed fault isolation area, the distributed intelligent terminal of the current node initiates a retrieval request, and sequentially traverses all connected available tie switch nodes and load nodes to be restored in the local distribution network topology map model of the downstream adjacent nodes, generating a retrieval information set containing the identifier, type, load capacity or available capacity and topological location information of each target node; when a downstream island boundary node is found, the island boundary switch is controlled to open through the distributed intelligent terminal where the island boundary node is located, and a corresponding island formation map identifier is generated in the topology map model of the island boundary node and the nodes within the island; S6. Each available tie switch node, based on the retrieved information set, takes the vertex in its local distribution network topology map model as the starting point, plans the restoration path according to the preset optimization principle, and according to the matching result of the path load and its own available capacity, first disconnects the end switch of the path and then closes the tie switch at the beginning of the path to restore the power supply to the load on the path; when the restoration path extends to the predefined island boundary, it controls the closing of the island boundary switch to realize the island grid-connected operation.

[0006] Furthermore, the nodes include at least power nodes, load nodes, distributed power nodes, tie switch nodes, and island boundary nodes and island nodes predefined according to the island operation planning scheme; The decision attributes associated with the vertex include at least the node type, and the decision attributes associated with the edge include at least the line current carrying capacity, switch status, electrical distance, and branch priority.

[0007] Furthermore, in the local power distribution network topology graph model, the visual attributes of the vertices and edges are linked and mapped with the decision attributes; The visual attributes include at least: distinguishing node types by different vertex shapes, mapping switch status by different edge colors, and indicating line current carrying capacity levels by different edge thicknesses.

[0008] Furthermore, the island boundary nodes are identified by cone-shaped vertices, and the island interior nodes are identified by frustum-shaped vertices; The vertex fill color of the nodes within the island dynamically maps to their island operation status or grid-connected operation status.

[0009] Furthermore, the current data includes the effective value of the current and the mutation rate.

[0010] Furthermore, the branch priority increases progressively from the end load node towards the main power source, with smaller values ​​indicating higher priority. The preset optimization principle is to minimize electrical distance and maximize branch priority.

[0011] A power supply restoration system for islanded active distribution networks based on dynamic topology mapping is used to implement the aforementioned power supply restoration method for islanded active distribution networks based on dynamic topology mapping. The system includes: multiple distributed intelligent terminals deployed at various nodes of the distribution network, and a high-speed fiber optic peer-to-peer communication network connecting each of the distributed intelligent terminals; wherein each of the distributed intelligent terminals includes: The topology mapping module is configured to execute step S1; and, The fault recovery decision module is configured to execute steps S2 to S6.

[0012] The beneficial effects of this invention are as follows: By constructing and updating the topology graph of the local distribution network connecting each node and its adjacent nodes in real time, this invention upgrades the distribution network topology information from static table storage to dynamic and visualized digital twins, fundamentally solving the core defects of traditional solutions such as unintuitive information association and delayed updates. Furthermore, by associating real-time current data with corresponding edges in the graph and performing graph-based matching with preset fault criteria, rapid and accurate fault location is achieved. Based on the switch identifiers pre-associated with vertices in the graph, upstream and downstream DTUs collaboratively execute tripping operations, completing fault isolation without master station intervention, completely eliminating centralized control... This invention addresses the risks of communication delays and single-point failures in active distribution networks. For islanded operation scenarios, the system intelligently determines fault locations and dynamically forms islands based on predefined island boundaries and node identifiers within the island, ensuring island operation stability. It traverses and retrieves available power sources and load nodes to be restored from the downstream vertex of the fault isolation area, providing a complete structured input for restoration decisions. Each available tie switch node plans restoration paths in parallel according to preset optimization principles based on the node retrieval information set, and executes switching operations to achieve precise load restoration at different levels. When the restoration path extends to the island boundary, the boundary switch is automatically closed, completing island grid-connected operation. This invention, through the deep integration of graph-based dynamic mapping and distributed collaborative decision-making, systematically solves the three major problems of slow fault response, inefficient topology management, and poor island adaptability in existing technologies, significantly improving the fault self-healing capability, power supply reliability, and operational adaptability of active distribution networks. Attached Figure Description

[0013] Figure 1 This is a flowchart of the present invention; Figure 2 This is the topology diagram of distribution network A. Detailed Implementation

[0014] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0015] As shown in the figure, the present invention provides a method for power supply restoration in an active distribution network with islanding based on dynamic mapping of topology graphs, comprising the following steps: S1. Based on distributed intelligent terminals deployed at various nodes of the distribution network and forming a peer-to-peer communication network through high-speed optical fibers, each distributed intelligent terminal constructs and updates in real time a local distribution network topology graph model centered on its own node and showing the connection relationships with adjacent nodes; the local distribution network topology graph model uses vertices to represent the node and its adjacent nodes, and edges to represent the connection lines between the node and its adjacent nodes, and the vertices and edges are associated with a set of information that dynamically maps the real-time state and decision attributes of the entities they represent; wherein, the vertices are also associated with switch identifiers; S2. Each distributed intelligent terminal collects real-time current data of each connecting line between this node and adjacent nodes in real time, and associates the current data with the corresponding edge in the local distribution network topology graph model of this node; performs graph matching between the associated current data and the fault criteria preset in the fault feature library, and identifies the fault edge and locks its two vertices in the local distribution network topology graph model of this node according to the matching result. S3. Based on the fault edge and its two vertices, and according to the switch identifier associated with the vertex, the distributed intelligent terminal corresponding to the upstream vertex of the fault edge controls the upstream switch to open, and notifies the distributed intelligent terminal corresponding to the downstream vertex of the fault edge through the peer-to-peer communication network; after receiving the notification, the distributed intelligent terminal corresponding to the downstream vertex controls the downstream switch to open; the upstream and downstream terminals update the state attributes of the corresponding edges in their respective local distribution network topology graph models to form a fault isolation area. S4. The downstream vertex of the fault isolation area determines whether the fault isolation area is within the planned island based on the predefined island boundary identifier and the node identifier within the island. If the downstream vertex of the fault isolation area is not within the planned island, then steps S5 and S6 are executed. If the downstream vertex of the fault isolation area is within the planned island, then the operation ends, keeping the original other planned island sub-graph intact, avoiding the accidental removal of the entire island due to a main network failure, and ensuring the power supply continuity of the load within the island. S5. Starting from the downstream vertex of the formed fault isolation area, the distributed intelligent terminal of the current node initiates a retrieval request, and sequentially traverses all connected available tie switch nodes and load nodes to be restored in the local distribution network topology map model of the downstream adjacent nodes, generating a retrieval information set containing the identifier, type, load capacity or available capacity and topological location information of each target node; when a downstream island boundary node is found, the distributed intelligent terminal where the island boundary node is located controls the island boundary switch to open, forming an island, ensuring the continuous, stable and autonomous operation of the island supported by distributed power source, and generating corresponding island formation map identifiers in the topology map model of the island boundary node and the nodes within the island; S6. Each available tie switch node, based on the retrieved information set, takes the vertex in its local distribution network topology map model as the starting point, plans the restoration path according to the preset optimization principle, and according to the matching result of the path load and its own available capacity, first disconnects the end switch of the path and then closes the tie switch at the beginning of the path to restore the power supply to the load on the path; when the restoration path extends to the predefined island boundary, it controls the closing of the island boundary switch to realize the island grid-connected operation.

[0016] In this embodiment of the invention, step S6 further includes: dynamically marking the restoration path and capacity matching information in the local distribution network topology map model where each node on the restoration path is located, so as to transform the abstract path planning and capacity calculation into intuitive visual information for maintenance personnel to verify.

[0017] This invention constructs and updates a topology graph of the local distribution network connecting each node and its adjacent nodes in real time, upgrading the distribution network topology information from static table storage to a dynamic and visualized digital twin. This fundamentally solves the core defects of traditional solutions, such as unintuitive information association and delayed updates. Furthermore, by associating real-time current data with corresponding edges in the graph and performing graph-based matching with preset fault criteria, rapid and accurate fault location is achieved. Based on the switch identifiers pre-associated to vertices in the graph, upstream and downstream DTUs collaboratively execute tripping operations, completing fault isolation without master station intervention and completely eliminating communication delays inherent in centralized control. To mitigate the risk of delayed or single-point-of-failure failures, and for islanded operation scenarios in active distribution networks, the system intelligently determines fault locations and dynamically forms islands based on predefined island boundaries and node identifiers within the island, ensuring island operation stability. Starting from the downstream vertex of the fault isolation area, it traverses and retrieves available power sources and load nodes to be restored, providing a complete structured input for restoration decisions. Each available tie switch node, based on the node retrieval information set, plans restoration paths in parallel according to preset optimization principles and executes switching operations, achieving precise load-level restoration. When the restoration path extends to the island boundary, the boundary switch is automatically closed, completing islanded grid-connected operation. This invention, through the deep integration of graph-based dynamic mapping and distributed collaborative decision-making, systematically solves the three major problems of slow fault response, inefficient topology management, and poor island adaptability in existing technologies, significantly improving the fault self-healing capability, power supply reliability, and operational adaptability of active distribution networks. In addition, each distributed intelligent terminal of the present invention only constructs and updates in real time a local power distribution network topology map model with its own node as the center and the connection relationship with adjacent nodes, rather than constructing a global map. This greatly reduces the data processing volume of each distributed intelligent terminal, and is also conducive to improving processing efficiency and reducing costs.

[0018] In this invention, the nodes include at least power supply nodes, load nodes, distributed power supply nodes, tie switch nodes, and island boundary nodes and island nodes predefined according to the island operation planning scheme; the decision attributes associated with the vertices include at least node type, and the decision attributes associated with the edges include at least line current carrying capacity, switch status, electrical distance, and branch priority.

[0019] In this embodiment of the invention, the visual attributes of vertices and edges are linked and mapped with the decision attributes. The visual attributes include at least: different vertex shapes to distinguish node types, different edge colors to map switch states, and different edge thicknesses to indicate line current carrying capacity levels. In this embodiment, a sphere represents a power supply node, a cube represents a DG node, a cylinder represents a tie switch node, a prism represents a load node, a cone represents an island boundary node, a frustum represents an island node, and a hemisphere represents an end load node; thick edges have a current carrying capacity ≥200A, and thin edges have a current carrying capacity <200A; green indicates that the corresponding line switch is in a closed state, and gray indicates that the corresponding line switch is in an open state; the vertex fill color of an island node dynamically maps its island operation state or grid-connected operation state: transparent color represents island operation state, and red color represents island grid-connected operation state. In this embodiment of the invention, the name and address of the current DTU, the names and addresses of adjacent DTUs, the upstream and downstream relationships with adjacent DTUs, and the branch priority are labeled next to the vertex edge.

[0020] In this invention, current data includes, but is not limited to, the effective value and mutation rate of the current. The fault feature library has built-in short-circuit current thresholds and current mutation rate thresholds for different line types. When the current threshold or current mutation rate threshold exceeds the corresponding preset value, a fault is determined to have occurred, and the faulty edge and its two vertices are then marked and locked. In this embodiment of the invention, a high-brightness visual (red pulse flashing) marker is used to lock the faulty edge and its two vertices, while the remaining nodes remain in a green connected state in the graph, continuously supplying power.

[0021] In some embodiments, the following approach is also adopted: the distributed intelligent terminals corresponding to the upstream and downstream vertices of the fault edge simultaneously perform the corresponding switch opening operation based on the switch identifier associated with each vertex.

[0022] In this embodiment of the invention, the branch priority increases progressively from the end load node towards the main power source, with smaller values ​​indicating higher priority. Specifically, the branch priority is set to 5 levels, with level 1 being the highest; the lowest-level branch has a priority of 1, and the level of the next higher-level branch is increased by 1, with the branch line from the main power source outlet having the highest priority.

[0023] In some embodiments, the preset optimization principle in step S6 is: critical loads take priority, followed by distance; or, distance takes priority, followed by maximum capacity.

[0024] When the branch priority increases progressively from the end load node towards the main power supply, with smaller values ​​indicating higher priority, in this embodiment of the invention, the preset optimization principle is the shortest electrical distance and the highest branch priority. That is, step S6 restores the path according to the shortest electrical distance and the highest branch priority (branch lines from the main power supply outlet). This optimization method, within the allowable capacity, maximizes restoration efficiency by using the shortest path and prioritizing the restoration of the most important loads.

[0025] This invention also provides a power supply restoration system for an active distribution network with islands based on dynamic topology mapping, used to implement the aforementioned power supply restoration method for an active distribution network with islands based on dynamic topology mapping, comprising: multiple distributed intelligent terminals (DTUs) deployed at various nodes of the distribution network, and a high-speed fiber optic peer-to-peer communication network connecting each of the distributed intelligent terminals; wherein each of the distributed intelligent terminals includes: The topology mapping module is configured to execute step S1; and, The fault recovery decision module is configured to execute steps S2 to S6.

[0026] In this embodiment of the invention, the network latency of the high-speed fiber optic peer-to-peer communication network is ≤20ms, the transmission rate of the distributed intelligent terminal is ≥100Mbps, and the response time is ≤10ms.

[0027] Example: Methods for restoring power supply to distribution network A. Figure 2 This is the topology diagram of distribution network A.

[0028] S1. Based on distributed intelligent terminals deployed at various nodes of the distribution network and forming a peer-to-peer communication network through high-speed optical fibers, each distributed intelligent terminal constructs and updates in real time a local distribution network topology map model centered on its own node and showing the connection relationships with adjacent nodes; the local distribution network topology map model uses vertices to represent the node and its adjacent nodes, and edges to represent the connection lines between the node and its adjacent nodes, and the vertices and edges are associated with a set of information that dynamically maps the real-time state and decision attributes of the entities they represent; wherein, the vertices are also associated with switch identifiers. In the diagram, spherical vertices mark power supply nodes, cube vertices mark DG1 nodes (DTU31), cylindrical vertices mark tie switch nodes (DTU11, DTU14, DTU22, DTU30), prism vertices mark load nodes (DTU1, DTU2, DTU3, DTU4, DTU5, DTU6, DTU7, DTU8, DTU9, DTU10, DTU12, DTU13, DTU15, DTU16, DTU17, DTU21, DTU23, DTU24, DTU25), conical vertices mark island boundary nodes (DTU18, DTU20, DTU27), frustum vertices mark island inner nodes (DTU19, DTU29), and hemispherical vertices mark end load nodes (DTU26, DTU28). Green edges indicate closed line switches, and gray edges indicate open line switches. Except for the tie switch outlet, all other lines are green.

[0029] S2. Each distributed intelligent terminal collects real-time current data of each connecting line between its own node and adjacent nodes, and associates the current data with the corresponding edge in the local distribution network topology graph model of its own node; the associated current data is then matched with the fault criteria preset in the fault feature library in a graph-based manner, and the fault edge is identified and its two vertices are locked in the local distribution network topology graph model of its own node according to the matching result. In the figure, DTU1 collects the current data at both ends of line L1-2 through the current detection unit, which are 18kA (upstream) and 17.8kA (downstream), respectively, with a current change rate of 8kA / ms, and marks the current characteristic value of the edge in the graph; the topology graph mapping module calls the fault feature library to determine whether the short-circuit fault condition is met (fault condition: short-circuit current threshold > 12kA or current change rate > 5kA / ms), and line L1-2 in the graph flashes with red pulses, locking the two vertices of the fault, DTU1 and DTU2; S3 and DTU1 issue commands to disconnect the upstream switch on the fault side, and DTU2 issues commands to disconnect the downstream switch on the fault side. In the diagram, line L1-2 (fault side) changes from green to gray. The upstream nodes of the fault (DTU1, DT8-DTU11) remain in a green connected state and continue to supply power.

[0030] S4 and DTU2 determine the location relationship between the fault isolation area and the planned island based on the predefined island boundary identifier and the node identifier within the island. If the map shows that the faulty line L1-2 is not within the planned island, then steps S5 and S6 are executed.

[0031] S5. Starting from the DTU2 vertex, the distributed intelligent terminal of the current node initiates a retrieval request, relaying the retrieval of all connected available tie switch nodes and load nodes to be restored in the local distribution network topology map model of the downstream adjacent nodes. Simultaneously, a retrieval information set is generated, containing the identifier, type, load capacity or available capacity, and topological location information of each target node. In the diagram, the downstream nodes are traversed along the blue trajectory, retrieving tie switch vertices DTU14, DTU22, and DTU30, DG1 vertex DTU31, and end load nodes DTU26 and DTU28. The downstream island boundary node DTU18 is retrieved. Following the downstream island boundary node sequence, island boundary vertices DTU18, DTU20, and DTU27 are automatically highlighted, and switches K1817, K2021, and K2728 are disconnected, forming an island (DTU18-DTU20, DTU27, DTU29).

[0032] S6 and DTU14 plan the green trajectory restoration path in the diagram (DTU14→DTU13→DTU12→DTU2→DTU3→DTU4→DTU5→DTU6), restoring DTU13 (load 3MVA), DTU12 (load 4MVA), DTU2 (load 4MVA), DTU3 (load 5MVA), DTU4 (load 6MVA), DTU5 (load 5MVA), and DTU6 (load 6MVA) in sequence, with a remaining capacity of 7MVA. Since DTU7 has already been restored in the green trajectory path of DTU30, the green trajectory restoration path of DTU14 is terminated. Switch K1413 is closed, and switch K00607 is opened, and the loads on the green trajectory path are restored to power supply. DTU22 plans a green track restoration path in the diagram (DTU22→DTU21→DTU20→DTU27→DTU28→DTU18), restoring DTU21 (load 5MVA) and DTU28 (load 1MVA) in sequence. The remaining capacity is 1MVA. Since the remaining capacity is insufficient to restore the load power supply of DTU17, the green track restoration path of DTU22 is terminated. Switches K2221, K2021, and K2728 are closed, while switch K1817 remains open. The loads on the green track path are restored to power supply and the islanded system is connected to the grid. DTU30 plans a green trajectory restoration path in the diagram (DTU30→DTU7→DTU23→DTU24→DTU25→DTU26), restoring DTU7 (load 2MVA), DTU23 (load 2MVA), DTU24 (load 3MVA), DTU25 (load 5MVA), and DTU26 (load 7MVA) in sequence, with a remaining capacity of 1MVA. Since all adjacent nodes of the load nodes on the green trajectory path are already in the green trajectory restoration path, the green trajectory restoration path of DTU30 is terminated. Switch K3017 is closed and switch K0706 is opened, and the loads on the green trajectory path are restored to power supply. Ultimately, only DTU17 (load 5MVA) failed to restore power because it exceeded the remaining capacity of the tie switch; the other 23 load nodes were powered normally.

Claims

1. A method for power supply restoration in active distribution networks with islanding based on dynamic mapping of topology graphs, characterized in that, Includes the following steps: S1. Based on distributed intelligent terminals deployed at various nodes of the distribution network and forming a peer-to-peer communication network through high-speed optical fibers, each distributed intelligent terminal constructs and updates in real time a local distribution network topology graph model centered on its own node and showing the connection relationships with adjacent nodes; the local distribution network topology graph model uses vertices to represent the node and its adjacent nodes, and edges to represent the connection lines between the node and its adjacent nodes, and the vertices and edges are associated with a set of information that dynamically maps the real-time state and decision attributes of the entities they represent; wherein, the vertices are also associated with switch identifiers; S2. Each distributed intelligent terminal collects real-time current data of each connecting line between this node and adjacent nodes in real time, and associates the current data with the corresponding edge in the local distribution network topology graph model of this node; performs graph matching between the associated current data and the fault criteria preset in the fault feature library, and identifies the fault edge and locks its two vertices in the local distribution network topology graph model of this node according to the matching result. S3. Based on the fault edge and its two vertices, according to the switch identifier associated with the vertex, the distributed intelligent terminal corresponding to the upstream vertex of the fault edge controls the upstream switch to open, and the distributed intelligent terminal corresponding to the downstream vertex of the fault edge controls the downstream switch to open. The state attributes of the corresponding edge are updated in their respective local distribution network topology graph models to form a fault isolation area. S4. The downstream vertex of the fault isolation region determines whether the fault isolation region is within the planned island based on the predefined island boundary identifier and the node identifier within the island. If the downstream vertex of the fault isolation region is not within the planned island, then steps S5 and S6 are executed. If the downstream vertex of the fault isolation region is within the planned island, then the operation ends. S5. Starting from the downstream vertex of the formed fault isolation area, the distributed intelligent terminal of the current node initiates a retrieval request, and sequentially traverses all connected available tie switch nodes and load nodes to be restored in the local distribution network topology map model of the downstream adjacent nodes, generating a retrieval information set containing the identifier, type, load capacity or available capacity and topological location information of each target node; when a downstream island boundary node is found, the island boundary switch is controlled to open through the distributed intelligent terminal where the island boundary node is located, and a corresponding island formation map identifier is generated in the topology map model of the island boundary node and the nodes within the island; S6. Each available tie switch node, based on the retrieved information set, takes the vertex in its local distribution network topology map model as the starting point, plans the restoration path according to the preset optimization principle, and according to the matching result of the path load and its own available capacity, first disconnects the end switch of the path and then closes the tie switch at the beginning of the path to restore the power supply to the load on the path; when the restoration path extends to the predefined island boundary, it controls the closing of the island boundary switch to realize the island grid-connected operation.

2. The power supply restoration method for active distribution networks with islanding based on dynamic mapping of topology graphs as described in claim 1, characterized in that, The nodes include at least power nodes, load nodes, distributed power nodes, tie switch nodes, and island boundary nodes and island nodes predefined according to the island operation planning scheme. The decision attributes associated with the vertex include at least the node type, and the decision attributes associated with the edge include at least the line current carrying capacity, switch status, electrical distance, and branch priority.

3. The power supply restoration method for active distribution networks with islanding based on dynamic mapping of topology graphs as described in claim 2, characterized in that, In the local power distribution network topology model, the visual attributes of the vertices and edges are linked and mapped to the decision attributes; The visual attributes include at least: distinguishing node types by different vertex shapes, mapping switch status by different edge colors, and indicating line current carrying capacity levels by different edge thicknesses.

4. The power supply restoration method for active distribution networks with islanding based on dynamic mapping of topology graphs as described in claim 3, characterized in that, The island boundary nodes are identified by cone-shaped vertices, and the island internal nodes are identified by frustum-shaped vertices. The vertex fill color of the nodes within the island dynamically maps to their island operation status or grid-connected operation status.

5. The power supply restoration method for active distribution networks with islanding based on dynamic mapping of topology graphs as described in claim 1, characterized in that, The current data includes the effective value of the current and the mutation rate.

6. The power supply restoration method for active distribution networks with islanding based on dynamic mapping of topology graphs as described in claim 2, characterized in that, The branch priority increases progressively from the end load node toward the main power source, with smaller values ​​indicating higher priority. The preset optimization principle is to minimize electrical distance and maximize branch priority.

7. A power supply restoration system for an active distribution network with islanding based on dynamic mapping of topology graphs, used to implement the power supply restoration method for an active distribution network with islanding based on dynamic mapping of topology graphs as described in any one of claims 1 to 6, characterized in that, include: Multiple distributed intelligent terminals deployed at various nodes of the power distribution network, and a high-speed fiber optic peer-to-peer communication network connecting each of the distributed intelligent terminals; wherein each of the distributed intelligent terminals includes: The topology mapping module is configured to execute step S1; and, The fault recovery decision module is configured to execute steps S2 to S6.

Images