A power distribution network fault active reconstruction method based on priority dynamic extension graph

Abstract

The application discloses a power distribution network fault active reconstruction method based on a priority dynamic continuation graph and belongs to the technical field of power distribution network fault recovery. The method first decomposes and maps the topology, load, operation state and recovery resource information of the power distribution network after a fault into a power grid prism mirror through power grid prism mapping, and calculates the effective abundance unified quantization resource support capability. Then, a priority dynamic continuation graph is constructed, with a recovery priority index as a node weight and a correlation between recovery units as an edge. Finally, the optimal recovery chain is searched, the flexible soft switch and the tie switch are coordinated to cooperatively perform reconstruction, and the power grid prism mirror and the continuation graph are dynamically updated after each recovery action. The application realizes unified expression of the state after a fault, dynamic evaluation of the recovery priority and collaborative optimization of multiple resources, and improves the efficiency and adaptability of the power distribution network fault recovery.

Description

Technical Field

[0001] This invention relates to a method for proactive reconfiguration of distribution network faults based on priority dynamic extension diagrams, belonging to the field of distribution network fault recovery technology. Background Technology

[0002] With the widespread integration of new energy power generation, distributed power sources, energy storage systems, and flexible interconnection equipment, the power distribution network is gradually developing into an active power distribution system with multi-element interaction between source, grid, load, and storage. The recovery process after a fault faces problems such as complex operating conditions, diverse recovery resources, and increased decision-making difficulty.

[0003] Existing distribution network fault recovery methods primarily rely on topology reconfiguration after fault location, restoring power supply through optimized switch combinations. These methods typically prioritize restoring load, reducing network losses, or minimizing the number of switch operations, making it difficult to fully utilize the recovery capabilities of flexible resources such as Standard Operating Procedures (SOPs) and distributed energy sources. Furthermore, post-fault distribution networks contain multi-dimensional state information, including topology, equipment, loads, and recovery resources. Existing methods lack a unified information representation mechanism and struggle to comprehensively consider the importance, recovery capabilities, recovery costs, and operational risks of recovery units, leading to inappropriate recovery path selection.

[0004] Existing research typically processes the above information separately. For example, electrical constraints are described using DistFlow equations (e.g., CN110350508B, "A Method for Constructing a Unified Model for Active Distribution Network Fault Recovery"), radial topological constraints are described using spanning tree models, or the Distribution Group (DG) is simplified to PQ nodes (e.g., CN116882139B, "An Active Distribution Network Component Repair Method and System Considering Network Reconfiguration and Rolling Optimization for Fault Repair"). While these methods can establish mathematical models, the various types of information are scattered across different constraints, lacking a unified and structured virtual state representation layer. This makes it difficult for subsequent decision-making processes to comprehensively and dynamically perceive the true availability of restored resources. In particular, when resources have uncertainties (e.g., communication failures, operational failure probabilities) or network reachability is limited by line capacity and voltage constraints, existing models struggle to quantify these factors into indicators that can be directly used for decision-making.

[0005] Therefore, a novel active reconfiguration technology is needed to achieve complete expression of the power grid state after a fault, dynamic evaluation of the recovery unit, and coordinated control of multiple types of recovery resources, so as to improve the efficiency and capability of distribution network fault recovery. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for active reconfiguration of distribution network faults based on priority dynamic extension graphs. By driving the active reconfiguration strategy through priority dynamic extension graphs, the method achieves coordinated optimization of multiple types of recovery resources and network topology, thereby improving the efficiency of distribution network recovery after a fault.

[0007] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0008] This invention provides a method for proactive fault reconfiguration in distribution networks based on priority dynamic extension diagrams, comprising the following steps:

[0009] Step S1: Collect power grid resource information of the distribution network after the fault;

[0010] Step S2: Construct a grid prism mapping mechanism for fault recovery, and generate a post-fault grid prism image based on the grid resource information of the distribution network after the fault.

[0011] Step S3: Based on the power grid prism mirror image, divide the recovery units and calculate the recovery priority index of each recovery unit. Using the recovery unit as a node, the recovery priority index as the node weight, and the relationship between the recovery units as the edge, construct a priority dynamic extension graph.

[0012] Step S4: Based on the priority dynamic extension graph, generate and optimize the recovery chain to obtain the optimal recovery chain;

[0013] Step S5: Perform recovery actions sequentially according to the optimal recovery chain to carry out active reconstruction.

[0014] The effects achieved by the above settings are as follows: In the above technical solution, step S1 provides a complete data foundation for all subsequent decisions by collecting power grid resource information of the distribution network after a fault; step S2 decomposes, maps, and merges multi-source heterogeneous power grid resource information through a power grid prism mapping mechanism to form a unified virtual power grid image, solving the problem of the lack of a unified expression mechanism for multi-dimensional state information of the distribution network after a fault in the existing technology; step S3 constructs a priority dynamic extension graph based on the virtual power grid image, using the recovery priority index as the node weight and the correlation between recovery units as the edge, transforming the traditional recovery order decision that relies on fixed rules or a single load level into a structured graph model, realizing the dynamic evaluation of recovery objects and the explicit expression of recovery relationships; step S4 generates and optimizes the recovery chain based on the priority dynamic extension graph, guiding the search direction through node weight constraints and edge constraints, avoiding redundant calculations caused by traversing all topology combinations in the traditional method, and improving decision-making efficiency; step S5 executes active reconstruction according to the optimal recovery chain, coordinating the coordinated action of flexible soft switches and tie switches, and unifying flexible power regulation and network topology adjustment into the execution framework. The above five steps form a complete technical chain from information collection, unified state expression, structured modeling, efficient decision-making to collaborative execution. This overcomes the shortcomings of existing technologies, such as fragmented state expression, static priority evaluation, and low search efficiency, and enables rapid, efficient, and adaptive proactive reconfiguration of the distribution network after a fault.

[0015] Furthermore, the method also includes:

[0016] Step S6: After each recovery action at the head of the optimal recovery chain is completed, return to step S1 to re-collect the current distribution network status resource information, and execute steps S2 to S5 in sequence to remap the power grid prism image, update the priority dynamic extension map, and generate a new optimal recovery chain until all recoverable areas are restored or the preset termination conditions are met.

[0017] The above setup achieves the following effect: The technical solution implements closed-loop dynamic updates in the recovery process through step S6. During distribution network fault recovery, after each recovery action, the network topology, load power supply status, and availability of remaining recovery resources all change, and the recovery conditions for subsequent areas to be recovered also change accordingly. Step S6 immediately re-collects the current state information after each recovery action, re-executes the prism mapping, and updates the priority dynamic extension graph, ensuring that subsequent recovery decisions are always based on the latest grid state, thus improving the recovery strategy's adaptability to the time-varying and uncertain nature of the grid state. Simultaneously, this rolling optimization mechanism ensures that only the first action of the current optimal recovery chain is executed each time, rather than the entire chain, allowing the optimization window to always move forward, achieving true dynamic adaptive reconfiguration.

[0018] Furthermore, in step S1, the power grid resource information includes topology resource information, load demand resource information, operating status resource information, and recovery resource information;

[0019] The topology resource information describes the network structure;

[0020] The load demand resource information describes the restoration needs of the power outage area;

[0021] The operational status resource information describes the power grid operational constraints;

[0022] The recovery resource information describes the flexible adjustment capability that can participate in fault recovery.

[0023] The above setup achieves the following effects: The technical solution categorizes post-fault distribution network resource information into four types: topology, load demand, operating status, and recovery resources, thus realizing hierarchical decoupling of complex grid conditions. By collecting multi-dimensional heterogeneous information in a categorized manner, topology reflects network connectivity, load reflects recovery demand, operating status reflects safety boundaries, and recovery resources reflect adjustability. This also provides a standardized information input format for subsequent decomposition and fusion in prism mapping. This classification method ensures the completeness and structure of information collection, avoiding decision-making biases caused by mixed or missing information.

[0024] Furthermore, in step S2, a grid prism mapping mechanism for fault recovery is constructed, which generates a post-fault grid prism image based on the grid resource information of the distribution network after the fault, including:

[0025] The topology resource information, load demand resource information, operation status resource information and recovery resource information of the distribution network after the fault are decomposed into topology resource units, load demand resource units, operation status resource units and recovery resource units, respectively.

[0026] The different resource units after separation are mapped to the virtual space to establish a one-to-one mapping relationship between the actual physical power grid resources and the virtual resources, realize the corresponding expression of different types of power grid resources in the virtual space, and form virtual topology resources, virtual load demand resources, virtual operating status resources and virtual recovery resources.

[0027] For each power outage area, calculate the effective abundance of recoverable resources;

[0028] The virtual topology resources, virtual load demand resources, virtual operating status resources, virtual recovery resources, and the effective abundance of recovery resources in each power outage area are integrated to generate a power grid prism image that corresponds to the actual distribution network state after the fault.

[0029] The above setup achieves the following effects: The proposed technical solution utilizes a power grid prism mapping mechanism, drawing inspiration from the principle of a prism decomposing and reconstructing complex optical information. First, the overall power grid state is decomposed into four independent resource units based on attributes, eliminating coupling between different information types. Then, each actual resource unit is mapped one-to-one into a virtual space, establishing a unified digital representation for different types of information. Finally, various virtual resources are fused and reconstructed with the effective abundance of recovery resources to form a complete virtual power grid image. This mechanism enables a complete, unified, and structured digital representation of the power grid's topology, load demand, operational constraints, and resource capabilities after a fault in the virtual space, providing a quantifiable state model foundation for subsequent recovery priority calculation and recovery chain search.

[0030] Furthermore, the effective abundance of the recovery resources is equal to the sum of the products of the equivalent recovery capacity of all available recovery resources in the region and the corresponding resource availability probability and network reachability coefficient, divided by the power demand to be recovered in the region.

[0031] Define the effective abundance of recovery resources for the i-th recovery unit at time t. for:

[0032] ;

[0033] Among them, R i (t) represents the set of candidate recovery resources that can support the i-th recovery unit; V r(t) represents the equivalent restored resource value of the r-th resource; p r (t)∈[0,1] represents the probability of resource availability; η ir (t)∈[0,1] represents the network reachability coefficient of resource to recovery unit i; D i (t) represents the power demand to be restored by the recovery unit; ε is a positive number to prevent the denominator from being zero;

[0034] Resource availability probability is determined by a combination of device health availability probability, communication success probability, and operation success probability; network reachability coefficient is determined based on topology connectivity, remaining line capacity, voltage constraints, and power flow security constraints.

[0035] The equivalent restored resource value V r (t) is determined according to the following rules:

[0036] For flexible soft switches, the remaining adjustable power range is taken as the equivalent recovery resource value;

[0037] For tie switches, the smaller value between the remaining upstream power supply capacity and the allowable transfer capacity of the recovery path is taken as the equivalent recovery resource value;

[0038] For distributed power sources, their current available output is taken as the equivalent recovery resource value;

[0039] For energy storage, the smaller of its power limit and the sustainable discharge power within a given recovery duration is taken as the equivalent recovery resource value.

[0040] The above settings achieve the following effect: The technical solution defines an effective abundance index for recovery resources, uniformly converting the capacity, availability probability, and network reachability of different types of recovery resources into the expected deliverable recovery capacity relative to a unit of recovery demand. This index has a clear physical meaning: when =0 indicates that there are no available resources; 0< <1 indicates insufficient resource support capacity; =1 indicates that resources are basically sufficient to meet the recovery needs; A value >1 indicates the existence of resource redundancy. This indicator unifies different types and characteristics of recovery resources (such as continuously adjustable SOPs, interconnecting switches with only switching states, distributed power sources with fluctuating output, and energy storage with limited capacity) under the same quantitative framework, solving the problem in existing technologies where multiple types of recovery resources lack unified quantitative standards and are difficult to compare and coordinate in the same decision-making model.

[0041] Further, step S3: Based on the power grid prism mirror image, divide the area into recovery units and calculate the recovery priority index for each recovery unit. Using the recovery unit as a node, the recovery priority index as the node weight, and the relationships between recovery units as edges, construct a priority dynamic extension graph, including:

[0042] Based on the power grid prism mirror image, the distribution network is divided into several recovery units; each recovery unit is a power supply area with independent recovery needs and capabilities after a fault.

[0043] For each recovery unit, indicators related to recovery decisions are extracted from the power grid prism mirror image, including the unit's importance, recoverability, recovery cost, and operational safety status. The importance indicator describes the unit's need for power restoration; the recoverability indicator, determined based on the availability of recovery resources, describes the expected support capacity of those resources for the region; the recovery cost indicator describes the number of operations, resource consumption, and recovery difficulty required during the recovery process; and the operational safety status indicator describes the impact of the recovery plan on node voltage, line power flow, and system safety constraints.

[0044] The recovery priority index for each recovery unit is calculated by combining the aforementioned importance index, recoverability index, recovery cost index, and operational safety status index.

[0045] Using the recovery unit as a node, the recovery priority index as the node weight, and the association between recovery units as edges, a priority dynamic extension graph is constructed.

[0046] The above setup achieves the following effects: By dividing the complex overall distribution network restoration problem into restoration units, the technical solution decomposes it into multiple sub-problems with independent restoration needs and capabilities, reducing decision-making complexity. Based on this, four types of restoration characteristic indicators are extracted from the power grid prism mirror image, comprehensively reflecting the importance, restoreability, restoration cost, and operational safety of the restoration unit, overcoming the deficiency of traditional methods that rely on a single factor for restoration priority evaluation. The four types of indicators are integrated into a comprehensive quantitative value through a restoration priority index, enabling comparable priority dimensions between different restoration units and providing a scientific numerical basis for subsequent node weighting in the graph model.

[0047] Furthermore, the recovery priority index for each recovery unit is calculated by comprehensively considering the importance index, recoverability index, recovery cost index, and operational safety status index, including:

[0048] The recovery priority index is expressed as:

[0049]

[0050] in, This represents the recovery priority index of the i-th recovery unit; Indicators representing the importance of recovery units; This indicates the recoverability index of the recovery unit; Indicates the cost of recovery; Indicates operational safety indicators; , , , These represent the weights of the corresponding indicators;

[0051] Among them, the importance index of the recovery unit is used to measure the value of load recovery in different areas, and is determined according to the load level, user type and power supply demand; the recovery cost index is used to measure the operational complexity and resource consumption during the recovery process; and the operation safety index is used to evaluate the system safety margin after the recovery plan is implemented.

[0052] The recoverability index is used to measure the support level of different recovery resources for a recovery unit, and is derived from the effective abundance of recovery resources. The recoverability index is defined as follows:

[0053] C i (t) = min{1, A i (t)}.

[0054] Furthermore, using the recovery unit as a node, the recovery priority index as the node weight, and the association relationships between recovery units as edges, a priority dynamic extension graph is constructed, including:

[0055] Based on the recovery priority index corresponding to each recovery unit, a dynamic priority extension diagram is constructed to describe the priority relationship and recovery dependency relationship between different recovery units; the dynamic priority extension diagram is represented as follows:

[0056] ;

[0057] Where N represents the set of recovery unit nodes, E represents the set of recovery dependency edges, and W represents the set of node weights;

[0058] In the priority dynamic extension graph, each node corresponds to a recovery unit, and the node weight is determined by the recovery priority index of that recovery unit:

[0059] ;

[0060] Where n represents the number of recovery units;

[0061] The edges in the priority dynamic extension graph are used to describe the relationships between recovery units. These relationships include topological connections, power supply path dependencies, and recovery resource support relationships. When any of these relationships exist between two recovery units, a connection edge is established between the corresponding nodes, thereby forming a structured recovery model with recovery priorities and dependencies. The recovery relationship edges are represented as follows:

[0062] ;

[0063] in, This represents the associated edge between recovery unit i and recovery unit j; Indicates topological relationships; Indicates power supply path dependency; This indicates the restoration of resource support relationships; This represents a multi-factor fusion function; if any relationship exists, an associated edge is established.

[0064] Among them, topology relationship is used to describe the network connection between recovery units; power supply path dependency relationship is used to describe the power supply path dependency of a certain recovery unit on other recovery units during the recovery process; recovery resource support relationship is used to describe the support relationship of recovery resources on different recovery units.

[0065] The above setup achieves the following effects: The technical solution provides a specific method for constructing a priority dynamic extension graph. Node weights are dynamically determined by the recovery priority index, reflecting which recovery unit is prioritized for recovery in the current state. Edges are connected through three types of relationships (topology connection, power supply path dependency, and resource support), reflecting the constraints of which recovery depends on which recovery unit. Compared with existing technologies, this graph model no longer treats the recovery order decision as an isolated priority ranking problem, but explicitly models the interdependencies between recovery objects as the topology of a graph. This allows the search and optimization of recovery paths to be performed under graph constraints, fundamentally avoiding the problem of invalid recovery paths caused by ignoring dependencies in traditional methods.

[0066] Furthermore, in step S4, based on the priority dynamic extension graph, a recovery chain is generated and optimized, including:

[0067] The node weights and edge connections in the priority dynamic extension graph are analyzed to generate candidate recovery chains that satisfy the node weight constraints and edge constraints. Each candidate recovery chain consists of multiple recovery actions in sequence.

[0068] Each candidate recovery chain is evaluated using a comprehensive evaluation function, and the candidate recovery chain that maximizes the value of the comprehensive evaluation function is selected as the optimal recovery chain.

[0069] The above setup achieves the following effects: The technical solution uses a priority-driven dynamic extension graph as the structured constraint space for recovery chain search. Recovery chains no longer rely on blindly traversing all topology reconstruction combinations. Instead, within the effective search space defined by the graph model, the search direction is guided by node weights, and the feasible path range is limited by edge constraints, generating a sequence of candidate recovery chains that satisfy priority order and dependencies. A comprehensive evaluation function is used to quantitatively evaluate the candidate chains from multiple dimensions, including recovery effect, recovery efficiency, recovery cost, and operational safety, and the optimal one is selected as the execution plan. Compared with existing technologies, this approach transforms the exponential topology combination search problem into a linear chain search problem under graph constraints, significantly reducing the search space, improving decision-making efficiency, and ensuring the overall optimality of the recovery plan.

[0070] Furthermore, the node weights and edge connections in the priority dynamic extension graph are analyzed to generate candidate recovery chains that satisfy the node weight constraints and edge constraints. Each candidate recovery chain consists of multiple recovery actions in sequence, including:

[0071] Obtain the node weights and edge connections in the priority dynamic extension graph, and generate all recovery chains that satisfy the node weight constraints and edge constraints as candidate recovery chains;

[0072] The node weight constraint is used to guide the recovery direction, indicating that in the recovery chain, the recovery priority index of the node corresponding to the recovery action that is ranked earlier is greater than the recovery priority index of the node corresponding to the recovery action that is ranked later.

[0073] The edge constraint is used to describe the recovery sequence conditions between different nodes, indicating that there must be a connecting edge in the priority dynamic extension graph between the two recovery units corresponding to adjacent recovery actions in the recovery chain.

[0074] The above setup achieves the following effects: The technical solution clarifies two core constraints in the candidate recovery chain generation process. Node weight constraints ensure that the recovery direction of the recovery chain always points to the recovery unit with the highest recovery value, allowing high-priority objects to obtain recovery resources first. Edge constraints ensure that there are actual physical or logical dependencies (topological connectivity, power supply path dependence, or resource support) between adjacent recovery actions in the recovery chain, avoiding the generation of invalid recovery schemes that do not meet actual power supply logic. These two types of constraints work together to ensure that the recovery chain search process follows both the priority-guided optimal direction and strictly satisfies the physical constraints and resource support conditions of power grid operation. This achieves rapid convergence of recovery paths while ensuring the feasibility of the scheme, further improving the efficiency of recovery decision-making.

[0075] Furthermore, each candidate recovery chain is evaluated using a comprehensive evaluation function, and the candidate recovery chain that maximizes the value of the comprehensive evaluation function is selected as the optimal recovery chain, including:

[0076] Based on the recovery effect, recovery cost, and operational safety constraints, different candidate recovery chains are comprehensively evaluated, and the optimal recovery chain that meets the constraints is selected:

[0077] ;

[0078] in, Indicates the optimal recovery chain. This represents the comprehensive evaluation function for the recovery chain;

[0079] Among them, the recovery chain comprehensive evaluation function The comprehensive evaluation function equation for different recovery chain schemes is as follows:

[0080]

[0081] in, As a metric for recovery effectiveness, the value of a recovery unit is measured using a recovery priority index. This is a recovery efficiency metric used to describe the speed at which recovery actions are executed; The recovery cost metric is used to describe the costs of switching operations and flexible resource allocation. To establish operational safety indicators and constrain the power grid's operational status after restoration; by maximizing the comprehensive evaluation function of the restoration chain, the optimal restoration chain that satisfies both restoration requirements and safety constraints is obtained. These represent the coefficients for each item.

[0082] Furthermore, in step S5, recovery actions are executed sequentially according to the optimal recovery chain to perform active reconstruction, including:

[0083] Recovery actions are executed sequentially according to the optimal recovery chain, with each recovery action coordinating power regulation by the flexible soft switch and topology adjustment by the tie switch;

[0084] The power adjustment of the flexible soft switch and the topology adjustment of the tie switch are executed in the same recovery action. That is, according to the instructions of the optimal recovery chain, the open and closed state of the tie switch is changed while adjusting the transmission power of the flexible soft switch, so as to achieve joint optimization of recovery resource adjustment and network structure reconstruction.

[0085] The above setup achieves the following effect: In this invention, each recovery action, according to the instructions of the optimal recovery chain, simultaneously adjusts the flexible transmission power of the SOP and changes the opening and closing state of the tie switch, enabling flexible power support and network structure adjustment to take effect jointly within the same time frame. This coordinated execution method can fully leverage the advantages of the SOP's fast response speed and high control precision, compensating for instantaneous power shortages through power regulation while adjusting the topology structure, avoiding transient power imbalances caused by a single topology adjustment, and improving the dynamic safety and continuity of power restoration during the recovery process.

[0086] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0087] (1) A power grid prism mapping mechanism is proposed. By decomposing, mapping and fusing multi-source heterogeneous resource information of the distribution network, and proposing an effective abundance index for recovery resources, the unified quantification of recovery resource capability, availability probability and network reachability is realized. A power grid prism mirror state model after the fault is constructed to provide state support for active recovery.

[0088] (2) A recovery priority index is proposed and a priority dynamic extension diagram is constructed to realize the dynamic evaluation of recovery units after failure and the structured expression of recovery relationship, providing a new model basis for recovery decision-making.

[0089] (3) A priority extension graph-driven active reconfiguration strategy is proposed. Recovery path optimization is achieved through recovery chain search, and multiple types of recovery resources such as SOP and tie switches are coordinated to improve the efficiency of power distribution network fault recovery. Attached Figure Description

[0090] Figure 1 It is the overall flowchart;

[0091] Figure 2 It is a power grid prism mapping and priority extension diagram. Detailed Implementation

[0092] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0093] Example 1:

[0094] This embodiment provides a method for proactive fault reconfiguration in distribution networks based on a priority dynamic extension graph, such as... Figure 1 As shown, it includes the following steps:

[0095] Step S1: Collect power grid resource information of the distribution network after the fault;

[0096] Step S2: Construct a grid prism mapping mechanism for fault recovery, and generate a post-fault grid prism image based on the grid resource information of the distribution network after the fault.

[0097] Step S3: Based on the power grid prism mirror image, divide the recovery units and calculate the recovery priority index of each recovery unit. Using the recovery unit as a node, the recovery priority index as the node weight, and the relationship between the recovery units as the edge, construct a priority dynamic extension graph.

[0098] Step S4: Based on the priority dynamic extension graph, generate and optimize the recovery chain to obtain the optimal recovery chain;

[0099] Step S5: Perform recovery actions sequentially according to the optimal recovery chain to carry out active reconstruction.

[0100] Implementation principle: This invention constructs a power grid prism mapping mechanism to decompose, map and fuse multi-dimensional resource information of the distribution network after a fault, and quantifies resource capability, availability probability and network reachability by restoring the effective abundance of resources.

[0101] By establishing a recovery priority index and constructing a priority dynamic extension diagram, the priority relationship of recovery units can be dynamically expressed. Furthermore, by driving the active reconfiguration strategy through the priority dynamic extension diagram, the collaborative optimization of multiple types of recovery resources such as SOPs and tie switches with the network topology can be achieved, thereby improving the efficiency of distribution network recovery after a fault.

[0102] This embodiment uses a power grid prism mapping mechanism to decompose, map, and fuse multi-source heterogeneous power grid resource information to form a unified virtual power grid image. This solves the problem of the lack of a unified expression mechanism for multi-dimensional state information of the distribution network after a fault in the existing technology. Based on the virtual power grid image, a priority dynamic extension graph is constructed. With the recovery priority index as the node weight and the correlation between recovery units as the edge, the traditional recovery order decision that relies on fixed rules or a single load level is transformed into a structured graph model, realizing the dynamic evaluation of recovery objects and the explicit expression of recovery relationships.

[0103] In this embodiment, step S4 generates and optimizes the recovery chain based on the priority dynamic extension graph. By guiding the search direction through node weight constraints and edge constraints, it avoids redundant calculations caused by the traditional method of traversing all topological combinations, thus improving decision-making efficiency.

[0104] In this embodiment, step S5 performs active reconfiguration according to the optimal recovery chain. By coordinating the coordinated actions of the flexible soft switch and the tie switch, flexible power regulation and network topology adjustment are integrated into the execution framework.

[0105] The above five steps form a complete technical chain from information collection, unified state expression, structured modeling, efficient decision-making to collaborative execution. This overcomes the shortcomings of existing technologies, such as fragmented state expression, static priority evaluation, and low search efficiency, and enables rapid, efficient, and adaptive proactive reconfiguration of the distribution network after a fault.

[0106] Example 2:

[0107] This embodiment proposes a method for proactive fault reconfiguration in distribution networks based on a priority dynamic extension graph, such as... Figure 1 and Figure 2 As shown, it includes the following steps:

[0108] Step 1: Construct a grid prism mapping mechanism for fault recovery to generate the post-fault grid prism mirror state. The specific steps are as follows:

[0109] (1) Power Grid Resource Deconstruction. Addressing the issue of multiple types of resource information and coupling relationships between different resources in the post-fault distribution network, the key resource information in the actual distribution network is first decomposed, breaking down the complex power grid state into resource units with independent physical meaning. Based on fault recovery requirements, distribution network resources are divided into topology resources, load demand resources, operational status resources, and recovery resources. Topology resources describe the network structure state after a fault, load demand resources describe the recovery needs of different regions, operational status resources describe power grid operational constraints, and recovery resources describe the flexible adjustment capabilities that can participate in fault recovery. After resource deconstruction, the post-fault distribution network state is represented as:

[0110]

[0111] Where G represents the overall state of the distribution network after the fault. , , , These represent topology resources, load demand resources, operating status resources, and recovery resources, respectively.

[0112] (2) Resource Information Mapping. For the different deconstructed resource units, a one-to-one mapping relationship is established between actual physical power grid resources and virtual resources, realizing the corresponding representation of different types of power grid resources in virtual space. The resource mapping process is represented as follows:

[0113]

[0114] in, This represents the i-th type of resource unit in the actual power grid. This represents the mapped virtual resource unit. This represents the resource mapping relationship. Through resource mapping, the topology, load, operation, and recovery resource states in the actual power grid are mapped to the virtual space, enabling the power grid prism mirror to reflect the actual operation state after a fault.

[0115] (3) Calculation of effective abundance of restored resources. In order to ensure that the prism mapping not only describes the resource category and state, but also reflects the actual support degree of each type of resource for a specific restoration unit, the effective abundance of restored resources of the i-th restoration unit at time t is defined. for:

[0116]

[0117] Among them, R i (t) represents the set of candidate recovery resources that can support the i-th recovery unit; V r (t) represents the equivalent restored resource value of the r-th resource; p r (t)∈[0,1] represents the probability of resource availability; η ir (t)∈[0,1] represents the network reachability coefficient of resource to recovery unit i; D i (t) represents the power demand to be restored by the recovery unit; ε is a very small positive number to prevent the denominator from being zero. The resource availability probability can be determined by combining the device health availability probability, communication success probability, and operation success probability; the network reachability coefficient is determined based on topology connectivity, line remaining capacity, voltage constraints, and power flow security constraints.

[0118] Different types of resources are converted using a unified equivalent recovery resource value: SOP is taken as its remaining adjustable power range, tie switches are taken as the smaller value between the remaining upstream power supply capacity and the allowable transfer capacity of the recovery path, distributed generation is taken as the current available output, and energy storage is taken as the smaller value between the power limit and the sustainable discharge power within a given recovery duration. This forms a recovery resource abundance with equal components for SOP, tie switches, distributed generation, and energy storage.

[0119] The effective abundance of restored resources reflects the expected deliverable restoration capacity corresponding to the unit restoration demand.

[0120] When A i When (t)=0, it means that there are no effective resources available for this recovery unit;

[0121] When 0 i When (t) < 1, it indicates that the expected resource support capacity is insufficient;

[0122] When A i When (t)=1, it indicates that the expected resource support capacity basically meets the recovery needs;

[0123] When A i ​When (t)>1, it indicates that there is a certain amount of recovery resource redundancy.

[0124] This indicator is dynamically updated based on fault status, network topology, and resource operating status, and serves as an important status attribute of the virtual recovery resource layer.

[0125] (4) Power Grid Prism Mirror Fusion and Reconstruction. After mapping various resource units, fusion and reconstruction are performed based on the relationships between different resources, recombining multiple virtual resource units to form the post-fault power grid prism mirror. The power grid prism mirror is represented as:

[0126]

[0127] Where G* represents the grid prism image formed by grid prism mapping, V top V load V state V res These represent virtual topology resources, virtual load demand resources, virtual operating status resources, and virtual recovery resources, respectively. F represents the multidimensional resource fusion relationship. The recovery resource abundance spectrum, which consists of the effective abundance of recovery resources and its components, is fused along with the virtual recovery resource layer as an attribute.

[0128] Integrated reconfiguration maps the topology, load demand, operational status, and recovery resource information of the distribution network after a fault into virtual topology, virtual load, virtual operational status, and virtual recovery resource information, respectively. Secondly, based on the distribution recovery unit number, a correspondence is established between different types of virtual information, and the effective abundance of recovery resources for each recovery unit is associated with the corresponding virtual recovery resource information. Finally, the above virtual information is combined according to the correspondence to generate a grid prism image at time t.

[0129] By integrating and reconstructing virtual resources, a grid prism mirror image corresponding to the actual distribution network state after a fault is formed, achieving a unified expression of fault state, recovery requirements, recovery capabilities, and resource redundancy.

[0130] Step 2: Based on the mirrored state of the power grid prism, construct a priority dynamic extension graph model. The specific steps are as follows:

[0131] (1) Determining Recovery Units. Based on the post-fault grid prism mirror state formed by grid prism mapping, the distribution network is divided into recovery units according to the fault impact range, network topology, and distribution of recovery resources. A recovery unit is a power supply area with independent recovery needs and capabilities after a fault. It can represent either the load area where power loss occurred or the target area to be restored through network reconfiguration, SOP power regulation, or distributed resource support. By dividing the network into recovery units, the complex distribution network recovery process is transformed into a priority decision-making problem among multiple recovery units.

[0132] (2) Extracting recovery feature information. For each recovery unit, information related to recovery decision-making is extracted from the power grid prism mapping model, including the importance of the recovery unit, its recoverability, recovery cost, and operational safety status. Among them, the importance is used to describe the degree of demand for power supply restoration of the recovery unit; the recoverability is determined according to the effective abundance of recovery resources and is used to describe the expected support capacity of existing SOPs, tie switches, distributed energy resources, and energy storage resources for the region; the recovery cost is used to describe the number of operations, resource consumption, and recovery difficulty required during the recovery process; and the operational safety status is used to describe the impact of the recovery plan on node voltage, line power flow, and system safety constraints after execution.

[0133] (3) Establish a recovery priority index. Based on the above recovery characteristic information, a recovery priority index is established as a unified evaluation index in the fault recovery process. The recovery priority index comprehensively considers the impact of different recovery factors on the recovery value and recovery difficulty, and realizes the dynamic quantification of the recovery priority of different recovery units.

[0134] The recovery priority index is expressed as:

[0135]

[0136] in, This represents the recovery priority index of the i-th recovery unit; Indicators representing the importance of recovery units; This indicates the recoverability index of the recovery unit; Indicates the cost of recovery; Indicates operational safety indicators; , , , These represent the weights of the corresponding indicators.

[0137] Among them, the importance index of the recovery unit is used to measure the value of load recovery in different areas, which can be determined according to the load level, user type and power supply demand; the recoverability index is used to measure the support level of different recovery resources for the recovery unit, and is obtained by normalizing the effective abundance of recovery resources; the recovery cost index is used to measure the operational complexity and resource consumption during the recovery process; and the operation safety index is used to evaluate the system safety margin after the recovery plan is implemented.

[0138] To ensure that the effective abundance of restored resources maintains the same dimension as other evaluation indicators in the restoration priority index, the recoverability index C will be... i (t) is defined as:

[0139] C i (t) = min{1, A i (t)}

[0140] Because A i (t) can be greater than 1, therefore C needs to be controlled. i (t) is between [0,1].

[0141] By restoring the priority index, the traditional single load priority ranking is transformed into a multi-factor integrated evaluation, enabling the recovery unit to dynamically adjust the recovery priority according to changes in fault status and resource allocation.

[0142] (4) Construction of the priority dynamic extension graph. Based on the recovery priority index corresponding to each recovery unit, a priority dynamic extension graph model is constructed to describe the priority relationship and recovery dependency relationship between different recovery units. The priority dynamic extension graph is represented as follows:

[0143]

[0144] Where N represents the set of recovery unit nodes, E represents the set of recovery dependency edges, and W represents the set of node weights.

[0145] In the priority dynamic extension graph, each node corresponds to a recovery unit, and the node weight is determined by the recovery priority index of that recovery unit:

[0146]

[0147] Where n represents the number of recovery units.

[0148] The edges in the graph describe the relationships between recovery units, including topological connections, power path dependencies, and recovery resource support. When two recovery units are associated with recovery conditions, connecting edges are established between the corresponding nodes, thus forming a structured recovery model with recovery priorities and dependencies. Recovery relationship edges are represented as follows:

[0149]

[0150] in, This represents the associated edge between recovery unit i and recovery unit j; Indicates topological relationships; Indicates power supply path dependency; This indicates the restoration of resource support relationships; This represents a multi-factor fusion function.

[0151] Among them, the topology relationship is used to describe the network connection between recovery units; the power supply path dependency relationship is used to describe the power supply path dependency of a certain recovery unit on other recovery units during the recovery process; and the recovery resource support relationship is used to describe the support relationship of recovery resources such as SOP, tie switches and distributed energy to different recovery units.

[0152] Through the above process of constructing node weights and associated edges, the priority dynamic extension graph can simultaneously express the importance of recovery units, recovery order, and resource dependencies, transforming the traditional recovery problem based on topology traversal into a structured decision problem based on recovery priority relationships, thus providing a model foundation for subsequent recovery chain generation and proactive reconstruction.

[0153] Step 3: Based on the priority dynamic extension graph, generate and optimize the recovery chain, coordinate multiple types of recovery resources to perform active reconstruction, and dynamically update the power grid prism mirror state and the priority dynamic extension graph during the recovery process. The specific steps are as follows:

[0154] (1) Priority Dynamic Extension Graph Analysis and Recovery Target Determination. Based on the priority dynamic extension graph model, the recovery units after the fault are analyzed, and the priority recovery units are determined according to the node weights and the recovery dependencies between nodes. Unlike the traditional method of determining the recovery order based on load level or fixed rules, this invention comprehensively considers the importance, recoverability, and recovery constraints of the recovery units through the priority dynamic extension graph, so that the recovery target can be dynamically adjusted with changes in fault state and resource allocation.

[0155] Based on the node weights in the priority dynamic extension graph, the recovery units are sorted from largest to smallest according to their recovery priority index:

[0156]

[0157]

[0158] in, This represents the priority sorting sequence of recovery units; n represents the number of recovery units. This represents the recovery unit located at the nth position after sorting; Represents recovery unit The corresponding recovery priority index;

[0159] By analyzing the priority dynamic extension diagram, the recovery sequence relationship between different recovery units is determined, providing a basis for decision-making in the generation of subsequent recovery paths.

[0160] (2) Recovery chain generation and optimization. Based on the node relationships and edge connections in the priority dynamic extension graph, a recovery chain consisting of multiple recovery actions is constructed. The recovery chain is used to describe the relationship between the selection of recovery units, the recovery order, and the configuration of recovery resources during the fault recovery process.

[0161] Sorted by recovery unit priority sequence The initial recovery chain is generated as follows:

[0162]

[0163] Where L represents the recovery chain, A m This represents the recovery action corresponding to the m-th recovery unit node.

[0164] During the recovery chain generation process, node weight constraints and edge constraints are set based on the node weights and edge connections in the priority-dynamic extension graph. Node weight constraints guide the recovery direction, while edge constraints limit the feasible range of recovery paths, thus avoiding the generation of numerous invalid recovery schemes caused by traditional methods that rely solely on network topology for combined searches.

[0165] Node weights are determined by a recovery priority index. During the recovery chain generation process, the priority direction of recovery units is determined based on the node weights, ensuring that the recovery process prioritizes nodes with higher recovery value for extended search. The node weight constraint is expressed as:

[0166]

[0167] in, and These represent the recovery priority indices of recovery unit i and recovery unit j, respectively. When recovery unit i has a higher recovery priority, node i is given priority during the recovery chain generation process, thus avoiding the problem of blindly searching according to line number or topological distance in traditional methods.

[0168] Edge constraints are used to describe the order of recovery between different recovery units. Since some recovery areas cannot be directly restored through switching operations and require power channels, power support, or recovery resources from other areas, recovery chains cannot be arbitrarily connected and must satisfy the edge constraint relationships in the priority dynamic extension graph. Recovery path constraints are represented as follows:

[0169]

[0170] in, This indicates that there is a valid recovery association between recovery unit i and recovery unit j, and E represents the set of edges in the priority dynamic extension graph. When there is no valid association between two recovery units, it is prohibited to directly establish a recovery path in the recovery chain.

[0171] Applying both node weight constraints and edge constraints to the recovery chain search process yields candidate recovery chains:

[0172] ;

[0173] Where L represents the candidate recovery chain, Let N represent the recovery actions corresponding to node i and node j, N represent the set of recovery unit nodes, and E represent the set of recovery dependencies.

[0174] Through the aforementioned constraint mechanism, the recovery chain generation process no longer traverses all possible topology reconfiguration combinations. Instead, it searches for feasible recovery paths based on recovery value and recovery relationships within the effective search space defined by the priority dynamic extension graph. On one hand, node weight constraints ensure that the recovery process prioritizes high-value recovery units; on the other hand, edge constraints ensure that the recovery path meets the actual power supply logic and resource support conditions, thereby reducing invalid topology searches and improving the efficiency of fault recovery decision-making.

[0175] Based on the recovery effect, recovery cost, and operational safety constraints, different candidate recovery chains are comprehensively evaluated, and the optimal recovery chain that meets the constraints is selected:

[0176]

[0177] in, Indicates the optimal recovery chain. This represents the comprehensive evaluation function for the recovery chain.

[0178] Among them, the recovery chain comprehensive evaluation function This function is used to evaluate the overall performance of different recovery chain solutions. It comprehensively considers recovery effectiveness, recovery efficiency, recovery cost, and operational safety. The specific equation for the comprehensive evaluation function of the recovery chain is as follows:

[0179]

[0180] in, As a metric for recovery effectiveness, the value of a recovery unit is measured using a recovery priority index. This is a recovery efficiency metric used to describe the speed at which recovery actions are executed; The recovery cost metric is used to describe the costs of switching operations and flexible resource allocation. To establish operational safety indicators and constrain the post-recovery grid operating state, the optimal recovery chain is obtained by maximizing the comprehensive evaluation function of the recovery chain, thus satisfying both recovery requirements and safety constraints.

[0181] (3) Resource recovery and network topology coordinated reconstruction. Based on the recovery action sequence determined by the optimal recovery chain, recovery resources such as SOPs and tie switches are uniformly incorporated into the active reconstruction process to achieve coordinated optimization of recovery resource adjustment and network topology adjustment.

[0182] Specifically, the SOP (Standard Operating Procedure) improves the power restoration capability of the power-loss area by changing the power support relationship between fault areas through flexible power regulation; the tie switch reconfigures the power supply path by changing the network connection relationship. Through the coordinated control of these two types of restoration resources, the restoration process is no longer limited to a single topology adjustment, but forms a comprehensive restoration mode that combines flexible resource regulation and network structure reconstruction.

[0183] (4) Dynamic updates during active reconfiguration. Due to the dynamic nature of the fault recovery process, the distribution network topology, load power supply status, and recovery resource status will all change after partial recovery actions are performed. Therefore, the power grid prism mapping status is updated again based on the results of the recovery actions, and the recovery priority index and priority dynamic extension diagram are further updated to achieve dynamic adjustment of the recovery scheme.

[0184] After each recovery action at the head of the optimal recovery chain is completed, return to step one to re-collect the current distribution network status resource information, and then execute steps two to three in sequence to remap the power grid prism image, update the priority dynamic extension map, and generate a new optimal recovery chain until all recoverable areas are restored or the preset termination condition (such as the maximum number of iterations) is reached.

[0185] The optimal recovery chain is used to provide the order of actions in the current state, but only the first recovery action of the chain is executed at a time. After the action is completed, the system immediately returns to S1, remaps the power grid state, updates the extension graph, and generates a new optimal recovery chain, without executing all the actions of the original recovery chain at once.

[0186] The update method for the priority dynamic extension graph is as follows:

[0187] Because the distribution network fault recovery process is a continuous evolution, the original fault state will change after partial recovery actions are performed. This includes changes in the power supply status of restored areas, reorganization of network topology, changes in the availability of restored resources, and changes in the recovery conditions of remaining power-loss areas. Therefore, a priority dynamic extension diagram built based on a fixed fault state cannot continuously reflect the priority relationship between recovery units at the current recovery stage, and the priority dynamic extension diagram needs to be updated in real time according to the recovery process.

[0188] After each recovery action is completed, the grid prism mapping is re-performed based on the new grid operating status. The topology resources, load demand resources, operating status resources, and recovery resource information in the grid prism image are updated. The effective abundance of recovery resources for each recovery unit is recalculated, and the recovery priority index for each recovery unit is recalculated based on the updated resource status and effective abundance.

[0189]

[0190] in, This represents the recovery priority index of the i-th recovery unit in the (t+1)-th recovery phase; This represents the updated image of the power grid prism. This indicates the relevant feature information of the recovery unit; This represents the function for calculating the recovery priority index.

[0191] Based on the new recovery priority index, the node weights in the priority dynamic extension graph are updated. Simultaneously, the associated edges in the priority dynamic extension graph are dynamically adjusted according to changes in topology, power supply paths, and recovery resource configuration.

[0192]

[0193] in, This represents the priority dynamic extension graph corresponding to the (t+1)th recovery stage; , , These represent the updated set of nodes, the set of edges, and the set of node weights, respectively.

[0194] Through the above dynamic update process, the priority dynamic extension diagram can be continuously adjusted as the fault recovery process progresses, realizing the real-time change expression of recovery unit priority, recovery dependency relationship and recovery resource support relationship. This avoids the problem of fixed execution after one optimization in traditional recovery strategies, and improves the adaptability and effectiveness of the proactive refactoring process.

[0195] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0196] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0197] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0198] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0199] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for proactive fault reconfiguration in distribution networks based on priority dynamic extension diagrams, characterized in that, Includes the following steps: Step S1: Collect power grid resource information of the distribution network after the fault; Step S2: Construct a grid prism mapping mechanism for fault recovery, and generate a post-fault grid prism image based on the grid resource information of the distribution network after the fault. Step S3: Based on the power grid prism mirror image, divide the recovery units and calculate the recovery priority index of each recovery unit. Using the recovery unit as a node, the recovery priority index as the node weight, and the relationship between the recovery units as the edge, construct a priority dynamic extension graph. Step S4: Based on the priority dynamic extension graph, generate and optimize the recovery chain to obtain the optimal recovery chain; Step S5: Perform recovery actions sequentially according to the optimal recovery chain to carry out active reconstruction.

2. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 1, characterized in that, The method further includes: Step S6: After the first recovery action of each optimal recovery chain is completed, return to step S1 to re-collect the current distribution network status resource information, and execute steps S2 to S5 in sequence to remap the power grid prism image, update the priority dynamic extension map and generate a new optimal recovery chain until all recoverable areas are restored or the preset termination conditions are met.

3. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 1, characterized in that, In step S1, the power grid resource information includes topology resource information, load demand resource information, operating status resource information, and recovery resource information; The topology resource information describes the network structure; The load demand resource information describes the restoration needs of the power outage area; The operational status resource information describes the power grid operational constraints; The recovery resource information describes the flexible adjustment capability that can participate in fault recovery.

4. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 3, characterized in that, In step S2, a grid prism mapping mechanism for fault recovery is constructed, which generates a post-fault grid prism image based on the grid resource information of the distribution network after the fault, including: The topology resource information, load demand resource information, operation status resource information and recovery resource information of the distribution network after the fault are decomposed into topology resource units, load demand resource units, operation status resource units and recovery resource units, respectively. The different resource units after separation are mapped to the virtual space to establish a one-to-one mapping relationship between the actual physical power grid resources and the virtual resources, realize the corresponding expression of different types of power grid resources in the virtual space, and form virtual topology resources, virtual load demand resources, virtual operating status resources and virtual recovery resources. For each power outage area, calculate the effective abundance of recoverable resources; The virtual topology resources, virtual load demand resources, virtual operating status resources, virtual recovery resources, and the effective abundance of recovery resources in each power outage area are integrated to generate a power grid prism image that corresponds to the actual distribution network state after the fault.

5. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 4, characterized in that, The effective abundance of recovery resources is equal to the sum of the products of the equivalent recovery capacity of all available recovery resources in the region and the corresponding resource availability probability and network reachability coefficient, divided by the power demand to be recovered in the region. Define the effective abundance of recovery resources for the i-th recovery unit at time t. for: ； Among them, R i (t) represents the set of candidate recovery resources that can support the i-th recovery unit; V r (t) represents the equivalent restored resource value of the r-th resource; p r (t)∈[0,1] represents the probability of resource availability; η ir (t)∈[0,1] represents the network reachability coefficient of resource to recovery unit i; D i (t) represents the power demand to be restored by the recovery unit; ε is a positive number to prevent the denominator from being zero; Resource availability probability is determined by a combination of device health availability probability, communication success probability, and operation success probability; network reachability coefficient is determined based on topology connectivity, remaining line capacity, voltage constraints, and power flow security constraints. The equivalent restored resource value V r (t) is determined according to the following rules: For flexible soft switches, the remaining adjustable power range is taken as the equivalent recovery resource value; For tie switches, the smaller value between the remaining upstream power supply capacity and the allowable transfer capacity of the recovery path is taken as the equivalent recovery resource value; For distributed power sources, their current available output is taken as the equivalent recovery resource value; For energy storage, the smaller of its power limit and the sustainable discharge power within a given recovery duration is taken as the equivalent recovery resource value.

6. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 1, characterized in that, Step S3: Based on the power grid prism mirror image, divide the area into recovery units and calculate the recovery priority index for each recovery unit. Using the recovery unit as a node, the recovery priority index as the node weight, and the relationships between recovery units as edges, construct a priority dynamic extension graph, including: Based on the power grid prism mirror image, the distribution network is divided into several recovery units; each recovery unit is a power supply area with independent recovery needs and capabilities after a fault. For each recovery unit, indicators related to recovery decisions are extracted from the power grid prism mirror image, including the importance index, recoverability index, recovery cost index, and operational safety status index of the recovery unit. Among them, the importance index describes the degree of demand for power restoration of the recovery unit; the recoverability index is determined based on the effective abundance of recovery resources and is used to describe the expected support capacity of resources for the region; the recovery cost index describes the number of operations, resource consumption, and recovery difficulty required during the recovery process; and the operational safety status index describes the impact of the recovery plan on node voltage, line power flow, and system safety constraints after execution. The recovery priority index for each recovery unit is calculated by combining the aforementioned importance index, recoverability index, recovery cost index, and operational safety status index. Using the recovery unit as a node, the recovery priority index as the node weight, and the association between recovery units as edges, a priority dynamic extension graph is constructed.

7. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 6, characterized in that, The recovery priority index for each recovery unit is calculated by combining the aforementioned importance indicators, recoverability indicators, recovery cost indicators, and operational safety status indicators, including: The recovery priority index is expressed as: ； in, This represents the recovery priority index of the i-th recovery unit; Indicators representing the importance of recovery units; This indicates the recoverability index of the recovery unit; Indicates the cost of recovery; Indicates operational safety indicators; , , , These represent the weights of the corresponding indicators; Among them, the importance index of the recovery unit is used to measure the value of load recovery in different areas, and is determined according to the load level, user type and power supply demand; the recovery cost index is used to measure the operational complexity and resource consumption during the recovery process; and the operation safety index is used to evaluate the system safety margin after the recovery plan is implemented. The recoverability index is used to measure the support level of different recovery resources for a recovery unit, and is derived from the effective abundance of recovery resources. The recoverability index is defined as follows: C i (t) = min{1, A i (t)}。 8. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 7, characterized in that, Using the recovery unit as a node, the recovery priority index as the node weight, and the association relationships between recovery units as edges, a priority dynamic extension graph is constructed, including: Based on the recovery priority index corresponding to each recovery unit, a dynamic priority extension diagram is constructed to describe the priority relationship and recovery dependency relationship between different recovery units; the dynamic priority extension diagram is represented as follows: ； Where N represents the set of recovery unit nodes, E represents the set of recovery dependency edges, and W represents the set of node weights; In the priority dynamic extension graph, each node corresponds to a recovery unit, and the node weight is determined by the recovery priority index of that recovery unit: ； Where n represents the number of recovery units; The edges in the priority dynamic extension graph are used to describe the relationships between recovery units. These relationships include topological connections, power supply path dependencies, and recovery resource support relationships. When any of these relationships exist between two recovery units, a connection edge is established between the corresponding nodes, thereby forming a structured recovery model with recovery priorities and dependencies. The recovery relationship edges are represented as follows: ； in, This represents the associated edge between recovery unit i and recovery unit j; Indicates topological relationships; Indicates power supply path dependency; This indicates the restoration of resource support relationships; This represents a multi-factor fusion function; if any relationship exists, an associated edge is established. Among them, topology relationship is used to describe the network connection between recovery units; power supply path dependency relationship is used to describe the power supply path dependency of a certain recovery unit on other recovery units during the recovery process; recovery resource support relationship is used to describe the support relationship of recovery resources on different recovery units.

9. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 1, characterized in that, In step S4, based on the priority dynamic extension graph, a recovery chain is generated and optimized, including: The node weights and edge connections in the priority dynamic extension graph are analyzed to generate candidate recovery chains that satisfy the node weight constraints and edge constraints. Each candidate recovery chain consists of multiple recovery actions in sequence. Each candidate recovery chain is evaluated using a comprehensive evaluation function, and the candidate recovery chain that maximizes the value of the comprehensive evaluation function is selected as the optimal recovery chain. The node weight constraint is used to guide the recovery direction, indicating that in the recovery chain, the recovery priority index of the node corresponding to the recovery action that is ranked earlier is greater than the recovery priority index of the node corresponding to the recovery action that is ranked later. The edge constraint is used to describe the recovery sequence conditions between different nodes, indicating that there must be a connecting edge in the priority dynamic extension graph between the two recovery units corresponding to adjacent recovery actions in the recovery chain.

10. The active fault reconfiguration method for distribution networks based on priority dynamic extension diagrams according to claim 9, characterized in that, Each candidate recovery chain is evaluated using a comprehensive evaluation function, and the candidate recovery chain that maximizes the value of the comprehensive evaluation function is selected as the optimal recovery chain, including: Based on the recovery effect, recovery cost, and operational safety constraints, different candidate recovery chains are comprehensively evaluated, and the optimal recovery chain that meets the constraints is selected: ； in, Indicates the optimal recovery chain. This represents the comprehensive evaluation function for the recovery chain; Among them, the recovery chain comprehensive evaluation function The comprehensive evaluation function equation for different recovery chain schemes is as follows: ； in, As a metric for recovery effectiveness, the value of a recovery unit is measured using a recovery priority index. This is a recovery efficiency metric used to describe the speed at which recovery actions are executed; The recovery cost metric is used to describe the costs of switching operations and flexible resource allocation. To establish operational safety indicators and constrain the power grid's operational status after restoration; by maximizing the comprehensive evaluation function of the restoration chain, the optimal restoration chain that satisfies both restoration requirements and safety constraints is obtained. These represent the coefficients for each item.

Images