A power distribution network key node identification and adaptive island division method considering disaster influence

Abstract

The application discloses a kind of considering disaster influence's distribution network key node identification and adaptive island division method, belong to distribution network resilience promotion and post-disaster recovery field. Including: establish the line fault probability model of distribution network under extreme disaster influence, obtain the fault probability of each line in disaster scene;Comprehensive importance index is constructed to node;According to the comprehensive importance of node, the node of distribution network is sorted, and the key node under disaster scene is identified;With key node power supply guarantee and system loss load minimum as target, build distribution network adaptive island division model, obtain island division scheme;According to actual fault state and resource available state after disaster, island structure, distributed power output, energy storage charging and discharging power and load recovery sequence are dynamically optimized.The method can improve the power supply guarantee capability of key node and important load of distribution network under extreme disaster conditions.

Description

Technical Field

[0001] This invention relates to the field of distribution network resilience enhancement and post-disaster recovery, specifically to a method for identifying critical nodes and isolating distribution networks that takes into account the impact of extreme disasters. Background Technology

[0002] In recent years, with the intensification of global climate change, extreme weather events such as typhoons, rainstorms, and lightning strikes have become more frequent, posing a serious threat to the safe and stable operation of the power system. As a key link directly supplying power to users, the distribution network is more prone to line faults, equipment damage, and large-scale power outages under the impact of extreme disasters, making it one of the most vulnerable parts of the power system.

[0003] Under extreme weather conditions, power distribution line faults exhibit significant spatiotemporal uncertainty, with their probability closely related to various factors such as wind speed, rainfall intensity, and geographical environment. For example, in typhoon disasters, fault types such as line breaks and tower collapses show a non-linear growth trend with increasing wind speed, potentially triggering cascading faults and leading to large-scale power outages. Therefore, identifying critical nodes in the system before or during a disaster and implementing targeted control measures is one of the core issues in enhancing the resilience of power distribution networks.

[0004] To address the aforementioned issues, existing research mainly focuses on two aspects: Firstly, methods for identifying critical nodes in distribution networks, typically based on network topology or electrical parameters, using indicators such as node degree, betweenness, or voltage sensitivity to assess node importance; secondly, methods for islanding and operating distribution networks, which, through network reconfiguration or distributed generation support, divide the system into multiple independently operating islands to ensure power supply to critical loads. However, existing technologies still have the following shortcomings: (1) Most key node identification methods are based on static topology and do not fully consider the impact of extreme disasters on the probability of line failure, making it difficult to reflect the true importance of nodes in disaster scenarios; (2) Islanding methods usually aim at power restoration or network reconstruction, but lack a coordination mechanism with the results of critical node identification, resulting in insufficient protection of critical loads in islanding results; (3) Most existing methods separate the pre-disaster analysis, operation during the disaster and post-disaster recovery process, lack a unified resilience enhancement framework, and are difficult to adapt to the dynamic evolution of extreme weather.

[0005] With the large-scale integration of distributed power sources, energy storage devices, and flexible loads, traditional distribution networks are gradually transforming into active distribution networks. These systems possess islanding and self-healing capabilities, providing new technical means for power restoration under disaster conditions. However, the complexity of the system structure and operation mode has increased significantly, making the collaborative optimization of critical node identification and islanding in disaster environments more complex. Therefore, it is necessary to propose a method for critical node identification and islanding in distribution networks that considers the impact of extreme disasters. This method organically combines disaster modeling, node importance assessment, and islanding to enhance the resilience of the distribution network throughout the pre-disaster, during-disaster, and post-disaster processes, thereby improving the reliability of power supply to critical loads and reducing the risk of system load loss. Summary of the Invention

[0006] This invention addresses the problems in existing methods for improving the resilience of distribution networks, such as the independent identification of critical nodes and islanding, insufficient characterization of disaster impacts, and a lack of dynamic coordination in post-disaster recovery strategies. It proposes a method for identifying critical nodes and islanding in distribution networks that considers the impact of disasters. This method establishes a line fault probability model under extreme disasters, constructs a comprehensive node importance index, identifies critical nodes in disaster scenarios, and forms an adaptive islanding scheme based on these critical nodes. This improves the critical load protection capability and post-disaster recovery capability of the distribution network under extreme weather conditions.

[0007] This invention is achieved through the following technical solution: This invention discloses a method for identifying critical nodes and isolating distribution networks considering disaster impacts, comprising the following steps: Step 1) Modeling the probability of power grid failure under extreme disasters: Obtain meteorological data, power grid topology data, line parameters and tower parameters for extreme disasters, establish the mapping relationship between disaster intensity and power line failure probability, and obtain the failure probability of each line under the action of disaster.

[0008] Step 2) Calculation of the overall importance of nodes considering the impact of disasters: Based on the line fault probability, node load importance, node power supply support capacity and node topology location, a node overall importance index is constructed to quantitatively assess the importance of distribution network nodes in disaster scenarios.

[0009] Step 3) Identification of key nodes in the distribution network: Based on the comprehensive importance index of the nodes, the nodes in the distribution network are sorted, and the set of key nodes is determined by threshold screening or sorting screening.

[0010] Step 4) Construction of distribution network islanding model based on key nodes: With the goal of ensuring power supply to key nodes, and combining distributed power sources, energy storage and controllable load resources, construct an optimized distribution network islanding model to realize active islanding in disaster scenarios.

[0011] Step 5) Dynamic reconstruction and recovery optimization of isolated networks after disaster: Based on the actual fault status and resource availability after the disaster, the isolated structure is dynamically adjusted, and the output of distributed power sources, the charging and discharging power of energy storage and the order of load recovery are optimized to achieve coordinated recovery of the distribution network after the disaster.

[0012] As a further aspect of the present invention, step 1) In extreme weather disaster scenarios, the relationship between disaster intensity and line fault probability is first established. Taking typhoon disasters as an example, the typhoon wind speed field can be determined by the typhoon center location, maximum wind speed radius, maximum wind speed, and typhoon movement speed. For lines in the distribution network... ij At any given moment t The wind speed received is recorded as: In the formula, V ij,t For the line ij At any moment t The equivalent wind speed; r ij,t For the line ij Distance to the center of the typhoon; R t The radius of the typhoon's maximum wind speed; V max,t This is the maximum wind speed of the typhoon.

[0013] Furthermore, the mechanical stress on the line is calculated based on the wind load: In the formula, For the line ij At any moment t The wind load it receives; k w This is the wind load factor; D ij The diameter of the wire; θ ij,t It is the angle between the wind direction and the line direction.

[0014] The total load of the line is expressed as: In the formula, G ij This refers to the gravity load generated by the line's own weight.

[0015] The wind-induced stress along the line is: In the formula, k g This is the load conversion factor.

[0016] A line fault occurs when the wind-induced stress exceeds the line's tensile strength. Assuming the line's tensile strength follows a normal distribution, the line fault probability can be expressed as: Further written as: In the formula, For the line ij At any moment t The probability of failure; For the tensile strength of the line; and These represent the mean and standard deviation of the line's tensile strength, respectively.

[0017] Alternatively, a data-driven approach can be used to establish the relationship between wind speed and line failure rate, and the XGBoost model can be trained using historical wind speeds, line operating status, and fault records. In the formula, F XGBoost (⋅) represents the trained failure rate prediction model; Ω ij,t It is a set of characteristic variables including line type, terrain conditions, tower parameters, and historical fault frequency.

[0018] This step enables the conversion of disaster meteorological data into line fault probability, providing disaster risk input for subsequent identification of key nodes.

[0019] As a further aspect of the present invention, step 2) After obtaining the line failure probability, a node disaster risk index is constructed. For nodes... i The set of associated lines is denoted as The node disaster risk index is: In the formula, R i For nodes i Disaster risk indicators; α ij For the line ij For nodes i Influence weight; For the line ij The probability of failure.

[0020] The node load importance index is defined as follows: In the formula, L i For nodes i The load importance index; λi Assigning node load level weights; For nodes i At any moment t Active load.

[0021] The node power supply support capability index is defined as follows: In the formula, For nodes i Power supply support capability; For nodes i The output of the connected distributed power source; and These are the energy storage discharge power and the charging power, respectively.

[0022] The topological importance of nodes can be characterized by node degree, betweenness, or electrical distance. For example, the node degree metric is: In the formula, T i For nodes i The topological importance index; d i For nodes i The degree of connectivity.

[0023] Taking into account disaster risk, load importance, power supply support capacity, and topological importance, a comprehensive node importance index is constructed: In the formula, I i For nodes i Overall importance; w 1, w 2, w 3, w 4 is the weighting coefficient, and it satisfies: Using the above indicators, the traditional static topology node importance identification can be extended into a dynamic critical node identification method that considers disaster evolution, load level, and resource support capacity.

[0024] As a further aspect of the present invention, step 3) Based on the node comprehensive importance obtained in step 2) I i Sort all nodes of the distribution network in descending order: In the formula, N This represents the total number of nodes in the distribution network. I (k)For the sorted number k The importance of each node.

[0025] The set of key nodes is determined using a threshold filtering method: In the formula, A set of key nodes; I th Identify thresholds for key nodes.

[0026] Alternatively, a proportional screening method can be used to determine the set of key nodes: In the formula, rank( I i ) is a node i Importance ranking number; ρ Select the proportion for key nodes.

[0027] This step can identify nodes that have a significant impact on the resilience of the distribution network in disaster scenarios, including important load nodes, high-risk nodes, and nodes with power supply support capabilities.

[0028] As a further aspect of the present invention, step 4) To ensure power supply to critical nodes and minimize system load shedding, an islanding optimization model based on critical nodes is constructed. The distribution network is divided into several islands, denoted as . ,node i Is it an isolated island? m The state variables are: line ij Whether the state variable remains closed after the island division is: With the goal of minimizing the weighted load loss and the risk of critical node power outage, the objective function is: In the formula, For nodes i At any moment t The amount of load loss; Assigning weights based on load importance; z i For the power supply state variables of critical nodes, if the node i If power is obtained ,otherwise ; β This represents the penalty coefficient for critical node failure.

[0029] The power balance constraints within the island are: In the formula, For the isolated island m The set of nodes within; For the isolated island m Internal circuit loss.

[0030] The node load recovery constraint is: The output constraints of distributed power sources are: Energy storage operation constraints are: In the formula, It is in a state of energy storage charge; and These are the energy storage charging efficiency and the discharging efficiency, respectively.

[0031] The voltage constraint is: The line capacity constraint is: The connectivity constraints for isolated islands are: In the formula, For the isolated island m Internal line collection, For the isolated island m Number of internal nodes.

[0032] For critical nodes, set priority power supply constraints: If the power supply resources within the isolated island meet the power supply requirements This step ensures that island partitioning is no longer based solely on network topology, but rather on safeguarding critical nodes under disaster impacts, thus achieving island partitioning geared towards enhanced resilience.

[0033] As a further aspect of the present invention, step 5) After a disaster occurs, the distribution network topology is updated based on the actual line fault status: In the formula, For disaster relief lines ij The connection status; This represents the initial connection status of the lines before the disaster. This is a line fault state variable; if the line is faulty, then... ,otherwise .

[0034] During the post-disaster recovery phase, a dynamic recovery optimization objective is constructed by combining the isolated operation status, mobile energy storage access status, and maintenance resource status: In the formula, For the line ij Repair completion time; For mobile emergency resources r scheduling costs; γ and η These are the weighting coefficients.

[0035] Mobile energy storage access constraints are: In the formula, Indicates mobile energy storage r Is it at the moment? t Access Node i .

[0036] The output constraint of mobile energy storage is: The maintenance status update constraint is: In the formula, For the line ij At any moment t The status variable indicating whether the repair is complete.

[0037] During post-disaster recovery, priority should be given to restoring power to critical nodes and their associated power lines: The higher the overall importance of a node, the higher its load recovery priority.

[0038] This step achieves a dynamic transformation from initial islanding to resilient islanding through post-disaster dynamic network reconstruction, mobile energy storage scheduling, and maintenance sequence optimization, thereby improving the speed of critical load recovery and reducing the total load loss during the entire disaster process.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows: The method of this invention can identify critical nodes by utilizing disaster failure probability, load importance, power supply support capability, and topology importance, and perform islanding and dynamic post-disaster recovery with the power supply guarantee of critical nodes as the core, thereby reducing the system load failure rate, improving the critical load guarantee rate, and shortening the system recovery time. Attached Figure Description

[0040] Figure 1 A diagram showing the main influencing factors of power transmission and distribution line faults; Figure 2 This is a schematic diagram illustrating the evolution of the operating status of the power distribution network under disaster scenarios. Figure 3 A flowchart for assessing the resilience of distribution networks and generating fault scenarios under the impact of disasters; Figure 4 Framework diagram for optimizing disaster prevention and post-disaster recovery model; Figure 5 Schematic diagram of the results of power distribution line reinforcement and islanding; Figure 6 This is a flowchart of the method of the present invention. Detailed Implementation

[0041] This embodiment provides a method for identifying critical nodes and dividing islands in a distribution network that considers the impact of disasters, including the following steps: Step 1): Modeling the probability of power distribution network line faults under extreme disasters; Step 2): Comprehensive importance assessment of distribution network nodes considering disaster impact; Step 3): Identification of key nodes based on their overall importance; Step 4): Construction of a distribution network islanding model based on key nodes; Step 5): Dynamic reconstruction and emergency recovery optimization of isolated disaster areas; Step 6): Evaluation of the effect of improving the resilience of the distribution network.

[0042] This embodiment focuses on an active distribution network under typhoon disaster conditions. The active distribution network includes a set of nodes. Route Collection Distributed power supply collection Energy storage collection and load set In this context, nodes represent busbars or load connection points in the distribution network, lines represent electrical connections between nodes, distributed power sources include photovoltaic, wind power, and controllable distributed power sources, energy storage includes stationary energy storage and mobile emergency energy storage, and loads include ordinary loads and critical loads.

[0043] In this embodiment, the extreme disaster type is selected as a typhoon disaster. The main steps include: Step 1): Modeling the probability of power distribution line faults under the influence of extreme disasters.

[0044] In this embodiment, a typhoon wind field model is first established based on the typhoon's path, maximum wind speed, movement speed, and spatial location of power lines, and the wind speed impact on each line at different times is calculated. Let the typhoon be at time... t The central position is Its initial position is Movement speed is The direction of movement is The location of the typhoon center can be represented as: In the formula, , They are time points t The horizontal and vertical coordinates of the typhoon center; , These are the coordinates of the initial center of the typhoon; This refers to the typhoon's movement speed; This represents the typhoon's direction of movement.

[0045] For any line in the distribution network Let its equivalent geographical coordinates be... Then the line The distance from the center of the typhoon is: In the formula, For the line At any moment t Distance from the center of the typhoon.

[0046] In this embodiment, the typhoon wind speed at point ijijij on the line can be expressed as: In the formula, For the line At any moment t The equivalent wind speed; For a moment t Maximum wind speed of the typhoon; R t The radius of the typhoon's maximum wind speed; Wind speed attenuation coefficient. The wind speed model may also employ the Holland wind field model, a modified Holland wind field model, or a spatial interpolation model based on meteorological observation data.

[0047] Furthermore, the wind load on the line is calculated based on the wind speed experienced by the line. At any moment t The wind load is expressed as: In the formula, For the line At any moment t The wind load it receives; This is the wind load factor; The equivalent diameter of the line; It is the angle between the wind direction and the line direction.

[0048] Considering the line's self-weight load, the total line load can be expressed as: In the formula, For the line At any moment t Total load borne; For the line The self-weight load.

[0049] Furthermore, the wind-induced stress along the line is expressed as: In the formula, For the line At any moment t Equivalent stress; This is the load-stress conversion factor.

[0050] When the wind-induced stress on the power line exceeds its tensile strength, a line breakage fault occurs. Let the tensile strength of the power line be... Follows a normal distribution: In the formula, This represents the average tensile strength of the line. This represents the standard deviation of the tensile strength of the line.

[0051] Then the line At any moment t The probability of a wire breakage is: To elaborate further: In the formula, For the line At any moment t The probability of a wire breakage.

[0052] In this embodiment, the collapse of the tower under typhoon conditions is also considered. Let the total bending moment on the tower be: in: In the formula, The total bending moment of the tower; The bending moment generated by the wind load on the pole; The additional bending moment transferred from the conductor wind load to the tower; The wind load on the tower; The equivalent stress height of the tower; For the first Horizontal wind load transmitted by the conductor; For the first The height of the conductor suspension point from the tower foundation.

[0053] Determine the bending strength of the tower Follows a normal distribution: The probability of tower collapse is: Right now: In the formula, For the line The associated towers at the time t The probability of tower collapse failure.

[0054] Taking into account both line breakage and tower collapse faults, the line ij The overall failure probability can be expressed as: In the formula, For the line The overall failure probability under the influence of typhoon disasters.

[0055] Furthermore, in this embodiment, the XGBoost model can also be used for data-driven correction of the line fault probability. Let the input feature vector be: In the formula, This refers to the line length; The diameter of the wire; Topographical features of the area where the route is located; For line type or tower type; This represents the number of historical failures.

[0056] The data-driven failure probability prediction model is then expressed as: In the formula, The trained XGBoost failure rate prediction model; This represents the line fault probability predicted by the model.

[0057] In this embodiment, the failure probability calculated by the physical model can be fused with the prediction result of the XGBoost model to obtain the final failure probability: In the formula, The probability of line failure after merging; The fusion weights between the physical model and the data-driven model, and .

[0058] Step 2): Comprehensive importance assessment of distribution network nodes considering the impact of disasters.

[0059] After obtaining the fault probability of each line, this embodiment further constructs a comprehensive importance index for distribution network nodes. This comprehensive importance index considers node disaster risk, load importance, power supply support capacity, and network topology importance simultaneously, and is used to characterize the degree of impact of nodes on the resilience of the distribution network under extreme disaster scenarios.

[0060] (1) Nodal disaster risk indicators: For any node Let the node be The set of connected lines is Then the node At any moment t The disaster risk indicators are: In the formula, For nodes At any moment t Disaster risk indicators; For nodes A set of connected lines; For the line For nodes Influence weight; For the line At any moment t The probability of failure.

[0061] The influence weight of the line can be expressed as: In the formula, Line load rate; This represents the line length.

[0062] (2) Node load importance index: The nodal load importance index is used to characterize the impact of nodal load power loss on social and system operations. Its calculation formula is as follows: In the formula, For nodes At any moment t The load importance index; Weighting based on load level; For nodes At any moment t The load power.

[0063] In this embodiment, the load level weight can be set according to the load type. Ordinary loads correspond to lower weights, while important industrial loads, communication loads, medical loads, and emergency command loads correspond to higher weights. For example: In the formula, Indicates normal load. Indicates important load, This indicates a critical load. The above values ​​are merely a preferred embodiment and do not constitute a limitation on the scope of protection of this invention.

[0064] (3) Node power supply support capability indicators: Node power supply capability is used to characterize the value of a node's power supply support during islanded operation or fault recovery. For nodes Its power supply support capacity is expressed as: In the formula, For nodes At any moment t Power supply support capability; For nodes The output of the connected distributed power source; For nodes Energy storage discharge power; For nodes Energy storage charging power.

[0065] Considering the actual availability of distributed power sources and energy storage, the above formula can be further modified as follows: In the formula, For distributed power availability; For nodes Distributed power generation capacity; Energy storage availability; For nodes Maximum discharge power of the stored energy.

[0066] (4) Node topological importance index: Node topological importance describes the positional value of a node within the distribution network structure. In this embodiment, betweenness centrality is used to represent node topological importance. In the formula, For nodes The topological importance index; For nodes s To the node rThe number of shortest paths; For nodes s To the node r The shortest path passes through the nodes The number of paths.

[0067] To avoid calculation bias caused by different units of measurement of the indicators, this embodiment normalizes the above indicators: In the formula, , , , These are the normalized node disaster risk index, load importance index, power supply support capacity index, and topology importance index.

[0068] Finally, the overall importance index of nodes is expressed as: In the formula, For nodes At any moment t Overall importance; The weighting coefficients are respectively for disaster risk, load importance, power supply support capacity, and topology importance, and satisfy the following: In this embodiment, the weighting coefficients can be determined using the analytic hierarchy process (AHP), entropy weighting, or a combination of weighting methods. When using the combination of weighting methods, the weights can be expressed as: In the formula, These are the subjective weights obtained using the analytic hierarchy process. The objective weights obtained by the entropy weight method; The subjective and objective weighting coordination coefficient, and .

[0069] Step 3): Identification of key nodes based on the overall importance of nodes.

[0070] In this embodiment, based on the node comprehensive importance obtained in step 2), Sort all nodes in the distribution network in descending order: In the formula, For a moment t After sorting, the firstk The overall importance of each node; N This represents the total number of nodes in the distribution network.

[0071] In this embodiment, the set of key nodes can be determined using a threshold method: In the formula, For a moment t The set of key nodes; Identify thresholds for key nodes.

[0072] Preferably, the threshold can be set as follows: In the formula, This represents the average importance of all nodes. The standard deviation of the overall importance of all nodes; This is the threshold adjustment coefficient.

[0073] In another implementation, a proportional screening method can also be used to determine the set of key nodes: In the formula, For nodes Importance ranking number; Select the proportion for key nodes.

[0074] The key nodes identified by the above methods include not only important load nodes, but also high disaster risk nodes, important topological connection nodes, and nodes with distributed power generation or energy storage support capabilities.

[0075] Step 4): Constructing a distribution network islanding model based on key nodes.

[0076] After identifying key nodes, this embodiment further constructs a distribution network islanding model based on key nodes. Assume the distribution network is divided into... M A few isolated islands, the islands are collectively called Define the node ownership variable: Define the closed state variables of the circuit: Define node power supply state variables: This embodiment aims to minimize the system's weighted unload and the penalty for critical node failure, and constructs an island partitioning objective function as follows: In the formula, FDefine the objective function for the isolated island; For the set of scheduling periods; For nodes At any moment t The amount of load loss; Node load weights; Penalty coefficient for failure to supply critical nodes; Assess the overall importance of nodes; Power supply state variables for nodes.

[0077] The node load supply relationship satisfies: In the formula, For nodes At any moment t The actual power supply load.

[0078] For any isolated island m The power balance constraint inside the island is: In the formula, For nodes Distributed power generation output; For nodes Energy storage discharge power; For nodes Energy storage charging power; For the line The meritorious trend; For the isolated island m Internal circuit loss.

[0079] The power flow constraints of the line are: In the formula, For the line susceptance; They are nodes ,node At any moment t The voltage phase angle.

[0080] The line capacity constraint is: In the formula, For the line The maximum transmission power. When the line When disconnected When the line is closed, the power flow is zero. The power flow of a line must not exceed its capacity limit.

[0081] The node voltage constraint is: In the formula, For nodes At any moment t The voltage amplitude; , These represent the lower and upper limits of the node voltage, respectively.

[0082] The output constraints of distributed power sources are: In the formula, , They are nodes The minimum and maximum active power output of the distributed power source.

[0083] Energy storage operation constraints are: In the formula, For nodes Energy storage at any time t The state of charge; To improve energy storage charging efficiency; For energy storage discharge efficiency; The scheduling time interval; , These are the lower and upper limits of the energy storage state of charge, respectively.

[0084] The node island ownership constraint is: In the formula, this constraint means that any node belongs to at most one island at any given time.

[0085] To ensure topological connectivity within the isolated islands, this embodiment uses spanning tree constraints to represent the island connectivity structure: In the formula, Indicates the line Is it an isolated island? m .

[0086] For critical nodes, this embodiment sets priority power supply constraints. When the power supply resources within the island meet the load requirements of the critical nodes, then: When power supply resources are insufficient to meet the power supply needs of all critical nodes, the penalty term for critical node power failure in the objective function is applied according to the overall importance of the nodes. The priority order for power supply is determined from high to low.

[0087] Step 5): Dynamic reconstruction and emergency recovery optimization of isolated areas after disaster.

[0088] Following a disaster, this embodiment further dynamically adjusts the islanded structure based on the actual fault status, emergency resource status, and repair progress. (The following is a partial translation of the original text: "Assume the line...") The post-disaster connectivity status is: In the formula, For the line In the aftermath of the disaster t The connection status; For the line Initial connection status before the disaster; This is a line fault state variable. When a line fault occurs... When the line is normal, .

[0089] The line maintenance status update constraint is: In the formula, For the line At any moment t The status variable indicating whether maintenance is complete. If the line... At any moment t Once the repairs are complete, ;otherwise .

[0090] In this embodiment, mobile emergency energy storage is introduced to participate in post-disaster recovery. Let the mobile emergency energy storage set be... Mobile energy storage r Whether to connect to the node The variables are: The same mobile energy storage device can connect to a node at most at any given time. In the formula, This refers to the state variables for mobile energy storage access.

[0091] The output constraint of mobile energy storage is: In the formula, For mobile energy storage r At any moment t ; output power; For mobile energy storage r Maximum output power.

[0092] The energy state update constraint for mobile energy storage is: In the formula, For mobile energy storage r At any moment t The remaining battery power; Power for charging mobile energy storage; This refers to the discharge power of mobile energy storage. , These refer to the charging and discharging efficiencies of mobile energy storage, respectively.

[0093] The optimization objective during the post-disaster recovery phase is to minimize the weighted load loss, maintenance time cost, and mobile energy storage dispatch cost throughout the recovery process. In the formula, G Optimize the objective function for post-disaster recovery; For the line Repair completion time; For mobile energy storage r scheduling costs; , These are the weighting coefficients for maintenance time costs and mobile energy storage dispatch costs, respectively.

[0094] During post-disaster recovery, the priority of node recovery is determined by the overall importance of the nodes: In the formula, For nodes At any moment t Recovery priority.

[0095] Therefore, the higher the overall importance of a node, the higher its priority in load restoration, power supply path repair, and mobile energy storage access. As lines are gradually repaired and the location of mobile energy storage changes, the islanding structure can be dynamically updated. When the interconnection lines between adjacent islands are repaired and voltage, power flow, and power balance constraints are met, adjacent islands are allowed to merge; when a line in a certain area fails again or its power supply capacity is insufficient, islanding can be redefined.

[0096] Step 6): Evaluation of the effect of improving the resilience of the distribution network.

[0097] In this embodiment, the resilience improvement effect of the method is evaluated by the system failure rate, critical load guarantee rate, and system recovery time.

[0098] The system load failure rate is defined as: In the formula, This represents the system load failure rate.

[0099] Critical load guarantee rate is defined as: In the formula, For critical load guarantee rate; As a key node At any moment t The actual power supply load.

[0100] System recovery time is defined as: In the formula, System recovery time; This is the allowable residual unload threshold.

[0101] The above evaluation indicators allow for a comparison of the effectiveness of traditional islanding methods and the method of this invention. Compared to methods that rely solely on topology or distributed power source location for islanding, this invention identifies critical nodes using disaster failure probability, load importance, power supply support capability, and topological importance. It then focuses on ensuring power supply to these critical nodes for islanding and dynamic post-disaster recovery, thereby reducing system load failure rate, improving critical load availability, and shortening system recovery time.

Claims

1. A method for identifying key nodes and adaptive islanding of a power distribution network considering disaster impact, characterized in that, Includes the following steps: Step 1): Modeling the probability of power grid line failure under extreme disasters: Obtain meteorological data, power grid topology data, line parameters, tower parameters and historical failure data for extreme disasters, establish the mapping relationship between disaster intensity and power grid line failure probability, and obtain the failure probability of each line under the action of disaster; Step 2): Comprehensive importance assessment of distribution network nodes considering disaster impact: Based on the line fault probability, node load importance, node power supply support capacity, and node topology importance, a comprehensive importance index for nodes is constructed to quantitatively assess the importance of distribution network nodes in disaster scenarios; Step 3): Key node identification based on the overall importance of nodes: Sort the distribution network nodes in descending order according to the overall importance index of nodes, and determine the set of key nodes by threshold screening or proportional screening. Step 4): Construction of distribution network islanding model based on key nodes: With the goal of ensuring power supply to key nodes and minimizing system load loss, and combining distributed power sources, energy storage and controllable load resources, construct an optimized distribution network islanding model to achieve active islanding in disaster scenarios; Step 5): Dynamic reconstruction and emergency recovery optimization of isolated power distribution networks after disasters: Based on the actual line fault status and resource availability status after the disaster, the isolated power structure is dynamically adjusted, and the output of distributed power sources, the charging and discharging power of energy storage, the access location of mobile energy storage, and the load recovery sequence are optimized to achieve coordinated recovery of the distribution network after disasters.

2. The method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 1, is characterized in that... In the step 1), when the extreme disaster is typhoon disaster, the line The equivalent wind speed at time t is represented as: In the formula, For the line At any moment The equivalent wind speed; The maximum wind speed of the typhoon at time t; For the line Distance to the center of the typhoon; Let be the radius of the typhoon's maximum wind speed at time t; The wind load on the line is calculated based on the equivalent wind speed of the line. At any moment t The wind load is expressed as: In the formula, For the line At any moment t The wind load it receives; This is the wind load factor; For the line The diameter of the conductor; The angle between the wind direction and the line direction; After considering the line's self-weight load, the total line load is expressed as: wherein is the line at time t total load; is the line dead load; The wind-induced stress of the line is expressed as: wherein is the line at time t wind-induced stress; is the load-stress conversion factor; If the tensile strength of the line obeys normal distribution, the line The probability of the broken line fault at time t is expressed as: In the formula, For the line At any moment t The probability of a wire breakage failure; For the tensile strength of the line; and These represent the mean and standard deviation of the line's tensile strength, respectively.

3. The method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 2, is characterized in that... The step 1) further comprises a tower collapse failure probability calculation, the tower at time t The total bending moment received is expressed as: in: In the formula, For the tower at time t The total bending moment; The bending moment generated by the wind load on the pole; The additional bending moment transferred from the conductor wind load to the tower; The wind load on the tower; This is the equivalent stress height of the tower; For the first n Horizontal wind load transmitted by the conductor; For the first n The height of the conductor suspension point from the tower foundation; The number of connecting wires to the tower; Assuming the bending strength of the tower follows a normal distribution, the probability of tower collapse is expressed as: In the formula, For the line The associated towers at time t The probability of tower collapse failure; For the bending strength of the tower; and These are the mean and standard deviation of the bending strength of the tower, respectively. line At any moment t The overall failure probability is expressed as: In the formula, is the line comprehensive failure probability based on the physical model.

4. The method of claim 3, wherein, Step 1) further includes using the XGBoost model to perform data-driven correction of the line fault probability, with the input feature vector represented as: In the formula, is the line At time t input feature vector; is the line length; is the wire diameter; is the terrain feature of the line area; is the line type or tower type; is the historical failure times; Data-driven failure probability is expressed as: In the formula, is the trained XGBoost failure rate prediction model; is the line failure probability predicted by the XGBoost model. The fault probability calculated by the physical model is fused with the fault probability predicted by the XGBoost model to obtain the final line fault probability: In the formula, is the fused line fault probability; is the fusion weight between the physical model and the data-driven model, and .

5. The method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 1, is characterized in that... The node disaster risk index in step 2) is expressed as follows: In the formula, For nodes At any moment t Disaster risk indicators; For nodes A set of connected lines; For the line For nodes Influence weight; For the line At any moment t The probability of failure; The node load importance index is expressed as follows: In the formula, is a node At time t The load importance index of the node at time is a node load level weight is a node At time t The load power of the node at time The node power supply support capability index is expressed as follows: In the formula, For nodes At any moment t Power supply support capability indicators; For nodes The output of the connected distributed power source; and They are nodes The discharge power and charging power of the energy storage.

6. The method of claim 5, wherein, The node topological importance index in step 2) is represented by betweenness centrality: In the formula, For nodes The topological importance index; For nodes s To the node r The number of shortest paths; For nodes s To the node r The shortest path passes through the nodes The number of paths; After normalizing the node disaster risk index, node load importance index, node power supply support capacity index, and node topology importance index, the comprehensive node importance index is expressed as: in: In the formula, For nodes At any moment t Overall importance; , , and These are the normalized node disaster risk index, node load importance index, node power supply support capacity index, and node topology importance index, respectively. , , , These are the corresponding weighting coefficients.

7. The method of claim 1, wherein, In step 3), the distribution network nodes are sorted in descending order according to their overall importance: In the formula, N is the total number of nodes in the power distribution network; is the time t is the comprehensive importance of the node after sorting k th node; The set of key nodes is determined using a threshold filtering method: In the formula, For a moment t The set of key nodes; Identify thresholds for key nodes; Alternatively, a proportional screening method can be used to determine the set of key nodes: In the formula, For nodes The overall importance ranking number; Select the proportion for key nodes.

8. The method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 1, is characterized in that... In step 4), the distribution network is divided into several islands, and the node affiliation variable is defined as follows: Define the circuit closure state variable as: Define the node power supply state variables as follows: The objective function for island partitioning is constructed with the goals of minimizing the system's weighted unload and the critical node's supply failure penalty: In the formula, F Define the objective function for the isolated island; T For the set of scheduling periods; N For distribution network nodes; For nodes Load weight; For nodes At any moment t Loss of load; Penalty coefficient for failure to supply critical nodes; For nodes At any moment t Overall importance; Power supply state variables for nodes.

9. A method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 8, is characterized in that... The island partitioning model in step 4) satisfies the power balance constraint within the island: In the formula, For the isolated island m The set of nodes within; For nodes At any moment t Distributed power output; and These are the energy storage discharge power and the charging power, respectively. For node load power; This refers to the node's load loss. For the isolated island m Internal line loss or equivalent switching power. The node load recovery constraint is: The output constraints of distributed power sources are: The energy storage state of charge constraints are: In the formula, For energy storage at any time t The state of charge; and These are the energy storage charging efficiency and the discharging efficiency, respectively. For energy storage capacity; This is the scheduling time interval.

10. The method for identifying critical nodes and adaptive islanding in a distribution network considering disaster impact, as described in claim 1, is characterized in that... In step 5), the distribution network topology is updated based on the post-disaster line fault status. The post-disaster connectivity status is represented as follows: In the formula, For the line At any moment t Post-disaster connectivity status; For the line The initial connection state; For the line At any moment t The fault state variable, when a line fault occurs. ,otherwise ; The post-disaster dynamic recovery optimization objective is represented as: In the formula, G Optimize the objective function for dynamic post-disaster recovery; R A collection of mobile emergency resources; For mobile emergency resources r At any moment t scheduling costs; For the line Repair completion time; and These are the weighting coefficients; In step 5), the mobile energy storage access state variable is represented as follows: And it satisfies: The output constraint of mobile energy storage is: In the formula, and Mobile energy storage r At any moment t The discharge power and charging power; and These are the maximum discharge power and maximum charging power of mobile energy storage, respectively. In step 5), the priority of post-disaster load recovery is determined by the overall importance of nodes. For any node... and nodes If the following conditions are met: Then node Load restoration priority is higher than node Load restoration priority; By reconstructing the network dynamically after a disaster, scheduling mobile energy storage, and optimizing the maintenance sequence, a dynamic transformation from initial islanding to resilient islanding can be achieved, improving the power supply guarantee capability of critical nodes and reducing the load loss throughout the disaster process.

Images