A dynamic reconstruction method for distribution network disaster based on prospective-predictive algorithm

Abstract

The present invention discloses a method for dynamic reconstruction of a distribution network during a disaster based on a prospective-pre-evaluation algorithm, which belongs to the technical field of distribution network reconstruction. The method for dynamic reconstruction of a distribution network during a disaster based on a prospective-pre-evaluation algorithm includes: inputting the initial operating state of the distribution network; obtaining the system state at the current moment #imgabs0#; generating a set of all feasible actions at the current moment #imgabs1#; for each action #imgabs2#, executing an evaluation process, constructing future disaster scenarios using the Monte Carlo method, simulating the system evolution process in combination with basic strategies, and evaluating the expected cost of the action #imgabs3# in the entire decision cycle; selecting the action with the smallest expected cost as the optimal action at the current moment, and updating the system state #imgabs4#; updating the time step, if the current time step #imgabs5#, returning to S2 to continue decision-making, otherwise terminating the execution of the algorithm. The reconstruction method described in the present invention optimizes the current decision and takes into account the impact of the current decision on the future, thereby achieving efficient and rapid dynamic reconstruction decisions during disasters and improving the reliability and resilience of the distribution network.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network reconstruction, and in particular to a method for dynamic reconstruction of a distribution network in a disaster based on a look-ahead and pre-calculation algorithm. Background Art

[0002] In recent years, the frequent occurrence of extreme weather events (such as typhoons, heavy rains, and earthquakes) has posed a severe challenge to the safe and stable operation of distribution networks. The spatiotemporal uncertainty of disasters leads to complex challenges in distribution networks, such as line failures, power outages, and load imbalances. Dynamically adjusting distribution network topology during disasters to rapidly restore power and enhance system resilience has become a research priority in the power industry.

[0003] Traditional distribution network reconfiguration methods are mostly based on deterministic models, assuming known disaster parameters (such as fault location and load demand). These methods struggle to cope with the uncertainties of real-time disaster scenarios. These methods typically employ static optimization strategies and lack dynamic adjustment capabilities. These methods are unable to adapt to dynamic changes such as the expansion of fault coverage, load fluctuations, and unstable distributed generation output during a disaster. This results in insufficient robustness in reconfiguration solutions, potentially leading to secondary power outages and inefficient power supply.

[0004] As the penetration of distributed power sources (such as photovoltaic and wind power) in distribution networks continues to increase, the demand for power supply to isolated islands during disasters has increased significantly. However, traditional methods have limitations in island delineation, power balancing, and topology validity verification, making it difficult to balance radial topology constraints with the flexible utilization of distributed power sources. Furthermore, existing technologies often struggle to meet the time requirements for real-time scheduling during disasters due to high computational complexity when dealing with large-scale distribution networks. In particular, the efficiency of traditional optimization algorithms (such as mixed integer programming) decreases significantly in high-dimensional state spaces, making it difficult to achieve rapid response in dynamic decision-making.

[0005] Therefore, there is an urgent need for a dynamic reconfiguration method for distribution networks during disasters that can effectively address disaster uncertainty while balancing real-time performance and optimization quality. This approach aims to improve power supply reliability, recovery efficiency, and robustness during disasters. This paper proposes a dynamic reconfiguration method based on a forward-looking and pre-evaluation algorithm. By incorporating Monte Carlo scenario simulation and basic strategy evolution, this method quantifies the impact of uncertainty and optimizes multi-stage decision-making, providing a new technical approach to addressing these issues. Summary of the Invention

[0006] The purpose of the present invention is to provide a method for dynamic reconstruction of distribution network during disasters based on a look-ahead and pre-recorded algorithm. In view of the spatiotemporal uncertainty of the disaster location, the proposed look-ahead and pre-recorded algorithm not only effectively optimizes the current decision, but also simulates the impact of the current decision on the future through the Monte Carlo method, thereby realizing efficient and rapid dynamic reconstruction decision-making during disasters and improving the reliability and resilience of the distribution network.

[0007] To achieve the above object, the present invention provides a method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead-predictive algorithm, comprising the following steps:

[0008] S1, initialization time step t , input the initial operating state of the distribution network;

[0009] S2. Get the current system status ;

[0010] S3. According to the current system status , generate the set of all feasible actions at the current moment ;

[0011] S4. For each action , execute the assessment process, use the Monte Carlo method to construct future disaster scenarios, and combine the basic strategy to simulate the system evolution process and evaluate the action expected costs throughout the decision cycle;

[0012] S5. Among all actions, select the action with the lowest expected cost , execute it as the optimal action at the current moment and update the system state ;

[0013] S6, update the time step, if the current time step , return to S2 to continue decision-making, otherwise terminate the algorithm execution.

[0014] Preferably, in S1, the initialization time step t =1, the initial operating state of the distribution network includes topology, load distribution, distributed generation configuration and disaster impact information.

[0015] Preferably, in said S2, the system state includes the line switch state vector, the fault line set, the fault state vector, the on-off state vector and the current time step.

[0016] Preferably, in said S3, a single-step look-ahead mechanism is adopted, based on the current system state , generate the set of all feasible actions at the current moment Specifically, the action includes the combined operation of the tie switch and the section switch. The feasible action set satisfies the radial topology constraint. At the same time, each connected area contains at least one distributed power generation node. The feasible action set satisfies the radial topology constraint as follows:

[0017] ;

[0018] ;

[0019] ;

[0020] ;

[0021] Where, express t Time branch The on-off state, To disconnect, For closure; Indicates t Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates t Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates the number of power sources forming an island, including the main grid; Represents all nodes in the distribution network, represents the set of all branches of the distribution network, Representation and Node The set of connected adjacent nodes, T Indicates a time period.

[0022] Preferably, in said S4, for each action , perform the assessment process, including the following steps:

[0023] S41. Sampling from Uncertain Disaster Models Using Monte Carlo Methods A disaster scenario is used to simulate the disaster development path after the action is executed;

[0024] S42. Execute actions in each disaster scenario , from the current state Entering a new state , and calculate the action Immediate cost ;

[0025] S43. In each disaster scenario, Starting from now, adopt basic strategies for future Evolution simulation of the time period, accumulating its operating costs ;

[0026] S44. Calculate the total cost for each disaster scenario , and calculate The average value of the total cost of each disaster scenario is used to obtain the current action Expected cost .

[0027] Preferably, the specific process of S41 is:

[0028] According to the current action The corresponding topological structure is used to identify the set of distribution lines in a closed state after the action is executed, and the Monte Carlo method is used to sample disaster scenarios based on the disaster uncertainty model; in each round of sampling, for the above-mentioned closed distribution lines, according to their corresponding disaster failure probabilities, a pseudo-random sampling method is used to determine whether they have failed in the current scenario, only considering the failure possibility of closed lines, and ignoring the failure impact of other unclosed lines; the above sampling process is repeated to generate a predetermined number of disaster scenario samples, each sample consisting of a set of line faults, and the disaster scenario serves as the environmental input for the subsequent action effect evaluation.

[0029] Preferably, the instant cost in S42 The calculation process is:

[0030] In each disaster scenario, perform actions , a new distribution network topology under the influence of the disaster is obtained. A graph structure model is constructed based on the new distribution network topology. At each load node, it is calculated whether permanent power outages have occurred due to disasters and operations. Furthermore, temporary island areas that are temporarily powered by distributed power sources due to the inability to access the main power supply are identified:

[0031] ;

[0032] ;

[0033] Where, Indicates permanent power loss, Indicates temporary island loss, represents the weight factor of the load node, express Time Node Load power; Representation node Power failure status indication, 1 indicates permanent power failure, 0 indicates normal power supply; express Time Node Whether it is in a temporary island state, 1 means the node is in a temporary island state, 0 means it is not in an island state;

[0034] Combined with the current load power, load importance factor and power failure status, calculate the immediate cost of the action in the current scenario.

[0035] ;

[0036] Where, The time required for the transfer operation, and are the penalty coefficients for permanent power outage and temporary power shortage respectively.

[0037] Preferably, the accumulated operating cost in S43 The calculation process is:

[0038] At the moment , by the basic strategy Select an action:

[0039] ;

[0040] Through the system state transfer function, simulate the system operation:

[0041] ;

[0042] Where, express External disturbance variables at time instant;

[0043] After calculating the real-time running cost After that, all future operating costs are accumulated:

[0044] ;

[0045] Where, Indicates the The cumulative operating cost calculated by the basic strategy starting from the current action in each scenario.

[0046] Preferably, in said S44, the action Expected cost The calculation formula is:

[0047] .

[0048] Preferably, in S5, the system compares the expected costs of all actions, selects the action with the smallest expected cost as the optimal action at the current moment, and uses the updated state as the starting state at the next moment to enter the next round of decision-making cycle.

[0049] The advantages and positive effects of the dynamic reconstruction method for distribution network disasters based on a forward-looking and pre-evaluation algorithm described in the present invention are: the present invention takes into account the spatiotemporal uncertainty of the disaster location, while optimizing the current action selection, and uses the Monte Carlo method to evaluate its impact on the future evolution path, thereby achieving more robust dynamic reconstruction decisions and improving the reliability and resilience of the distribution network during disasters. By introducing a heuristic forward-looking and pre-evaluation algorithm, the present invention obtains a near-optimal, high-quality feasible solution at the expense of global optimality, achieving rapid response and dynamic adjustment, and better meeting the dual requirements of real-time scheduling and practicality during disasters.

[0050] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of a method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to the present invention;

[0052] Figure 2 This is a flow chart of distribution network evaluation based on the prospective-predictive algorithm of the present invention;

[0053] Figure 3 A topological diagram in an embodiment of the present invention;

[0054] Figure 4 The dynamic reconstruction process in the embodiment of the present invention t 0 Time period topology diagram;

[0055] Figure 5 The dynamic reconstruction process in the embodiment of the present invention t 0 +0.5h Time period topology diagram;

[0056] Figure 6 The dynamic reconstruction process in the embodiment of the present invention t 0 +1h Time period topology diagram;

[0057] Figure 7 The dynamic reconstruction process in the embodiment of the present invention t 0 +1.5h Time period topology diagram;

[0058] Figure 8 The dynamic reconstruction process in the embodiment of the present invention t 0 +2h Time period topology diagram;

[0059] Figure 9The dynamic reconstruction process in the embodiment of the present invention t 0 +2.5h Time period topology diagram;

[0060] Figure 10 The dynamic reconstruction process in the embodiment of the present invention t 0 +3h Time period topology diagram;

[0061] Figure 11 The dynamic reconstruction process in the embodiment of the present invention t 0 +3.5h Time period topology diagram. DETAILED DESCRIPTION

[0062] It will be understood by those skilled in the art that, unless expressly stated otherwise, the singular forms "a", "an", "said" and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the description of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, "connected" or "coupled" as used herein may include wireless connections or couplings. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0063] In this application, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. In the event of any inconsistency, the meaning described in this specification or the meaning derived from the contents recorded in this specification shall prevail. In addition, the terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0064] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0065] like Figure 1 As shown, a method for dynamic reconstruction of distribution network during disaster based on the prospective-predictive algorithm includes the following steps:

[0066] S1, initialization time step t , input the initial operating status of the distribution network.

[0067] Initialize time step t=1, the initial operating state of the distribution network includes topology, load distribution, distributed power configuration and disaster impact information.

[0068] S2. Get the current system status .

[0069] The system state mainly includes the following five key variables, which are used to describe the distribution network at any time step: The following operating characteristics:

[0070] Line switch state vector :express The on / off status of all section switches and tie switches in the system at all times, express Time switch In closed state, express Time switch In the disconnected state, this state determines the current topology of the distribution network.

[0071] Fault line collection :Record The set of line numbers that are known to have faults at any given moment reflects the direct damage caused by the disaster to the distribution network.

[0072] Fault state vector : Used to mark whether each line has a fault. express Timeline A malfunction occurs. express The line is normal at this time.

[0073] On-off state vector : Indicates the on / off status of all lines in the system at the current moment. express Timeline conduction, express Timeline Disconnected. This state is determined by both the switching operation and the fault effect.

[0074] Current time step : Indicates the current decision-making stage. It is initially 1 and increases after each round of decision-making until it reaches the end of the decision-making cycle.

[0075] Through S1 and S2, current system status information is obtained in real time, including the currently observable fault location, the power supply status of the main and distributed power sources, and the status of operable tie switches. This status information serves as the fundamental input for action decisions and assessments, reflecting the system damage and operating status as the disaster evolves.

[0076] S3. According to the current system status , generate the set of all feasible actions at the current moment .

[0077] Adopting a single-step look-ahead mechanism, based on the current system status , generate the set of all feasible actions at the current moment The feasible action set is the set of feasible actions that meet the radial topology constraints. The actions include the combined operation of the tie switch and the section switch. To ensure the feasibility of the system structure after the operation, the topology structure validity check must be performed on each candidate action. Specifically:

[0078] The generated topology should satisfy the radial topology constraint, that is, the structure of each connected area should be a tree graph;

[0079] At the same time, each connected area contains at least one distributed power supply node to ensure that the isolated area has independent power supply capabilities.

[0080] The set of feasible actions that satisfy the radial topology constraint is:

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] Where, express t Time branch The on-off state, =0 is disconnected, =1 is closed; Indicates t Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates t Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates the number of power sources forming an island, including the main grid; Represents all nodes in the distribution network, represents the set of all branches of the distribution network, Representation and Node The set of connected adjacent nodes, T Indicates a time period.

[0086] In S3, according to the current system state, an action generation mechanism is used to construct a set of feasible actions, covering all possible distribution network structure reconstruction schemes, ensuring that the search space covers all potential feasible paths.

[0087] S4. For each action , execute the assessment process, use the Monte Carlo method to construct future disaster scenarios, and combine the basic strategy to simulate the system evolution process and evaluate the action Expected costs throughout the decision cycle.

[0088] like Figure 2 As shown, for each action , perform the assessment process, including the following steps:

[0089] S41. Sampling from Uncertain Disaster Models Using Monte Carlo Methods A disaster scenario is used to simulate the disaster development path after the action is executed.

[0090] According to the current action The corresponding topological structure is used to identify the set of distribution lines in a closed state after the action is executed, and based on the disaster uncertainty model, the Monte Carlo method is used to sample disaster scenarios. Specifically, in each round of sampling, for the above-mentioned closed distribution lines, according to their corresponding disaster failure probabilities, a pseudo-random sampling method is used to determine whether they have failed in the current scenario. Only the failure possibility of the closed line is considered, and the failure impact of other unclosed lines is ignored. The above sampling process is repeated to generate a predetermined number of disaster scenario samples. Each sample consists of a set of line fault sets, which represent the possible disaster development paths under the distribution topology after the action is executed. The above multiple disaster scenarios serve as environmental inputs for the evaluation of the effectiveness of subsequent actions.

[0091] S42. Execute actions in each disaster scenario , from the current state Entering a new state , and calculate the action Immediate cost .

[0092] Immediate costs The calculation process is:

[0093] Execute actions for each disaster scenario generated , that is, implementing the corresponding switching operation under the current system state to obtain the new distribution network topology under the influence of the disaster. Based on the new distribution network topology, a graph structure model is constructed. The accessibility of each load node in the power system to the main power node and distributed power node is evaluated through the breadth-first search method, and the power supply capacity is determined accordingly.

[0094] For the new topology, calculate at each load node whether permanent power outages have occurred due to disasters and operations, and further identify temporary island areas where the main power source is unavailable but is temporarily supplied by distributed generation:

[0095] ;

[0096] ;

[0097] Where, Indicates permanent power loss, Indicates temporary island loss, represents the weight factor of the load node, express Time Node Load power; Representation node Power failure status indication, 1 indicates permanent power failure, 0 indicates normal power supply; express Time Node Whether it is in a temporary island state, 1 means the node is in a temporary island state, 0 means it is not in an island state;

[0098] Combined with the current load power, load importance factor, and power outage status, the immediate cost of the action in the current scenario is calculated, including the economic loss of the permanent power outage load and the transient power shortage penalty caused by the island power transfer. The final immediate cost assessment result of the action in this disaster scenario is as follows:

[0099] ;

[0100] Where, The time required for the transfer operation, and are the penalty coefficients for permanent power outage and temporary power shortage respectively.

[0101] S43. In each disaster scenario, Starting from now, adopt basic strategies for future Evolution simulation of the time period, accumulating its operating costs .

[0102] In each scene In, from the state Starting with the basic strategy (Base Policy) for the future Evolution simulation of the time period, accumulating its operating costs . Cumulative operating costs The calculation process is:

[0103] At the moment , by the basic strategy Select an action:

[0104] ;

[0105] Through the system state transfer function, simulate the system operation:

[0106] ;

[0107] Where, express External disturbance variables at time instant;

[0108] After calculating the real-time running cost After that, all future operating costs are accumulated:

[0109] ;

[0110] Where, Indicates the The cumulative operating cost calculated by the basic strategy starting from the current action in each scenario.

[0111] S44. Calculate the total cost for each disaster scenario , and calculate The average value of the total cost of each disaster scenario is used to obtain the current action Expected cost .

[0112] action Expected cost The calculation formula is:

[0113] .

[0114] To accurately assess the future operational effects of each candidate action, S4 introduces a Monte Carlo scenario generation method to simulate multiple future disaster evolution paths. In each scenario, starting from the state after the action is executed, the basic strategy is invoked to simulate the future evolution of the system, and its operating costs are accumulated over the entire rolling horizon. This evaluation process comprehensively considers the uncertainty of future scenarios and quantifies the performance stability and power supply capacity of each action under different disaster scenarios.

[0115] S5. Among all actions, select the action with the lowest expected cost , execute it as the optimal action at the current moment and update the system state .

[0116] The system compares the expected costs of all actions, selects the action with the smallest expected cost as the optimal action at the current moment, and uses the updated state as the starting state for the next moment to enter the next round of decision-making cycle.

[0117] S6, update the time step, if the current time step , return to S2 to continue decision-making, otherwise terminate the algorithm execution.

[0118] Through the synergistic effect of the above steps, the proposed dynamic reconstruction method for distribution network disasters based on the foresight-rehearsal strategy can dynamically adjust the power supply structure in the multi-stage time domain after the disaster occurs, taking into account uncertainty and cost optimization, and effectively improving the system's recovery efficiency and operational robustness.

[0119] As a specific example, the present invention is further described in detail in one of the embodiments.

[0120] The simulations in this example were written using MATLAB R2021a and run on a laptop with a 64-bit operating system, a 2.4GHz quad-core CPU, and 8GB of memory. The dynamic reconfiguration model for distribution network disasters based on the look-ahead-and-preview strategy was implemented using YALMIP and solved using the CPLEX commercial solver.

[0121] In this embodiment, a 148-node distribution network system is used as an example to verify the effectiveness of the dynamic reconstruction method of the distribution network disaster based on the prospective-preview strategy proposed in this invention. Its topology is as follows: Figure 3 shown.

[0122] The test system consists of two substations, four transformers, and eight feeders. The system voltage level is 12.66 kV, and the reference power is 10 MVA. There are two substation nodes and 146 load nodes within the system. Nodes 16, 32, 53, 54, 76, 78, 85, 86, 88, 90, 91, 92, 93, 96, 100, 102, 105, 118, 120, 129, 132, and 140 are critical load nodes. It is assumed that all lines in the system are equipped with sectionalizers, which are normally closed (shown by solid lines); tie switches are initially open (shown by dashed lines). Nodes 16, 53, 86, 102, 118, and 132 are connected to distributed generation (DGs) with capacities of 500 kW, 100 kW, 500 kW, 100 kW, 500 kW, and 400 kW, respectively.

[0123] To test the effectiveness of post-disaster reconstruction, we used a four-hour post-disaster period as the research period and assumed that the distribution network was reconfigured every 0.5 hours. Table 1 shows the statistics of faulty lines in the system.

[0124] Table 1 Statistics of faulty lines in different time periods

[0125] ;

[0126] During the disaster, the system dynamically responds to the gradually exposed fault location and adopts the proposed disaster reconstruction method based on the prospective-pre-rehearsal algorithm to perform topology adjustment in different time periods. In each time period, the system generates the current optimal switching operation plan based on the current fault status and load demand, taking into account the optimization direction of the objective function. The final dynamic reconstruction optimization plan of this embodiment is shown in Table 2 below, and the corresponding final dynamic reconstruction process topology change diagram is shown in Figure 4-11 shown.

[0127] Table 2 Final dynamic reconstruction solution

[0128] ;

[0129] As can be seen from the table, in the early stage of the system, the global power supply structure optimization is mainly achieved through local reconstruction to avoid the generation of islands. When the load distribution and fault range expand to a certain extent, the system t 0 +3 h It actively forms an isolated island area with independent power supply capability to maximize the continuity of load supply, verifying the dynamic adaptability and scheduling intelligence of the present invention during the evolution process of disasters.

[0130] In summary, the present invention has the following advantages compared to the prior art:

[0131] Currently, research on improving distribution network resilience, both domestically and internationally, focuses primarily on the pre-disaster and post-disaster phases, relying on pre-disaster equipment reinforcement measures and post-disaster emergency resource deployment. Research on dynamic adjustments during disasters remains relatively weak, lacking methodological support that balances real-time performance with high-quality decision-making.

[0132] This paper considers the spatiotemporal uncertainty of disaster locations and proposes a method for dynamic reconfiguration of distribution networks during disasters based on a look-ahead and pre-evaluation algorithm. This method optimizes the current action selection while utilizing Monte Carlo methods to assess its impact on future evolution paths. This enables more robust dynamic reconfiguration decisions and improves the reliability and resilience of the distribution network during disasters.

[0133] In addition, the present invention uses the YALMIP toolbox in MATLAB for modeling and calls the CPLEX solver for solving. On the basis of ensuring the modeling expression ability, it effectively improves the solution efficiency and has strong algorithm adaptability, which can meet the real-time scheduling needs during the disaster development process.

[0134] Compared to traditional distribution network reconfiguration strategy solutions based on mixed-integer second-order cone programming (MISOCP) methods, the proposed method significantly reduces computation time. Traditional MISOCP methods often struggle to obtain high-quality solutions within a limited timeframe in disaster-related reconfiguration scenarios due to the large number of integer variables and high solution complexity, and may even fail to reach acceptable solution boundaries. However, this method, by introducing a heuristic look-ahead and pre-calculation algorithm, achieves a near-optimal, high-quality feasible solution at a moderate cost of global optimality, enabling rapid response and dynamic adjustment. This approach better meets the dual requirements of real-time scheduling and practicality during disasters.

[0135] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0136] From the above description of the embodiments, it is clear that those skilled in the art will clearly understand that the present invention can be implemented using software and a necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, or an optical disk, and includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various embodiments of the present invention, or portions thereof.

[0137] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for dynamic reconstruction of distribution network during disaster based on prospective-predictive algorithm, characterized in that: The following steps are involved: S1, initialization time step , input the initial operating state of the distribution network; S2. Get the current system status ; S3. According to the current system status , generate the set of all feasible actions at the current moment ; S4. For each action , execute the assessment process, use the Monte Carlo method to construct future disaster scenarios, and combine the basic strategy to simulate the system evolution process and evaluate the action expected costs throughout the decision cycle; S5. Among all actions, select the action with the lowest expected cost , execute it as the optimal action at the current moment and update the system state ; S6, update the time step, if the current time step , return to S2 to continue decision-making, otherwise terminate the algorithm execution; In S3, a single-step look-ahead mechanism is used to determine the current state of the system. , generate the set of all feasible actions at the current moment Specifically, the action includes the combined operation of the tie switch and the section switch. The feasible action set satisfies the radial topology constraint. At the same time, each connected area contains at least one distributed power node. The feasible action set satisfies the radial topology constraint as follows: ； ； ； ； Where, express Time branch The on-off state, To disconnect, For closure; Indicates Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates Time branch Milestone With node The relationship between parent and child nodes, Representation node is a node The parent node of , otherwise 0; Indicates the number of power sources forming an island, including the main grid; Represents all nodes in the distribution network, represents the set of all branches of the distribution network, Representation and Node The set of connected adjacent nodes, Indicates a time period; In S4, for each action , perform the assessment process, including the following steps: S41. Sampling from Uncertain Disaster Models Using Monte Carlo Methods A disaster scenario is used to simulate the disaster development path after the action is executed; S42. Execute actions in each disaster scenario , from the current state Entering a new state , and calculate the action Immediate cost ; S43. In each disaster scenario, Starting from now, adopt basic strategies for future Evolution simulation of the time period, accumulating its operating costs ; S44. Calculate the total cost for each disaster scenario , and calculate The average value of the total cost of each disaster scenario is used to obtain the current action Expected cost .

2. The method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to claim 1, characterized in that: In S1, the initialization time step ,The initial operating state of the distribution network includes topology ,structure, load distribution, distributed generation configuration and disaster ,impact information.

3. The method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to claim 1, characterized in that: In S2, the system state includes the line switch state vector, the fault line set, the fault state vector, the on-off state vector and the current time step.

4. The method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to claim 1, characterized in that: The specific process of S41 is as follows: According to the current action The corresponding topological structure,identifies the set of distribution lines in the closed state after the action is executed,and uses the Monte Carlo method to sample disaster scenarios based on the,disaster uncertainty model; In each round of sampling, for the above-mentioned closed distribution lines, a pseudo-random sampling method is used to determine whether they have failed in the current scenario based on their corresponding disaster failure probabilities. Only the failure possibility of the closed lines is considered, and the failure impact of other unclosed lines is ignored. The above sampling process is repeated to generate a predetermined number of disaster scenario samples. Each sample consists of a set of line faults. The disaster scenario serves as the environmental input for the evaluation of the effectiveness of subsequent actions.

5. The method for dynamic reconstruction of distribution network during disaster based on the look-ahead and pre-calculation algorithm according to claim 4 is characterized in that: The instant cost of S42 The calculation process is: In each disaster scenario, perform actions , a new distribution network topology under the influence of the disaster is obtained. A graph structure model is constructed based on the new distribution network topology. At each load node, it is calculated whether permanent power outages have occurred due to disasters and operations. Furthermore, temporary island areas that are temporarily powered by distributed power sources due to the inability to access the main power supply are identified: ； ； Where, Indicates permanent power loss, Indicates temporary island loss, represents the weight factor of the load node, express Time Node Load power; Representation node Power failure status indication, 1 indicates permanent power failure, 0 indicates normal power supply; express Time Node Whether it is in a temporary island state, 1 means the node is in a temporary island state, 0 means it is not in an island state; Combined with the current load power, load importance factor and power failure status, calculate the immediate cost of the action in the current scenario. ; Where, The time required for the transfer operation, and are the penalty coefficients for permanent power outage and temporary power shortage respectively.

6. The method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to claim 5, characterized in that: The S43 cumulative operating costs The calculation process is: At the moment , by the basic strategy Select an action: ； Through the system state transfer function, simulate the system operation: ； Where, express External disturbance variables at time instant; After calculating the real-time running cost After that, all future operating costs are accumulated: ； Where, Indicates the The cumulative operating cost calculated by the basic strategy starting from the current action in each scenario.

7. The method for dynamic reconstruction of distribution network during disaster based on the look-ahead and pre-calculation algorithm according to claim 6, characterized in that: In the above S44, the action Expected cost The calculation formula is: 。 8. The method for dynamic reconstruction of a distribution network during a disaster based on a look-ahead and pre-calculation algorithm according to claim 1, characterized in that: In S5, the system compares the expected costs of all actions, selects the action with the smallest expected cost as the optimal action at the current moment, and uses the updated state as the starting state at the next moment to enter the next round of decision-making cycle.

Images