A method for post-disaster emergency recovery decision of urban distribution network based on machine-like intuition

Abstract

This invention provides a post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition. The method includes: initializing decision parameters and basic data of the urban power distribution network to obtain the maximum load set without considering energy constraints; identifying restored and unrestored loads based on the maximum load set satisfying power flow constraints; generating a set of emergency recovery candidate loads through permutation, combination, and supplementation strategies; calculating the emergency recovery benefit corresponding to each feasible emergency recovery candidate load set; selecting the feasible emergency recovery candidate load set with the highest benefit as the forward-looking emergency recovery plan for the current period; and updating the restored load status, unrestored load status, and available remaining energy data of the power distribution network to complete a single-round online dynamic decision-making process. This invention not only effectively optimizes the current decision but also considers the impact of the current decision on the future through simulation, achieving efficient and rapid load recovery decision-making and improving the reliability and resilience of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of power system analysis technology, and in particular to a post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition. Background Technology

[0002] With the continuous integration of various distributed power sources, such as photovoltaics, electric vehicles, distributed energy storage systems, and combined cooling, heating, and power (CCHP), into the power grid, new opportunities have emerged to enhance the rapid power restoration capabilities of the distribution network after disasters. However, existing technologies employ emergency control strategies developed offline for typical operating scenarios based on anticipated fault scenarios. These strategies may be unsuitable for real-world disaster scenarios due to factors such as repair time, newly discovered faults, relocation time, and uncertainties in wind and solar power output. Consequently, they struggle to cope with multi-point disturbances characterized by strong online random dynamic changes.

[0003] However, with the continuous improvement of power system automation, today's urban distribution networks have made significant progress in information sensing, data acquisition, and intelligent control compared to traditional distribution networks. For example, the application of distribution automation systems, smart sensors, IoT technology, and big data analytics enables the power grid to monitor its operating status in real time and obtain load distribution information. These technological advancements provide data support and implementation possibilities for online decision-making based on "machine intuition," making the post-disaster distribution network recovery process more flexible and efficient.

[0004] In recent years, influenced by climate change and international instability, the frequency and intensity of extreme events have increased, posing challenges to the safe operation of urban power grids. Major power outages caused by natural disasters, cascading failures, and cyberattacks have drawn widespread attention. Firstly, major power outages disconnect the urban distribution network from the upper-level power grid, disrupting industrial production and commercial activities, impacting residents' daily lives, and causing enormous economic losses. Prolonged power outages affecting critical infrastructure and service systems such as healthcare, nursing homes, street lighting, and communications jeopardize basic survival and livelihood security, as well as social stability. As a crucial component of the power system, the post-disaster recovery capability of the distribution network directly impacts the stable operation of society and the public's quality of life. Therefore, research on rapid power restoration methods for distribution networks after extreme events is of significant strategic importance.

[0005] In summary, existing power distribution network disaster recovery strategies primarily rely on pre-defined offline optimization plans. However, due to the complexity of disasters and the existence of various complex and difficult-to-describe constraints in offline optimization, recovery efficiency is low, and it may even be difficult to execute in some extreme cases. Summary of the Invention

[0006] The embodiments of the present invention provide a post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition, so as to effectively improve the reliability and resilience of urban power distribution networks.

[0007] To achieve the above objectives, the present invention adopts the following technical solution.

[0008] (Corresponds to the claims)

[0009] As can be seen from the technical solutions provided by the embodiments of the present invention above, the look-ahead algorithm proposed by the present invention not only effectively optimizes the current decision, but also realizes efficient and rapid load recovery decision by simulating the impact of the current decision on the future, thereby improving the reliability and resilience of the distribution network.

[0010] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of the invention. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 The flowchart illustrates a post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition, as provided in this embodiment of the invention. Detailed Implementation

[0013] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0014] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated 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 say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0015] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0016] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0017] This invention combines the operational characteristics of distribution networks with a dynamic decision-making technology for emergency recovery of large urban distribution networks after disasters, taking into account personnel dispatching experience. Based on "machine intuition," it constructs an online decision-making method suitable for the post-disaster recovery of distribution networks and utilizes machine intuition to achieve efficient online decision-making, thereby improving the response speed and decision quality of post-disaster recovery.

[0018] A flowchart of a post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition is provided in this embodiment of the invention. Figure 1 As shown, the processing steps include the following:

[0019] Step S1. Initialize the decision parameters and basic data of the urban power distribution network. Without considering energy constraints, construct a single-time power flow optimization model to determine the maximum load set that satisfies the power flow constraints.

[0020] Step S2. Identify recovered and unrecovered loads based on the maximum load set satisfying power flow constraints, and obtain the current system state. The cost-effectiveness of each unrestored load is calculated and sorted to obtain a sorted set of unrestored loads. Simultaneously, the available effective surplus energy in the current urban power distribution network is calculated.

[0021] Step S3. Based on the maximum load set determined in step S1, first construct the least conservative emergency recovery candidate load set, then expand the emergency recovery candidate load set through permutation, combination and supplementation strategies, and generate the emergency recovery candidate load set using a single-step look-ahead strategy.

[0022] Step S4. Perform power flow feasibility verification on each emergency recovery candidate load set generated in step S3, introduce power flow constraints and node voltage constraints to eliminate infeasible emergency recovery candidate load sets, and calculate the output of each generator for the remaining feasible emergency recovery candidate load sets;

[0023] Step S5. Perform emergency recovery simulation on each feasible emergency recovery candidate load set, calculate the emergency recovery benefit corresponding to each feasible emergency recovery candidate load set, and select the feasible emergency recovery candidate load set with the highest benefit as the emergency recovery prospective plan for the current period.

[0024] Step S6. Based on the emergency recovery prospective plan for the current time period determined in Step S5, update the data on the restored load status, unrestored load status, and available remaining energy of the distribution network to complete a single round of online dynamic decision-making. If the system time step does not reach the preset target, return to Step S2 to enter the next round of decision-making.

[0025] Specifically, the process of constructing the single-time power flow optimization model and determining the maximum load set in step S1 is as follows:

[0026] S1.1: Obtain post-disaster topology data of the urban power distribution network, including node parameters, branch parameters, generator parameters, and load parameters, and clarify the node relationships and branch connection relationships of the urban power distribution network;

[0027] S1.2: Construct a single-time power flow optimization model without considering energy constraints based on the nodal power balance principle. The expression of this single-time power flow optimization model is:

[0028]

[0029] Where N is the total number of distribution network nodes; , These are the active power and reactive power of the branch between node i and node j, respectively. , These represent the active power output and reactive power output of the generator at node i, respectively. , These are the active power and reactive power of the load at node i, respectively.

[0030] S1.3: Voltage constraint:

[0031]

[0032] in Let be the voltage amplitude at node i; , These are the minimum and maximum voltage values ​​at node i, typically taken as 0.95 and 1.05 per unit.

[0033] S1.4: With the goal of maximizing the weighted load recovery time, the objective function of the single-time power flow optimization model is set as follows:

[0034]

[0035] in, The importance coefficient of node i is determined according to the load level. The value is 100 for level 1 load (hospitals, emergency command centers, etc.), 10 for level 2 load (commercial complexes, key industrial production loads, etc.), and 1 for level 3 load (ordinary residential loads, general commercial loads, etc.). Let be the recovery state variable of the load at node i. =1 indicates that the load of node i has been restored. =0 indicates that the load on node i has not been restored;

[0036] The maximum load set that satisfies all power flow constraints is obtained. Since energy constraints are not considered in the optimization solution, the solution obtained under sufficient energy conditions is the maximum recoverable load set. In the context of emergency recovery, the number of recoverable loads will inevitably be less than this set. Therefore, the set can serve as a preliminary judgment for subsequent recovery.

[0037] Specifically, step S2 involves obtaining the current state of the distribution network system. The calculation and ranking process for the cost-effectiveness of each unrestored load is as follows:

[0038] S2.1: Obtain the state of the distribution network system at the current decision-making moment. The status of the power distribution network system It includes two core dimensions:

[0039] (1) Load recovery status: Identify the recovery variables of each load node. (Restored / Not Restored), forming a set of restored loads and a set of not restored loads;

[0040] (2) Energy surplus status: Statistics on the current available surplus energy of all power sources (including distributed generators, energy storage devices, emergency power supply vehicles, etc.) in the distribution network. With rated output power Where g is the power supply number, g=1,2,...,G, and G is the total number of power supplies:

[0041] S2.2: Define load cost-effectiveness To calculate the cost-effectiveness ratio of load restoration benefits to restoration costs, for each unrestored load i in the unrestored load set, the calculation formula is as follows:

[0042]

[0043] in, The importance coefficient of the load at node i. Let be the active power of the load at node i. The principle is to restore loads that are more important and consume less energy first.

[0044] S2.3: Based on cost-effectiveness Sort the unrestored loads in the unrestored load set in descending order to obtain the sorted unrestored load sequence [L1, L2, ..., L]. m ], where m is the total number of loads that have not been restored, and the full ;

[0045] S2.4: Calculate the effective surplus energy of the distribution network For each power supply g, considering its rated output power limitation, the maximum energy it can release during the power outage recovery cycle is the product of its rated power and the total power outage duration. The effective remaining energy is the smaller value between the power supply's current actual remaining energy and this maximum releaseable energy. The effective remaining energy of the entire network is the sum of the effective remaining energies of all power supplies. The specific calculation steps are as follows:

[0046] (1) Calculate the effective residual energy of a single power source g. :

[0047]

[0048] in, Let g be the current actual remaining energy of the power source g; The rated output power of power supply g; The remaining power outage duration is the total time from this moment to the preset power restoration completion time.

[0049] (2) Calculate the available effective surplus energy in the current urban power distribution network. :

[0050]

[0051] Specifically, the process of generating the emergency recovery candidate load set in step S3 is as follows:

[0052] S31: Determine the set of least conservative emergency recovery candidate loads This set is not the set obtained in step S1 (because energy constraints were not considered), but is obtained by combining energy constraints and cost-effectiveness ranking and filtering. The specific process is as follows:

[0053] S311: Within the maximum load set satisfying the power flow obtained in S1, the sorted unrecovered load set and effective surplus energy obtained through S2 are... ;

[0054] S312: Sequentially evaluate the unrecovered load sequence [L1, L2, ..., L...] m Whether the energy of the load system in [ ] can meet its full-time recovery needs, for load L k Calculate the energy required for its continuous operation throughout the entire timeframe after recovery in the current decision-making period. ,in Let be the active power of load k, T be the estimated total power outage duration, and t be the current time. If Then the load L k Added to the emergency recovery candidate load set and update the effective remaining energy. ,like If the selection fails, the screening process stops, ultimately yielding the least conservative set of emergency recovery candidate loads. Its load quantity is n, that is .

[0055] S313: Final result This is the least conservative set of emergency recovery candidate loads, with a load count of n, i.e. ;

[0056] S32: Since obtaining only one solution may be insufficient to cover the optimal situation, the least conservative emergency recovery candidate load set is selected to obtain multiple more likely recovery solutions. The lowest cost-performance ratio among the [0.3n] loads (where n is...) The number of loads), denoted as a subset. ,right The loads in the data are permuted and combined to generate multiple non-empty subsets.

[0057] S33: Will Remove The load sets other than those above are merged to obtain the basic candidate subset, and finally an empty set is added, where the empty set corresponds to the scheme that does not restore any new load.

[0058] S34: Finally, the set of emergency recovery candidate loads is obtained. , where k is the total number of candidate loads for emergency recovery.

[0059] Specifically, the power flow feasibility verification process for each emergency recovery candidate load set in step S4 adopts an optimization judgment logic of "multi-constraint + minimum network loss optimization", which is as follows:

[0060] S41: The objective function of the optimization model for constructing the emergency recovery candidate load set, with the goal of minimizing network loss, is:

[0061]

[0062] in, Total network loss of the distribution network; Let be the current in branch ij; Let be the resistance of branch ij;

[0063] S42: Specify the set of constraints for feasibility verification, including:

[0064] (1) Power flow constraints: The Newton-Raphson method is used to solve the power flow equations corresponding to the emergency recovery candidate load set to ensure power flow convergence and satisfy the branch power constraints.

[0065]

[0066] (2) Node voltage constraints: The voltage magnitude of all nodes must satisfy:

[0067]

[0068] (3) Generator output constraints: The active and reactive power output of each generator must be within the rated range, i.e.:

[0069]

[0070] Because some generators have high power output but low remaining energy, to prevent their energy from being depleted too quickly, the generator output is set to not exceed the average of its remaining energy and the remaining time period.

[0071]

[0072] in, The remaining energy of the generator. For the remaining time, This represents the actual maximum output of the generator.

[0073] (4) Operational constraints

[0074] Node power balance constraint: The total power flowing out of node i is equal to the sum of the power flowing into node i, expressed as:

[0075]

[0076] in, The set of nodes directly connected to node i; The active and reactive power flowing from node i to node j; The active and reactive power flowing from node j to node i;

[0077] Node injection power constraint: The power injected into node i from the outside is equal to the power emitted by the power source connected to it minus the online load power of that node, expressed as:

[0078]

[0079] S43: Substitute the emergency recovery candidate load set obtained in S3 into the objective function of the optimization model of the above emergency recovery candidate load set, and solve for the above constraints. Since the load recovery scheme has been determined in the previous step, the integer variables in the optimization problem have been determined. Therefore, with the goal of minimizing network loss, call the solver to determine how each generator should output power. Only in this way can network loss be minimized and energy utilization maximized.

[0080] If the model has a solution, the corresponding set of emergency recovery candidate loads is determined to be a feasible set of emergency recovery candidate loads, and the output of each generator in the solution results is output. (i.e., the optimal output that satisfies the goal of minimizing network loss); if the model has no solution (there is a constraint conflict or no feasible solution space), then the corresponding emergency recovery candidate load set is determined to be an infeasible emergency recovery candidate load set and is removed;

[0081] Specifically, the total benefit calculation process of the pre-simulation and feasible emergency recovery candidate load set described in step S5 is based on the multiple feasible emergency recovery candidate load sets selected in step S4. After verifying the feasibility of restoring the unrestored loads according to the principle of prioritizing loads with high importance and low power consumption, the benefit calculation is completed, and the emergency recovery candidate load set with the highest benefit is selected as the emergency recovery prospective plan for the current period. The details are as follows:

[0082] S51: For each set of feasible emergency recovery candidate loads (i.e. feasible forward-looking schemes) selected in step S4, build a pre-simulation model for emergency recovery of the distribution network; based on the current state, including load recovery status, power supply effective energy remaining status, and unrecovered load sequence (excluding loads already included in the current feasible forward-looking schemes), conduct pre-simulation according to the set heuristic basic strategy to determine the potential benefits of each forward-looking scheme.

[0083] Basic strategy: For the remaining limited effective energy, prioritize the restoration of loads with high load importance and low power consumption. Determine whether the load can be restored at a later time. If it can be restored, record the restoration time and restoration benefits until all remaining loads are unrestoreable.

[0084] S52: Run the pre-simulation model corresponding to each feasible forward-looking scheme in sequence. The core is to verify the energy feasibility of the unrestored load according to the principle of cost-effectiveness first. The specific process is as follows:

[0085] S521: Extract the remaining unrecovered load sequence corresponding to the current feasible forward-looking scheme, that is, remove the recovered load of the current scheme from the unrecovered load sequence after sorting in step S2. The sequence satisfy , where m t The number of remaining unrestored loads corresponding to the current plan.

[0086] S522: Effective residual energy calculated based on step S2 Based on the expected generator output for each subsequent period, calculate the total available energy for each subsequent period. (s is the subsequent time period number, s=1,2,...,Tt)

[0087]

[0088] in The initial effective remaining energy in time period s, The estimated active power output of the g-th generator during the s-th time period;

[0089] S523: Press The order of sorting is used to determine whether each load can recover in a later time period: for loads Calculate the energy required for its recovery. ,in For the recovery period, if a certain period exists. ,satisfy Then determine the load in the first... Time period recovery, record recovery time is T- +1, and update subsequent available energy. If the subsequent time periods all meet the requirements If the load cannot be restored temporarily, the assessment of subsequent loads will be stopped.

[0090] S524: Record the set of potential benefits corresponding to the current feasible forward-looking solutions and their corresponding recovery periods.

[0091] S53: Construct a total benefit evaluation model for each feasible forward-looking solution, total benefit Forward benefits of the current plan Potential benefits of rehearsal recovery The sum is calculated using the following formula:

[0092]

[0093] S54: Calculate forward-looking returns This refers to the benefit of restoring load through currently feasible forward-looking solutions during the current decision-making period, expressed by the following formula:

[0094]

[0095] in: This is the importance coefficient of the load. Let be the recovery state variable of the load at node i. =1 indicates that the load of node i has been restored. =0 indicates that the load on node i has not been restored; Recovery time for restoring load

[0096] S55: Calculate potential revenue This refers to the expected benefits brought by the simulated recovery load corresponding to the currently feasible forward-looking solutions. It is calculated based on the importance of the simulated recoverable load and the expected recovery time, using the following formula:

[0097]

[0098] in The expected recovery time for load i is the duration from the start of the expected recovery period to the end of the rehearsal.

[0099] S56: Calculate the total revenue for each feasible forward-looking solution. The option with the highest total revenue is selected as the set of feasible emergency recovery candidate loads for the current period. .

[0100] Specifically, the data update process described in step S6 is as follows:

[0101] Based on the forward-looking scheme determined in step S5, update the data on the restored load status, unrestored load status, and available remaining energy of the distribution network to complete a single round of online dynamic decision-making. If the system time step does not reach the preset target, return to step S2 to enter the next round of decision-making.

[0102] S61: Update the recovered load set The set of feasible emergency recovery candidate loads will be determined. Load addition ,Right now ;

[0103] S62: Update the unrecovered load set Remove forward-looking schemes from the original set of unrestored loads. The load in, i.e. ;

[0104] S63: Update available remaining energy

[0105] Through the synergistic effect of the above steps, the proposed machine intuition-based power distribution network post-disaster recovery method can dynamically adjust the power supply structure in the multi-stage time domain after a disaster, taking into account uncertainty and cost optimization, and effectively improving the system's recovery efficiency and operational robustness.

[0106] As a specific example, the invention will be further described in detail in one embodiment.

[0107] In this embodiment, an IEEE 13-node distribution network system is used as an example to verify the effectiveness of the distribution network post-disaster recovery method based on cumulative machine intuition strategy proposed in this invention.

[0108] The test system contains four generators located at nodes 1, 2, 9, and 7, with capacities of 600kW, 800kW, 1000kW, and 800kW respectively. There are eight load nodes in total, with nodes 6 and 8 being primary load nodes, nodes 4, 9, and 11 being secondary load nodes, and nodes 2, 5, and 12 being ordinary load nodes and important load nodes. Five time periods are set up.

[0109] During the post-disaster recovery process, the system adopts the proposed machine-intuitive post-disaster recovery method, adjusting the load recovery plan in different time periods to maximize the weighted duration of the load and generate the current optimal operation plan. The optimized plan implemented is shown below:

[0110] Number of time periods Load restoration plan 1 6、8、9、11 2 6、8、9、11 3 4、6、8、9、11 4 4、6、8、9、11 5 4、6、8、9、11

[0111] As can be seen from the table, the system prioritizes the restoration of loads with higher importance in the early stage. In the third period, since the remaining capacity of the system is sufficient to continue restoration, supplementary restoration is carried out to ensure the maximization of the objective function, thereby maximizing the continuity of load supply and verifying the dynamic adaptability and scheduling intelligence of the present invention in the post-disaster process.

[0112] In summary, the method of this invention has the advantages of strong real-time performance, high robustness, and the ability to exhibit "intuitive prediction" similar to human maintenance personnel in deterministic scenarios. It can significantly improve the post-disaster recovery efficiency of urban power distribution networks and reduce economic losses to the system and society.

[0113] The method of this invention is based on machine intuition and comprehensively considers the experience of personnel scheduling in large-scale urban power distribution networks. It is a dynamic decision-making technology for post-disaster emergency recovery, enhancing the system's ability to recover critical loads after a disaster, thereby improving the resilience of the urban power distribution network and increasing the response speed and decision-making quality during disaster recovery. The method of this invention combines the operating characteristics of the power distribution network to construct an optimized model suitable for disaster recovery and utilizes machine intuition to achieve efficient online decision-making.

[0114] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0115] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part 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, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0116] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0117] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A post-disaster emergency recovery decision-making method for urban power distribution networks based on machine intuition, characterized in that, include: Initialize the decision parameters and basic data of the urban power distribution network, and construct a single-time power flow optimization model without considering energy constraints to obtain the maximum load set without considering energy constraints; Based on the maximum load set satisfying power flow constraints, identify restored and unrestored loads, obtain the current system state, calculate the cost-effectiveness of each unrestored load and sort them to obtain a sorted set of unrestored loads; calculate the available effective surplus energy in the current urban power distribution network; Based on the sorted set of unrecovered loads, a set of least conservative emergency recovery candidate loads is constructed. Then, the set of least conservative emergency recovery candidate loads is expanded through permutation, combination and supplementation strategies to generate a set of emergency recovery candidate loads. Power flow feasibility is verified for each set of emergency recovery candidate loads. Power flow constraints and node voltage constraints are introduced to eliminate infeasible emergency recovery candidate load sets. The output of each generator is calculated for the remaining set of feasible emergency recovery candidate loads. Emergency recovery simulation is performed on each feasible emergency recovery candidate load set. The emergency recovery benefit corresponding to each feasible emergency recovery candidate load set is calculated. The feasible emergency recovery candidate load set with the highest benefit is selected as the emergency recovery prospective plan for the current period. Based on the emergency recovery prospective plan for the current period, update the data on the restored load status, unrestored load status, and available remaining energy of the distribution network to complete a single round of online dynamic decision-making.

2. The method according to claim 1, characterized in that, The initialization of decision parameters and basic grid data for the urban distribution network, without considering energy constraints, constructs a single-time power flow optimization model to obtain the maximum load set without considering energy constraints, including: Obtain post-disaster topology data of the urban power distribution network, including node parameters, branch parameters, generator parameters, and load parameters, to clarify the node relationships and branch connection relationships of the urban power distribution network; A single-time power flow optimization model without considering energy constraints is constructed based on the nodal power balance principle. The expression of this single-time power flow optimization model is as follows: Where N is the total number of distribution network nodes; , They are nodes With nodes The active and reactive power of the branches between them; , They are nodes The active and reactive power output of the generator; , These are the active power and reactive power of the load at node i, respectively. The voltage constraint of the single-time power flow optimization model is set as follows: in Let be the voltage amplitude at node i; , They are nodes The minimum and maximum voltage; With the goal of maximizing the weighted load recovery time, the objective function of the single-time power flow optimization model is set as follows: in, For nodes The importance factor of a load is determined based on its load level. For nodes The recovery state variables of the load. =1 indicates a node Load restoration =0 indicates a node The load has not been restored; Based on the voltage constraints, the maximum load set without considering energy constraints is obtained by solving the objective function that satisfies the single-time power flow optimization model.

3. The method according to claim 2, characterized in that, The method described above identifies restored and unrestored loads based on the maximum load set satisfying power flow constraints, obtains the current system state, calculates the cost-effectiveness of each unrestored load, and sorts them to obtain a sorted set of unrestored loads, including: Obtain the state of the distribution network system at the current decision-making moment. The status of the power distribution network system This includes: load recovery status and energy remaining status, wherein the load recovery status includes: recovery variables of each load node. , This includes both restored and unrestored loads, forming a set of restored loads and a set of unrestored loads. Define load cost-effectiveness To calculate the cost-effectiveness ratio of load restoration benefits to restoration costs, for each unrestored load i in the set of unrestored loads, calculate its cost-effectiveness. The calculation formula is: in, For nodes Importance coefficient of load, For nodes The active power of the load; Based on cost performance Sort the unrestored loads in the unrestored load set in descending order to obtain the sorted unrestored load sequence [L1, L2, ..., L]. m ], where m is the total number of loads that have not been restored, and the full .

4. The method according to claim 3, characterized in that, The calculation of the available effective surplus energy in the current urban power distribution network includes: Statistics on the current available surplus energy of all power sources in the distribution network With rated output power Where g is the power supply number, g=1,2,...,G, and G is the total number of power supplies: Calculate the effective residual energy of a single power source g. : in, Let g be the current actual remaining energy of the power source g; The rated output power of power supply g; The remaining power outage duration is the total time from this moment to the preset power restoration completion time. Calculate the available effective surplus energy in the current urban power distribution network. : 。 5. The method according to claim 4, characterized in that, The process of constructing the least conservative emergency recovery candidate load set based on the sorted unrecovered load set, and then expanding the least conservative emergency recovery candidate load set through permutation, combination, and supplementation strategies to generate an emergency recovery candidate load set, includes: The sorted unrecovered load sequence [L1, L2, ..., L] is judged sequentially. m Whether the energy of the load system in [ ] can meet its full-time recovery needs, for load L k Calculate the energy required for its continuous operation throughout the entire timeframe after recovery in the current decision-making period. ,in Let be the active power of load k, T be the estimated total power outage duration, and t be the current time. Then the load L k Added to the emergency recovery candidate load set and update the effective remaining energy. ,like If the selection fails, the screening process stops, and the set of least conservative emergency recovery candidate loads is obtained. Its load quantity is n, that is ; Select the least conservative emergency recovery candidate load set Mid-range cost performance The lowest [0.3n] loads are denoted as the subset. ,right The load in the data is permuted and combined to generate multiple non-empty subsets; Will Remove The load sets other than those specified are merged to obtain a basic candidate subset, and finally an empty set is added, where the empty set corresponds to the scheme of not restoring any new load, thus obtaining the emergency recovery candidate load set. ,in This represents the total number of candidate load sets for emergency recovery.

6. The method according to claim 5, characterized in that, The aforementioned process involves verifying the power flow feasibility of each emergency recovery candidate load set, introducing power flow constraints and node voltage constraints to eliminate infeasible emergency recovery candidate load sets, and calculating the output of each generator for the remaining feasible emergency recovery candidate load sets. The objective function of the optimization model for constructing the emergency recovery candidate load set, with the goal of minimizing network loss, is as follows: in, Total network loss of the distribution network; branch road The current; branch road The resistance; The constraints for feasibility verification of each emergency recovery candidate load set include: (1) Power flow constraints: The Newton-Raphson method is used to solve the power flow equations corresponding to the emergency recovery candidate load set to ensure power flow convergence and satisfy the branch power constraints. (2) Node voltage constraints: The voltage magnitude of all nodes must satisfy: (3) Generator output constraints: The active and reactive power output of each generator must be within the rated range, i.e.: The generator output is set not to exceed the average of its remaining energy and the remaining time period, i.e. in, The remaining energy of the generator. For the remaining time, This represents the maximum output of the actual generator. (4) Operational constraints Node power balance constraint: The total power flowing out of node i is equal to the sum of the power flowing into node i, expressed as: in, The set of nodes directly connected to node i; The active and reactive power flowing from node i to node j; The active and reactive power flowing from node j to node i; Node injection power constraint: The power injected into node i from the outside is equal to the power emitted by the power source connected to it minus the online load power of that node, expressed as: Each emergency recovery candidate load set is sequentially substituted into the objective function of the optimization model for that set to solve. If the model has a solution, the corresponding emergency recovery candidate load set is determined to be a feasible emergency recovery candidate load set, and the generator output in the solution results is output. If the model has no solution, the corresponding set of emergency recovery candidate loads is determined to be an infeasible set of emergency recovery candidate loads and is removed.

7. The method according to claim 6, characterized in that, The aforementioned emergency recovery simulation exercise for each feasible emergency recovery candidate load set, calculating the emergency recovery benefit corresponding to each feasible emergency recovery candidate load set, and selecting the feasible emergency recovery candidate load set with the highest benefit as the emergency recovery prospective solution for the current period includes: For each set of feasible emergency recovery candidate loads selected, a distribution network emergency recovery simulation model is built. Based on the load recovery status, the remaining effective energy status of the power source, and the sequence of unrecovered loads, a simulation is performed according to the set heuristic basic strategy to determine the potential benefits of each set of feasible emergency recovery candidate loads. Run the pre-simulation model corresponding to each set of feasible emergency recovery candidate loads in sequence, verify the energy feasibility of the unrecovered loads according to the principle of cost-effectiveness first, and record the potential benefit set and corresponding recovery period of the current set of feasible emergency recovery candidate loads. Construct a total benefit evaluation model for each feasible emergency recovery candidate load set, and the total benefit... Forward-looking benefits for the current set of feasible emergency recovery candidate loads Potential benefits of rehearsal recovery The sum is calculated using the following formula: Calculate forward returns This refers to the benefit derived from restoring the current set of feasible emergency recovery candidate loads during the current decision-making period, expressed by the formula: in: This is the importance coefficient of the load. Let be the recovery state variable of the load at node i. =1 indicates that the load of node i has been restored. =0 indicates that the load on node i has not been restored; The recovery time for restoring the load; Calculate potential returns The expected benefits from the simulated recovery loads corresponding to the feasible emergency recovery candidate load set are calculated based on the importance of the simulated recoverable loads and the expected recovery time, using the following formula: in The expected recovery time for load i is the duration from the start of the expected recovery period to the end of the rehearsal. Calculate the total revenue for each feasible set of emergency recovery candidate loads. The set of feasible emergency recovery candidate loads that maximizes total revenue is selected as the set of feasible emergency recovery candidate loads for the current time period. .

8. The method according to claim 7, characterized in that, The process of updating the restored load status, unrestored load status, and available remaining energy data of the distribution network based on the emergency recovery prospective plan for the current time period, and completing a single round of online dynamic decision-making, includes: Update the set of restored loads in the distribution network. The set of feasible emergency recovery candidate loads will be determined. Load addition ,Right now ; Update the set of unrestored loads in the distribution network Remove forward-looking schemes from the original set of unrestored loads. The load in, i.e. ; Update the available surplus energy of the distribution network The system completes a single round of online dynamic decision-making. If the system time step does not reach the preset target, the above process is repeated to enter the next round of decision-making.

9. The method according to claim 7, characterized in that, The aforementioned process of sequentially running the pre-simulation model corresponding to each feasible emergency recovery candidate load set to verify the energy feasibility of the unrecovered loads according to the principle of cost-effectiveness priority includes: Extract the remaining unrecovered load sequence corresponding to the current feasible emergency recovery candidate load set. The sequence satisfy , where m t This represents the number of remaining unrestored loads corresponding to the current plan; Based on effective surplus energy Based on the expected generator output for each subsequent period, calculate the total available energy for each subsequent period. s is the number of the subsequent time period, s=1,2,...,Tt; in The initial effective remaining energy in time period s, The estimated active power output of the g-th generator during the s-th time period; according to The order of sorting is used to determine whether each load can recover in a later time period: for loads Calculate the energy required for its recovery. ,in For the recovery period, if a certain period exists. ,satisfy Then determine the load in the first... Time period recovery, record recovery time is T- +1, and update subsequent available energy. If the subsequent time periods all meet the requirements If the load cannot be restored temporarily, the assessment of subsequent loads will be stopped. Record the potential revenue set corresponding to the current set of feasible emergency recovery candidate loads and their corresponding recovery periods.

Images