Three-layer planning method, system and equipment for energy storage type intelligent soft switch and medium

Abstract

The invention relates to the technical field of power system planning, and discloses a three-layer planning method, system, equipment and medium for an energy storage type intelligent soft switch, and the method comprises the steps: determining a candidate position through upper layer toughness optimization, economically and optimally configuring the capacity through a middle layer, and achieving active defense and closed loop feedback through lower layer real-time optimization. The method has the effects of coordinating disaster prevention toughness, economy and operation flexibility, forming a dynamic optimization mechanism, and remarkably improving the power supply reliability and the economic operation level of the high-proportion new energy power distribution network under extreme disasters.

Description

Technical Field

[0001] This invention relates to the field of power system planning technology, and in particular to a three-layer planning method, system, equipment and medium for energy storage-type intelligent soft switch. Background Technology

[0002] With the continuous growth in installed capacity and power generation of renewable energy sources such as wind and solar power, they are gradually replacing thermal power as the main energy source in the new power system. Against the backdrop of high-penetration renewable energy integration, the distribution network faces a dual challenge: the uncertainty of power supply capacity due to the randomness of renewable energy output, and the operational risks caused by extreme natural disasters. Therefore, it is urgent to ensure the reliability of the distribution system through resilience enhancement and operational strategy optimization.

[0003] Flexible power distribution devices, represented by smart soft open points (SOPs), have attracted much attention due to their flexible power regulation capabilities. SOPs can replace traditional tie switches, achieving dynamic balance of active power between feeders. Under fault conditions, they provide continuous and precise reactive power support through control mode switching, effectively maintaining voltage / frequency stability in non-faulty areas and ensuring uninterrupted power supply. However, SOPs are built based on fully controlled power electronic devices, with initial investment costs as high as 2.3-3.5 times that of conventional equipment. Therefore, it is urgent to conduct research on the coordinated optimization of SOP site selection and capacity setting to maximize its technical benefits while minimizing its total life-cycle investment cost.

[0004] However, the increasing penetration rate of distributed power sources in the current distribution network has exacerbated the uncertainty of system operation. Traditional energy storage-type smart soft switch (ESOP) planning methods have three shortcomings: First, the coupled impact of extreme weather disasters and wind and solar uncertainties is ignored, resulting in insufficient resilience of the planning scheme under disaster scenarios; second, the optimization model is difficult to coordinate the conflict of multiple objectives such as disaster resilience, economy and operational flexibility; and finally, the disaster response strategy is passive, lacks dynamic defense mechanism, and cannot actively adjust the operation strategy according to the warning level.

[0005] To address the coupling problem between disasters and uncertainties, this invention constructs a meteorological disaster chain model and Monte Carlo fault scenario simulation to generate typical wind-solar-disaster coupling scenarios and establishes a scenario-vulnerability correlation matrix to quantify node vulnerability. To address the multi-objective conflict problem, a three-layer collaborative planning framework (resilience optimization layer, economic optimization layer, and flexibility optimization layer) is designed. To address the passive defense problem, a three-stage active defense strategy is embedded in the lower-level real-time optimization, and the operating mode is dynamically switched according to the meteorological warning level to achieve a full-process disaster response of "prevention-emergency-recovery". Summary of the Invention

[0006] In view of the above-mentioned existing problems, the present invention provides a three-layer planning method, system, device and medium for energy storage intelligent soft switch.

[0007] Therefore, the technical problem solved by the present invention is: To address the aforementioned technical problems, this invention provides the following technical solution: a three-layer planning method for energy storage-type intelligent soft switching, comprising: using an upper-layer extreme weather resilience optimization module, based on meteorological disaster chain data and power grid topology parameters, generating fault scenarios using Monte Carlo simulation, constructing a scenario-vulnerability correlation matrix to quantify the comprehensive vulnerability of nodes, and outputting a set of ESOP candidate locations; Through the mid-level typical daily economic optimization module, an initial wind and solar power output scenario library is generated based on the ESOP candidate location set. Typical wind and solar-disaster coupling scenario sets are extracted through Wasserstein distance metric and backward reduction technology. Under the constraint of ensuring network power flow security, the optimal location and capacity configuration of ESOP are solved. Through the lower-level real-time flexibility optimization module, meteorological and power grid data are collected in real time. When the wind speed or ice thickness exceeds the warning threshold, a three-stage active defense strategy is activated. Under normal operation, the flexibility index is optimized, and real-time disaster loss data is fed back to the upper-level module to update the candidate location set. The operating cost data is fed back to the middle-level module to correct the economic model weight coefficient, forming a two-way dynamic optimization closed loop.

[0008] As a preferred embodiment of the three-layer planning method for energy storage-type intelligent soft switching described in this invention, the ESOP candidate location set includes input typhoon path probability distribution and icing growth model meteorological disaster chain data and power grid topology parameters, and generates multiple fault scenarios through Monte Carlo simulation; For each fault scenario, calculate the load loss, voltage overrun, and load recovery time of the nodes, and construct a scenario-vulnerability correlation matrix; Based on the scenario-vulnerability correlation matrix, a multi-objective optimization algorithm is used to calculate the overall vulnerability of nodes; Nodes are sorted by overall vulnerability and selected as the ESOP candidate location set.

[0009] As a preferred embodiment of the three-layer planning method for energy storage-type intelligent soft switch described in this invention, the step of solving the optimal location and capacity configuration of ESOP includes generating an initial wind and solar power output scenario library based on the ESOP candidate location set. By using Wasserstein distance to measure the historical wind and solar power output distribution characteristics and employing backward reduction techniques to iteratively eliminate the scenarios with the lowest probability weights, a set of typical wind and solar-disaster coupled scenarios is extracted. Considering the uncertainties of wind and solar power, this paper aims to find the optimal location and capacity configuration of ESOP under the constraints of ensuring network flow security, with the goals of return on investment and overall cost.

[0010] As a preferred embodiment of the three-layer planning method for energy storage intelligent soft switch described in this invention, the lower-layer real-time flexibility optimization module includes real-time data collection of wind speed, icing thickness and line load rate through an online monitoring system. When the wind speed or ice thickness exceeds the warning threshold, the three-stage active defense strategy of prevention control, emergency control and recovery control shall be activated immediately. Under normal operating conditions, optimize the system's flexibility in tracking load fluctuations; The real-time generated disaster loss data, including load reduction and recovery time, is fed back to the upper-level module, while the operating cost data is fed back to the middle-level module.

[0011] As a preferred embodiment of the three-layer planning method for energy storage-type intelligent soft switching described in this invention, the three-stage active defense strategy includes: Prevention and control phase: When the wind speed or ice thickness exceeds the first-level warning threshold, the energy storage is fully charged to SOC=100% and non-critical loads are pre-cut. Emergency control phase: If a disaster is confirmed, switch the SOP to the maximum reactive power support mode, switch the energy storage to constant pressure mode, and activate fault location and clear the fault. Recovery and control phase: If the fault has been cleared, network reconstruction and directional power transmission are performed, and loads are restored in stages based on load importance.

[0012] As a preferred embodiment of the three-layer planning method for energy storage-type intelligent soft switch described in this invention, the step of feeding back operating cost data to the middle-layer module includes feeding back disaster loss data generated in real time by the lower-layer module to the upper-layer module, triggering an annual update of the ESOP candidate location set; The operating cost data generated in real time by the lower-level module is fed back to the middle-level module to correct the weight coefficients of the economic model; By forming a dynamic optimization closed loop through two-way feedback, the planning scheme continuously approaches the global optimal solution during iteration.

[0013] As a preferred embodiment of the three-layer planning method for energy storage-type intelligent soft switch described in this invention, the method of solving the optimal location and capacity configuration of ESOP further includes using a second-order cone planning algorithm to handle non-convex nonlinear constraints, relaxing the SOP constraints, and transforming the SOP capacity constraints, active power constraints, and loss constraints into a second-order cone form. The constraints on the operation of the distribution network are relaxed, and the active and reactive power balance constraints, Ohm's law constraints, branch head power constraints and system security constraints are transformed into second-order cone forms. By constructing a cone optimization model with mathematical closure, a solver is used to efficiently solve for the optimal location and capacity configuration of the ESOP.

[0014] This invention provides a three-layer planning system for energy storage-type intelligent soft switching.

[0015] As a preferred embodiment of the three-layer planning system for energy storage intelligent soft switch described in this invention, it includes an upper-layer extreme weather resilience optimization module: responsible for long-term planning, assessing the vulnerability of each node in the power grid by simulating extreme weather fault scenarios, identifying high-risk areas, and outputting a set of candidate installation locations for energy storage intelligent soft switches; Mid-level typical daily economic optimization module: responsible for mid-term configuration. Based on the candidate locations provided by the upper layer, it comprehensively considers the uncertainty of wind and solar power output and economic objectives, and solves the optimal location and capacity setting scheme of ESOP under the constraint of satisfying the safe operation of the power grid. The lower-level real-time flexibility optimization module is responsible for short-term operation and real-time control. By monitoring meteorological and power grid data, it executes proactive defense strategies during disaster warnings, optimizes the system's operational flexibility under normal conditions, and feeds back real-time operational data to the upper and middle-level modules, driving the entire system to perform dynamic updates and optimizations.

[0016] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a three-level planning method for an energy storage type intelligent soft switch.

[0017] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a three-layer planning method for an energy storage type intelligent soft switch.

[0018] The beneficial effects of this invention are as follows: It is the first to create a resilience-economy-flexibility closed-loop optimization system. The upper layer integrates meteorological disaster chain models and Monte Carlo fault simulation, and quantifies the comprehensive risk of nodes through a scenario vulnerability matrix to output highly vulnerable candidate areas. The middle layer integrates wind and solar uncertainty scenario reduction technology to meet the investment return of operators and the cost targets of power distribution companies. The lower layer monitors disaster warnings in real time and feeds back operational data to form a two-way dynamic optimization closed loop.

[0019] Based on real-time feedback of disaster loss data, the baseline value of the vulnerable node set is updated annually. An innovative two-stage mechanism integrating configuration optimization and dynamic response is employed: in the configuration stage, a multi-objective model is established using sensitive nodes as input. Maximizing social benefit is the objective function, achieving an optimal trade-off between cost and reliability.

[0020] Develop a prevention-emergency-recovery integrated control strategy. During the prevention phase, ensure full charging of energy storage and reactive power support at SOP. During the emergency phase, perform topology reconfiguration and targeted power supply to critical loads. During the recovery phase, activate energy storage for black start and tiered recovery. Optimize flexibility indicators during normal operation to build a full-scenario defense network. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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.

[0022] Figure 1 This is a schematic flowchart of a three-layer planning method for an energy storage-type intelligent soft switch, provided as an embodiment of the present invention.

[0023] Figure 2 This is a diagram showing the connection location of a three-layer planning method for energy storage-type intelligent soft switch in a distribution network, provided as an embodiment of the present invention.

[0024] Figure 3 The diagram shows the IEEE-33 node example structure of a three-layer planning method for energy storage-type intelligent soft switching, provided as an embodiment of the present invention.

[0025] Figure 4 This is a SOP (Start of Operation) addressing and capacity determination flowchart for a three-layer planning method for energy storage-type intelligent soft switches, provided as an embodiment of the present invention.

[0026] Figure 5 This invention provides a solution for the node comprehensive vulnerability of a three-layer planning method for energy storage-type smart soft switches, as an embodiment of the present invention.

[0027] Figure 6 This invention provides a process for generating uncertain wind and solar scenarios using a three-layer planning method for energy storage-type intelligent soft switching, as an embodiment of the present invention.

[0028] Figure 7 This invention provides a three-stage active defense strategy for a three-layer planning method for an energy storage-type intelligent soft switch, as an embodiment of the present invention. Detailed Implementation

[0029] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0030] Example 1, referring to Figures 1-3 This is the first embodiment of the present invention, which provides a three-layer planning method for energy storage-type intelligent soft switching, including: S1: Through the upper-level extreme weather resilience optimization module, based on meteorological disaster chain data and power grid topology parameters, Monte Carlo simulation is used to generate fault scenarios, construct a scenario-vulnerability correlation matrix to quantify the comprehensive vulnerability of nodes, and output a set of ESOP candidate locations.

[0031] S2: Through the mid-level typical daily economic optimization module, an initial wind and solar power output scenario library is generated based on the ESOP candidate location set. Typical wind and solar-disaster coupling scenario sets are extracted through Wasserstein distance metric and backward reduction technology. Under the constraint of ensuring network power flow security, the optimal location and capacity configuration of ESOP are solved.

[0032] S3: The lower-level real-time flexibility optimization module collects meteorological and power grid data in real time. When the wind speed or ice thickness exceeds the warning threshold, it activates a three-stage active defense strategy. Under normal operation, it optimizes the flexibility index and feeds back real-time disaster loss data to the upper-level module to update the candidate location set. It also feeds back operating cost data to the middle-level module to correct the economic model weight coefficients, forming a two-way dynamic optimization closed loop.

[0033] It should be noted that the topology access structure of the flexible multi-state switch described in this method in the distribution network is as follows: Figure 2 As shown, two sets of voltage source converter (VSC) units are connected to the feeder end nodes respectively, and are electrically isolated via DC-side capacitors. This topology enables independent controllability of each VSC AC port through DC bus capacitor isolation, achieving dynamic power flow balance between feeders based on active power coordination control. Under fault conditions, the DC capacitors provide dynamic reactive power support, and the power supply continuity in non-fault areas is maintained through VSC output voltage amplitude-frequency coordinated adjustment. The implementation process of this method is based on a three-layer collaborative architecture, forming a closed-loop optimization system from disaster prevention to real-time response. The specific process is as follows: Figure 3 As shown.

[0034] Example 2, refer to Figures 4-7 As an embodiment of the present invention, based on the above embodiment, a three-layer planning method for energy storage intelligent soft switching is provided.

[0035] Furthermore, in this embodiment, step S1 uses the upper-level extreme weather resilience optimization module to generate fault scenarios based on meteorological disaster chain data and power grid topology parameters using Monte Carlo simulation, constructs a scenario-vulnerability correlation matrix to quantify the comprehensive vulnerability of nodes, and outputs a set of ESOP candidate locations. Specific steps include S101-S104: Inputting meteorological disaster chain data such as typhoon path probability distribution and icing growth dynamics model, along with power grid topology parameters, Monte Carlo simulations generate 1,000 fault scenarios. For each scenario, three key indicators are precisely quantified: load loss, voltage exceedance, and load recovery time. A scenario-vulnerability correlation matrix is ​​constructed, and the non-dominated sorting genetic algorithm (NSGA-III) is used to solve for the comprehensive vulnerability of nodes, selecting candidate locations for Emergency Response Points (ESOPs). Simultaneously, the results are dynamically updated annually based on real-time disaster loss data from lower-level feedback, ensuring continuous coverage of evolving risk hotspots within the planned area.

[0036] S101: In M=1,000 fault scenarios, perform a power flow calculation for each scenario m, traversing all nodes i∈[1,33], and statistically analyzing the load loss P_loss, voltage overrun U_yx, and load recovery time T_restore for each node in the Monte Carlo simulation. The load loss is the amount of load that must be reduced to restore the system to normal operation in that scenario; it is directly solved as an optimization variable in the power flow calculation. The voltage overrun is the value by which the node voltage exceeds the normal operating voltage in scenario m, U_yx=|U(i)−1.0|. The load recovery time can be estimated based on a lookup table pre-established using disaster intensity and maintenance resources.

[0037] In an optional embodiment, the load recovery time can also be estimated using a linear regression model. Specifically, historical disaster event data, including disaster intensity (e.g., wind speed, icing thickness), maintenance resource availability (e.g., number of maintenance teams, equipment availability), and actual load recovery time, are collected as training samples. A linear regression algorithm is used, with disaster intensity and maintenance resources as independent variables and load recovery time as the dependent variable, to train a linear prediction model. The model parameters are determined by least squares fitting. For a new fault scenario, the disaster intensity and maintenance resource parameters of that scenario are input, and the estimated load recovery time is directly output through the trained linear regression model.

[0038] In another optional embodiment, load recovery time can also be estimated using expert system rules. Specifically, based on domain expert knowledge, a set of rules is defined, such as "if the disaster intensity is high and maintenance resources are low, then the load recovery time is long." These rules cover different combinations of disaster intensity and maintenance resources, and assign a recovery time level or specific value to each combination. For each fault scenario, the corresponding rule in the rule base is matched based on the real-time monitored disaster intensity and maintenance resource status.

[0039] S102: The scenario-vulnerability correlation matrix D is the core data structure connecting Monte Carlo fault scenarios with node vulnerability. The rows of the matrix represent different random fault scenarios M=1000, and the columns represent various vulnerability indicators to be evaluated (load loss, voltage exceedance, load recovery time). Simultaneously, the vulnerability indicators are vectors of length N (the number of system nodes). The scenario-vulnerability correlation matrix stores the results of each simulation experiment, providing a data foundation for subsequent analysis of the comprehensive performance of each node under different disaster impacts.

[0040] S103: Based on the matrix D above, calculate a comprehensive vulnerability score (F_i) for each node i. This score comprehensively reflects the node's "contribution" to load loss, voltage instability, and slow recovery under numerous random disaster scenarios. Therefore, the multi-objective optimization problem is solved using the Fast Non-Dominated Ranking Algorithm (NSGA-III), an elitist strategy, and a crowding comparison operator. in, , , These represent the correlation degree of load loss, the correlation degree of self-voltage instability, and the correlation degree of recovery time, respectively, with i being the variable index.

[0041] The specific process for calculating the overall vulnerability of nodes is as follows: Each node has three objective function values ​​(F1_i, F2_i, F3_i), calculated according to the formula above. The NSGA-III algorithm is used to perform a non-dominated sort of all nodes. Within the same rank, it's impossible to compare which node is better. NSGA-III outputs the node rankings (Rank 1, Rank 2, ...), where nodes in Rank 1 are Pareto optimal. See the detailed process below. Figure 5 .

[0042] In an optional embodiment, the overall vulnerability score F_i of a node can also be calculated using a weighted summation method, specifically, For each node, the three objective function values ​​(F1_i, F2_i, F3_i) are subjected to min-max normalization to eliminate the influence of dimensions, ensuring that all index values ​​fall within the [0,1] interval. Based on system requirements, fixed weights are assigned to the three indicators (e.g., 0.5 for load loss correlation, 0.3 for self-voltage instability, and 0.2 for recovery time correlation), with a sum of weights of 1. For each node, the normalized index value is multiplied by its corresponding weight, and the sum is used to obtain the comprehensive vulnerability score F_i.

[0043] In another alternative embodiment, the overall vulnerability score F_i of a node can also be calculated using the TOPSIS method. Specifically, all nodes are treated as solutions, and three objective function values ​​(F1_i, F2_i, F3_i) are used as attributes to construct a decision matrix. The positive ideal solution (optimal value of each indicator) and the negative ideal solution (worst value of each indicator) for each attribute are calculated. For each node, its Euclidean distance to both the positive and negative ideal solutions is calculated. The overall vulnerability score F_i of each node is calculated based on the relative proximity formula (distance to the negative ideal solution divided by the sum of the distances to the positive and negative ideal solutions), with a higher score indicating greater vulnerability.

[0044] Nodes located at lower ranks (higher positions) have higher overall vulnerability. Therefore, the overall vulnerability of a node can be simply defined as the reciprocal of its non-dominant ranking rank or by its ranking rank, F_i = 1 / Rank_i.

[0045] All nodes are sorted from highest to lowest based on F_i, and the higher the ranking of the node, the higher its overall vulnerability.

[0046] S104: Based on budget and planning requirements, pre-determine the number K of candidate locations to be screened, and select the top K nodes with the highest overall vulnerability ranking to form the ESOP candidate location set. These nodes are the weakest and most critical points in the system.

[0047] Furthermore, in this embodiment, step S2, through the mid-level typical daily economic optimization module, generates an initial wind and solar power output scenario library based on the ESOP candidate location set, refines a typical wind and solar-disaster coupling scenario set through Wasserstein distance metric and backward reduction technique, and solves for the optimal ESOP location and capacity configuration under the constraint of ensuring network power flow security. Specific steps include S201-S204: Based on the candidate node set output from the upper layer, wind and solar uncertainty processing techniques are integrated. The Wasserstein distance technique is used to measure the historical wind and solar power output distribution characteristics, generating an initial scene library. Five typical disaster-wind-solar coupling scenarios, such as high wind / snow / low solar power combinations, are extracted using a backward reduction technique. Under the constraint of ensuring network power flow security, a second-order cone algorithm is used to solve for the optimal location and capacity configuration of the ESOP (Emergency Optimal Power Provider).

[0048] S201: N sets of historical data on photovoltaic, wind power, and load, constructing a random sample set. , and Construct using Dirac functions Reference probability distribution , It can be viewed as an estimate of the true probability distribution, while the Wasserstein distance represents the distance between the true probability distribution and the reference probability distribution. In the formula: Represents the true probability distribution; A random variable representing the true distribution; It is a random variable that follows a probability distribution.

[0049] Based on the above definition, construct the following... Centered on, A Wasserstein sphere with radius , i.e. In the formula: Indicates the support probability distribution All possible values ​​in the space, It is the set of Wasserstein spheres.

[0050] S202: Based on the historical power output dataset of wind and solar power, N initial scenes are generated by random sampling according to a normal distribution; the probability distribution of each scene is initialized and the target number of scenes to be reduced, n, is set; the backward reduction algorithm is used to iteratively eliminate the scene with the smallest probability weight until the number of remaining scenes meets the constraint n (see the detailed iteration process). Figure 6 The final output is a set of typical landscape power output scenarios that are probabilistically representative.

[0051] In an optional embodiment, the initial scene can also be generated through historical data resampling. Specifically, a data record is randomly selected from the historical wind and solar power output dataset as a sampling sample. Using sampling with replacement, the historical data is randomly sampled N times repeatedly, with each sample containing one complete data record (including photovoltaic, wind power, and load data), forming N initial scenes. The probability of each initial scene is set to 1 / N to ensure a uniform probability distribution.

[0052] In another optional embodiment, the initial scenarios can also be generated through uniform sampling. Specifically, historical wind and solar power output data is analyzed to determine the minimum and maximum values ​​of each variable (such as photovoltaic output and wind power output), and the value range of each variable is defined. Within the value range of each variable, N sets of scenario values ​​are randomly generated using a uniform distribution, and combined to form an initial scenario library. The probability of each initial scenario is set to 1 / N to ensure probability consistency.

[0053] The backward reduction algorithm steps are as follows: definition ,in Let be the wind speed at time t in a given scenario. For any scenario... , where j is the variable index, and we need to calculate the Kantorovich Distance (KD) between the two scenes. In this method, KD can represent the absolute value of the difference in wind speed: Search and Scene Scenes with the smallest distance ,Right now and calculate the scene Weighting: The weighting process is repeated to find the scenario with the minimum wind speed among all scenarios. ,Right now Scenarios requiring wind speed reduction Modify the scene probability to Meanwhile, the total wind speed scenario is modified to N=N-1.

[0054] Repeat the above steps until the number of scenes is reduced to the target value.

[0055] In an optional embodiment, scene reduction can also be implemented using a forward selection algorithm. Specifically, a scene is randomly selected from the initial scene library as the first typical scene and added to the typical scene set. The average distance between each remaining scene and the current typical scene set is calculated, and the scene with the largest average distance is added to the typical scene set. This iterative addition step is repeated until the number of typical scenes reaches the target value n. Then, the probability of each typical scene is adjusted to match the probability distribution of the original scene set (e.g., allocated according to the sum of probabilities of its cluster).

[0056] In another alternative embodiment, scene reduction can also be achieved through a clustering algorithm. Specifically, the initial scene library is divided into n clusters using the K-means clustering algorithm, where n is the number of target scenes. From each cluster, the scene closest to the cluster center is selected as the typical scene. The sum of the probabilities of all scenes in each cluster is assigned to the corresponding typical scene as its new probability value.

[0057] S203: System Constraints ①SOP constraints SOP capacity constraints: SOP active power constraints: Loss constraints: in, , , These represent the active power, total capacity, and active power loss transmitted through the SOP port, respectively. This is the SOP loss factor.

[0058] ② Distribution network operation constraints Active and reactive power balance constraints: Ohm's Law constraint: Branch head power constraint: System security constraints: in, , The active and reactive power flowing through the branch are, Let r be the set of all branches connected to node i. ij Let I be the resistance of the branch connecting nodes i and j. ji The current flowing through the branch circuit is ji. Let be the active power flowing from node i to node k. The reactance of the branch connecting nodes i and j. Let i be the reactive power flowing from node i to node k. Let be the voltage magnitudes at point i and node j; To inject the sum of active contributions; , , These represent the distributed power sources on the node, the active power injected by the SOP, and the active power loss, respectively. Reactive power is the same as above. , These are the upper and lower limits of the system's allowable voltage. This is the maximum current allowed to flow through the branch.

[0059] S204: Since the power flow equations and SOP constraints contain non-convex nonlinear terms, this invention adopts a second-order cone programming algorithm. By constructing a cone optimization model with mathematical closure, the computational efficiency of the GUROBI solver is improved by 3.2 times while ensuring the accuracy of the approximate solution to the original problem.

[0060] SOP constraint relaxation: Relaxation of distribution network operation constraints: Furthermore, in this embodiment, step S3, the lower-level real-time flexibility optimization module, collects meteorological and power grid data in real time. When the wind speed or icing thickness exceeds the warning threshold, it activates a three-stage active defense strategy. Under normal operation, it optimizes the flexibility index and feeds back real-time disaster loss data to the upper-level module to update the candidate location set. It also feeds back operating cost data to the middle-level module to correct the economic model weight coefficients, forming a two-way dynamic optimization closed loop. The specific steps include: The system continuously collects data on wind speed, icing thickness, and line load rate through a wide-area measurement system, enabling real-time system optimization and proactive defense. When wind speed > 25 m / s or icing > 15 mm is detected, a three-tiered defense strategy is immediately activated: Level 1 alert triggers full charging of energy storage to 100% SOC and pre-shedding of non-critical loads; further assessment determines whether a disaster has occurred. If confirmed, Level 2 alert is triggered, switching SOP to maximum reactive power support mode, switching energy storage to constant voltage mode, and initiating fault location and clearing; if the fault has been cleared, Level 3 alert is triggered, executing network reconfiguration and directional power transmission, restoring loads in stages based on importance to ensure uninterrupted power supply to critical loads. Specific procedures are detailed below. Figure 7 Under normal operating conditions, flexibility indicators are optimized, with a focus on improving the system's ability to track load fluctuations. Disaster loss data such as load reduction and recovery time generated in real time at this layer are fed back to the upper layer to trigger annual planning updates, while operating cost data is fed back to the middle layer, forming a two-way closed-loop optimization mechanism.

[0061] Example 3 is the third embodiment of the present invention, which differs from the previous two embodiments in that: This embodiment also provides an energy storage-type intelligent soft-switching three-layer planning system, including: Upper-level extreme weather resilience optimization module: responsible for long-term planning, assessing the vulnerability of each node in the power grid by simulating extreme weather failure scenarios, identifying high-risk areas, and outputting a set of candidate installation locations for energy storage-type smart soft switches; Mid-level typical daily economic optimization module: responsible for mid-term configuration. Based on the candidate locations provided by the upper layer, it comprehensively considers the uncertainty of wind and solar power output and economic objectives, and solves the optimal location and capacity setting scheme of ESOP under the constraint of satisfying the safe operation of the power grid. The lower-level real-time flexibility optimization module is responsible for short-term operation and real-time control. By monitoring meteorological and power grid data, it executes proactive defense strategies during disaster warnings, optimizes the system's operational flexibility under normal conditions, and feeds back real-time operational data to the upper and middle-level modules, driving the entire system to perform dynamic updates and optimizations.

[0062] This embodiment also provides an electronic device, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the energy storage type intelligent soft switch three-layer planning method proposed in the above embodiment.

[0063] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a three-layer planning method for energy storage-type intelligent soft switching as proposed in the above embodiment.

[0064] The storage medium proposed in this embodiment and the three-layer planning method for implementing an energy storage type intelligent soft switch proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0065] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. 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 computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0066] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A three-layer planning method for energy storage-type intelligent soft switching, characterized in that: include, Through the upper-level extreme weather resilience optimization module, based on meteorological disaster chain data and power grid topology parameters, Monte Carlo simulation is used to generate fault scenarios, construct a scenario-vulnerability correlation matrix to quantify the comprehensive vulnerability of nodes, and output a set of ESOP candidate locations; Through the mid-level typical daily economic optimization module, an initial wind and solar power output scenario library is generated based on the ESOP candidate location set. Typical wind and solar-disaster coupling scenario sets are extracted through Wasserstein distance metric and backward reduction technology. Under the constraint of ensuring network power flow security, the optimal location and capacity configuration of ESOP are solved. Through the lower-level real-time flexibility optimization module, meteorological and power grid data are collected in real time. When the wind speed or ice thickness exceeds the warning threshold, a three-stage active defense strategy is activated. Under normal operation, the flexibility index is optimized, and real-time disaster loss data is fed back to the upper-level module to update the candidate location set. The operating cost data is fed back to the middle-level module to correct the economic model weight coefficient, forming a two-way dynamic optimization closed loop.

2. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 1, characterized in that: The ESOP candidate location set includes input typhoon path probability distribution and icing growth model meteorological disaster chain data and power grid topology parameters, and generates multiple fault scenarios through Monte Carlo simulation. For each fault scenario, calculate the load loss, voltage overrun, and load recovery time of the nodes, and construct a scenario-vulnerability correlation matrix; Based on the scenario-vulnerability correlation matrix, a multi-objective optimization algorithm is used to calculate the overall vulnerability of nodes; Nodes are sorted by overall vulnerability and selected as the ESOP candidate location set.

3. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 2, characterized in that: The process of finding the optimal location and capacity configuration of ESOPs includes generating an initial wind and solar power output scenario library based on the set of candidate ESOP locations. By using Wasserstein distance to measure the historical wind and solar power output distribution characteristics and employing backward reduction techniques to iteratively eliminate the scenarios with the lowest probability weights, a set of typical wind and solar-disaster coupled scenarios is extracted. Considering the uncertainties of wind and solar power, this paper aims to find the optimal location and capacity configuration of ESOP under the constraints of ensuring network flow security, with the goals of return on investment and overall cost.

4. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 3, characterized in that: The lower-level real-time flexibility optimization module includes the ability to collect wind speed, icing thickness and line load rate data in real time through an online monitoring system. When the wind speed or ice thickness exceeds the warning threshold, the three-stage active defense strategy of prevention control, emergency control and recovery control shall be activated immediately. Under normal operating conditions, optimize the system's flexibility in tracking load fluctuations; The real-time generated disaster loss data, including load reduction and recovery time, is fed back to the upper-level module, while the operating cost data is fed back to the middle-level module.

5. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 4, characterized in that: The three-stage active defense strategy includes, Prevention and control phase: When the wind speed or ice thickness exceeds the first-level warning threshold, the energy storage is fully charged to SOC=100% and non-critical loads are pre-cut. Emergency control phase: If a disaster is confirmed, switch the SOP to the maximum reactive power support mode, switch the energy storage to constant pressure mode, and activate fault location and clear the fault. Recovery and control phase: If the fault has been cleared, network reconstruction and directional power transmission are performed, and loads are restored in stages based on load importance.

6. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 5, characterized in that: The step of feeding back operating cost data to the middle-level module includes feeding back disaster loss data generated in real time by the lower-level module to the upper-level module, triggering an annual update of the ESOP candidate location set; The operating cost data generated in real time by the lower-level module is fed back to the middle-level module to correct the weight coefficients of the economic model; By forming a dynamic optimization closed loop through two-way feedback, the planning scheme continuously approaches the global optimal solution during iteration.

7. The three-layer planning method for energy storage-type intelligent soft switching as described in claim 6, characterized in that: The solution for the optimal location and capacity configuration of the ESOP also includes using a second-order cone programming algorithm to handle non-convex nonlinear constraints, relaxing the SOP constraints, and transforming the SOP capacity constraints, active power constraints, and loss constraints into a second-order cone form. The constraints on the operation of the distribution network are relaxed, and the active and reactive power balance constraints, Ohm's law constraints, branch head power constraints and system security constraints are transformed into second-order cone forms. By constructing a cone optimization model with mathematical closure, a solver is used to efficiently solve for the optimal location and capacity configuration of the ESOP.

8. A three-layer planning system for energy storage intelligent soft switching, employing the three-layer planning method for energy storage intelligent soft switching as described in any one of claims 1 to 7, characterized in that, include: Upper-level extreme weather resilience optimization module: responsible for long-term planning, assessing the vulnerability of each node in the power grid by simulating extreme weather failure scenarios, identifying high-risk areas, and outputting a set of candidate installation locations for energy storage-type smart soft switches; Mid-level typical daily economic optimization module: responsible for mid-term configuration. Based on the candidate locations provided by the upper layer, it comprehensively considers the uncertainty of wind and solar power output and economic objectives, and solves the optimal location and capacity setting scheme of ESOP under the constraint of satisfying the safe operation of the power grid. The lower-level real-time flexibility optimization module is responsible for short-term operation and real-time control. By monitoring meteorological and power grid data, it executes proactive defense strategies during disaster warnings, optimizes the system's operational flexibility under normal conditions, and feeds back real-time operational data to the upper and middle-level modules, driving the entire system to perform dynamic updates and optimizations.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the three-layer planning method for energy storage intelligent soft switching as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the three-layer planning method for energy storage intelligent soft switching as described in any one of claims 1 to 7.

Images