Pre-disaster, post-disaster and total scheduling method for community multi-element shared energy storage system under typhoon disaster

Abstract

According to the pre-disaster, post-disaster and total scheduling method for the community multi-element shared energy storage system under the typhoon disaster, in the pre-disaster stage, through optimal layout of entity energy storage stations, mobile energy storage equipment and an electric vehicle cluster, the operation toughness of the system is guaranteed, meanwhile, the deployment cost is minimized, and resource optimal configuration is achieved; in the post-disaster stage, the recovery time, the energy loss and the scheduling cost are taken as optimization targets, constraints such as load recovery priority are combined, and the emergency power supply demand is quickly responded. A closed-loop iterative optimization mechanism is formed by feeding back a toughness index calculated by a post-disaster power supply recovery trajectory to a pre-disaster optimization objective function, a particle swarm optimization algorithm and mixed integer linear programming are adopted for joint solution, and the solution precision and efficiency are both considered.

Description

Technical Field

[0001] The present invention relates to the field of power supply, and more specifically to a pre-disaster, post-disaster and overall dispatching method for a community multi-shared energy storage system under typhoon disasters. Background Art

[0002] With the frequent occurrence of extreme weather events, how to achieve rapid recovery and efficient dispatch of community power systems in the context of natural disasters such as typhoons has become a key research direction in the power system. Currently, a large number of studies focus on how to establish emergency response mechanisms after disasters such as typhoons. The focus is on how to optimize resource scheduling strategies based on the failure rate model established before the disaster, and quickly deploy emergency repair forces to restore damaged power lines. Most existing solutions focus on optimizing dispatch paths and determining the repair order in the post-disaster phase, using heuristic algorithms to deploy emergency repair teams and emergency resources. Most current research still focuses on post-disaster emergency remediation, lacking pre-disaster energy storage equipment capacity planning and resource spatial layout design. At the same time, there is a lack of systematic modeling for the coordinated layout of various resources within the community during the pre-disaster phase. This results in delayed energy storage dispatch response and uneven distribution of emergency resource supply when disasters occur, making it difficult to achieve a balanced design of dynamic spatial configuration and response capabilities of multiple resources before a disaster. Existing solutions mostly use single shared energy storage as an isolated dispatching unit, ignoring the energy complementarity and dispatch coupling relationship between physical energy storage stations and virtual resources, making it difficult to build a generalized shared energy storage system, resulting in low efficiency of post-disaster resource supply and conflicts in dispatching strategies, which limits further improvement of system resilience. Faced with complex dynamic environments such as post-disaster traffic congestion and road damage, the path planning of existing dispatching models is generally based on the static shortest assumption, lacking the integrated modeling of disaster information, resulting in both feasibility and efficiency limitations in the actual post-disaster implementation of dispatching solutions. Although some studies have considered the return of electricity by electric vehicles through V2G, the modeling of EV user behavior is rough, and the role of EV as flexible distributed energy storage cannot be effectively brought into play, making it difficult to form an effective power support force field in disaster situations. Summary of the Invention

[0003] In order to solve the problems of poor recovery efficiency and disaster resilience of existing distribution systems in disaster scenarios, the present invention proposes a pre-disaster, post-disaster and overall scheduling method for a community multi-shared energy storage system under typhoon disasters.

[0004] In order to achieve the above technical effects, the technical solutions of the present invention are as follows: A pre-disaster dispatching method for a community multi-shared energy storage system during a typhoon disaster. The community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage dispatching center. The virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster. The physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage dispatching center. The pre-disaster dispatching method includes the following steps: Step 1: Establish a pre-disaster objective function. This function aims to optimize the layout of physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-faceted shared energy storage system before a typhoon occurs, minimizing acquisition and deployment costs while ensuring system operational resilience. Step 2: Establish a constraint function for pre-disaster optimization, which includes physical energy storage constraints, distribution network topology constraints, and power flow constraints; Step 3: Optimize the pre-disaster objective function.

[0005] A post-disaster dispatching method for a community multi-shared energy storage system during a typhoon disaster. The community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage dispatching center. The virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster. The physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage dispatching center. The post-disaster dispatching method includes the following steps: Step 1: Establish a post-disaster objective function. The post-disaster objective function is to quickly restore the physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-shared energy storage system after a typhoon, minimizing recovery time, energy loss, and scheduling costs. A post-disaster scheduling model is established based on the objective function. Step 2: Establish a constraint function for post-disaster optimization, which includes load restoration priority constraints, spatiotemporal dynamic scheduling constraints for mobile energy storage, constraints on individual and cluster electric vehicles, and constraints on parking lots and charging piles. Step 3: Solve the post-disaster dispatch model.

[0006] A method for pre-disaster, post-disaster, and overall dispatching of a community multi-shared energy storage system during a typhoon disaster includes the following steps: Based on the post-disaster scheduling solution, the power supply restoration trajectory of each node in each time period is extracted, and the corresponding resilience index of the particle is calculated. The resilience index is fed back into the pre-disaster objective function. Combined with the total investment cost, the fitness value of the particle is calculated. The particle's historical optimal position and global optimal position are updated using the particle swarm optimization algorithm, and speed and position updates are performed. If the improvement in the fitness value of several consecutive generations of particles is less than the set convergence threshold, the early stopping mechanism is triggered to terminate the iteration early. Otherwise, the next generation of optimization is executed, forming an iterative cycle of repeated linkage between pre-disaster and post-disaster until the maximum number of generations or the early stopping condition is met. When the particle swarm optimization iteration terminates, the optimal pre-disaster and post-disaster deployment plan corresponding to the particle with the best fitness value in the current particle swarm is output.

[0007] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: The present invention proposes a pre-disaster, post-disaster and overall dispatching method for a community multi-shared energy storage system under typhoon disasters. In the pre-disaster stage, by optimizing the layout of physical energy storage stations, mobile energy storage equipment and electric vehicle clusters, the deployment cost is minimized while ensuring the resilience of the system operation, thereby achieving optimal resource allocation. In the post-disaster stage, the recovery time, energy loss and dispatching cost are used as optimization targets, combined with constraints such as load recovery priority, to quickly respond to emergency power supply needs. By feeding back the resilience index calculated from the post-disaster power supply recovery trajectory to the pre-disaster optimization objective function, a closed-loop iterative optimization mechanism is formed, and a particle swarm optimization algorithm and mixed integer linear programming are used for joint solution, taking into account both solution accuracy and efficiency. In addition, by integrating physical and virtual energy storage resources, the system scheduling flexibility and resource utilization efficiency are improved, and the optimal energy flow is achieved through reasonable scheduling, which significantly reduces redundant deployment and energy waste. While improving resource utilization overall, disaster recovery capabilities and economic rationality are taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A diagram of a multi-shared energy storage system according to an embodiment of the present invention.

[0009] Figure 2 This is a two-stage flow chart of EV user evacuation and emergency repair according to an embodiment of the present invention.

[0010] Figure 3 This is a diagram of a transportation-distribution network coordinated scheduling model according to an embodiment of the present invention.

[0011] Figure 4 This is a rolling optimization flowchart shown in an embodiment of the present invention.

[0012] Figure 5 This is a flowchart of the overall solution of the total scheduling method shown in an embodiment of the present invention. DETAILED DESCRIPTION

[0013] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0014] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The singular forms "a," "the," and "the" used in this invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0015] It should be understood that although the terms "first," "second," "third," etc. may be used in the present invention to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0016] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Example 1 This embodiment provides a pre-disaster dispatching method for a community multi-shared energy storage system during a typhoon disaster. The community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage dispatching center. The virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster. The physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage dispatching center. The pre-disaster dispatching method includes the following steps: Step 1: Establish a pre-disaster objective function. This function aims to optimize the layout of physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-faceted shared energy storage system before a typhoon occurs, minimizing acquisition and deployment costs while ensuring system operational resilience. Step 2: Establish a constraint function for pre-disaster optimization, which includes physical energy storage constraints, distribution network topology constraints, and power flow constraints; Step 3: Optimize the pre-disaster objective function.

[0018] Furthermore, the present invention has constructed a community post-disaster recovery strategy system covering the trinity of "pre-disaster resource coordination deployment - post-disaster scheduling path optimization - multi-shared energy storage integration". This strategy not only breaks through the scheduling barriers between traditional isolated energy storage and temporary power sources, but also significantly improves the autonomous recovery capability and controllable scheduling level of the community power grid in extreme disaster scenarios. Through the construction of physical and virtual shared energy storage systems and the introduction of path accessibility optimization mechanisms, resource utilization efficiency and scheduling economy are improved, and a theoretical basis and practical guidance are provided for urban communities to cope with natural disasters such as typhoons, significantly improving the community's self-sustaining ability and resilience recovery under disasters, such as Figure 1 Shown is a diagram of a multi-shared energy storage system.

[0019] The overhead line system in the distribution network is composed of conductors and power towers, and its failure probability is significantly affected by extreme wind speeds. To accurately describe the evolution of equipment failure in typhoon scenarios, it is necessary to establish probabilistic failure models for conductors and towers that vary with wind speed, taking into account the structural characteristics and load-bearing differences of conductors and towers. Among them, as components directly exposed to the wind field, the failure risk of conductors mainly comes from swinging and breaking under the action of wind pressure; while the tower structure may experience overall overturning or fracture damage under high wind loads, showing typical nonlinear failure characteristics. To this end, this paper uses three-segment lognormal distribution functions to construct their failure probability models, and further integrates them into a calculation formula for the overall failure rate of the line segment based on series structure logic.

[0020] Conductors are the components of overhead transmission networks most directly exposed to wind, and their stability significantly impacts the operational reliability of power lines. At low wind speeds, conductors experience only slight oscillation and wear, and their failure rate remains at a stable lower limit. However, as wind speeds increase, the stability of the conductor structure gradually weakens, and especially after exceeding a critical wind speed, the failure rate increases significantly. Therefore, the impact of typhoon wind speeds on distribution line conductors was investigated by combining historical accident data and wind speed evolution patterns.

[0021] The conductor fault probability model expression is:

[0022] in, represents the probability of conductor failure, represents the minimum basic failure rate, Indicates the measured wind speed, represents the standard normal distribution function, 、 Indicates the initial wind speed of fault growth and the critical wind speed boundary of complete failure, Indicates the critical wind speed value for conductor structure stability, It represents the conductor fault growth factor; Due to the coupling characteristics of structural load-bearing and wind load, transmission towers are prone to structural instability or even collapse under strong wind conditions, resulting in functional interruption of the entire line section; the tower structure is usually affected by wind and exhibits obvious suddenness and criticality. Therefore, a three-segment log-normal distribution function is also used for fitting modeling.

[0023] The tower failure probability model expression is:

[0024] in, represents the tower failure probability, Indicates the turning wind speed value of the tower structure stability, 、 It represents the critical wind speed boundary where the tower failure probability begins to increase and the failure probability saturates. Indicates the fault sensitivity coefficient of the tower structure; Considering that each section of the distribution line is composed of conductors and towers, when the conductor breaks or any tower collapses, the line section will become unavailable. It is necessary to combine the failure rates of the two and conduct a joint model to calculate the availability level of the entire line. It consists of multiple towers and a section of wire. The wire and the tower are connected by wind speed. The failure probabilities are and .

[0025] The overall fault probability model of the line section is established based on the conductor fault probability model and the tower fault probability model, and its expression is:

[0026] in, represents the overall failure probability of the line segment, Indicates line segment The number of tower poles included in the above, It is a collection of distribution network lines.

[0027] To ensure the rapid response of the distribution system and restore power to critical loads after a disaster, this study aims to optimize the layout of power and energy storage resources within a community, starting from a pre-disaster pre-configuration perspective, with the goal of minimizing the combined costs of resource allocation and emergency dispatch. This phase of resource optimization focuses on the efficient deployment of diverse energy sources within the community, providing strong support for subsequent post-disaster dispatch through the rational planning of resource acquisition and deployment strategies.

[0028] This stage mainly includes the following three steps: (1) Capacity planning of physical energy storage stations: including the design of energy storage power and capacity to ensure the basic power supply capacity of the community's main power supply; (2) Scale optimization of parking lots and V2G charging piles: considering the site selection, number of charging piles, and the capacity of electric vehicles to enhance the flexible scheduling capability of the system; (3) Pre-deployment decision of mobile energy storage equipment: pre-deployment of key nodes of the power grid to ensure that mobile energy storage vehicles can access the island network as soon as possible after the disaster to carry out the task of restoring power supply.

[0029] Before a disaster strikes, the system must establish a comprehensive and responsive energy network based on the potential impact areas predicted by meteorological disaster warnings. The goal is to optimally deploy resources within the community, including physical energy storage, electric vehicles, and mobile energy storage devices, before a typhoon strikes, ensuring system resilience while minimizing acquisition and deployment costs.

[0030] In an optional embodiment, the pre-disaster objective function is as follows:

[0031] in, represents the cost of the physical energy storage station, represents the cost of parking lots and charging piles, represents the deployment cost of mobile energy storage, Represents the toughness index, represents the trade-off coefficient.

[0032] In an optional embodiment, the pre-disaster objective function is optimized. The energy storage station power and maximum capacity are used to optimize the energy storage station capacity configuration. The parking lot capacity and the number of charging piles are used to optimize the parking lot size. The connection relationship of the mobile energy storage devices is used to optimize the pre-deployment of mobile energy storage. The cost of the physical energy storage station includes investment cost and operation and maintenance cost. The cost expression of the physical energy storage station is:

[0033] in, represents the cost of the physical energy storage station, represents the discount rate coefficient, Indicates the service life of the physical energy storage station in years. 、 Represent the power and maximum capacity of the energy storage station respectively, 、 denote the unit power and capacity investment costs respectively, Indicates the unit power operation and maintenance cost, in RMB / kW per year; The cost of shared energy storage stations includes investment and operation and maintenance costs. In order to improve the independent operation capability of the community main power supply after a disaster, this study introduces a capacity optimization model for shared energy storage stations. The goal is to reasonably determine the energy storage power and capacity , achieving stable island power supply capabilities and balancing initial investment with long-term operation and maintenance costs. It determines the scale of shared energy storage station construction and affects the overall power supply capacity of the power grid.

[0034] Furthermore, the investment cost includes power investment cost and capacity investment cost. This cost function adopts an annuity factor discount mechanism to evenly distribute the one-time investment over the service life and calculate the equivalent annual investment cost of long-term investment. This can reflect the annualized economic efficiency of long-term investment and ensure that the shared energy storage station not only meets capacity requirements but also has a good return on investment. The cost of a parking lot and its charging piles includes the land cost and the construction cost of the charging piles. The cost expression of a parking lot and its charging piles is:

[0035] in, represents the cost of parking lots and charging piles, represents the unit land cost, represents the parking lot capacity, i.e. the number of electric vehicles that can be parked. Indicates the comprehensive area occupied by a single parking space. represents the construction cost of a single charging pile, Indicates the number of charging piles; The cost of parking lots and charging stations, including land costs and charging station construction costs, determines the deployment capacity of electric vehicles and affects the flexible scheduling after disasters. Therefore, the capacity planning of community parking lots and the construction of charging facilities are integrated into the cost function. The mobile energy storage deployment cost expression is:

[0036] in, represents the deployment cost of mobile energy storage, is a 0-1 variable, indicating whether the mobile energy storage device k is connected to the node n. represents the unit mobile energy storage deployment cost; The deployment cost of mobile energy storage can fully utilize the rapid response advantage of mobile energy storage and determine the pre-disaster deployment of mobile energy storage; The toughness index expression is:

[0037] in, Represents the toughness index, represents the total critical load demand within the scheduling period, represents the probability of disaster occurrence, represents the importance weight of the load node, represents the power supply required by the key load node n at time t, It indicates the actual power obtained by the key load node n at time t.

[0038] It is a resilience indicator that can calculate the power supply gap of key loads and evaluate the overall resilience of the power grid based on the probability of disasters.

[0039] Furthermore, in the pre-disaster objective function, is the trade-off coefficient, when When the power grid is increased, the focus is on the resilience of the power grid, and the cost is allowed to increase to improve the disaster recovery capability; when When the size is reduced, the focus is on minimizing construction costs and reducing the emphasis on post-disaster recovery capabilities. The value of can avoid sacrificing post-disaster recovery capabilities due to excessive compression of pre-disaster construction costs with the goal of minimizing comprehensive costs.

[0040] Pre-disaster decision variables can be divided into three categories: (1) 、 Responsible for optimizing the capacity configuration of energy storage stations, which is a continuous optimization variable and determines the power supply capacity and support duration of energy storage equipment; (2) 、 Responsible for optimizing the parking lot size, which is an integer optimization variable and determines the deployment capacity of V2G resources; (3) Responsible for the pre-deployment of mobile energy storage, which is a binary deployment variable. It determines the degree of power supply readiness for key nodes after a disaster in the pre-disaster phase.

[0041] This objective function comprehensively considers multiple dimensions, including resource acquisition, deployment, operations and maintenance, and disaster risk trade-offs. Through systematic modeling and multi-objective trade-offs, it fully embodies the multi-source, heterogeneous collaborative optimization philosophy of the pre-disaster planning phase. By integrating shared energy storage stations, mobile energy storage systems, and electric vehicles into the model framework, a total cost minimization model for the overall resource supply system is constructed, providing an optimal cost-performance solution for subsequent scheduling.

[0042] In an optional embodiment, the constraint function includes physical energy storage constraint, distribution network topology constraint and power flow constraint. The physical energy storage constraint expression is:

[0043]

[0044]

[0045]

[0046]

[0047]

[0048] in, Indicates the unit energy storage system capacity of the shared energy storage station, in kWh. represents the actual storage capacity of the energy storage station at time t, / Respectively represent the charging / discharging power of the energy storage station at time t, Indicates the rated power of the energy storage station, / Represent the charge / discharge efficiency respectively; The distribution network topology constraint expression is:

[0049]

[0050]

[0051]

[0052] in, It is a 0-1 variable that describes the line connection status. Indicates the number of distribution network nodes, represents the power flow variable, represents all adjacent nodes of node i, Represents a 0-1 variable, indicating whether node i is a power node or connected to a power node. represents all downstream nodes of node i, represents all upstream nodes of node i, represents the load demand of node i; The distribution network power flow constraint expression is:

[0053]

[0054]

[0055]

[0056]

[0057]

[0058] in, Indicates the active power provided by the distribution network, Indicates the active power of mobile energy storage, represents the active power transmitted from node j to i, represents the load demand, represents the active power transmitted from node i to node j, / is the line loss, indicating that the power from the main grid and distributed energy must meet the load and flow output of each node. represents the square of the voltage modulus, Indicates the circuit breaker status 0-1 variable, represents the slack variable.

[0059] Furthermore, the entity shared energy storage constraint expression is: (1) (2) (3) (4) (5) (6) in, Indicates the unit energy storage system capacity of the shared energy storage station, in kWh. represents the actual storage capacity of the energy storage station at time t, / Respectively represent the charging / discharging power of the energy storage station at time t, Indicates the rated power of the energy storage station, / Represent the charge / discharge efficiency respectively; Furthermore, formula (1) represents the capacity constraint of the physical energy storage station, where is the unit energy storage system capacity of the shared energy storage station, in kWh. Formula (2) represents the actual state of charge constraint of the energy storage station, which is used to limit the charge and discharge energy of the shared energy storage station to not exceed its rated SOC range to prevent overcharging or over-discharging of the energy storage. is the actual storage capacity of the energy storage station at time t. Formulas (3) and (4) are the charging and discharging power constraints of the energy storage station, which limit the instantaneous power during the charging and discharging process to avoid overload operation. / are the charging / discharging power of the energy storage station at time t, is the rated power of the energy storage station. Formula (5) ensures that charging and discharging behaviors will not be carried out simultaneously. Formula (6) represents the energy dynamic balance constraint, which is used to update the energy storage state and express the charge change of the energy storage station between consecutive moments, where / are the charging / discharging efficiency, respectively, to ensure that the energy storage at time t+1 depends on the energy storage at time t and the charge / discharge amount at time t.

[0060] During multi-period post-disaster recovery, the topology of the distribution network continuously changes as faults are repaired and resources are accessed. To adapt to the dynamic evolution of the number of isolated islands in the distribution network and ensure that the tree-like structure of the distribution network is maintained during system operation, this paper introduces line state variables, node connectivity constraints, and acyclicity constraints to characterize the feasible topology of the distribution network at any time.

[0061] The distribution network topology constraint expression is: (7) (8) (9) (10) in, It is a 0-1 variable that describes the line connection status. Indicates the number of distribution network nodes, represents the power flow variable, represents all adjacent nodes of node i, Represents a 0-1 variable, indicating whether node i is a power node or connected to a power node. represents all downstream nodes of node i, represents all upstream nodes of node i, represents the load demand of node i; Formula (7) represents the topological acyclic constraint, which ensures that the distribution network maintains a tree topology and that the system has no closed loops. It is a 0-1 variable that describes the line connection status. is the number of distribution network nodes. Formula (8) represents the power supply direction constraint, where is a flow variable that describes the flow direction. This constraint uses the large M method to constrain the flow value to take effect when the connectivity state is 1, otherwise it is forced to 0, ensuring that the flow direction is affected by the topology switch. Formula (9) represents the node connectivity constraint, where are all adjacent nodes of node i, Is a 0-1 variable, indicating whether node i is a power node or connected to a power node. If it is 1, it means that node i needs to be powered, ensuring that the power supply or the node connected to the power supply has a connected branch. If it is 0, it means that node i is not necessarily connected, and nodes that are allowed to be powered off are not required to be connected. Formula (10) is the subnet power balance constraint. During the post-disaster recovery process, the power balance needs to be ensured within each island. are all downstream nodes of node i, are all upstream nodes of node i, is the load demand of node i, and the power supply is provided by the upstream power flow.

[0062] Based on the topological structure, this patent adopts a relaxed second-order cone DistFlow power flow model, comprehensively considering multiple types of energy supply from the main grid, shared energy storage, mobile energy storage, etc., and models the voltage, current and power transmission to ensure the physical feasibility and safety of the power system operation.

[0063] The distribution network power flow constraint expression is: (11) (12) (13) (14) (15) (16) in, Indicates the active power provided by the distribution network, Indicates the active power of mobile energy storage, represents the active power transmitted from node j to i, represents the load demand, represents the active power transmitted from node i to node j, / is the line loss, indicating that the power from the main grid and distributed energy must meet the load and flow output of each node. represents the square of the voltage modulus, Indicates the circuit breaker status 0-1 variable, represents the slack variable; Formulas (11) and (12) represent the active and reactive power balance constraints of the nodes, where Active power provided to the distribution network, is the active power of mobile energy storage, is the active power transmitted from node j to node i, is the load demand, is the active power transmitted from node i to node j, / is the line loss, indicating that the power from the main grid and distributed energy must meet the load and flow output of each node. Formula (13) represents the node voltage constraint, where is the square of the voltage modulus, which stipulates that the node voltage must be kept within the specified upper and lower limits to avoid voltage being too high or too low. Formula (14) represents the line current constraint, which stipulates that the line current does not exceed the rated current of the equipment, where is the line switch state 0-1 variable. Formula (15) expresses the line voltage drop constraint, where is a slack variable, allowing a certain error, Formula (16) represents the line power constraint and its second-order cone relaxation, which limits the line transmission power to a safe range.

[0064] In an optional embodiment, optimizing the pre-disaster objective function includes the following steps: The particle swarm optimization algorithm is used to optimize the pre-disaster objective function; The current particle pre-disaster deployment plan is generated based on the current particle code. The resilience index of the plan is calculated based on the load recovery capability and scheduling cost corresponding to each particle. While ensuring the resilience of system operation, the objective function is to minimize the total investment cost and the indirect losses caused by insufficient load recovery after the disaster. The particle swarm position is iteratively optimized through particle swarm speed update and position adjustment until the optimization target is reached and the iteration ends.

[0065] In the pre-disaster stage, this patent proposes a pre-dispatch strategy based on typhoon disaster warning and failure probability models. By analyzing historical data and meteorological warning information, a predictive model of the failure rate of power grid facilities is constructed to complete the coordinated planning and allocation of multiple types of resources in the community. This includes: optimizing the capacity and power design of shared energy storage stations, rationally planning the site selection and scale construction of several parking lots in the community, and pre-deploying mobile energy storage resources. At the same time, the number and layout of charging piles configured in the parking lot are considered to build a power resource network with scheduling potential before the disaster. This strategy can form the optimal layout of resource spatial distribution before the disaster occurs, providing a guarantee for subsequent rapid response and load recovery after the disaster.

[0066] Example 2 This embodiment provides a post-disaster scheduling method for a community multi-shared energy storage system during a typhoon disaster. The community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage scheduling center. The virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster. The physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage scheduling center. The post-disaster scheduling method includes the following steps: Step 1: Establish a post-disaster objective function. The post-disaster objective function is to quickly restore the physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-shared energy storage system after a typhoon, minimizing recovery time, energy loss, and scheduling costs. A post-disaster scheduling model is established based on the objective function. Step 2: Establish a constraint function for post-disaster optimization, which includes load restoration priority constraints, spatiotemporal dynamic scheduling constraints for mobile energy storage, constraints on individual and cluster electric vehicles, and constraints on parking lots and charging piles. Step 3: Solve the post-disaster dispatch model.

[0067] After the disaster, the distribution network lost main grid power, and several lines failed. Based on the severity of the disaster, the emergency command center dispatched resources such as mobile energy storage deployed within the community and electric vehicles parked in parking lots, connecting them to the corresponding nodes to provide power restoration to the distribution network, achieving optimal distribution of power in time and space. Dynamic adjustments were made to the motion switches to optimize the distribution network topology, while shared energy storage equipment within the community was requisitioned to maximize power restoration for critical loads. This leveraged the advantages of multiple distributed resources within the community, including physical shared energy storage, mobile energy storage, and electric vehicles, to achieve phased, multi-resource coordinated recovery and dispatch of the post-disaster power grid.

[0068] This stage mainly includes the following three steps: (1) Island power supply operation of physical shared energy storage stations: When the main grid fails after a disaster, the island operation mode is first activated to provide stable support capabilities for the load in the adjacent area; (2) Dynamic scheduling of mobile energy storage vehicles: Establish a scheduling path optimization mechanism based on the traffic integration coefficient to characterize the accessibility and response time of mobile energy storage at different stages after the disaster; (3) Electric vehicle cluster scheduling modeling: Model the individual EV path, access node, and power selection, and then construct a cluster scheduling model for EV through the Minkowski summation method.

[0069] In order to coordinate the comprehensive goals of recovery time, power supply and demand, and resource allocation, and ensure that the post-disaster distribution system can respond quickly and restore power supply to key load nodes, this study starts from the perspective of post-disaster multi-source emergency collaborative dispatch and constructs a multi-objective collaborative recovery optimization model with the comprehensive goals of minimizing recovery time, energy loss, and dispatching costs. On the basis of ensuring the rapid recovery of key loads, it takes into account both dispatching economy and grid operation efficiency.

[0070] In an optional embodiment, the post-disaster objective function is as follows:

[0071] in, represents the load importance weight, represents the power grid restoration time in region i, Energy loss refers to the loss of energy supply due to power outages, that is, unmet electricity demand. represents the dispatch cost, including the dispatch cost of EV and mobile energy storage; The energy loss expression is:

[0072]

[0073] in, represents energy loss, represents the power demand of load node n at time t, represents the actual power supply provided by the energy storage devices (EV, MES and shared energy storage) to the load node n at time t; Dispatch cost, which is used to calculate the total cost of mobile energy storage and electric vehicles to complete the emergency power supply task during the dispatch process, including equipment driving costs and the cost of subsidies for EV owners participating in the dispatch; The scheduling cost expression is:

[0074] in, represents the scheduling cost, 、 Indicates the total distance traveled by mobile energy storage and electric vehicles, 、 are the unit dispatching costs of mobile energy storage and electric vehicles, Represents the cost of incentive subsidies given to EV owners.

[0075] Furthermore, post-disaster decision variables can be divided into three categories: (1) MES: Responsible for determining the MES operation trajectory, which is a binary variable and constructs a spatial dynamic scheduling path; 、 They are responsible for controlling the MES charging and discharging state and optimizing its charging and discharging power, respectively. They are binary behavioral variables and continuous variables, used for dynamic power balance modeling. (2) EV: Responsible for determining the EV position change, which is a binary variable reflecting the start and end status of the path; 、 They are responsible for controlling the EV charging and discharging status and optimizing its charging and discharging power, which are binary behavior variables and continuous variables respectively; Responsible for constraining parking lot capacity, which is a binary variable to prevent parking lot overload; (3) Responsible for controlling the recovery of critical loads. It is a continuous variable that reflects the degree of recovery distribution of each node under the load priority recovery strategy.

[0076] This objective function comprehensively considers multiple dimensions, including the distribution system's load recovery speed, the energy supply-demand gap, and the cost of multi-source resource scheduling, to construct a multi-indicator collaborative optimization model that reflects the resilience-oriented and economic trade-offs in resource allocation during the post-disaster period. By introducing key load priority weights and electric vehicle and mobile energy storage scheduling behavior variables, a rapid response and scheduling execution model for community emergency resources after a disaster is formed. The post-disaster objective function combines the islanding operation of shared energy storage stations, the discharge response of electric vehicles, and the path optimization scheduling strategy for mobile energy storage to construct an integrated recovery model that couples the "time-space-energy" multi-domain, ultimately achieving multi-objective collaborative recovery in resource scheduling and power restoration.

[0077] With the accelerated advancement of transportation electrification, the increasing proportion of electric vehicles can form a distributed mobile energy storage network with large energy storage capacity and flexible topology. It can also conduct two-way energy interaction with the distribution network through charging piles with V2G (Vehicle-to-Grid) capabilities at the site. It can play a reverse energy supply role in scenarios such as post-disaster main grid outages and load instability in isolated areas. In post-disaster scheduling, it can assist emergency mobile energy storage to support the recovery of the distribution network.

[0078] Assuming that the EV charging stations and EVs in each parking lot within the community meet the "vehicle-grid interaction" requirements, power can be transmitted to the grid. After a disaster, the mobility of EVs can be fully utilized to achieve small-scale self-healing. EVs connected to V2G charging stations can then provide power to the load through battery discharge.

[0079] EVs are user-side resources, and their ability to effectively participate in post-disaster recovery scheduling depends on the rationality and safety of scheduling instructions. To guide EV users in participating in power restoration after a typhoon disaster, the post-disaster behavior of EV users is divided into an evacuation phase and a response phase. When a typhoon makes landfall and the disaster begins, the community enters the evacuation phase: Assuming that all parking lots in the community have good evacuation capabilities, an EV user compensation mechanism can be used to guide EVs traveling on the road and near parking lots to seek refuge at the parking lots. In other words, during the evacuation phase, EVs that were originally on the road seek refuge at the nearest parking lot; EVs that were already parked at the parking lot stop charging.

[0080] It is necessary to screen the power of all EVs parked at parking nodes after evacuation guidance so that there is enough power to participate in unified scheduling in the subsequent response phase:

[0081] in, Indicates the amount of electricity when the EV enters the parking lot node. Indicates on-site emergency repair time, approximately 3-5 hours.

[0082] At the same time, an incentive compensation mechanism should be established for EV owners participating in the dispatch, which consists of two parts. 1) Positive incentive: According to the discharge amount output by EV to the system during the post-disaster dispatch process, the discharge incentive unit price Give rewards; 2) Negative settlement: After the post-disaster dispatch is completed, considering that the owner needs to leave the parking lot, the system will replenish the EV to restore it to full power after the total emergency repair is completed. The charging amount is the EV's upper limit minus the amount of power when the owner parks the EV in the parking lot. The electricity price is lower than the market price. According to the replenishment cost coefficient Charge the car owner. The incentive compensation mechanism is as follows:

[0083] in, is the discharge capacity compensation coefficient, is the market electricity price, is the electricity price subsidy coefficient.

[0084] When the intensity of the typhoon damage decreases to the point where it no longer has the ability to damage the urban distribution network, the community enters the response phase. EVs that meet the power screening conditions in the parking lot will participate in the restoration of the distribution network power supply under the guidance of the incentive mechanism. The two-stage flow chart of EV users taking refuge and emergency repair is as follows: Figure 2 shown.

[0085] In order to ensure the feasibility of the scheduling model, it is necessary to limit the range of electric vehicle scheduling nodes, defining that electric vehicles can only be dispatched in parking nodes with V2G capabilities. Dispatching is carried out between the V2G charging piles in the parking lot to recharge the electricity after the disaster. The relevant variables are as follows: etc. Only in the parking lot node collection The above definition is as follows:

[0086] EV individuals must meet power constraints during post-disaster scheduling.

[0087] In actual post-disaster dispatch, the energy storage characteristics and dispatchability of individual EVs are affected by numerous factors. Because EV user behavior patterns are difficult to accurately predict, the dispatch of individual EVs is subject to significant uncertainty, making stable energy storage management difficult. Furthermore, directly modeling each EV introduces a large number of variables, increasing the complexity of the solution. Furthermore, predicting the travel characteristics of individual electric vehicles is currently extremely difficult.

[0088] The EV cluster modeling method can effectively reduce the impact of individual uncertainty on the overall system and provide more stable energy storage scheduling capabilities. Minkowski summation is an addition method applicable to Euclidean space. Using this idea, EV clusters can be aggregated into equivalent generalized energy storage devices. The upper and lower boundary parameters of the EV cluster determine its potential as a flexible load resource. Through this method, the characteristics of individual EVs can be uniformly described at the cluster level, reducing the model dimension and giving the EV cluster good scheduling flexibility, thereby improving its controllability as a distributed energy storage resource.

[0089] The shared energy storage model for an EV cluster is as follows:

[0090] in, represents the state of charge (SOC) of the e-th EV when it arrives at the charging station, represents a 0-1 variable, judging whether the EV arrives at the shared energy storage site at time t. represents the amount of electricity in the electric vehicle at time t, / represents the charge and discharge power at time t, / Indicates a 0-1 variable, indicating whether the electric vehicle is in the charging or discharging state. When calculating the virtual shared energy storage model of the EV cluster, it is also necessary to calculate ,considering the impact of a single EV on the overall SOC of the EV cluster when entering and exiting the parking lot,,it ensures that the system can accurately track the energy boundary changes while the cluster,dynamically changes.

[0091] To enhance the independent power supply capabilities of community power grids in isolated post-disaster conditions, a physical shared energy storage system model was constructed. By incorporating a shared economy operating model, shared energy storage devices can serve as a primary power interaction within the community before a disaster, optimizing daily power distribution for the entire community. Post-disaster, due to their dedicated power supply lines, they can serve as a temporary primary power source for the community in the event of a minor disaster. This can form an island with nearby loads, maintaining continuous power supply throughout the island repair period, effectively enhancing the resilience of the grid's disaster prevention and mitigation capabilities.

[0092] In an optional embodiment, the constraint function includes a load recovery priority constraint, a mobile energy storage spatiotemporal dynamic scheduling constraint, an electric vehicle individual and cluster constraint, and a parking lot and charging pile constraint. The load recovery priority constraint expression is:

[0093]

[0094] in, represents the load recovery degree variable, / Indicates the maximum load of the node under normal operation; The spatiotemporal dynamic scheduling constraints of mobile energy storage include node scheduling constraints and operating power constraints. The node scheduling constraint expression is:

[0095] in, Indicates that the mobile energy storage moves from i to j at time t, which is a 0-1 variable, indicating that the scheduling path of the mobile energy storage in different time periods should maintain consistency; The operating power constraint expression is:

[0096]

[0097]

[0098] in, represents the amount of electricity of the mobile energy storage device e at time t, Indicates the self-discharge rate of mobile energy storage, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the mobile energy storage is in the charging or discharging state; The constraints on individual and cluster electric vehicles include the individual EV power constraint and the upper and lower bound constraints of the EV cluster. The expression of the individual EV power constraint is:

[0099]

[0100]

[0101] in, represents the amount of electricity in the electric vehicle at time t, represents the self-discharge rate of electric vehicles, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the electric vehicle is in the charging or discharging state. is a 0-1 variable, which determines whether node i is a parking lot node; The upper and lower boundary constraint expressions of the EV cluster are:

[0102] in, represents the total number of EV individuals in the EV cluster; The constraint expressions of parking lots and charging piles are:

[0103]

[0104] in, is a 0-1 variable, indicating whether the electric car e is in the parking lot node n. represents the maximum capacity of parking lot node n, represents the set of all parking lot nodes, is the number of charging piles in parking lot node n.

[0105] Furthermore, during the post-disaster recovery phase, the loads of different nodes may not be fully restored in a short period of time due to the limitations of energy resources, emergency repair progress, and traffic accessibility. In order to reasonably arrange the resource scheduling order, a priority scheduling mechanism based on load importance is introduced. By assigning load importance weight coefficients to different load nodes, To control the degree of recovery of each load , so that the more important the nodes are, the higher the degree of recovery will be, guiding the scheduling system to give priority to restoring the load of important nodes, forming a hierarchical and gradual recovery strategy.

[0106] Specifically, during the typhoon disaster warning phase, the system issues evacuation instructions and recommended nearby shelters to community residents through multiple channels, including community broadcasts, text message alerts, and app push notifications. Pedestrians and electric vehicles can seek shelter in the nearest parking lot; residents in fixed locations such as buildings, schools, and shopping malls must remain inside awaiting rescue. During the post-disaster emergency repair phase, priority is given to lighting and communications power supply at these nodes to ensure essential life needs and the smooth flow of rescue information. Furthermore, medical institutions and emergency command centers, which require continuous equipment operation, can be designated as special nodes. When prioritizing loads, these critical nodes can be given the highest priority to prioritize the safety of community residents. Only after completing emergency rescue efforts for all evacuees will the system begin resource allocation, emergency repairs, and power restoration at non-rigid nodes, such as commercial and industrial facilities, to maximize overall disaster rescue and post-disaster recovery efficiency.

[0107] The load recovery priority constraint expression is: (17) (18) in, represents the load recovery degree variable, / Indicates the maximum load of the node under normal operation; Formula (17) represents the load power recovery degree, where is the load recovery degree variable, , control the proportion of load recovery, / is the maximum load of the node under normal operation. Formula (18) represents the recovery priority control constraint. To ensure that high-importance nodes are restored first, set The lower limit and load weight The greater the load weight, the higher the degree of recovery required. This model allows the dispatch system to prioritize power needs at key nodes like medical care, communications, and emergency shelters when power resources are limited, maximizing the overall resilience and recovery efficiency of the community.

[0108] Mobile energy storage (MES) has flexible spatial transfer capabilities and stable energy storage and replay characteristics. It can be connected to the power distribution cabinet in a plug-and-play manner. It can be used as a distributed power source to cooperate with shared energy storage stations to provide emergency power support for isolated islands. Its advantages are gradually becoming apparent in post-disaster emergency power restoration. The mobile energy storage in this article is considered to be the emergency repair mobile energy storage of the power supply bureau. In a grid-transportation network convergence system, the dispatch state of mobile energy storage is determined by both its traffic state and its charge and discharge state, exhibiting spatiotemporal coupling characteristics. Therefore, a spatiotemporal dynamic dispatch model for mobile energy storage is established, consisting of three main components: path continuity constraints, traffic flow coefficient modeling, and energy state constraints. It is assumed that mobile energy storage does not consume electricity during dispatch.

[0109] The spatiotemporal dynamic scheduling constraints of mobile energy storage include node scheduling constraints and operating power constraints. The node scheduling constraint expression is:

[0110] in, Indicates that the mobile energy storage moves from i to j at time t. It is a 0-1 variable, indicating that the scheduling path of the mobile energy storage in different time periods should maintain consistency, that is, the end point of any scheduling period is the starting point of the next period, and no scheduling interruption or jump behavior should occur; Since the real-time traffic capacity of the traffic network is affected by real-time road conditions and various emergencies, it is highly uncertain. To ensure the reliability of the passage of emergency mobile energy storage equipment in disaster scenarios, a traffic integration coefficient is constructed to reflect the spatial scheduling relationship between the actual travel time of mobile energy storage and the disaster damage coefficient between various road sections after the disaster. To this end, based on the BPR (Bureau of Public Roads) delay function, its original modeling ideas for traffic flow are rewritten into a modeling method for the degree of road damage, and a traffic integration coefficient model suitable for post-disaster scenarios is constructed, which is expressed as follows: (19) (20) Formula (19) represents the traffic integration coefficient, which is used to quantify the road accessibility of the post-disaster traffic network, where is the disaster damage coefficient, with a value between 0 and 1, where 0 means unobstructed traffic and 1 means complete obstruction. is the weight coefficient, which regulates the impact of traffic delay. It is a nonlinear growth index that controls the steepness of the curve. By introducing the traffic integration coefficient , combining the impact of road accessibility and disaster damage to describe the actual road conditions; Formula (20) represents the actual travel time constraint after the disaster, which is used to calculate the equivalent travel time after the disaster, where is the actual travel time of road ij after the disaster, is the actual physical distance between nodes, is the ideal travel speed. This model enables optimal path selection and travel time estimation, providing a time-optimized basis for dispatching routes. After receiving dispatch instructions, the mobile energy storage system selects the optimal path based on the actual travel time between nodes under disaster conditions, achieving the shortest possible travel time to the target node.

[0111] Mobile energy storage must also meet the following operating power constraints during the dispatch process. The operating power constraint expression is: (twenty one) (twenty two) (twenty three) in, represents the amount of electricity of the mobile energy storage device e at time t, Indicates the self-discharge rate of mobile energy storage, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the mobile energy storage is in the charging or discharging state; Formula (21) represents the power boundary constraint, where is the amount of electricity in the mobile energy storage device e at time t, which keeps the SOC of the mobile energy storage device within a reasonable range to prevent overcharging or over-discharging. Formula (22) represents the dynamic balance constraint of electricity, describing the energy update process of the mobile energy storage at each time, where is the self-discharge rate of mobile energy storage, / is the charge and discharge power at time t, / is the charge and discharge efficiency. Formula (23) represents the charge and discharge power and behavior constraints, where / It is a 0-1 variable, indicating whether the mobile energy storage is in the charging or discharging state, preventing the mobile energy storage device from charging and discharging at the same time to avoid energy conflicts.

[0112] Constraints on individual and cluster electric vehicles include individual EV power constraints and upper and lower bound constraints on the EV cluster. Individual EVs must meet the following power constraints during post-disaster scheduling: (twenty four) (25) (26) in, represents the amount of electricity in the electric vehicle at time t, represents the self-discharge rate of electric vehicles, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the electric vehicle is in the charging or discharging state. is a 0-1 variable, which determines whether node i is a parking lot node; Formula (24) represents the power boundary constraint, where is the amount of electricity in the electric vehicle at time t, to prevent overcharging or over-discharging. Formula (25) represents the dynamic balance constraint of electricity, describing the energy update process of the electric vehicle at each time, where is the self-discharge rate of electric vehicles, / is the charge and discharge power at time t, / is the charge and discharge efficiency. Formula (26) represents the charge and discharge power and behavior constraints, where / It is a 0-1 variable, indicating whether the electric vehicle is in the charging or discharging state, preventing charging and discharging at the same time to avoid energy conflicts. It is a 0-1 variable that determines whether node i is a parking lot node and restricts EVs to dispatch and charge and discharge only between parking lot nodes.

[0113] A bottom-up modeling strategy is adopted to first characterize the energy boundary, energy evolution behavior and charging and discharging strategy of a single EV, then the behavioral boundary of the EV cluster is equivalently encapsulated, and the EV group is further abstracted into a virtual shared energy storage unit.

[0114] The upper and lower boundary constraint expressions of the EV cluster are:

[0115] in, Represents the total number of EVs in the EV cluster. By determining the upper and lower boundaries of the EV cluster, the EV cluster can be coordinated and controlled as a whole in terms of power scheduling and power management, avoiding the complexity of vehicle-by-vehicle modeling while also preserving the scheduling characteristics of EV devices at the boundary constraint level. In order to reasonably control the spatial deployment of electric vehicles and the access behavior of charging pile nodes in the community and ensure that the EV scheduling process meets the facility capacity, the parking lot and charging pile constraint expressions are as follows: (27) (28) in, is a 0-1 variable, indicating whether the electric car e is in the parking lot node n. represents the maximum capacity of parking lot node n, represents the set of all parking lot nodes, is the number of charging piles in parking lot node n.

[0116] Formula (27) represents the parking lot capacity constraint, where is a 0-1 variable, indicating whether the electric car e is in the parking lot node n. is the maximum capacity of parking lot node n, is the set of all parking lot nodes. At the same time, the total number of EVs entering parking lot node n does not exceed its maximum capacity limit. Formula (28) represents the charging pile resource constraint, where is the number of charging piles in parking lot node n. The number of electric vehicles actually participating in charging and discharging at parking lot node n does not exceed the number of charging piles configured in the parking lot.

[0117] In an optional embodiment, solving the post-disaster scheduling model includes the following steps: The system process is divided into multiple rolling optimization cycles. Prediction and optimization are performed based on the current system status in each cycle. The rolling optimization cycle includes a prediction period and an execution period. With the objective function of minimizing load loss, recovery time, energy loss and scheduling cost, a post-disaster scheduling model is established based on the constraint function of post-disaster optimization. The post-disaster scheduling model is solved by the Gurobi solver to output the power supply restoration trajectory, resource movement path and time schedule for each node and time period. The optimization of each cycle only implements the scheduling of the execution period. After the scheduling is completed, the model is updated and the prediction and optimization of the next cycle are started until all post-disaster recovery tasks are completed.

[0118] Furthermore, a rolling optimization method is used to realize the optimization of transportation-distribution network coordinated dispatch and the execution of restoration plan. The transportation-distribution network coordinated dispatch framework is as follows: Figure 3 shown.

[0119] In order to deal with the uncertainty issues such as resource status changes, traffic road accessibility changes, and demand load fluctuations in the post-disaster power load recovery process, this patent proposes to introduce a rolling optimization method in the post-disaster overall system optimization and scheduling process. The specific process is as follows Figure 4 shown.

[0120] In the rolling optimization phase, the system first Obtain various energy storage resources (including mobile energy storage units, electric vehicle clusters, shared energy storage systems) and distribution network operating status information as initial input. Then, the system faces a fixed forecast period in the future. In the short term, a rolling optimization model is constructed based on factors such as current resource availability, traffic accessibility, and load recovery requirements. The forecast period is generally several times the execution period, that is, each optimization considers several future periods. , but only the first one is actually executed In each forecast period, the optimization is carried out to comprehensively minimize multiple objectives such as load recovery time, energy loss and scheduling cost. After the optimization model is solved, the system only performs the execution period. Scheduling decisions within.

[0121] After the control time domain is executed, the system updates the operating status of all resources, node load recovery status and traffic road accessibility information based on the actual execution results, and rolls the time window by one time step. , with a new starting point The prediction time domain and optimization model are reconstructed to execute the next round of optimization and scheduling. This repetitive cycle forms a rolling scheduling closed loop of "prediction-optimization-execution-feedback" to achieve dynamic adaptation to complex post-disaster environmental changes and continuous updating of scheduling plans.

[0122] After a disaster, the system enters a dynamic scheduling phase, with the main grid failing and some lines damaged. Based on disaster feedback, the emergency command center rapidly activates the post-disaster resource scheduling mechanism, centrally dispatching three types of distributed energy storage resources: shared energy storage stations, mobile energy storage devices, and electric vehicles. Shared energy storage stations, leveraging their stable capacity and dedicated power lines, are first activated as temporary primary power sources in areas capable of island operation, providing continuous power to surrounding critical loads. Mobile energy storage devices, leveraging their access flexibility and centralized capacity, prioritize power supply to islanded areas while meeting accessibility and energy constraints. Electric vehicle clusters utilize V2G charging stations for fixed-point power feedback, employing incentive mechanisms to guide user participation in discharge tasks. The scheduling path, taking into account the post-disaster traffic damage coefficient, constructs a traffic integration factor and establishes a dynamic optimal path selection model for post-disaster operations, further ensuring the feasibility and time efficiency of resource scheduling.

[0123] Example 3 Based on Examples 1 and 2, this embodiment proposes a method for pre-disaster, post-disaster, and overall dispatching of a community multi-shared energy storage system during a typhoon disaster, including the following steps: Based on the post-disaster scheduling solution, the power supply restoration trajectory of each node in each time period is extracted, and the corresponding resilience index of the particle is calculated. The resilience index is fed back into the pre-disaster objective function. Combined with the total investment cost, the fitness value of the particle is calculated. The particle's historical optimal position and global optimal position are updated using the particle swarm optimization algorithm, and speed and position updates are performed. If the improvement in the fitness value of several consecutive generations of particles is less than the set convergence threshold, the early stopping mechanism is triggered to terminate the iteration early. Otherwise, the next generation of optimization is executed, forming an iterative cycle of repeated linkage between pre-disaster and post-disaster until the maximum number of generations or the early stopping condition is met. When the particle swarm optimization iteration terminates, the optimal pre-disaster and post-disaster deployment plan corresponding to the particle with the best fitness value in the current particle swarm is output.

[0124] Furthermore, the two-stage optimization model is jointly solved by the particle swarm optimization algorithm and the mixed integer linear programming method.

[0125] Furthermore, the overall solution framework adopts a two-stage optimization model that combines upper-level pre-disaster resource deployment optimization with lower-level post-disaster dynamic resource scheduling. The flow chart is as follows: Figure 5 As shown, the problem is solved by combining the particle swarm optimization (PSO) algorithm with the mixed integer linear programming (MILP) method. The specific solution steps are as follows: (1) Model initialization and parameter setting: Set the resource load requirements, initial resource allocation plan and disaster scenario prediction parameters of each node in the community. Initialize the particle swarm optimization related parameters and set the particle swarm size. , maximum number of iterations , speed updates inertia weight , individual learning factor , group learning factor , convergence judgment threshold The initial iteration number is set to Each particle encodes a set of pre-disaster resource deployment decision plans, including shared energy storage power and capacity, parking lot size, number of charging piles, and pre-deployed nodes of mobile energy storage devices.

[0126] (2) Disaster scenario prediction and generation: Based on typhoon warnings, historical disaster statistics, and the geographic location of community nodes and the traffic network structure, the system predicts possible distribution network failures, road disruptions, and traffic delays during disasters. Disaster scenario samples are randomly generated, including the power load damage and traffic damage at each node. The generated scenario data is then passed as input to the lower-level post-disaster dynamic resource scheduling optimization module.

[0127] (3) Pre-disaster resource allocation optimization: In each generation of particles, the corresponding resource pre-configuration plan is generated based on the pre-disaster deployment decision encoded by the current particle. The basic feasibility of the deployment plan is quickly evaluated. If there is an obviously infeasible situation, the particle is directly eliminated. After screening, the lower-level post-disaster dynamic rolling optimization scheduling module is called for the feasible plan to simulate the dynamic resource scheduling process under the corresponding disaster scenario, calculate the load recovery capability and scheduling cost corresponding to each particle, and evaluate the resilience level of this plan. The goal of pre-disaster optimization is to comprehensively minimize the total investment cost and the indirect losses caused by insufficient post-disaster load recovery. In the particle swarm optimization process, the particle swarm position is continuously iterated through the speed update and position adjustment of the particle swarm. After each iteration is completed, the execution .

[0128] (4) Solution for post-disaster dynamic resource scheduling: After a disaster occurs, a rolling optimization mechanism is introduced based on the resource deployment plan of particle transmission and the actual post-disaster fault situation. The system recovery process is divided into multiple rolling optimization cycles, and prediction and optimization are performed based on the current system status in each cycle. Post-disaster scheduling aims to minimize the load loss, resource scheduling cost and total recovery time, and a mixed integer linear programming (MILP) model is constructed. The constraints considered include distribution network topology and flow constraints, energy evolution constraints of mobile energy storage equipment and electric vehicle clusters, and actual accessibility restrictions of the traffic network. The scheduling optimization model is solved by the commercial solver Gurobi, and the power supply recovery trajectory, resource movement path and time schedule of each node and time period are output. Each round of optimization only implements the scheduling decision of the execution period. After completion, the resource remaining status and traffic accessibility data are updated, and the next round of prediction and optimization is entered until all post-disaster recovery tasks are completed. A soft penalty mechanism is introduced in the post-disaster scheduling process. For hard constraints that cannot be met due to disasters, slack variables are allowed to be introduced to improve the feasibility and convergence speed of the post-disaster scheduling problem.

[0129] (5) Particle fitness update and population evolution: Based on the post-disaster scheduling solution, extract the power supply trajectory of each node in each time period , calculate the toughness index corresponding to the particle ,Will Feedback is fed back into the pre-disaster optimization objective function, and the fitness value of the particle is calculated based on the total investment cost of resource deployment. According to the PSO standard mechanism, the historical optimal position and global optimal position of the particle are updated, and the speed and position are updated. If the fitness improvement of several consecutive generations of particles is less than the set convergence threshold , the early stopping mechanism is triggered to terminate the iteration early; otherwise, the next generation of evolution will continue, forming an iterative cycle of repeated linkage before and after the disaster until the maximum number of generations or the early stopping condition is met.

[0130] (6) Output of the optimal deployment plan: When the particle swarm optimization iteration terminates, the pre-disaster resource deployment plan and post-disaster dynamic recovery scheduling trajectory corresponding to the particle with the best fitness in the current particle swarm are output, including the configuration of shared energy storage stations, parking lot and charging pile layout, mobile energy storage deployment decisions, as well as post-disaster MES, EV driving scheduling and charging and discharging arrangements, providing decision-making support for improving community resilience and optimizing disaster response.

[0131] This example proposes a pre-disaster collaborative deployment model for generalized shared energy storage through modeling the coordinated deployment of multiple resources before a disaster. By integrating physical shared energy storage, mobile energy storage, and electric vehicle resources, the spatial distribution and capacity planning of these multiple resources is accomplished under the guidance of meteorological warnings. Furthermore, by optimizing parking lot location and scale before a disaster, the system possesses a highly potential distributed recovery interface after a disaster. This strategy significantly shortens the post-disaster startup time and effectively improves the pre-deployment capability of emergency resources and the resilience of the power grid; the virtual-real integrated shared energy storage dispatching platform, in response to the "isolated point" dispatching limitations of traditional energy storage systems, builds a "physical + virtual" integrated shared energy storage platform, abstracts mobile energy storage equipment and electric vehicle clusters into virtual energy storage nodes, realizes unified resource dispatching on the same platform, breaks through the dispatching barriers between physical energy storage and flexible resources, and improves the system dispatching flexibility and resource utilization efficiency; embeds a disaster-aware path planning mechanism, introduces a traffic integration coefficient based on disaster feedback, dynamically perceives the post-disaster road accessibility and traffic efficiency, and ensures that mobile energy storage equipment can reach the demand node in the shortest actual time. This mechanism significantly improves the execution efficiency and accessibility of post-disaster resource dispatching It provides a guarantee for dynamic scheduling under the condition of continuously changing disaster conditions; electric vehicle user behavior modeling and incentive mechanism, constructs a two-stage EV response model during and after the disaster, and guides EVs to enter suitable parking lots for risk avoidance and participate in post-disaster discharge through the mechanism of guidance in the evacuation stage and incentive in the response stage. It establishes an incentive mechanism to mobilize user participation, reasonably models EV users, and provides stable distributed energy storage support for post-disaster recovery; cost and resource utilization efficiency optimization, systematically considers multiple costs, introduces multiple types of resource economic factors in the objective function setting, achieves optimal energy flow through reasonable scheduling strategies, significantly reduces redundant deployment and energy waste, and coordinates the scheduling of multiple types of energy storage to avoid duplicate construction investment, while improving resource utilization as a whole and taking into account disaster recovery capabilities and economic rationality.

[0132] This embodiment has constructed a community post-disaster recovery strategy system covering the trinity of "pre-disaster resource coordination deployment - post-disaster scheduling path optimization - multi-shared energy storage integration". This strategy not only breaks down the scheduling barriers between traditional isolated energy storage and temporary power sources, but also significantly improves the autonomous recovery capability and controllable scheduling level of the community power grid in extreme disaster scenarios. By constructing physical and virtual shared energy storage systems and introducing a path accessibility optimization mechanism, resource utilization efficiency and scheduling economy are improved, and a theoretical basis and practical guidance are provided for urban communities to cope with natural disasters such as typhoons, significantly improving the community's self-sustaining ability and resilience recovery in disasters.

[0133] Each embodiment of the present invention is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiment. The device embodiment described above is merely exemplary. The modules described as separate components may or may not be physically separated. When implementing the scheme of the present invention, the functions of each module can be implemented in the same one or more software and / or hardware. It is also possible to select some or all of the modules according to actual needs to achieve the purpose of the scheme of this embodiment.

[0134] Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A method for pre-disaster dispatching of a community multi-shared energy storage system during a typhoon disaster, wherein the community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage dispatching center, wherein the virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster, and the physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage dispatching center, characterized in that: The pre-disaster dispatch method includes the following steps: Step 1: Establish a pre-disaster objective function. This function aims to optimize the layout of physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-faceted shared energy storage system before a typhoon occurs, minimizing acquisition and deployment costs while ensuring system operational resilience. Step 2: Establish a constraint function for pre-disaster optimization, which includes physical energy storage constraints, distribution network topology constraints, and power flow constraints; Step 3: Optimize the pre-disaster objective function.

2. The method for pre-disaster scheduling of a community multi-shared energy storage system under typhoon disasters according to claim 1, characterized in that: The pre-disaster objective function is as follows: in, represents the cost of the physical energy storage station, represents the cost of parking lots and charging piles, represents the deployment cost of mobile energy storage, Represents the toughness index, represents the trade-off coefficient.

3. The method for pre-disaster scheduling of a community multi-shared energy storage system under typhoon disasters according to claim 2, characterized in that: The pre-disaster objective function is optimized. The energy storage station power and maximum capacity are used to optimize the energy storage station capacity configuration. The parking lot capacity and the number of charging piles are used to optimize the parking lot size. The connection relationship of mobile energy storage devices is used to optimize the pre-deployment of mobile energy storage. The cost of the physical energy storage station includes investment cost and operation and maintenance cost. The cost expression of the physical energy storage station is: in, represents the cost of the physical energy storage station, represents the discount rate coefficient, Indicates the service life of the physical energy storage station in years. 、 Represent the power and maximum capacity of the energy storage station respectively, 、 denote the unit power and capacity investment costs respectively, Indicates the unit power operation and maintenance cost, in RMB / kW per year; The cost of a parking lot and its charging piles includes the land cost and the construction cost of the charging piles. The cost expression of a parking lot and its charging piles is: in, represents the cost of parking lots and charging piles, represents the unit land cost, represents the parking lot capacity, i.e. the number of electric vehicles that can be parked. Indicates the comprehensive area occupied by a single parking space. represents the construction cost of a single charging pile, Indicates the number of charging piles; The mobile energy storage deployment cost expression is: in, represents the deployment cost of mobile energy storage, is a 0-1 variable, indicating whether the mobile energy storage device k is connected to the node n. represents the unit mobile energy storage deployment cost; The toughness index expression is: in, Represents the toughness index, represents the total critical load demand within the scheduling period, represents the probability of disaster occurrence, represents the importance weight of the load node, represents the power supply required by the key load node n at time t, It indicates the actual power obtained by the key load node n at time t.

4. The method for pre-disaster scheduling of a community multi-shared energy storage system under a typhoon disaster according to claim 1, characterized in that: The constraint function includes physical energy storage constraint, distribution network topology constraint and power flow constraint. The physical energy storage constraint expression is: in, Indicates the unit energy storage system capacity of the shared energy storage station, in kWh. represents the actual storage capacity of the energy storage station at time t, / Respectively represent the charging / discharging power of the energy storage station at time t, Indicates the rated power of the energy storage station, / Represent the charge / discharge efficiency respectively; The distribution network topology constraint expression is: in, It is a 0-1 variable that describes the line connection status. Indicates the number of distribution network nodes, represents the power flow variable, represents all adjacent nodes of node i, Represents a 0-1 variable, indicating whether node i is a power node or connected to a power node. represents all downstream nodes of node i, represents all upstream nodes of node i, represents the load demand of node i; The distribution network power flow constraint expression is: in, Indicates the active power provided by the distribution network, Indicates the active power of mobile energy storage, represents the active power transmitted from node j to i, represents the load demand, represents the active power transmitted from node i to node j, / is the line loss, indicating that the power from the main grid and distributed energy must meet the load and flow output of each node. represents the square of the voltage modulus, Indicates the circuit breaker status 0-1 variable, represents the slack variable.

5. The method for pre-disaster scheduling of a community multi-shared energy storage system under a typhoon disaster according to any one of claims 1 to 4, characterized in that: Optimizing the pre-disaster objective function includes the following steps: The particle swarm optimization algorithm is used to optimize the pre-disaster objective function; The current particle pre-disaster deployment plan is generated based on the current particle code. The resilience index of the plan is calculated based on the load recovery capability and scheduling cost corresponding to each particle. While ensuring the resilience of system operation, the objective function is to minimize the total investment cost and the indirect losses caused by insufficient load recovery after the disaster. The particle swarm position is iteratively optimized through particle swarm speed update and position adjustment until the optimization target is reached and the iteration ends.

6. A post-disaster dispatch method for a community multi-shared energy storage system during a typhoon disaster, wherein the community multi-shared energy storage system includes a physical energy storage station, a virtual energy storage station, and an energy storage dispatch center, wherein the virtual energy storage station includes a mobile energy storage device and an electric vehicle cluster, and the physical energy storage station, the mobile energy storage device, and the electric vehicle cluster are all communicatively connected to the energy storage dispatch center, characterized in that: The post-disaster deployment method includes the following steps: Step 1: Establish a post-disaster objective function. The post-disaster objective function is to quickly restore the physical energy storage stations, mobile energy storage devices, and electric vehicle cluster resources within the community's multi-shared energy storage system after a typhoon, minimizing recovery time, energy loss, and scheduling costs. A post-disaster scheduling model is established based on the objective function. Step 2: Establish a constraint function for post-disaster optimization, which includes load restoration priority constraints, spatiotemporal dynamic scheduling constraints for mobile energy storage, constraints on individual and cluster electric vehicles, and constraints on parking lots and charging piles. Step 3: Solve the post-disaster dispatch model.

7. The post-disaster dispatching method for a community multi-shared energy storage system under a typhoon disaster according to claim 6, characterized in that: The post-disaster objective function is as follows: in, represents the load importance weight, represents the power grid restoration time in region i, Energy loss refers to the loss of energy supply due to power outages, that is, unmet electricity demand. represents the dispatch cost, including the dispatch cost of EV and mobile energy storage; The energy loss expression is: in, represents energy loss, represents the power demand of load node n at time t, represents the actual power supply provided by the energy storage device to the load node n at time t; The scheduling cost expression is: in, represents the scheduling cost, 、 Indicates the total distance traveled by mobile energy storage and electric vehicles, 、 are the unit dispatching costs of mobile energy storage and electric vehicles, Represents the cost of incentive subsidies given to EV owners.

8. The post-disaster dispatching method for a community multi-shared energy storage system under a typhoon disaster according to claim 6, characterized in that: The constraint functions include load recovery priority constraints, mobile energy storage spatiotemporal dynamic scheduling constraints, electric vehicle individual and cluster constraints, and parking lot and charging pile constraints. The load recovery priority constraint expression is: in, represents the load recovery degree variable, / Indicates the maximum load of the node under normal operation; The spatiotemporal dynamic scheduling constraints of mobile energy storage include node scheduling constraints and operating power constraints. The node scheduling constraint expression is: in, Indicates that the mobile energy storage moves from i to j at time t, which is a 0-1 variable, indicating that the scheduling path of the mobile energy storage in different time periods should maintain consistency; The operating power constraint expression is: in, represents the amount of electricity of the mobile energy storage device e at time t, Indicates the self-discharge rate of mobile energy storage, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the mobile energy storage is in the charging or discharging state; The constraints on individual and cluster electric vehicles include the individual EV power constraint and the upper and lower bound constraints of the EV cluster. The expression of the individual EV power constraint is: in, represents the amount of electricity in the electric vehicle at time t, represents the self-discharge rate of electric vehicles, / represents the charge and discharge power at time t, / Indicates the charge and discharge efficiency, / It is a 0-1 variable, indicating whether the electric vehicle is in the charging or discharging state. is a 0-1 variable, which determines whether node i is a parking lot node; The upper and lower boundary constraint expressions of the EV cluster are: in, represents the total number of EV individuals in the EV cluster; The constraint expressions of parking lots and charging piles are: in, is a 0-1 variable, indicating whether the electric car e is in the parking lot node n. represents the maximum capacity of parking lot node n, represents the set of all parking lot nodes, is the number of charging piles in parking lot node n.

9. The post-disaster dispatching method for a community multi-shared energy storage system under a typhoon disaster according to any one of claims 6 to 8, characterized in that: Solving the post-disaster dispatch model includes the following steps: The system process is divided into multiple rolling optimization cycles. Prediction and optimization are performed based on the current system status in each cycle. The rolling optimization cycle includes a prediction period and an execution period. With the objective function of minimizing load loss, recovery time, energy loss and scheduling cost, a post-disaster scheduling model is established based on the constraint function of post-disaster optimization. The post-disaster scheduling model is solved by the Gurobi solver to output the power supply restoration trajectory, resource movement path and time schedule for each node and time period. The optimization of each cycle only implements the scheduling of the execution period. After the scheduling is completed, the model is updated and the prediction and optimization of the next cycle are started until all post-disaster recovery tasks are completed.

10. A method for pre-disaster, post-disaster and overall dispatching of a community multi-shared energy storage system during a typhoon disaster, characterized in that: The steps include: Based on the post-disaster scheduling solution results described in claim 9, the power supply restoration trajectory of each node in each time period is extracted, the corresponding resilience index of the particle is calculated, and the resilience index is fed back to the pre-disaster objective function described in claim 6. Combined with the total investment cost, the fitness value of the particle is calculated, and the particle swarm optimization algorithm is used to update the historical optimal position and the global optimal position of the particle, and perform speed and position updates; if the fitness value improvement of several consecutive generations of particles is less than the set convergence threshold, the early stopping mechanism is triggered to terminate the iteration early, otherwise the next generation optimization is continued to be executed, forming an iterative cycle of repeated linkage before and after the disaster until the maximum number of generations or the early stopping condition is met; When the particle swarm optimization iteration terminates, the optimal pre-disaster and post-disaster deployment plan corresponding to the particle with the best fitness value in the current particle swarm is output.

Images