Power distribution network toughness improving method based on energy internet under mobile energy storage space-time optimization scheduling

Abstract

The invention discloses a power distribution network toughness improvement method based on an energy internet under mobile energy storage space-time optimization scheduling, and the method comprises the steps: building a double-layer scheduling model which comprises an upper-layer model and a lower-layer model; the upper layer model adopts a distributed robust optimization method, considers the uncertainty of fault nodes in extreme weather and the influence of traffic network passing time, and optimizes the configuration number and position scheme of mobile energy storage; the lower layer model constructs a multi-source cooperative recovery mixed integer quadratic cone programming model based on dynamic scheduling and time sequence output characteristics of mobile energy storage, an electric vehicle and a diesel generator, and solves an optimal load reduction scheme and an important load power recovery state; constructing a fault probability model based on the double-layer scheduling model; the method comprises the following steps: performing simulation analysis by improving an IEEE-33 node power distribution network, and verifying the effectiveness of the method; according to the method, the key problems of dynamic adaptive scheduling and multi-temporal-spatial-scale collaborative optimization are solved, the potential of mobile energy storage and multi-source collaboration is released, and a high-toughness power distribution network system can be constructed.

Description

Technical Field

[0001] This invention belongs to the field of distribution network dispatching technology, and in particular relates to a method for improving the resilience of distribution networks under the spatiotemporal optimization dispatching of the energy internet based on mobile energy storage. Background Technology

[0002] With the increasing frequency of extreme weather and natural disasters, power distribution networks face the severe challenge of escalating power outage risks, making the enhancement of grid resilience a research hotspot in the energy sector. In recent years, mobile energy storage systems, such as battery storage vehicles and hydrogen fuel cell vehicles, have been increasingly applied to power distribution network fault recovery scenarios due to their flexible deployment and rapid response capabilities. Statistics show that the global mobile energy storage market exceeded $50 billion in 2024, with significant growth in its application in power emergency response and disaster recovery, especially in areas prone to typhoons and wildfires. Meanwhile, multi-source collaborative recovery technology has significantly improved fault recovery efficiency by integrating distributed power sources, fixed energy storage, and load resources. However, existing technologies still face multiple bottlenecks: traditional two-layer scheduling models often employ static optimization strategies, making it difficult to adapt to dynamic changes in fault scenarios. Furthermore, the scheduling of mobile energy storage relies on manual experience or single-objective optimization, lacking comprehensive consideration of multi-dimensional constraints such as traffic conditions and energy storage capacity decay. Regarding collaborative mechanisms, the timing coordination between fixed and mobile energy storage charging and discharging is insufficient, often resulting in energy backflow or localized overload. Communication latency and data silos further constrain the effectiveness of multi-source collaboration. Although some studies have attempted to introduce reinforcement learning algorithms to optimize scheduling models, their training data is mostly based on idealized assumptions, resulting in limited generalization ability in real-world complex environments. Therefore, it is necessary to propose a method for improving the resilience of the distribution network of the energy internet based on spatiotemporal optimization scheduling of mobile energy storage to address these issues. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a method for improving the resilience of distribution networks based on the spatiotemporal optimization scheduling of energy internet, so as to break through the key technologies of dynamic adaptive scheduling and multi-spatiotemporal scale collaborative optimization, so as to fully release the potential of mobile energy storage and multi-source collaboration and build a highly resilient distribution network system.

[0004] To achieve the above-mentioned technical effects, the technical solution adopted by the present invention is as follows: A method for improving the resilience of the distribution network in the energy internet based on spatiotemporal optimization scheduling of mobile energy storage includes the following steps: Establish a two-layer scheduling model, including an upper-layer model and a lower-layer model; The upper-level model adopts the sub-Bruker optimization method, which takes into account the uncertainty of fault nodes under extreme weather conditions and the impact of traffic network travel time, and optimizes the configuration quantity and location scheme of mobile energy storage. The lower-level model is based on the dynamic scheduling and time-series output characteristics of mobile energy storage, electric vehicles and diesel generators. It constructs a mixed integer quadratic cone programming model for multi-source collaborative recovery to solve the optimal load shedding scheme and the power recovery state of important loads. A failure probability model is constructed based on a two-level scheduling model; The effectiveness of the proposed method was verified through simulation analysis of an improved IEEE-33 node distribution network.

[0005] Preferably, in step S1, the objective function of the upper-level model is to minimize the mobile energy storage configuration cost, system risk cost, and load reduction cost, and its mathematical expression is: ; In the formula, For the set of load nodes; p A fuzzy set of the comprehensive norm of node faults under typhoon weather; Mathematical expectation of fuzzy sets with comprehensive norm; Defined as a 0-1 variable to characterize mobile energy storage devices and nodes. i The connection status between them.

[0006] Preferably, the constraints of the upper-level model include dynamic adjustment constraints of mobile energy storage, distribution network radiation topology constraints, load reduction constraints, distributed power output constraints, distribution network operation constraints, and spatiotemporal dynamic scheduling constraints of mobile energy storage.

[0007] Preferably, the dynamic adjustment constraint for mobile energy storage is: ; In the formula, during the energy storage deployment phase, the upper limit of the number of mobile energy storage units is determined by... express; Based on complex network theory, nodes and lines in a power system can be represented as a connected graph. ,Right now .in It is the set of power nodes in a power grid, where different nodes in the power grid represent different functional attributes, such as power generation, substation, and load. This represents the edge set of a power grid, indicating the lines in the power system. It also uses an adjacency matrix. The topological relationships between different nodes are represented by the following formula.

[0008]

[0009] In the formula: a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elementsa ij =1; otherwise a ij Then it is 0.

[0010] The topological constraints of the distribution network are: ; ; ; ; In the formula, The set of all branches in the entire power grid; a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; This represents the total number of nodes within a power grid; for a specific node... i The child and parent nodes connected to it are respectively categorized into sets. and ; Describes the virtual power values ​​transmitted through each branch; This represents a 0-1 variable, indicating the virtual source node added to the network optimization, thus effectively taking into account the number of isolated nodes generated during network reconstruction into the optimization model; M represents the virtual power output by each node; during the calculation of this constraint, M is used as a maximum value. The load reduction constraint is: ; ; In the formula, the power reduction of the active power of the power grid load is: The power reduction of reactive power of the power grid load is The maximum active power reduction of the power grid load is The maximum reactive power reduction of the power grid load is ; The output constraints of distributed power sources are: ; ; ; In the formula: the node sets M, E, D, and P represent mobile energy storage systems, electric vehicle charging stations, diesel generator sets, and photovoltaic power generation systems, respectively; Used to characterize different types of distributed power sources and specific nodes i The connection status between them; and These represent the active and reactive power outputs of the distributed power source, respectively; and and This specifies the maximum limits for the active and reactive power output of these power sources; among them, and The upper and lower limits of the power factor for distributed power sources are defined.

[0011] Preferably, the operating constraints of the distribution network are: ; ; ; ; ; ; ; In the formula: and These are the active and reactive power of the branch circuit, respectively. a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; and These are the branch resistance and reactance, respectively; and These are the squares of the node voltage and branch current, respectively; the maximum value of the square of the node voltage is... The minimum value of the square of the node voltage is ; This is the maximum value of the square of the branch current.

[0012] When using the Big M method, it is crucial to pay close attention to the value of M. A value that is too small may result in an infeasible solution, while a value that is too large may cause constraints to fail due to computer precision limitations. For example, when M is extremely large (10... 20 ), then the corresponding If it equals 10 -18 ( If the value is close to 1), the constraint may also be relaxed.

[0013] The nonlinear constraint among voltage, current, and power is addressed using a second-order cone relaxation technique, transforming the nonlinear constraint into a second-order cone form. The constraint conditions are as follows: ; In the formula, and These are the active and reactive power of the branch circuit, respectively. This is the square term of the branch current; This is the maximum value of the square of the branch current.

[0014] Preferably, in the spatiotemporal dynamic scheduling constraints of mobile energy storage, the mobile energy storage unit is considered. i From node j To node k The transportation time, and in and During the node installation time, mobile energy storage units i At the node j With nodes k The travel time between them is The equivalent travel distance is and the actual driving speed is The spatiotemporal dynamic scheduling constraints of mobile energy storage are expressed as follows: ; ; ; In the formula, For connecting nodes j and nodes k The length of the road; This is the theoretical maximum speed assumed when there is no traffic flow; while c It represents specific disaster conditions.

[0015] Preferably, the objective function of the lower-level model is to minimize the load reduction power during the fault period, and its mathematical expression is: ; In the formula: A set of periods of failure; Let y be the set of system operation decision variables under the scheduling scheme; y is the corresponding scheduling scheme. Let t be the active power reduction of the load at node i at time t.

[0016] Preferably, the constraints of the lower-level model include the charging and discharging power and state of charge constraints of mobile energy storage, the constraints of the electric vehicle charging decision model, the output constraints of distributed power sources, and the operation constraints of the distribution network.

[0017] Preferably, the failure probability model is as follows: Two assumptions are established: The average wind speed at the same location in different directions follows a single type of extreme value distribution; The model parameters for different directions at the same location are independent of each other, and the wind speed data samples for each direction are used to independently estimate the model parameters. The GEV distribution is adopted, and the distribution function is as follows: ; In the formula, , , These are the position parameter, scale parameter, and shape parameter, respectively. The scale parameter is greater than zero. v represents the extreme wind speed; overhead transmission lines have inherent limitations in their ability to withstand external forces. Based on the theory of metal deformation, the line fault probability model is as follows: ; In the formula, The probability of transmission line failure under strong winds; v The actual wind speed that the transmission line experiences; V This is the design value for the maximum wind speed that the transmission line can withstand.

[0018] Preferably, the method uses an improved IEEE-33 node distribution network in the simulation analysis, and the specific parameters include: Rated voltage, upper voltage limit, lower voltage limit; Installation and configuration time for mobile energy storage; Ideal vehicle speed for mobile energy storage under zero traffic conditions.

[0019] The beneficial effects of this invention are as follows: 1. To address the issue of improving the resilience of distribution networks in the context of power grid and transportation network coupling, this invention proposes a two-layer scheduling model of mobile energy storage and multi-source collaborative recovery to improve the resilience of distribution networks. By establishing a two-layer optimization model and conducting case studies, the effectiveness of the proposed method in improving the resilience of distribution networks is verified.

[0020] 2. To address the uncertainty of fault nodes under extreme weather conditions, this invention proposes a two-layer scheduling model combining mobile energy storage and multi-source collaborative recovery. This model enhances the grid resilience in the face of extreme weather events. Solving the model using the C&CG algorithm yields a mobile energy storage scheduling strategy under extreme weather conditions. This strategy not only accelerates the deployment of mobile energy storage devices for emergency load restoration but also minimizes necessary load reductions, improving the overall economic efficiency of the power system and demonstrating the effectiveness of the model in enhancing the power system's ability to cope with extreme events.

[0021] 3. This invention improves the accuracy and robustness of mobile energy storage scheduling under extreme weather conditions by simulating the uncertainty of extreme weather and using the distributed robust method. In contrast, traditional robust optimization only considers the case with the highest branch failure rate in the worst scenario, ignoring valuable probabilistic information. Furthermore, the extreme scenarios used are often unlikely to occur in reality, so the proposed solutions are more conservative but require higher load reduction costs. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the two-layer model structure of mobile energy storage and multi-source collaborative recovery in this invention; Figure 2 This is a schematic diagram of the "power grid-transportation network" coupling network of the energy internet in an embodiment of the present invention; Figure 3 This is a functional curve diagram of the energy internet distribution network system in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the dynamic scheduling of mobile energy storage in a transportation network according to an embodiment of the present invention; Figure 5 This is a schematic diagram of a modified IEEE-33 node distribution network system in an embodiment of the present invention; Figure 6 This is a schematic diagram of the dynamic scheduling results of mobile energy storage in an embodiment of the present invention; Figure 7 This is a schematic diagram illustrating the relationship between the state of charge and access location of mobile energy storage in an embodiment of the present invention; Figure 8 This is a schematic diagram of the charging and discharging power of electric vehicle charging piles at different time periods in an embodiment of the present invention; Figure 9 This is a schematic diagram of the generator's active power output during the fault phase in an embodiment of the present invention; Figure 10 This is a schematic diagram of load power and recovery ratio for each time period in an embodiment of the present invention; Figure 11 This is a comparison chart of the changes in active power at different time periods in Method 1 and Method 2 in this embodiment of the invention; Figure 12This is a comparison chart of the solution time of Method 3 and Method 2 in the MATLAB simulation platform in the embodiments of the present invention. Detailed Implementation

[0023] Example 1: like Figure 1 As shown, the method for improving the resilience of the distribution network of the energy internet based on the spatiotemporal optimization scheduling of mobile energy storage includes the following steps: Establish a two-layer scheduling model, including an upper-layer model and a lower-layer model; The upper-level model adopts the sub-Bruker optimization method, which takes into account the uncertainty of fault nodes under extreme weather conditions and the impact of traffic network travel time, and optimizes the configuration quantity and location scheme of mobile energy storage. The lower-level model is based on the dynamic scheduling and time-series output characteristics of mobile energy storage, electric vehicles and diesel generators. It constructs a mixed integer quadratic cone programming model for multi-source collaborative recovery to solve the optimal load shedding scheme and the power recovery state of important loads. A failure probability model is constructed based on a two-level scheduling model; The effectiveness of the proposed method was verified through simulation analysis of an improved IEEE-33 node distribution network.

[0024] Preferably, in step S1, the objective function of the upper-level model is to minimize the mobile energy storage configuration cost, system risk cost, and load reduction cost, and its mathematical expression is: ; In the formula, For the set of load nodes; p A fuzzy set of the comprehensive norm of node faults under typhoon weather; Mathematical expectation of fuzzy sets with comprehensive norm; Defined as a 0-1 variable to characterize mobile energy storage devices and nodes. i The connection status between them.

[0025] Preferably, the constraints of the upper-level model include dynamic adjustment constraints of mobile energy storage, distribution network radiation topology constraints, load reduction constraints, distributed power output constraints, distribution network operation constraints, and spatiotemporal dynamic scheduling constraints of mobile energy storage.

[0026] Preferably, the dynamic adjustment constraint for mobile energy storage is: ; In the formula, during the energy storage deployment phase, the upper limit of the number of mobile energy storage units is determined by... express; The topological constraints of the distribution network are: ; ; ; ; In the formula, The set of all branches in the entire power grid; This represents the total number of nodes within a power grid; for a specific node... i The child and parent nodes connected to it are respectively categorized into sets. and ; Describes the virtual power values ​​transmitted through each branch; M represents the virtual power output by each node; during the calculation of this constraint, M is used as a maximum value. The load reduction constraint is: ; ; In the formula, the power reduction of the active power of the power grid load is: The power reduction of reactive power of the power grid load is The maximum active power reduction of the power grid load is The maximum reactive power reduction of the power grid load is ; The output constraints of distributed power sources are: ; ; ; In the formula: the node sets M, E, D, and P represent mobile energy storage systems, electric vehicle charging stations, diesel generator sets, and photovoltaic power generation systems, respectively; Used to characterize different types of distributed power sources and specific nodes i The connection status between them; and These represent the active and reactive power outputs of the distributed power source, respectively; and and This specifies the maximum limits for the active and reactive power output of these power sources; among them, and The upper and lower limits of the power factor for distributed power sources are defined.

[0027] Preferably, the operating constraints of the distribution network are: ; ; ; ; ; ; ; In the formula: and These are the active and reactive power of the branch circuit, respectively. a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; and These are the branch resistance and reactance, respectively; and These are the squares of the node voltage and branch current, respectively; the maximum value of the square of the node voltage is... The minimum value of the square of the node voltage is ; This is the maximum value of the square of the branch current.

[0028] When using the Big M method, it is crucial to pay close attention to the value of M. A value that is too small may result in an infeasible solution, while a value that is too large may cause constraints to fail due to computer precision limitations. For example, when M is extremely large (10... 20 ), then the corresponding If it equals 10 -18 ( If the value is close to 1), the constraint may also be relaxed.

[0029] The nonlinear constraint among voltage, current, and power is addressed using a second-order cone relaxation technique, transforming the nonlinear constraint into a second-order cone form. The constraint conditions are as follows: ; In the formula, and These are the active and reactive power of the branch circuit, respectively. This is the square term of the branch current; This is the maximum value of the square of the branch current.

[0030] Preferably, in the spatiotemporal dynamic scheduling constraints of mobile energy storage, the mobile energy storage unit is considered. i From node j To node k The transportation time, and in and During the node installation time, mobile energy storage units i At the node jWith nodes k The travel time between them is The equivalent travel distance is and the actual driving speed is The spatiotemporal dynamic scheduling constraints of mobile energy storage are expressed as follows: ; ; ; In the formula, For connecting nodes j and nodes k The length of the road; This is the theoretical maximum speed assumed when there is no traffic flow; while c It represents specific disaster conditions.

[0031] Preferably, the objective function of the lower-level model is to minimize the load reduction power during the fault period, and its mathematical expression is: ; In the formula: A set of periods of failure; Let y be the set of system operation decision variables under the scheduling scheme; y is the corresponding scheduling scheme. Let t be the active power reduction of the load at node i at time t.

[0032] Preferably, the constraints of the lower-level model include the charging and discharging power and state of charge constraints of mobile energy storage, the constraints of the electric vehicle charging decision model, the output constraints of distributed power sources, and the operation constraints of the distribution network.

[0033] Preferably, the failure probability model is as follows: Two assumptions are established: The average wind speed at the same location in different directions follows a single type of extreme value distribution; The model parameters for different directions at the same location are independent of each other, and the wind speed data samples for each direction are used to independently estimate the model parameters. The GEV distribution is adopted, and the distribution function is as follows: ; In the formula, , , These are the position parameter, scale parameter, and shape parameter, respectively. The scale parameter is greater than zero. v represents the extreme wind speed; overhead transmission lines have inherent limitations in their ability to withstand external forces. Based on the theory of metal deformation, the line fault probability model is as follows: ; In the formula, The probability of transmission line failure under strong winds; v The actual wind speed that the transmission line experiences; V This is the design value for the maximum wind speed that the transmission line can withstand.

[0034] Preferably, the method uses an improved IEEE-33 node distribution network in the simulation analysis, and the specific parameters include: Rated voltage, upper voltage limit, lower voltage limit; Installation and configuration time for mobile energy storage; Ideal vehicle speed for mobile energy storage under zero traffic conditions.

[0035] Example 2: This embodiment provides a strategy for improving the resilience of distribution networks using a two-layer scheduling model combining mobile energy storage and multi-source collaborative recovery. It includes the following steps: 1. Establish a two-level scheduling model: Step 1: Construct a two-layer optimization model for improving the resilience of the distribution network in the energy internet; Step 2: Consider the upper-level model of mobile energy storage dispatch in the energy internet distribution network under distributed bar optimization; Step 3: Consider the post-disaster multi-source collaborative recovery energy internet distribution network lower-level dispatch optimization model; Step 4: Conduct simulation analysis by improving the IEEE-33 node distribution network to make energy storage layout decisions.

[0036] 2. Construct a multi-source collaborative model: (1) Considering the inclusion of distributed energy sources such as portable energy storage units, photovoltaic devices, electric vehicle charging facilities (EVS), and diesel engines in the distribution network of the energy internet, the coupling model of the integrated framework of the power grid and the transportation network is as follows: Figure 2 As shown.

[0037] (2) A two-layer model resilience enhancement strategy is proposed in the model to enhance the resilience of the distribution network through prevention and recovery measures. Specifically, it involves the preparation before a disaster occurs and the recovery process after the disaster. This process is reflected in the function curve. Figure 3The t0 to t1 and t3 to t4 stages are described in detail. The upper layer of the model considers the potential failure risk of load nodes under extreme weather conditions and, given the unknown timing and duration of the disaster, effectively deploys and schedules mobile energy storage units to ensure their rapid intervention after a disaster, supporting load recovery. The lower layer of the model considers a comprehensive recovery strategy encompassing multiple energy resources in the post-disaster phase. By rationally scheduling various resources, including mobile energy storage and electric vehicles, it achieves efficient energy allocation, prioritizes the power supply needs of critical loads, and further enhances the disaster response capabilities of the distribution network. Figure 1 The described fault recovery phase.

[0038] Example 3: This embodiment provides a detailed explanation of the model's working principle: 1. Two-level scheduling model: When introducing the upper-level model of mobile energy storage scheduling in the distribution network of the Energy Internet under the distributed bar optimization, the objective function, dynamic adjustment constraints of mobile energy storage, distribution network radiation topology constraints, load shedding constraints, distributed power generation output constraints, distribution network operation constraints, and spatiotemporal dynamic scheduling constraints of mobile energy storage are obtained respectively; including the following sub-steps: 1.1 Objective Function: During the distribution network failure phase under extreme weather conditions, the uncertainty of load failure nodes within the distribution network is assessed. In cases where a faulty branch causes the distribution network to disconnect from the main grid, the duration of the failure can be predicted based on the severity of the disaster and the amount of available repair resources. The upper-level model of its mobile energy storage dispatch model uses the configuration cost of mobile energy storage as the objective function to optimize the configuration of the mobile energy storage system and ensure that system risk costs and load reduction costs are minimized after a failure.

[0039] 1.2 Constraints: 1.2.1 Dynamic Adjustment Constraints for Mobile Energy Storage. Considering the limited availability of mobile energy storage resources before extreme disasters; furthermore, the equipment parameters of all mobile energy storage units must remain consistent, and at most one such device can be connected to each power node.

[0040] 1.2.2 Distribution Network Radiation Topology Constraints. This provides a more flexible and accurate method for simulating the dynamic response of distribution networks during the post-disaster reconstruction phase.

[0041] 1.2.3 Load reduction constraints are formulated under the premise that the load power factor remains constant.

[0042] 1.2.4 Output constraints of distributed power sources.

[0043] 1.2.5 Distribution network operation constraints: The Big-M method was used to relax the voltage equations to adapt to the dynamic changes in this topology.

[0044] 1.2.6 Spatiotemporal Dynamic Scheduling Constraints of Mobile Energy Storage; In systems coupled with power grids and transportation networks, the scheduling of mobile energy storage systems depends on their charging / discharging states and their transportation status within the transportation network, exhibiting significant spatiotemporal coupling characteristics. When considering mobile energy storage units... i From node j To node k The transportation time, and in and When considering node installation time, this section constructs a spatiotemporal dynamic scheduling model for mobile energy storage. This model assumes no energy loss during transportation. A schematic diagram of the spatiotemporal dynamic scheduling is shown below. Figure 4 As shown.

[0045] Because real-time traffic conditions and unexpected events can easily affect the transportation network, the actual travel time of mobile energy storage systems on the same road segment may vary under different disaster conditions and road surface conditions. To ensure the reliability of emergency mobile energy storage devices in disaster situations, this embodiment addresses the impact of extreme disasters such as typhoons on traffic flow, using a traffic fusion coefficient to describe the spatial distribution relationship between the actual travel speed and equivalent travel distance of mobile energy storage units on a specific road segment. Among these, the mobile energy storage unit... i At the node j With nodes k The travel time between them is The equivalent travel distance is and the actual driving speed is It can be represented as: ; ; ; In the formula: For connecting nodes j and nodes k The length of the road; This is the theoretical maximum speed assumed when there is no traffic flow; while c This represents the situation under specific disaster conditions.

[0046] 2. Failure probability model: The analysis begins with establishing two assumptions: (1) The average wind speed in different directions at the same location follows a single type of extreme value distribution.

[0047] (2) The model parameters of different directions at the same location are independent of each other, and the wind speed data samples of each direction are used to independently estimate the model parameters.

[0048] This section uses the GEV distribution, whose distribution function is as follows: ; In the formula: , , These are the position parameter, scale parameter, and shape parameter, respectively. The scale parameter must be greater than zero. v represents the extreme wind speed.

[0049] Overhead transmission lines have inherent limitations in their ability to withstand external forces. Based on the theory of metal deformation, this study adopts the following line fault probability model: ; In the formula, The probability of transmission line failure under strong winds; v The actual wind speed that the transmission line experiences; V This is the design value for the maximum wind speed that the transmission line can withstand.

[0050] 3. Multi-source collaborative model: When considering the upper-level model of mobile energy storage dispatch in the energy internet distribution network under distributed bar optimization, the objective function, constraints of the charging decision model, and electric vehicle charging and discharging constraints are obtained, and the mathematical calculation model is performed. This includes the following sub-steps: 3.1 Objective Function: In the lower-level model, by setting photovoltaic output as a deterministic scenario and aiming to minimize load reduction power during a fault, a multi-source cooperative post-disaster grid recovery optimization model is constructed. The decision variables of the lower-level model include the charging and discharging power of mobile energy storage and its state of charge constraints, the charging and discharging limitations of electric vehicles, the output of distributed generation, and the load reduction power at a specific moment. Furthermore, the output of distributed generation and the operation of the distribution network must also satisfy its constraints, and the form of these constraints is consistent with that of the upper-level model. Its objective function is as follows: ; In the formula: A set of periods of failure; Let y be the set of system operation decision variables under the scheduling scheme; y is the corresponding scheduling scheme. Let t be the active power reduction of the load at node i at time t.

[0051] 3.2 Constraints: The charging and discharging power and state of charge constraints of mobile energy storage. The above spatiotemporal dynamic scheduling constraints can be expressed as: The constraints of the electric vehicle user charging decision model can be stated as: ; ; ; ; ; ; ; ; ; ; ; In the formula: This refers to the collection of mobile energy storage systems that connect to the system when a disaster occurs; t0 is the time when the fault occurs. and These are the charging and discharging flags for mobile energy storage, with a value of 1 indicating the charging and discharging status, respectively. and These are the active power for charging and discharging mobile energy storage, respectively. and These are the charging and discharging reactive power of mobile energy storage, respectively. and These are the upper limits of the active power for charging and discharging mobile energy storage, respectively. and These are the upper limits of the reactive power for charging and discharging of mobile energy storage, respectively. and These are the charging and discharging efficiencies of mobile energy storage, respectively. Let be the energy storage capacity of the mobile energy storage at time t. and These represent the upper and lower limits of the energy storage capacity for mobile energy storage, respectively.

[0052] After receiving a dispatch instruction, if mobile energy storage needs to be supplied by a node... j Transfer to node k To perform charging and discharging, the optimal route is first selected based on the degree of congestion in the traffic network under disaster conditions, in order to achieve the shortest travel time. Next destination node k Before mobile energy storage reaches the node. k Previously, i.e., time interval At that time, it is related to the node k The connection status is always 0. This means that at any given time, a mobile energy storage system is limited to connecting to a maximum of a single node.

[0053] Example 4: The specific analysis process provided in this embodiment is as follows: use Figure 5 The improved IEEE-33 node distribution network shown is simulated and analyzed. The power supply and system operating parameters are shown in Table 1 and Table 2, respectively. The topology of the transportation network is the same as that of the power grid. The road distance between adjacent electrical nodes is 2km. Considering the average speed of mobile energy storage in the transportation network, this example only considers the dispatch of two mobile energy storage units.

[0054] Assuming a typhoon occurs at 01:00, and considering the load recovery within 24 hours, the worst-case scenario probability distribution is obtained after applying Bruker bar fuzzy set constraints. Based on this worst-case probability distribution, energy storage deployment decisions are made. Mobile energy storage has sufficient time to reach the fault node, and all mobile energy storage devices and electric vehicles in charging piles are fully charged. To verify the advantages of the proposed mobile energy storage and multi-source collaborative recovery two-layer scheduling model in improving distribution network resilience, the mobile energy storage scheduling method considering Bruker bar optimization in this embodiment is compared with the mobile energy storage scheduling method considering robust optimization.

[0055] Method 1: During the fault period, robust optimization is used for dynamic scheduling of mobile energy storage and Benders decomposition algorithm is used for calculation; Method 2: During the fault period, mobile energy storage is dynamically scheduled using Benders decomposition optimization and calculated using the Benders decomposition algorithm; Method 3: During the fault period, the mobile energy storage is dynamically scheduled using the split-blob bar optimization and calculated using the C&CG algorithm, which is the method in this embodiment.

[0056] The simulation results are analyzed as follows: Following the impact of extreme weather events, the coordinated restoration of distributed energy resources is crucial for rebuilding power supply. This embodiment explores in detail the dynamic scheduling strategies of two mobile energy storage units (referred to as mobile energy storage S1 and mobile energy storage S2) under different scenarios. The relevant power parameters are shown in Table 1 below, and the system operating parameters are shown in Table 2 below. The specific scheduling results are shown in Table 3, where discharge power is represented as a positive number and charging power as a negative number.

[0057] Table 1: Relevant power supply parameters:

[0058] Table 2: System Operating Parameters

[0059] Table 3: Dynamic scheduling results of mobile energy storage:

[0060] Furthermore, the complex relationship between the output power, current state of charge, and geographical location of these mobile energy storage units is presented in detail. Figure 6 and Figure 7 In addition, the charging and discharging power of electric vehicle charging stations at different times is as follows: Figure 8 As shown, the generator's active power output at different times during the fault period is as follows: Figure 9 As shown, a precise visualization analysis was also performed on the power supply load and the recovery rate of different types of loads within each time period, such as... Figure 10 As shown.

[0061] The results of Method 1 and Method 2 in this paper, including the load shedding power of critical loads, total load shedding power, average recovery rate of critical loads, and total scheduling cost, are shown in Table 4.

[0062] Table 4: Comparison of active power reduction and cost between Method 1 and Method 2:

[0063] The active power and the reduced active power were compared at different time periods, and the results are shown in [the table below]. Figure 11 From Table 4, Figure 10 It can be seen that Method 1 and Method 2 set up two mobile energy storage devices and respectively adopted robust optimization method and distributed robust optimization method to carry out multi-distributed resource collaborative restoration of load power consumption from 1:00 to 24:00. Since Method 2 adopts robust optimization for dynamic scheduling of mobile energy storage during the fault period, compared with Method 1, which adopts distributed robust optimization for dynamic scheduling of mobile energy storage during the fault period, the amount of important load reduction is reduced by 36.31kW, the total load reduction is reduced by 288.9012kW, the important load recovery rate is increased by 0.58%, and the cost of scheduling multi-source resource restoration is reduced by 689.2745 yuan.

[0064] A comparison of the solution time of Method 3 and Method 2 on the MATLAB simulation platform is shown in the following results. Figure 12 Table 5 shows a comparison of the solution algorithm and solution time. Table 5: Comparison of solution time and solution algorithm between Method 2 and Method 3

[0065] Depend on Figure 12 As shown in Table 5, the C&CG algorithm used in Method 3 has a significant advantage in solution time, which is reduced to within 0.02s.

[0066] This embodiment focuses on mobile energy storage and multi-source collaboration. Addressing the issue of distribution systems relying on the upstream power grid to meet load demands during extreme weather events, it proposes a two-layer scheduling model combining mobile energy storage and multi-source collaboration to enhance distribution network resilience. The main conclusions are as follows: (1) Compared with Method 1, Method 2 not only reduces the amount of critical load reduction and the overall system load reduction, but also improves the critical load recovery rate. Under the premise of ensuring the safe and stable operation of the power system, Method 2 uses a robust optimization method to perform uninterrupted dispatch of mobile energy storage during the fault phase. The dispatch scheme is more in line with the actual power system fault scenario, reducing the overall load reduction cost of the power system. Therefore, Method 2 is more economical than Method 1.

[0067] (2) By using the C&CG algorithm, not only can the limitations of the Benders algorithm in dealing with specific problems be overcome, but the efficiency and speed of model solving can also be effectively improved.

Claims

1. A method for improving the resilience of distribution networks in the energy internet based on spatiotemporal optimization scheduling of mobile energy storage, characterized in that, Includes the following steps: Establish a two-layer scheduling model, including an upper-layer model and a lower-layer model; The upper-level model adopts the sub-Bruker optimization method, which takes into account the uncertainty of fault nodes under extreme weather conditions and the impact of traffic network travel time, and optimizes the configuration quantity and location scheme of mobile energy storage. The lower-level model is based on the dynamic scheduling and time-series output characteristics of mobile energy storage, electric vehicles and diesel generators. It constructs a mixed integer quadratic cone programming model for multi-source collaborative recovery to solve the optimal load shedding scheme and the power recovery state of important loads. A failure probability model is constructed based on a two-level scheduling model; The effectiveness of the proposed method was verified through simulation analysis of an improved IEEE-33 node distribution network.

2. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 1, is characterized in that... In step S1, the objective function of the upper-level model is to minimize the mobile energy storage configuration cost, system risk cost, and load reduction cost, and its mathematical expression is: ； In the formula, For the set of load nodes; p A fuzzy set of the comprehensive norm of node faults under typhoon weather; Mathematical expectation of fuzzy sets with comprehensive norm; Defined as a 0-1 variable to characterize mobile energy storage devices and nodes. i The connection status between them.

3. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 2, is characterized in that... The constraints of the upper-level model include dynamic adjustment constraints for mobile energy storage, distribution network radiation topology constraints, load reduction constraints, distributed power output constraints, distribution network operation constraints, and spatiotemporal dynamic scheduling constraints for mobile energy storage.

4. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 3, is characterized in that... The dynamic adjustment constraints for mobile energy storage are: ； In the formula, during the energy storage deployment phase, the upper limit of the number of mobile energy storage units is determined by... express; Based on complex network theory, nodes and lines in a power system can be represented as a connected graph. ,Right now ;in It is the set of power nodes in a power grid, where different nodes in the power grid represent different functional attributes, such as power generation, substation, and load. The edge set of the power grid represents the lines in the power system; it is also represented by an adjacency matrix. The topological relationships between different nodes are represented by the following formula: In the formula: a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; The topological constraints of the distribution network are: ； ； ； ； In the formula, The set of all branches in the entire power grid; a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; This represents the total number of nodes within a power grid; for a specific node... i The child and parent nodes connected to it are respectively categorized into sets. and ; Describes the virtual power values ​​transmitted through each branch; This represents a 0-1 variable, indicating the virtual source node added to the network optimization, thus effectively taking into account the number of isolated nodes generated during network reconstruction into the optimization model; M represents the virtual power output by each node; during the calculation of this constraint, M is used as a maximum value. The load reduction constraint is: ； ； In the formula, the power reduction of the active power of the power grid load is: The power reduction of reactive power of the power grid load is ; The maximum active power reduction of the power grid load is The maximum reactive power reduction of the power grid load is ; The output constraints of distributed power sources are: ； ； ； In the formula: the node sets M, E, D, and P represent mobile energy storage systems, electric vehicle charging stations, diesel generator sets, and photovoltaic power generation systems, respectively; Used to characterize different types of distributed power sources and specific nodes i The connection status between them; and These represent the active and reactive power outputs of the distributed power source, respectively; and and This specifies the maximum limits for the active and reactive power output of these power sources; among them, and The upper and lower limits of the power factor for distributed power sources are defined.

5. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 4, is characterized in that... The operating constraints of the distribution network are: ； ； ； ； ； ； ； In the formula: and These are the active and reactive power of the branch circuit, respectively. a ij In the matrix The elements; when the point in the formula i With nodes j When there is a connection between elements a ij =1; otherwise a ij Then it is 0; and These are the branch resistance and reactance, respectively; and These are the squares of the node voltage and branch current, respectively; the maximum value of the square of the node voltage is... The minimum value of the square of the node voltage is ; The maximum value of the square of the branch current; The nonlinear constraint among voltage, current, and power is addressed using a second-order cone relaxation technique, transforming the nonlinear constraint into a second-order cone form. The constraint conditions are as follows: ； In the formula, and These are the active and reactive power of the branch circuit, respectively. This is the square term of the branch current; This is the maximum value of the square of the branch current.

6. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 5, is characterized in that... In the spatiotemporal dynamic scheduling constraints of mobile energy storage, mobile energy storage units are considered. i From node j To node k The transportation time, and in and During the node installation time, mobile energy storage units i At the node j With nodes k The travel time between them is The equivalent travel distance is and the actual driving speed is The spatiotemporal dynamic scheduling constraints of mobile energy storage are expressed as follows: ； ； ； In the formula, For connecting nodes j and nodes k The length of the road; This is the theoretical maximum speed assumed when there is no traffic flow; while c It represents specific disaster conditions.

7. The method for improving the resilience of the distribution network based on the spatiotemporal optimization scheduling of mobile energy storage for the energy internet, as described in claim 1, is characterized in that... The objective function of the lower-level model is to minimize the load reduction power during the fault period, and its mathematical expression is: ； In the formula: A set of periods of failure; Let y be the set of system operation decision variables under the scheduling scheme; y is the corresponding scheduling scheme. Let t be the active power reduction of the load at node i at time t.

8. The method for improving the resilience of the distribution network of the energy internet based on spatiotemporal optimization scheduling of mobile energy storage as described in claim 7, characterized in that, The constraints of the lower-level model include the charging and discharging power and state of charge constraints of mobile energy storage, the constraints of the electric vehicle charging decision model, the output constraints of distributed power sources, and the operation constraints of the distribution network.

9. The method for improving the resilience of the distribution network of the energy internet based on spatiotemporal optimization scheduling of mobile energy storage as described in claim 1, characterized in that, The failure probability model is as follows: Two assumptions are established: The average wind speed at the same location in different directions follows a single type of extreme value distribution; The model parameters for different directions at the same location are independent of each other, and the wind speed data samples for each direction are used to independently estimate the model parameters. The GEV distribution is adopted, and the distribution function is as follows: ； In the formula, , , These are the position parameter, scale parameter, and shape parameter, respectively. The scale parameter is greater than zero. v represents the extreme wind speed; overhead transmission lines have inherent limitations in their ability to withstand external forces. Based on the theory of metal deformation, the line fault probability model is as follows: ； In the formula, The probability of transmission line failure under strong winds; v The actual wind speed that the transmission line experiences; V This is the design value for the maximum wind speed that the transmission line can withstand.

10. The method for improving the resilience of the distribution network of the energy internet based on spatiotemporal optimization scheduling of mobile energy storage as described in claim 1, characterized in that, The method described uses an improved IEEE-33 node distribution network in the simulation analysis, with specific parameters including: Rated voltage, upper voltage limit, lower voltage limit; Installation and configuration time for mobile energy storage; Ideal vehicle speed for mobile energy storage under zero traffic conditions.

Images