Power distribution network post-disaster parallel recovery method based on dynamic boundary microgrid

Abstract

The invention discloses a power distribution network post-disaster parallel recovery method based on a dynamic boundary micro-grid, and is characterized in that the power distribution network parallel recovery method based on cooperation of a mobile emergency generator (MEG) and a transformer substation comprises the following steps: S1, on the basis of post-disaster fault isolation, carrying out node cell division on a power distribution network, and establishing a cell-level power distribution network recovery model; s2, fault equipment repair, switching operation and MEG scheduling are jointly considered, a dynamic boundary temporary micro-grid is constructed, and an energy supply recovery path is determined; and S3, establishing a power distribution network parallel recovery model based on dynamic boundary microgrid and substation energy supply path collaboration, and realizing gradual recovery and final access to a main network of the power distribution network. According to the post-disaster parallel recovery method for the power distribution network, the problem of time scale difference of multi-resource scheduling can be effectively solved, the power failure time of key loads is shortened, and the post-disaster recovery efficiency and the system toughness of the power distribution network are improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network post-disaster recovery and emergency power supply technology, and in particular to a parallel post-disaster recovery method for power distribution networks based on dynamic boundary microgrids, specifically involving mobile emergency generator access for power supply, temporary microgrid construction, and coordinated switching operation and fault repair. Background Technology

[0002] Under the influence of sudden events such as natural disasters, extreme weather, and external damage, power distribution network lines, switching equipment, and critical nodes are prone to varying degrees of damage, leading to localized or large-scale power outages. Since the power distribution network directly serves end users, its post-disaster recovery speed and sequence directly affect the continuity of residents' lives, public services, and essential production activities. Therefore, how to efficiently organize the power distribution network recovery process and shorten outage time under post-disaster conditions has become a crucial issue in power distribution network operation and emergency management.

[0003] Existing post-disaster recovery methods for power distribution networks typically focus on restoring power supply to substations, gradually expanding the power supply range downstream through fault isolation, equipment repair, and segmented power transmission. While this approach demonstrates high maturity and reliability in engineering practice, the overall recovery process exhibits a clear sequential characteristic: power restoration to downstream areas can only commence after the upstream power supply path is fully restored and ready for transmission. When the number of post-disaster faults is large and the repair work cycle is lengthy, this sequential recovery method can easily lead to prolonged periods of power outages in some areas or critical loads, significantly limiting overall recovery efficiency.

[0004] On the other hand, in post-disaster emergency response, temporary power supply equipment such as mobile emergency power generation units are often used to provide emergency power support for local areas or important users. However, in existing technologies, temporary power supply methods mostly exist in the form of "independent emergency protection," and their access location, power supply range, and operating time are usually determined by manual experience, lacking systematic coordination with the overall restoration process of the distribution network. Temporary power supply areas are often simply cut off after the main grid power supply is restored, failing to form an orderly connection with the main grid restoration path, thus making it difficult to fully realize the potential advantages of temporary power supply in shortening power outage time and improving overall restoration efficiency.

[0005] Furthermore, the post-disaster recovery of distribution networks typically involves multiple operational activities, including repair of faulty equipment, switching operations, temporary power supply connection, and power transmission to the main grid. These operations differ significantly in terms of time scale, execution conditions, and inter-constraint relationships. Without a unified recovery coordination mechanism, problems such as mutual constraints on recovery actions and unreasonable recovery sequences can easily arise, thereby affecting the feasibility and efficiency of the recovery process.

[0006] In summary, existing power distribution network post-disaster recovery technologies still primarily rely on a single-path, sequential recovery approach. There is a lack of effective coordination mechanisms between temporary power supply and main grid power supply, making it difficult to achieve parallel implementation of multiple recovery methods. In complex post-disaster scenarios, this type of recovery model still has significant shortcomings in shortening outage time, ensuring power supply to critical loads, and improving overall recovery efficiency. Therefore, it is necessary to propose a new power distribution network post-disaster recovery method that allows temporary power supply paths and main grid power supply paths to proceed in parallel during the recovery process, and to achieve their gradual integration through reasonable coordination and control, thereby improving the efficiency and reliability of power distribution network post-disaster recovery. Summary of the Invention

[0007] Symbols and definitions:

[0008] A. Exponents, sets, and matrices:

[0009]

[0010] B. Variables and constants

[0011]

[0012] C. Abbreviation

[0013]

[0014] The purpose of this invention is to address the problems of the conventional sequential recovery method used in the post-disaster recovery of distribution networks, which leads to long recovery cycles, difficulty in effectively shortening the outage time of critical loads, and a lack of unified coordination mechanism between temporary power supply and main grid power supply. This invention proposes a parallel recovery method for distribution networks after disasters. By introducing a recovery mechanism that promotes the parallel advancement of temporary power supply and main grid power supply during the post-disaster recovery process, the distribution network can simultaneously restore power to certain areas in advance while repairing faulty equipment and gradually restoring power supply paths. Furthermore, the invention enables dynamic adjustment and gradual grid connection of temporary power supply areas during the recovery process, thereby improving the overall efficiency and coordination of the distribution network's post-disaster recovery process and enhancing the feasibility and reliability of post-disaster power supply restoration.

[0015] To achieve the above objectives, this invention provides a parallel post-disaster recovery method for distribution networks based on dynamic boundary microgrids (MEGs), characterized by a parallel recovery method for distribution networks based on MEG and substation coordination, comprising the following steps:

[0016] S1. Based on post-disaster fault isolation, the distribution network is divided into node cells to establish a cell-level distribution network recovery model.

[0017] S2. Taking into account faulty equipment repair, switching operations, and MEG scheduling, construct a dynamic boundary temporary microgrid and determine the power restoration path;

[0018] S3. Establish a parallel recovery model for the distribution network based on the coordination of dynamic boundary microgrids and substation power supply paths to achieve the gradual recovery of the distribution network and its eventual connection to the main grid.

[0019] Preferred MEG-based dynamic microgrid recovery model:

[0020] (1) Dynamic Boundary Microgrid Power Supply Model

[0021] a) Energy supply path constraints

[0022] The MEG's power supply path will not pass through substation cells, nor will it traverse all cells. It will only pass through some cells, and each cell allows at most one MEG power supply path to pass through, ensuring the radial operation of the power grid.

[0023]

[0024]

[0025]

[0026]

[0027] In the formula: xEM m,n represents the closed state of the switching circuit between cells m and n under the action of MEG, and the power supply sequence is from m to n; N represents the set of node cells; N MG N represents the set of cells that can be accessed by MEG; SBS This represents a substation node cell.

[0028] b) Energy supply time constraints

[0029] If a path for MEG cell recovery exists, then the recovery time of the load cell depends on the closing time of the switches between cells:

[0030]

[0031] Linearization is as follows:

[0032]

[0033]

[0034]

[0035] In the formula: AEMm and AEMn are the power supply times of cells m and n, respectively; ASW,EMm,n is the time when the switch between cells m and n is closed; β m,n It is an auxiliary variable.

[0036] (2) MEG scheduling model

[0037] a) Movement path constraints

[0038] The MEG is retrieved from the warehouse, completes all power supply tasks, and eventually returns to the warehouse. Each access node can only connect to one MEG.

[0039]

[0040]

[0041]

[0042] In the formula: V MG and E MG They represent non-directional networks G and G respectively. MG (V MG E MG The set of nodes and branches in the network; e and f represent nodes in the network; xMG e,f,mg indicates whether MEG is scheduled from node e to node f; DP represents the set of all warehouses and maintenance stations.

[0043] b) Shift-time constraints

[0044]

[0045]

[0046] In the formula: AMG,AR e,mg represents the time when the generator mg arrives at node e; Tsty e,mg represents the time when mg stays at node e to supply power; TTRA e,f,mg represents the time required for mg to move from power supply node e to power supply node f; M is a large number used in the model building and linearization process.

[0047] (3) Coupled model

[0048] Only nodes with MEG access will serve as the starting point for MEG power supply path restoration. MEGs will only leave after being connected to the main grid, and the cells used as the restoration starting point will not be in other MEG power supply paths.

[0049]

[0050]

[0051]

[0052]

[0053] In the formula: is the node in cell m that can be connected to MEG; AES m is the moment when cell m is connected to the substation power supply path.

[0054] Preferred substation-based recovery model:

[0055] (1) Energy supply path and time model

[0056] a) Energy supply path constraints

[0057] The power restoration path for substations is similar to that of MEG, but the power restoration path for substations must pass through all cells, and ultimately all cells are connected to the main power grid for power supply:

[0058]

[0059]

[0060]

[0061] In the formula, xES m,n represents the closed state of the switch line between cells m and n in the substation path, and the power supply sequence is from m to n.

[0062] b) Energy supply time constraints

[0063] The power restoration time under the substation path depends on the switching closing time between cells:

[0064]

[0065]

[0066] In the formula, AES m and AES n represent the power supply time of cells m and n under the substation path, respectively; ASW,ESm,n represents the time when the switch between cells m and n is closed under the substation path; TSW m,n is the time required for the switch between cells m and n to close.

[0067] (2) Maintenance personnel scheduling model

[0068] a) Maintenance Path Model

[0069] Maintenance personnel depart from the repair station to the fault location to perform repair tasks, with only one team of personnel going to each fault location. After completing all repair tasks, the maintenance personnel return to the repair station.

[0070]

[0071]

[0072]

[0073] In the formula, V MC and E MC They represent non-directional networks G and G respectively. MC (V MC E MC The set of nodes and branches in the network; k and l represent nodes in the network; xMC k,l,mc indicates whether maintenance personnel mc is dispatched from fault point k to fault point l.

[0074] b) Maintenance event model

[0075]

[0076]

[0077]

[0078] In the formula, AMC,AR k,mc represents the time when maintenance personnel mc arrives at node k; AMC,RP k represents the time when the faulty equipment at node k is repaired; TRP k,mc represents the time required for personnel mc to repair the equipment at node k; and TTRA k,l,mc represents the time required for personnel mc to travel from node k to node l.

[0079] (3) Coupled model

[0080] Power can only be supplied to the main grid after all faults within the cell have been repaired.

[0081]

[0082] Before connecting to the MEG for power, all faults within the cell must be repaired. Equation (29) can be rewritten as follows:

[0083]

[0084] (30) Linearized as:

[0085]

[0086] In the formula, VMC m represents the set of faulty devices in node cell m; A m This represents the actual recovery time of cell m.

[0087] Preferred, MEG-based and substation-based coupled recovery models:

[0088] (1) Recovery path coupling model

[0089] a) Actual path constraints:

[0090] The actual MEG recovery path must be a part of the substation recovery path:

[0091]

[0092] b) Switch type constraints:

[0093] Switches are divided into quasi-synchronous parallel switches and non-quasi-synchronous parallel switches. Quasi-synchronous parallel switches can synchronize the voltage amplitude on both sides and can be closed directly. However, for non-quasi-synchronous parallel switches, if it is a MEG restoration line, it can be closed. If it is a substation restoration line, it can also be closed. It should be avoided to perform switch closing between different power supply lines.

[0094]

[0095] In the formula, This is a set of non-quasi-synchronous parallel switches.

[0096] (2) Recovery time-coupled model

[0097] a) Constraint on the actual action time of the switch:

[0098] When an MEG recovery path exists, the actual closing time of the switch is the same as the closing time of the switch under the MEG power supply path. When no MEG recovery path exists but a substation recovery path exists, the actual closing time of the switch is determined by the power supply recovery sequence. If no power supply path exists, the switch will not close.

[0099]

[0100]

[0101] (35) Linearized as:

[0102]

[0103] In the formula, ASW m,n represents the actual closing time of the switch between cells m and n; V SW For the set of operating switches; ASBS m is the actual time when cell m is connected to the substation.

[0104] b) Constraint on actual cell recovery time:

[0105] Similar to the actual closing time of a switch, the actual recovery time of a cell is also related to whether there is a MEG power supply path. The existence of a MEG power supply recovery path enables the node cell to restore power supply in advance before connecting to the main grid.

[0106]

[0107]

[0108] (38) Linearized as:

[0109]

[0110] c) Actual time constraints for cell access to substation:

[0111] The actual time a cell connects to the substation also depends on whether a MEG power supply path exists. If a MEG power supply path exists from cell m to cell n, the switches between them will close in advance, connecting to the substation power supply path together. If no MEG power supply path exists, then it depends on the actual switch closing time between cells m and n.

[0112]

[0113]

[0114]

[0115] (41) Linearized as:

[0116]

[0117] (3) Power-on state constraints

[0118] a) Energy supply time model:

[0119] If node cell m is at time A m If power is restored, the load in node cell m and distributed generation can be restored in T. c Recovery occurs after a time delay, which is mainly related to the grid connection time of controllable loads and the time required for synchronous grid connection of distributed generation; for uncontrollable load c in node cell m, its power supply time is the same as that of node cell m:

[0120]

[0121]

[0122] In the formula: t c The power supply time for device c; For t c A set; Ωc m is the set of controllable loads and distributed generation; Ωc m is the set of loads and node cells in node m.

[0123] b) Energy supply state model:

[0124] The operating state of the distribution network changes continuously with the grid connection of loads and distributed generation. Therefore, it is unnecessary to verify the safety constraints of the distribution network at every moment; it is sufficient to verify the safety operating constraints at each power supply event. Considering only the time t at each power supply event... c The load and distributed generation power supply status, the power supply status of switched lines, the power supply status of unswitched lines, and the MEG connection status:

[0125]

[0126]

[0127]

[0128]

[0129] (46)-(49) are linearized as follows:

[0130]

[0131]

[0132]

[0133]

[0134] In the formula: xEGS c,t represents the load or distributed generation c at time t. The power supply status; xEGS m,n,t represents the switching circuit mn at time t. The power supply status; i and j represent nodes in the distribution network; B m Let xEGS represent the set of lines in node cell m; xEGS i,j,t represents the line ij at time t. The power supply status.

[0135] Preferred safety operation constraints during power distribution network restoration:

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142] In the formula: I and B are the sets of distribution network nodes and branches; P i,j,t and Q i,j,t Let Pij represent the active and reactive power of line ij at time t; Pg,t and Qg,t represent the output of the distributed generation at node i at time t; Pmg,t and Qmg,t represent the output of the MEG at node i at time t; xEGS,L,t and xEGS,G,t are 0-1 variables, representing the load of node i and the power supply status of the distributed generation at time t; PLi is the active load of node i; f L U is the load power factor; i,t R represents the voltage magnitude at node i at time t; i,j ,Xi,j Let represent the resistance and reactance of line ij; U0 be the reference voltage; Pmax i,j ,Qmax i,j be the active and reactive power transmission capacity of line ij; Pmax i , Qmax i be the upper limit of active and reactive power output of distributed generation at node i; Umax i ,Umin i be the upper and lower limits of voltage amplitude at node i; n(i) and m(i) be the set of parent and child nodes of node i.

[0143] Preferred parallel recovery model for dynamic boundary microgrid resilient distribution systems:

[0144] Based on the above model, a parallel recovery model for a dynamic boundary microgrid resilient distribution system can be obtained, with the objective function being:

[0145]

[0146] In the formula: tLi represents the power supply time of the load at node i, ωi i The importance of the load on node i.

[0147] This model is driven by substations and MEG multiple power sources, and considers the dynamic construction of temporary microgrids during the recovery process, which are restored in parallel with the top-down power supply path of the main grid. At the same time, the model systematically considers various dynamic processes and their coupling relationships, such as maintenance personnel scheduling, switch switching, and MEG scheduling. It also solves the challenges caused by the differences in different operation time scales by using an event-driven method based on variable time steps, and can generate topology snapshots and switch operation sequences of the distribution network under parallel recovery. Attached Figure Description

[0148] Figure 1 This invention provides a framework for the post-disaster recovery method of power distribution networks.

[0149] Figure 2 This is a schematic diagram of the IEEE 123 node distribution network used in the embodiments of the present invention;

[0150] Figure 3 This is a diagram showing the movement path of maintenance personnel obtained by the method in this embodiment of the invention;

[0151] Figure 4 These are the MEG movement paths and time maps obtained by the method in this embodiment of the invention;

[0152] Figure 5 This is a diagram showing the switch closing sequence and timing obtained by the method in this embodiment of the invention;

[0153] Figure 6 This is a diagram showing the power restoration sequence of cell 10 obtained by the method in this embodiment of the invention;

[0154] Figure 7 This is a diagram showing the switching sequence and timing of the method in this embodiment of the invention under non-parallel recovery.

[0155] Figure 8 This is a cell recovery time diagram under different recovery methods according to an embodiment of the present invention.

[0156] Figure 9 This is a comparison chart of power and energy consumption under different recovery methods according to an embodiment of the present invention. Detailed Implementation

[0157] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0158] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0159] This invention aims to provide a parallel post-disaster recovery method for distribution networks. By simultaneously introducing temporary power supply paths and main grid power supply paths during the post-disaster recovery process, the method enables the distribution network to restore power to certain areas in advance while faulty equipment is repaired and power supply paths are gradually restored. It also achieves a gradual connection between temporary power supply and main grid power supply during the recovery process. This method uses node cells as the basic unit of post-disaster recovery, combining mobile emergency power generation units, temporary microgrid construction, and substation-led power restoration paths to form a post-disaster recovery process that utilizes multiple recovery methods in parallel. This improves the efficiency of distribution network post-disaster recovery and shortens the outage time of critical loads.

[0160] In the case study, the CPLEX solver was called on the GAMS computer language platform to optimize and solve the model, verifying the effectiveness of the proposed model.

[0161] A. Test System and Parameter Settings

[0162] This example uses an improved IEEE 123 node distribution network system, including one substation, four MEG access nodes, six distributed generation units, and 16 remote control switches, including 11 quasi-synchronous switches. The total load is 3235kW, with a critical load of 1060kW. It assumes 15 lines are simultaneously faulty across three sites, each with two maintenance personnel. Site 1 has two MEGs participating in dispatch, each with a maximum power of 200W. Details are as follows... Figure 2 As shown.

[0163] The model contains a large number of 0-1 variables, especially in the maintenance personnel scheduling model, which places a significant burden on the model solution. To reduce the computational burden and improve solution efficiency, faulty equipment can be clustered and assigned to personnel at specific stations. The clustering results are shown in Table 1. Furthermore, the maintenance personnel's repair schedule for faulty equipment is shown in Table 2, and the switch operation time is shown in Table 3.

[0164] Table 1 Clustering of Faulty Equipment

[0165] TABLE.1 Clustering results of faulted devices

[0166]

[0167] Table 2 Repair time for faulty equipment

[0168] TABLE.2 Repair times of faulted devices

[0169]

[0170] Table 3 Switch action time

[0171] TABLE.3 Switch operation times

[0172]

[0173] B. Results Analysis

[0174] Solving the above model using the proposed method takes 352 seconds. The obtained MEG movement path and maintenance personnel movement path are as follows: Figure 3 and Figure 4 As shown in the diagram. Each color represents a maintenance worker path or a MEG vehicle path.

[0175] In addition, the arrival and departure times of maintenance personnel are shown in Table 4. The power supply sequence under MEG operation is shown in Table 5, and the closing time and actual closing time of the switch in the MEG path are shown in Table 6. Figure 5 As shown in Table 7, the actual power supply time and actual substation connection time of the cells are as follows.

[0176] Table 4 Maintenance Personnel Dispatch Schedule

[0177] TABLE.4 Repair crew scheduling timetable

[0178]

[0179] Table 5 Energy Supply Sequence under MEG Effect

[0180] TABLE.5 Energization sequence under MEG support

[0181]

[0182] Table 6. Switch closing time and actual closing time in the MEG path

[0183] TABLE.6 Switching closing times and actual closing times along MEGpaths

[0184]

[0185] Table 7. Actual substation connection time and actual power supply time for cells

[0186] TABLE.7 Actual substation connection times and actual energization times of cells

[0187]

[0188] like Figure 5 As shown, the substation and MEG simultaneously serve as restoration power sources, supplying power to the de-energized loads in parallel. Through a synchronous switch, the dynamic boundary microgrid is ultimately connected to the substation-based power supply path, achieving complete restoration of the distribution system. At t=71min, MEG1 and MEG2 are connected to node cells 13 and 14 respectively, and together with the substation located at node cell 15, they perform parallel restoration of the distribution system. Subsequently, at t=113min, the boundary of the MEG1-based microgrid expands from node cell 13 to node cells 13+4+3, while the boundary of the MEG2-based microgrid expands from node cell 14 to node cells 14+10, and is connected to the substation power supply path through the synchronous switch between nodes 10 and 7. At this time, node cell 10 switches from MEG2 power supply to substation power supply. Subsequently, MEG2 is dispatched to node cell 11 at t=157min to restore node cell 1. As the recovery process progressed, the boundaries of the microgrids based on MEG1 and MEG2 gradually expanded, and finally, at t=499min, they were all connected to the substation power supply path, completing the full recovery of the power distribution system.

[0189] By constructing a dynamic boundary microgrid, some node cells can be restored before being connected to the substation, thus their actual power supply time is earlier than their connection time to the substation. Figure 6 Taking the power transmission process of node cell 10 as an example: at t=19min, the MEG reaches node 350, enabling NC14 to recover and serving as the starting point for the dynamic boundary microgrid recovery path; at t=38min, the repair team completes the repair of lines 109–110, and at t=68min, lines 101–105 are repaired; subsequently, at t=71min, switches 300–350 close, enabling node cell 10 to achieve power transmission through the MEG recovery path, and with the support of the distributed power source located at node 105, all loads within node cell 10 are restored, thus forming a temporary microgrid composed of node cell 10 and node cell 4; at t=103min, node cell 7 reconnects to the substation at node 15 and restores power supply; finally, at t=109min... When switch 97–197 is closed, node cells 10 and 14 are reconnected to the main network and powered by the substation. At this time, the MEG completes its task at this location, leaves node 350, and proceeds to the next node to continue performing subsequent tasks.

[0190] C. Comparative Analysis of Model Recovery Strategies

[0191] The proposed model for parallel restoration of distribution network power supply using MEG and substations is compared with a non-parallel restoration model that only restores distribution network power supply using substations to verify the superiority of the proposed model. Table 8 and Figure 7 The power supply path restoration sequence is based on the operation of the substation alone.

[0192] Table 8 Power Supply Sequence under Substation Operation

[0193] TABLE.8 Energization sequence under substation-only restoration

[0194]

[0195] As shown in Table 8 and Figure 7As shown, since the substation is the sole starting point of the power transmission path in the distribution system, when only the substation restoration path is used, the restoration of a node cell depends not only on the completion of its own fault repair but also on waiting for its upstream node cells to complete their restoration before power can be restored. For example, for node cell 1, in the model considering only the substation restoration path, the time required to restore it to normal operation is 330 minutes; however, in the restoration model that incorporates MEG and substation coordination, node cell 1 can achieve restoration in 157 minutes, significantly shortening the power outage duration.

[0196] By comparing the parallel recovery model and the non-parallel recovery model in terms of node cell recovery time, recovered load power, and recovered power, the results are as follows: Figure 8 and Figure 9 As shown. From Figure 8 It can be seen that the recovery time of more than half of the node cells was shortened to varying degrees under the parallel recovery model. Figure 9 This indicates that the power recovery rate of the parallel recovery model is significantly faster than that of the non-parallel recovery model. Within the same time frame, the cumulative recovery power of the parallel recovery model is consistently higher than that of the non-parallel recovery model. Furthermore, the total recovered power under the parallel recovery model reaches 15,970.84 kWh, which is 7.15% higher than the 14,905.08 kWh of the non-parallel recovery model.

Claims

1. A method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids, characterized in that, The parallel restoration method for distribution networks based on MEG and substation coordination comprises the following steps: S1. Based on post-disaster fault isolation, the distribution network is divided into node cells to establish a cell-level distribution network recovery model. S2. Taking into account faulty equipment repair, switching operations, and MEG scheduling, construct a dynamic boundary temporary microgrid and determine the power restoration path; S3. Establish a parallel recovery model for the distribution network based on the coordination of dynamic boundary microgrids and substation power supply paths to achieve the gradual recovery of the distribution network and its eventual connection to the main grid.

2. The method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids according to claim 1, characterized in that, MEG-based dynamic microgrid recovery model: (1) Dynamic Boundary Microgrid Power Supply Model a) Energy supply path constraints The MEG's power supply path will not pass through substation cells, nor will it traverse all cells. It will only pass through some cells, and each cell allows at most one MEG power supply path to pass through, ensuring the radial operation of the power grid. In the formula: xEM m,n represents the closed state of the switching circuit between cells m and n under the action of MEG, and the power supply sequence is from m to n; N represents the set of node cells; N MG N represents the set of cells that can be accessed by MEG; SBS This represents a substation node cell. b) Energy supply time constraints If a path for MEG cell recovery exists, then the recovery time of the load cell depends on the closing time of the switches between cells: Linearization is as follows: In the formula: AEMm and AEMn are the power supply times of cells m and n, respectively; ASW,EMm,n is the time when the switch between cells m and n is closed; β m,n It is an auxiliary variable. (2) MEG scheduling model a) Movement path constraints The MEG is retrieved from the warehouse, completes all power supply tasks, and eventually returns to the warehouse. Each access node can only connect to one MEG. In the formula: V MG and E MG They represent non-directional networks G and G respectively. MG (V MG E MG The set of nodes and branches in the network; e and f represent nodes in the network; xMG e,f,mg indicates whether MEG is scheduled from node e to node f; DP represents the set of all warehouses and maintenance stations. b) Shift-time constraints In the formula: AMG,AR e,mg represents the time when the generator mg arrives at node e; Tsty e,mg represents the time when mg stays at node e to supply power; TTRA e,f,mg represents the time required for mg to move from power supply node e to power supply node f; M is a large number used in the model building and linearization process. (3) Coupled model Only nodes with MEG access will serve as the starting point for MEG power supply path restoration. MEGs will only leave after being connected to the main grid, and the cells used as the restoration starting point will not be in other MEG power supply paths. In the formula: is the node in cell m that can be connected to MEG; AES m is the moment when cell m is connected to the substation power supply path.

3. The method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids according to claim 1, characterized in that, Substation-based recovery model: (1) Energy supply path and time model a) Energy supply path constraints The power restoration path for substations is similar to that of MEG, but the power restoration path for substations must pass through all cells, and ultimately all cells are connected to the main power grid for power supply: In the formula, xES m,n represents the closed state of the switch line between cells m and n in the substation path, and the power supply sequence is from m to n. b) Energy supply time constraints The power restoration time under the substation path depends on the switching closing time between cells: In the formula, AES m and AES n represent the power supply time of cells m and n under the substation path, respectively; ASW,ES m,n represents the time when the switch between cells m and n is closed under the substation path; TSW m,n is the time required for the switch between cells m and n to close. (2) Maintenance personnel scheduling model a) Maintenance Path Model Maintenance personnel depart from the repair station to the fault location to perform repair tasks, with only one team of personnel going to each fault location. After completing all repair tasks, the maintenance personnel return to the repair station. In the formula, V MC and E MC They represent non-directional networks G and G respectively. MC (V MC E MC The set of nodes and branches in the network; k and l represent nodes in the network; xMC k,l,mc indicates whether maintenance personnel mc is dispatched from fault point k to fault point l. b) Maintenance event model In the formula, AMC,AR k,mc represents the time when maintenance personnel mc arrives at node k; AMC,RP k represents the time when the faulty equipment at node k is repaired; TRP k,mc represents the time required for personnel mc to repair the equipment at node k; and TTRA k,l,mc represents the time required for personnel mc to travel from node k to node l. (3) Coupled model Power can only be supplied to the main grid after all faults within the cell have been repaired. Before connecting to the MEG for power, all faults within the cell must be repaired. Equation (29) can be rewritten as follows: (30) Linearized as: In the formula, VMC m represents the set of faulty devices in node cell m; A m This represents the actual recovery time of cell m.

4. The method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids according to claim 1, characterized in that, MEG-based and substation-based coupled recovery models: (1) Recovery path coupling model a) Actual path constraints: The actual MEG recovery path must be a part of the substation recovery path: b) Switch type constraints: Switches are divided into quasi-synchronous parallel switches and non-quasi-synchronous parallel switches. Quasi-synchronous parallel switches can synchronize the voltage amplitude on both sides and can be closed directly. However, for non-quasi-synchronous parallel switches, if it is a MEG restoration line, it can be closed. If it is a substation restoration line, it can also be closed. It should be avoided to perform switch closing between different power supply lines. In the formula, This is a set of non-quasi-synchronous parallel switches. (2) Recovery time-coupled model a) Constraint on the actual action time of the switch: When an MEG recovery path exists, the actual closing time of the switch is the same as the closing time of the switch under the MEG power supply path. When no MEG recovery path exists but a substation recovery path exists, the actual closing time of the switch is determined by the power supply recovery sequence. If no power supply path exists, the switch will not close. (35) Linearized as: In the formula, ASW m,n represents the actual closing time of the switch between cells m and n; V SW For the set of operating switches; ASBS m is the actual time when cell m is connected to the substation. b) Constraint on actual cell recovery time: Similar to the actual closing time of a switch, the actual recovery time of a cell is also related to whether there is a MEG power supply path. The existence of a MEG power supply recovery path enables the node cell to restore power supply in advance before connecting to the main grid. (38) Linearized as: c) Actual time constraints for cell access to substation: The actual time a cell connects to the substation also depends on whether a MEG power supply path exists. If a MEG power supply path exists from cell m to cell n, the switches between them will close in advance, connecting to the substation power supply path together. If no MEG power supply path exists, then it depends on the actual switch closing time between cells m and n. (41) Linearized as: (3) Power-on state constraints a) Energy supply time model: If node cell m is at time A m If power is restored, the load in node cell m and distributed generation can be restored in T. c Recovery occurs after a time delay, which is mainly related to the grid connection time of controllable loads and the time required for synchronous grid connection of distributed generation; for uncontrollable load c in node cell m, its power supply time is the same as that of node cell m: In the formula: t c The power supply time for device c; For t c A set; Ωcm is the set of controllable loads and distributed generation; Ωcm is the set of loads and node cells in node m. b) Energy supply state model: The operating state of the distribution network changes continuously with the grid connection of loads and distributed generation. Therefore, it is unnecessary to verify the safety constraints of the distribution network at every moment; it is sufficient to verify the safety operating constraints at each power supply event. Considering only the time t at each power supply event... c The load and distributed generation power supply status, the power supply status of switched lines, the power supply status of unswitched lines, and the MEG connection status: (46)-(49) are linearized as follows: In the formula: xEGS c,t represents the load or distributed generation c at time t. The power supply status; xEGS m,n,t represents the switching circuit mn at time t. The power supply status; i and j represent nodes in the distribution network; B m Let xEGS represent the set of lines in node cell m; xEGS i,j,t represents the line ij at time t. The power supply status.

5. The method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids according to claim 1, characterized in that, Safety constraints during power distribution network restoration: In the formula: I and B are the sets of distribution network nodes and branches; P i,j,t and Q i,j,t Let Pij represent the active and reactive power of line ij at time t; Pg,t and Qg,t represent the output of the distributed power source at node i at time t; Pmg,t and Qmg,t represent the output of the MEG at node i at time t; xEGS,L,t and xEGS,G,t are 0-1 variables, representing the load of node i and the power supply status of the distributed power source at time t. PLi represents the active load at node i; fi L U is the load power factor; i,t R represents the voltage magnitude at node i at time t; i,j ,X i,j Let Uij be the resistance and reactance of line ij; U0 be the reference voltage. Pmax i,j ,Qmax i,j represents the active and reactive power transmission capacity of line ij. Pmax i and Qmax i are the upper limits of active and reactive power output of distributed generation at node i. Umax i and Umin i are the upper and lower limits of the voltage amplitude of node i; n(i) and m(i) are the sets of parent and child nodes of node i.

6. The method for parallel post-disaster recovery of distribution networks based on dynamic boundary microgrids according to claim 1, characterized in that, Parallel recovery model for a dynamic boundary microgrid resilient distribution system: Based on the above model, a parallel recovery model for a dynamic boundary microgrid resilient distribution system can be obtained, with the objective function being: In the formula: tLi represents the power supply time of the load at node i, ωi i The importance of the load on node i. This model is driven by substations and MEG multiple power sources, and considers the dynamic construction of temporary microgrids during the recovery process, which are restored in parallel with the top-down power supply path of the main grid. At the same time, the model systematically considers various dynamic processes and their coupling relationships, such as maintenance personnel scheduling, switch switching, and MEG scheduling. It also solves the challenges caused by the differences in different operation time scales by using an event-driven method based on variable time steps, and can generate topology snapshots and switch operation sequences of the distribution network under parallel recovery.

Images