Power distribution network toughness improvement method based on V2G technology under high-proportion distributed energy access

By building an important load level model and an electric vehicle scheduling model, combined with the constraints of distributed energy and transportation network, the problem of insufficient synergy effect in the recovery process of the distribution network under extreme high temperature disasters is solved, and the robustness of the distribution network is achieved and rapid improvement, reducing power loss time and regulation costs.

CN118920541APending Publication Date: 2024-11-08LIUPANSHUI NORMAL UNIV
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202410965590.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In the case of mass failure of extreme high temperature disasters, the distribution network fails to fully consider the synergy between the electric vehicle dispatching process and island operation, distributed power supply and emergency repair teams during the recovery process, and under high proportion of new energy access , there are insufficient measures to improve the resilience of distribution networks, especially the suppression of the impact on the volatility of new energy and the lack of comprehensive indicators.

Method used

Build a distribution network model and electric vehicle scheduling model based on important load levels, and propose a multi-objective scheduling model, including maximum load recovery, minimum recovery time and minimum regulation cost, combining the safety constraints of distributed energy, transportation network and V2G sites and Running constraints are used to solve them using Dijkstra algorithm and multi-objective genetic algorithm.

Benefits of technology

It has achieved the improvement of the robustness, rapidity and redundancy of the distribution network in extreme events, reduced the loss time and regulation cost, and enhanced the anti-interference ability of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118920541A_ABST
    Figure CN118920541A_ABST
Patent Text Reader

Abstract

The invention discloses a power distribution network toughness improvement method based on a V2G technology under high-proportion distributed energy access, relates to the technical field of power systems, and is technically characterized in that the method proposes a multi-target toughness scheduling strategy based on the V2G technology by constructing a power distribution network security constraint in a recovery process. A multi-target scheduling model and a solving method are provided by establishing a power distribution network model based on load importance levels and a V2G scheduling model based on a scheduling process of an electric vehicle, and constructing safety constraints and V2G operation constraints in a power distribution network recovery process. According to the invention, the improved PGamp is adopted; a G69 verification model and a traffic network simplified model are verified, quantitative indexes of robustness, rapidity and redundancy are established and compared, and the result shows that the V2G station can effectively supplement power loss of the load after the island is formed, the load recovery amount is increased, and the power loss time is shortened.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method for improving the resilience of a distribution network based on V2G technology under high-proportion distributed energy access. Background Art

[0002] In order to effectively improve the power system's ability to cope with "low probability, high impact" extreme events, the "resilience improvement" of the distribution network has become a research hotspot. Resilience is mainly divided into three stages: pre-disaster defense, disaster resistance and post-disaster recovery. Among them, post-disaster recovery measures mainly rely on network reconstruction, island division, mobile emergency power generation vehicles and repair teams. As an important part of energy conservation and emission reduction in the national transportation field, electric vehicles (EVs) can participate in load recovery after a disaster by using EVs to interact with the grid (vehicle to grid, V2G, electric vehicles connected to the grid) to quickly transfer electricity to the load to be restored and reduce the pressure on the distribution network lines.

[0003] Scholars at home and abroad have verified the role of V2G technology in improving the resilience of distribution networks. The existing technology discloses the use of EVs to supply power to important loads after a power outage, and analyzes the feasibility of EVs participating in the recovery of distribution networks after disasters from an economic perspective; the existing literature discloses the use of limited resources to optimize their transfer paths and power supply recovery range, thereby reducing distribution network losses. The existing literature also proposes a strategy for using V2G technology to improve resilience under typhoon disasters, and proposes multiple indicators to evaluate the ability of V2G to participate in resilience recovery. In addition, due to the complexity of the EV participation scheduling model, it is necessary to consider reconstruction, island division, mobile emergency power generation vehicles and repair teams. The existing literature also constructs a two-stage framework for post-disaster recovery. In the first stage, the damaged lines, power generation and transportation components are clustered to reduce the optimization time of the second stage; in the second stage, distributed power sources, network reconstruction and maintenance personnel are coordinated to perform post-disaster recovery of the distribution network. The existing literature also establishes a comprehensive optimal recovery model that considers the switching sequence of remote control and manual switches, the optimal scheduling sequence of operators and maintenance personnel, and the load recovery power supply sequence to ensure the minimum recovery time of the distribution network.

[0004] The above existing technologies have made certain progress in the assessment and improvement of distribution network resilience, but they rarely consider the synergy between the EV dispatch process and distribution network resilience improvement measures such as island operation, distributed generation (DG), and repair teams in dealing with group failures caused by extreme high temperature disasters. With the access of distributed energy, the distribution network is susceptible to the volatility of new energy during the recovery process. V2G technology can obtain stable electricity within a certain period of time, so it can be used as a means to smooth out the volatility of new energy. However, there is little consideration of related content in the existing technologies. Secondly, the most commonly used goal is to minimize load losses or minimize power outages, and there are few comprehensive indicators.

[0005] Therefore, in the context of a new distribution system with a high proportion of new energy access, this invention proposes a method to improve the resilience of V2G technology in the recovery of distribution networks. Establish quantitative indicators of robustness, rapidity and redundancy to quantify system resilience; build models of various resilience resources such as distributed power sources in new distribution systems; study the safety constraints and operation constraints of distributed energy, distribution networks, transportation networks and V2G sites, and build a multi-objective scheduling model with maximum load recovery, minimum recovery time and minimum control cost for solution. Summary of the invention

[0006] The purpose of the present invention is to solve the above problems and provide a method for improving the resilience of a distribution network based on V2G technology under high-proportion distributed energy access.

[0007] In order to achieve the above object, the technical solution of the present invention is as follows:

[0008] The present invention provides a method for improving the resilience of a distribution network based on V2G technology under high-proportion distributed energy access. The method is: by establishing a distribution network model based on load importance levels and a V2G scheduling model based on the scheduling process of electric vehicles, safety constraints and V2G operation constraints in the distribution network recovery process are constructed, and a multi-objective scheduling model and a solution method are proposed; specifically, the following steps are included:

[0009] S1. Construct a distribution network restoration framework with multi-distributed resource collaboration:

[0010] Based on the basic structure of the distribution network with distributed energy access, combined with the transportation network, a photovoltaic, wind, and electric vehicle coordinated energy network is formed, including:

[0011] S1-1. Define distribution network resilience indicators: define system resilience indicators from three dimensions: robustness, rapidity and redundancy;

[0012] S1-2, build a distribution network restoration framework;

[0013] S2. Construct a distribution network model with coordinated multi-distributed resources:

[0014] S2-1. Construct distribution network model:

[0015] (1) Distribution network load model:

[0016] The weight of the node is determined according to the proportion of Class I, Class II, and Class III loads in the node, expressed as:

[0017] ω i =(s 1i k 1i +s 2i k 2i +s 3i k 3i )ω ∑

[0018]

[0019] In the formula, ω i is the weight of node i; s 1i 、s 2i 、s 3i , k 1i , k 2i , k 3i are the class I, class II and class III loads contained in node i and the proportion of class I, class II and class III loads to the total load of the node, respectively. 1i +k 2i +k 3i =1;ω ∑ The power S of node i i S ∑ ratio.

[0020] (2) Distribution network topology model:

[0021] The node weighted tree model T(V, E, W) is used to describe the topological structure of the distribution network under study; V, E, and W represent the nodes, edges, and node weights in the structure respectively; then W can be expressed as:

[0022] W(v i )=ω i (S Gi -S Li );

[0023] In the formula, S Gi , S Li The nodes v i The connected line pairs node v i The sum of the injected power of and the sum of the power carried by vi;

[0024] S2-2, Distributed Power Supply Model:

[0025] (1) Distributed photovoltaic model:

[0026] p PV (t) = η PV S PV G PV (t);

[0027] Where: η PV is the conversion efficiency; S PV is the photovoltaic array area; G PV (t) is the solar radiation intensity at time t;

[0028] (2) Distributed wind model:

[0029] It is expressed as follows:

[0030]

[0031] Where: P R is the rated power of the fan; v r , v in , v out are the rated wind speed, cut-in wind speed and cut-out wind speed of the wind turbine, which approximately obey the Weibull distribution;

[0032] S2-3, V2G participates in the capacity recovery model:

[0033] (1) Road network model:

[0034] Use Dijkstra algorithm to determine the shortest driving path;

[0035] G=(V tr ,E tr );

[0036] Where: G is the road network set; V tr To simplify the set of nodes in the back road network; E tr To simplify the set of connections between nodes; the generated matrix D(G) represents the direct distance between nodes in the road network;

[0037] (2) EV driving time model:

[0038] A. EV driving time model:

[0039] Assuming that the EV user's initial location and remaining battery capacity both satisfy normal distribution, the time it takes for an EV to reach its destination is calibrated by the Federal Highway Administration of the United States as

[0040]

[0041] Where: Q is the traffic volume of a certain road section, Q0 is the traffic volume of the road section under normal weather conditions; I p , Vp are visibility and rainstorm intensity respectively, τ0, τ1, τ2 are coefficients; μ is the friction coefficient, which is determined by the rainstorm intensity, road congestion and slipperiness; L k is the average body length of EV, C TP is the road capacity, α TP , ξ are calibration parameters; the average vehicle speed is χ k ; L i,j is the distance from the i-th point to the j-th node; assuming the statistical time is ΔT, then when When the vehicle reaches the destination, it is in the parking state, otherwise it is in the driving state;

[0042] B. EV's willingness to evacuate:

[0043] The evacuation willingness coefficient β is introduced to quantify the number of EVs responding to evacuation dispatch;

[0044] N ref =βN ev.total ;

[0045] Where: N ref N is the number of EVs participating in evacuation dispatch; ev.total is the total number of EVs in the system;

[0046] C. EV evacuation dispatch model:

[0047]

[0048] N sta (t) = N ref +N sta (t-1);

[0049] Where: N road is the number of EVs on the road; N sta is the number of EVs in the V2G station during the evacuation phase; d i,j is the driving distance from the i-th EV to the j-th shelter, which can be determined by the Dijkstra algorithm; P is the set of shelters; the above formula means that the EV goes to the nearest shelter for refuge;

[0050] (3) EV Capacity Model

[0051] It is assumed that the SOC value of the total EV before each trip satisfies the normal distribution; to reduce battery damage, it is assumed that the battery is charged when the battery power is lower than Y1 until the battery power reaches Y2 or above;

[0052] S3. Distribution network optimization dispatching model:

[0053] S3-1. Optimization goals

[0054] (1) Maximum recovery load

[0055] The goal is to restore the maximum load of the distribution network:

[0056]

[0057] In the formula, z is the set of isolated islands formed; i is the node in the isolated island; P Li is the load power of node i;

[0058] (2) Minimize power outage time

[0059] The time when the load loses power after a fault is a combination of the transportation time of the emergency repair resources, the maintenance time of the maintenance personnel, and the time when the EV user begins to participate in the recovery after arriving at the V2G site; the time when the system actively decouples to form an island is short and can be ignored; the distribution of emergency repair resources has a clear starting point, and the end point set can be obtained after the fault occurs. Its minimum path can be determined by Dijkstra. The transportation time from the qx node to the jth node can be calculated by referring to the above content The time T for the repair personnel to arrive at the fault point and repair the fault re It is related to the severity of the fault and the skills of the personnel. To simplify the calculation, this article takes a fixed value. The minimum power outage time of the system can be approximately expressed as:

[0060]

[0061] It represents the maximum time for EVs at different positions in the load to reach the V2G station; represents the time when EV transfers from node i to V2G at node j;

[0062] (3) Minimize deployment costs

[0063] The cost of emergency dispatch mainly includes the cost of wind and solar power abandonment of distributed power sources C DG , Switching loss S l , the cost of compensation from the service aggregator to the user C v2g ;

[0064] A. Cost of power abandonment of distributed power generation C DG

[0065] The system takes into account wind power and photovoltaic power generation, so the wind and photovoltaic abandonment can be written as:

[0066]

[0067] Where: C PV , C PW N is the penalty coefficient for abandoning solar power and wind power;PV 、N PW are the number of photovoltaic generators and wind turbines included in the system, respectively; The f-th photovoltaic generator and the m-th wind turbine generator are Δt The actual capacity value and capacity forecast value of the time period;

[0068] B. Switching loss S l

[0069] Assuming that all distribution network lines are equipped with soft switches, the fewer the switch actions, the shorter the recovery cycle, and the lower the cost of using the intermediate contact switch; and the number of times the switch is used is approximately proportional to the power supply distance; therefore, the paper uses the distance between the power source and the load receiving its power supply to represent the cost of using the switch;

[0070]

[0071] In the formula, S l is the cost of the switch that supplies power to load l; n pj The number of power clusters (i.e., the divided areas with power supplies inside) that need to be operated to supply power to load l; d lj is the distance between the jth V2G station and the load l powered by it;

[0072] C. The cost of compensation from the service aggregator to the user v2g

[0073]

[0074] Where: N all.re is the total number of nodes; β v2g and β re are the compensation price per unit response power and the compensation price per single response behavior respectively; Q N is the total number of regions participating in the scheduling; is the total capacity of EVs in each area participating in the dispatch, N EV.re is the number of EVs participating in the response;

[0075] S3-2. Constraints

[0076] A. Distribution Network Security Constraints

[0077] (1) Unit output constraints

[0078] For distributed power sources in the distribution network, the unit output should meet the following requirements:

[0079] u r P gmin <P g <u r P gmax

[0080] u r Q gmin <Q g <u r Q gmax ;

[0081] Where P g , Q g are the active and reactive power of power source g respectively; P gmin , P gmax , Q gmin , Q gma x is the lower and upper limits of active and reactive output of g respectively; u r is the investment factor, 1 means g is invested, 0 means g is not invested;

[0082] (2) Power balance constraints within an island

[0083] Wind power and photovoltaic power generation have strong volatility, which can be compensated by stable power sources such as diesel engines. The expression is:

[0084]

[0085]

[0086] In the formula, G and NG are the stable power source and unstable power source set respectively; P Gi , Q Gi , Q Gjmin , P Gimin are respectively the actual active power output, reactive power output and minimum power output of the stable power supply; P Gjmax and P Gjmin are the maximum and minimum output of unstable power supply respectively; P Lk , Q Lk They are the active and reactive power of the load in the island respectively; it is worth noting that when the number of EVs in the V2G site changes over time, it can be considered as an unstable power source when participating in power supply. When the EVs in the V2G are in charging state, the V2G site is considered as a load.

[0087] (3) Node voltage constraints

[0088] U imin ≤U i ≤U imax ;

[0089] Where U i , U imax , U imin are the node voltage amplitude and its upper and lower limits respectively; i=1,2,…,ne;

[0090] (4) Reactive power constraints and phase constraints of the line during the restoration process

[0091] Considering the linearized AC power flow constraint and assuming that the branch ratio is 1, we have:

[0092]

[0093] The above formula represents the branch active and reactive transmission power constraint;

[0094]

[0095] The above formula is the active power constraint of line transmission;

[0096]

[0097] The above formula is the reactive power constraint of line transmission;

[0098]

[0099] The above formula is a phase constraint;

[0100] In the formula, S b is the reference power; B ij , G ij is the mutual susceptance and mutual conductance of nodes i and j; θ i,t 、V i,t are the phase and voltage value of bus i respectively; θ + ,θ - are the upper and lower limits of the phase difference respectively; They are the upper and lower limits of active and reactive power allowed to flow through line ij respectively;

[0101] B. V2G operation constraints

[0102] (1) Constraints on the number of EVs in a V2G station

[0103] 0≤N sta,i (t)≤N EV.V2G (i,j)∈V;

[0104] The above formula indicates that during the post-disaster recovery process, the number of EVs accepted by each V2G site should be less than the total capacity of the site;

[0105]

[0106] The above formula states that the number of EVs accepted by all V2G stations during the post-disaster recovery process should be less than the maximum number of dispatchable EVs in the urban area;

[0107] Where: N sta,i (t) is the total number of vehicles in the i-th V2G station, NEV.V2G is the maximum number of EVs that a V2G site can accommodate;

[0108] (2) V2G site charging and discharging power constraints

[0109] 0≤S FV2G ≤N EV.V2G ×S v2g ;

[0110] S v2g is the charging and discharging power of a single charging pile in V2G, S FV2G is the discharge power that the V2G station can provide, and is the sum of the remaining power of all EVs in the station. The above formula indicates that the power that each V2G can provide at the same time cannot exceed the power that all charging piles in the station can discharge at the same time.

[0111] S3-3. Multi-objective model transformation and solution

[0112] (1) Multi-objective optimization model

[0113] The above-mentioned multi-objective optimization scheduling model of resilient resources is described as follows:

[0114]

[0115] (2) Solution of multi-objective optimization model

[0116] For the multi-objective model in the formula (1) above, first convert f1 into the standard minimization form:

[0117]

[0118] f2 is only related to the EV dispatch time, emergency repair time, etc., and each time is independent of each other; secondly, for the f3 control cost, the longer the EV dispatch time, the more battery capacity is consumed and the higher the cost, so f3 is positively correlated with f2; therefore, the objective function f2 can be optimized first and then the multi-objective genetic algorithm II can be used to solve it.

[0119] Furthermore, in step S1-1, robustness reflects the power supply capability of the distribution system to maintain a high load level in the face of disturbances; the load reduction ratio of the distribution network is used to characterize the robustness of the system; that is:

[0120]

[0121] Where: N is the number of total load nodes in the distribution network; ΔL n is the load loss of distribution network node n; L n is the load of node n before the extreme event occurs.

[0122] Furthermore, in step S1-1, a rapidity index is constructed to characterize how fast the load recovers to the optimal state; the system recovery time is mainly composed of the time for the EV to reach the V2G site, the time for the emergency repair equipment to arrive at the fault location, and the maintenance time, ignoring the power adjustment time of distributed renewable energy generation and the time for the system to be disconnected to form an island; the rapidity index is expressed as:

[0123] R T =min(T,(T qx +T re ));

[0124] Where: T is the actual time for EV to arrive at the V2G station in the traffic network; T qx The actual time it takes to transport the repair materials to the fault location; T re This is the maintenance time for emergency repair personnel.

[0125] Furthermore, in step S1-1, redundancy is used to enhance the anti-interference capability of the system through system backup, with the load-power number being used as a redundancy indicator.

[0126]

[0127] In the formula, R K is the redundancy index, N S is the total number of load nodes, M ij is the total number of paths from load i to load j, Q l is the reliable set of all links.

[0128] Furthermore, step S1-2 specifically includes the following steps:

[0129] (1) Divide the distribution network and urban transportation network into multiple grids to facilitate analysis and recovery of internal EV capacity and traffic conditions;

[0130] (2) By connecting to distributed energy sources such as distributed photovoltaic and wind power generation devices, the distribution network actively decouples and forms an island;

[0131] (3) Dispatching EVs to V2G sites to participate in load recovery until the entire distribution network resumes power supply.

[0132] Compared with the prior art, this solution has the following beneficial effects:

[0133] The method of the present invention proposes a multi-objective resilient scheduling strategy based on V2G technology by constructing distribution network safety constraints in the recovery process; by establishing a distribution network model based on load importance levels and a V2G scheduling model based on the scheduling process of electric vehicles, safety constraints and V2G operation constraints in the distribution network recovery process are constructed, and a multi-objective scheduling model and a solution method are proposed; an improved PG&G69 verification model and a simplified transportation network model are used for verification, and quantitative indicators of robustness, rapidity and redundancy are established for comparison, so that V2G sites can effectively supplement the power outage that can be reduced after the formation of an island, thereby increasing the load recovery amount and reducing the power outage time. BRIEF DESCRIPTION OF THE DRAWINGS

[0134] Figure 1 It is the basic structure of the distribution network considering the access of distributed energy in the embodiment of the present invention;

[0135] Figure 2 This is a flow chart of urban distribution network restoration in an embodiment of the present invention;

[0136] Figure 3 is a flow chart of the NSGAII algorithm in an embodiment of the present invention;

[0137] Figure 4 It is a simplified model of the topology and transportation network of the improved PG&E69 node test system in the embodiment of the present invention;

[0138] Figure 5 In the embodiment of the present invention, the distribution network is actively decoupled into island operation;

[0139] Figure 6 is the V2G scheduling and emergency repair optimization result in the embodiment of the present invention;

[0140] Figure 7 is a load recovery variation curve in an embodiment of the present invention. DETAILED DESCRIPTION

[0141] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solution of the present invention will be further described in detail below in conjunction with the embodiments of the present invention and the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0142] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below in conjunction with the embodiments.

[0143] Example:

[0144] The embodiment of the present invention provides a method for improving the resilience of a distribution network based on V2G technology under high-proportion distributed energy access. The method quantifies the system resilience by establishing quantitative indicators of robustness, rapidity and redundancy; constructs a model of various resilience resources such as distributed power sources in a new distribution system; studies the safety constraints and operation constraints of distributed energy, distribution networks, transportation networks and V2G sites, and constructs a multi-objective scheduling model of maximum load recovery, minimum recovery time and minimum control cost for solution; specifically:

[0145] 1. Distribution network restoration framework with multi-distributed resource collaboration

[0146] like Figure 1 As shown, it gives the basic structure of the distribution network considering the access of distributed energy. This structure is based on the distribution network and combines the transportation network to form an energy network that coordinates photovoltaic, wind power, and electric vehicles.

[0147] 1.1 Distribution network resilience indicators

[0148] The present invention defines system resilience indicators from three dimensions: robustness, rapidity and redundancy.

[0149] 1.1.1 Robustness indicators

[0150] Robustness reflects the ability of the distribution system to maintain a high load level in the face of disturbances. The present invention uses the load reduction ratio of the distribution network to characterize the robustness of the system.

[0151]

[0152] Where: N is the number of total load nodes in the distribution network; ΔL n is the load loss of distribution network node n; L n is the load of node n before the extreme event occurs.

[0153] 1.1.2 Rapidity Index

[0154] The present invention constructs a rapidity index to characterize how fast the load recovers to the optimal state. The system recovery time is mainly composed of the time it takes for the EV to reach the V2G site, the time it takes for the emergency repair equipment to reach the fault site, and the maintenance time, ignoring the power adjustment time of distributed new energy generation and the time it takes for the system to be disconnected to form an island; the rapidity index is expressed as:

[0155] R T =min(T,(T qx +T re ));

[0156] Where: T is the actual time for EV to arrive at the V2G station in the traffic network; T qx The actual time it takes to transport the repair materials to the fault location; Tre This is the maintenance time for emergency repair personnel.

[0157] 1.1.3 Redundancy Index

[0158] Redundancy refers to enhancing the anti-interference capability of the system through system backup. The present invention uses the load-power number as a redundancy indicator.

[0159]

[0160] In the formula, R K is the redundancy index, N S is the total number of load nodes, M ij is the total number of paths from load i to load j, Q l is the reliable set of all links.

[0161] 1.2. Distribution Network Restoration Framework

[0162] The present invention is constructed as Figure 1 The distribution network restoration framework is shown in the figure. The specific process is as follows:

[0163] (1) Divide the distribution network and urban traffic network into multiple grids to facilitate the analysis and recovery of internal EV capacity and traffic road conditions.

[0164] (2) By connecting to distributed energy sources such as distributed photovoltaic and wind power generation devices, the distribution network actively decouples and forms an island.

[0165] (3) Dispatching EVs to V2G sites to participate in load recovery until the entire distribution network resumes power supply.

[0166] 2. Distribution network model with multi-distributed resource collaboration

[0167] 2.1 Distribution Network Model

[0168] 2.1.1 Distribution network load model

[0169] The weight of the node is determined according to the proportion of Class I, Class II, and Class III loads in the node, expressed as:

[0170] ω i =(s 1i k 1i +s 2i k 2i +s 3i k 3i )ω ∑

[0171]

[0172] In the formula, ω i is the weight of node i; s1i 、s 2i 、s 3i , k 1i , k 2i , k 3i are the class I, class II and class III loads contained in node i and the proportion of class I, class II and class III loads to the total load of the node, respectively. 1i +k 2i +k 3i =1;ω ∑ The power S of node i i S ∑ ratio.

[0173] 2.1.2 Distribution network topology model

[0174] The present invention refers to reference

[20] and adopts the node weighted tree model T (V, E, W) to describe the distribution network topology structure under study. Where V, E, and W represent the nodes, edges, and node weights in the structure, respectively. Then W can be expressed as:

[0175] W(v i )=ω i (S Gi -S Li ) (4)

[0176] In the formula, S Gi , S Li The nodes v i The connected line pairs node v i The sum of the injected powers and v i The sum of the power carried;

[0177] 2.2 Distributed Power Supply Model

[0178] 2.2.1 Distributed photovoltaic model

[0179] p PV (t) = η PV S PV G PV (t) (5)

[0180] Where: η PV is the conversion efficiency; S PV is the photovoltaic array area; G PV (t) is the solar radiation intensity at time t.

[0181] 2.2.2 Distributed wind model

[0182] It is expressed as follows:

[0183]

[0184] Where: P R is the rated power of the fan; v r , v in , v out are the rated wind speed, cut-in wind speed and cut-out wind speed of the wind turbine, respectively, which approximately obey the Weibull distribution [18-19].

[0185] 2.3 V2G Participation Recovery Capacity Model

[0186] 2.3.1 Road Network Model

[0187] The present invention adopts Dijkstra algorithm to determine the shortest driving path.

[0188] G=(V tr ,E tr ) (7)

[0189] Where: G is the road network set; V tr To simplify the set of nodes in the back road network; E tr To simplify the connection set between nodes, the generator matrix D(G) represents the direct distance between nodes in the road network.

[0190] 2.3.2 EV driving time model

[0191] (1) EV driving time model

[0192] Assuming that the EV user's initial location and remaining battery capacity both satisfy normal distribution, the time it takes for an EV to reach its destination is calibrated by the Federal Highway Administration of the United States as

[0193]

[0194] Where: Q is the traffic volume of a certain road section, and Q0 is the traffic volume of the road section under normal weather conditions. p 、V p are visibility and rainstorm intensity, respectively, τ0, τ1, τ2 are coefficients. μ is the friction coefficient, which is determined by the rainstorm intensity, road congestion and slipperiness. k is the average body length of EV, C TP is the road capacity, α TP , ξ are calibration parameters. The average vehicle speed is χ k .L i,j is the distance from the i-th point to the j-th node. Assuming the statistical time is ΔT, when When the vehicle reaches the destination, it is in a parked state, otherwise it is in a driving state.

[0195] (2) EV evacuation intention

[0196] The evacuation willingness coefficient β is introduced to quantify the number of EVs responding to evacuation dispatch.

[0197] N ref =βN ev.total (8)

[0198] Where: N ref N is the number of EVs participating in evacuation dispatch; ev.total is the total number of EVs in the system.

[0199] (3) EV evacuation dispatch model

[0200]

[0201] N sta (t) = N ref +N sta (t-1)

[0202] Where: N road is the number of EVs on the road; N sta is the number of EVs in the V2G station during the evacuation phase; d i,j is the driving distance from the i-th EV to the j-th shelter, which can be determined by the Dijkstra algorithm; P is the set of shelters. The above formula indicates that the EV goes to the nearest shelter for refuge.

[0203] 2.3.3 EV Capacity Model

[0204] It is assumed that the SOC value of the total EV before each trip satisfies the normal distribution. To reduce battery damage, it is assumed that the battery is charged when the battery power is lower than Y1 until the battery power is Y2 or above, as shown in Table 1:

[0205] 1. Electric vehicles cannot be charged during driving and are in a power consumption state. The driving variable is 0 and the state variable is 0.

[0206] 2. In the docking state, SOC is less than Y1, the electric vehicle is neither charged nor discharged, the driving variable is 2, and the state variable is 0.

[0207] 3. When the vehicle is parked, the SOC is discharged between Y1 and Y2 until Y1, at which time the driving variable is 2 and the state variable is 2.

[0208] 4. In the docking state, under normal conditions and post-disaster recovery conditions, if SOC is greater than Y2, the electric vehicle is discharged, the driving variable is 2, and the state variable is 0.

[0209] Table 1 Electric vehicle charging and discharging behavior

[0210]

[0211] The electric capacity of an electric vehicle at time t and at time t-1 is related to the charging and discharging state, which can be written as:

[0212]

[0213] is the remaining battery capacity of the nth electric car at time t; P c , P f are the charging and discharging power of the electric vehicle within the calculation time Δt, in kW; W100 is the power consumption of the electric vehicle for 100 km, in kW·h; L m,n,Δt (t-1) is the driving distance of electric vehicles in the time period Δt. Then, the EV cluster capacity in different grids can be obtained.

[0214] 2.3.3 Real-time status model of available capacity of V2G sites

[0215]

[0216] Where: Sstatus is the collection of real-time status information of each EV after the evacuation dispatch is completed; Nv2g is the total number of EVs in the nth V2G station; is the set of remaining power of all EVs in the station; W is the set of driving states.

[0217] 3. Distribution network optimization dispatch model

[0218] 3.1 Optimization goals

[0219] 3.1.1. Maximum recovery load

[0220] The goal is to restore the maximum load of the distribution network

[0221]

[0222] In the formula, z is the set of isolated islands formed; i is the node in the isolated island; P Li is the load power of node i.

[0223] 3.1.2 Minimize power outage time

[0224] The time for the system to actively decouple and form an island is short and can be ignored; the distribution of emergency repair resources has a clear starting point, and the end point set can be obtained after the failure occurs. The minimum path can be determined by Dijkstra. The transportation time from the qx node to the jth node can be calculated by referring to the above content The time T for the repair personnel to arrive at the fault point and repair the fault re It is related to the severity of the fault and the skills of the personnel. To simplify the calculation, this article takes a fixed value. The minimum power outage time of the system can be approximately expressed as:

[0225]

[0226] It represents the maximum time for EVs at different positions in the load to reach the V2G station; represents the time when EV transfers from node i to V2G at node j;

[0227] 3.1.2 Minimize deployment costs

[0228] The cost of emergency dispatch mainly includes the cost of wind and solar power abandonment of distributed power sources C DG , Switching loss S l , the cost of compensation from the service aggregator to the user C v2g .

[0229] (1) The cost of power abandonment of distributed generation C DG

[0230] The system takes into account wind power and photovoltaic power generation, so the wind and photovoltaic abandonment can be written as:

[0231]

[0232] Where: C PV , C PW N is the penalty coefficient for abandoning solar power and wind power; PV 、N PW are the number of photovoltaic generators and wind turbines included in the system, respectively; The f-th photovoltaic generator and the m-th wind turbine generator are Δt The actual capacity value and the predicted capacity value for the time period.

[0233] (2) Switching loss S l

[0234] The fewer the switch actions, the shorter the recovery cycle and the lower the cost of the intermediate contact switch. The number of times the switch is used is approximately proportional to the power supply distance. Therefore, the paper uses the distance between the power source and the load that receives its power to represent the cost of using the switch.

[0235]

[0236] In the formula, S l is the cost of the switch that supplies power to load l; n pj The number of power clusters (i.e., the divided areas with power supplies inside) that need to be operated to supply power to load l; d lj is the distance between the jth V2G station and the load l powered by it.

[0237] (3) The cost C that the service aggregator compensates to users v2g

[0238]

[0239]

[0240] Where: N all.re is the total number of nodes; β v2g and β re They are the compensation price per unit response power and the compensation price per single response behavior; QN is the total number of regions participating in the scheduling; is the total capacity of EVs in each area participating in the dispatch, N EV.re is the number of EVs participating in the response.

[0241] 3.2 Constraints

[0242] 3.2.1. Distribution network security constraints

[0243] (1) Unit output constraints

[0244] For distributed power sources in the distribution network, the unit output should meet

[0245]

[0246] Where P g , Q g are the active and reactive power of power source g respectively; P gmin , P gmax , Q gmin , Q gmax g are the lower and upper limits of active and reactive output respectively; u r is the investment factor, 1 means g is invested, 0 means g is not invested.

[0247] (2) Power balance constraints within an island

[0248] Wind power and photovoltaic power generation have strong volatility, which can be compensated by stable power sources such as diesel engines. The expression is:

[0249]

[0250] In the formula, G and NG are the stable power source and unstable power source set respectively; P Gi , Q Gi , Q Gjmin , P Gimin are respectively the actual active power output, reactive power output and minimum power output of the stable power supply; P Gjmax and P Gjmin are the maximum and minimum output of unstable power supply respectively; P Lk , Q Lkare the active and reactive power of the load in the island respectively.

[0251] (3) Node voltage constraints

[0252] U imin ≤U i ≤U imax (twenty two)

[0253] Where U i , U imax , U imin are the node voltage amplitude and its upper and lower limits respectively; i=1,2,…,ne.

[0254] (4) Reactive power constraints and phase constraints of the line during the restoration process

[0255] Considering the linearized AC power flow constraint and assuming that the branch ratio is 1, we have:

[0256]

[0257] Formula (23) represents the active and reactive power transmission constraints of the branch; Formula (24) and Formula (25) are the active and reactive power constraints of the line transmission; Formula (26) is the phase constraint. b is the reference power; B ij , G ij is the mutual susceptance and mutual conductance of nodes i and j; θ i,t 、V i,t are the phase and voltage value of bus i respectively; θ + ,θ - are the upper and lower limits of the phase difference respectively; They are the upper and lower limits of active and reactive power allowed to flow through line ij respectively.

[0258] 3.2.2 V2G operation constraints

[0259] (1) Constraints on the number of EVs in a V2G station

[0260] 0≤N sta,i (t)≤N EV.V2G (i,j)∈V (27)

[0261]

[0262] Where: N sta,i (t) is the total number of vehicles in the i-th V2G station, N EV.V2G is the maximum number of EVs that a V2G site can accommodate. Formula (27) indicates that during the post-disaster recovery process, the number of EVs accepted by each V2G site should be less than the total capacity of the site. Formula (28) indicates that during the post-disaster recovery process, the number of EVs accepted by all V2G sites should be less than the maximum number of EVs that can be dispatched in the urban area.

[0263] (2) V2G site charging and discharging power constraints

[0264] 0≤S FV2G ≤N EV.V2G ×S v2g (30)

[0265] S v2g is the charging and discharging power of a single charging pile in V2G, S FV2G is the discharge power that the V2G station can provide, and is the sum of the remaining power of all EVs in the station. The above formula indicates that at the same time, the power that each V2G can provide cannot exceed the power discharged by all charging piles in the station at the same time.

[0266] 3.3 Multi-objective model transformation and solution

[0267] 3.3.1 Multi-objective optimization model

[0268] The above-mentioned multi-objective optimization scheduling model of resilient resources is described as follows:

[0269]

[0270] 3.3.2. Solution of multi-objective optimization model

[0271] For the multi-objective model in formula (31), the present invention first converts f1 into a standard minimization form:

[0272]

[0273] f2 is only related to the EV dispatch time, emergency repair time, etc., and each time is independent. Secondly, for the f3 control cost, the longer the EV dispatch time, the more battery capacity is consumed and the higher the cost, so f3 is positively correlated with f2. Therefore, the objective function f2 can be optimized first and then the multi-objective genetic algorithm II can be used to solve it. The solution process is as follows:

[0274] 4. Example Analysis

[0275] Based on the improved PG&E69 node test system, simulation analysis is performed on the topology of the power distribution network and the transportation network. Figure 4As shown. In the traffic network, a total of 300 EVs can be dispatched. Distributed new energy is connected to the distribution network nodes 5, 18, 36, and 52, of which DG1 and DG2 are distributed photovoltaics, DG3 is a distributed wind power station, and DG4 is an emergency diesel unit; there are V2G stations at nodes 11, 24, and 38, and each V2G station has a total of 30 charging piles. The charging and discharging power of a single charging pile is 7kW, and the charging and discharging characteristics of Y1 are 5kW·h and Y2 are 20kW·h; 35 and 54 are storage locations for faulty materials. In the simulation analysis, the distance and topology of the traffic network are given. It is assumed that the operation and repair time of the fault are 10 minutes respectively. The computer configuration used for the example analysis in this chapter is: IntelCore i5-10210U 1.60GHz, 16GB RAM, 64-bit operating system.

[0276] Assume that in an extreme event, the distribution network lines (3-59), (11-12), and (23-24) fail, the main grid is disconnected, and part (11-12) of the traffic network is inaccessible. At this time, reference

[20] actively decouples the distribution network to form the following Figure 5 The island shown.

[0277] Set up the following 3 options:

[0278] Solution 1: The distribution network restoration method proposed in the present invention considering distributed renewable energy generation and EV scheduling;

[0279] Solution 2: Remove the distributed renewable energy generation in Solution 1, that is, only V2G participates in the recovery;

[0280] Solution 3: Remove the V2G site in Solution 1, that is, only distributed renewable energy generation is involved in the restoration.

[0281] 4.1 Optimization Results

[0282] 4.1.1 Multi-objective optimization results

[0283] The load recovery amount and recovery index of different schemes are shown in Table 1.

[0284] Table 1 Comparison of optimization results

[0285]

[0286] Compared with the above schemes, the total load recovery amount is lower when only EV participates in the recovery. This is because there are fewer V2G sites and the charging power of a single charging pile is fixed. When V2G is fully powered, it is constrained by the line flow, and its transmission line cannot work at full load, resulting in low recovery amount; only distributed renewable energy generation has a more obvious effect on improving the resilience of the distribution system (the load recovery amount is increased by 218.1kW compared with scheme 2), and the cost of distributed renewable energy generation is low, which is conducive to reducing the deployment cost. However, by comparing schemes 1 and 3, it can be seen that the output of distributed renewable energy generation is limited (the optimization result is that the distributed renewable energy generation is fully generated and the line flow is limited). V2G participating in the distribution network recovery can provide electricity for the reduced load. Comparing the quantitative indicators, the robustness of the scheme proposed in this paper is 0.12 within the duration of this example, indicating that the scheme proposed in this paper can greatly enhance the system's anti-interference ability. The scheme comparison results verify the effectiveness of the proposed resilient resource collaborative scheduling method. In addition, due to the addition of EV scheduling in scheme 1, the deployment cost is relatively high. For decision makers, the process of compromise between the distribution network recovery amount and the emergency deployment cost in the distribution system.

[0287] 4.1.2 Detailed analysis of the solution of the present invention

[0288] A detailed analysis of the solution of the present invention shows that the shortest fault recovery time is 18.23 minutes. The V2G input time and its power supply range are shown in the figure below.

[0289] Figure 6 The blue load is the load inside the island, and the green part is the load powered by V2G after the V2G site is connected. T represents the time when the V2G site starts to participate in the recovery after accepting the EV, that is, the time when the EV is dispatched to the V2G site; T qx T represents the transportation time of transporting the repair materials to the fault point. re Indicates that repair time will be rushed.

[0290] Combination Figure 6 , Figure 7 It can be seen that at the beginning of the fault, the main grid and distributed power sources were the main power sources, the system was operating in multiple small island modes, and the load recovery was low. As EVs were dispatched to the V2G site for power supply after 3.19 minutes, the load recovery increased, corresponding to Figure 5 There were multiple turning points until 18.32 minutes later when the line (11-12) was repaired and all loads in the distribution network were powered normally.

[0291] In order to verify the influence of the V2G site location on the restoration of the distribution network, the V2G site location was changed and optimized according to the scheme of the present invention. The results are compared in the following table.

[0292] Table 2 Comparison of optimization results

[0293]

[0294] After the upstream V2G site was transferred to node 23, the subsequent capacity of the V2G site was not considered during the decoupling, and the system was decoupling at node (23-24). When the subsequent EV was transferred to node 23, it could not provide power to the downstream node, thus reducing the total load recovery. In sub-scheme 3, due to the line fault between nodes 12 and 11, the V2G site at node 12 could not provide power to node 11 until the line was repaired 26.98 minutes later. At this time, there were 24 EVs in the V2G site at node 12, with a dispatchable capacity of 341kW, which could meet the power requirements of nodes 11, 12, and 59-69, and the distribution network load was restored.

[0295] 5. Conclusion

[0296] This paper focuses on the problem of improving the resilience of new distribution systems under extreme events, establishes distribution network resilience indicators, and constructs a multi-objective optimization model and performs decoupling solutions by studying distribution network safety constraints and V2G operation constraints. The results show that distributed new energy has a low cost and can effectively reduce the system's load loss ratio and regulation cost when the distribution network fails, but it cannot guarantee the power supply of other loads after forming an island; EV can effectively transfer power to the power-off load.

[0297] The above specific embodiments are merely explanations of the present invention and are not limitations of the present invention. After reading this specification, those skilled in the art may make modifications to the embodiments without any creative contribution as needed. However, such modifications are protected by the patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A method to improve the resilience of distribution network based on V2G technology under high proportion of distributed energy access. Its characteristics are: The method is as follows: by establishing a distribution network operation model based on a load level weight model and a V2G dispatching model based on the dispatching process of electric vehicles, safety constraints and V2G operation constraints in the distribution network restoration process are constructed, and a multi-objective dispatching model and a solution method are proposed; The specific steps include: S1. Construct a distribution network restoration framework with multi-distributed resource collaboration: Based on the basic structure of the distribution network with distributed energy access, combined with the transportation network, a photovoltaic, wind, and electric vehicle coordinated energy network is formed, including: S1-1. Define distribution network resilience indicators: define system resilience indicators from three dimensions: robustness, rapidity and redundancy; S1-2, build a distribution network restoration framework; S2. Construct a distribution network model with coordinated multi-distributed resources: S2-1. Construct distribution network model: (1) Distribution network load model: The weight of the node is determined according to the proportion of Class I, Class II, and Class III loads in the node, expressed as: oh i =(s 1i k 1i +s 2i k 2i +s 3i k 3i )oh ∑ In the formula, ω i is the weight of node i; s 1i 、s 2i 、s 3i , k 1i , k 2i , k 3i are the class I, class II and class III loads contained in node i and the proportion of class I, class II and class III loads to the total load of the node, respectively. 1i +k 2i +k 3i =1;ω ∑ The power S of node i i S ∑ ratio. (2) Distribution network topology model: The node weighted tree model T(V, E, W) is used to describe the topological structure of the distribution network under study; V, E, and W represent the nodes, edges, and node weights in the structure respectively; then W can be expressed as: W(v i )=ω i (S Gi -S Li ); In the formula, S Gi , S Li The nodes v i The connected line pairs node v i The sum of the injected powers and v i The sum of the power carried; S2-2, Distributed Power Supply Model: (1) Distributed photovoltaic model: p PV (t)=η PV S PV G PV (t); Where: η PV is the conversion efficiency; S PV is the photovoltaic array area; G PV (t) is the solar radiation intensity at time t; (2) Distributed wind model: It is expressed as follows: Where: P R is the rated power of the fan; v r , v in , v out are the rated wind speed, cut-in wind speed and cut-out wind speed of the wind turbine, which approximately obey the Weibull distribution; S2-3, V2G participates in the capacity recovery model: (1) Road network model: Use Dijkstra algorithm to determine the shortest driving path; G=(V tr ,E tr ); Where: G is the road network set; V tr To simplify the set of nodes in the back road network; E tr To simplify the set of connections between nodes; the generated matrix D(G) represents the direct distance between nodes in the road network; (2) EV driving time model: A. EV driving time model: Assuming that the EV user's initial location and remaining battery capacity both satisfy normal distribution, the time it takes for an EV to reach its destination is calibrated by the Federal Highway Administration of the United States as Where: Q is the traffic volume of a certain road section, Q 0 is the traffic volume of the road section under normal weather conditions; I p 、V p are visibility and rainstorm intensity, τ 0 , τ 1 , τ 2 is the coefficient; μ is the friction coefficient, which is determined by the intensity of the rainstorm, the degree of road congestion and the degree of slipperiness; L k is the average body length of EV, C TP is the road capacity, α TP , ξ are calibration parameters; the average vehicle speed is χ k ; L i,j is the distance from the i-th point to the j-th node; assuming the statistical time is ΔT, then when When the vehicle reaches the destination, it is in the parking state, otherwise it is in the driving state; B. EV's willingness to evacuate: The evacuation willingness coefficient β is introduced to quantify the number of EVs responding to evacuation dispatch; N ref =βN ev.total ; Where: N ref N is the number of EVs participating in evacuation dispatch; ev.total is the total number of EVs in the system; C. EV evacuation dispatch model: N sta (t)=N ref +N sta (t-1); Where: N road is the number of EVs on the road; N sta is the number of EVs in the V2G station during the evacuation phase; d i,j is the driving distance from the i-th EV to the j-th refuge station, which can be determined by the Dijkstra algorithm; P is the set of refuge stations; D near It means that the EV goes to the nearest shelter for shelter; (3) EV Capacity Model It is assumed that the SOC value of the total EV before each trip satisfies the normal distribution; to reduce battery damage, it is assumed that the battery is charged when the battery power is lower than Y1 until the battery power reaches Y2 or above; S3. Distribution network optimization dispatching model: S3-1. Optimization goals (1) Maximum recovery load The goal is to restore the maximum load of the distribution network: In the formula, z is the set of isolated islands formed; i is the node in the isolated island; P Li is the load power of node i; (2) Minimize power outage time The time when the load loses power after a fault is a combination of the transportation time of the emergency repair resources, the maintenance time of the maintenance personnel, and the time when the EV user begins to participate in the recovery after arriving at the V2G site; the time when the system actively decouples to form an island is short and can be ignored; the distribution of emergency repair resources has a clear starting point, and the end point set can be obtained after the fault occurs. Its minimum path can be determined by Dijkstra. The transportation time from the qx node to the jth node can be calculated by referring to the above content The time T for the repair personnel to arrive at the fault point and repair the fault re It is related to the severity of the fault and the skills of the personnel. To simplify the calculation, this article takes a fixed value. The minimum power outage time of the system can be approximately expressed as: max(T i j ) represents the maximum time for EVs at different positions in the load to reach the V2G station; T i j represents the time when EV transfers from node i to V2G at node j; (3) Minimize deployment costs The cost of emergency dispatch mainly includes the cost of wind and solar power abandonment of distributed power sources C DG , Switching loss S l , the cost of compensation from the service aggregator to the user C v2g ; A. Cost of power abandonment of distributed power generation C DG The system takes into account wind power and photovoltaic power generation, so the wind and photovoltaic abandonment can be written as: Where: C PV , C PW N is the penalty coefficient for abandoning solar power and wind power; PV 、N PW are the number of photovoltaic generators and wind turbines included in the system, respectively; is the actual capacity value and predicted capacity value of the f-th photovoltaic generator and the m-th wind turbine in the Δt period; B. Switching loss S l Assuming that all distribution network lines are equipped with soft switches, the fewer the switch actions, the shorter the recovery cycle, and the lower the cost of the intermediate contact switch; and the number of times the switch is used is approximately proportional to the power supply distance; therefore, the paper uses the distance between the power source and the load receiving its power supply to represent the cost of using the switch; In the formula, S l The cost of the switch to supply power to load l; n pj The number of power clusters (i.e., the divided areas with power supplies inside) that need to be operated to supply power to load l; d lj is the distance between the jth V2G station and the load l powered by it; C. The cost of compensation from the service aggregator to the user v2g Where: N all.re is the total number of nodes; β v2g and β re They are the compensation price per unit response power and the compensation price per single response behavior; QN is the total number of regions participating in the scheduling; is the total capacity of EVs in each area participating in the dispatch, N EV.re is the number of EVs participating in the response; S3-2. Constraints A. Distribution Network Security Constraints (1) Unit output constraints For distributed power sources in the distribution network, the unit output should meet the following requirements: u r Q gmin <P g <u r Q gmax u r Q gmin <Q g <u r Q gmax ; Where P g , Q g are the active and reactive power of power source g respectively; P gmin , P gmax , Q gmin , Q gmax g are the lower and upper limits of active and reactive output respectively; u r is the investment factor, 1 means g is invested, 0 means g is not invested; (2) Power balance constraints within an island Wind power and photovoltaic power generation have strong volatility, which can be compensated by stable power sources such as diesel engines. The expression is: In the formula, G and NG are the stable power source and unstable power source set respectively; P Gi , Q Gi , Q Gjmin , P Gimin are respectively the actual active power output, reactive power output and minimum power output of the stable power supply; P Gjmax and P Gjmin are the maximum and minimum output of unstable power supply respectively; P Lk , Q Lk They are the active and reactive power of the load in the island respectively; it is worth noting that when the number of EVs in the V2G site changes over time, it can be considered as an unstable power source when participating in power supply. When the EVs in the V2G are in charging state, the V2G site is considered as a load. (3) Node voltage constraints IN imin ≤U i ≤U imax ; Where U i , U imax , U imin are the node voltage amplitude and its upper and lower limits respectively; i=1,2,…,ne; (4) Reactive power constraints and phase constraints of the line during the restoration process Considering the linearized AC power flow constraint and assuming that the branch ratio is 1, we have: The above formula represents the branch active and reactive transmission power constraint; The above formula is the active power constraint of line transmission; The above formula is the reactive power constraint of line transmission; The above formula is a phase constraint; In the formula, S b is the reference power; B ij , G ij is the mutual susceptance and mutual conductance of nodes i and j; θ i,t 、V i,t are the phase and voltage value of bus i respectively; θ + ,θ - are the upper and lower limits of the phase difference respectively; They are the upper and lower limits of active and reactive power allowed to flow through line ij respectively; B. V2G operation constraints (1) Constraints on the number of EVs in a V2G station 0≤N sta,i (t)≤N EV.V2G (i,j)∈V; The above formula indicates that during the post-disaster recovery process, the number of EVs accepted by each V2G site should be less than the total capacity of the site; The above formula states that the number of EVs accepted by all V2G stations during the post-disaster recovery process should be less than the maximum number of dispatchable EVs in the urban area; Where: N sta,i (t) is the total number of vehicles in the i-th V2G station, N EV.V2G is the maximum number of EVs that a V2G site can accommodate; (2) V2G site charging and discharging power constraints 0≤S FV2G ≤N EV.V2G ×S v2g ; S v2g is the charging and discharging power of a single charging pile in V2G, S FV2G is the discharge power that the V2G station can provide, and is the sum of the remaining power of all EVs in the station. The above formula indicates that the power that each V2G can provide at the same time cannot exceed the power that all charging piles in the station can discharge at the same time. S3-3. Multi-objective model transformation and solution (1) Multi-objective optimization model The above-mentioned multi-objective optimization scheduling model of resilient resources is described as follows: (2) Solution of multi-objective optimization model For the multi-objective model in the formula (1) above, firstly, f 1 Converted to standard minimized form: f 2 It is only related to the EV dispatch time, emergency repair time, etc., and each time is independent of each other; secondly, for f 3 In the regulation cost, the longer the EV dispatch time is, the more battery capacity is consumed and the higher the cost is. 3 With f 2 Positive correlation; therefore, we can first optimize the objective function f 2 Then the multi-objective genetic algorithm II is further used to solve the problem.

2. The method for improving the resilience of a distribution network based on V2G technology under high proportion of distributed energy access as claimed in claim 1, Its characteristics are: In step S1-1, robustness reflects the power supply capability of the distribution system to maintain a high load level in the face of disturbances; the load reduction ratio of the distribution network is used to characterize the robustness of the system; that is: Where: N is the number of total load nodes in the distribution network; ΔL n is the load loss of distribution network node n; L n is the load of node n before the extreme event occurs.

3. The method for improving the resilience of a distribution network based on V2G technology under high proportion of distributed energy access as claimed in claim 1, Its characteristics are: In step S1-1, a rapidity index is constructed to characterize how fast the load recovers to the optimal state; the system recovery time is mainly composed of the time it takes for the EV to reach the V2G station, the time it takes for the emergency repair equipment to arrive at the fault location, and the maintenance time, ignoring the power adjustment time of distributed renewable energy generation and the time it takes for the system to be disconnected to form an island; the rapidity index is expressed as: R T =min(T,(T qx +T re )); Where: T is the actual time for EV to arrive at the V2G station in the traffic network; T qx The actual time it takes to transport the repair materials to the fault location; T re This is the maintenance time for emergency repair personnel.

4. The method for improving the resilience of a distribution network based on V2G technology under high proportion of distributed energy access as claimed in claim 1, Its characteristics are: In step S1-1, redundancy enhances the anti-interference capability of the system through system backup, with the load-power number as the redundancy index. In the formula, R K is the redundancy index, N S is the total number of load nodes, M ij is the total number of paths from load i to load j, Q l is the reliable set of all links.

5. The method for improving the resilience of a distribution network based on V2G technology under high proportion of distributed energy access as claimed in claim 1, Its characteristics are: Step S1-2 specifically includes the following steps: (1) Divide the distribution network and urban transportation network into multiple grids to facilitate analysis and recovery of internal EV capacity and traffic conditions; (2) By connecting to distributed energy sources such as distributed photovoltaic and wind power generation devices, the distribution network actively decouples and forms an island; (3) Dispatching EVs to V2G sites to participate in load recovery until the entire distribution network resumes power supply.

Citation Information

Cited By

  • Method for improving and reinforcing toughness of coupling system of power distribution network and traffic network in extreme weather

    CN119253620A

  • Power distribution area emergency supply insurance method and system based on vehicle network space-time matching, and medium

    CN119382132A

  • Emergency supply guarantee method, system and medium for distribution station area based on vehicle-grid time-space matching

    CN119382132B