Power distribution network dispatching method considering vehicle network interaction mode division

Abstract

The invention relates to a power distribution network dispatching method considering vehicle network interaction mode division, which comprises the following steps of: establishing a road network model, enabling road network nodes and power distribution network nodes to be in joint correspondence, and constructing a road network-power distribution network coupling model; dividing EV user types and constructing corresponding travel feature models; establishing respective response models based on response modes of participating in vehicle network interaction under subjective willingness of different EV users; and based on the established model and distributed photovoltaic and energy storage in the distribution network, with the minimum deviation between the network loss cost, the EV scheduling cost and the bus voltage as the target, constructing a day-ahead-real-time two-stage power distribution network operation optimization scheduling problem considering EV automobile scheduling mode diversification, and solving the problem to obtain a real-time scheduling scheme. Compared with the prior art, the method has the advantages that the EV connected into the power distribution network is called to participate in power distribution network dispatching, the power flow distribution of the power distribution network is effectively adjusted, the network loss caused by power flow delivery due to power supply of the main network is reduced, and the operation cost of the power distribution network is further reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network optimization and dispatching, and in particular to a two-stage distribution network optimization and dispatching strategy considering different EV participation dispatching modes under the background of large-scale access of electric vehicles (EV) to urban distribution networks via vehicle to grid (V2G) interaction technology. Background Art

[0002] The promotion and construction of large-scale charging stations in cities has led to an increasing rate of EV access, and the power grid faces the risk of random load access. The risk brought by this uncertainty may be further amplified by the disordered charging behavior of EVs in large-scale charging stations, posing a huge threat to the reliability of the power system. Therefore, the research on the orderly charging of EVs connected to the power grid has received attention in recent years. Li Jinpeng, Feng Hua, Chen Xiaogang, etc. considered the multiple uncertainties after EVs were connected to the distribution network, established a distributed robust optimization model to predict the spatiotemporal distribution of EV charging loads, and intuitively managed the overall reliability of the system. Jia Shicheng, Liao Kai, Yang Jianwei, etc. considered the differences in charging loads at the spatiotemporal level and proposed a distribution network regional division method based on spectral clustering to effectively reduce the operating cost of the distribution network. F.Chen, F.Li, R.Hong, M.Guo, Z.Dai and R.Mo designed a control strategy for orderly charging of electric vehicles under the premise of considering the continuity of charging for electric vehicle users, which improved the operating efficiency of the distribution network and reduced the electricity cost of electric vehicle users. The above literature only considers the uncertainty of EVs as loads during charging, does not give play to the energy storage characteristics of EVs, and cannot fully play the supporting potential of EVs for grid dispatch optimization.

[0003] The introduction of V2G technology allows electric vehicles to accept charging and discharging scheduling when connected to charging piles. They are no longer just connected to the distribution network system as power loads. When necessary, they can also inject the power stored in their batteries into the power grid, thereby providing flexible power support Shang YT, Yu, H, Niu SY, et al. On the other hand, by actively and orderly regulating the charging and discharging of EVs, the grid-side load can be smoothed and the system abundance level can be improved Lü Si, Wei Zhinong, Sun Guoqiang, et al. In current research, many scholars have conducted in-depth research and discussion on the application of V2G technology in power systems. For example, Wang Ying, He Jinghan, Xu Yin, etc. and Su Su, Wei Cunhao, Chen Qifang, etc. have dispatched emergency power generation vehicles to supply power to regional power grids, but have not fully explored the potential of a large number of small EVs in urban distribution networks as distributed energy storage resources. Ge Xiaolin, Cao Shipeng, Fu Yang, etc. established a V2G-based electric vehicle scheduling model with the goal of minimizing charging costs and network loss costs, proving the great potential of EVs as flexible resources in improving the flexibility and reliability of power systems. Jin Guobin, Li Shuang, Li Guoqing and others applied EV clusters to emergency optimal dispatching of AC / DC hybrid distribution networks. Their experimental results showed that by relying on the spatiotemporal energy controllable characteristics of EVs, emergency optimal dispatching can be achieved in terms of economy and reliability of distribution networks. However, they did not further discuss the impact of user subjective willingness on EV participation in dispatching response.

[0004] In summary, the current research on EVs participating in the optimal dispatching of distribution networks through V2G charging piles under normal conditions still focuses on dispatching EVs as centralized loads, or simply dispatching individual EVs as distributed energy storage. The impact of the subjective willingness differences of different types of EV users on the degree of EV participation in dispatching has not been classified. Under the subjective influence of users, the change in the vehicle-grid interaction mode will affect the distribution network dispatching plan by changing the electrical characteristics of individual EVs when they are connected to the distribution network through V2G charging piles. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a distribution network scheduling method that takes into account the division of vehicle-grid interaction modes in a typical urban scenario.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A distribution network dispatching method considering the division of vehicle-grid interaction modes, the method steps comprising:

[0008] Obtain the vehicle OD travel probability matrix, establish a road network model, and jointly correspond the road network nodes with the distribution network nodes to construct a road network-distribution network coupling model;

[0009] Classify EV users into types based on travel modes, and construct travel characteristic models for corresponding user types; wherein the EV user classification types include: private car users, taxi users, and bus users;

[0010] According to the classification of EV user types, based on the response modes of different EV users participating in different degrees of vehicle-grid interaction, respective response models are established; wherein the response modes include: full response EV, partial response EV and non-participation response EV;

[0011] Based on the established model, taking into account the distributed photovoltaic and energy storage in the distribution network, with the goal of minimizing network loss cost, EV scheduling cost and bus voltage deviation, a two-stage optimal scheduling of the distribution network day-ahead and real-time operation considering the diversity of EV vehicle scheduling modes is constructed, and the real-time scheduling problem of the distribution network day-ahead and real-time operation is solved to obtain a real-time scheduling plan.

[0012] As a preferred technical solution, the road network model in the road network-distribution network coupling model is as follows:

[0013]

[0014] Where: P(R) is the node in the road network, n i represents the i-th road node, k is the number of road network nodes, E(R) is the road section, L road is the road length matrix, V road is the road resistance coefficient matrix, l ij For node n i With node n j Length between, v ij is the direct path resistance coefficient between node i and node j;

[0015] EV vehicles travel in the road network, enter traffic nodes according to the given optimal route to perform charging and discharging actions, and transmit charging load information from the traffic nodes to the corresponding distribution network nodes;

[0016] The load information in the road network nodes is stored in the corresponding distribution network node load information matrix through the mapping function, realizing the coupling of urban road network-distribution network information.

[0017] As a preferred technical solution, the travel characteristics of the private car user are as follows:

[0018] The probability density function expression of the first trip time of a private car user on the day is as follows:

[0019]

[0020] Where: T s1 is the starting time of the first trip, α is the scale parameter, β is the first shape parameter, and μ is the second shape parameter;

[0021] The time that private vehicles stay in residential and work areas follows a Gaussian distribution, and the time that they stay in commercial areas follows an exponential distribution.

[0022] As a preferred technical solution, the travel characteristics of the taxi users are as follows:

[0023] The probability density function expression of the first trip time of a taxi user on the same day is as follows:

[0024]

[0025] Where: T s1 is the starting time of the first trip, α is the scale parameter, β is the first shape parameter, and μ is the second shape parameter;

[0026] When the SOC of the taxi user's vehicle battery reaches the minimum charging threshold, the taxi user drives to the nearest charging station for charging. The docking time can be expressed as:

[0027]

[0028] Where: S car,n is the battery capacity of the nth vehicle, is the maximum capacity that the battery of the nth vehicle can accept, is the battery capacity when the nth vehicle joins the network, P ch It is the charging power of the charging pile per unit time.

[0029] As a preferred technical solution, the bus user considers the fixed charging load characteristics of the bus station during non-business hours.

[0030] As a preferred technical solution, the charging and discharging model of the fully responsive EV is as follows;

[0031]

[0032] Where: is the capacity of the nth EV in period t, and are the charging and discharging power of the nth EV in period t, and They represent the charging and discharging efficiency of EV respectively;

[0033] The vehicle dispatch cost is expressed as:

[0034]

[0035] Where: N1 represents the total number of EVs that can be fully dispatched, is the electricity price purchased by the distribution network from EV at time t, EV receives full dispatch subsidy per unit time.

[0036] As a preferred technical solution, the partial response EV is used as a large-scale flexible load resource, and the charging and discharging model of the EV is:

[0037]

[0038] The charging constraints when connected to a charging pile are:

[0039]

[0040] Through dispatching instructions, when the total EV charging load in a dispatching cycle remains unchanged, part of the charging load is transferred to other time periods within the parking time range;

[0041] The running time before transfer is The acceptable time period after transfer is Then the dispatch cost of EV that can be used as flexible load is:

[0042]

[0043] Where: C tr is the dispatch cost per unit power transfer, is the load power of the nth EV transferred within the time period.

[0044] As a preferred technical solution, the non-participating EV is a rigid load that is not subject to distribution network regulation.

[0045] The charging model of accessing the charging station is the same as the flexible load charging model:

[0046]

[0047] The charging constraints are:

[0048]

[0049] As a preferred technical solution, the day-ahead phase of the two-stage optimization scheduling of the distribution network day-ahead and real-time operation is to construct a multi-objective function of the distribution network day-ahead with the goal of minimizing the distribution network operation cost and the load fluctuation:

[0050] Minimize the total cost of operating charging stations and urban distribution networks:

[0051]

[0052] In the formula, is the total operating cost of the distribution network in the day-ahead stage; C buy is the cost of power purchased by the distribution network from the upper power grid, C PV is the cost of abandoned rooftop photovoltaic power generation, CESS is the operating cost of the energy storage power station, C loss is the network loss cost, F EV The total cost of dispatching EVs;

[0053] The dispatch objects are the power purchased by the distribution network from the upper main grid, the output power of rooftop photovoltaic power plants, and the charging and discharging power of energy storage power stations;

[0054]

[0055] F EV =F EV，1 +F EV,2

[0056] Where: τ is the scheduling period; Δt is the scheduling time interval; λ t The real-time electricity purchase price; is the active power purchased by the distribution network from the upper power grid during period t; RL is the capacity price; is the capacity purchased by the distribution network from the upper power grid at time t; c PV Penalty for abandoning light for the unit;Ω PV It is the set of nodes connected to photovoltaic in the distribution network; is the maximum active power generated by the photovoltaic power plant at node i at time t; is the actual power of wind power and photovoltaic power connected to the grid at node i at time t; c ESS is the energy storage charging and discharging cost coefficient; and are the charging and discharging power of the energy storage in period t; J is the total number of distribution network nodes; r ij is the resistance of branch ij; U i,t is the voltage of node i at time t; P ij,t and Q ij,t is the active and reactive power flowing through branch ij at time t; C e Penalty for unit active network loss;

[0057] Considering the grid base load and EV charging and discharging load, minimizing daily load fluctuations:

[0058]

[0059] Where: is the mean square error of the load in the day-ahead phase, indicating the load fluctuation of the distribution network; is the foundation load at time t; are the charging power and discharging power of the nth EV respectively; P av is the average daily load.

[0060] As a preferred technical solution, in the real-time stage of the two-stage optimization scheduling of the distribution network day-ahead and real-time operation, the distribution network scheduling optimization is only performed for the period when the EV response situation changes and the period thereafter, with the goal of minimizing the distribution network operation cost and the load fluctuation. The objective function is:

[0061]

[0062] Where: is the total operating cost of the distribution network in real time; is the mean square error of load in the real-time stage.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] The distribution network dispatching strategy method proposed in the present invention considers the division of vehicle-grid interaction modes. By analyzing the vehicle traffic data of each node, the OD travel probability matrix is ​​obtained, and an urban road network model is created. The road network nodes are jointly matched with the distribution network nodes to construct an urban road network-distribution network coupling model. The vehicle usage data and travel data of urban residents are analyzed, and different EV user types are divided. Through various fitting method experiments, the optimal fitting function is found, and the power consumption characteristics of different types of EV users are simulated by Monte Carlo. According to the division of different types of EV users, different response modes of participating in vehicle-grid interaction under the subjective will of different users are proposed, and respective response models are recommended, so that the EV vehicle participation in the distribution network dispatching mode can be more accurately divided. The EVs connected to the network are fully mobilized to participate in the distribution network dispatching, the distribution of the distribution network flow is effectively adjusted, and the network loss caused by the outward flow caused by the power supply of the main network is reduced, thereby reducing the operating cost of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic diagram of the road-network coupling relationship;

[0066] Figure 2 A histogram of user travel state transition probabilities;

[0067] Figure 3 This is a schematic diagram of the vehicle scheduling model;

[0068] Figure 4 This is a schematic diagram of the process of solving the day-ahead-real-time scheduling problem;

[0069] Figure 5 It is a schematic diagram of the coupling model of road network 26 nodes and distribution network 33 nodes;

[0070] Figure 6 It is the photovoltaic output curve;

[0071] Figure 7 Schematic diagram of EV output and load demand under two-stage scheduling;

[0072] Figure 8 It is a schematic diagram of node voltage comparison;

[0073] Fig. 9 This is a schematic diagram for comparing network losses in different time periods. DETAILED DESCRIPTION

[0074] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0075] Example 1

[0076] This paper proposes a distribution network dispatching strategy method that considers the division of vehicle-grid interaction modes in a typical urban scenario, the core of which mainly includes the following contents:

[0077] S1. By analyzing the vehicle traffic data of each node, the OD travel probability matrix is ​​obtained, and the urban road network model is established. The road network nodes are matched with the distribution network nodes to construct the urban road network-distribution network coupling model.

[0078] S2. Analyze the car usage and travel data of urban residents, classify different EV user types, find the optimal fitting function through various fitting method experiments, and perform Monte Carlo simulation on the electricity consumption characteristics of different types of EV users.

[0079] S3. Based on the classification of different types of EV users, different response modes for different users to participate in vehicle-grid interaction under their subjective willingness are proposed, and respective response models are recommended, so that a more accurate classification of EV vehicles participating in distribution network scheduling modes can be achieved.

[0080] S4. Based on the above model, taking into account distributed photovoltaic and energy storage in the distribution network, considering the economics of distribution network operation, in order to enhance the reliability of distribution network operation, with the goal of minimizing network loss cost, EV scheduling cost and bus voltage deviation, the distribution network operation is designed to consider the diversified EV vehicle scheduling mode. The day-ahead-real-time two-stage optimization scheduling. Finally, the Cplex solver is used to solve the distribution network optimization scheduling problem with vehicle-grid interaction.

[0081] (1) Urban road network-distribution network coupling model

[0082] like Figure 2 As shown in the figure, the road network is an important carrier that reflects the spatiotemporal characteristics of EVs, and is also the basis for characterizing their travel characteristics and charging load requirements. This paper divides urban areas into three main categories: residential areas, work areas, and commercial areas, and models them using graph theory. The road network model is as follows:

[0083]

[0084] Where: N(R) is the node in the road network, n i represents the i-th road node, k is the number of road network nodes, E(R) is the road section, L road is the road length matrix, V road is the road resistance coefficient matrix, l ij For node n i With node n j Length between, v ij is the direct path resistance coefficient between node i and node j.

[0085] EV vehicles travel in the road network and enter traffic nodes according to the given optimal route to perform charging and discharging operations, and transmit charging load information from the traffic nodes to the corresponding distribution network nodes. Figure 1 As shown. Information processing is performed on the distribution network side through the mapping function f: The load information in the road network nodes Store the corresponding distribution network node load information matrix In order to realize the information coupling of urban road network and distribution network.

[0086] (2) OD travel probability matrix

[0087] The concept of random walk is introduced, which means that an object moves randomly and gradually within a certain range in space. That is, assuming that an object is located at a node in the network, the object moves to other nodes with a fixed probability of movement. In the traffic network, the random movement of vehicles between road sections can be regarded as random walks. Therefore, the EV traffic volume on each road section in each period can be obtained by analyzing the urban road data published by the transportation department, and then the OD travel probability corresponding to each traffic node can be obtained using Transcad software. Figure 3 The OD travel probability matrix is:

[0088]

[0089] Where: a ij Represents the vehicle with node n i Travel from the starting point to node n j The probability of being the destination.

[0090] (3) Analysis of EV travel characteristics

[0091] The behavioral habit factors of EV users are closely related to the characteristics of charging load. Accurate analysis of the above influencing factors is the theoretical basis for accurately predicting charging load. Based on the travel survey report, the present invention selects the travel data of urban residents and divides EV vehicles into three types according to travel mode: private cars, taxis and buses. The probability distribution statistics of travel data in a known form are obtained through data sorting, and then the travel variables to be fitted are fitted and analyzed for probability modeling. The probability density function expression of the first travel time of a private car user on the day is as follows:

[0092]

[0093] Where: T s1 is the starting time of the first trip, α is the scale parameter, β is the first shape parameter, and μ is the second shape parameter.

[0094] The time that private vehicles stay in residential and work areas follows a Gaussian distribution, as shown in formula (4), and the time that they stay in commercial areas follows an exponential distribution, as shown in formula (5).

[0095]

[0096] Taxi users mainly pick up passengers, so the probability of their first trip of the day is the same as that of private cars. At the same time, based on the characteristics of taxis carrying passengers, the parking time during operation is negligible. When the vehicle battery SOC reaches the minimum charging threshold, it will drive to the nearest charging station for parking and charging until the battery SOC reaches the maximum value and then leaves. Therefore, its parking time can be expressed as:

[0097]

[0098] Where: S car,n is the battery capacity of the nth vehicle, is the maximum capacity that the battery of the nth vehicle can accept, is the battery capacity when the nth vehicle joins the network, P ch It is the charging power of the charging pile per unit time.

[0099] Bus users are managed by their bus operating companies. They charge their buses at fixed company parking lots during non-business hours. Their daily operations are strictly carried out according to the prescribed timetable and operating routes. The daily power consumption of their vehicles is basically the same. Therefore, only the fixed charging load characteristics during non-business hours at the bus stations are considered.

[0100] (4) V2G charging and discharging model

[0101] Compared with the original disorderly charging EVs, the EV cluster connected to the distribution network through V2G charging piles realizes the unified scheduling of EV charging and discharging status. EVs can determine the charging and discharging status according to the system's scheduling instructions. The V2G charging and discharging model is as follows:

[0102]

[0103] Where: is the capacity of the nth EV in period t, and are the charging and discharging power of the nth EV in period t, and They represent the charging and discharging efficiency of EV respectively.

[0104] The V2G charging and discharging process meets the constraints:

[0105]

[0106] Where: and are the upper and lower limits of EV battery capacity, and are the upper and lower limits of the charging power of the n-th EV and are the upper and lower limits of the charging power and discharging power of the n-th EV respectively.

[0107] EV users who accept charging dispatch will report the charging target power set by the dispatch center. To ensure that the battery charge level is sufficient when the vehicle leaves:

[0108]

[0109] At the same time, the acceptable scheduling time for the vehicle after reaching the charging target can be obtained:

[0110]

[0111] The vehicle battery meets the following requirements during the dispatch process:

[0112]

[0113] Where: It is the minimum capacity that the battery of the nth vehicle can accept.

[0114] (5) Fully responsive to EV

[0115] During the EV's dispatchable time, if the user chooses to fully cooperate with the dispatch center's control, the EV's discharge requirements to the grid can be realized through V2G equipment. That is, the vehicle is completely idle during the parking time and can fully respond to dispatch instructions. Private cars and buses are the main representatives.

[0116] In the statistics of family travel data, private cars generally travel between work and home, and most of them are divided into two situations: working during the day and staying at home at night. The vehicles are parked in a certain place for a long time and are idle. Buses only need to ensure that they have sufficient power when they are put into use the next day. After parking at the company's charging station at night, they are also idle for a long time, so they can also participate in full dispatch.

[0117] In the case of full dispatch, EV can be connected to the distribution network node where the charging station is located as a generalized energy storage and receive unified dispatch from the dispatch center. Its charging and discharging model is consistent with the V2G charging and discharging model (7). By giving full play to the energy storage characteristics of vehicles, it can not only smooth out the peak and valley differences in loads and maintain power balance, but also bring a certain improvement to the photovoltaic absorption capacity of urban rooftops. When used as a generalized energy storage, the vehicle dispatch cost can be expressed as:

[0118]

[0119] Where: N1 represents the total number of EVs that can be fully dispatched, is the electricity price purchased by the distribution network from EV at time t, EV receives full dispatch subsidy per unit time.

[0120] (6) Partial response EV

[0121] There are some EV users who are unwilling to participate in the V2G mode, but can accept orderly charging. These EVs can be used as large-scale flexible load resources, that is, under the premise of meeting the travel needs of users, the battery power of the vehicle reaches the target power when the vehicle leaves the charging pile. The charging and discharging model of EV as a flexible load is:

[0122]

[0123] When this part of EV load interacts with the distribution network, its power demand can be flexibly changed during the parking period. EV can determine the charging time according to the system's scheduling instructions. As a flexible load, the charging constraints of EV charging when connected to the charging pile are:

[0124]

[0125] Through the dispatch instruction, the total EV charging load can be transferred to other time periods within the parking time range while the total EV charging load remains unchanged within a dispatch cycle. The acceptable time period after transfer is The dispatch cost of EVs that can be used as flexible loads is

[0126]

[0127] Where: C tris the dispatch cost per unit power transfer, is the load power of the nth EV transferred within the time period.

[0128] (7) Not participating in the response EV

[0129] Some EV users do not participate in grid dispatching during the charging process, mainly taxi users who use vehicle driving as a means of operation. These users also only have charging as their main demand and do not perform discharge operations. Therefore, their charging model for accessing charging stations is consistent with the flexible load charging model (13). This type of EV load can be regarded as a rigid load, that is, a load that must guarantee electricity demand under normal circumstances and cannot be regulated by the distribution network.

[0130] Since the actual demand of this type of user is to minimize the parking time of the vehicle and put the vehicle into operation as soon as possible, only the fast charging model is selected, so the charging constraint is:

[0131]

[0132] (8) Day-ahead-real-time scheduling strategy

[0133] 1) Day-ahead stage

[0134] Assuming that the travel plan and response mode of each EV remain unchanged in the day-ahead stage, the day-ahead multi-objective function of the distribution network is constructed with the goal of minimizing the distribution network operation cost and load fluctuation.

[0135] The objective function for minimizing the distribution network operation cost is shown in formula (17), which is to minimize the total operating cost of the charging station and the urban distribution network, including the cost of purchasing electricity from the upper power grid, the cost of rooftop photovoltaic power generation connected to the distribution network, the charging and discharging cost of the connected energy storage station, the network loss cost of the distribution network, and the EV scheduling cost.

[0136]

[0137] In the formula, is the total operating cost of the distribution network in the day-ahead stage; C buy is the cost of power purchased by the distribution network from the upper power grid, C PV is the cost of abandoned rooftop photovoltaic power generation, C ESS is the operating cost of the energy storage power station, C loss is the network loss cost, F EV The total cost of dispatching EVs.

[0138] At the same time, the dispatching objects are the power purchased by the distribution network from the upper main grid, the output power of rooftop photovoltaics, and the charging and discharging power of energy storage power stations.

[0139]

[0140] FEV =F EV，1 +F EV,2 (twenty two)

[0141] Where: τ is the scheduling period; Δt is the scheduling time interval; λ t The real-time electricity purchase price; is the active power purchased by the distribution network from the upper power grid during period t; RL is the capacity price; is the capacity purchased by the distribution network from the upper power grid at time t; c PV Penalty for abandoning light for the unit;Ω PV It is the set of nodes connected to photovoltaic in the distribution network; is the maximum active power generated by the photovoltaic power plant at node i at time t; is the actual power of wind power and photovoltaic power connected to the grid at node i at time t; c ESS is the energy storage charging and discharging cost coefficient; and are the charging and discharging power of the energy storage in period t; J is the total number of distribution network nodes; r ij is the resistance of branch ij; U i,t is the voltage of node i at time t; P ij,t and Q ij,t is the active and reactive power flowing through branch ij at time t; C e Penalty for unit active network loss.

[0142] The distribution network load in the present invention mainly considers the basic load in the network and the EV charging and discharging load, and the minimum daily load fluctuation objective function is:

[0143]

[0144] Where: is the mean square error of the load in the day-ahead phase, indicating the load fluctuation of the distribution network; P av is the daily average load; is the foundation load at time t; are the charging power and discharging power of the nth EV respectively.

[0145] 2) Real-time stage

[0146] In the real-time stage, the distribution network scheduling optimization is only performed for the period when the EV response situation changes and the period thereafter. Its scheduling target is consistent with the day-ahead, and the objective function is:

[0147]

[0148] Where: is the total operating cost of the distribution network in real time; is the mean square error of load in the real-time stage.

[0149] 3) Solution process

[0150] The solution process of the above scheduling model is as follows: Figure 4 shown.

[0151] First, import common load data, road network information, EV characteristic parameters, and photovoltaic data;

[0152] In the recent stage:

[0153] Generate travel data for the nth EV; calculate the travel power consumption and determine whether the EV is connected to a charging station;

[0154] If the EV is not connected to the charging pile, the calculation returns to calculate the travel power consumption of n+1 EVs; if the nth EV is connected to the charging pile, its parking time is generated and it is determined whether it participates in the scheduling;

[0155] If EV participates in the dispatch; if EV does not participate in the dispatch, charging load is generated;

[0156] Record the scheduling cycle and repeat the above steps until all EVs are traversed;

[0157] The Cplex solver is used to optimize the day-ahead scheduling plan.

[0158] In real-time:

[0159] In the current period, determine whether the response of each EV changes;

[0160] If the EV response changes, the response mode of the corresponding EV is updated, the distribution network node number where it is located is obtained, and the charging and discharging plan is determined;

[0161] After completing the traversal and update of the response status of all EVs, the day-ahead scheduling plan is optimized and the next moment is entered;

[0162] After the day-ahead stage scheduling plan optimization for the current period is completed, the real-time stage scheduling plan is output.

[0163] (9) Constraints

[0164] 1) Power flow constraints

[0165] The DistFlow model is used to calculate the power flow of the distribution network and obtain the power flow distribution. The specific calculation model is as follows:

[0166]

[0167] Where: z:j→z represents the set of terminal nodes z with node j as the means; is the active and reactive output of the load and distributed generation at node j at time t; Iij,t is the current on branch ij at time t; r ij is the resistance of the ij branch; x ij is the reactance of the ij branch; U j,t is the voltage at node j at time t.

[0168] The computation constraints are:

[0169]

[0170] Where: U1 is the equilibrium node voltage; U ref is the reference value of U1; I ij,max Indicates the maximum current value allowed on the ij branch.

[0171] In the power flow equation of DistFlow, there are equality constraints of quadratic nonlinear terms, which will make the optimization problem non-convex. Therefore, the second-order cone relaxation method is used to convert equation (29) into a standard second-order cone constraint (30) for convexification.

[0172]

[0173] 2) Transmission constraints of tie lines

[0174]

[0175] Where P g,min , P g,max It is the upper and lower limits of active power transmitted from the upper power grid to the distribution network; Q g,min , Q g,max It is the upper and lower limits of reactive power transmitted from the upper power grid to the distribution network.

[0176] 3) Energy storage system operation constraints

[0177]

[0178] Where: is the energy storage capacity at time t; and Respectively represent the charging and discharging efficiency of energy storage; E ESS,min and E ESS,max are the maximum and minimum capacity of energy storage respectively; and They are the starting and ending power of energy storage respectively; and They are respectively the energy storage charging and discharging power at time t; and are the maximum charging and discharging power of energy storage respectively.

[0179] 4) Power balance constraints

[0180]

[0181] Example 2

[0182] This embodiment provides one of the implementation examples of the distribution network dispatching strategy method considering the division of vehicle-grid interaction modes as described above. The road network model is based on 26 road network nodes and 45 road data in a certain urban area, and combined with the IEEE 33-node distribution network system, a road network coupling model is established. Figure 6 As shown. Taking the typical daily load curve of the urban area as an example, the peak, valley and flat electricity price data are used. The detailed parameters are shown in Table 1. The rated voltage of the distribution network is 12.66kV, the rated capacity is 10MW, the upper and lower limits of the voltage are 0.95 and 1.05 (per unit value), and the maximum current that the branch can withstand is 500A. Photovoltaic is connected to the distribution network through nodes 8, 14, 17, 20, 23, and 26. The rated power of photovoltaic access at the cut-off point is 0.2MW. The power curves of photovoltaic power generation devices connected to different nodes are the same. The timing output diagram is shown in Figure 7 As shown in Table 2, the unit network loss cost of the distribution network is 1.0 yuan / (kW·h). The capacity electricity price is 0.0114 yuan / (kW·h). The energy storage power station is connected from nodes 6 and 12. The parameters are shown in Table 2.

[0183] Table 1 Time-of-use electricity price

[0184]

[0185] Table 2 Energy storage parameters

[0186]

[0187] The 1000 EVs in the area are taken as an example for research. The EV parameters are detailed in Table 3. Considering the EV load transfer problem, the initial power SOC of the EV is set to a normal distribution with an expectation of 0.8 and a variance of 1. The day-ahead scheduling cycle is set to 24 hours, with a time interval of 1 hour, divided into 24 time periods, and the real-time scheduling time interval is 15 minutes, divided into 96 time periods, and the cycle for EV to accept charging and discharging scheduling is 15 minutes.

[0188] Table 3 EV parameters

[0189]

[0190] Figure 7 The figure shows the power and load demand transformation of EVs in the two stages of the proposed method, the mean square error of daily load in the real-time scenario is reduced by 1579.7 compared with the day-ahead scenario, and the charging and discharging cost is reduced by 10084.67 yuan. This shows that through the two-stage optimization scheduling, EVs can participate more fully in the power system scheduling, which is more conducive to improving the stability and security of the power system and reducing the operating costs of the distribution network.

[0191] In order to further verify the proposed two-stage optimization scheduling strategy's ability to regulate the node voltage and network energy transmission loss of the urban distribution network, a comparison method is adopted to compare the node voltage and network loss of the distribution network before and after the proposed strategy is adopted. The node voltage comparison results are shown in Figure 2. Figure 8 As shown in the figure, it can be seen that before the proposed optimization strategy is adopted, the minimum voltage value of the distribution network node drops to 0.9, and there is an obvious under-limit, which has a huge impact on the stable operation of the distribution network; after the proposed strategy is adopted, the minimum voltage of the distribution network node only drops to 0.95, the node voltage transformation is within the safe range, and the node voltage deviation is improved. The network loss comparison results are shown in Fig. 9 As shown in the figure, it is obvious that the proposed method can mobilize the EVs connected to the grid to participate in the distribution network scheduling, effectively adjust the power flow distribution of the distribution network, reduce the network loss caused by the power flow transmission caused by the main grid power supply, and thus reduce the operation cost of the distribution network.

[0192] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A distribution network dispatching method considering the division of vehicle-grid interaction modes, characterized in that: The method steps include: Obtain the vehicle OD travel probability matrix, establish a road network model, and jointly correspond the road network nodes with the distribution network nodes to construct a road network-distribution network coupling model; Classify EV users into types based on travel modes, and construct travel characteristic models for corresponding user types; wherein the EV user classification types include: private car users, taxi users, and bus users; According to the classification of EV user types, based on the response modes of different EV users participating in different degrees of vehicle-grid interaction, respective response models are established; wherein the response modes include: full response EV, partial response EV and non-participation response EV; Based on the established model, taking into account the distributed photovoltaic and energy storage in the distribution network, with the goal of minimizing network loss cost, EV scheduling cost and bus voltage deviation, a two-stage optimal scheduling of the distribution network day-ahead and real-time operation considering the diversity of EV vehicle scheduling modes is constructed, and the real-time scheduling problem of the distribution network day-ahead and real-time operation is solved to obtain a real-time scheduling plan.

2. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The road network model in the road network-distribution network coupling model is as follows: Where: P(R) is the node in the road network, n i represents the i-th road node, k is the number of road network nodes, E(R) is the road section, L road is the road length matrix, V road is the road resistance coefficient matrix, l ij For node n i With node n j Length between, v ij is the direct path resistance coefficient between node i and node j; EV vehicles travel in the road network, enter traffic nodes according to the given optimal route to perform charging and discharging actions, and transmit charging load information from the traffic nodes to the corresponding distribution network nodes; The load information in the road network nodes is stored in the corresponding distribution network node load information matrix through the mapping function, realizing the coupling of urban road network-distribution network information.

3. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The travel characteristics of the private car users are as follows: The probability density function expression of the first trip time of a private car user on the day is as follows: Where: T s1 is the starting time of the first trip, α is the scale parameter, β is the first shape parameter, and μ is the second shape parameter; The time that private vehicles stay in residential and work areas follows a Gaussian distribution, and the time that they stay in commercial areas follows an exponential distribution.

4. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The travel characteristics of taxi users are as follows: The probability density function expression of the first trip time of a taxi user on the same day is as follows: Where: T s1 is the starting time of the first trip, α is the scale parameter, β is the first shape parameter, and μ is the second shape parameter; When the SOC of the taxi user's vehicle battery reaches the minimum charging threshold, the taxi user drives to the nearest charging station for charging. The docking time can be expressed as: Where: S car,n is the battery capacity of the nth vehicle, is the maximum capacity that the battery of the nth vehicle can accept, is the battery capacity when the nth vehicle joins the network, P ch It is the charging power of the charging pile per unit time.

5. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The bus users consider the fixed charging load characteristics of the bus station during non-business hours.

6. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 3, characterized in that: The charging and discharging model of the fully responsive EV is as follows; Where: is the capacity of the nth EV in period t, and are the charging and discharging power of the nth EV in period t, and They represent the charging and discharging efficiency of EV respectively; The vehicle dispatch cost is expressed as: Where: N1 represents the total number of EVs that can be fully dispatched, is the electricity price purchased by the distribution network from EV at time t, EV receives full dispatch subsidy per unit time.

7. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The partial response EV is used as a large-scale flexible load resource, and the charging and discharging model of EV is: The charging constraints when connected to a charging pile are: Through dispatching instructions, when the total EV charging load in a dispatching cycle remains unchanged, part of the charging load is transferred to other time periods within the parking time range; The running time before transfer is The acceptable time period after transfer is Then the dispatch cost of EV that can be used as flexible load is: Where: C tr is the dispatch cost per unit power transfer, is the load power of the nth EV transferred within the time period.

8. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: The non-response EV is a rigid load that is not subject to distribution network regulation. The charging model of accessing the charging station is the same as the flexible load charging model: The charging constraints are:

9. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 1, characterized in that: In the day-ahead phase of the two-stage optimization scheduling of the distribution network day-ahead and real-time operation, a multi-objective function of the distribution network day-ahead is constructed with the goal of minimizing the distribution network operation cost and the load fluctuation: Minimize the total cost of operating charging stations and urban distribution networks: In the formula, is the total operating cost of the distribution network in the day-ahead stage; C buy is the cost of power purchased by the distribution network from the upper power grid, C PV is the cost of abandoned rooftop photovoltaic power generation, C ESS is the operating cost of the energy storage power station, C loss is the network loss cost, F EV The total cost of dispatching EVs; The dispatch objects are the power purchased by the distribution network from the upper main grid, the output power of rooftop photovoltaic power plants, and the charging and discharging power of energy storage power stations; F EV =F EV，1 +F EV,2 Where: τ is the scheduling period; Δt is the scheduling time interval; λ t The real-time electricity purchase price; is the active power purchased by the distribution network from the upper power grid during period t; RL is the capacity price; is the capacity purchased by the distribution network from the upper power grid at time t; c PV Penalty for abandoning light for the unit;Ω PV It is the set of nodes connected to photovoltaic in the distribution network; is the maximum active power generated by the photovoltaic power plant at node i at time t; is the actual power of wind power and photovoltaic power connected to the grid at node i at time t; c ESS is the energy storage charging and discharging cost coefficient; and are the charging and discharging power of the energy storage in period t; J is the total number of distribution network nodes; r ij is the resistance of branch ij; U i,t is the voltage of node i at time t; P ij,t and Q ij,t is the active and reactive power flowing through branch ij at time t; C e Penalty for unit active network loss; Considering the grid base load and EV charging and discharging load, minimizing daily load fluctuations: Where: is the mean square error of the load in the day-ahead phase, indicating the load fluctuation of the distribution network; is the foundation load at time t; are the charging power and discharging power of the nth EV respectively; P av is the average daily load.

10. A distribution network dispatching method considering the division of vehicle-grid interaction modes according to claim 9, characterized in that: In the real-time stage of the two-stage optimization scheduling of the distribution network day-ahead and real-time operation, the distribution network scheduling optimization is only performed during the period when the EV response situation changes and the period thereafter, with the goal of minimizing the distribution network operation cost and the load fluctuation. The objective function is: Where: is the total operating cost of the distribution network in real time; is the mean square error of load in the real-time stage.

Images