A power distribution network dynamic reconstruction method containing an electric vehicle and an intelligent soft switch

Abstract

The application relates to a power distribution network dynamic reconstruction method containing electric vehicles and intelligent soft switches, and the method comprises the following steps: step S1: based on the time distribution function, the driving mileage distribution function and the initial SOC of the electric vehicle (EV) connected to the power grid, simulating an unordered charging model of the EV; step S2: for the unordered EV charging load, adopting a Lagrange decentralized optimization algorithm and a virtual electricity price to orderly schedule the EV load; step S3: establishing an SOP model under power distribution network reconstruction, taking the sum of the network loss cost, the SOP operation cost, the switch action cost and the wind and light abandonment loss cost as an objective function, transforming the nonlinear constraint by adopting a second-order cone relaxation, and constructing a mixed integer second-order cone programming model of the power distribution network reconstruction; and step S4: performing power distribution network dynamic reconstruction through the established model. The method is beneficial to improving the rationality and economy of power distribution network operation.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power distribution network, and particularly relates to a power distribution network dynamic reconstruction method containing electric vehicles and intelligent soft switches. BACKGROUND

[0002] With the gradual increase of the scale of DG and EV accessing the power distribution network, the existing power distribution network is difficult to cope with the problems of large operation loss and low operation economy of the power distribution network caused by the changes on the source-load side. Therefore, it is necessary to find a suitable method to orderly schedule the EV before accessing the power distribution network, so as to reduce the impact on the power distribution network. At the same time, it is also necessary to find a suitable method to optimize the power flow of the power distribution network and improve the utilization rate of renewable energy DG.

[0003] The access of EV charging load will increase the network loss of the power distribution network and cause the voltage drop of the terminal nodes of the power distribution network, and orderly scheduling the EV charging load before accessing the power distribution network can reduce the adverse effects on the operation of the power distribution network, thereby improving the economy of the operation of the power distribution network. The existing related researches mostly use centralized optimization algorithm for the orderly scheduling of EV load, which has the disadvantage of long calculation time. Virtual price is completely based on the curve characteristics of the basic load of the power distribution network to develop, and can flexibly develop dynamic price to guide the charging of EV load. The present application uses Lagrange decentralized optimization algorithm and virtual price to orderly schedule EV, which improves the calculation speed and also improves the effect of peak load shifting of EV load.

[0004] Power distribution network reconstruction is one of the ways to improve the operation economy of the power distribution network by changing the network topology. Power distribution network reconstruction is divided into static reconstruction and dynamic reconstruction. Static reconstruction is a study of a single time section, and the load is assumed to be constant, which has the disadvantage of being unable to cope with the dynamic changes of the load of the existing power distribution network, and has great limitations. Dynamic reconstruction is a study of a power distribution network operation cycle, and the load is dynamically changed in the operation cycle, so dynamic reconstruction is more suitable for the operation characteristics of the existing power distribution network.

[0005] In recent years, the development of power electronic devices has also provided new ideas for the operation optimization of the power distribution network. SOP, as a switch that can optimize the line power flow of the power distribution network and improve the flexibility of the power distribution network, has appeared in the research on the optimal operation of the power distribution network, but there are few studies on replacing the tie-line switch in power distribution network reconstruction with SOP. The present application will use SOP to replace the tie-line switch in power distribution network reconstruction, change the physical results of traditional power distribution network reconstruction, and study the influence of the access position and access quantity of SOP on the power flow optimization effect of the power distribution network and the economy of the power distribution network. SUMMARY

[0006] The application aims to provide a power distribution network dynamic reconstruction method containing electric vehicles and intelligent soft switches, which is beneficial to improve the rationality and economy of power distribution network operation.

[0007] To achieve the above-mentioned purpose, the technical solution adopted by the application is as follows: a power distribution network dynamic reconstruction method containing electric vehicles and intelligent soft switches, comprising the following steps:

[0008] Step S1: based on the time distribution function, the driving mileage distribution function and the initial SOC of the electric vehicle EV accessing the power grid, the disordered charging model of the EV is simulated;

[0009] Step S2: for the disordered EV charging load, the Lagrange decentralized optimization algorithm and the virtual electricity price are used for the orderly scheduling of the EV load;

[0010] Step S3: the SOP model under the power distribution network reconstruction is established, the sum of the network loss cost, the SOP operation cost, the switch action cost and the wind and light abandonment loss cost is taken as the objective function, the second-order cone relaxation is used for the transformation of the nonlinear constraint, and the mixed integer second-order cone programming model of the power distribution network reconstruction is constructed;

[0011] Step S4: the dynamic reconstruction of the power distribution network is carried out through the established model.

[0012] Further, in step S1, the method for accessing the power distribution network of the EV is determined according to the travel rule of the EV:

[0013] Step S11: the region of the power distribution network is functionally divided according to its distance from the transformer node, and is divided into commercial area, office area and residential area;

[0014] Step S12: it is assumed that in the morning, 80% of the EVs go from the residential area to the office area, 15% of the EVs go from the residential area to the commercial area, and 5% of the EV users continue to stay in the residential area; after work in the afternoon, 95% of the EVs return to the residential area, and 5% of the EVs continue to stay in the office area; the load distribution problem of the EV accessing the power distribution network all follows this rule.

[0015] Further, in step S2, the process of orderly scheduling of the disordered EV charging load using the Lagrange decentralized optimization algorithm and the virtual electricity price is as follows:

[0016] Step S21: first, the virtual electricity price is formulated according to the basic load information of the power distribution network; the scheduling period of the virtual electricity price is T=24h, the scheduling time interval is ΔT=0.25h, and the formulation of the virtual electricity price is represented by formula (1):

[0017]

[0018] In the formula, γ(P l,a(t), t) is the virtual electricity price at time t; is the formulation coefficient of virtual electricity price; + represents max{0, μ}; l,a P (t) represents the total load of power grid at time t, l,a P (t) = P l,bc P (t) = P l,EV (t) + P e (t) = P

[0019] Step S22: orderly scheduling of EV based on the simulated EV disorderly charging load of EV travel habits; first, a centralized optimization scheduling model is established, and the objective function is the minimum charging cost F of EV:

[0020]

[0021] In the formula: N is the total number of EVs; P e represents the charging power; is the charging state of the ith EV at time t:

[0022]

[0023] Step S23: setting the constraints of EV orderly scheduling;

[0024] 1) EV user's charging demand constraint

[0025]

[0026] In the formula: S e,i is the expected SOC of the ith EV; η is the charging efficiency of the EV; C c is the battery capacity of the EV;

[0027] 2) Time constraint of EV accepting scheduling

[0028] Suppose that the EV is only scheduled by the virtual electricity price in the period of connecting to the power grid, so:

[0029]

[0030] In the formula: t = 1, 2,..., t c1,i ,t l,i +1,..., t c2,i ; i = 1, 2,..., N; t c1,i , t c2,i represents the time when the ith EV connects to the power grid for the first and second time; t l,i represents the time when the ith EV is disconnected from the power grid;

[0031] 3) System power constraint

[0032] At any moment, the total load value is less than the upper limit of the load that the distribution network can bear:

[0033]

[0034] In the formula: P l,m is the upper limit of the load that the distribution network can bear;

[0035] Step S24: Steps S22 and S23 establish the centralized optimization solution model. If it is changed to a Lagrangian relaxation decentralized optimization model, the systematic constraint (6) needs to be added to the objective function as a penalty term. Thus, the Lagrangian relaxation problem of the original problem without the systematic constraint is obtained, as shown in formula (7):

[0036]

[0037] The above formula is further transformed to obtain the sub-problem of each EV, as shown in formula (8):

[0038]

[0039] The dual problem of the original problem is obtained from the Lagrangian relaxation problem In the dual problem, λ is a variable; the value of λ obtained from formula (8) is substituted into the original and dual problems. If the difference between the objective functions of the original and dual problems is within the range allowed by the precision gap, the original problem can be solved by the dual problem. If the difference is not within the range allowed by the precision gap, λ is updated.

[0040] Step S25: If the difference between the objective functions obtained by the original and dual problems is not within the range allowed by the precision gap, the sub-gradient method is used to update λ. The sub-gradient method is to assign an initial value to λ when k=1, and then iterate along the direction of the sub-gradient. K represents the iteration number, and the iteration formula is shown in formula (9):

[0041]

[0042] In the formula: m and n are constants, χ(k)=1 / (m+n·k) is the iteration step size; is the sub-gradient, and λ is updated until the dual gap precision meets the requirements.

[0043] Further, in step S3, the second-order cone relaxation is used to transform the nonlinear constraint, and the process of constructing the mixed integer second-order cone programming model of the distribution network reconstruction is as follows:

[0044] ​Step S31: The power distribution network dynamic reconstruction model is configured to minimize the sum of the network loss cost, the SOP operation cost, the switch action cost and the wind and light abandonment loss cost, and is expressed by formula (10):

[0045]

[0046] In the formula, B line is a branch set of the power distribution network, C loss is a network loss coefficient, is a current on the line ij at time t, r ij is a resistance of the line ij; B sop is a SOP access node set, C soploss is a SOP operation cost coefficient, is an active power flowing through the SOP access node at time t; C switch is a switch action cost coefficient, is a line switch state at time t, is 0 / 1 indicating that the line is disconnected and closed; B wind , B pv respectively represent a wind access node set and a light access node set, respectively represent a maximum wind output and a maximum light output, respectively represent a wind active power actually used by the power distribution network and a light active power actually used by the power distribution network;

[0047] Step S32: After setting the objective function, a DistFlow power flow model of the power distribution network dynamic reconstruction is constructed:

[0048]

[0049]

[0050]

[0051]

[0052]

[0053]

[0054] In the formula, f(j), s(j) respectively are an upper branch node set and a lower branch node set of node j, respectively are active power and reactive power flowing through the line ij at time t, x ij is a reactance of the line ij, respectively are active power and reactive power injected into node j at time t, respectively are active power and reactive power of charging and discharging of the energy storage at node j at time t, Pj(t) represents all active loads at node j at time t, including the base load and EV load of the distribution network node, Qj(t) represents the reactive power flowing through the SOP access node at time t, Qj(t) represents all reactive loads at node j at time t, and Vi(t) represents the voltage at node i;

[0055] Step S33: the model of the SOP is established as follows:

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] In the formula: Pj(t) represents all active loads at node j at time t, including the base load and EV load of the distribution network node, Qj(t) represents the reactive power flowing through the SOP access node at time t, Qj(t) represents all reactive loads at node j at time t, and

[0062] Step S34: the nonlinear constraints of formula (11), (12), (15) and (16) are removed by letting to eliminate the square term in the formula; for formula (16), the second-order cone relaxation is adopted to make it become constraint (23); in order to make constraint (15) true for both the disconnected and closed branches, the large M method relaxation is adopted for transformation, so that constraint (15) and (16) are replaced by constraint (22) and (23);

[0063]

[0064]

[0065] In the formula: M is a very large positive number;

[0066] For nonlinear constraints formula (18), (19), (20) and (21), the second-order cone relaxation is also adopted to linearize the solution, and the relaxed formula is represented by (24), (25), (26) and (27):

[0067]

[0068]

[0069]

[0070]

[0071] Compared with the prior art, the present application has the following beneficial effects:

[0072] 1) The use of the Lagrange distributed optimization algorithm and virtual electricity price for EV load proposed in the present application can effectively realize the orderly charging of EV load, improve the effect of peak load shifting of EV load, and on the other hand, can reduce the electricity cost of the vehicle owner on the premise of meeting the electricity demand of the vehicle owner;

[0073] 2) The use of SOP instead of the tie line switch of the traditional power distribution network reconstruction can improve the economy of the power distribution network operation and the voltage amplitude of the power distribution network node, and the access of different numbers of SOPs has different effects on the improvement of the economy of the power distribution network;

[0074] 3) In the research of power distribution network reconstruction operation, the orderly scheduling of SOP and EV is considered, which can effectively reduce the operation cost of the power distribution network and improve the rationality and economy of the power distribution network operation. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 is the optimization scheduling flowchart of the method of the present application;

[0076] Figure 2 is the curve graph of the base load of the power distribution network superimposed with the unordered EV charging load in the embodiment of the present application;

[0077] Figure 3 is the curve graph of the base load of the power distribution network superimposed with the ordered EV charging load in the embodiment of the present application;

[0078] Figure 4 is the power distribution network region division graph in the embodiment of the present application;

[0079] Figure 5 is the node voltage comparison graph of the power distribution network in the embodiment of the present application. DETAILED DESCRIPTION

[0080] The present application will be further described below in combination with the drawings and embodiments.

[0081] It should be pointed out that the following detailed description is exemplary and is intended to provide further description of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as generally understood by those skilled in the art to which the present application belongs.

[0082] It is to be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, steps, operations, elements, components, and / or groups thereof, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0083] The embodiment provides a power distribution network dynamic reconstruction method containing an electric vehicle and an intelligent soft switch, and comprises the following steps:

[0084] Step S1: based on the time distribution function, the driving distance distribution function and the initial SOC of the electric vehicle (EV) connected to the power grid, a disordered charging model of the EV is simulated, and a method for connecting the EV to the power distribution network is determined.

[0085] In step S1, the method for connecting the EV to the power distribution network is determined according to the travel rule of the EV:

[0086] Step S11: the regions of the power distribution network are functionally divided according to their distances from the transformer node, and are divided into a commercial area, an office area and a residential area.

[0087] Step S12: the EV load is considered to be distributed. According to the travel habits of office workers, it can be reasonably assumed that 80% of the EVs go from the residential area to the office area in the morning, 15% of the EVs go from the residential area to the commercial area, and 5% of the EV users continue to stay in the residential area; after work in the afternoon, 95% of the EVs return to the residential area, and 5% of the EVs continue to stay in the office area; the load distribution problem of the EV connected to the power distribution network all follows this rule.

[0088] Step S2: the disordered EV charging load is orderly scheduled by using a Lagrange decentralized optimization algorithm and a virtual price.

[0089] In step S2, the process of orderly scheduling the disordered EV charging load by using the Lagrange decentralized optimization algorithm and the virtual price is as follows:

[0090] Step S21: first, the virtual price is formulated according to the basic load information of the power distribution network; the scheduling period of the virtual price is T=24h, the scheduling time interval is ΔT=0.25h, and the formulation of the virtual price is represented by formula (1):

[0091]

[0092] In the formula, γ (P l,a (t),t) is the virtual price at time t; is the formulation coefficient of the virtual price; [μ] + represents max{0, μ}; Pl,a (t) represents the total load of the power grid at time t, P l,a (t) = P l,bc (t) + P l,EV (t) ;

[0093] Step S22: orderly scheduling of EVs based on the simulated EV disorderly charging load according to the EV travel habits; first, a centralized optimization scheduling model is established, and the objective function is the minimum charging cost F of the EVs:

[0094]

[0095] In the formula, N is the total number of EVs; P e represents the charging power; is the charging state of the ith EV at time t:

[0096]

[0097] Step S23: setting the constraints of the orderly scheduling of EVs.

[0098] 1) charging demand constraint of EV users

[0099]

[0100] In the formula, S e,i is the expected SOC of the ith EV; η is the charging efficiency of the EV; C c is the battery capacity of the EV.

[0101] 2) time constraint of EVs accepting scheduling

[0102] It is assumed that the EVs are only scheduled by the virtual electricity price in the period of connecting to the power grid, so:

[0103]

[0104] In the formula, t = 1, 2,..., t c1,i ,t l,i + 1,..., t c2,i ; i = 1, 2,..., N; t c1,i , t c2,i represent the time when the ith EV connects to the power grid for the first time and the second time; t l,i represents the time when the ith EV is disconnected from the power grid.

[0105] 3) system power constraint

[0106] At any time, the total load value is less than the upper limit of the load that the power grid can bear:

[0107]

[0108] P = Pmax - Pmin l,m Pmax is the upper limit of the load that the power distribution network can bear.

[0109] Step S24: After the centralized optimization solution model is established in steps S22 and S23, if the Lagrange relaxation decentralized optimization model is to be changed, the systematic constraint (6) needs to be added as a penalty term to the objective function, and thus, the Lagrange relaxation problem of the original problem without the systematic constraint is obtained, as shown in formula (7):

[0110]

[0111] The sub-problem of each EV is obtained by further transforming the above formula, as shown in formula (8):

[0112]

[0113] The dual problem of the original problem is obtained from the Lagrange relaxation problem In the dual problem, λ is a variable; the value of λ obtained from formula (8) is substituted into the original and dual problems. If the difference between the objective functions of the original and dual problems is within the range allowed by the accuracy gap, the original problem can be solved by using the dual problem; if the difference is not within the range allowed by the accuracy gap, λ is updated.

[0114] Step S25: If the difference between the objective functions obtained by the original and dual problems is not within the range allowed by the accuracy gap, the sub-gradient method is used to update λ; the sub-gradient method is to assign an initial value to λ when k=1, and then iterate along the direction of the sub-gradient, where k represents the iteration number, and the iteration formula is shown in formula (9):

[0115]

[0116] In the formula, m and n are constants, χ(k)=1 / (m+n·k) is the iteration step size; is the sub-gradient, and λ is updated until the dual gap accuracy meets the requirements.

[0117] Step S3: The SOP model under the power distribution network reconstruction is established, the sum of the network loss cost, the SOP operation cost, the switch action cost, and the wind and light abandonment loss cost is taken as the objective function, the second-order cone relaxation is used to transform the nonlinear constraint, and the mixed integer second-order cone programming model of the power distribution network reconstruction is constructed.

[0118] In step S3, the process of using the second-order cone relaxation to transform the nonlinear constraint and constructing the mixed integer second-order cone programming model of the power distribution network reconstruction is as follows:

[0119] Step S31: The power distribution network dynamic reconstruction model is configured to minimize the sum of the network loss cost, the SOP operation cost, the switch action cost and the wind and light abandonment loss cost, which is expressed by formula (10):

[0120]

[0121] In the formula, B line is a branch set of the power distribution network, C loss is a network loss coefficient, is a current on the line ij at time t, r ij is a resistance of the line ij; B sop is a SOP access node set, C soploss is a SOP operation cost coefficient, is an active power flowing through the SOP access node at time t; C switch is a switch action cost coefficient, is a line switch state at time t, is 0 / 1 indicating that the line is disconnected and closed; B wind , B pv respectively represent a wind and light access node set, respectively represent a wind and light maximum output, respectively represent a wind and light active power actually used by the power distribution network;

[0122] Step S32: After setting the objective function, a DistFlow power flow model of the power distribution network dynamic reconstruction is constructed:

[0123]

[0124]

[0125]

[0126]

[0127]

[0128]

[0129] In the formula, f(j), s(j) respectively represent an upper branch node set and a lower branch node set of node j, respectively represent an active power and a reactive power flowing through the line ij at time t, x ij is a reactance of the line ij, respectively represent an active power and a reactive power injected into node j at time t, respectively represent an active power and a reactive power of the energy storage charging and discharging at node j at time t, Pj(t) represents all active loads at node j at time t, including the base load and EV load of the distribution network node, Qj(t) represents the reactive power flowing through the SOP access node at time t, Qj(t) represents all reactive loads at node j at time t, and Vi(t) represents the voltage at node i;

[0130] Step S33: the model of the SOP is established as follows:

[0131]

[0132]

[0133]

[0134]

[0135]

[0136] In the formula: Pj(t) represents all active loads at node j at time t, including the base load and EV load of the distribution network node, Pj(t) represents the reactive power flowing through the SOP access node at time t, Qj(t) represents all reactive loads at node j at time t, and

[0137] Step S34: the nonlinear constraints of formula (11), (12), (15) and (16) are removed by letting to eliminate the square term in the formula; for formula (16), the second-order cone relaxation is adopted to make it become constraint (23); in order to make constraint (15) true for both the disconnected and closed branch, the large M method relaxation is adopted for transformation, so that constraint (15) and (16) are replaced by constraint (22) and (23);

[0138]

[0139]

[0140] In the formula: M is a very large positive number;

[0141] For nonlinear constraints formula (18), (19), (20) and (21), the second-order cone relaxation is also adopted to linearize the solution, and the relaxed formula is represented by (24), (25), (26) and (27):

[0142]

[0143]

[0144]

[0145]

[0146] The application also provides a radial constraint, a switch action frequency constraint, a voltage and current constraint, a distributed power output constraint and an energy storage constraint.

[0147] The optimal scheduling method of the application can be represented by an optimal scheduling flowchart as shown in the figure. Figure 1

[0148] Step S4: dynamic reconstruction of the power distribution network through the established model.

[0149] To verify the effect of the application on the improvement of the economic efficiency of the power distribution network reconstruction operation, the example simulation is performed on the IEEE33 node system. The control group is selected as the traditional power distribution network reconstruction operation without using the method of the application. The comparison verification method process is as follows:

[0150] (1) To verify the effectiveness of the EV ordered scheduling method of the application combining the Lagrange relaxation decentralized algorithm and the virtual price, the power distribution network superimposed unordered load curve is compared with the power distribution network superimposed ordered load curve to verify the peak load shifting effect of the EV ordered scheduling method.

[0151] (2) To verify the effect of the application on the improvement of the economic efficiency of the power distribution network dynamic reconstruction, the simulation analysis is performed on the control group and the experimental group using the method of the application.

[0152] The MISOCP model of the power distribution network is solved by CPLEX on MATLAB. The rated voltage of the power distribution network is 12.66kV, the base capacity is 100MW, node 0 is the substation node, and the voltage unit value is 1.0. The capacity of the SOP is 1MW, and the loss coefficient is 0.02. The coefficients of the objective function in the reconstruction model are set as follows: the network loss coefficient C loss , the SOP operation cost coefficient C soploss , the abandoned wind and light cost coefficient C windloss and C pvloss are all set as 400 yuan / (MW·h), and the switch action cost coefficient C switch is set as 2 yuan / time.

[0153] ​For EV load, the present application assumes that there are 100 EVs participating in the dispatch cycle, and that the charging power of a single EV is 10 kW, the battery capacity is 30 kW·h, and the charging efficiency is 90%. The curve of the simulated EV unordered charging load superimposed with the base load of the power distribution network is shown in Figure 2 . After the EV ordered scheduling using the Lagrange decentralized optimization algorithm and the virtual electricity price proposed by the present application, the curve of the EV ordered charging load superimposed with the base load of the power distribution network is shown in Figure 3 . Comparing Figure 2 and Figure 3 , it can be seen that the EV ordered scheduling method proposed by the present application can not only meet the electricity demand of EV owners, but also realize peak load shifting of EV load and reduce the adverse effects of EV load on the power distribution network.

[0154] A reasonable EV load distribution method is used to connect the EV load to the power distribution network. First, the power distribution network is divided into commercial, office and residential areas, and charging station nodes are set for each of the three areas. The power distribution network area division of the present application is shown in Figure 4 . Then, according to the travel habits of EV owners, it is assumed that 80% of EVs will go from the residential area to the office area in the morning, 15% of EVs will go from the residential area to the commercial area, and 5% of EV users will continue to stay in the residential area. In the evening, 95% of EVs will return to the residential area, and 5% of EVs will continue to stay in the office area. After the distribution and connection of EV load are processed, the influence of the coordinated scheduling of EV and SOP on the economic efficiency of power distribution network reconfiguration can be studied. The present application sets six scenarios to study the influence of the coordinated scheduling of EV and SOP on the economic efficiency of power distribution network reconfiguration under 40% penetration rate.

[0155] Scenario 1: power distribution network reconfiguration without SOP;

[0156] Scenario 2: unordered EV connection and power distribution network reconfiguration without SOP;

[0157] Scenario 3: unordered EV connection and power distribution network reconfiguration with SOP connected to line L 7-20 ;

[0158] Scenario 4: ordered EV connection and power distribution network reconfiguration without SOP;

[0159] Scenario 5: ordered EV connection and power distribution network reconfiguration with SOP connected to line L 7-20 ;

[0160] Scenario 6: ordered EV and line L 7-20 and L 11-21 each connected to a SOP power distribution network reconfiguration.

[0161] Table 1. Power distribution network reconfiguration operation cost under multiple scenarios of coordinated optimization of EV and SOP

[0162]

[0163] As can be seen from Table 1, the six scenarios DG can be completely absorbed. At the same time, the access of EV will increase the network loss of the power distribution network, and the access of SOP can effectively reduce the network loss, thereby reducing the total cost of the operation of the power distribution network. The ordered scheduling of EV and the cooperation of SOP can further improve the economy of the operation of the power distribution network.

[0164] In order to verify the effect of SOP on the voltage of each node of the power distribution network, the reconstruction condition of the power distribution network without the access of SOP is set as a control group, and the experimental group is five groups, which are respectively the reconstruction conditions of the power distribution network with the access of five tie lines to a single SOP. The effect of SOP on the voltage of each node of the power distribution network is shown in Table 2. Figure 5 .

[0165] In summary, the power distribution network dynamic reconstruction method containing an electric vehicle and a smart soft switch provided by the present application can improve the peak load shifting effect of the access of EV load after the ordered scheduling of EV by using the Lagrange decentralized algorithm and the virtual price provided by the present application. At the same time, the ordered scheduling method of EV and the cooperation of SOP can effectively improve the rationality and economy of the operation of the power distribution network.

[0166] The above description is only the preferred embodiment of the present application, and is not intended to limit the other forms of the present application. Any person skilled in the art can use the disclosed technical content to make changes or modifications into equivalent embodiments. However, any simple modification, equivalent change and modification made according to the technical essence of the present application without departing from the technical solution of the present application still belongs to the protection scope of the present application.

Claims

1. A method for dynamic reconfiguration of a distribution network incorporating electric vehicles and intelligent soft switches, characterized in that, Includes the following steps: Step S1: Simulate the disordered charging model of EV based on the time distribution function, driving range distribution function and initial SOC of the electric vehicle (EV) connected to the power grid; Step S2: For disordered EV charging loads, the Lagrange distributed optimization algorithm and virtual electricity price are used to schedule the EV load in an orderly manner. Step S3: Establish the SOP model under distribution network reconfiguration, with the objective function being to minimize the sum of network loss cost, SOP operation cost, switch operation cost, and wind and solar curtailment loss cost. Second-order cone relaxation is used to transform the nonlinear constraints, and a mixed integer second-order cone programming model for distribution network reconfiguration is constructed. Step S4: Perform dynamic reconfiguration of the distribution network using the established model; In step S2, the process of orderly scheduling of EV charging load using the Lagrange distributed optimization algorithm and virtual electricity pricing for disordered EV charging load includes: Step S21: First, formulate the virtual electricity price based on the basic load information of the distribution network; assuming the scheduling cycle of the virtual electricity price is T = 24h and the scheduling time interval ΔT = 0.25h, the formulation of the virtual electricity price is expressed by formula (1): In the formula: γ(P) l,a (t),t) represents the virtual electricity price at time t; [μ] is the coefficient for setting virtual electricity prices. + P represents max{0,μ}; l,a (t) represents the total load of the distribution network at time t, P l,a (t)=P l,bc (t)+P l,EV (t); Step S22: Perform ordered scheduling of EVs based on the simulated disordered charging load of EVs according to EV travel habits; first, establish a centralized optimization scheduling model, whose objective function is to minimize the charging cost F of EVs. In the formula: N is the total number of EVs; P e Indicates charging power; The charging state of the i-th EV at time t: Step S23: Set constraints for ordered scheduling of EVs; 1) Charging demand constraints of EV users In the formula: S e,i Let be the expected SOC of the i-th EV; η be the charging efficiency of the EV; C c For EV battery capacity; 2) Time constraints for EVs to accept scheduling Assume that the EV is only subject to virtual electricity pricing during the period it is connected to the grid, then we have: In the formula: t=1,2,...,t c1,i ,t l,i +1,...,t c2,i i = 1, 2, ... N; t c1,i t c2,i Indicates the times when the i-th EV connects to the grid for the first and second time; t l,i This indicates the moment when the i-th EV is disconnected from the power grid; 3) System power constraints At any given moment, the total load must be less than the maximum load that the distribution network can handle. In the formula: P l,m This is the upper limit of the load that the distribution network can withstand.

2. The method for dynamic reconfiguration of a distribution network including electric vehicles and intelligent soft switches according to claim 1, characterized in that, In step S1, the method for determining the EV's connection to the power distribution network is based on its travel patterns: Step S11: Divide the distribution network area into commercial area, office area and residential area according to its distance from the substation node; Step S12: Assume that in the morning, 80% of EVs travel from residential areas to office areas, 15% of EVs travel from residential areas to commercial areas, and 5% of EV users remain in residential areas; after get off work in the afternoon, 95% of EVs will return to residential areas, and 5% of EVs will remain in office areas; the load distribution problem when EVs are connected to the power distribution network follows this rule.

3. The method for dynamic reconfiguration of a distribution network including electric vehicles and intelligent soft switches according to claim 1, characterized in that, After completing steps S21-S23, perform the following steps: Step S24: Steps S22 and S23 established a centralized optimization solution model. If it is to be changed to a Lagrange relaxation decentralized optimization model, the system power constraint needs to be added as a penalty term to the objective function. Thus, the Lagrange relaxation problem of the original problem without system constraints is obtained, as shown in Equation (7): Further transformation of the above equation yields the subproblem for each EV, as shown in equation (8): The dual problem derived from the Lagrange relaxation problem. In the dual problem, λ is a variable; the result obtained from equation (8) Substituting into the primal and dual problems, if the difference between the objective functions of the primal and dual problems is within the allowable range of the accuracy gap, then the primal problem can be solved using the dual problem; if the difference is not within the allowable range of the accuracy gap, then λ is updated. Step S25: If the difference between the objective functions obtained from the primal and dual problems is not within the allowable range of the accuracy gap, then the subgradient method is used to update λ. The subgradient method first assigns an initial value to λ when k=1, and then iterates along the direction of the subgradient, where k represents the number of iterations. The iteration formula is shown in equation (9): In the formula: m and n are constants, and χ(k) = 1 / (m + n·k) is the iteration step size; It is the subgradient, and Update λ until the dual gap accuracy meets the requirements.

Images