Layered and distributed charging and discharging scheduling method and system for electric power system containing electric vehicle

Abstract

The invention discloses a layered and distributed charging and discharging scheduling method and system for an electric power system containing an electric vehicle. The method comprises the following steps: establishing a day-ahead load scheduling model containing the electric vehicle; polymerizing the electric automobile, and introducing into an electric automobile aggregate layer; on the basis of an electric vehicle aggregate layer, generating an aggregate charging and discharging power scheduling instruction according to the upper layer part of the established day-ahead load scheduling model; the lower layer part of the established day-ahead load scheduling model comprises an aggregate layer and a user layer, and the electric vehicle aggregate layer responds to a scheduling instruction and issues a control signal to the user layer, so that distributed solution of charging and discharging scheduling is realized, and a user charging and discharging control result is obtained. The system comprises a model establishment module, an aggregation introduction module, a scheduling instruction generation module and a solving module. According to the method, the defects that aggregation is irreversible and aggregate power solutions are possibly infeasible on individuals in traditional electric vehicle scheduling can be overcome, user privacy can be well protected, and the problem solving and information storage pressure can be effectively relieved.

Description

Technical Field

[0001] The present invention belongs to the field of power system operation, and in particular relates to a hierarchical distributed charging and discharging scheduling method and system for a power system containing electric vehicles. Background Art

[0002] The electrification of terminal energy consumption is an important part of the energy consumption revolution, and replacing traditional fuel vehicles with electric vehicles is the main way to promote the electrification process at this stage. The transportation system consumes about one-third of the world's energy, so there is great potential for the development of electric vehicles and the use of electricity instead of oil. In addition to the huge number of new energy vehicles, electric vehicles are also a very representative type of load controllable resources in terms of nature. It can not only flexibly adjust the power consumption time according to the user's travel plan, but also discharge to the power grid when necessary and provide power support for the power grid. It also has a certain energy storage capacity. Therefore, in-depth exploration of the controllable characteristics of electric vehicles in the power grid is of great significance to the safe, economical and stable operation of the power system.

[0003] However, the large number of electric vehicles and the numerous constraints have brought certain difficulties and challenges to the source-load coordination at the regional power grid level. Aggregating the constraints of a single vehicle and modeling the electric vehicle cluster is an intuitive way to deal with it, which can greatly reduce the number of constraints and control the scale of the problem. It has also been widely used in existing studies. However, although the variable and constraint aggregation method provides a way to solve the problem, it still has many limitations: 1) It is difficult to deal with the discrete technical parameters of electric vehicles in reality (such as charging and discharging efficiency). Considering time-varying factors such as user travel, the equivalent average value of the parameters is also difficult to determine; 2) The aggregation method of some constraints is not accurate, and it is based on the premise of sacrificing some control flexibility; 3) It is impossible to avoid cross-charging, and only the scheduling plan at the cluster level can be obtained, and it is difficult to restore the results to each electric vehicle.

[0004] The main difficulty faced by load dispatching is how to consider more detailed and specific load constraints in the wide-area dispatching problem, and to maximize the feasibility of the load dispatching scheme while achieving large-scale source-load coordination. At the regional power system level, compared with conventional generators, load-side controllable resources such as dispatchable loads, electric vehicles, and distributed generation have a smaller capacity scale, but are large in number, widely distributed, and complex in characteristics, making it difficult to consider them uniformly under a centralized framework. At the same time, the overall scale of load-side controllable resources is huge, and its operation strategy will have a significant impact on the operation of the power grid and power generation resources.

[0005] In summary, how to achieve the coordination of resources on the load side and the power generation side, as well as the coordination within the load side resources, is an important issue to be solved in load dispatching, which is related to the economy of system operation and the feasibility of load regulation. In addition, in the existing research on cluster modeling and cluster-level charging and discharging control, users usually need to report vehicle technical parameters and travel plans to the cluster so that the dispatching agency can construct the model and implement the dispatching. This will bring a certain degree of threat and harm to the privacy protection of users, which should be avoided through the design of appropriate charging and discharging dispatching mechanism. Summary of the invention

[0006] The purpose of the present invention is to provide a hierarchical distributed charging and discharging scheduling method and system for a power system containing electric vehicles based on the current development status of the power system, targeting the massive load-side controllable resources represented by electric vehicles in the power grid, taking into account both source-load coordination and internal load coordination, and taking into account the privacy and security of users in load scheduling, so as to further study the operation control issues of the power grid.

[0007] In order to achieve the above object, the present invention adopts the following technical solution: A hierarchical distributed charging and discharging scheduling method for a power system including electric vehicles comprises the following steps: Step 1: Establish a day-ahead load dispatch model including electric vehicles; Step 2: To facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into the electric vehicle aggregation layer; Step 3: Based on the electric vehicle cluster layer in step 2, according to the upper part of the day-ahead load dispatch model established in step 1, the upper layer is the system dispatch layer, so as to achieve the coordination of the power system source, grid and load, and generate the dispatch instructions of the cluster charging and discharging power; Step 4: The lower layer of the established day-ahead load dispatch model includes the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the dispatch instruction and issues a control signal to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging dispatch, and obtain the user charging and discharging control results.

[0008] A further improvement of the present invention is that in step 2, in order to facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into an electric vehicle aggregation layer, including: By aggregating electric vehicles within a certain range, where a certain range is a series of electric vehicles within a charging station or geographically close, the segmentation characteristics of the problem are further clarified: (1) (2) In the formula, for Moment Electric Vehicle Aggregate The power drawn from the grid, For electric vehicles exist The power drawn from the grid during the period, They are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively.

[0009] A further improvement of the present invention is that, in step 3, based on the electric vehicle aggregate layer in step 2, according to the day-ahead load dispatch model established in step 1, the coordination of the power system source, grid and load is realized at the system dispatch layer, and a dispatch instruction for aggregate charging and discharging power is generated, including: Take aggregate power In order to link variables, the day-ahead load dispatch model is transformed into a two-layer problem, where the upper main problem is: (3) The constraints are: (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) In the formula, Generator sets For The upward and downward spare capacity available during the period, Generator sets The maximum and minimum output, Generator sets Minimum continuous downtime and start-up time, Generator sets Upward and downward climbing rates, is the time interval length; The system is The upward and downward spare capacity requirements that can be provided during the period, For Node exist Fixed load other than controllable load during the period, For electric vehicles exist The net power drawn from the grid during the period, is the line number, is the DC power flow distribution factor, For line Active transmission capacity, For Node The collection of electric vehicles connected to the grid, where for Moment Electric Vehicle Aggregate The power drawn from the grid, are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively; where (4)-(6) are the maximum and minimum output constraints of the unit, (7)-(10) are the continuous start-stop time constraints of the unit, (11)-(12) are the ramp constraints of the unit, (13)-(14) are the reserve capacity constraints, (15) is the power balance constraint, (16) is the active transmission capacity constraint of the transmission line, and (17) is the cluster variable constraint.

[0010] A further improvement of the present invention is that in step 4, the lower-level problem SP can be constructed as: The obtained aggregate power is feasibility checked and adjusted according to the check results. The underlying sub-problems shown below are constructed to check the feasibility of the aggregate charging and discharging scheme: (18) In the formula, is the power deviation between the upper aggregate power scheduling scheme and the feasible scheme, is a non-negative slack variable, is the dual multiplier of the corresponding constraint, Electric vehicles exist The charging and discharging power during the period, Electric vehicles Maximum charging and discharging power, is a 0-1 variable, indicating the grid-connected status of the electric vehicle. For electric vehicles exist Energy storage level at all times, Electric vehicles Maximum and minimum energy storage levels, Electric vehicles Charging and discharging efficiency, is a parameter, indicating that electric vehicles exist The energy consumed during driving.

[0011] A further improvement of the present invention is that in step 3 and step 4, the obtained two-layer problem is solved hierarchically using the Benders decomposition method, and the generated Benders cut is in the form of: (19).

[0012] A further improvement of the present invention is that, in step 4, the underlying aggregate subproblem is improved into the following form; since the square of the deviation is non-negative, a set of slack variables is introduced; (20) In the formula, To improve the objective function value of the subproblem, is a non-negative slack variable; The optimization problem of formula (20) can be further written as: (twenty one) In formula (21), the constraints include the EV charging and discharging power constraints, which are separated according to the vehicle, and the separated EV variables are only coupled through the squared objective function; When the objective function of problems (20) and (21) is not 0, construct the Benders cut to return the upper objective function: (twenty two) (twenty three) When problem (18) is rewritten as (20), the feasible domain does not change, but problem (20) is non-convex and does not satisfy the Slater constraint specification conditions. Therefore, the method of forming the Benders cut is not strict. The cut defined by equations (22)-(23) is called the approximate Benders cut.

[0013] A further improvement of the present invention is that, in step 4, the user privacy issue and the storage solution pressure caused by the increase in the number of electric vehicles in the cluster are considered, the form of the cluster sub-problem is improved, and finally a three-layer control mechanism and an electric vehicle charging and discharging load scheduling framework with underlying support for distributed computing are formed; The upper layer is system scheduling; The middle layer is the electric vehicle cluster that aggregates the charging and discharging power of each electric vehicle. The following formula is used to calculate the The wheel charging and discharging control signal is issued to the electric vehicles under its jurisdiction: (twenty four) In the formula, Electric vehicle aggregates The charge and discharge control parameters, For electric vehicles In the The charging control scheme reported by the round, For aggregate The charging and discharging power instructions obtained from system scheduling, For aggregate Released Wheel charging control signal; Each electric vehicle in the lower layer solves the following single vehicle charging control problem, and the solution is the first Cycle charge and discharge plan , report to the group to which it belongs: (25) When solving the subproblem, the variable of the electric vehicle charging and discharging plan under the electric vehicle aggregation calculation is used as the iterative convergence criterion: (26) when If the time is small enough, the iteration is considered to have converged, the cluster sends a confirmation signal to the electric vehicles under its jurisdiction, and calculates the charging and discharging power of the cluster to form an approximate directional system scheduling report.

[0014] The hierarchical distributed charging and discharging dispatching system of the power system including electric vehicles includes: Model building module, building a day-ahead load dispatch model including electric vehicles; Aggregation introduction module aggregates electric vehicles and introduces them into the electric vehicle aggregation layer to facilitate problem solving and system scheduling; The dispatch instruction generation module, based on the electric vehicle aggregation layer of the aggregation introduction module, generates the dispatch instruction of the aggregate charging and discharging power according to the upper part of the day-ahead load dispatch model established in the model establishment module, the upper layer is the system dispatch layer, so as to realize the coordination of the power system source, grid and load; The solution module establishes the lower layer of the day-ahead load scheduling model, including the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the scheduling instructions and issues control signals to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging scheduling, and obtain the user charging and discharging control results.

[0015] Compared with the prior art, the present invention has at least the following beneficial technical effects: 1) A hierarchical model is proposed to solve the problem that the aggregate power of electric vehicles obtained in the traditional model cannot be implemented to each electric vehicle. In the model of the present invention, large-scale electric vehicles will not significantly increase the scale of the problem or affect the feasibility of actual charging and discharging; 2) A distributed solution method is adopted, and the charging and discharging control plan of a single electric vehicle is optimized by the user himself. The user does not need to report the technical parameters and travel plan of the electric vehicle to the electric vehicle aggregate, but only needs to report the charging and discharging power plan to it anonymously; the aggregate does not need to establish a corresponding relationship between the collected charging and discharging plan and the user, which ensures that the user's privacy information can be fully protected; 3) A distributed method is used to solve the user's individual charging and discharging plan, so that the increasing number of electric vehicles will not significantly bring information aggregation, storage and processing pressure to the aggregate.

[0016] In summary, the present invention uses a hierarchical method to return the Benders cut to the main problem through sub-problems, which can make up for the shortcomings of irreversible aggregation in traditional electric vehicle scheduling and the infeasibility of aggregate power solutions in the individual charging and discharging scheduling of electric vehicles. By improving the sub-problems and constructing approximate Benders cuts, combined with the proposed three-layer scheduling framework, the distributed solution of the problem is realized, which makes up for the defects of traditional scheduling methods such as infringement of user privacy and high pressure of aggregate storage solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the method of the present invention; Figure 2 System net load curves under different control situations; Figure 3 Schematic diagram of the iterative process (V2G control) for different types of cuts; Figure 4 It is a schematic diagram of the two-layer charge and discharge scheduling structure; Figure 5 This is a schematic diagram of the three-layer charge and discharge scheduling structure finally proposed by the present invention; Figure 6 The overall process of the three-layer charge and discharge control method finally proposed by the present invention; Figure 7 It is a structural block diagram of the hierarchical distributed charging and discharging scheduling system of the electric vehicle power system of the present invention. DETAILED DESCRIPTION

[0018] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0019] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0020] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0021] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0022] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0023] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] Example 1 The present invention provides a hierarchical distributed charging and discharging scheduling method for a power system including electric vehicles, comprising the following steps: Step 1: Establish a day-ahead load dispatch model including electric vehicles; Step 2: To facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into the electric vehicle aggregation layer; Step 3: Based on the electric vehicle cluster layer in step 2, according to the upper part of the day-ahead load dispatch model established in step 1, the upper layer is the system dispatch layer, so as to achieve the coordination of the power system source, grid and load, and generate the dispatch instructions of the cluster charging and discharging power; Step 4: The lower layer of the established day-ahead load dispatch model includes the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the dispatch instruction and issues a control signal to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging dispatch, and obtain the user charging and discharging control results.

[0025] Example 2 like Figure 1 As shown, the hierarchical distributed charging and discharging scheduling method for a power system including electric vehicles provided by the present invention comprises the following steps: Step 1: Establish a day-ahead load dispatch model including electric vehicles; In the day-ahead load dispatch model, the objective function is to minimize the total system operation cost.

[0026] (27) In the formula, is the generator set number, is the generator set number, is the total number of time periods in the scheduling cycle, For generator sets The fuel cost function is For generator sets The startup costs, For generator sets The downtime cost, For generator sets exist The effort of the moment, It is a 0-1 variable, indicating the generator set exist The start and stop status of the time period, It is a 0-1 variable, indicating the generator set exist The starting action of the moment, It is a 0-1 variable, indicating the generator set exist Shutdown action at any time.

[0027] The constraints mainly include the following three parts: ① Generator set constraints (28) (29) (30) (31) (32) (33) (34) (35) (36) In the formula, Generator sets For The upward and downward spare capacity available during the period, Generator sets The maximum and minimum output, Generator sets Minimum continuous downtime and start-up time, Generator sets Upward and downward climbing rates, is the time interval length. Among them, equations (28)-(30) are the maximum and minimum output constraints of the unit, equations (31)-(34) are the continuous start-stop time constraints of the unit, and equations (35)-(36) are the ramp constraints of the unit.

[0028] ② Network operation constraints (37) (38) (39) (40) In the formula, The system is The upward and downward spare capacity requirements that can be provided during the period, For Node exist Fixed load other than controllable load during the period, For electric vehicles exist The net power drawn from the grid during the period, is the line number, is the DC power flow distribution factor, For line Active transmission capacity, For Node The set of electric vehicles connected to the grid. (37)-(38) are the reserve capacity constraints, (39) is the power balance constraint, and (40) is the active transmission capacity constraint of the transmission line.

[0029] ③ Electric vehicle restrictions (41) (42) (43) (44) (45) In the formula, Electric vehicles exist The charging and discharging power during the period, For electric vehicles exist The power drawn from the grid during the period, Electric vehicles Maximum charging and discharging power, is a 0-1 variable, indicating the grid-connected status of the electric vehicle. For electric vehicles exist Energy storage level at all times, Electric vehicles Maximum and minimum energy storage levels, Electric vehicles Charging and discharging efficiency, is a parameter, indicating that electric vehicles exist The energy consumed during driving during the period. Among them, (41) is the net power consumption constraint, (42)-(43) are the charging and discharging power constraints, (44) is the upper and lower limit constraints of the battery state of charge, and (45) is the state of charge change constraint.

[0030] From the perspective of model composition, the scale of the problem mainly depends on the number of electric vehicles. As the number of electric vehicles increases, the scale of the model will increase dramatically. At the same time, compared with the generator set, the power and energy scale of a single electric vehicle are extremely limited. If it is directly processed in the same model, the model will be somewhat pathological. From the perspective of structure, the model has significant block characteristics. Large-scale electric vehicle constraints and generator set constraints are coupled through a limited number of system constraints. The power generation optimization scheduling part (unit and system constraints) is only associated with electric vehicle constraints through charging and discharging power. Therefore, the solution to this problem is suitable for the decomposition and coordination algorithm-Benders decomposition method.

[0031] Step 2: To facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into the electric vehicle cluster layer The power drawn from the grid by the aggregate is: (46) Accordingly, constraints (39) and (40) can be rewritten as: (47) (48) At the same time, for polymer power: (49) In the formula, for Moment Electric Vehicle Aggregate The power drawn from the grid, They are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively.

[0032] Step 3: Based on the electric vehicle cluster layer in step 2, according to the upper part of the day-ahead load dispatch model established in step 1, the upper layer is the system dispatch layer, the purpose of which is to achieve the coordination of power system sources, grids and loads, and generate dispatch instructions for cluster charging and discharging power; Take aggregate power To connect the variables, the original problem is transformed into a two-layer problem. The upper layer is the system scheduling layer, and the upper master problem (Master problem, MP) is: (50) The constraints are: equations (28)-(38), equations (47)-(48), and equation (49).

[0033] This problem optimizes the system power generation cost under the conditions of considering the operating constraints of the generator sets, the system operating constraints and the simplified power constraints of the electric vehicle clusters. It is a variant of the traditional unit commitment problem. From a mathematical point of view, this problem belongs to a hybrid MIQP problem; if the fuel cost is piecewise linearized, the problem can be further transformed into a MILP problem, which is easy to solve. Since the number of electric vehicle clusters is relatively limited, the introduction of cluster power will not significantly increase the scale of the problem.

[0034] Step 4: The lower layer of the established day-ahead load dispatch model includes the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the dispatch instruction and issues a control signal to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging dispatch, and obtain the user charging and discharging control results.

[0035] In addition to the upper-level main problem, a large number of electric vehicle charging and discharging constraints and constraints (46) of the original model are classified into the lower level and dealt with in detail in the lower-level problem. Since the relevant constraints of electric vehicles are not directly considered, for a specific electric vehicle, the aggregate power is obtained by solving the upper-level main problem. It may not be feasible. Therefore, it is necessary to perform feasibility verification on the obtained aggregate power and make necessary adjustments based on the verification results. Considering the charging and discharging constraints of a single vehicle and equation (46), the following underlying sub-problem (SP) can be constructed for each electric vehicle aggregate to verify the feasibility of the aggregate charging and discharging scheme: (51) In the formula, is the power deviation between the upper aggregate power scheduling scheme and the feasible scheme, is a non-negative slack variable, is the dual multiplier of the corresponding constraint.

[0036] Solve the model, if the objective function is 0, which means For an electric vehicle cluster, the upper-level scheduling scheme is feasible, and the corresponding cluster power can be allocated to each electric vehicle (i.e., the solution to this problem is ); Otherwise, if the objective function value If it is greater than 0, it means that there is a certain deviation between the upper solution and the feasible solution, and the upper scheduling instruction is not feasible. Accordingly, the following Benders feasibility cuts are generated: (52) The above cuts will be substituted into the upper-level main problem as additional aggregate power constraints to adjust the power generation plan and aggregate charging and discharging plan. If there are still deviations, the next round of optimization scheduling and verification will be carried out, and the iteration will be continued until the above deviations are eliminated. Although the main problem does not explicitly consider the charging and discharging constraints of electric vehicles, the relevant constraints can be implicitly considered through the continuously generated Benders feasible cuts. In this framework, the upper-level problem is responsible for generating aggregate charging and discharging scheduling instructions, and the lower-level problem is responsible for verifying the feasibility of the scheduling instructions. When the upper-level instructions are not feasible, the lower-level feedback Benders feasible cut constraints are fed back to the upper-level solution to make it gradually feasible. The upper and lower layers coordinate with each other through Benders cuts to jointly solve feasible direct scheduling instructions.

[0037] The upper-level main problem is the relaxation of the original problem. Then, the system production cost obtained by solving the upper-level problem is Does not exceed the objective function value of the original problem ,Right now (53) After solving subproblem (51), a set of feasible charging schemes that meet the constraints of electric vehicles is obtained. The aggregate charging and discharging power values ​​obtained by the scheme are substituted back into the upper main problem and fixed. The solution satisfies the electric vehicle constraints, the generator set constraints and the system grid constraints at the same time and is a feasible solution to the original problem. The power generation cost corresponding to this solution Not less than the objective function value of the original problem ,Right now (54) Formulas (53) and (54) respectively define the lower and upper limits of the objective function value of the original problem. The iterative convergence criterion is that the convergence gap is 0. The convergence gap is defined as: (55) The hierarchical coordination algorithm consists of two layers, namely the upper system dispatch (SO) and the lower EV aggregator charge and discharge control layer. The charge and discharge dispatch structure is as follows: Figure 4 As shown. At the upper level, the system dispatcher comprehensively considers the unit constraints, system operation constraints and electric vehicle aggregate power constraints (original constraints and continuously generated Benders cuts) to achieve coordination between the charging and discharging of the generator set and the electric vehicle. At the lower level, the electric vehicle aggregate considers the charging and discharging constraints of a single electric vehicle in detail, verifies whether the aggregate power dispatch scheme obtained by the upper level is feasible, and if so, allocates it to each electric vehicle; if not, forms the corresponding Benders cut, which is returned to the upper level as the basis for adjusting the scheme. Interaction is achieved through multi-step iteration and information transmission between the upper and lower layers until a feasible source-load coordination dispatch scheme with a certain degree of accuracy is formed.

[0038] However, in the proposed two-layer scheduling framework, the electric vehicle cluster needs to collect and summarize the technical parameters and travel information of users, construct the underlying verification problem and solve it centrally. On the one hand, this design will have an adverse impact on user privacy protection; on the other hand, with the popularization of electric vehicles, the number of vehicles under the jurisdiction of the cluster continues to increase, which will bring information aggregation, storage and processing pressure to the cluster. It is necessary to further improve the form of the cluster subproblem and use approximate Benders cut to achieve distributed solution of the subproblem.

[0039] Rewrite equation (46) as: (56) The underlying aggregation subproblem can be rewritten as follows. Since the square of the deviation is non-negative, only a set of slack variables need to be introduced.

[0040] (57) In the formula, To improve the objective function value of the subproblem, is a non-negative slack variable.

[0041] The optimization problem of formula (57) can be further written as: (58) In equation (58), the constraints only include the charging and discharging power constraints of electric vehicles. The electric vehicle variables are separated according to the vehicles, and the separated electric vehicle variables are only coupled through the squared objective function. Compared with problem (51), this problem is easier to solve in a distributed manner.

[0042] When the objective function of problem (57) or (58) is not 0, construct the Benders cut to return the upper objective function: (59) (60) When problem (51) is rewritten as (57), the feasible domain does not change, but problem (57) is non-convex and does not satisfy the Slater constraint specification conditions. Therefore, the method of forming the Benders cut is not strict. The cut defined in equations (59)-(60) is called the approximate Benders cut, and the cut in equation (52) is called the original Benders cut.

[0043] Taking into further consideration issues such as user privacy and the speed of cluster solution, the sub-problem form is improved and combined with the proposed approximate Benders cut to form an electric vehicle charging and discharging load scheduling framework with a three-layer regulatory mechanism and underlying support for distributed computing, and an iterative method is used to achieve distributed solution of the improved sub-problem.

[0044] The distributed solution process of the improved sub-problem is as follows: 1) Initialization of electric vehicle clusters, the number of iterations is

[0045] 2) Initialization of each electric vehicle to form the initial charging and discharging power curve and report to the electric vehicle aggregate to which it belongs; 3) The electric vehicle cluster summarizes the charging and discharging power of each electric vehicle and calculates the The wheel charging and discharging control signal is issued to the electric vehicles under its jurisdiction: (61) In the formula, Electric vehicle aggregates The charge and discharge control parameters, For electric vehicles In the The charging control scheme reported by the round, For aggregate The charging and discharging power instructions obtained from system scheduling, For aggregate Released Round charging control signal.

[0046] 4) Each electric vehicle solves the following single vehicle charging control problem, and the solution is the first Cycle charge and discharge plan , report to the group to which it belongs: (62) 5) The variable of the electric vehicle charging and discharging plan under the electric vehicle cluster calculation is used as the iterative convergence criterion: (63) when If the time is small enough, the iteration is considered to have converged, and the cluster sends a confirmation signal to the electric vehicles under its jurisdiction, and calculates the charging and discharging power of the cluster to form an approximate directional system scheduling report. Otherwise, , go to step 3).

[0047] In the above process, if the charging and discharging parameter values ​​are set according to the following formula, the distributed process is convergent.

[0048] (64) In the formula, is the cardinality of the set, i.e. the number of elements it contains, here the number of electric vehicles governed by the aggregate.

[0049] The above conclusions are proved as follows.

[0050] First, according to As the optimal solution to problem (62), we can obtain: (65) If and only if hour, , equation (65) can be equal, the problem converges, and the proposition is proved. Rearranging equation (65), we can get: (66) On the other hand, due to (67) Therefore, there is (68) make (69) Considering (70) According to formula (68), we have (71) According to the Cauchy inequality, we have (72) From equation (71), combined with equations (66) and (72), we can get (73) When the condition in formula (64) is satisfied, Strictly less than 0. That is, as the iteration proceeds, the value of the objective function of problem (58) continues to decrease, and the calculation process converges.

[0051] The proposed three-layer charging and discharging scheduling framework is as follows Figure 5As shown, the system dispatch (SO) at the top layer has the same functions as the two-layer architecture, that is, through the solution of the main problem, the coordination of power generation units (sources), transmission resources (grids), and electric vehicles (loads) is achieved in a wide area, and the dispatching instructions for the charging and discharging power of the aggregate are generated. The electric vehicle aggregate is located in the middle layer (EVaggregator). It is no longer responsible for collecting the user's vehicle parameters and travel plans, but for guiding and coordinating the charging and discharging instructions issued by the electric vehicle tracking system under its jurisdiction. Based on the difference between the charging and discharging plans reported by the user and the charging and discharging dispatching instructions, they generate charging and discharging control signals and guide users to adjust the charging and discharging plans of the vehicles, rather than directly issuing charging and discharging power instructions to users. Electric vehicle agents (EV agents) are introduced at the bottom layer. They are owned by users, who are responsible for their maintenance and operation, and can be used as part of an electric vehicle energy management system or a home energy management system. The method flow chart is shown in Figure 6 shown.

[0052] Example 3 1) Analysis of the benefits of charging and discharging scheduling The example uses the adapted IEEE-RTS1979, and the maximum load of the system is corrected to 2550MW. It is assumed that the total number of electric vehicles in the adapted IEEE-RTS is 50,000; electric vehicles are distributed at each load node in the same proportion as the basic power load. The types of electric vehicles and their proportions as well as the main parameters of each type of vehicle are shown in Table 1. It is assumed that the charging efficiency of electric vehicles is 0.95 and the discharging efficiency is 0.9 to keep the parameters consistent and facilitate constraint aggregation. User travel data and travel characteristics statistics can refer to relevant survey reports. Based on the above assumptions and data, the technical parameters of 50,000 users' electric vehicles and their travel plans within a day are sampled and simulated. The total electricity demand for electric vehicle travel is about 950MWh. The following three cases, fast charging (Case I), simple charging control (Case II) and V2G control (Case III) are compared and studied; in the latter two cases, the hierarchical charging and discharging load scheduling method based on Benders decomposition proposed in this chapter is adopted.

[0053] Table 1 Distribution of electric vehicle types and main technical parameters

[0054] In the fast charging case, the charging demand of users is first simulated and counted, and then the charging load is substituted into the upper model for power generation scheduling; here, the latter two cases adopt the two-layer charging and discharging load scheduling model proposed in this paper. Figure 2 The power generation cost and net load curve of the whole system under the three control modes are compared respectively.

[0055] Compared with the fast charging situation, after the introduction of simple charging control, the total cost of system power generation is reduced to $725412.9; after considering V2G, the total cost can be further reduced to $723841.6. At the same time, due to the effect of charging and discharging control on the load curve, the net load curve of the system becomes smooth and the load fluctuation is suppressed. Correspondingly, the start-stop frequency and start-stop cost of conventional units are also reduced to a certain extent. After the introduction of charging and discharging control, the operating efficiency of thermal power units has also been significantly improved. After the implementation of simple charging control and V2G control, the unit thermal power energy cost dropped to 13.79$ / MWh and 13.75$ / MWh respectively.

[0056] Table 2 Comparison of production costs under different control situations

[0057] In general, the proposed two-layer model can smooth load fluctuations, improve the operating efficiency of thermal power units, and reduce system production costs. It is a feasible and effective method for charging and discharging loads of electric vehicles. At the same time, compared with simple charging and discharging control, V2G control has more potential and advantages in reducing system production costs and improving system operating efficiency.

[0058] 2) Comparison of different cuts This section analyzes and compares the performance and characteristics of the original Benders cut and the approximate Benders cut proposed in this paper. Table 3 shows the power generation cost of the system under simple charging control and V2G control when using different types of Benders cuts. As shown in the table, the results obtained using the approximate cut are very close to those obtained using the original Benders cut. From the perspective of optimizing the value of the objective function and reducing the power generation cost of the system, the approximate cut proposed in this paper can achieve performance similar to that of the original Benders cut, which provides a feasible implementation plan for stratified charging and discharging control.

[0059] Table 3 Considering different cutting time system production costs

[0060] Figure 3 The iterative processes of the two cuts under the V2G control method are compared, that is, the change of the gap value with the number of iterations.

[0061] As the iteration proceeds, the gap value shows an overall downward trend, with only some small oscillations occurring locally. The main reason for the oscillations is that the upper-level master problem (whether considering the Benders cut or the fixed aggregate charge and discharge power) is a large MILP problem in form. Limited by the current solution technology and solution means, its exact optimal solution cannot be determined, and only an approximate optimal solution with a certain degree of optimality can be obtained. This will cause certain errors in the calculation of the gap. To some extent, the introduction of corresponding errors and oscillations of the gap during the iteration process can be avoided by continuously improving the solution accuracy of the upper-level master problem.

[0062] Table 4 compares the number of iterations required to obtain the optimization results in Table 3.

[0063] Table 4 Comparison of iteration times

[0064] Overall, the number of iterations required for the approximate cut is less. The qualitative analysis is: in the objective function of forming the approximate cut, different degrees of deviation are given different degrees of weight, and the greater the deviation, the greater the weight. In the case of minimizing the total deviation as the objective function, the larger deviation will be quickly limited, thereby accelerating the entire iterative convergence process. Therefore, the approximate Benders cut proposed in this paper can not only be calculated through a distributed method, which is beneficial to the protection of user privacy, but also has certain potential in accelerating the iterative process.

[0065] Example 4 like Figure 7 As shown, the hierarchical distributed charging and discharging scheduling system for a power system including electric vehicles provided by the present invention comprises: Model building module, building a day-ahead load dispatch model including electric vehicles; Aggregation introduction module aggregates electric vehicles and introduces them into the electric vehicle aggregation layer to facilitate problem solving and system scheduling; The dispatch instruction generation module, based on the electric vehicle aggregation layer of the aggregation introduction module, generates the dispatch instruction of the aggregate charging and discharging power according to the upper part of the day-ahead load dispatch model established in the model establishment module, the upper layer is the system dispatch layer, so as to realize the coordination of the power system source, grid and load; The solution module establishes the lower layer of the day-ahead load scheduling model, including the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the scheduling instructions and issues control signals to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging scheduling, and obtain the user charging and discharging control results.

[0066] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0067] In addition, it should be understood that although this specification is described in accordance with the implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A hierarchical distributed charging and discharging scheduling method for a power system including electric vehicles, characterized in that: The following steps are involved: Step 1: Establish a day-ahead load dispatch model including electric vehicles; Step 2: To facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into the electric vehicle aggregation layer; Step 3: Based on the electric vehicle cluster layer in step 2, according to the upper part of the day-ahead load dispatch model established in step 1, the upper layer is the system dispatch layer, so as to achieve the coordination of the power system source, grid and load, and generate the dispatch instructions of the cluster charging and discharging power; Step 4: The lower layer of the established day-ahead load dispatch model includes the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the dispatch instruction and issues a control signal to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging dispatch, and obtain the user charging and discharging control results.

2. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 1 is characterized in that: In step 2, to facilitate problem solving and system scheduling, electric vehicles are aggregated and introduced into the electric vehicle aggregation layer, including: By aggregating electric vehicles within a certain range, where a certain range is a series of electric vehicles within a charging station or geographically close, the segmentation characteristics of the problem are further clarified: (1) (2) In the formula, for Moment Electric Vehicle Aggregate The power drawn from the grid, For electric vehicles exist The power drawn from the grid during the period, They are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively.

3. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 2 is characterized in that: In step 3, based on the electric vehicle cluster layer in step 2, according to the day-ahead load dispatch model established in step 1, the coordination of power system source, grid and load is realized at the system dispatch layer, and the dispatch instructions of cluster charging and discharging power are generated, including: Take aggregate power In order to link variables, the day-ahead load dispatch model is transformed into a two-layer problem, where the upper main problem is: (3) The constraints are: (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) In the formula, Generator sets For The upward and downward spare capacity available during the period, Generator sets The maximum and minimum output, Generator sets Minimum continuous downtime and start-up time, Generator sets Upward and downward climbing rates, is the time interval length; The system is The upward and downward spare capacity requirements that can be provided during the period, For Node exist Fixed load other than controllable load during the period, For electric vehicles exist The net power drawn from the grid during the period, is the line number, is the DC power flow distribution factor, For Line Active transmission capacity, For Node The collection of electric vehicles connected to the grid, where for Moment Electric Vehicle Aggregate The power drawn from the grid, are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively; where (4)-(6) are the maximum and minimum output constraints of the unit, (7)-(10) are the continuous start-stop time constraints of the unit, (11)-(12) are the ramp constraints of the unit, (13)-(14) are the reserve capacity constraints, (15) is the power balance constraint, (16) is the active transmission capacity constraint of the transmission line, and (17) is the cluster variable constraint.

4. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 3 is characterized in that: In step 4, the lower-level problem SP can be constructed as: The obtained aggregate power is feasibility checked and adjusted according to the check results. The underlying sub-problems shown below are constructed to check the feasibility of the aggregate charging and discharging scheme: (18) In the formula, is the power deviation between the upper aggregate power scheduling scheme and the feasible scheme, is a non-negative slack variable, is the dual multiplier of the corresponding constraint, Electric vehicles exist The charging and discharging power during the period, Electric vehicles Maximum charging and discharging power, is a 0-1 variable, indicating the grid-connected status of the electric vehicle. For electric vehicles exist Energy storage level at all times, Electric vehicles Maximum and minimum energy storage levels, Electric vehicles Charging and discharging efficiency, is a parameter, indicating that electric vehicles exist The energy consumed during driving.

5. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 4 is characterized in that: In step 3 and step 4, the obtained two-layer problem is solved hierarchically using the Benders decomposition method, and the generated Benders cut is in the form of: (19)。 6. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 5, characterized in that: In step 4, the underlying aggregate subproblem is improved to the following form; since the square of the deviation is non-negative, a set of slack variables is introduced; (20) In the formula, To improve the objective function value of the subproblem, is a non-negative slack variable; The optimization problem of formula (20) can be further written as: (21) In formula (21), the constraints include the EV charging and discharging power constraints, which are separated according to the vehicle, and the separated EV variables are only coupled through the squared objective function; When the objective function of problems (20) and (21) is not 0, construct the Benders cut to return the upper objective function: (22) (23) When problem (18) is rewritten as (20), the feasible domain does not change, but problem (20) is non-convex and does not satisfy the Slater constraint specification conditions. Therefore, the method of forming the Benders cut is not strict. The cut defined by equations (22)-(23) is called the approximate Benders cut.

7. The hierarchical distributed charging and discharging scheduling method for a power system containing electric vehicles according to claim 6 is characterized in that: In step 4, considering the user privacy issue and the storage solution pressure caused by the increase in the number of electric vehicles in the cluster, the form of the cluster sub-problem is improved, and finally a three-layer control mechanism and a bottom-level distributed computing-supported electric vehicle charging and discharging load scheduling framework are formed; The upper layer is system scheduling; The middle layer is the electric vehicle cluster that aggregates the charging and discharging power of each electric vehicle. The following formula is used to calculate the The wheel charging and discharging control signal is issued to the electric vehicles under its jurisdiction: (24) In the formula, Electric vehicle aggregates The charge and discharge control parameters, For electric vehicles In the The charging control scheme reported by the round, For aggregate The charging and discharging power instructions obtained from system scheduling, For aggregate Released Wheel charging control signal; Each electric vehicle in the lower layer solves the following single vehicle charging control problem, and the solution is the first Cycle charge and discharge plan , report to the group to which it belongs: (25) When solving the subproblem, the variable of the electric vehicle charging and discharging plan under the electric vehicle aggregation calculation is used as the iterative convergence criterion: (26) when If the time is small enough, the iteration is considered to have converged, the cluster sends a confirmation signal to the electric vehicles under its jurisdiction, and calculates the charging and discharging power of the cluster to form an approximate directional system scheduling report.

8. A hierarchical distributed charging and discharging dispatching system for a power system including electric vehicles, characterized in that: include: Model building module, building a day-ahead load dispatch model including electric vehicles; Aggregation introduction module aggregates electric vehicles and introduces them into the electric vehicle aggregation layer to facilitate problem solving and system scheduling; The dispatch instruction generation module, based on the electric vehicle aggregation layer of the aggregation introduction module, generates the dispatch instruction of the aggregate charging and discharging power according to the upper part of the day-ahead load dispatch model established in the model establishment module, the upper layer is the system dispatch layer, so as to realize the coordination of the power system source, grid and load; The solution module establishes the lower layer of the day-ahead load scheduling model, including the aggregate layer and the user layer. The electric vehicle aggregate layer responds to the scheduling instructions and issues control signals to the user layer to ensure the feasibility of the charging and discharging plan of a single electric vehicle, realize the distributed solution of the charging and discharging scheduling, and obtain the user charging and discharging control results.

9. The hierarchical distributed charging and discharging dispatching system for the electric power system including electric vehicles according to claim 8, characterized in that: In the aggregation introduction module, electric vehicles are aggregated and introduced into the electric vehicle aggregation layer to facilitate problem solving and system scheduling, including: By aggregating electric vehicles within a certain range, where a certain range is a series of electric vehicles within a charging station or geographically close, the segmentation characteristics of the problem are further clarified: (1) (2) In the formula, for Moment Electric Vehicle Aggregate The power drawn from the grid, For electric vehicles exist The power drawn from the grid during the period, They are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively.

10. The hierarchical distributed charging and discharging dispatching system for the electric power system including electric vehicles according to claim 9, characterized in that: In the dispatch instruction generation module, based on the electric vehicle aggregation layer of the aggregation introduction module, according to the day-ahead load dispatch model established in the model establishment module, the coordination of the power system source, grid and load is realized at the system dispatch layer, and the dispatch instruction of the aggregate charging and discharging power is generated, including: Take aggregate power In order to link variables, the day-ahead load dispatch model is transformed into a two-layer problem, where the upper main problem is: (3) The constraints are: (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) In the formula, Generator sets For The upward and downward spare capacity available during the period, Generator sets The maximum and minimum output, Generator sets Minimum continuous downtime and start-up time, Generator sets Upward and downward climbing rates, is the time interval length; The system is The upward and downward spare capacity requirements that can be provided during the period, For Node exist Fixed load other than controllable load during the period, For electric vehicles exist The net power drawn from the grid during the period, is the line number, is the DC power flow distribution factor, For Line Active transmission capacity, For Node The collection of electric vehicles connected to the grid, where for Moment Electric Vehicle Aggregate The power drawn from the grid, are the upper and lower limits of power estimated according to the number and type of electric vehicles in the cluster, respectively; where (4)-(6) are the maximum and minimum output constraints of the unit, (7)-(10) are the continuous start-stop time constraints of the unit, (11)-(12) are the ramp constraints of the unit, (13)-(14) are the reserve capacity constraints, (15) is the power balance constraint, (16) is the active transmission capacity constraint of the transmission line, and (17) is the cluster variable constraint.

Images