Accurate evaluation method for interactive aggregation adjustable capability of electric vehicle and power distribution network

Abstract

The invention relates to the technical field of electric power system operation and control, in particular to an electric vehicle and power distribution network interactive aggregation adjustable capability accurate evaluation method, which comprises the following steps: firstly, based on a similar energy storage model, establishing a feasible region model of a single electric vehicle; secondly, preprocessing the data of the electric vehicle; next, based on the preprocessed data and the feasible region model of the single electric vehicle, establishing a polyhedron-based aggregation model; and finally, considering physical characteristics and combining with a polyhedron-based aggregation model to establish an aggregation feasible region model of the electric vehicle. According to the electric vehicle and power distribution network interactive aggregation adjustable capability accurate evaluation method provided by the invention, large-scale electric vehicles can be effectively aggregated to participate in energy and standby markets.

Description

Technical Field

[0001] The present application relates to the field of power system operation and control technology, and in particular to a method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks. Background Art

[0002] As more and more distributed photovoltaic and distributed wind power generation are connected to the distribution network, the uncertainty of renewable energy has brought safety problems such as voltage over-limit and line congestion to the operation of the distribution network. It is necessary to further make full use of demand-side flexibility resources such as electric vehicles in the distribution network to cope with the challenges brought by the increase in the penetration rate of distributed renewable energy. Although large-scale electric vehicles can provide more flexible and adjustable capabilities for the distribution network, if each electric vehicle is directly dispatched by the dispatch center, it will lead to problems such as the scheduling model dimension being too large and the communication conditions not being able to meet the requirements. How to achieve effective aggregation of electric vehicles is the key to realizing the use of their flexibility conditions to cope with the access of renewable energy.

[0003] In existing technical research, the aggregation method of electric vehicles is mainly to establish the feasible domain of a single vehicle through the relevant parameters of the electric vehicle, and then realize the aggregation of large-scale electric vehicles based on the power and energy projection summation method. However, this method will overestimate the adjustable capacity of aggregation and cannot guarantee the reliability of the energy and backup services provided by electric vehicles. The method that can achieve accurate aggregation is to calculate the Minerkovsky sum of the feasible domain of a single vehicle, but this method is too inefficient. As the number of electric vehicles increases, its calculation time increases exponentially, and it cannot be used for the aggregation of large-scale electric vehicles.

[0004] In order to avoid the problem of low computational efficiency in accurate aggregation, some studies first use the basic model to approximate the feasible domain of a single electric vehicle, and then obtain the approximate aggregate feasible domain by summing the relevant shape coefficients of the basic model. These include box-based aggregation methods and symmetric polygon-based aggregation methods. However, because the shape characteristics of the box and symmetric polygons are far from the flexible and adjustable model of actual electric vehicles, the aggregation efficiency of these two methods is low, resulting in a large loss of flexibility of electric vehicles in the process of approximate aggregation, and the efficient use of the flexibility of electric vehicles is not achieved.

[0005] Therefore, there is currently a lack of efficient methods for aggregating large numbers of electric vehicles within a limited time while ensuring the reliability of the energy and backup services provided by electric vehicles. Summary of the invention

[0006] The embodiment of the present application provides a method for accurately evaluating the adjustable capacity of electric vehicles and distribution networks in interactive aggregation. The proposed method for accurately evaluating the adjustable capacity of electric vehicles can effectively aggregate large-scale electric vehicles to participate in the energy and backup markets; the proposed polyhedron-based aggregation method can achieve reliable aggregation of large-scale electric vehicles, and ensure the reliability of the energy and backup provided by electric vehicles; the proposed method considering the physical properties of batteries can overcome the calculation errors of energy levels in the polyhedron-based aggregation method, and achieve accurate calculation of electric vehicle scheduling instructions.

[0007] To solve the above technical problems, an embodiment of the present application provides a method for accurately evaluating the interactive aggregated adjustable capacity of electric vehicles and distribution networks, comprising the following steps: first, based on a quasi-energy storage model, a feasible domain model of a single electric vehicle is established; then, the data of the electric vehicle is preprocessed; next, based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established; finally, considering the physical characteristics and combining the polyhedron-based aggregation model, an aggregated feasible domain model of the electric vehicle is established.

[0008] In some exemplary embodiments, a feasible domain model of a single electric vehicle is established based on a quasi-energy storage model, including: treating a single electric vehicle as a flexible energy storage unit when connected to the power grid, and scheduling the power of the charging process of the electric vehicle under the premise that its charging demand can be met when the electric vehicle leaves; based on the quasi-energy storage model, calculating the flexible feasible domain when the single electric vehicle is connected to the power grid, and establishing a feasible domain model for the single electric vehicle.

[0009] In some exemplary embodiments, according to the energy storage-like model shown in formula (1) to formula (5), the flexibility feasible domain when a single electric vehicle is connected to the power grid is calculated; (1) (2) (3) (4) (5) Among them, i is the number of the electric vehicle; is the charge state of the electric vehicle; is the charging / discharging power of the electric vehicle; The charging and discharging efficiency of electric vehicles; is the arrival / departure time of the electric vehicle; is the maximum charge / discharge power of the electric vehicle; Maintain the upper / lower limits of the electric vehicle's state of charge; The upper / lower limits of the battery capacity allowed for electric vehicles; It is the target power of electric vehicle when it is off-grid.

[0010] In some exemplary embodiments, the feasible domain of a single electric vehicle defined by equations (1) to (5) can be expressed as a linear constraint set of polyhedral nature: , as shown below: (6) In the formula, ; and is the coefficient matrix, which is defined as follows: (7) In the formula, for T dimensional identity matrix; and The lower triangular elements are all and The lower triangular matrix of ; .

[0011] In some exemplary embodiments, the data of electric vehicles are preprocessed, including: excluding inflexible electric vehicles with a charging time of less than one hour from the cluster to be aggregated; normalizing the arrival and departure times of the electric vehicles by rounding them to the nearest 15-minute interval; and clustering the electric vehicles according to their arrival and departure times using a k-means algorithm.

[0012] In some exemplary embodiments, the arrival time of all electric vehicles in each cluster is set to the latest arrival time of all electric vehicles in the cluster, and the departure time is set to the earliest departure time of all electric vehicles to achieve the unification of the feasible domain dimensions.

[0013] In some exemplary embodiments, the clustering formula is as follows: (8) (9) in, and They represent the unified arrival and departure times of all electric vehicles in the cluster after preprocessing; represents the number of electric vehicles in the cluster.

[0014] In some exemplary embodiments, a polyhedron-based aggregation model is established based on preprocessed data and a feasible domain model of a single electric vehicle, including: based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based electric vehicle aggregation method is used to establish a polyhedron-based aggregation model; a feasible domain Minkowski sum shown in formula (10) is used to represent a model for accurately aggregating electric vehicles; (10) In the formula, It means to find the Minkowski sum; .

[0015] In some exemplary embodiments, a polyhedron-based electric vehicle aggregation method is used to establish a polyhedron-based aggregation model, including: selecting a basic polyhedron; translating and scaling the basic polyhedron to approximate the original feasible domain of each electric vehicle; and achieving aggregation by summing the translation coefficient and the scaling coefficient; wherein the basic polyhedron is represented as follows: (11) (12) (13) In the formula, and is the coefficient matrix of the basic type; Under the constraints of the inner approximation, the optimal scaling and translation coefficients are determined to maximize the utilization of the EV flexible model as follows: (14) (15) Among them, formula (14) indicates that the objective function is to maximize the volume of the approximate feasible domain; formula (15) indicates that the approximate feasible domain must be within the range of the original feasible domain; is the scaling factor; is the translation coefficient.

[0016] In some exemplary embodiments, after the objective function is modified by adding the minimum value of the linearized charge and discharge power to the objective function, the proposed polyhedron-based electric vehicle aggregation method considering physical properties can obtain the translation coefficient with the least impact on the physical constraints while ensuring that the feasible domain volume is consistent with the original method. The specific model is expressed as follows: (16) (17) (18) In the formula, is the original objective function; To obtain the translation coefficient that minimizes the violation of physical constraints under the optimal aggregation feasible domain volume; is the minimum value of the charge and discharge power after linearization; Representation Matrix No. to List; Represents a variable.

[0017] The technical solution provided by the embodiments of the present application has at least the following advantages: The embodiment of the present application provides a precise evaluation method for the interactive aggregation adjustable capacity of electric vehicles and distribution networks, comprising the following steps: first, based on a class energy storage model, a feasible domain model of a single electric vehicle is established; then, the data of the electric vehicle is preprocessed; next, based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established; finally, considering the physical properties and combining the polyhedron-based aggregation model, an aggregated feasible domain model of the electric vehicle is established. The precise evaluation method for the aggregate adjustable capacity of electric vehicles proposed in the present application can effectively aggregate a large number of electric vehicles into an equivalent energy storage model to participate in the optimal scheduling of the distribution network, avoiding high-dimensional models and excessive communication requirements. The proposed polyhedron-based aggregation method can complete the reliable aggregation of electric vehicles within a limited time, because the shape characteristics of the polyhedron used are very similar to the feasible domain model of the electric vehicle, so the utilization rate of the flexibility of electric vehicles is greatly improved compared with the existing method. The proposed method considering the physical properties of the battery can overcome the calculation error of the energy level in the polyhedron-based aggregation method, and realize the accurate calculation of the dispatching instructions by the dispatching center. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplifications do not constitute limitations on the embodiments. Unless otherwise stated, the pictures in the drawings do not constitute proportional limitations.

[0019] Figure 1 A flowchart of a method for accurately evaluating the interactive aggregation adjustable capacity of an electric vehicle and a distribution network provided in one embodiment of the present application. DETAILED DESCRIPTION

[0020] As can be seen from the background technology, there is currently a lack of an efficient method for aggregating large-scale electric vehicles within a limited time while ensuring the reliability of the energy and backup services provided by the electric vehicles.

[0021] In order to solve the above technical problems, the embodiment of the present application provides a method for accurately evaluating the adjustable capacity of electric vehicles and distribution networks in interactive aggregation, including the following steps: first, based on a quasi-energy storage model, a feasible domain model of a single electric vehicle is established; then, the data of the electric vehicle is preprocessed; next, based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established; finally, considering the physical characteristics and combining the polyhedron-based aggregation model, an aggregated feasible domain model of the electric vehicle is established. The embodiment of the present application provides a method for accurately evaluating the adjustable capacity of electric vehicles and distribution networks in interactive aggregation. The proposed method for accurately evaluating the adjustable capacity of electric vehicles in aggregate can effectively aggregate large-scale electric vehicles to participate in the energy and backup markets; the proposed polyhedron-based aggregation method can achieve reliable aggregation of large-scale electric vehicles and ensure the reliability of the energy and backup provided by electric vehicles; the proposed method considering the physical characteristics of batteries can overcome the calculation error of the energy level in the polyhedron-based aggregation method and achieve accurate calculation of the dispatching instructions of electric vehicles.

[0022] The following will describe the various embodiments of the present application in detail with reference to the accompanying drawings. However, it will be appreciated by those skilled in the art that in the various embodiments of the present application, many technical details are provided in order to enable the reader to better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solution claimed in the present application can be implemented.

[0023] See also Figure 1 The embodiment of the present application provides a method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks, including the following steps: Step S1: Based on the energy storage model, a feasible domain model of a single electric vehicle is established.

[0024] Step S2: pre-processing the electric vehicle data.

[0025] Step S3: establishing a polyhedron-based aggregation model based on the preprocessed data and the feasible domain model of a single electric vehicle.

[0026] Step S4: Considering the physical characteristics and combining the polyhedron-based aggregation model, an aggregation feasible domain model of electric vehicles is established.

[0027] In some embodiments, based on a quasi-energy storage model, a feasible domain model of a single electric vehicle is established, including: treating a single electric vehicle as a flexible energy storage unit when connected to the power grid, and scheduling the power of the charging process of the electric vehicle under the premise that its charging demand can be met when the electric vehicle leaves; based on the quasi-energy storage model, calculating the flexible feasible domain when the single electric vehicle is connected to the power grid, and establishing a feasible domain model of the single electric vehicle.

[0028] In some embodiments, according to the energy storage-like model shown in formula (1) to formula (5), the flexibility feasible domain when a single electric vehicle is connected to the power grid is calculated; (1) (2) (3) (4) (5) Among them, i is the number of the electric vehicle; is the charge state of the electric vehicle; is the charging / discharging power of the electric vehicle; The charging and discharging efficiency of electric vehicles; is the arrival / departure time of the electric vehicle; is the maximum charge / discharge power of the electric vehicle; Maintain the upper / lower limits of the electric vehicle's state of charge; The upper / lower limits of the battery capacity allowed for electric vehicles; It is the target power of electric vehicle when it is off-grid.

[0029] In some embodiments, the feasible domain of a single electric vehicle defined by equations (1) to (5) can be expressed as a set of linear constraints of polyhedral nature: , as shown below: (6) In the formula, ; and is the coefficient matrix, which is defined as follows: (7) In the formula, for T dimensional identity matrix; and The lower triangular elements are all and The lower triangular matrix of ; .

[0030] Because the polyhedron-based aggregation method needs to approximate the feasible domain by scaling the basic polyhedron in proportion, it requires that the dimensions of the feasible domains of different electric vehicles (EVs) remain consistent. In order to achieve dimensional consistency of the feasible domain, the original data needs to be preprocessed. First, EVs with less flexibility and charging time of less than one hour are excluded from the cluster to be aggregated. Then, the arrival and departure times of EVs are normalized by rounding them to the nearest 15-minute interval. Finally, the k-means algorithm is used to cluster EVs according to their arrival and departure times. To ensure that it is an internal approximate aggregation of the original feasible domain, the arrival time of all EVs in each cluster is set to the latest arrival time of all EVs in the cluster, and the departure time is set to the earliest departure time of all EVs to achieve the unification of the feasible domain dimensions.

[0031] In some embodiments, the data of electric vehicles are preprocessed, including: excluding inflexible electric vehicles with a charging time of less than one hour from the cluster to be aggregated; normalizing the arrival and departure times of the electric vehicles by rounding them to the nearest 15-minute interval; and clustering the electric vehicles according to their arrival and departure times using a k-means algorithm.

[0032] In some exemplary embodiments, the arrival time of all electric vehicles in each cluster is set to the latest arrival time of all electric vehicles in the cluster, and the departure time is set to the earliest departure time of all electric vehicles to achieve the unification of the feasible domain dimensions.

[0033] In some exemplary embodiments, the clustering formula is as follows: (8) (9) in, and They represent the unified arrival and departure times of all electric vehicles in the cluster after preprocessing; represents the number of electric vehicles in the cluster.

[0034] In some exemplary embodiments, a polyhedron-based aggregation model is established based on preprocessed data and a feasible domain model of a single electric vehicle, including: based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based electric vehicle aggregation method is used to establish a polyhedron-based aggregation model; a feasible domain Minkowski sum shown in formula (10) is used to represent a model for accurately aggregating electric vehicles; (10) In the formula, It means to find the Minkowski sum; .

[0035] However, finding the Minkowski sum of a high-dimensional polyhedron is a non-deterministic polynomial time hard problem, and the efficiency of the solution cannot be guaranteed when aggregating large-scale EVs, so it is not suitable for aggregating large-scale EVs. Because the Minkowski sum of a polyhedron obtained by translating and scaling the same basic polyhedron can be obtained by directly summing the translation coefficients and scaling coefficients of each polyhedron, thus avoiding the problem of non-deterministic polynomial time hardness. The idea of ​​the polyhedron-based EV aggregation method is to first select a basic polyhedron, approximate the original feasible domain of each EV by translating and scaling this basic polyhedron, and then achieve aggregation by summing the translation coefficients and scaling coefficients.

[0036] The efficiency of the polyhedron-based aggregation method in utilizing flexibility depends on the similarity between the basic form and the original feasible domain. In order to make the basic form more consistent with the shape of the original feasible domain and fully utilize the flexibility, it is first necessary to determine the basic form by averaging the parameters of all EV feasible domains in the cluster. , so that the shape of the basic model will be close to the shape of the original feasible domain, thereby reducing the waste of flexibility and retaining the characteristics of the energy storage model. The basic model is shown in formula (11).

[0037] In some embodiments, a polyhedron-based electric vehicle aggregation method is used to establish a polyhedron-based aggregation model, including: selecting a basic polyhedron; translating and scaling the basic polyhedron to approximate the original feasible domain of each electric vehicle; and achieving aggregation by summing the translation coefficient and the scaling coefficient; wherein the basic polyhedron is represented as follows: (11) (12) (13) In the formula, and is the coefficient matrix of the basic type; Under the constraints of the inner approximation, the optimal scaling and translation coefficients are determined to maximize the utilization of the EV flexible model as follows: (14) (15) Among them, formula (14) indicates that the objective function is to maximize the volume of the approximate feasible domain; formula (15) indicates that the approximate feasible domain must be within the range of the original feasible domain; is the scaling factor; is the translation coefficient.

[0038] The Farkas lemma can be used to transform the aggregation model shown in formula (14) and formula (15) into a linear model for solution. The definition of Farkas lemma is as follows: If the inequality All feasible solutions of satisfy the inequality , then there must exist a matrix G At the same time, the constraints and constraints , and vice versa is also true.

[0039] The specific steps of transforming and solving the aggregation model using Farkas' lemma are as follows: First, by transforming the variables and Re-expressed as and , transform the models shown in formula (14) and formula (15) into the following form: (16) (17) (18) (19) From formula (17) to formula (19), we can see that any inequality The solutions all satisfy According to Farkas's lemma, if formula (17) holds, then there must exist a matrix At the same time, the constraints and , and vice versa is also true.

[0040] According to the above derivation, formula (17) can be equivalently replaced by linear constraints (21) and (22). In this way, the original aggregation model can be expressed as the following linear model using Farkas lemma. The optimal solution based on this linear model is and , the optimal scaling factor and translation factor and They can be expressed as and .

[0041] (20) (twenty one) (twenty two) After solving the above model to obtain the optimal translation coefficient and scaling coefficient for each EV, the Minkowski sum of the approximate feasible domain can be directly obtained by calculating the translation coefficient and scaling coefficient and The sum of the points is expressed as: (twenty three) In the formula, is the aggregation feasible domain; is the approximate feasible region of a single EV.

[0042] When considering the aggregation model of physical characteristics, the original polyhedron-based aggregation method will produce mandatory charging and discharging power constraints that violate the physical characteristics of EV due to translation, resulting in the energy level calculated by the theoretical model being lower than the energy level calculated by the actual physical model. The polyhedron-based aggregation method based on physical characteristics proposed on the basis of the method introduced in the previous section aims to minimize the mandatory power constraints that violate physical characteristics due to translation while keeping the aggregation feasible domain volume consistent with the original method, thereby enhancing the accuracy of the energy level calculated by the theoretical model. First, by analyzing the original aggregation model shown in equations (20) to (22), it can be found that the variable The value of is uniquely determined by formula (21), so that and The feasible domain of is only determined by equation (22).

[0043] In order to obtain the translation coefficient that minimizes the violation of physical constraints under the optimal aggregation feasible domain volume , it is necessary to add an expression related to the minimum value of the shifted charge and discharge power in the objective function .

[0044] Compared with the original minimum charge and discharge power constraint expression , the denominator of the added expression M A relatively large positive number acts as a weighting factor to ensure that the added expression is relative to the original objective function can be ignored. Thus, the scaling factor obtained by improving the model is Keeping consistent with the original model, the optimality of the feasible domain volume is ensured. The value of is approximately 1 and M is a very large number, so from Removing it will not affect the solution of the model. Therefore, It can be simplified to a linear form .

[0045] In some embodiments, considering physical characteristics and combining with the polyhedron-based aggregation model, an aggregation feasible domain model of electric vehicles is established, including: after the objective function is modified by adding the minimum value of the linearized charge and discharge power in the objective function, the proposed polyhedron-based electric vehicle aggregation method considering physical characteristics can obtain the translation coefficient with the least impact on physical constraints while ensuring that the feasible domain volume is consistent with the original method. The specific model is expressed as follows: (twenty four) (25) (26) In the formula, is the original objective function; To obtain the translation coefficient that minimizes the violation of physical constraints under the optimal aggregation feasible domain volume; is the minimum value of the charge and discharge power after linearization; Representation Matrix No. to List; Represents a variable.

[0046] Based on the above technical solutions, the embodiment of the present application provides a precise evaluation method for the interactive aggregation adjustable capacity of electric vehicles and distribution networks, including the following steps: first, based on a class energy storage model, a feasible domain model of a single electric vehicle is established; then, the data of the electric vehicle is preprocessed, an aggregation model based on a polyhedron is established, an aggregation model considering physical properties is established, and an aggregation feasible domain model of the electric vehicle is established. The precise evaluation method for the aggregate adjustable capacity of electric vehicles proposed in the present application can effectively aggregate a large number of electric vehicles into an equivalent energy storage model to participate in the optimal scheduling of the distribution network, avoiding high-dimensional models and excessive communication requirements. The proposed polyhedron-based aggregation method can complete the reliable aggregation of electric vehicles within a limited time, because the shape characteristics of the polyhedron used are very similar to the feasible domain model of the electric vehicle, so compared with the existing method, the utilization rate of the flexibility of electric vehicles is greatly improved. The proposed method considering the physical properties of the battery can overcome the calculation error of the energy level in the polyhedron-based aggregation method, and realize the accurate calculation of the dispatching instructions by the dispatching center.

[0047] Those skilled in the art can understand that the above-mentioned embodiments are specific examples for implementing the present application, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of the present application. Any person skilled in the art can make their own changes and modifications without departing from the spirit and scope of the present application, so the scope of protection of the present application shall be based on the scope defined in the claims.

Claims

1. A precise evaluation method for the interactive aggregation adjustable capacity of electric vehicles and distribution networks, characterized in that: The following steps are involved: Based on the energy storage model, a feasible domain model of a single electric vehicle is established; Preprocessing of electric vehicle data; Based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established; Considering the physical characteristics and combining the polyhedron-based aggregation model, an aggregation feasible domain model of electric vehicles is established.

2. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 1 is characterized in that: Based on the energy storage model, a feasible domain model of a single electric vehicle is established, including: Treating a single electric vehicle connected to the grid as a flexible energy storage unit, the power of the electric vehicle's charging process is dispatched while ensuring that its charging needs are met when the electric vehicle leaves the grid. Based on the energy storage-like model, the flexibility feasible domain of a single electric vehicle when connected to the power grid is calculated, and a feasible domain model of a single electric vehicle is established.

3. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 2 is characterized in that: According to the energy storage-like model shown in formula (1) to formula (5), the flexibility feasible domain when a single electric vehicle is connected to the grid is calculated; （1） （2） （3） （4） （5） Among them, i is the number of the electric vehicle; is the charge state of the electric vehicle; is the charging / discharging power of the electric vehicle; The charging and discharging efficiency of electric vehicles; is the arrival / departure time of the electric vehicle; is the maximum charge / discharge power of the electric vehicle; Maintain the upper / lower limits of the electric vehicle's state of charge; The upper / lower limits of the battery capacity allowed for electric vehicles; It is the target power of electric vehicle when it is off-grid.

4. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 3 is characterized in that: The feasible domain of a single electric vehicle defined by equations (1) to (5) can be expressed as a set of linear constraints that are polyhedral in nature: , as shown below: （6） In the formula, ; and is the coefficient matrix, which is defined as follows: （7） In the formula, for T dimensional identity matrix; and The lower triangular elements are all and The lower triangular matrix of ; .

5. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 1 is characterized in that: Preprocessing of electric vehicle data, including: Exclude inflexible electric vehicles with charging times of less than one hour from being included in the clusters to be aggregated; Arrival and departure times of electric vehicles were normalized by rounding them to the nearest 15-minute interval; The k-means algorithm is used to cluster electric vehicles according to their arrival and departure times.

6. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 5 is characterized in that: The arrival time of all electric vehicles in each cluster is set as the latest arrival time of all electric vehicles in the cluster, and the departure time is set as the earliest departure time of all electric vehicles to achieve the unification of the feasible domain dimensions.

7. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 5 is characterized in that: The clustering formula is as follows: （8） （9） in, and They represent the unified arrival and departure times of all electric vehicles in the cluster after preprocessing; represents the number of electric vehicles in the cluster.

8. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 1 is characterized in that: Based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established, including: Based on the preprocessed data and the feasible domain model of a single electric vehicle, a polyhedron-based aggregation model is established by using the polyhedron electric vehicle aggregation method; The feasible domain Minkowski sum shown in formula (10) is used to represent the model for accurate aggregation of electric vehicles; （10） In the formula, It means to find the Minkowski sum; .

9. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 8 is characterized in that: The polyhedron-based electric vehicle aggregation method is adopted to establish a polyhedron-based aggregation model, including: Select a basic polyhedron; The basic model is translated and scaled to approximate the original feasible domain of each electric vehicle; Aggregation is achieved by summing the translation and scaling factors; Among them, the basic type is as follows: （11） （12） （13） In the formula, and is the coefficient matrix of the basic type; Under the constraints of the inner approximation, the optimal scaling and translation coefficients are determined to maximize the utilization of the EV flexible model as follows: （14） （15） Among them, formula (14) indicates that the objective function is to maximize the volume of the approximate feasible domain; formula (15) indicates that the approximate feasible domain must be within the range of the original feasible domain; is the scaling factor; is the translation coefficient.

10. The method for accurately evaluating the interactive aggregation adjustable capacity of electric vehicles and distribution networks according to claim 1 is characterized in that: Considering the physical characteristics and combining the polyhedron-based aggregation model, an aggregation feasible domain model of electric vehicles is established, including: After the objective function is modified by adding the minimum value of the linearized charge and discharge power to the objective function, the proposed polyhedron-based electric vehicle aggregation method considering physical characteristics can obtain the translation coefficient with the minimum impact on physical constraints while ensuring that the feasible domain volume is consistent with the original method. The specific model is expressed as follows: （16） （17） （18） In the formula, is the original objective function; To obtain the translation coefficient that minimizes the violation of physical constraints under the optimal aggregation feasible domain volume; is the minimum value of the charge and discharge power after linearization; Representation Matrix No. to List; Represents a variable.

Images