An electric vehicle aggregator energy-frequency modulation bidding optimization method and device, electronic equipment and medium

Abstract

The present application relates to the technical field of battery management, and more particularly to an electric vehicle aggregator energy-frequency modulation bidding optimization method and device, electronic equipment and medium, in the method, first, EV historical charging and travel data and user survey data are acquired and preprocessed; second, a flexible index system including adjustable capacity, online time and charging power is constructed to form a cluster chino polyhedral flexible domain constraint; third, a fuzzy neural network is used to evaluate user response willingness, and a dispatchable user set is determined according to a threshold value; in a multi-price scenario, the total expected income is maximized as the target, the charging and discharging bidding power, the upward / downward frequency modulation capacity and the subsidy decision quantity of each period are jointly optimized, and the power, capacity dynamics, frequency modulation margin and cluster flexible domain constraints are satisfied, and the energy and frequency modulation resource competition-complementary overall allocation is realized. It is suitable for demand response resource aggregation transaction, improves the bidding executability and income stability, and reduces the examination punishment risk.

Description

Technical Field

[0001] This invention relates to the field of battery management technology, and in particular to a method, apparatus, electronic device, and medium for optimizing energy-frequency modulation bidding for electric vehicles. Background Technology

[0002] Electric vehicles possess controllable charging and discharging capabilities, enabling them to participate in ancillary services such as peak shaving and valley filling, and frequency regulation as demand response resources. With the growth in electric vehicle ownership and the improvement of electricity market mechanisms, electric vehicle aggregators can aggregate dispersed electric vehicle resources to form adjustable capacity that meets market access requirements, and submit bids and organize their execution in the energy and frequency regulation markets.

[0003] In existing technologies, one type of method focuses on charge-discharge arbitrage in the energy market, while another type focuses on capacity gains or response performance in the frequency regulation market. However, many solutions model the energy market and the frequency regulation market separately, failing to fully reflect the competitive relationship and coupling constraints between the two markets for the same charge-discharge power and battery state of energy resources. This can easily lead to problems of uncoordinated allocation of bidding resources in multi-time periods and multi-price fluctuation scenarios, thereby affecting overall returns and feasibility of implementation.

[0004] On the other hand, the availability of electric vehicle resources is not only limited by objective conditions such as battery energy state, charging and discharging power limits, and on-grid duration, but also by subjective factors such as user sensitivity to discharge scheduling, frequency regulation participation, and subsidies. Some existing solutions often treat users as completely obedient to scheduling or simplify user wishes, leading to a larger deviation between "bid application and actual execution," which in turn causes risks such as insufficient performance, performance penalties, or additional compensation costs.

[0005] Furthermore, if the adjustable range is characterized only by static SOC boundaries or simple capacity limits, without combining historical charging and travel behavior to form a feasible domain constraint that reflects comprehensive factors such as adjustable capacity, on-grid time, and charging power, the available flexibility may be overestimated, making it difficult for the bidding results to meet the energy state boundary and power constraints in actual scheduling, thus reducing the reliability of the strategy.

[0006] Therefore, there is an urgent need for a method that can optimize joint bidding between the energy market and the frequency regulation market, taking into account the objective flexibility boundaries of electric vehicle clusters and users' subjective willingness to respond, in order to improve the feasibility of bidding strategies and the stability of returns. Summary of the Invention

[0007] This invention provides a method, apparatus, electronic device, and medium for optimizing energy-frequency modulation bidding for electric vehicle aggregators, which can effectively solve the problems in the background art.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] An energy-frequency modulation bidding optimization method for electric vehicle aggregators includes the following steps:

[0010] Acquire historical charging and travel data of electric vehicles, and obtain survey data to characterize user response intentions, as well as multi-scenario parameters for energy market prices and frequency regulation market capacity prices;

[0011] Based on the historical charging and travel data, an electric vehicle flexibility evaluation index system including adjustable capacity, on-grid time, and charging power is constructed, and a flexible domain constraint for electric vehicle clusters is established.

[0012] Based on the aforementioned flexibility evaluation index system and the aforementioned flexible domain constraints of the electric vehicle cluster, the flexible domain parameters used to define the energy state boundary of the electric vehicle cluster are calculated.

[0013] A user response willingness assessment model is established based on the survey data, outputting the response willingness of each user, and generating a willingness indicator variable based on the response willingness to determine users who are acceptable for adjustment.

[0014] Using the flexible domain parameters and the willingness indicator variables as constraint inputs, a joint bidding optimization model for the energy market and frequency regulation market of electric vehicle aggregators is established under a multi-market price scenario.

[0015] Solve the joint bidding optimization model of the energy market and frequency regulation market for electric vehicle aggregators. With the goal of maximizing the total expected revenue of electric vehicle aggregators under the multi-scenario parameters, output the decision quantities related to charging bidding power, discharging bidding power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidies in the energy market for each time period, and form the joint bidding application results of electric vehicle aggregators in the energy market and frequency regulation market.

[0016] Furthermore, the historical charging and travel data, as well as the survey data, are preprocessed, including at least data cleaning, missing data processing, and parameterization.

[0017] Missing data processing includes mean imputation or data reconstruction.

[0018] Furthermore, in the electric vehicle flexibility evaluation index system, the adjustable capacity includes at least the average grid connection capacity obtained based on data cluster u, where data cluster u represents a set of charging records of the same user under similar behavioral patterns, and the average grid connection capacity is used to characterize the statistical features of the adjustable capacity.

[0019] Furthermore, the average grid access capacity is obtained by averaging the product of the grid access SOC of each charge within the data cluster u and the battery capacity of the electric vehicle.

[0020] The average inbound capacity under the u-th data cluster is calculated using the following formula:

[0021] ;

[0022] in, For a user's i-th charging SOC within the u-th data cluster, the network connection is defined. Indicates the battery capacity of an electric vehicle. This represents the number of charging records under the u-th data cluster.

[0023] Furthermore, the user response willingness evaluation model is a fuzzy neural network model, comprising a fuzzy rule layer and a defuzzification layer, wherein the output of node i in the defuzzification layer satisfies the following formula:

[0024] ;

[0025] Where x, y, z, u are input variables. The set of adaptive parameters for the nodes; and the desired response of the output layer satisfies the following equation:

[0026] ;

[0027] in, Let be the initial trigger strength of the i-th fuzzy rule. Let be the contribution value of the i-th fuzzy rule to the output result. The output result is the response intention, where n is the number of rule nodes.

[0028] Furthermore, the stated willingness to respond denoted as μ i Generate a willingness indicator variable based on the stated willingness to respond. The user's acceptance of the control is represented as a Boolean variable using a threshold, as shown in the following formula:

[0029] ;

[0030] in, This indicates that user i accepts regulation. This indicates that user i does not accept regulation, μ i The magnitude of user i's willingness to respond, as predicted by the fuzzy neural network;

[0031] In the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market, only... The charging and discharging power and frequency regulation capacity of electric vehicles are used as schedulable decision variables or included in the effective callable capacity.

[0032] Furthermore, the objective function of the joint bidding optimization model for the electric vehicle aggregator energy market and frequency regulation market is:

[0033] ;

[0034] Where s is the scene number, S is the total number of scenes, t is the time period number, T is the number of time periods, and a day is divided into 24 time periods, with each time period consisting of 1 hour. P s Let be the probability of the s-th scenario. , , These represent the revenue of electric vehicle aggregators participating in the energy market, the revenue of electric vehicle aggregators participating in the frequency modulation auxiliary service market, and the cost of electric vehicle aggregators participating in the energy-frequency modulation auxiliary service market, respectively, within the t-th time period under the s-th scenario.

[0035] Furthermore, electric vehicle aggregators will benefit from the energy market in scenario s and time period t. Calculated using the following formula:

[0036] ;

[0037] Where N is the number of electric vehicles in the electric vehicle cluster. , Let be the charging power and discharging power of the i-th electric vehicle in the t-th time period, respectively. The electricity price for electric vehicle aggregators under scenario s and time period t; Here, Δt represents the energy market price under scenario s and time period t; Δt is the time interval. This is an indicator variable for intention.

[0038] Furthermore, This includes the costs of electric vehicle aggregators in the energy market within scenario s and time period t. Calculated using the following formula:

[0039] ;

[0040] in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period. For energy market quotes under scenario s and time period t, , Let be the charging power and discharging power of the i-th electric vehicle in the t-th time period, respectively, and Δt be the time interval. This is an indicator variable for intention.

[0041] Furthermore, the revenue of electric vehicle aggregators in the frequency modulation auxiliary market for scenario s and time period t satisfies the following formula:

[0042] ;

[0043] Frequency modulation market capacity revenue during the t-th time period in the s-th scenario Calculated using the following formula:

[0044] ;

[0045] in, For scenario s and time period t, the frequency regulation capacity price is... , Let be the upward and downward frequency regulation capacities provided by the electric vehicle aggregator to dispatch vehicle i during the t-th time period, respectively. This is an indicator variable for intention.

[0046] Furthermore, This includes the frequency modulation market costs for electric vehicle aggregators in the t-th time period of the s-th scenario. Calculated using the following formula:

[0047] ;

[0048] in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period. , These represent the upward and downward frequency modulation capacities provided by the electric vehicle aggregator for dispatching vehicle i during the t-th time period.

[0049] Furthermore, the following constraints are imposed on the charging and discharging power of each electric vehicle:

[0050] ;

[0051] ;

[0052] ;

[0053] The charging and discharging power submitted by the electric vehicle aggregator in time period t is obtained by aggregating the power of each electric vehicle:

[0054] ;

[0055] in, The upper limit of charging power for the i-th electric vehicle. Let be the upper limit of the discharge power of the i-th electric vehicle.

[0056] Furthermore, the joint bidding optimization model uses the charging power, discharging power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy decision amount of each electric vehicle in each time period as optimization variables, and satisfies the charging and discharging power constraints of each electric vehicle, the dynamic constraints of battery energy state, the frequency regulation power margin constraints, the frequency regulation capacity coupling constraints, and the flexible domain constraints of the electric vehicle cluster.

[0057] Furthermore, the adjustable capacity includes at least the average grid connection capacity obtained based on data cluster u, where data cluster u represents a set of charging records of the same user under similar behavior patterns, and the average grid connection capacity is obtained by averaging the product of the grid connection SOC of each charging session within data cluster u and the battery capacity of the electric vehicle.

[0058] The dynamic constraint on battery energy state and the coupling constraint on frequency modulation capacity satisfy the following equation:

[0059] ;

[0060] ;

[0061] ;

[0062] ;

[0063] ;

[0064] ;

[0065] in, Let be the battery capacity of the i-th electric vehicle during time period t; , These are the charging efficiency coefficient and the discharging efficiency coefficient, respectively. Let be the power of the i-th electric vehicle within time period t; , These are the upper and lower limits of capacity. , These represent the upward and downward frequency modulation capacities provided by the electric vehicle aggregator for dispatching vehicle i during the t-th time period.

[0066] Furthermore, the electric vehicle aggregator provides upward frequency modulation bidding capacity for vehicles scheduled during time period t. With downward frequency modulation bidding capacity Up-frequency modulation capacity provided by each electric vehicle in time period t With down-modulation capacity The aggregation yields the following formula:

[0067] ;

[0068] Total capacity of electric vehicle cluster in time period t As shown in the following formula:

[0069] ;

[0070] ;

[0071] in, The electric vehicle cluster satisfies the flexible domain constraints of the Chino polyhedron form. , and These represent the center point, generator, and maximum scaling factor of the flexible domain of the electric vehicle cluster in the form of a Chino polyhedron within time period t.

[0072] An electric vehicle aggregator energy-frequency modulation bidding optimization device, comprising:

[0073] The data acquisition module is used to acquire historical charging and travel data of electric vehicles, survey data to characterize user response intentions, and multi-scenario parameters of energy market prices and frequency regulation market capacity prices;

[0074] The flexibility characterization module is used to construct an electric vehicle flexibility evaluation index system that includes adjustable capacity, on-grid time and charging power based on the historical charging data and travel data, establish flexible domain constraints for electric vehicle clusters, and calculate the flexible domain parameters used to limit the energy state boundary of electric vehicle clusters.

[0075] The response willingness assessment module is used to establish a user response willingness assessment model based on the survey data, output the response willingness of each user, and generate a willingness indicator variable based on the response willingness to determine the set of users that can be controlled.

[0076] The joint bidding optimization module is used to establish and solve the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market under the multi-market price scenario, using the flexible domain parameters and the willingness indicator variables as constraint inputs.

[0077] The output module is used to output the energy market charging bid power, discharging bid power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy decision amount for each time period, forming the joint bidding application results.

[0078] Furthermore, it also includes a data preprocessing module, used to perform data cleaning, missing data processing, and parameterization on the historical charging data, travel data, and survey data; wherein, the missing data processing includes mean interpolation or data reconstruction.

[0079] Furthermore, in the flexibility evaluation index system constructed by the flexibility characterization module, the adjustable capacity includes at least the average network access capacity obtained based on data cluster u, where data cluster u represents a set of charging records of the same user under similar behavioral patterns, and the average network access capacity is used to characterize the statistical features of the adjustable capacity.

[0080] Furthermore, the flexibility characterization module obtains the average grid-connected capacity by averaging the product of the grid-connected state of charge (SOC) and the electric vehicle battery capacity for each charge within the data cluster u.

[0081] Furthermore, the response willingness assessment module uses a fuzzy neural network to assess user response willingness. The fuzzy neural network includes a fuzzy rule layer and a defuzzification layer. The defuzzification layer outputs the contribution value of each rule node based on the input variables and node adaptive parameters. The output layer outputs the user response willingness based on the trigger strength of each fuzzy rule and the contribution value.

[0082] Furthermore, the response willingness assessment module generates a willingness indicator variable based on user response willingness and a preset threshold, and determines the set of users that can be controlled based on the willingness indicator variable. The joint bidding optimization module only allows electric vehicles corresponding to the user set to participate in the decision-making of charging and discharging power and frequency regulation capacity or to be included in the effective callable capacity.

[0083] Furthermore, the joint bidding optimization module establishes the joint bidding optimization model with the goal of maximizing the total expected revenue of electric vehicle aggregators under multiple scenario parameters. The total expected revenue is jointly determined by the energy market revenue, frequency regulation auxiliary market revenue, and the cost of participating in the energy-frequency regulation auxiliary service market at each time period under each price scenario.

[0084] Furthermore, the joint bidding optimization module includes an energy market revenue calculation submodule and an energy market cost calculation submodule. The energy market revenue calculation submodule is used to determine the energy market revenue based on the charging power, discharging power, electricity sales price, energy market quotation, and time interval of each electric vehicle in each time period. The energy market cost calculation submodule is used to determine the energy market cost based on the subsidy price paid to users subject to regulation in each time period.

[0085] Furthermore, the joint bidding optimization module includes a frequency modulation market revenue calculation submodule and a frequency modulation market cost calculation submodule. The frequency modulation market revenue calculation submodule is used to determine the frequency modulation auxiliary market revenue based on the frequency modulation capacity price and the upward and downward frequency modulation capacity provided by each electric vehicle. The frequency modulation market cost calculation submodule is used to determine the frequency modulation market cost based on the subsidy price paid to users accepting regulation in each time period.

[0086] Furthermore, the joint bidding optimization module uses the charging power, discharging power, upward frequency modulation capacity, downward frequency modulation capacity, and user subsidy decision amount of each electric vehicle in each time period as optimization variables, and satisfies the charging and discharging power constraints of each electric vehicle, the dynamic constraints of battery energy state, the frequency modulation power margin constraints, the frequency modulation capacity coupling constraints, and the flexible domain constraints of electric vehicle clusters.

[0087] Furthermore, the joint bidding optimization module is also used to apply the following constraints to the charging and discharging power of each electric vehicle: the charging power of each electric vehicle in each time period does not exceed the corresponding charging power limit, the discharging power does not exceed the corresponding discharging power limit, and the charging state and the discharging state are mutually exclusive; and the charging power of each electric vehicle is aggregated to obtain the charging bidding power for time period t, and the discharging power of each electric vehicle is aggregated to obtain the discharging bidding power for time period t.

[0088] Furthermore, the joint bidding optimization module is also used for:

[0089] The battery state of energy for time period t is determined based on the battery state of energy for time period t-1, the charging power, discharging power, charging efficiency, discharging efficiency for time period t, and the time interval.

[0090] The upward frequency modulation capacity and downward frequency modulation capacity are limited according to the charging and discharging operating point and power limit of each electric vehicle, and the upward frequency modulation capacity and downward frequency modulation capacity are mutually exclusive.

[0091] The upward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the upward frequency modulation bidding capacity for time period t, and the downward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the downward frequency modulation bidding capacity for time period t.

[0092] Where Et represents the total capacity of the electric vehicle cluster in time period t, and Et satisfies the flexible domain constraint of the electric vehicle cluster in the form of a chino polyhedron, which is characterized by the flexible domain center point, the generator matrix and the maximum scaling factor.

[0093] An electronic device includes a processor and a memory, wherein the memory stores a computer program that, when executed by the processor, implements the method described above.

[0094] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0095] The technical solution of this invention can achieve the following technical effects:

[0096] This invention establishes a joint bidding optimization model for the energy market and the frequency regulation market, and simultaneously makes decisions on charging / discharging bidding power, upward / downward frequency regulation bidding capacity, and subsidy-related decisions for each time period under the same optimization framework. This unifies and coordinates the occupation relationship of energy and frequency regulation for the same power and energy state resources, avoids resource conflicts caused by independent optimization of separate markets, and thus improves the overall revenue level and resource allocation efficiency of aggregators.

[0097] Based on historical charging and travel data, a flexibility assessment index system is constructed and a cluster flexibility domain constraint is formed. Further, the flexibility domain parameters that limit the energy state boundary of the cluster are obtained. These parameters are used as constraint inputs for the optimization model, which can effectively limit the scheduling results from exceeding the cluster's feasible domain, reduce the risk of execution failure due to overestimation of available flexibility, and improve the feasibility and reliability of the bidding scheme.

[0098] By using a fuzzy neural network to output user response intention μ_i and generating an intention indicator variable θ_i based on a threshold to determine the set of users that can be controlled, only resources that meet the intention conditions are included in the schedulable decision and effective capacity statistics. This reduces the risk of insufficient performance caused by low-willing users being unable to respond from the source, and improves the consistency between bidding results and actual execution, as well as the stability of revenue. Attached Figure Description

[0099] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0100] Figure 1 This is a schematic diagram illustrating the mechanism by which electric vehicles participate in the energy-frequency regulation market operation according to the present invention.

[0101] Figure 2 This is a schematic diagram of the electric vehicle user flexibility evaluation index system of the present invention;

[0102] Figure 3 This is a schematic diagram illustrating an example of the flexible domain of an electric vehicle in an embodiment of the present invention;

[0103] Figure 4 This is a schematic diagram illustrating the construction of the upper boundary of the flexible domain in an embodiment of the present invention;

[0104] Figure 5 This is a schematic diagram of the construction of the lower boundary of the flexible domain in a scenario according to an embodiment of the present invention;

[0105] Figure 6 This is a schematic diagram of the construction of the lower boundary of the flexible domain in scenario two of the embodiments of the present invention. Detailed Implementation

[0106] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0107] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0108] This invention discloses a method and apparatus for optimizing energy-frequency modulation bidding for electric vehicle aggregators, such as... Figure 1 The diagram illustrates the operational mechanism of electric vehicles (EVs) participating in the energy-frequency regulation market. It addresses a scenario where EV aggregators (EVAs) simultaneously participate in joint bidding for both the energy market and the frequency regulation ancillary service market. The model integrates the objective flexibility of EV clusters with users' subjective willingness to respond, thereby maximizing the total expected revenue under uncertain multi-price scenarios by outputting bidding power, frequency regulation capacity, and subsidy decision quantities. Joint energy-frequency regulation bidding explicitly depicts the competition and complementarity between the two markets, which helps increase the overall revenue ceiling for EVAs. Incorporating user willingness to respond into the bidding optimization as an executable constraint / weight reduces the performance risk caused by "bidding-execution deviation." By using subsidy prices as a decision quantity and influencing willingness prediction, a closed-loop optimization of "subsidy-willingness-available resources-revenue" is achieved, improving the strategy's feasibility and robustness.

[0109] In this embodiment, historical charging and travel data of electric vehicles are first acquired, along with survey data representing user response intentions, and multi-scenario parameters for energy market prices and frequency regulation market capacity prices are obtained. The historical charging and travel data, as well as the survey data, are preprocessed, including at least data cleaning, missing data processing, and parameterization. Mean imputation is used for a small amount of missing data, and reconstruction is performed based on existing data for a large amount of missing data.

[0110] Specifically, historical charging and travel data are cleaned: obvious outliers such as SOC exceeding limits, timestamp errors, and power sign confusion are removed, and the time granularity is standardized; the day is divided into 24 time periods, with each time period consisting of 1 hour.

[0111] Missing values ​​are derived by utilizing the behavior of the same user in adjacent time periods, the statistical distribution of similar vehicles, or the consistency of charging pile records to ensure the sample completeness of flexible domain construction and willingness assessment. Market price scenario parameters can be obtained by clustering / sampling multiple scenario sets based on historical market price series, and a probability is assigned to each scenario for subsequent expected return maximization objectives.

[0112] In the specific implementation process, basic user attributes, including gender, age, income, vehicle usage, and battery capacity, are used for sample statistics and heterogeneity characterization; user behavioral attributes and subsidy psychological thresholds, including onboarding SOC, expected SOC, baseline SOC, onboarding time, onboard time, charging power, and acceptable subsidy price, are used for the design and training of input features for fuzzy neural networks.

[0113] In this embodiment, an electric vehicle flexibility evaluation index system including adjustable capacity, on-grid time, and charging power is constructed based on historical charging and travel data, and flexible domain parameters of the electric vehicle cluster are established. Based on the flexibility evaluation index system and the flexible domain constraints of the electric vehicle cluster, flexible domain parameters used to limit the energy state boundary of the electric vehicle cluster are calculated.

[0114] Specifically, such as Figure 2 As shown, in the electric vehicle flexibility evaluation index system, the adjustable capacity includes at least the average grid connection capacity obtained based on the data cluster u. The data cluster u represents the set of charging records of the same user under similar behavior patterns. The average grid connection capacity is used to characterize the statistical features of the adjustable capacity and is obtained by taking the average of the product of the grid connection SOC of each charging in the data cluster u and the electric vehicle battery capacity.

[0115] The average inbound capacity under the u-th data cluster is calculated using the following formula:

[0116] ;

[0117] in, For a user's i-th charging SOC within the u-th data cluster, the network connection is defined. Indicates the battery capacity of an electric vehicle. This represents the number of charging records under the u-th data cluster.

[0118] Furthermore, a flexible domain is constructed for a single electric vehicle. A flexible domain represents the area within which the EV can perform charging and discharging scheduling, such as... Figure 3 As shown in the figure, the slopes of the hypotenuses in the upper and lower boundaries correspond to the charging and discharging power, respectively. A slope of zero indicates zero power. The number, location, and shape of the flexible domain for each EV user reflect the distribution and magnitude of user flexibility.

[0119] In the specific implementation process, see Figure 3As shown, the flexible domain of the γth EV user in the uth data cluster is defined by the upper boundary. and lower boundary Composition. In this example, the upper boundary. This indicates that the battery will be charged to full capacity immediately after joining the network, then charging will stop, and finally the battery will disconnect from the network. (Lower boundary) This indicates that after joining the network, the battery first discharges to the baseline capacity, then stops discharging, and finally undergoes forced charging before disconnecting from the network to meet the user's travel needs. (See diagram) , , , It can be obtained directly from the indicator system. , , , Calculated using the following formulas respectively:

[0120] ;

[0121] ;

[0122] ;

[0123] ;

[0124] in, This refers to the end time of EV charging to maximum battery capacity under data cluster u. This value represents the cluster center of data cluster u after clustering by network access time, and it represents the user's habitual network access time. This represents the time it takes for the EV to discharge to the reference capacity under data cluster u. For discharge power, The start time of the forced charging process of the EV under data cluster u. The last time a data cluster under u was disconnected from the network. Let u be the average online time of the vehicle corresponding to the u-th data cluster. Let be the average charging power of the vehicle corresponding to the u-th data cluster.

[0125] The variables in the above formula, together with the EV flexibility index, constitute the upper and lower boundaries of the EV flexibility domain. Due to limitations such as charging power, on-grid time, and grid capacity, these variables will vary, resulting in different shapes for the EV flexibility domain. By analyzing the relationships between the variables in the above formula and calculating and connecting several key points on the upper and lower boundaries, the EV flexibility domain can be obtained.

[0126] The upper boundary of the flexible domain can be calculated using the following formula. If the offline time... Later than or equal to the time it takes to fully charge This indicates that the EV can be fully charged before leaving the grid. In this case, the upper boundary is as follows: Figure 4 As shown by the midline A, and conversely, as shown by line B.

[0127] ;

[0128] in, The off-grid capacity after charging but not fully charged is calculated using the following formula:

[0129] ;

[0130] The construction of the lower boundary differs from that of the upper boundary and needs to be considered in two scenarios: Scenario 1: EVs can be discharged directly after being connected to the grid; Scenario 2: EVs need to be charged before discharging after being connected to the grid. In Scenario 1, the average grid-connected capacity in data cluster u... Higher than the baseline capacity The lower boundary is then calculated using the following formula:

[0131] ;

[0132] in, The off-grid capacity when the battery is disconnected before reaching the reference capacity is calculated using the following formula:

[0133] ;

[0134] Time required to discharge to the reference capacity Earlier than or equal to the start time of the forced charging process This indicates that the EV can discharge to its reference capacity before being taken off the grid. It then enters the forced charging phase, charging to the desired off-grid capacity. At this time, the lower boundary is as follows Figure 5 As shown by line A. Conversely, the time required to discharge to the reference capacity. Later than the start time of the forced charging process But earlier than or equal to the time of disconnection This means that the EV is immediately recharged after the discharge process ends, but it cannot be guaranteed that the EV will discharge to the baseline capacity during the discharge process. At this point, the lower boundary is as follows: Figure 5 As shown by line B, this represents the time required to discharge to the reference capacity. Later than the time of disconnection At that time, the lower boundary is as follows Figure 5 As shown by the midline C.

[0135] In scenario two, the average inbound capacity in data cluster u Below the baseline capacity This means that EVs need to be charged after being registered on the network before they can be scheduled for operation. The inflection point at the lower boundary in Scenario 2 is calculated using the following formula:

[0136] ;

[0137] in, The time required for an EV to reach its base capacity and then cease charging after being connected to the grid is calculated using the following formula:

[0138] ;

[0139] When the energy replenishment time ends Earlier than or equal to the start time of the forced charging process At this point, the EV can undergo a forced charging process before being disconnected from the grid, charging to the user's desired off-grid capacity. At this time, the lower boundary is as follows Figure 6 As shown by the midline A, and conversely, as shown by the midline B.

[0140] By calculating and connecting the points in the upper and lower boundaries according to the above rules, the boundary lines can be formed. By combining the upper and lower boundaries, the flexible domain of a user at a certain network access time can be obtained, thereby realizing the flexible characterization of a single user.

[0141] The flexible domain is approximated using the Chino polyhedron, and the approximation results are aggregated using Minkowski summation to obtain the flexible domain of the electric vehicle cluster.

[0142] In this embodiment, to fully characterize the adjustable capacity and time domain boundary, the baseline capacity, expected off-grid capacity, average on-grid time, and average charging power are further calculated; baseline capacity: the minimum capacity to meet travel needs and protect battery health; expected off-grid capacity: reflects the user's SOC preference when off-grid; average on-grid time: reflects the sustainable scheduling time window; average charging power: reflects the response power capability.

[0143] Base capacity To ensure the minimum battery capacity required for user travel, the following formula is used:

[0144] ;

[0145] in, The minimum SOC value to protect battery health; The SOC consumed for the b-th trip; This represents the number of travel records.

[0146] Expected AFLO capacity of EV users under the u-th data cluster of a certain user Calculated by the following formula:

[0147] ;

[0148] in, The off-grid SOC for the i-th charge under the u-th data cluster.

[0149] The average online time for a user in the u-th data cluster is calculated by the following formula:

[0150] ;

[0151] in, The on-network time for the i-th charge under the u-th data cluster.

[0152] The average charging power of a user in the u-th data cluster is calculated by the following formula:

[0153] ;

[0154] in, The charging power for the i-th charge under the u-th data cluster. This indicates that the charging power of the same user varies depending on when they join the network.

[0155] In practical implementation, the construction of the flexible domain for a single EV involves constructing a feasible scheduling area for each electric vehicle in the "time-energy (or SOC)" space, using factors such as grid access time, grid exit time, grid access SOC, expected SOC / off-grid SOC, baseline capacity, and power limit as boundaries. The slope of the flexible domain boundary corresponds to the charging and discharging power constraints; a slope of zero indicates zero power.

[0156] Two typical scenarios are: Scenario A: After grid connection, the device can directly discharge to participate in frequency regulation / energy trading; Scenario B: After grid connection, the device must first be charged to the base capacity CapminCapmin before it can discharge or provide upward frequency regulation capability. By comparing the sequential relationship between the "off-grid time", "forced charging end time", and "discharge start time", the segmented shape of the flexible domain boundary is determined.

[0157] The clustered flexible domain aggregates the individual flexible domains of multiple EVs at the cluster level and represents the flexible domain parameters for time period t using a Chino polyhedron (commonly in the form of a center point + generator). , and This is used to constrain the energy state of the cluster from going out of bounds.

[0158] In this embodiment, a user response willingness assessment model is established based on survey data, outputs the response willingness of each user, and generates a willingness indicator variable based on the response willingness to determine the set of users that can be controlled.

[0159] Specifically, to establish a user response willingness assessment model, the first step is to identify the influencing factors, collect user questionnaires, analyze and rank the importance of the influencing factors to users, select the factors with higher importance as input variables of the fuzzy neural network, and use the degree of user response willingness as the output variable.

[0160] Secondly, membership functions are constructed, with the expected SOC value falling within the [0,1] interval. Membership functions are determined based on survey data, categorizing users into high, medium, and low categories, and establishing corresponding membership functions. For subsidized electricity prices, user attitudes towards subsidized electricity prices are analyzed using questionnaire data, categorizing users into high, medium, and low subsidized electricity price categories, and establishing membership functions for each category. For grid connection time, based on questionnaire data, grid connection time is divided into long (more than 4 hours) and short (less than 4 hours) categories, and corresponding membership functions are established. For grid connection SOC, the value falls within the [0,1] interval, and based on survey data, it is categorized into high, medium, and low categories, and membership functions are established.

[0161] This paper designs a fuzzy neural network structure, which typically consists of five layers: an input layer, a fuzzification layer, a fuzzy rule layer, a defuzzification layer, and an output layer. To reduce model complexity, based on questionnaire analysis results, four main influencing factors—expected SOC, subsidy price, network time, and initial SOC—are selected as the training inputs for the fuzzy neural network.

[0162] Input layer: When evaluating EV user willingness, this paper uses four nodes to accept data input based on expected SOC, subsidy price, network time, and network access SOC, which serve as the basis for fuzzification and rule construction.

[0163] The fuzzification layer transforms the four parameter variables—expected SOC, subsidy price, network time, and entry SOC—into linguistic variables. Based on the questionnaire survey analysis, the initial domain of expected SOC is [0,1], and the linguistic set is [L,M,H]; the initial domain of subsidy price is [0,2], and the linguistic set is [L,M,H]; the initial domain of network time is [0,12], and the linguistic set is [L,H]; and the initial domain of entry SOC is [0,1], and the linguistic set is [L,M,H], where [L,M,H] represent low, medium, and high linguistic variables, respectively. Based on the different linguistic sets of each variable, the fuzzy layer constructs four nodes to calculate the membership degree of each variable in the fuzzy linguistic set, meaning each node includes the fuzzy linguistics under its respective fuzzy function. Now, assume that node i has the following fuzzy function:

[0164] ;

[0165] ;

[0166] ;

[0167] ;

[0168] Where x, y, z, u are the inputs to node i, such as expected SOC, subsidy price, etc. , , , It is a fuzzy set. , , , These are membership functions. , , , The membership value is used to measure whether variables x, y, z, u belong to a certain group or group. , , , The degree to which the input belongs is determined by calculating the membership degree of the input's expected SOC in each language set, such as the degree to which it belongs to high / medium / low expected SOC.

[0169] Fuzzy Rule Layer: Each node in this layer is associated with one of the fuzzy sets of nodes in the previous layer. Since the four nodes in the fuzzification layer contain 3, 3, 2, and 3 fuzzy language sets respectively, the fuzzy rule layer has a total of 54 nodes. For example, node i represents: high expected SOC, high subsidy price, high network time, and high onboard SOC. This requires the fuzzy neural network to construct 54 inference rules. The fuzzy rule layer multiplies the outputs of the fuzzy layers. The confidence level of the i-th rule output by node i is shown in the following formula:

[0170] ;

[0171] To ensure uniformity in the strength of the fuzzy rules output by each variable, the obtained confidence levels are normalized, as shown in the following formula:

[0172] ;

[0173] Where n is the number of nodes, This represents the credibility normalization result of the i-th rule, which is used for rule matching.

[0174] Further analysis of the impact on EV users' willingness to respond, combined with the directionality of various factors, yields some fuzzy inference rules designed in this paper, as shown in Table 1 below. For example, the first rule indicates that when users have low expected SOC, low subsidy price, short network time, and low initial SOC, their willingness to respond is low; the third rule indicates that when users have low expected SOC, low subsidy price, short network time, and high initial SOC, their willingness to respond is moderate; and the sixth rule indicates that when network time is longer, users' willingness to respond increases. The meanings of the remaining rules are the same as above.

[0175] Table 1 Fuzzy Rules

[0176]

[0177] Defuzzification Layer: The user response willingness evaluation model is a fuzzy neural network model, which includes a fuzzy rule layer and a defuzzification layer. The output of node i in the defuzzification layer satisfies the following formula:

[0178] ;

[0179] Where x, y, z, u are input variables. The set of adaptive parameters for the nodes; and the desired response of the output layer satisfies the following equation:

[0180] ;

[0181] in, Output the results to reflect the user's response intentions, where n is the number of rule nodes.

[0182] Furthermore, user response intentions denoted as μ i The threshold value is used to represent the user's acceptance of the control as a Boolean variable. As shown in the following formula:

[0183] ;

[0184] in, This indicates that user i accepts regulation. This indicates that user i does not accept regulation, μ i The fuzzy neural network predicts the response intention of user i; in the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market, only the charging and discharging power and frequency regulation capacity of the corresponding electric vehicle are allowed as dispatchable decision variables or included in the effective callable capacity.

[0185] In the specific implementation process, input variables may include: onboard SOC, expected SOC / offboard SOC, baseline SOC, onboard time, onboard time, charging power, and subsidy price, etc. Among them, the subsidy variable not only affects users' psychological acceptance but is also one of the key decision-making factors for subsequent optimization. The fuzzification layer can set low, medium, and high levels for the input variables, and the rule output is linearly combined and adaptively calibrated by the defuzzification layer; the response intention is obtained by summarizing the results from the output layer. The training method can adopt a hybrid learning approach combining least squares and gradient descent or other equivalent methods, with the goal of minimizing the error between the predicted intention and the questionnaire / historical response behavior, thereby improving the reliability of the intention prediction.

[0186] Obtained through thresholding Subsequent optimizations can directly remove users with low willingness or reduce their available capacity, making the bidding strategy more consistent with actual execution and reducing assessment risks.

[0187] In this embodiment, a joint bidding optimization model for energy and frequency regulation is performed. Flexible domain parameters and willingness indicator variables are used as constraint inputs to establish a joint bidding optimization model for the energy market and frequency regulation market of electric vehicle aggregators under a multi-market price scenario. The joint bidding optimization model uses the charging power, discharging power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy-related decision quantities of each electric vehicle at each time period as optimization variables, satisfying constraints on the charging and discharging power of each electric vehicle, dynamic constraints on battery energy state, frequency regulation power margin constraints, and flexible domain constraints of the electric vehicle cluster. Solving the joint bidding optimization model for the energy market and frequency regulation market aims to maximize the total expected revenue of the electric vehicle aggregator under multiple scenario parameters. The output includes the charging bidding power, discharging bidding power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy-related decision quantities in the energy market at each time period, generating a corresponding electric vehicle cluster scheduling plan. Specifically, the charging bidding power and discharging bidding power are obtained by aggregating the charging power and discharging power of each electric vehicle, and the upward frequency regulation capacity and downward frequency regulation capacity are obtained by aggregating the upward frequency regulation capacity and downward frequency regulation capacity of each electric vehicle.

[0188] With the objective of maximizing the total expected revenue of electric vehicle aggregators under multiple market price scenarios, the model outputs the energy market charging and discharging bidding power, frequency regulation bidding capacity, and decision quantities related to user subsidies for each time period; the objective function of the joint bidding optimization model is:

[0189] ;

[0190] Where s is the scene number, S is the total number of scenes, t is the time period number, T is the number of time periods, and a day is divided into 24 time periods, with each time period consisting of 1 hour. P s Let be the probability of the s-th scenario. , , These represent the revenue of electric vehicle aggregators participating in the energy market, the revenue of electric vehicle aggregators participating in the frequency modulation auxiliary service market, and the cost of electric vehicle aggregators participating in the energy-frequency modulation auxiliary service market, respectively, within the t-th time period under the s-th scenario.

[0191] Electric vehicle aggregators' participation in the energy market benefits Calculated using the following formula:

[0192] ;

[0193] Where N is the number of electric vehicles in the electric vehicle cluster. , Let be the charging and discharging power of the i-th electric vehicle in the t-th time period, respectively. The electricity price for electric vehicle aggregators under scenario s and time period t; Here are the energy market quotes for scenario s and time period t; Δt is the time interval.

[0194] Costs of electric vehicle aggregators in the energy market Calculated using the following formula:

[0195] ;

[0196] in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period.

[0197] Subsidized prices On the one hand, it enters the cost item; on the other hand, it can serve as an input variable for the fuzzy neural network, influencing μ. i That is, influence This, along with available resources, forms a closed-loop optimization mechanism of "subsidy adjustment - change in willingness - change in available resources - change in revenue".

[0198] Furthermore, the revenue generated by electric vehicle aggregators participating in the frequency modulation auxiliary market satisfies the following formula:

[0199] ;

[0200] Frequency modulation market capacity revenue during the t-th time period in the s-th scenario Calculated using the following formula:

[0201] ;

[0202] in, For scenario s and time period t, the frequency regulation capacity price is... , These represent the upward and downward frequency regulation capacities provided by the electric vehicle aggregator for dispatching vehicle i during the t-th time period;

[0203] Frequency modulation market cost in the t-th time period under the s-th scenario Calculated using the following formula:

[0204] ;

[0205] in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period.

[0206] Will The introduction of frequency modulation capacity revenue calculation ensures that only the capacity contributed by users with high willingness is included in the effective revenue, thus avoiding insufficient actual performance due to "paper capacity".

[0207] Furthermore, the following constraints are imposed on the charging and discharging power of each electric vehicle:

[0208] ;

[0209] ;

[0210] ;

[0211] The charging and discharging power submitted by the electric vehicle aggregator in time period t is obtained by aggregating the power of each electric vehicle:

[0212] ;

[0213] in, The upper limit of charging power for the i-th electric vehicle. Let be the upper limit of the discharge power of the i-th electric vehicle.

[0214] The dynamic constraints on electric vehicle battery capacity and the coupling constraints on frequency regulation capacity satisfy the following equation:

[0215] ;

[0216] The relationship between the vehicle's net power and its charging / discharging power components satisfies the following equation:

[0217] ;

[0218] Apply a power margin constraint to the frequency modulation capacity:

[0219] ;

[0220] ;

[0221] The mutual exclusion constraints for up and down frequency modulation are:

[0222] ;

[0223] The upper and lower limits of battery capacity are constrained as follows:

[0224] ;

[0225] in, Let be the battery capacity of the i-th electric vehicle during time period t; , These are the charging efficiency coefficient and the discharging efficiency coefficient, respectively. Let be the power of the i-th electric vehicle within time period t; , These are the upper and lower limits of capacity.

[0226] The frequency modulation bidding capacity is obtained by aggregating the frequency modulation capacities of all electric vehicles:

[0227] ;

[0228] The electric vehicle cluster capacity meets the following requirements:

[0229] ;

[0230] ;

[0231] in, For electric vehicle cluster capacity, , and These represent the center point, generator, and maximum scaling factor of the flexible domain of the Cino polyhedral electric vehicle cluster within time period t. Ensuring this by limiting the scaling factor ensures... It does not exceed the boundaries of the flexible domain.

[0232] In this embodiment, model solving and bidding output are performed. After solving the optimization problem, the bidding power for charging and discharging in the energy market, the bidding capacity for frequency regulation, and the subsidy decision amount for each time period are output, and the bidding application results and the corresponding electric vehicle cluster scheduling plan are formed.

[0233] Because the model contains mutually exclusive constraints and possible piecewise / threshold variables, this embodiment can linearize the mutually exclusive relationships by introducing 0-1 variables, such as the Big-M method, to construct a mixed-integer optimization model; alternatively, equivalent decomposition algorithms or heuristic algorithms can be used for solving. In engineering implementation, commercial or open-source solvers can be called to complete the day-ahead calculation.

[0234] In the specific implementation process, EVA participates in the day-ahead pre-clearing: receiving the frequency regulation demand signal and bidding information for the next day from the dispatch center; calculating and submitting the energy bidding power and frequency regulation bidding capacity for each time period based on the model of this invention; after obtaining the winning bid capacity for each time period of the day-ahead clearing, it enters the intraday actual clearing: receiving the real-time clearing signal and frequency regulation demand signal, performing secondary correction on the response signal according to the flexible domain constraints and user response willingness, and allocating the corrected regulation demand to vehicles within the set of acceptable regulation users for execution; finally completing the clearing settlement and execution feedback, and the execution data can be fed back to update the willingness model and flexible domain parameters to form rolling optimization.

[0235] This invention further discloses a joint bidding optimization device for electric vehicle aggregators participating in the energy-frequency regulation market, taking into account flexibility and user responsiveness, comprising:

[0236] The data acquisition module is used to acquire historical charging and travel data of electric vehicles, survey data to characterize user response intentions, and multi-scenario parameters of energy market prices and frequency regulation market capacity prices;

[0237] The flexibility characterization module is used to construct an electric vehicle flexibility evaluation index system that includes adjustable capacity, on-grid time and charging power based on the historical charging data and travel data, establish flexible domain constraints for electric vehicle clusters, and calculate the flexible domain parameters used to limit the energy state boundary of electric vehicle clusters.

[0238] The response willingness assessment module is used to establish a user response willingness assessment model based on the survey data, output the response willingness of each user, and generate a willingness indicator variable based on the response willingness to determine the set of users that can be controlled.

[0239] The joint bidding optimization module is used to establish and solve the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market under the multi-market price scenario, using the flexible domain parameters and the willingness indicator variables as constraint inputs.

[0240] The output module is used to output the energy market charging bid power, discharging bid power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy decision amount for each time period, forming the joint bidding application results.

[0241] It also includes a data preprocessing module, used to perform data cleaning, missing data processing, and parameterization on the historical charging data, travel data, and survey data; wherein, the missing data processing includes mean interpolation or data reconstruction.

[0242] In the flexibility assessment index system constructed by the flexibility characterization module, the adjustable capacity includes at least the average network access capacity obtained based on data cluster u. The data cluster u represents the charging record set of the same user under similar behavior patterns, and the average network access capacity is used to characterize the statistical features of the adjustable capacity.

[0243] The flexibility characterization module obtains the average grid-connected capacity by averaging the product of the grid-connected state of charge (SOC) and the electric vehicle battery capacity for each charge within the data cluster u.

[0244] The response willingness assessment module uses a fuzzy neural network to assess user response willingness. The fuzzy neural network includes a fuzzy rule layer and a defuzzification layer. The defuzzification layer outputs the contribution value of each rule node based on the input variables and node adaptive parameters. The output layer outputs the user response willingness based on the trigger strength of each fuzzy rule and the contribution value.

[0245] The response willingness assessment module generates a willingness indicator variable based on user response willingness and a preset threshold, and determines the set of users that can be controlled based on the willingness indicator variable. The joint bidding optimization module only allows electric vehicles corresponding to the user set to participate in the decision-making of charging and discharging power and frequency regulation capacity or to be included in the effective callable capacity.

[0246] The joint bidding optimization module establishes the joint bidding optimization model with the goal of maximizing the total expected revenue of electric vehicle aggregators under multiple scenario parameters. The total expected revenue is jointly determined by the energy market revenue, frequency regulation auxiliary market revenue, and the cost of participating in the energy-frequency regulation auxiliary service market at each time period under each price scenario.

[0247] The joint bidding optimization module includes an energy market revenue calculation submodule and an energy market cost calculation submodule. The energy market revenue calculation submodule is used to determine the energy market revenue based on the charging power, discharging power, electricity sales price, energy market quotation, and time interval of each electric vehicle in each time period. The energy market cost calculation submodule is used to determine the energy market cost based on the subsidy price paid to users subject to regulation in each time period.

[0248] The joint bidding optimization module includes a frequency modulation market revenue calculation submodule and a frequency modulation market cost calculation submodule. The frequency modulation market revenue calculation submodule is used to determine the frequency modulation auxiliary market revenue based on the frequency modulation capacity price and the upward and downward frequency modulation capacity provided by each electric vehicle. The frequency modulation market cost calculation submodule is used to determine the frequency modulation market cost based on the subsidy price paid to users accepting regulation in each time period.

[0249] The joint bidding optimization module uses the charging power, discharging power, upward frequency modulation capacity, downward frequency modulation capacity, and user subsidy decision amount of each electric vehicle in each time period as optimization variables, and satisfies the charging and discharging power constraints of each electric vehicle, the dynamic constraints of battery energy state, the frequency modulation power margin constraints, the frequency modulation capacity coupling constraints, and the flexible domain constraints of electric vehicle clusters.

[0250] The joint bidding optimization module is also used to apply the following constraints to the charging and discharging power of each electric vehicle: the charging power of each electric vehicle in each time period shall not exceed the corresponding charging power limit, the discharging power shall not exceed the corresponding discharging power limit, and the charging state and the discharging state shall be mutually exclusive; and the charging power of each electric vehicle shall be aggregated to obtain the charging bidding power for time period t, and the discharging power of each electric vehicle shall be aggregated to obtain the discharging bidding power for time period t.

[0251] The joint bidding optimization module is also used for:

[0252] The battery state of energy for time period t is determined based on the battery state of energy for time period t-1, the charging power, discharging power, charging efficiency, discharging efficiency for time period t, and the time interval.

[0253] The upward frequency modulation capacity and downward frequency modulation capacity are limited according to the charging and discharging operating point and power limit of each electric vehicle, and the upward frequency modulation capacity and downward frequency modulation capacity are mutually exclusive.

[0254] The upward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the upward frequency modulation bidding capacity for time period t, and the downward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the downward frequency modulation bidding capacity for time period t.

[0255] Where Et represents the total capacity of the electric vehicle cluster in time period t, and Et satisfies the flexible domain constraint of the electric vehicle cluster in the form of a chino polyhedron, which is characterized by the flexible domain center point, the generator matrix and the maximum scaling factor.

[0256] The present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, which, when executed by the processor, implements the above-described method.

[0257] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0258] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0259] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0260] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0261] In the embodiments provided by this invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0262] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for optimizing energy-frequency modulation bidding for electric vehicle aggregators, characterized in that, Includes the following steps: Acquire historical charging and travel data of electric vehicles, and obtain survey data to characterize user response intentions, as well as multi-scenario parameters for energy market prices and frequency regulation market capacity prices; Based on the historical charging and travel data, an electric vehicle flexibility evaluation index system including adjustable capacity, on-grid time, and charging power is constructed, and a flexible domain constraint for electric vehicle clusters is established. Based on the aforementioned flexibility evaluation index system and the aforementioned flexible domain constraints of the electric vehicle cluster, the flexible domain parameters used to define the energy state boundary of the electric vehicle cluster are calculated. Based on the survey data, a user response willingness assessment model is established, outputting the response willingness of each user, and a willingness indicator variable is generated based on the response willingness to determine the users who can be controlled. Using the flexible domain parameters and the willingness indicator variables as constraint inputs, a joint bidding optimization model for the energy market and frequency regulation market of electric vehicle aggregators is established under a multi-market price scenario. Solve the joint bidding optimization model of the energy market and frequency regulation market for electric vehicle aggregators. With the goal of maximizing the total expected revenue of electric vehicle aggregators under the multi-scenario parameters, output the decision quantities related to charging bidding power, discharging bidding power, upward frequency regulation capacity, downward frequency regulation capacity and user subsidies in the energy market for each time period, and form the joint bidding application results of electric vehicle aggregators in the energy market and frequency regulation market. Among them, the total capacity of the electric vehicle cluster in time period t As shown in the following formula: ； ； in, The electric vehicle cluster satisfies the flexible domain constraints of the Chino polyhedron form. , and These represent the center point, generator, and maximum scaling factor of the flexible domain of the Cino polyhedral electric vehicle cluster within time period t. Let be the battery capacity of the i-th electric vehicle during time period t.

2. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 1, characterized in that, The historical charging and travel data, as well as the survey data, are preprocessed. The preprocessing includes at least data cleaning, missing data processing, and parameterization. Missing data processing includes mean imputation or data reconstruction.

3. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 1, characterized in that, In the electric vehicle flexibility evaluation index system, the adjustable capacity includes at least the average grid connection capacity obtained based on data cluster u, where data cluster u represents the set of charging records of the same user under similar behavior patterns, and the average grid connection capacity is used to characterize the statistical features of the adjustable capacity.

4. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 3, characterized in that, The average grid-connected capacity is obtained by averaging the product of the grid-connected SOC of each charge within the data cluster u and the electric vehicle battery capacity. The average inbound capacity of the u-th data cluster is calculated using the following formula: ； in, For a user's i-th charging SOC within the u-th data cluster, the network connection is defined. Indicates the battery capacity of an electric vehicle. This represents the number of charging records under the u-th data cluster.

5. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 1, characterized in that, The user response willingness evaluation model is a fuzzy neural network model, which includes a fuzzy rule layer and a defuzzification layer. The output of node i in the defuzzification layer satisfies the following formula: ； Where x, y, z, u are input variables. The set of adaptive parameters for the nodes; and the desired response of the output layer satisfies the following equation: ； in, Let be the initial trigger strength of the i-th fuzzy rule. Let be the contribution value of the i-th fuzzy rule to the output result. The output result is the response intention, where n is the number of rule nodes.

6. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 5, characterized in that, The response intention denoted as μ i Generate a willingness indicator variable based on the stated willingness to respond. The user's acceptance of the control is represented as a Boolean variable using a threshold, as shown in the following formula: ； in, This indicates that user i accepts regulation. This indicates that user i does not accept regulation, μ i The magnitude of user i's willingness to respond, as predicted by the fuzzy neural network; In the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market, only... The charging and discharging power and frequency regulation capacity of electric vehicles are used as schedulable decision variables or included in the effective callable capacity.

7. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 1, characterized in that, The objective function of the joint bidding optimization model for the energy market and frequency regulation market of electric vehicle aggregators is: ； Where s is the scene number, S is the total number of scenes, t is the time period number, T is the number of time periods, and a day is divided into 24 time periods, with each time period consisting of 1 hour. P s Let be the probability of the s-th scenario. , , These represent the revenue of electric vehicle aggregators participating in the energy market, the revenue of electric vehicle aggregators participating in the frequency modulation auxiliary service market, and the cost of electric vehicle aggregators participating in the energy-frequency modulation auxiliary service market, respectively, within the t-th time period under the s-th scenario.

8. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 7, characterized in that, Electric vehicle aggregators' energy market revenue in scenario s and time period t Calculated using the following formula: ； Where N is the number of electric vehicles in the electric vehicle cluster. , Let be the charging power and discharging power of the i-th electric vehicle in the t-th time period, respectively. The electricity price for electric vehicle aggregators under scenario s and time period t; Here, Δt represents the energy market price under scenario s and time period t; Δt is the time interval. This is an indicator variable for intention.

9. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 7, characterized in that, This includes the costs of electric vehicle aggregators in the energy market within scenario s and time period t. Calculated using the following formula: ； in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period. For energy market quotes under scenario s and time period t, , Let be the charging power and discharging power of the i-th electric vehicle in the t-th time period, respectively, and Δt be the time interval. This is an indicator variable for intention.

10. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 7, characterized in that, The revenue of an electric vehicle aggregator in the frequency modulation auxiliary market for scenario s and time period t satisfies the following formula: ； Frequency modulation market capacity revenue during the t-th time period in the s-th scenario Calculated using the following formula: ； in, For scenario s and time period t, the frequency regulation capacity price is... , Let be the upward and downward frequency regulation capacities provided by the electric vehicle aggregator to dispatch vehicle i during the t-th time period, respectively. This is an indicator variable for intention.

11. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 7, characterized in that, This includes the frequency modulation market costs for electric vehicle aggregators in the t-th time period of the s-th scenario. Calculated using the following formula: ； in, Let be the subsidy price offered by the electric vehicle aggregator to user i during the t-th time period. , These represent the upward and downward frequency modulation capacities provided by the electric vehicle aggregator for dispatching vehicle i during the t-th time period.

12. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 8, characterized in that, The following constraints are imposed on the charging and discharging power of each electric vehicle: ； ； ； The charging and discharging power submitted by the electric vehicle aggregator in time period t is obtained by aggregating the power of each electric vehicle: ； in, The upper limit of charging power for the i-th electric vehicle. Let be the upper limit of the discharge power of the i-th electric vehicle.

13. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 1, characterized in that, The joint bidding optimization model uses the charging power, discharging power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy decision amount of each electric vehicle in each time period as optimization variables, and satisfies the charging and discharging power constraints of each electric vehicle, the dynamic constraints of battery energy state, the frequency regulation power margin constraints, the frequency regulation capacity coupling constraints, and the flexible domain constraints of the electric vehicle cluster.

14. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 13, characterized in that, The adjustable capacity includes at least the average network access capacity obtained based on data cluster u, where data cluster u represents a set of charging records of the same user under similar behavior patterns. The average network access capacity is obtained by averaging the product of the network access SOC of each charging session within data cluster u and the battery capacity of the electric vehicle. The dynamic constraint on battery energy state and the coupling constraint on frequency modulation capacity satisfy the following equation: ； ； ； ； ； ； in, , These are the charging efficiency coefficient and the discharging efficiency coefficient, respectively. Let be the power of the i-th electric vehicle within time period t; , These are the upper and lower limits of capacity. , These represent the upward and downward frequency modulation capacities provided by the electric vehicle aggregator for dispatching vehicle i during the t-th time period.

15. The electric vehicle aggregator energy-frequency modulation bidding optimization method according to claim 14, characterized in that, Electric vehicle aggregators provide upward frequency modulation bidding capacity for vehicles dispatched during time period t. With downward frequency modulation bidding capacity Up-frequency modulation capacity provided by each electric vehicle in time period t With down-modulation capacity The result of aggregation is shown in the following formula: 。 16. An electric vehicle aggregator energy-frequency modulation bidding optimization device, characterized in that, include: The data acquisition module is used to acquire historical charging and travel data of electric vehicles, survey data to characterize user response intentions, and multi-scenario parameters of energy market prices and frequency regulation market capacity prices; The flexibility characterization module is used to construct an electric vehicle flexibility evaluation index system that includes adjustable capacity, on-grid time and charging power based on the historical charging data and travel data, establish flexible domain constraints for electric vehicle clusters, and calculate the flexible domain parameters used to limit the energy state boundary of electric vehicle clusters. The response willingness assessment module is used to establish a user response willingness assessment model based on the survey data, output the response willingness of each user, and generate a willingness indicator variable based on the response willingness to determine the set of users that can be controlled. The joint bidding optimization module is used to establish and solve the joint bidding optimization model of the electric vehicle aggregator energy market and frequency regulation market under the multi-market price scenario, using the flexible domain parameters and the willingness indicator variables as constraint inputs. The output module is used to output the energy market charging bid power, discharging bid power, upward frequency regulation capacity, downward frequency regulation capacity, and user subsidy decision amount for each time period, forming the joint bidding application results; Among them, the total capacity of the electric vehicle cluster in time period t As shown in the following formula: ； ； in, The electric vehicle cluster satisfies the flexible domain constraints of the Chino polyhedron form. , and These represent the center point, generator, and maximum scaling factor of the flexible domain of the electric vehicle cluster in the form of a Chino polyhedron within time period t.

17. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 16, characterized in that, It also includes a data preprocessing module, used to perform data cleaning, missing data processing, and parameterization on the historical charging data, travel data, and survey data; wherein, the missing data processing includes mean interpolation or data reconstruction.

18. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 16, characterized in that, In the flexibility assessment index system constructed by the flexibility characterization module, the adjustable capacity includes at least the average network access capacity obtained based on data cluster u. The data cluster u represents the charging record set of the same user under similar behavior patterns, and the average network access capacity is used to characterize the statistical features of the adjustable capacity.

19. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 18, characterized in that, The flexibility characterization module obtains the average grid-connected capacity by averaging the product of the grid-connected state of charge (SOC) and the electric vehicle battery capacity for each charge within the data cluster u.

20. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 16, characterized in that, The response willingness assessment module uses a fuzzy neural network to assess user response willingness. The fuzzy neural network includes a fuzzy rule layer and a defuzzification layer. The defuzzification layer outputs the contribution value of each rule node based on the input variables and node adaptive parameters. The output layer outputs the user response willingness based on the trigger strength of each fuzzy rule and the contribution value.

21. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 20, characterized in that, The response willingness assessment module generates a willingness indicator variable based on user response willingness and a preset threshold, and determines the set of users that can be controlled based on the willingness indicator variable. The joint bidding optimization module only allows electric vehicles corresponding to the user set to participate in the decision-making of charging and discharging power and frequency regulation capacity or to be included in the effective callable capacity.

22. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 16, characterized in that, The joint bidding optimization module establishes the joint bidding optimization model with the goal of maximizing the total expected revenue of electric vehicle aggregators under multiple scenario parameters. The total expected revenue is jointly determined by the energy market revenue, frequency regulation auxiliary market revenue, and the cost of participating in the energy-frequency regulation auxiliary service market at each time period under each price scenario.

23. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 22, characterized in that, The joint bidding optimization module includes an energy market revenue calculation submodule and an energy market cost calculation submodule. The energy market revenue calculation submodule is used to determine the energy market revenue based on the charging power, discharging power, electricity sales price, energy market quotation, and time interval of each electric vehicle in each time period. The energy market cost calculation submodule is used to determine the energy market cost based on the subsidy price paid to users subject to regulation in each time period.

24. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 22, characterized in that, The joint bidding optimization module includes a frequency modulation market revenue calculation submodule and a frequency modulation market cost calculation submodule. The frequency modulation market revenue calculation submodule is used to determine the frequency modulation auxiliary market revenue based on the frequency modulation capacity price and the upward and downward frequency modulation capacity provided by each electric vehicle. The frequency modulation market cost calculation submodule is used to determine the frequency modulation market cost based on the subsidy price paid to users accepting regulation in each time period.

25. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 22, characterized in that, The joint bidding optimization module uses the charging power, discharging power, upward frequency modulation capacity, downward frequency modulation capacity, and user subsidy decision amount of each electric vehicle in each time period as optimization variables, and satisfies the charging and discharging power constraints of each electric vehicle, the dynamic constraints of battery energy state, the frequency modulation power margin constraints, the frequency modulation capacity coupling constraints, and the flexible domain constraints of electric vehicle clusters.

26. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 25, characterized in that, The joint bidding optimization module is also used to apply the following constraints to the charging and discharging power of each electric vehicle: the charging power of each electric vehicle in each time period shall not exceed the corresponding charging power limit, the discharging power shall not exceed the corresponding discharging power limit, and the charging state and the discharging state shall be mutually exclusive. The charging power of each electric vehicle is aggregated to obtain the charging bid power for time period t, and the discharging power of each electric vehicle is aggregated to obtain the discharging bid power for time period t.

27. The electric vehicle aggregator energy-frequency modulation bidding optimization device according to claim 25, characterized in that, The joint bidding optimization module is also used for: The battery state of energy for time period t is determined based on the battery state of energy for time period t-1, the charging power, discharging power, charging efficiency, discharging efficiency for time period t, and the time interval. The upward frequency modulation capacity and downward frequency modulation capacity are limited according to the charging and discharging operating point and power limit of each electric vehicle, and the upward frequency modulation capacity and downward frequency modulation capacity are mutually exclusive. The upward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the upward frequency modulation bidding capacity for time period t, and the downward frequency modulation capacity provided by each electric vehicle is aggregated to obtain the downward frequency modulation bidding capacity for time period t.

28. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program that, when executed by the processor, implements the method according to any one of claims 1 to 15.

29. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 15.

Images