A Method and System for Multi-Time Regulation of EVA for Ancillary Services

Abstract

This invention discloses a method and system for bidirectional energy interaction and multi-period regulation of electric vehicle aggregators (EVs) for ancillary services, relating to the fields of smart grid and electric vehicle aggregator operation optimization technology. The method includes: calculating the energy and power feasible domain boundaries of the EV cluster within the scheduling cycle based on the framework of EV aggregator (EVA) participation in ancillary services; establishing a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation; combining charging demand response strategy (CBDR) and discharging demand response strategy (DBDR) to determine the user participation ratio; constructing an optimization model with maximizing the operating revenue of EV aggregators as the objective function; and obtaining the optimal charging and discharging scheduling scheme by solving this optimization model. This method employs a composite strategy when the EV access scale expands, helping to improve regulation capacity and user acceptance, and enhancing the stability of revenue.

Description

Technical Field

[0001] This invention relates to the field of smart grid and electric vehicle aggregator operation optimization technology, specifically to an EVA bidirectional energy interaction multi-time period control method and system for ancillary services. Background Technology

[0002] Against the backdrop of a global energy structure transition towards low-carbon development, the green development of the transportation and power sectors, as significant sources of carbon emissions, is crucial for addressing climate change. The large-scale adoption of electric vehicles (EVs) provides the power system with flexible distributed energy storage resources, which, theoretically, can participate in electricity market transactions and frequency regulation services through aggregators, thereby enhancing the grid's capacity to absorb renewable energy.

[0003] Existing research focuses on dispatch strategies for electric vehicle aggregators (EVAs) participating in the ancillary services market, primarily addressing demand response, incentive mechanisms, and optimized dispatch models. Methods such as dynamic pricing, deep learning prediction, and robust optimization are used to guide EV users to participate in grid peak shaving and valley filling, thereby improving photovoltaic absorption capacity and grid stability. Related work encompasses technologies such as price-based demand response, dispatch strategies based on EV adjustability, user behavior modeling, and uncertainty handling, achieving significant results in profit optimization, vehicle-to-grid (V2G) coordination, and V2G dispatch. However, when researching dispatch strategies for EVA participation in the ancillary services market, numerous uncertainties can significantly impact the final revenue.

[0004] While existing technologies have made significant progress in demand response, incentive mechanisms, and optimized scheduling, several shortcomings remain: First, most studies focus on single-market scenarios, lacking systematic research on EVA scheduling strategies in multi-market scenarios, making it difficult to adapt to complex market environments; second, insufficient consideration is given to the energy boundaries of electric vehicles during charging and discharging and their impact on peak-shaving capabilities, resulting in inaccurate characterization of controllable capacity and response boundaries; furthermore, the handling of user-involved uncertainties is still inadequate, easily leading to under-response or over-response, affecting scheduling effectiveness and stability; simultaneously, existing methods still need improvement in optimizing aggregation effects and enhancing scheduling robustness, limiting the flexibility and adaptability of strategies in large-scale practical applications. Summary of the Invention

[0005] To address the problem that existing technologies do not adequately consider the energy boundaries of electric vehicles during charging and discharging and their impact on peak-shaving capabilities, resulting in inaccurate characterization of controllable capacity and response boundaries, this invention proposes a bidirectional energy interaction multi-time-period regulation method and system for auxiliary services (EVs). By analyzing the controllable power regions and response boundaries of individual and cluster EVs, the invention clarifies the power adjustment capabilities and their evolutionary characteristics under different response strategies. An optimization model incorporating user price response intentions is constructed, thereby solving the problems existing in the prior art.

[0006] A method for bidirectional energy interaction and multi-time-period regulation of EVA for ancillary services includes the following steps:

[0007] Based on the framework of electric vehicle aggregator EVA participating in auxiliary services, the scheduling feasible domain of a single electric vehicle (EV) under different charging and discharging paths is constructed. The charging and discharging boundaries of the EV are determined based on the scheduling feasible domain of the EV. Based on the charging and discharging boundaries of the EV, a probability sampling method and EVA aggregation model are used to generate EV clusters of sufficient size, and the energy and power feasible domain boundaries of the EV clusters within the scheduling cycle are calculated.

[0008] A user willingness discrimination model is established to determine the probability of EV users' willingness to participate in regulation; and the user participation ratio is determined by combining the charging demand response strategy CBDR and the discharging demand response strategy DBDR. The CBDR is used as the charging incentive discount based on the difference between the real-time electricity price and the EVA price, and the EV cluster response ratio is calculated based on the user's sensitivity function to the charging incentive price; the DBDR is used to calculate the discharging subsidy price based on the battery's discharge depth and cycle count, and the EV cluster response ratio is calculated based on the user's response function to the discharging subsidy price.

[0009] By combining the energy and power feasible domain boundary of the EV cluster within the scheduling cycle, the probability of EV users' willingness to participate in regulation, and the user participation ratio, the effective power capacity of EVA that is actually adjusted and reduced in each time period is determined.

[0010] Using effective power capacity as a constraint on regulation capability, an optimization model is constructed with the objective function of maximizing the operating revenue of electric vehicle aggregators; by solving this optimization model, the optimal charging and discharging scheduling scheme is obtained.

[0011] Furthermore, the step of using the difference between the real-time electricity price and the EVA price as the charging incentive discount, and calculating the proportion of EV cluster response CBDR based on the user's sensitivity function to the charging incentive price, specifically includes the following steps:

[0012] By using the difference between the real-time electricity price and the EVA price as a charging incentive discount, when an EV user has a charging need, the charging incentive price Δ for the current period is applied. D E ( t Choose whether to participate in CBDR;

[0013] If the user chooses to respond to CBDR, they will enjoy the electricity price discount provided by EVA, which will regulate the charging process; EVA in the first... i The control period during EV charging is as follows:

[0014] ;

[0015] In the formula, t i,min For the first i The fastest charging time for an EV to reach the desired SOC value; t i,CBDR For the first i The moment when the EV responds to CBDR regulation ends; SOC i,s For the first i SOC of an EV upon network access; m i,CBDR A constant variable of 0-1 is used to represent the response state, and its value of 1 indicates the first... i The EV responds to CBDR; a value of 0 indicates that it does not respond to CBDR. B e This represents the battery's full capacity. t i,s For the first i The time when an EV is registered with the network; t i,d For the first i The time when an EV leaves the network; SOC i,des For the first i Target SOC value of an EV when charging ends S e ; For EV in time t Efficiency during charging; P+ i ( t ) is the first i Maximum charging power limit for EVs; Δ t For time intervals;

[0016] If the user does not respond to CBDR, it indicates that they are not sensitive to the incentive discount price, choose not to accept EVA regulation, and directly charge autonomously until the SOC reaches the expected target value.

[0017] Based on the user's sensitivity function to charging incentive prices, the proportion of EV clusters responding to CBDR is calculated; the calculation process is expressed as follows:

[0018] ;

[0019] ;

[0020] In the formula, fCBDR i The function is used to represent the response function of CBDR; a This indicates the user's sensitivity to the incentive price; b This indicates the baseline response ratio.

[0021] Furthermore, the calculation of the discharge subsidy price based on battery wear caused by the depth of discharge and the number of cycles specifically includes the following steps:

[0022] Depth of discharge of the battery D The relationship between the number of cycles and battery wear is expressed as follows:

[0023] ;

[0024] ;

[0025] ;

[0026] In the formula: D The depth of discharge for a single EV participating in DBDR; f N It is the number of discharge cycles participating in the DBDR strategy; G ( D The total charge / discharge energy is 100%. The average depth of discharge; D j The specific depth of discharge for EVA within any given time period; N D Maximum permissible depth of discharge; R loss For battery wear and tear costs; R c Battery purchase cost;

[0027] The formula for calculating the discharge subsidy price is as follows:

[0028] ;

[0029] In the formula, R 1 represents the current day discharge subsidy price, which serves as the base subsidy price for EV users participating in DBDR.

[0030] Furthermore, the framework based on the participation of electric vehicle aggregator EVA in auxiliary services constructs the scheduling feasible region of a single electric vehicle (EV) under different charging and discharging paths. The charging and discharging boundaries of the EV are determined based on the scheduling feasible region of the EV. The determination process specifically includes the following steps:

[0031] The feasible region for scheduling a single electric vehicle (EV) is represented as follows:

[0032] ;

[0033] ;

[0034] In the formula, and These represent increases and decreases in power for EVs, respectively.P 1( t )~ P 4( t These represent the response power of EV in response states I to IV, respectively.

[0035] The boundary between EV charging and discharging is then represented as:

[0036] ;

[0037] ;

[0038] In the formula, SOC i ( t )for t Time of the first i SOC of an EV; For the first i EVs in Charging time within a given time period; R ev The values ​​1, 0, and -1 represent the EV charging, idle, and discharging states, respectively. , The first i EVs in t The upper and lower limits of SOC at any given time; SOC i,des For the first i Target SOC value of an EV when charging ends S e ; B e This represents the battery's full capacity. t i,s For the first i The time when an EV is registered with the network; t i,d For the first i The time when an EV leaves the network; P c,max and P d,max This represents the maximum charge / discharge power boundary. For the first i EVs in time t The power during charging; , EV at time t Efficiency during charging and discharging; For the first i EVs in time t Power during discharge; This represents the minimum battery capacity for an EV.

[0039] Furthermore, the calculation process for the energy and power feasible domain boundaries of the EV cluster within the scheduling cycle is expressed as follows:

[0040] ;

[0041] In the formula, , They are respectively t The upper and lower limits of energy for the EV cluster at any given time; N The number of EVs that EVA can control; , They are respectively t The upper and lower limits of the power of the EV cluster at any given time; m i A constant variable between 0 and 1, its value of 1 indicates t Time of the first i The EV is in a charging state, and its value is 0. t Time of the first i The EV is in a discharge state.

[0042] Furthermore, the establishment of a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation specifically includes the following steps:

[0043] Definition of the first i The psychological threshold for electric vehicles in Taiwan is i i The current price incentive level is Δ p Then the probability of the vehicle participating in regulation. p i Represented as:

[0044] ;

[0045] in, α This represents the kurtosis coefficient of the intention function;

[0046] Based on the probability of vehicle control participation p i By performing Monte Carlo sampling on all agent individuals to determine the actual set of responding vehicles, the overall effective participation rate of the cluster can be calculated. r Represented as:

[0047] ;

[0048] In the formula, s i Indicates the first i The participation status of the vehicles, if the sampling result is less than p i ,but s i =1, otherwise si =0;

[0049] Incorporating user preferences to adjust capacity for auxiliary services P up,eff ( t ) and reduced capacity P down,eff ( t ) is represented as:

[0050] ;

[0051] ;

[0052] In the formula, r up (t), r down (t) represents the effective user engagement rate in the scenarios of upward and downward adjustments, respectively. P up,max ( t ), P down,max ( t ) represents the maximum adjustable power under the boundary conditions.

[0053] Furthermore, the optimization model is constructed based on energy balance constraints, power constraints, energy boundary constraints, electric vehicle quantity constraints, battery state constraints, and real-time power constraints, with expenditures including day-ahead electricity purchase cost, real-time peak shaving cost, discharge subsidies, and battery loss compensation, and revenues including capacity revenue and energy revenue.

[0054] Furthermore, the objective function is expressed as:

[0055] ;

[0056] Among them, EVA's electricity purchase cost in the day-ahead market F 1 is represented as: , R DA ( t (Time period) t The day-ahead market transaction price of electricity; P DA ( t (Time period) t The planned power output set by EVA in the daytime; T For the entire scheduling period of EVA; Δ t For time intervals;

[0057] The real-time peak shaving cost F2 is expressed as: , and For time period tReal-time market transaction electricity prices within the country; P up ( t )and P down ( t This refers to the real-time upward and downward peaking power of the EVA.

[0058] Discharge subsidies and battery loss compensation are expressed as follows: , For the first i EVs during the time period t Battery energy loss due to internal discharge; R loss For battery wear and tear costs; R 1 represents the day-ahead discharge subsidy price; For the first i EVs in time t The power during charging;

[0059] EVA's capacity revenue from participating in the peak-shaving market is expressed as follows: , R P The electricity price per unit of peak-shaving capacity; C a The peak-shaving capacity that EVA is committed to providing during peak hours; oh Peak shaving service coefficient;

[0060] The energy revenue that EVA receives from participating in peak shaving in the ancillary services market is expressed as follows: , and These refer to the real-time upward and downward adjustments of peak-shaving energy electricity prices in the market during the day. E UP ( t )and E Down ( t ) are respectively t The upward and downward peak-shaving energy of the EV cluster during the time period.

[0061] Furthermore, the energy balance constraint, power constraint, energy boundary constraint, electric vehicle quantity constraint, battery state constraint, and real-time power constraint are specifically expressed as follows:

[0062] EVA constraints include: , , , and ;in, For time intervals; and The first i EVs in tThe charging and discharging power at any given moment; or c and or d The first i The charging and discharging efficiency of EVs For EVA in t Energy at time -1; and for t The charging and discharging power of EVA at time -1; for t The EV is constantly connected to the energy of the EVA; for t The energy that EV leaves EVA at any moment; N This represents the set of EV quantities in EVA. For EVA in t Maximum charging power at any given time; For the first i Maximum charging power limit for EVs For EVA in t The maximum discharge power at any given moment; For the first i Maximum discharge power limit for EVs; and They represent t The maximum and minimum rechargeable capacity of EVA at any given time;

[0063] The energy boundary constraint is expressed as: , E - ( t )for t The lower limit of EV cluster energy at any given time. E + ( t )for t The maximum energy capacity of the EV cluster at any given time; P + ( t ), P - ( t ) are respectively t The upper and lower limits of the power of the EV cluster at any given time;

[0064] The EV quantity constraint is expressed as: , N act The number of EVs that are actually charged and participate in regulation; N max The maximum number of EVs that EVA can control;

[0065] The battery state constraints are expressed as follows: , The minimum SOC value required to participate in DBDR; This represents the ideal state value after intervention and regulation.

[0066] The power constraint is expressed as: , P ( t This refers to the net power of EVA in real-time regulation; and

[0067] The real-time upward and downward peak-shaving power of EVA is greater than 0 during real-time control.

[0068] The present invention also includes an EVA bidirectional energy interaction multi-time period control system for ancillary services, comprising:

[0069] The acquisition module is used to construct the scheduling feasible domain of a single electric vehicle (EV) under different charging and discharging paths based on the framework of EV aggregator EVA participating in auxiliary services. Based on the scheduling feasible domain of the EV, the charging and discharging boundaries of the EV are determined. Based on the charging and discharging boundaries of the EV, a probability sampling method and EVA aggregation model are used to generate EV clusters of sufficient size, and the energy and power feasible domain boundaries of the EV clusters within the scheduling cycle are calculated.

[0070] The calculation module is used to establish a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation; and to determine the user participation ratio by combining the charging demand response strategy CBDR and the discharging demand response strategy DBDR. The CBDR is used to calculate the proportion of EV cluster response to CBDR based on the difference between the real-time electricity price and the EVA price, and the user's sensitivity function to the charging incentive price. The DBDR is used to calculate the discharging subsidy price based on the battery's discharge depth and cycle count, and the user's response function to the discharging subsidy price is used to calculate the proportion of EV cluster response to DBDR.

[0071] The determination module is used to combine the energy and power feasible domain boundary of the EV cluster within the scheduling cycle, the probability of EV users' willingness to participate in regulation, and the user participation ratio to determine the effective power capacity of EVA that is actually adjusted and reduced in each time period.

[0072] The optimization module is used to construct an optimization model with the effective power capacity as the control capability constraint and the objective function of maximizing the operating revenue of electric vehicle aggregators. By solving the optimization model, the optimal charging and discharging scheduling scheme is obtained.

[0073] This invention provides a method for bidirectional energy interaction and multi-time period control of EVA for ancillary services, which has the following beneficial effects:

[0074] This invention constructs the scheduling feasible domain of a single EV and aggregates the energy and power feasible domain boundaries of the EV cluster. This allows for precise quantification of the upper and lower limits of the physical regulation capacity available to the electric vehicle aggregator (EVA) within the scheduling cycle. It overcomes the problem of insufficient consideration of the energy boundaries of EVs during charging and discharging processes and their impact on peak-shaving capacity, which leads to inaccurate characterization of controllable capacity and response boundaries. By introducing a joint response mechanism of CBDR and DBDR and integrating user price incentive response intention modeling, the uncertainty of user behavior is incorporated into the quantitative analysis framework. This enables the EVA to more accurately predict the actual available response resources, effectively mitigating revenue fluctuations caused by user intention uncertainty. This method not only flexibly releases bidirectional regulation capacity and smooths energy curve fluctuations during multi-period peak shaving but also significantly improves load tracking capability and scheduling flexibility. Comparison of revenue range fluctuations under different strategies reveals that the composite response mechanism, while releasing greater regulation capacity, can improve user acceptance and enhance revenue stability, providing technical support for the EVA to maximize revenue and ensure stable system operation under large-scale EV access conditions. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the four response states of EV in an embodiment of the present invention;

[0076] Figure 2 This is a schematic diagram of the EVA participation in the ancillary services market transaction framework in an embodiment of the present invention;

[0077] Figure 3 This is a schematic diagram of the EVA bidirectional energy interaction multi-time period control method for auxiliary services in an embodiment of the present invention. Detailed Implementation

[0078] 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.

[0079] This invention addresses the scheduling optimization problem of electric vehicles (EVs) in the multi-time-period ancillary service market involving aggregators (EVAs). It proposes a bidirectional energy interaction multi-time-period control strategy for EVAs oriented towards ancillary services. By analyzing the controllable power range and response boundary of individual and clustered EVs, the invention clarifies the power adjustment capabilities and their evolutionary characteristics under different response strategies. An optimization model incorporating user price response intentions is constructed, and a mapping relationship between discount coefficients and satisfaction is introduced to achieve dynamic load boundary adjustment, reflecting the nonlinear characteristics of user behavior. A composite strategy is employed as the scale of EV access expands to help improve regulation capacity and user acceptance, thereby enhancing revenue stability.

[0080] like Figure 3As shown, the method specifically includes the following steps:

[0081] S1. Based on the EVA participation ancillary service framework, the scheduling feasible domain of a single EV under different charging and discharging paths is characterized by combining the initial SOC state of the EV, and the charging and discharging boundaries are clarified; based on the probability sampling method and the EVA aggregation model, an EV cluster of sufficient size is generated and its energy and power feasible domain boundaries are superimposed and calculated; the scheduling feasible domain of electric vehicles (EVs) refers to the range of adjustable charging and discharging power of the electric vehicle cluster under the premise of meeting user needs, grid safety and market rules.

[0082] like Figure 1 As shown, for the EV cluster response boundary, random sampling analysis is performed on the EV cluster response capability boundary based on probability sampling and EVA aggregation model. The specific steps are as follows:

[0083] a) Based on the probability model of relevant parameters in EV uncertainty analysis and modeling, determine the charging power, SOC capacity and charging capacity of a single EV, and determine its charging and discharging state.

[0084] b) Based on the response boundary of a single EV, the energy boundary of a single EV is determined by equations (1) and (2).

[0085] (1)

[0086] (2)

[0087] in, and This represents the maximum charge / discharge power boundary.

[0088] c) By iteratively executing the sampling process a)-b) until the preset EV sample size is reached, the energy and power feasible domain boundaries of the EV cluster are determined based on equation (3).

[0089] (3)

[0090] (1) EV response boundary analysis:

[0091] (4)

[0092] In the formula, and These represent increases and decreases in power for EVs, respectively. P 1( t )~ P 4( t ) represent the response power of EV in response states I to IV, respectively.

[0093] (2) Boundary analysis of single-unit EV response:

[0094] (5)

[0095] (6)

[0096] In the formula, SOC i ( t )for t Time of the first i SOC of an EV; For the first i EVs in Charging time within a given time period; R ev The values ​​1, 0, and -1 represent the EV charging, idle, and discharging states, respectively. , The first i EVs in t The upper and lower limits of SOC at any given time; SOC i,des For the first i The target SOC value of an EV when charging ends (i.e. S e ); B e This represents the battery's full capacity. t i,s For the first i The time when an EV is registered with the network; t i,d For the first i The time when an EV leaves the network; P c,max and P d,max This represents the maximum charge / discharge power boundary. For the first i EVs in time t The power during charging; , EV at time t Efficiency during charging and discharging; For the first i EVs in time t Power during discharge; This represents the minimum battery capacity for an EV.

[0097] (3) EV cluster response boundary analysis:

[0098] (7)

[0099] In the formula, , They are respectively t The upper and lower limits of energy for the EV cluster at any given time;N The number of EVs can be adjusted using EVA; , They are respectively t The upper and lower limits of the power of the EV cluster at any given time; m i A constant variable between 0 and 1, its value of 1 indicates t Time of the first i The EV is in a charging state, and its value is 0. t Time of the first i The EV is in a discharge state.

[0100] S2. In the study of EV participation in auxiliary service regulation, a user willingness discrimination model based on Monte Carlo sampling is introduced. Through multi-agent modeling, a psychological threshold is set for virtual EV individuals to determine their willingness to participate in regulation.

[0101] EV user willingness modeling:

[0102] In terms of probabilistic modeling of intentions, let the first... i The psychological threshold for electric vehicles in Taiwan is i i The current price incentive level is Δ p Then the probability of the vehicle participating in regulation. p i It can be represented as:

[0103] (8)

[0104] in, α The kurtosis coefficient of the willingness function reflects the user's sensitivity to price changes. Based on the above probability model, Monte Carlo sampling is performed on all agent individuals to determine the actual set of responding vehicles. If the total number of vehicles in the cluster is N, then the overall effective participation rate of the cluster is... r It can be represented as:

[0105] (9)

[0106] In the formula, s i Indicates the first i The participation status of the vehicles, if the sampling result is less than p i but s i =1, otherwise s i =0.

[0107] After incorporating user preferences, the actual capacity that can be increased for auxiliary service regulation can be adjusted. P up,eff ( t ) and reduced capacity Pdown,eff ( t It can be determined by both maximum physical capacity and willingness to respond:

[0108] (10)

[0109] (11)

[0110] In the formula, r up (t), r down (t) represents the effective user engagement rate in the scenarios of upward and downward adjustments, respectively. P up,max ( t ), P down,max ( t ) represents the maximum adjustable power under the boundary conditions.

[0111] S3. Design a charging demand response strategy (CBDR) based on price incentives and a discharging demand response strategy (DBDR) based on discharging subsidies. Calculate the participation ratio under the two strategies using a user price sensitivity function. Combine user willingness with the physical adjustable boundary to provide a basis for EVA to schedule EV charging and discharging behavior in real time.

[0112] (1) Charge-type demand response strategy (CBDR): By using the difference between the real-time electricity price and the EVA price as a charging incentive discount, when EV users have charging needs, the charging incentive price Δ can be applied to the EV user. D E ( t Users can choose whether to participate in CBDR. If a user chooses to participate in CBDR, they can enjoy the electricity price discount provided by EVA, which will then make reasonable adjustments during the charging process. If a user does not participate in CBDR, it indicates that they are not sensitive to the incentive discount price, choose not to accept EVA's adjustments, and directly charge autonomously until the SOC reaches the desired target value.

[0113] For EV clusters, the proportion of CBDR responses N CBDR Represented as:

[0114] (12)

[0115] (13)

[0116] In the formula, The function is used to represent the response function of CBDR; a This represents the user's sensitivity to the incentive price, and is set to 0.2. bThis represents the baseline response ratio, set to 0.05.

[0117] If the EV user chooses to respond to CBDR, EVA can [do something] in the [time period]. i The control period during EV charging is as follows:

[0118] (14)

[0119] In the formula, t i,min For the first i The fastest charging time for an EV to reach the desired SOC value; t i,CBDR For the first i The moment when the EV responds to CBDR regulation ends; SOC i,s For the first i SOC of an EV upon network access; m i,CBDR A constant variable of 0-1 is used to represent the response state, and its value of 1 indicates the first... i The EV responds to CBDR, and a value of 0 indicates that it does not respond to CBDR.

[0120] (2) Discharge-type demand response strategy (DBDR): During DBDR, the EV battery is affected by the charge / discharge rate, depth of discharge, and number of charge / discharge cycles. This invention mainly considers the impact of depth of discharge and number of cycles on the performance of the EV battery. The relationship between the depth of discharge (D) and the number of cycles can be expressed by formulas (15)-(17).

[0121] (15)

[0122] (16)

[0123] (17)

[0124] In the formula: D The depth of discharge for a single EV participating in DBDR; f N It is the number of discharge cycles participating in the DBDR strategy; G ( D The total charge / discharge energy is 100%. The average depth of discharge; D j The specific depth of discharge of EVA within a certain time period; N D Maximum permissible depth of discharge; R loss For battery wear and tear costs; R cFor the cost of purchasing batteries.

[0125] The specific formula for calculating the discharge subsidy price is as follows:

[0126] (18)

[0127] In the formula, R 1 represents the day-ahead discharge subsidy price, which can serve as the base subsidy price for EV users to participate in DBDR.

[0128] Considering the differences in the willingness of EV users to respond, the proportion of EVs responding to the DBDR strategy can be derived from formulas (19) and (20).

[0129] (19)

[0130] (20)

[0131] In the formula, The function is used to represent the response function to DBDR; the user response constant is calculated based on electricity price fluctuations. K It is 2.5; ϴ >1 means the exponent is 1.5; R e This refers to the electricity price after real-time subsidies.

[0132] (3) EVA performs real-time scheduling during the EV charging control period:

[0133] (twenty one)

[0134] In the formula, t i,DBDR For the first i The end time of DBDR regulation for each EV; n i This is a binary variable; if its value is 1, it represents the [number]th [element]. i The EV responded to either CBDR or DBDR, with a value of 0 indicating that it did not participate in either regulation; similarly, d i Used to identify the response status, its value of 1 indicates the first response. i The EV responds to CBDR or DBDR, with a value of 0 indicating no response.

[0135] (4) EVA Costs: The day-ahead electricity purchase cost F1 of EVA is shown below:

[0136] (twenty two)

[0137] In the formula, R DA (t (Time period) t The day-ahead market transaction price of electricity; P DA ( t (Time period) t The planned power output set by EVA in the daytime; T For the entire scheduling period of EVA; Δ t This refers to the time interval.

[0138] In the process of EVA participating in peak shaving, its costs include upward peak shaving costs and downward peak shaving costs, which can be specifically expressed as:

[0139] (twenty three)

[0140] In the formula, and For time period t Real-time market transaction electricity prices within the country; P up ( t )and P down ( t This refers to the real-time upward and downward peaking power of the EVA.

[0141] During the regulation process, considering that some EV users respond to DBDR, EVA will spend a portion of the cost as a discharge subsidy and battery discharge loss compensation, as shown in equation (24):

[0142] (twenty four)

[0143] In the formula, For the first i EVs during the time period t Battery energy loss due to internal discharge.

[0144] S4. Taking the maximization of EVA's operating revenue during peak shaving as the objective function, and comprehensively considering expenditures such as day-ahead electricity purchase cost, real-time peak shaving cost, discharge subsidies and battery loss compensation, as well as revenues such as capacity revenue and energy revenue, an optimization model is established based on EVA energy balance constraints, power constraints, energy boundary constraints, EV quantity constraints, battery state constraints and real-time power constraints to solve the optimal charging and discharging scheduling scheme that meets the travel needs of EV users and the system peak shaving requirements.

[0145] (1) EVA compensation income:

[0146] The capacity revenue of EVA participating in the peak shaving market can be expressed as:

[0147] (25)

[0148] In the formula,R P The electricity price per unit of peak-shaving capacity; C a The peak-shaving capacity that EVA is committed to providing during peak hours; oh The peak-shaving service factor.

[0149] EVA responds to the needs of the power system by adjusting the charging and discharging behavior of EVs, and its peak-shaving energy is shown in equation (26):

[0150] (26)

[0151] In the formula, E UP ( t )and E Down ( t ) are respectively t The upward and downward peak-shaving energy of the EV cluster during the time period; A ( i , t For the ancillary services market t The first time period released i One peak-shaving indicator signal; T m for t The number of time intervals for peak-shaving signals within a given period. Based on this formula, the energy revenue of EVA participating in peak-shaving in the ancillary services market can be expressed as:

[0152] (27)

[0153] In the formula, and These refer to the real-time upward and downward adjustments to peak-shaving energy electricity prices in the market during the day.

[0154] (2) Taking the maximum operating profit F from EVA as the optimization objective, the objective function can be expressed as:

[0155] (28)

[0156] In the formula, the objective function consists of EVA expenditure costs and revenue sources in formula (22) - formula (27).

[0157] (3) Constraints:

[0158] ①EVA constraint:

[0159] (29)

[0160] (30)

[0161] (31)

[0162] (32)

[0163] (33)

[0164] In the formula, For the time interval; in equation (26) and The first i EVs in t The charging and discharging power at any given moment; or c and or d The first i The charging and discharging efficiency of an EV. Equation (27) is the energy balance constraint. For EVA in t Energy at time -1; and for t The charging and discharging power of EVA at time -1; for t The EV is constantly connected to the energy of the EVA; for t The energy that EV leaves EVA at any moment; N Let represent the set of EV quantities in EVA. In equation (27) t At any given moment, the energy of EVA is equal to t EVA energy, EV charging energy, and EV connection energy at time -1 t Discharge energy at time -1 t The energy difference between EV and EVA at time t. In equation (28) For EVA in t Maximum charging power at any given time; For the first i Maximum charging power limit for EVs. In equation (29) For EVA in t The maximum discharge power at any given moment; For the first i The maximum discharge power limit of an EV. In equation (30) and They represent t The model defines the maximum and minimum rechargeable capacity of the EVA at any given time. It quantifies the energy and aggregate response boundaries of EV cluster charging, ensuring energy and power balance of the EVA when EV users connect and disconnect.

[0165] ② Energy boundary constraints:

[0166] (34)

[0167] In the formula, E - ( t )for t The lower limit of EV cluster energy at any given time. E + ( t )for t The maximum energy capacity of the EV cluster at any given time; , They are respectively t The upper and lower limits of the power of the EV cluster at any given time.

[0168] ③EV quantity constraints

[0169] Considering the number of charging stations and the capacity of EVA, the number of EVs that can be aggregated by EVA at one time must not exceed the maximum vehicle capacity of EVA:

[0170] (35)

[0171] In the formula, N act The number of EVs that are actually charged and participate in regulation; N max The maximum number of EVs that can be regulated for EVA.

[0172] ④ Battery state constraints:

[0173] Since some users participate in DBDR, and considering the travel needs of EV users, it is necessary to ensure that SOC participates in regulation without affecting user travel:

[0174] (36)

[0175] In the formula, The minimum SOC value required to participate in DBDR; This represents the ideal state value after intervention and regulation.

[0176] ⑤ Power constraint:

[0177] EVA needs to adjust its power output in real time according to changes in the peak-shaving signal. Its net power output must be adjusted within any given time period. P ( t The discharge power must not be lower than the maximum discharge power for the current period, nor exceed the maximum charging power. Specific constraints are as follows:

[0178] (37)

[0179] In the formula, P ( tThis refers to the net power of EVA in real-time regulation; and The real-time upward and downward peak-shaving power of EVA is greater than 0 during real-time control.

[0180] like Figure 2 As shown, in EVA optimized scheduling, this invention introduces a joint response mechanism of CBDR and DBDR, and integrates user price incentive response willingness modeling. This not only enables flexible release of bidirectional adjustment capabilities and smoothing of energy curve fluctuations during multi-period peak shaving, but also significantly improves load tracking capabilities and scheduling flexibility. A comparison of revenue range fluctuations under different strategies reveals that the composite response mechanism, while releasing greater adjustment capacity, can improve user acceptance and enhance revenue stability. Combined with a dynamic matching compensation strategy based on user profiles (determining EVA compensation plans for different time periods based on the characteristics of EV user willingness modeling), it can effectively alleviate revenue fluctuations caused by willingness uncertainty, providing technical support for maximizing EVA revenue and ensuring system stability under large-scale EV access conditions.

[0181] Based on the same inventive concept, this invention also proposes a bidirectional energy interaction multi-time period control system for EVA (Electronic Valve Auxiliary Services), comprising:

[0182] The acquisition module is used to construct the scheduling feasible domain of a single electric vehicle (EV) under different charging and discharging paths based on the framework of EV aggregator EVA participating in auxiliary services. Based on the scheduling feasible domain of the EV, the charging and discharging boundaries of the EV are determined. Based on the charging and discharging boundaries of the EV, a probability sampling method and EVA aggregation model are used to generate EV clusters of sufficient size, and the energy and power feasible domain boundaries of the EV clusters within the scheduling cycle are calculated.

[0183] The calculation module is used to establish a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation; and to determine the user participation ratio by combining the charging demand response strategy CBDR and the discharging demand response strategy DBDR. CBDR is used to calculate the proportion of EV cluster response to CBDR based on the difference between the real-time electricity price and the EVA price as the charging incentive discount, and the proportion of EV cluster response to CBDR is calculated based on the user's sensitivity function to the charging incentive price. DBDR is used to calculate the discharging subsidy price based on the battery's discharge depth and cycle count, and the proportion of EV cluster response to DBDR is calculated based on the user's response function to the discharging subsidy price.

[0184] The determination module is used to combine the energy and power feasible domain boundary of the EV cluster within the scheduling cycle, the probability of EV users' willingness to participate in regulation, and the user participation ratio to determine the effective power capacity of EVA that is actually adjusted and reduced in each time period.

[0185] The optimization module is used to construct an optimization model with the effective power capacity as the control capability constraint and the objective function of maximizing the operating revenue of electric vehicle aggregators. By solving the optimization model, the optimal charging and discharging scheduling scheme is obtained.

[0186] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for bidirectional energy interaction and multi-time-period regulation of EVA for ancillary services, characterized in that, Includes the following steps: Based on the framework of electric vehicle aggregator EVA participating in auxiliary services, the scheduling feasible domain of a single electric vehicle (EV) under different charging and discharging paths is constructed. The charging and discharging boundaries of the EV are determined based on the scheduling feasible domain of the EV. Based on the charging and discharging boundaries of the EV, a probability sampling method and EVA aggregation model are used to generate EV clusters of sufficient size, and the energy and power feasible domain boundaries of the EV clusters within the scheduling cycle are calculated. A user willingness discrimination model is established to determine the probability of EV users' willingness to participate in regulation; and the user participation ratio is determined by combining the charging demand response strategy CBDR and the discharging demand response strategy DBDR. The CBDR is used as the charging incentive discount based on the difference between the real-time electricity price and the EVA price, and the EV cluster response ratio is calculated based on the user's sensitivity function to the charging incentive price; the DBDR is used to calculate the discharging subsidy price based on the battery's discharge depth and cycle count, and the EV cluster response ratio is calculated based on the user's response function to the discharging subsidy price. By combining the energy and power feasible domain boundary of the EV cluster within the scheduling cycle, the probability of EV users' willingness to participate in regulation, and the user participation ratio, the effective power capacity of EVA that is actually adjusted and reduced in each time period is determined. Using effective power capacity as a constraint on regulation capability, an optimization model is constructed with the objective function of maximizing the operating revenue of electric vehicle aggregators; by solving this optimization model, the optimal charging and discharging scheduling scheme is obtained.

2. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 1, characterized in that, The method of using the difference between the real-time electricity price and the EVA price as the charging incentive discount, and calculating the proportion of EV cluster response CBDR based on the user's sensitivity function to the charging incentive price, specifically includes the following steps: By using the difference between the real-time electricity price and the EVA price as a charging incentive discount, when an EV user has a charging need, the charging incentive price Δ for the current period is applied. D E ( t Choose whether to participate in CBDR; If the user chooses to respond to CBDR, they will enjoy the electricity price discount provided by EVA, which will regulate the charging process; EVA in the first... i The control period during EV charging is as follows: ； In the formula, t i,min For the first i The fastest charging time for an EV to reach the desired SOC value; t i,CBDR For the first i The moment when the EV response to CBDR regulation ends; SOC i,s For the first i SOC of an EV upon network access; m i,CBDR A constant variable of 0-1 is used to represent the response state, and its value of 1 indicates the first... i The EV responds to CBDR; a value of 0 indicates that it does not respond to CBDR. B e This represents the battery's full capacity. t i,s For the first i The time when an EV is registered with the network; t i,d For the first i The time when an EV leaves the network; SOC i,des For the first i Target SOC value of an EV when charging ends S e ; For EV in time t Efficiency during charging; For the first i Maximum charging power limit for EVs; Δ t For time intervals; If the user does not respond to CBDR, it indicates that they are not sensitive to the incentive discount price, choose not to accept EVA regulation, and directly charge autonomously until the SOC reaches the expected target value. Based on the user's sensitivity function to charging incentive prices, the proportion of EV clusters responding to CBDR is calculated; the calculation process is expressed as follows: ； ； In the formula, The function is used to represent the response function of CBDR; This indicates the user's sensitivity to the incentive price; b This indicates the baseline response ratio.

3. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 1, characterized in that, The calculation of the discharge subsidy price based on battery wear caused by the depth of discharge and the number of cycles includes the following specific steps: Depth of discharge of the battery D The relationship between the number of cycles and battery wear is expressed as follows: ； ； ； In the formula: D The depth of discharge for a single EV participating in DBDR; f N It is the number of discharge cycles participating in the DBDR strategy; G ( D The total charge / discharge energy is 100%. The average depth of discharge; D j The specific depth of discharge for EVA within any given time period; N D Maximum permissible depth of discharge; R loss For battery wear and tear costs; R c Battery purchase cost; The formula for calculating the discharge subsidy price is as follows: ； In the formula, R 1 represents the current day discharge subsidy price, which serves as the base subsidy price for EV users participating in DBDR.

4. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 1, characterized in that, The framework based on the participation of electric vehicle aggregator (EVA) in auxiliary services constructs the scheduling feasible region of a single electric vehicle (EV) under different charging and discharging paths. The charging and discharging boundaries of the EV are determined based on its scheduling feasible region. The determination process specifically includes the following steps: The feasible region for scheduling a single electric vehicle (EV) is represented as follows: ； ； In the formula, and These represent increases and decreases in power for EVs, respectively. P 1( t )~ P 4( t These represent the response power of EV in response states I to IV, respectively. The boundary between EV charging and discharging is then represented as: ； ； In the formula, SOC i ( t )for t Time of the first i SOC of an EV; For the first i EVs in Charging time within a given time period; R ev The values ​​1, 0, and -1 represent the EV charging, idle, and discharging states, respectively. , The first i EVs in t The upper and lower limits of SOC at any given time; SOC i,des For the first i Target SOC value of an EV when charging ends S e ; B e This represents the battery's full capacity. t i,d For the first i The time when an EV leaves the network; For the first i EVs in time t The power during charging; , EV at time t Efficiency during charging and discharging; For the first i EVs in time t The power during discharge; This represents the minimum battery capacity for an EV.

5. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 4, characterized in that, The calculation process for the energy and power feasible region boundaries of the EV cluster within the scheduling period is expressed as follows: ； In the formula, E ( t ), E ( t ) are respectively t The upper and lower limits of energy for the EV cluster at any given time; N The number of EVs that EVA can control; P + ( t ), P - ( t ) are respectively t The upper and lower limits of the power of the EV cluster at any given time; m i A constant variable between 0 and 1, its value of 1 indicates t Time of the first i The EV is in a charging state, and its value is 0. t Time of the first i The EV is in a discharging state; The first i EVs in t The upper limit of charging power and the lower limit of discharging power at any given time.

6. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 1, characterized in that, The establishment of a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation specifically includes the following steps: Definition of the first i The psychological threshold for electric vehicles in Taiwan is θ i The current price incentive level is Δ p Then the probability of the vehicle participating in regulation. p i Represented as: ； in, α This represents the kurtosis coefficient of the intention function; Based on the probability of vehicle control participation p i By performing Monte Carlo sampling on all agent individuals to determine the actual set of responding vehicles, the overall effective participation rate of the cluster can be calculated. ρ Represented as: ； In the formula, s i Indicates the first i The participation status of the vehicles, if the sampling result is less than p i ,but s i =1, otherwise s i =0; N Represents the set of EV quantities in EVA; Incorporating user preferences to adjust capacity for auxiliary services P up,eff ( t ) and reduced capacity P down,eff ( t ) is represented as: ； ； In the formula, ρ up (t), ρ down (t) represents the effective user engagement rate in the scenarios of upward and downward adjustments, respectively. P up,max ( t ), P down,max ( t ) represents the maximum adjustable power under the boundary conditions.

7. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 1, characterized in that, The optimization model is constructed based on energy balance constraints, power constraints, energy boundary constraints, electric vehicle quantity constraints, battery state constraints, and real-time power constraints, with expenditures including day-ahead electricity purchase cost, real-time peak shaving cost, discharge subsidies, and battery loss compensation, and revenues including capacity revenue and energy revenue.

8. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 7, characterized in that, The objective function is expressed as: ； Among them, EVA's electricity purchase cost in the day-to-day market F 1 is represented as: , R DA ( t (Time period) t The day-ahead market transaction price of electricity; P DA ( t (Time period) t The planned power output set by EVA in the daytime; T For the entire scheduling period of EVA; Δ t For time intervals; The real-time peak shaving cost F2 is expressed as: , and For time period t Real-time market transaction electricity prices within the country; P up ( t )and P down ( t This refers to the real-time upward and downward peaking power of the EVA. Discharge subsidies and battery loss compensation are expressed as follows: , For the first i EVs during the time period t Battery energy loss due to internal discharge; R loss For battery wear and tear costs; R 1 represents the day-ahead discharge subsidy price; For the first i EVs in time t The power during charging; N Represents the set of EV quantities in EVA; EVA's capacity revenue from participating in the peak-shaving market is expressed as follows: , R P The electricity price per unit of peak-shaving capacity; C a The peak-shaving capacity that EVA is committed to providing during peak hours; ω Peak shaving service coefficient; The energy revenue that EVA receives from participating in peak shaving in the ancillary services market is expressed as follows: , and These refer to the real-time upward and downward adjustments of peak-shaving energy electricity prices in the market during the day. E UP ( t )and E Down ( t ) are respectively t The upward and downward peak-shaving energy of the EV cluster during the time period.

9. The EVA bidirectional energy interaction multi-time period control method for ancillary services according to claim 8, characterized in that, The energy balance constraint, power constraint, energy boundary constraint, electric vehicle quantity constraint, battery state constraint, and real-time power constraint are specifically expressed as follows: EVA constraints include: , , , and ;in, For time intervals; and The first i EVs in t The charging and discharging power at any given moment; η c and η d The first i The charging and discharging efficiency of EVs For EVA in t Energy at time -1; and for t -1 time point EVA charging and discharging power; for t The EV is constantly connected to the energy of the EVA; for t The energy that EV leaves EVA at any moment; N This represents the set of EV quantities in EVA. For EVA in t Maximum charging power at any given time; For the first i Maximum charging power limit for EVs For EVA in t The maximum discharge power at any given moment; For the first i Maximum discharge power limit for EVs; and They represent t The maximum and minimum rechargeable capacity of EVA at any given time; The energy boundary constraint is expressed as: , E - ( t )for t The lower limit of EV cluster energy at any given time. E + ( t )for t The maximum energy capacity of the EV cluster at any given time; P + ( t ), P - ( t ) are respectively t The upper and lower limits of the power of the EV cluster at any given time; The EV quantity constraint is expressed as: , N act The number of EVs that are actually charged and participate in regulation; N max The maximum number of EVs that EVA can control; The battery state constraints are expressed as follows: , The minimum SOC value required to participate in DBDR; This represents the ideal state value after intervention and regulation. The power constraint is expressed as: , P ( t This refers to the net power of EVA in real-time regulation; and The real-time upward and downward peak-shaving power of EVA is greater than 0 during real-time control.

10. An EVA bidirectional energy interaction multi-time period control system for ancillary services, characterized in that, include: The acquisition module is used to construct the scheduling feasible domain of a single electric vehicle (EV) under different charging and discharging paths based on the framework of EV aggregator EVA participating in auxiliary services. Based on the scheduling feasible domain of the EV, the charging and discharging boundaries of the EV are determined. Based on the charging and discharging boundaries of the EV, a probability sampling method and EVA aggregation model are used to generate EV clusters of sufficient size, and the energy and power feasible domain boundaries of the EV clusters within the scheduling cycle are calculated. The calculation module is used to establish a user willingness discrimination model to determine the probability of EV users' willingness to participate in regulation; and to determine the user participation ratio by combining the charging demand response strategy CBDR and the discharging demand response strategy DBDR. The CBDR is used to calculate the proportion of EV cluster response to CBDR based on the difference between the real-time electricity price and the EVA price, and the user's sensitivity function to the charging incentive price. The DBDR is used to calculate the discharging subsidy price based on the battery's discharge depth and cycle count, and the user's response function to the discharging subsidy price is used to calculate the proportion of EV cluster response to DBDR. The determination module is used to combine the energy and power feasible domain boundary of the EV cluster within the scheduling cycle, the probability of EV users' willingness to participate in regulation, and the user participation ratio to determine the effective power capacity of EVA that is actually adjusted and reduced in each time period. The optimization module is used to construct an optimization model with the effective power capacity as the control capability constraint and the objective function of maximizing the operating revenue of electric vehicle aggregators. By solving the optimization model, the optimal charging and discharging scheduling scheme is obtained.

Images