V2G aggregator optimization scheduling method considering electric vehicle demand response

Abstract

The present invention discloses a V2G aggregator optimization scheduling method considering electric vehicle demand response, which relates to the field of electric vehicle optimization scheduling. It optimizes scheduling by establishing a day-ahead optimization scheduling model and an intraday optimization scheduling model, and adds a breach of contract emergency solution. It also establishes an electric vehicle response comprehensive evaluation index considering the economic efficiency of the V2G aggregator and the reliability of EVs, and performs scheduling with the optimal value of this index as the goal. In the future, a reward mechanism is set up for the V2G aggregator to issue rewards to EV users who abide by the contract, so as to motivate relevant users to abide by the contract. On the basis of meeting the demand response, the execution of the V2G aggregator's scheduling instructions is improved, the breach of contract risk is reduced, and the benefits of both parties are increased.

Description

Technical Field

[0001] The present invention relates to the field of electric vehicle optimization scheduling, and in particular to a V2G aggregator optimization scheduling method considering electric vehicle demand response. Background Art

[0002] With the increase in sales and utilization of electric vehicles (EVs), EVs, as a new type of load, can participate in grid demand response as a flexible demand-side resource. However, since traditional grid demand response usually uses power plants to increase their own spare capacity and build energy storage to achieve demand response, such methods are too costly and the cost of participating in peak-shaving and frequency regulation is too high, it is feasible to use EVs as demand-side resources to participate in grid peak-shaving and frequency regulation demand response regulation and control, and to participate in grid demand response in the form of vehicle-to-grid (V2G) aggregators.

[0003] However, there are many uncertainties in using V2G aggregators to participate in grid demand response. When V2G participates in demand response, the charging time and location of electric vehicles are affected by the travel habits and needs of the owners. In addition, the willingness of EV owners to participate in V2G services and the battery status of electric vehicles will affect the potential of EVs to participate in demand response to a certain extent. These uncertainties make it impossible for EVs to participate in demand response in real time and accurately, which not only affects the quality of grid operation, but also cannot guarantee the interests of V2G aggregators and users. Therefore, an effective demand response optimization scheduling strategy is needed to regulate EVs.

[0004] However, strategies for EV participation in demand response are usually carried out on a single time scale. For example, relevant strategies are used to formulate methods for EV participation in day-ahead demand response. However, due to the strong randomness of EVs and too many influencing factors, it is difficult to achieve the target effect by relying solely on day-ahead demand response, and it is difficult to achieve real-time and accurate demand response. Therefore, it is necessary to study the V2G aggregator optimization scheduling strategy considering electric vehicle demand response at multiple time scales, and provide solutions for the uncertainty factors of EV users, while ensuring that EV demand response can achieve peak shaving and valley filling of the power grid, and at the same time improve the benefits of V2G aggregators and EVs participating in demand response. Summary of the Invention

[0005] The purpose of the present invention is: to address the above-mentioned problems, the present invention provides a V2G aggregator optimization scheduling method that takes into account the demand response of electric vehicles. It optimizes scheduling by establishing a day-ahead optimization scheduling model and an intraday optimization scheduling model, and adds a breach of contract emergency solution. It also establishes a comprehensive evaluation index of electric vehicle response that takes into account the economy of the V2G aggregator and the reliability of EVs. Scheduling is carried out with the optimal value of this index as the goal, and a reward mechanism is set up for the V2G aggregator in the future to issue rewards to EV users who abide by the contract, so as to motivate relevant users to abide by the contract. On the basis of meeting the demand response, the execution of the V2G aggregator's scheduling instructions is improved, the risk of breach of contract is reduced, and the benefits of both parties are increased.

[0006] The technical solution adopted in the present invention is as follows:

[0007] The V2G aggregator optimization scheduling method considering electric vehicle demand response is implemented in the following steps:

[0008] Step 1: EV users send demand responses to the grid through V2G aggregators, and the grid issues a winning curve based on the demand responses.

[0009] Step 2: The V2G aggregator obtains the winning bid curve from Step 1 and establishes a day-ahead optimization scheduling model based on the credit-power comprehensive evaluation index.

[0010] Step 3: Establish an intraday optimal scheduling model through the comprehensive evaluation index of electric vehicle response of V2G aggregator economy and EV reliability;

[0011] Step 4, solving the day-ahead optimization scheduling model and the intraday optimization scheduling model established in steps 2 and 3;

[0012] Step 5: Establish a demand response result evaluation and compliance reward mechanism for the future.

[0013] Furthermore, the credit-power comprehensive evaluation index in step 2 includes the user's historical credit index and power evaluation index;

[0014] The user's historical credit index is as follows:

[0015]

[0016] Where R ij,its is the initial credit, R ijrec is the recommended credit, k is the number of direct transactions between electric vehicle i and V2G aggregator j, and m is the threshold of the number; T e is the total number of periods in which electric vehicle i participates in the demand credit of V2G aggregator j in a month; and are the charging and discharging power bids of electric vehicle i under V2G aggregator j in period t within one month, and are the actual charging and discharging powers of electric vehicle i under V2G aggregator j in period t within a month;

[0017] Combining the above formula, the power evaluation index is the ratio of the EV's declared power to the average of the EV's total declared power, as shown in the following formula:

[0018]

[0019] Where, and are the dispatchable charging and discharging powers reported by electric vehicle i in V2G aggregator j on the previous day; is the total number of EVs that submitted demand response on the day before, The dispatchable charging and discharging power reported on the V2G aggregation on that date, i.e. the total power reported by EVs;

[0020] In summary, the credit-power comprehensive evaluation index is as follows:

[0021]

[0022] Where w1 and w2 are the day-ahead weight coefficients.

[0023] Furthermore, in step 2, a day-ahead optimization scheduling model is established based on the credit-power comprehensive evaluation index, and the objective function is to maximize the comprehensive evaluation index value, that is, maxf day ;

[0024] The constraints in the day-ahead optimization scheduling model include charging and discharging power constraints, charging and discharging equation constraints, battery safety constraints, travel constraints, and time constraints.

[0025] The specific charging and discharging power constraint is that the bid power issued by the V2G aggregator shall not be higher than the dispatchable power declared by the EV, as shown in the following formula:

[0026]

[0027] The charge and discharge equation constraints are as follows:

[0028]

[0029] Where S t is the battery state of EV at time t, η x,t and η d,t The charging efficiency and discharging efficiency of EV;

[0030] The battery safety constraints are as follows:

[0031]

[0032] Where, They are the maximum safe charge and minimum safe charge of EV batteries respectively;

[0033] The travel constraints are as follows:

[0034]

[0035] Where: D t represents the travel demand of EV at time t;

[0036] The specific time constraint is: let the charging and discharging time period of EV be [T start , T end ], the time constraint is expressed as:

[0037] T start ≤t≤T end (10)

[0038] Where T start With T end Respectively represent the start time and end time of the EV charging and discharging time period.

[0039] Furthermore, the comprehensive evaluation index of electric vehicle response of EV reliability in step 3 is as follows:

[0040] EV user response credibility index As follows:

[0041]

[0042] Where, is the user compliance rate, as shown in the following formula:

[0043]

[0044] Where, and are the dispatchable charging and discharging powers fed back by electric vehicle i in V2G aggregator j within the day;

[0045] The power evaluation index of the intraday optimization scheduling model is as follows:

[0046]

[0047] In summary, the comprehensive evaluation index of electric vehicle response for EV reliability is as follows:

[0048]

[0049] Where w 1,in 、w 2,in and w 3,inis the weight coefficient of the comprehensive evaluation index of EV cluster response; This is the number of EVs reported after the emergency response mechanism was activated; It is the flag indicating whether electric vehicle i in V2G aggregator j is an emergency response vehicle in period t. If so, Take 1, otherwise take 0; is the time priority of EV, as follows:

[0050]

[0051] Where t ij is the estimated arrival time of the electric vehicle i at the charging station, max(t ij ) and min(t ij ) are the maximum and minimum values ​​of the estimated arrival time reported by all EVs participating in the intraday optimization scheduling.

[0052] Furthermore, the establishment of the intraday optimization scheduling model in step 3 is as follows:

[0053] The objective function is defined as the maximum comprehensive evaluation index value of the electric vehicle response of EV reliability, that is, maxf in ;

[0054] The intraday optimization scheduling model specifically includes emergency response constraints, supply and demand constraints, and related constraints in the day-ahead scheduling model;

[0055] The emergency response constraint is specifically to determine if an EV defaults within a day: within a certain time period t, the grid allocates a certain amount of demand response to the V2G aggregator, which must be equal to the dispatchable power reported by the EV within the day. If they are not equal, the default is determined and the default emergency response plan is initiated.

[0056] The supply and demand constraints are as follows:

[0057]

[0058] Where, are the charging and discharging power bids issued by V2G aggregator j to electric vehicle i within a day;

[0059] The charge and discharge equation constraints are as follows:

[0060]

[0061] The battery safety constraints and time constraints are the same as those in the day-ahead optimization scheduling model;

[0062] The travel constraints are as follows:

[0063]

[0064] Furthermore, the specific emergency response constraint plan is as follows:

[0065] The emergency response constraints are as follows:

[0066]

[0067]

[0068] Where, are all invited time periods t after time period t for electric vehicle i in V2G aggregator j. sur The total bid charging and discharging power; are the charging and discharging powers of the demand response of electric vehicle i in the recent V2G aggregator’s winning bid; are the charging and discharging powers reported by EV i in V2G aggregator j within a day after the emergency mechanism is activated; is the battery capacity of electric vehicle i in the V2G aggregator; are the upper and lower limits of the battery safety constraints of electric vehicle i in V2G aggregator j at time t, is the normalization coefficient.

[0069] Furthermore, in the emergency response constraint scheme, if the demand response shortfall cannot be met, the demand response shortfall can be made up by issuing emergency demand response invitations to other idle electric vehicles within the area where the V2G aggregator is located.

[0070] Furthermore, step 4 is specifically to use the CPLEX solver to solve the day-ahead optimization scheduling model to obtain the day-ahead scheduling strategy; and use deep reinforcement learning to solve the intraday optimization scheduling model. Due to the influence of emergency response constraints, the optimization scheduling strategy needs to be updated in real time according to the feedback during the intraday solution.

[0071] Furthermore, the subsequent demand response result evaluation and compliance reward mechanism in step 5 specifically include the benefit evaluation of electric vehicles participating in demand response and the benefit evaluation of V2G aggregators;

[0072] Benefit assessment of electric vehicles participating in demand response:

[0073] The benefits of electric vehicles participating in demand response are shown in the following formula:

[0074]

[0075] Where, is the profit of electric vehicle i participating in demand response under V2G aggregator j within one day; T d is the number of time periods throughout the day; are the charging and discharging electricity revenue, battery loss cost, credit reward, and default penalty of electric vehicle i in V2G aggregator j participating in demand response in period t;

[0076] The charge and discharge benefits are defined as follows:

[0077]

[0078] Where r ij G is the revenue sharing ratio agreed between electric vehicle i and V2G aggregator j; d,sub , G c,sub are the charging and discharging unit prices of EVs participating in demand response; are the actual charging and discharging power of electric vehicle i participating in demand response under V2G aggregator j;

[0079] The battery loss cost is defined as follows:

[0080]

[0081] Where c bat is the battery cost, C max is the battery capacity, D t The discharge depth of the electric vehicle during period t;

[0082] V2G aggregator revenue assessment: V2G aggregator revenue includes revenue sharing from EVs participating in demand response and default costs;

[0083] The revenue share of EVs participating in demand response can be expressed as follows:

[0084]

[0085] The default cost of the V2G aggregator is defined as the default fee that the V2G aggregator j needs to pay to the grid during period t as follows:

[0086]

[0087] Where, ρ V2G is the default settlement unit price between the grid and the V2G aggregator, and Δt is the time when the V2G aggregator defaults.

[0088] Furthermore, the contract compliance reward mechanism is as follows:

[0089] The remaining fee after deducting the default fee paid to the grid from the default fee collected by the V2G aggregator is used as the electric vehicle credit reward, which is issued according to the actual demand response volume participated by the EV, as shown in the following formula:

[0090]

[0091] Where: Indicates that the result is rounded down; A flag indicating that the demand response quantity is qualified; is the default fee that electric vehicle i needs to pay to the V2G aggregator during time period t; is the default fee that V2G aggregator j needs to pay to the grid during time period t; is the total number of EVs with defaults, in is the number of EVs whose compliance rate is less than 1, The number of EVs that defaulted within the day and won the bid in the optimized scheduling within the day.

[0092] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0093] The present invention provides a V2G aggregator optimized scheduling method that considers electric vehicle demand response. The scheduling of electric vehicles is divided into two processes: day-ahead scheduling and intraday scheduling. In day-ahead scheduling, the V2G aggregator, based on its winning bids for grid ancillary services, defines a credit-power comprehensive evaluation index and dispatches EVs with the goal of optimizing the comprehensive evaluation index for each EV. Furthermore, intraday scheduling is performed. To avoid excessive demand response defaults, a default contingency solution is provided for the V2G aggregator. If an EV participating in demand response during day-ahead scheduling is unable to participate in real time, idle EVs in the surrounding area that have previously participated in demand response are invited to participate. A comprehensive electric vehicle response evaluation index that considers the V2G aggregator's economic efficiency and EV reliability is established, and scheduling is performed with the goal of optimizing this index. A reward mechanism is subsequently implemented for the V2G aggregator, rewarding compliant EV users to encourage compliance. This improves the execution of the V2G aggregator's scheduling instructions, reduces default risk, and increases the benefits for both parties, while meeting demand response requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] Figure 1 is a flow chart of a V2G aggregator optimization scheduling method considering electric vehicle demand response according to the present invention;

[0095] Figure 2 This is a result diagram of the day-ahead intraday optimization scheduling in Example 3 of the V2G aggregator optimization scheduling method considering electric vehicle demand response of the present invention. DETAILED DESCRIPTION

[0096] The present invention will be described in detail below with reference to the accompanying drawings.

[0097] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0098] Example 1

[0099] V2G aggregator optimization dispatch method considering electric vehicle demand response, such as Figure 1 As shown, the specific implementation steps are as follows:

[0100] Step 1: After the grid initiates a demand response invitation, EV users apply to participate in the demand response through the relevant platform and reserve the time and power available for participation in advance. The V2G aggregator submits the capacity and price information for participating in the demand response to the grid. Based on this, the grid dispatch center issues the winning bid curve to each responding entity according to the demand response application.

[0101] Step 2: The V2G aggregator obtains the winning bid curve from Step 1 and establishes a day-ahead optimization scheduling model based on the credit-power comprehensive evaluation index.

[0102] Step 3: Establish an intraday optimal scheduling model through the comprehensive evaluation index of electric vehicle response of V2G aggregator economy and EV reliability;

[0103] Step 4, solving the day-ahead optimization scheduling model and the intraday optimization scheduling model established in steps 2 and 3;

[0104] Step 5: Establish a demand response result evaluation and compliance reward mechanism for the future.

[0105] Example 2

[0106] This embodiment provides a specific process for establishing a day-ahead optimization scheduling model, an intraday optimization scheduling model, and subsequent solutions, as follows:

[0107] In step 2, in the day-ahead optimization scheduling, after receiving the winning bid curve issued by the power grid, the V2G aggregator optimizes the scheduling of EV users in its jurisdiction who have applied to participate in demand response and whose credit scores meet the requirements based on the credit-power comprehensive evaluation index;

[0108] The definition of the credit-power comprehensive evaluation index is divided into two aspects. One aspect is the user's historical credit index, as shown in the following formula:

[0109]

[0110] Where R ij,its is the initial credit, R ijrecis the recommended credit, k is the number of direct transactions between electric vehicle i and V2G aggregator j, and m is the threshold of the number, which is set to 20 in this embodiment; T e is the total number of periods in which electric vehicle i participates in the demand credit of V2G aggregator j in a month; and are the charging and discharging power bids of electric vehicle i under V2G aggregator j in period t within one month, and are the actual charging and discharging powers of electric vehicle i under V2G aggregator j in period t within a month;

[0111] In combination with the above formula, on the other hand, the power evaluation index is defined as the ratio of the EV's declared power to the average value of the EV's total declared power, as shown in the following formula:

[0112]

[0113] Where, and are the dispatchable charging and discharging powers reported by electric vehicle i in V2G aggregator j on the previous day; is the total number of EVs that submitted demand response on the day before, The dispatchable charging and discharging power reported on the V2G aggregation on that date, i.e. the total power reported by EVs;

[0114] In summary, the credit-power comprehensive evaluation index is obtained as follows:

[0115]

[0116] Where w1 and w2 are the day-ahead weight coefficients;

[0117] In summary, this embodiment establishes a day-ahead optimization scheduling model based on the credit-power comprehensive evaluation index, and the objective function is to maximize the comprehensive evaluation index value, that is, maxf day ;

[0118] The constraints in the day-ahead optimization scheduling model include charging and discharging power constraints, charging and discharging equation constraints, battery safety constraints, travel constraints, and time constraints.

[0119] The specific charging and discharging power constraint is that the bid power issued by the V2G aggregator shall not be higher than the dispatchable power declared by the EV, as shown in the following formula:

[0120]

[0121] The charge and discharge equation constraint means that there is a certain relationship between the charge and discharge power of the EV and the change in the battery state, as shown in the following equation:

[0122]

[0123] Where S t is the battery state of EV at time t, η c,t and η d,t The charging efficiency and discharging efficiency of EV;

[0124] Battery safety constraints mean that the battery state of an EV must be within a certain range to ensure safe operation of the battery, as shown in the following formula:

[0125]

[0126] Where, They are the maximum safe charge and minimum safe charge of EV batteries respectively;

[0127] Travel constraints mean that the charging and discharging behavior of EVs must meet their travel needs, as shown in the following formula:

[0128]

[0129] Where: D t It represents the travel demand of EV at time t, which can be charging demand or discharging demand;

[0130] Time constraint means that the charging and discharging behavior of EV must be completed within a specific time period. Let the charging and discharging time period of EV be [T start , T end ], the time constraint is expressed as:

[0131] T start ≤t≤T end (10)

[0132] Where T start With T end Respectively represent the start time and end time of the EV charging and discharging time period.

[0133] In this embodiment, the comprehensive evaluation index of electric vehicle response of EV reliability in step 3 is as follows:

[0134] In the user's historical credit index R ij Based on the user compliance rate Combined to obtain EV user response credibility index As follows:

[0135]

[0136] Where, The user compliance rate is used to guide EVs to provide intraday feedback as much as possible according to the day-ahead bid amount. It is defined as the ratio of the dispatchable power fed back by EVs to the day-ahead bid power, as shown in the following formula:

[0137]

[0138] Where, and are the dispatchable charging and discharging powers fed back by electric vehicle i in V2G aggregator j within the day;

[0139] Referring to the power evaluation index defined in the day-ahead scheduling model, the power evaluation index of the intraday optimization scheduling model is as follows:

[0140]

[0141] In summary, the comprehensive evaluation index of electric vehicle response for EV reliability is as follows:

[0142]

[0143] Where w 1,in 、w 2,in and w 3,in is the weight coefficient of the comprehensive evaluation index of EV cluster response; This is the number of EVs reported after the emergency response mechanism was activated; It is the flag indicating whether electric vehicle i in V2G aggregator j is an emergency response vehicle in period t. If so, Take 1, otherwise take 0; is the time priority of EV, as follows:

[0144]

[0145] Where t ij is the estimated arrival time of the electric vehicle i at the charging station, max(t ij ) and min(t ij ) are the maximum and minimum values ​​of the estimated arrival time reported by all EVs participating in the intraday optimization scheduling;

[0146] Based on the above content, the establishment of the intraday optimization scheduling model is as follows:

[0147] The objective function is defined as the maximum comprehensive evaluation index value of the electric vehicle response of EV reliability, that is, maxf in ;

[0148] The intraday optimization scheduling model specifically includes emergency response constraints, supply and demand constraints, and related constraints in the day-ahead scheduling model;

[0149] The emergency response constraint is specifically to determine if an EV defaults within a day. The specific judgment process is as follows: within a certain period of time t, the grid allocates a certain amount of demand response to the V2G aggregator, which must be the same as the dispatchable power reported by the EV within the day. If they are not the same, it is considered a default and the default emergency response plan is activated.

[0150] Utilize electric vehicles that have accepted invitations in the past to supplement the demand response shortfall. However, participating EVs need to charge and discharge under the premise of meeting their own pre-dispatch instructions for subsequent time periods. When V2G aggregators dispatch EVs, they need to consider whether they have time periods after the current time period to participate in demand response and consider battery safety constraints to avoid safety issues. The specific emergency response constraint plan is as follows:

[0151] The emergency response constraints are as follows:

[0152]

[0153] Where, are all invited time periods t after time period t for electric vehicle i in V2G aggregator j. sur The total bid charging and discharging power; are the charging and discharging powers of the demand response of electric vehicle i in the recent V2G aggregator’s winning bid; are the charging and discharging powers reported by EV i in V2G aggregator j within a day after the emergency mechanism is activated; is the battery capacity of electric vehicle i in the V2G aggregator; are the upper and lower limits of the battery safety constraints of electric vehicle i in V2G aggregator j at time t, is the normalization coefficient.

[0154] If the above scheme cannot meet the demand response shortfall, an emergency demand response invitation will be issued to other idle electric vehicles within a certain range in the area where the V2G aggregator is located to make up the demand response shortfall;

[0155] The supply and demand constraints are as follows:

[0156]

[0157] Where, are the charging and discharging power bids issued by V2G aggregator j to electric vehicle i within a day;

[0158] The charge and discharge equation constraints are as follows:

[0159]

[0160] The battery safety constraints and time constraints are the same as those in the day-ahead optimization scheduling model;

[0161] The travel constraints are as follows:

[0162]

[0163] In this embodiment, step 4 specifically involves using the CPLEX solver to solve the day-ahead optimization scheduling model to obtain the day-ahead scheduling strategy; using deep reinforcement learning to solve the intraday optimization scheduling model. Due to the influence of emergency response constraints, the optimization scheduling strategy needs to be updated in real time according to feedback during the intraday solution;

[0164] The process of solving the intraday optimization scheduling model using deep reinforcement learning is as follows:

[0165] Build a deep convolutional neural network (DCN) model, select deep Q-learning (DQN) as the reinforcement learning algorithm, initialize the deep convolutional neural network parameters and reinforcement learning algorithm parameters, train the model, iterate the strategy, optimize the model structure based on feedback, and update the optimization strategy in real time based on newly collected information;

[0166] The future demand response result evaluation and compliance reward mechanism in step 5 of this embodiment specifically includes the benefit evaluation of electric vehicles participating in demand response and the benefit evaluation of V2G aggregators;

[0167] Benefit assessment of electric vehicles participating in demand response:

[0168] The benefit of electric vehicles participating in demand response is the economic benefit obtained from the transaction between them and the V2G aggregator minus their default costs, where the economic benefit is shown as follows:

[0169]

[0170] Where, is the profit of electric vehicle i participating in demand response under V2G aggregator j within one day; T d is the number of time periods throughout the day; are the charging and discharging electricity revenue, battery loss cost, credit reward, and default penalty of electric vehicle i in V2G aggregator j participating in demand response in period t;

[0171] The charge and discharge benefits are defined as follows:

[0172]

[0173] Where r ij G is the revenue sharing ratio agreed between electric vehicle i and V2G aggregator j; d,sub , G c,sub are the charging and discharging unit prices of EVs participating in demand response; are the actual charging and discharging power of electric vehicle i participating in demand response under V2G aggregator j;

[0174] The battery wear cost is defined as follows:

[0175]

[0176] Where c bat is the battery cost, C max is the battery capacity, D t The discharge depth of the electric vehicle during period t;

[0177] V2G aggregator revenue assessment: V2G aggregator revenue includes revenue sharing from EVs participating in demand response and default costs;

[0178] The revenue share of EVs participating in demand response can be expressed as follows:

[0179]

[0180] The default cost of the V2G aggregator is defined as the default fee that the V2G aggregator j needs to pay to the grid during period t as follows:

[0181]

[0182] Where, ρ V2G is the default settlement unit price between the grid and the V2G aggregator, and Δt is the time when the V2G aggregator defaults.

[0183] To encourage electric vehicle users to comply with their contracts and participate in demand response, the designed electricity compliance reward mechanism is as follows:

[0184] The remaining fee after deducting the default fee paid to the grid from the default fee collected by the V2G aggregator is used as the electric vehicle credit reward, which is issued according to the actual demand response volume participated by the EV, as shown in the following formula:

[0185]

[0186] Where: Indicates that the result is rounded down. When it is 1, the user compliance rate of electric vehicle i is qualified; The flag bit for the qualified demand response quantity, 1 indicates that the demand response quantity of electric vehicle i in time period t is qualified, and 0 indicates that it is unqualified; is the default fee that electric vehicle i needs to pay to the V2G aggregator during time period t; is the default fee that V2G aggregator j needs to pay to the grid during time period t; is the total number of EVs with defaults, in is the number of EVs whose compliance rate is less than 1, The number of EVs that defaulted within the day and won the bid in the optimized scheduling within the day.

[0187] Example 3

[0188] Specific case analysis: This embodiment designs two scenarios to verify the effectiveness of the proposed optimized scheduling strategy. Since the proposed strategy designs an emergency response mechanism, the effects of whether this mechanism is enabled are compared:

[0189] Scenario 1: No emergency response mechanism is considered. That is, the V2G aggregator does not perform emergency response scheduling for the demand response shortfall caused by the default of its subordinate EVs.

[0190] Scenario 2: Considering the emergency response mechanism, the proposed method is used for scheduling. When a default occurs, the emergency response mechanism is used to deal with the demand response shortage and rewards are given.

[0191] Table 1 Demand response benefits and default costs for EV users and V2G aggregators

[0192]

[0193] The optimization scheduling results for the day before are as follows: Figure 2 As shown in the figure, only the scheduling status of some EV users is displayed, which shows the role of the proposed emergency response mechanism. For example, EV users 2, 12, and 13 did not submit their demand response plans on the previous day. However, during the intraday scheduling, due to the default of users 8 and 15, their intraday feedback values ​​were 0, resulting in a demand response shortage. Therefore, the V2G aggregator issued a certain amount of demand response power to the idle EV users 2, 12, and 13, enabling them to participate in the demand response and meet the shortage.

[0194] Therefore, the present invention considers the V2G aggregator optimization scheduling strategy of electric vehicle demand response, effectively solves the problem of demand response shortage by setting up an emergency response mechanism, and can significantly reduce the risk of default, increase the benefits of V2G aggregators and EV users, and stimulate healthy competition among EV users.

[0195] The day-ahead optimization scheduling model established based on the comprehensive evaluation index of EV user credit and power and the intraday optimization scheduling model established based on the comprehensive evaluation index of electric vehicle response of V2G aggregator economy and EV reliability can select EVs with high user credit index and large demand response volume for pre-scheduling; on the basis of meeting demand response within the day, it can improve the execution of V2G aggregator scheduling instructions and reduce the risk of default.

[0196] The principles and implementation methods of the present invention are described herein using specific embodiments. The description of the above embodiments is intended only to facilitate understanding of the method and core concept of the present invention. It should be noted that those skilled in the art may make various improvements and modifications to the present invention without departing from the principles of the present invention, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A V2G aggregator optimization scheduling method considering electric vehicle demand response, characterized by: Please follow the steps below to implement: Step 1: EV users send demand responses to the grid through V2G aggregators, and the grid issues a winning curve based on the demand responses. Step 2: The V2G aggregator obtains the winning bid curve from Step 1 and establishes a day-ahead optimization scheduling model based on the credit-power comprehensive evaluation index, which includes the user's historical credit index and power evaluation index. The user's historical credit index is as follows: Where R ij,its is the initial credit, R ijrec is the recommended credit, k is the number of direct transactions between electric vehicle i and V2G aggregator j, and m is the threshold of the number; T e is the total number of periods in which electric vehicle i participates in the demand credit of V2G aggregator j in a month; and are the charging and discharging power bids of electric vehicle i under V2G aggregator j in period t within one month, and are the actual charging and discharging powers of electric vehicle i under V2G aggregator j in period t within a month; Combining the above formula, the power evaluation index is the ratio of the EV's declared power to the average of the EV's total declared power, as shown in the following formula: Where, and are the dispatchable charging and discharging powers reported by electric vehicle i in V2G aggregator j on the previous day; is the total number of EVs that submitted demand response on the day before, The dispatchable charging and discharging power reported on the V2G aggregation on that date, i.e. the total power reported by EVs; In summary, the credit-power comprehensive evaluation index is as follows: Where w1 and w2 are the day-ahead weight coefficients; The day-ahead optimization scheduling model is established based on the credit-power comprehensive evaluation index, and the objective function is to maximize the comprehensive evaluation index value, that is, maxf day ; The constraints in the day-ahead optimization scheduling model include charging and discharging power constraints, charging and discharging equation constraints, battery safety constraints, travel constraints, and time constraints. The specific charging and discharging power constraint is that the bid power issued by the V2G aggregator shall not be higher than the dispatchable power declared by the EV, as shown in the following formula: The charge and discharge equation constraints are as follows: Where S t is the battery state of EV at time t, η c,t and η d,t The charging efficiency and discharging efficiency of EV; The battery safety constraints are as follows: Where, They are the maximum safe charge and minimum safe charge of EV batteries respectively; The travel constraints are as follows: Where: D t represents the travel demand of EV at time t; The specific time constraint is: let the charging and discharging time period of EV be [T start , T end ], the time constraint is expressed as: T start ≤t≤T end (10) Where T start With T end They represent the start and end time of the EV charging and discharging time periods respectively; Step 3: Establish an intraday optimal scheduling model through the comprehensive evaluation index of electric vehicle response of V2G aggregator economy and EV reliability; Step 4, solving the day-ahead optimization scheduling model and the intraday optimization scheduling model established in steps 2 and 3; Step 5: Establish a demand response result evaluation and compliance reward mechanism for the future.

2. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 1 is characterized in that: The comprehensive evaluation index of electric vehicle response of EV reliability in step 3 is as follows: EV user response credibility index As follows: Where, is the user compliance rate, as shown in the following formula: Where, and are the dispatchable charging and discharging powers fed back by electric vehicle i in V2G aggregator j within the day; The power evaluation index of the intraday optimization scheduling model is as follows: In summary, the comprehensive evaluation index of electric vehicle response for EV reliability is as follows: Where w 1,in 、w 2,in and w 3,in is the weight coefficient of the comprehensive evaluation index of EV cluster response; This is the number of EVs reported after the emergency response mechanism was activated; It is the flag indicating whether electric vehicle i in V2G aggregator j is an emergency response vehicle in period t. If so, Take 1, otherwise take 0; is the time priority of EV, as follows: Where t ij is the estimated arrival time of the electric vehicle i at the charging station, max(t ij ) and min(T ij ) are the maximum and minimum values ​​of the estimated arrival time reported by all EVs participating in the intraday optimization scheduling.

3. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 2 is characterized in that: The establishment of the intraday optimization scheduling model in step 3 is as follows: The objective function is defined as the maximum comprehensive evaluation index value of the electric vehicle response of EV reliability, that is, maxf in ; The intraday optimization scheduling model specifically includes emergency response constraints, supply and demand constraints, and related constraints in the day-ahead scheduling model; The emergency response constraint is specifically to determine if an EV defaults within a day: within a certain time period t, the grid allocates a certain amount of demand response to the V2G aggregator, which must be equal to the dispatchable power reported by the EV within the day. If they are not equal, the default is determined and the default emergency response plan is initiated. The supply and demand constraints are as follows: Where, are the charging and discharging power bids issued by V2G aggregator j to electric vehicle i within a day; The charge and discharge equation constraints are as follows: The battery safety constraints and time constraints are the same as those in the day-ahead optimization scheduling model; The travel constraints are as follows:

4. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 3 is characterized in that: The specific emergency response constraint plan is as follows: The emergency response constraints are as follows: Where, are all invited time periods t after time period t for electric vehicle i in V2G aggregator j. sur The total bid charging and discharging power; are the charging and discharging powers of the demand response of electric vehicle i in the recent V2G aggregator’s winning bid; are the charging and discharging powers reported by EV i in V2G aggregator j within a day after the emergency mechanism is activated; is the battery capacity of electric vehicle i in the V2G aggregator; are the upper and lower limits of the battery safety constraints of electric vehicle i in V2G aggregator j at time t, is the normalization coefficient.

5. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 4 is characterized in that: In the emergency response constraint scheme, if the demand response shortfall cannot be met, the demand response shortfall can be made up by issuing emergency demand response invitations to other idle electric vehicles within the area where the V2G aggregator is located.

6. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 1 is characterized in that: Specifically, step 4 includes using the CPLEX solver to solve the day-ahead optimization scheduling model to obtain the day-ahead scheduling strategy; and using deep reinforcement learning to solve the intraday optimization scheduling model. Due to the influence of emergency response constraints, the optimization scheduling strategy needs to be updated in real time according to feedback during the intraday solution.

7. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 1 is characterized in that: The subsequent demand response result evaluation and compliance reward mechanism in step 5 specifically include the profit evaluation of electric vehicles participating in demand response and the profit evaluation of V2G aggregators; Benefit assessment of electric vehicles participating in demand response: The benefits of electric vehicles participating in demand response are shown in the following formula: Where, is the profit of electric vehicle i participating in demand response under V2G aggregator j within one day; T d is the number of time periods throughout the day; are the charging and discharging electricity revenue, battery loss cost, credit reward, and default penalty of electric vehicle i in V2G aggregator j participating in demand response in period t; The charge and discharge benefits are defined as follows: Where r ij G is the revenue sharing ratio agreed between electric vehicle i and V2G aggregator j; d,sub , G c,sub are the charging and discharging unit prices of EVs participating in demand response; are the actual charging and discharging power of electric vehicle i participating in demand response under V2G aggregator j; The battery loss cost is defined as follows: Where c bat is the battery cost, C max is the battery capacity, D t The discharge depth of the electric vehicle during period t; V2G aggregator revenue assessment: V2G aggregator revenue includes revenue sharing from EVs participating in demand response and default costs; The revenue share of EVs participating in demand response can be expressed as follows: The default cost of the V2G aggregator is defined as the default fee that the V2G aggregator j needs to pay to the grid during period t as follows: Where, ρ V2G is the default settlement unit price between the grid and the V2G aggregator, and Δt is the time when the V2G aggregator defaults.

8. The V2G aggregator optimization scheduling method considering electric vehicle demand response according to claim 7 is characterized in that: The compliance reward mechanism is as follows: The remaining fee after deducting the default fee paid to the grid from the default fee collected by the V2G aggregator is used as the electric vehicle credit reward, which is issued according to the actual demand response volume participated by the EV, as shown in the following formula: Where: Indicates that the result is rounded down; A flag indicating that the demand response quantity is qualified; is the default fee that electric vehicle i needs to pay to the V2G aggregator during time period t; is the default fee that V2G aggregator j needs to pay to the grid during time period t; is the total number of EVs with defaults, in is the number of EVs whose compliance rate is less than 1, The number of EVs that defaulted within the day and won the bid in the optimized scheduling within the day.

Images