Electric vehicle peak regulation service energy storage optimization method based on column sum constraint generation algorithm

Abstract

The invention discloses an electric vehicle peak regulation service energy storage optimization method based on a column sum constraint generation algorithm, and relates to the technical field of energy storage optimization configuration in an electric power system, and the method comprises the steps: constructing an electric vehicle aggregator peak regulation auxiliary service energy storage optimization configuration model; converting the model into a two-stage robust optimization model; solving the two-stage robust optimization model based on a column sum constraint generation algorithm; and performing result analysis and sensitivity analysis to obtain an aggregator peak regulation and storage allocation scheme. The method comprises the following steps: constructing an electric vehicle aggregator peak regulation auxiliary service energy storage optimization configuration model, converting the model into a two-stage robust optimization model, solving the two-stage robust optimization model based on a column sum constraint generation algorithm, and carrying out result analysis and sensitivity analysis to obtain an aggregator peak regulation distribution and storage scheme. Therefore, a more effective aggregator peak regulation auxiliary service market strategy is guided, and the purposes of improving the operation income of the aggregator and reducing the deviation punishment of the peak regulation auxiliary service market are achieved.

Description

Technical Field

[0001] This invention relates to the field of energy storage optimization configuration technology in power systems, and in particular to an energy storage optimization method for electric vehicle peak shaving services based on column and constraint generation algorithms. Background Technology

[0003] With the increasing proportion of renewable energy generation, its intermittency and uncertainty place higher demands on the stable operation of the power grid. Electric vehicle aggregators (EVs) can centrally manage the batteries of multiple EVs, allowing them to participate in electricity market transactions as a whole, including but not limited to peak shaving, frequency regulation, and backup power ancillary services. However, the demand response of EV users participating in EV aggregators is also uncertain, leading to deviation penalties in meeting ancillary service demands. Configuring energy storage can effectively mitigate load fluctuations caused by the uncertainty of EV demand response. Simultaneously, market-side electricity prices are also uncertain; energy storage can effectively alleviate the decision-making risks in bidding for ancillary services. Therefore, how to consider the impact of multiple uncertainties and formulate the optimal energy storage configuration strategy with the goal of maximizing the peak shaving revenue of EV aggregators has become an urgent problem to be solved. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide an energy storage optimization method for electric vehicle peak shaving services based on a column and constraint generation algorithm. By constructing an energy storage optimization configuration model for peak shaving auxiliary services of electric vehicle aggregators, the model is transformed into a two-stage robust optimization model. The two-stage robust optimization model is solved based on the column and constraint generation algorithm, and the results and sensitivity analysis are performed to obtain the peak shaving energy storage allocation scheme of aggregators. This guides more effective aggregator peak shaving auxiliary service market strategies to improve aggregator operating revenue and reduce the market deviation penalty for peak shaving auxiliary services.

[0005] The technical solution adopted in this invention is: an optimization method for peak-shaving energy storage services for electric vehicles based on a column and constraint generation algorithm, the method comprising: S1, Construct an energy storage optimization configuration model for peak shaving auxiliary services of electric vehicle aggregators; S2 transforms the model into a two-stage robust optimization model; S3, a two-stage robust optimization model is solved based on a column and constraint generation algorithm; S4. Perform result analysis and sensitivity analysis to obtain the aggregator peak shaving and storage scheme.

[0006] Further, step S1 includes the following steps: Step S11: Establish the objective function of the peak-shaving auxiliary service energy storage optimization configuration model for electric vehicle aggregators; Step S12: Establish day-ahead regulatory constraints for electric vehicle aggregators; Step S13: Establish intraday control constraints for electric vehicle aggregators; Step S14: Establish daily reporting constraints for electric vehicle aggregators; Step S15: Establish intraday deviation constraints for electric vehicle aggregators.

[0007] Further, step S11 includes the following steps: The objective function for modeling the peak-shaving market energy storage optimization configuration of electric vehicle aggregators is the average daily total cost C. total Minimum, C total Daily investment costs for energy storage, etc. Typical Electricity Purchase Costs in the Daily Energy Market Typical daily peak-shaving ancillary service market recent winning bid revenue R EVA and typical daily peak-shaving ancillary service market response deviation loss It consists of four parts, namely: The expressions for each part are as follows: In the formula, ρ represents the total cost of energy storage configuration; ρ represents the discount rate; n represents the expected lifespan of the energy storage; C P Indicates the unit power cost of energy storage batteries and related equipment; C E P represents the unit capacity cost of energy storage batteries; ESS,max Indicates the rated power of the configured energy storage; E ESS,max Indicates the rated capacity of the configured energy storage; This indicates the electricity purchase cost for electric vehicle aggregators in the electricity market; This indicates the actual daily adjustment power of the electric vehicle aggregator; The electric vehicle aggregator represents the electricity purchase price for time period t in the electricity market; Δt represents the unit of time; R EVA This indicates the recent winning bid revenue in the typical daily peak-shaving ancillary services market; K t P represents the market coefficient during the clearing period. t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; P t base This indicates the baseline load of peak-shaving power in time period t, calculated and verified by the North China Power Grid-Load Storage Platform based on test results. This indicates the peak-shaving clearing price in the peak-shaving ancillary services market; T peak Indicates peak-shaving period; This represents the market response deviation loss for peak-shaving ancillary services on a typical day.

[0008] Further, step S12 includes the following steps: Electric vehicles are divided into two categories: Category I electric vehicles use fixed-power charging, and Category II electric vehicles use sequential charging. For Category I electric vehicles, the control mode for vehicle i is as follows: In the formula, P represents the actual charging power of electric vehicle i in time period t. i rated The rated charging power for electric vehicle i; η represents the battery charge of electric vehicle i in time period t; i E represents the charging efficiency of electric vehicle i; i,or E represents the initial charge of electric vehicle i; i,ex Δt represents the expected off-grid charge of electric vehicle i; Δt represents the unit time. For the second type of electric vehicle, the control mode of vehicle i is as follows: In the formula, This indicates the day-ahead controlled charging power of the second type of electric vehicle i during time period t; For the day-ahead regulation of the second type of electric vehicle i during time period t; E i,or E represents the initial charge of electric vehicle i; i,ex Δt represents the expected off-grid charge of electric vehicle i; Δt represents the unit of time. The first and second categories of electric vehicles can be merged into one category, and electric vehicles can be modeled uniformly as follows: In the formula, and These refer to the daytime controlled power and battery capacity of electric vehicles, respectively. This represents the pre-scheduled charging reservation information for electric vehicles, and is an integer variable between 0 and 1, where 1 represents the first type of electric vehicle and 0 represents the second type of electric vehicle. Let Δt be the actual charging power of electric vehicle i as a first-class electric vehicle during time period t; Δt represents the unit time duration.

[0009] Further, step S13 includes the following steps: The intraday regulation constraints for electric vehicle aggregators are divided into two parts: intraday regulation constraints for electric vehicles and intraday regulation constraints for energy storage. The intraday control constraints for electric vehicles are basically the same as the daytime control constraints, except for the changes in constraints caused by the introduction of uncertain variables, as follows: In the formula, P represents the actual controlled charging power of electric vehicle i during the daytime period t; i rated The rated charging power for electric vehicle i; E i,or E represents the initial charge of electric vehicle i; i,ex This represents the expected off-grid charge of electric vehicle i; The daily electricity consumption of the second type of electric vehicle i during time period t; and These refer to the daily adjustable power and battery level of the electric vehicle, respectively. For the uncertain variable of the intraday orderly charging contract for electric vehicles, a value of 1 is assigned to represent the first type of electric vehicle, while 0 represents the second type of electric vehicle; Let be the actual charging power of electric vehicle i as a first-class electric vehicle during time period t; Δt represents the unit time duration. The intraday control constraints for energy storage are as follows: In the formula, P ESS,max Indicates the rated power of the configured energy storage; E ESS,max Indicates the rated capacity of the configured energy storage; P t ch P t dis These represent the energy storage charging power and discharging power, respectively; η ch η dis These represent the energy storage charging efficiency and discharging efficiency, respectively. Δt represents the real-time daily energy storage capacity during time period t; Δt represents the unit duration.

[0010] Further, step S14 includes the following steps: The electric vehicle aggregator recently won a bid that meets the following constraints regarding power output: In the formula, the first line indicates that the day-ahead regulation power of the electric vehicle aggregator consists of the sum of the actual power of all internal electric vehicles and energy storage electric vehicle aggregators; the second line indicates that the bid power during peak shaving periods is not less than the baseline load power; P t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; N EV This indicates the total number of electric vehicles within the electric vehicle aggregator. For the daytime regulated power of electric vehicles; P t baseThis represents the baseline load of peak-shaving power during time period t, calculated and verified by the North China Power Grid-Load-Storage Platform based on test results; T peak Indicates the peak-shaving period.

[0011] Further, step S15 includes: The actual power regulation by electric vehicle aggregators must meet the following constraints: In the formula, the first line indicates that the actual power of the electric vehicle aggregator is the sum of the actual power of all internal electric vehicles and energy storage; the second line indicates that the actual power during the peak-shaving period is not less than the baseline load power and not greater than the day-ahead bid power. This indicates the actual daily power adjustment of the electric vehicle aggregator; N EV This indicates the total number of electric vehicles within the electric vehicle aggregator. For the daytime regulated power of electric vehicles; P t ch P t dis These represent the energy storage charging power and discharging power, respectively; P t base This represents the baseline load of peak-shaving power during time period t, calculated and verified by the North China Power Grid-Load-Storage Platform based on test results; P t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; T peak Indicates the peak-shaving period.

[0012] Furthermore, the model is transformed into a two-stage robust optimization model, including the following steps: Step S31: Construct the uncertainty set of the two-stage robust optimization model; The model is improved using robust optimization methods, and relevant box-type uncertainty sets are constructed: In the formula, This represents the electricity purchase price for electric vehicle aggregators in the electricity market during time period t. This indicates the clearing price in the peak-shaving ancillary services market; This indicates the projected electricity market price and peak-shaving market clearing price. This is a binary variable representing whether there is a deviation in electricity prices during time period t; a value of 1 indicates that a deviation has occurred. energy , Γ peak Indicates the maximum number of periods during which electricity prices deviate; β energy β peak Indicates the degree of deviation in electricity prices; This is a binary variable representing whether the electric vehicle user is complying with the contract; a value of 1 indicates compliance. EV This represents the number of electric vehicles that the aggregator can control; β EV This indicates the maximum percentage of electric vehicles that do not comply with the contract. Step S32: Linearize some of the constraints; The rated power formula P for configuring energy storage ESS,max Linearize the following formula: Introducing intermediate variables The rated capacity E of the configured energy storage ESS,max The linearization is expressed by the following formula: Step S33: Reconstruct the model into a two-stage robust optimization model; The formula that minimizes the average daily total cost The reconstruction into a two-stage robust optimization model is transformed into a two-stage robust optimization model; the two-stage robust optimization model is as follows: In the formula, y is a continuous variable in the first-stage optimization problem; x is a continuous variable in the second-stage optimization problem; and u is an uncertain scenario variable. The specific expression is:

[0013] Furthermore, the two-stage robust optimization model is solved based on a column and constraint generation algorithm, including: The two-stage robust optimization model is decomposed into a main problem and sub-problems. In the main problem, the influence of uncertain variables is not considered first, and it is solved as a deterministic optimization. Based on deterministic optimization solutions, relatively severe scenarios and corresponding sub-problem decision variables and constraints are continuously added, thereby continuously improving the upper and lower bounds of the objective function until the algorithm converges.

[0014] Furthermore, results analysis and sensitivity analysis were conducted to derive the aggregator's peak shaving and storage scheme, including: Inputting data such as peak shaving market, predicted clearing price of electricity in the power market, and internal electric vehicle charging data of electric vehicle aggregators, the algorithm uses column and constraint generation to solve the problem, performs sensitivity analysis, and obtains the optimal energy storage configuration scheme for the peak shaving ancillary service market of aggregators by setting appropriate robust parameters.

[0015] The beneficial effects of this invention are as follows: By constructing an optimized configuration model for peak shaving ancillary services and energy storage for electric vehicle aggregators, the model is transformed into a two-stage robust optimization model. The two-stage robust optimization model is solved based on a column and constraint generation algorithm. The results and sensitivity analysis are then performed to obtain aggregator peak shaving and energy storage allocation schemes. This guides more effective aggregator peak shaving ancillary service market strategies, thereby improving aggregator operating revenue and reducing market deviation penalties for peak shaving ancillary services. Attached Figure Description

[0016] Figure 1 This is a flowchart of the algorithm of the present invention.

[0017] Figure 2 This is the prediction result of the market clearing price and market coefficient for peak-shaving ancillary services in this invention; Figure 3 This is a flowchart of the algorithm for generating columns and constraints for aggregator peak shaving and storage in this invention; Figure 4 This is a graph showing the results of how cost, rated energy storage power, and rated energy storage capacity change with robust parameters in this invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0019] See appendix Figure 1-4 A method for optimizing energy storage for peak-shaving services of electric vehicles based on a column and constraint generation algorithm, the method comprising: S1, Construct an energy storage optimization configuration model for peak shaving auxiliary services of electric vehicle aggregators; S2 transforms the model into a two-stage robust optimization model; S3, a two-stage robust optimization model is solved based on a column and constraint generation algorithm; S4. Perform result analysis and sensitivity analysis to obtain the aggregator peak shaving and storage scheme.

[0020] As the global energy structure accelerates its transition to a low-carbon model, the large-scale grid connection of renewable energy sources such as wind and solar power poses a severe challenge to the peak-shaving capacity of the power system due to their inherent intermittency and volatility. Simultaneously, the rapid popularization of electric vehicles has spurred the development of massive distributed energy storage resources. Their battery systems can flexibly participate in grid regulation through orderly charging technology, providing peak-shaving ancillary services to the power system. Against this backdrop, electric vehicle aggregators, as key hubs connecting dispersed electric vehicle resources with the electricity market, urgently need to optimize the allocation of energy storage resources to mitigate the negative impacts of multiple uncertainties and achieve coordinated scheduling of large-scale electric vehicle clusters, thereby improving peak-shaving efficiency and economic benefits.

[0021] Step S1 includes the following steps: Step S11: Establish the objective function of the peak-shaving auxiliary service energy storage optimization configuration model for electric vehicle aggregators; Step S12: Establish day-ahead regulatory constraints for electric vehicle aggregators; Step S13: Establish intraday control constraints for electric vehicle aggregators; Step S14: Establish daily reporting constraints for electric vehicle aggregators; Step S15: Establish intraday deviation constraints for electric vehicle aggregators.

[0022] Further, step S1 includes the following steps: Step S11: Establish the objective function of the peak-shaving auxiliary service energy storage optimization configuration model for electric vehicle aggregators; Step S12: Establish day-ahead regulatory constraints for electric vehicle aggregators; Step S13: Establish intraday control constraints for electric vehicle aggregators; Step S14: Establish daily reporting constraints for electric vehicle aggregators; Step S15: Establish intraday deviation constraints for electric vehicle aggregators. Step S11 includes the following steps: The objective function for modeling the peak-shaving market energy storage optimization configuration of electric vehicle aggregators is the average daily total cost C. total Minimum, C total Daily investment costs for energy storage, etc. Typical Electricity Purchase Costs in the Daily Energy Market Typical daily peak-shaving ancillary service market recent winning bid revenue R EVA and typical daily peak-shaving ancillary service market response deviation loss It consists of four parts, namely: The expressions for each part are as follows: In the formula, ρ represents the total cost of energy storage configuration; ρ represents the discount rate; n represents the expected lifespan of the energy storage; C P Indicates the unit power cost of energy storage batteries and related equipment; C E P represents the unit capacity cost of energy storage batteries; ESS,max Indicates the rated power of the configured energy storage; E ESS,max Indicates the rated capacity of the configured energy storage; This indicates the electricity purchase cost for electric vehicle aggregators in the electricity market; This indicates the actual daily adjustment power of the electric vehicle aggregator; The electric vehicle aggregator represents the electricity purchase price for time period t in the electricity market; Δt represents the unit of time; R EVA This indicates the recent winning bid revenue in the typical daily peak-shaving ancillary services market; K t P represents the market coefficient during the clearing period. t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; P t base This indicates the baseline load of peak-shaving power in time period t, calculated and verified by the North China Power Grid-Load Storage Platform based on test results. This indicates the peak-shaving clearing price in the peak-shaving ancillary services market; T peak Indicates peak-shaving period; This represents the market response deviation loss for peak-shaving ancillary services on a typical day.

[0023] Step S12 includes the following steps: Electric vehicles are divided into two categories: Category I electric vehicles use fixed-power charging, and Category II electric vehicles use sequential charging. For Category I electric vehicles, the control mode for vehicle i is as follows: In the formula, P represents the actual charging power of electric vehicle i in time period t. i rated The rated charging power for electric vehicle i; η represents the battery charge of electric vehicle i in time period t; i E represents the charging efficiency of electric vehicle i; i,or E represents the initial charge of electric vehicle i; i,ex Δt represents the expected off-grid charge of electric vehicle i; Δt represents the unit of time. For the second type of electric vehicle, the control mode of vehicle i is as follows: In the formula, This indicates the day-ahead controlled charging power of the second type of electric vehicle i during time period t; For the day-ahead regulation of the second type of electric vehicle i during time period t; E i,or E represents the initial charge of electric vehicle i; i,ex Δt represents the expected off-grid charge of electric vehicle i; Δt represents the unit of time. The first and second categories of electric vehicles can be merged into one category, and electric vehicles can be modeled uniformly as follows: In the formula, and These refer to the daytime controlled power and battery capacity of electric vehicles, respectively. This represents the pre-scheduled charging reservation information for electric vehicles, and is an integer variable between 0 and 1, where 1 represents the first type of electric vehicle and 0 represents the second type of electric vehicle. Let Δt be the actual charging power of electric vehicle i as a first-class electric vehicle during time period t; Δt represents the unit time duration.

[0024] Step S13 includes the following steps: The intraday regulation constraints for electric vehicle aggregators are divided into two parts: intraday regulation constraints for electric vehicles and intraday regulation constraints for energy storage. The intraday control constraints for electric vehicles are basically the same as the daytime control constraints, except for the changes in constraints caused by the introduction of uncertain variables, as follows: In the formula, P represents the actual controlled charging power of electric vehicle i during the daytime period t; i rated The rated charging power for electric vehicle i; E i,or E represents the initial charge of electric vehicle i; i,ex This represents the expected off-grid charge of electric vehicle i; The daily electricity consumption of the second type of electric vehicle i during time period t; and These refer to the daily adjustable power and battery level of the electric vehicle, respectively. For the uncertain variable of the intraday orderly charging contract for electric vehicles, a value of 1 is assigned to represent the first type of electric vehicle, while 0 represents the second type of electric vehicle; Let be the actual charging power of electric vehicle i as a first-class electric vehicle during time period t; Δt represents the unit time duration. The intraday control constraints for energy storage are as follows: In the formula, P ESS,max Indicates the rated power of the configured energy storage; E ESS,max Indicates the rated capacity of the configured energy storage; P t ch P t dis These represent the energy storage charging power and discharging power, respectively; η ch η dis These represent the energy storage charging efficiency and discharging efficiency, respectively. Δt represents the real-time daily energy storage capacity during time period t; Δt represents the unit duration.

[0025] Step S14 includes the following steps: The electric vehicle aggregator recently won a bid that meets the following constraints regarding power output: In the formula, the first line indicates that the day-ahead regulation power of the electric vehicle aggregator consists of the sum of the actual power of all internal electric vehicles and energy storage electric vehicle aggregators; the second line indicates that the bid power during peak shaving periods is not less than the baseline load power; P t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; N EV This indicates the total number of electric vehicles within the electric vehicle aggregator. For the daytime regulated power of electric vehicles; P t base This represents the baseline load of peak-shaving power during time period t, calculated and verified by the North China Power Grid-Load-Storage Platform based on test results; T peak Indicates the peak-shaving period.

[0026] Step S15 includes: The actual power regulation by electric vehicle aggregators must meet the following constraints: In the formula, the first line indicates that the actual power of the electric vehicle aggregator is the sum of the actual power of all internal electric vehicles and energy storage; the second line indicates that the actual power during the peak-shaving period is not less than the baseline load power and not greater than the day-ahead bid power. This indicates the actual daily power adjustment of the electric vehicle aggregator; N EV This indicates the total number of electric vehicles within the electric vehicle aggregator. For the daytime regulated power of electric vehicles; P t ch P t dis These represent the energy storage charging power and discharging power, respectively; P t base This represents the baseline load of peak-shaving power during time period t, calculated and verified by the North China Power Grid-Load-Storage Platform based on test results; P t bid This indicates the power output of electric vehicle aggregators in the peak-shaving auxiliary service market during time period t; T peak Indicates the peak-shaving period.

[0027] Transforming the model into a two-stage robust optimization model includes the following steps: Step S31: Construct the uncertainty set of the two-stage robust optimization model; The model is improved using robust optimization methods, and relevant box-type uncertainty sets are constructed: In the formula, This represents the electricity purchase price for electric vehicle aggregators in the electricity market during time period t. This indicates the clearing price in the peak-shaving ancillary services market; This indicates the projected electricity market price and peak-shaving market clearing price. This is a binary variable representing whether there is a deviation in electricity prices during time period t; a value of 1 indicates that a deviation has occurred. energy , Γ peak Indicates the maximum number of periods during which electricity prices deviate; β energy β peak Indicates the degree of deviation in electricity prices; This is a binary variable representing whether the electric vehicle user is complying with the contract; a value of 1 indicates compliance. EV This represents the number of electric vehicles that the aggregator can control; β EV This indicates the maximum percentage of electric vehicles that do not comply with the contract. Step S32: Linearize some of the constraints; The rated power formula P for configuring energy storage ESS,max Linearize the following formula: Introducing intermediate variables The rated capacity E of the configured energy storage ESS,max The linearization is expressed by the following formula: Step S33: Reconstruct the model into a two-stage robust optimization model; The formula that minimizes the average daily total cost The reconstruction into a two-stage robust optimization model is transformed into a two-stage robust optimization model; the two-stage robust optimization model is as follows: In the formula, y is a continuous variable in the first-stage optimization problem; x is a continuous variable in the second-stage optimization problem; and u is an uncertain scenario variable. The specific expression is:

[0028] Solving a two-stage robust optimization model based on a column and constraint generation algorithm includes: The two-stage robust optimization model is decomposed into a main problem and sub-problems. In the main problem, the influence of uncertain variables is not considered first, and it is solved as a deterministic optimization.

[0029] Based on deterministic optimization solutions, relatively severe scenarios and corresponding sub-problem decision variables and constraints are continuously added, thereby continuously improving the upper and lower bounds of the objective function until the algorithm converges.

[0030] S3, based on the column and constraint generation algorithm, solves the two-stage robust optimization model, including the following steps:

[0031] Step S31: Decompose the two-stage robust optimization model into a main problem and sub-problems.

[0032] Substituting the variables into the objective function, we obtain the complete two-stage robust optimization model as follows: The main question is as follows: The sub-problems are as follows: in, It is a 0-1 variable, and a value of 1 indicates that the time period t is a peak-shaving period.

[0033] Step S32: Transform the subproblem minmax structure into min structure using kkt conditions.

[0034] Introducing the Lagrange function, as follows: Simultaneously, intermediate variables are introduced to linearize the KKT conditions, as follows: Step S33: Iterative solution of the main problem and subproblems. The main problem is initially solved as a deterministic optimization, without considering the influence of uncertain variables. Based on the deterministic optimization solution, increasingly severe scenarios and corresponding subproblem decision variables and constraints are continuously added, thereby continuously improving the upper and lower bounds of the objective function until the algorithm converges.

[0035] Results and sensitivity analyses were conducted to derive aggregator peak shaving and storage solutions, including: Inputting data such as peak shaving market, predicted clearing price of electricity in the power market, and internal electric vehicle charging data of electric vehicle aggregators, the algorithm uses column and constraint generation to solve the problem, performs sensitivity analysis, and obtains the optimal energy storage configuration scheme for the peak shaving ancillary service market of aggregators by setting appropriate robust parameters.

[0036] Step S41: Input peak shaving market, electricity market forecast clearing price, and electric vehicle charging data from the electric vehicle aggregator.

[0037] The electric vehicle (EV) charging data internally generated by the EV aggregator is based on simulations of 1000 EV charging data points generated according to user travel patterns. The electricity market clearing price is derived from historical data. For example... Figure 2 The diagram shows the peak-shaving market clearing electricity price.

[0038] Step S42: Solve using a column and constraint generation algorithm. For example... Figure 3 The flowchart shown is the algorithm for generating columns and constraints for aggregator peak shaving and storage. According to Γ energy =24,Γ peak =12,β energy =β peak =0.15, β EV Solve using a value of 0.2.

[0039] Step S43: Perform sensitivity analysis and obtain the optimal energy storage configuration scheme for the aggregator peak shaving ancillary service market by setting appropriate robust parameters.

[0040] When robustness parameters change, the cost, rated power of energy storage, and rated capacity of energy storage change, such as Figure 4 The graph shows how cost, rated energy storage power, and rated energy storage capacity vary with robust parameters. Figure 4 The units on the central vertical axis, from left to right, are yuan, kWh, and kW.

[0041] As the robustness parameter increases, all three parameters, though fluctuating, continue to increase. Therefore, the worst-case energy storage configuration is selected, i.e., E = 1787 kWh and P = 818.6 kW.

[0042] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An electric vehicle peak shaving service energy storage optimization method based on column and constraint generation algorithm, characterized in that, The method comprises: S1, constructing an electric vehicle aggregator peak shaving auxiliary service energy storage optimization configuration model; S2, converting the model into a two-stage robust optimization model; S3, solving the two-stage robust optimization model based on a column and constraint generation algorithm; S4, performing result analysis and sensitivity analysis to obtain an aggregator peak shaving and storage scheme.

2. The column-based and constraint generation algorithm-based optimization method for electric vehicle peak shaving service energy storage according to claim 1, characterized in that, The step S1 comprises the following steps: Step S11, establishing an objective function of the electric vehicle aggregator peak shaving auxiliary service energy storage optimization configuration model; Step S12, establishing a day-ahead regulation constraint condition of the electric vehicle aggregator; Step S13, establishing an intra-day regulation constraint condition of the electric vehicle aggregator; Step S14, establishing a day-ahead reporting constraint condition of the electric vehicle aggregator; Step S15, establishing an intra-day deviation constraint condition of the electric vehicle aggregator.

3. The column-based and constraint generation algorithm-based optimization method for electric vehicle peak shaving service energy storage according to claim 2, characterized in that, The step S11 comprises the following steps: The objective function of the modeling of the optimal configuration of the peak market energy storage of the aggregator of electric vehicles is the daily total cost C total min, C total The daily value investment cost of the energy storage and the like The electricity purchase cost of the typical day energy market The day-ahead winning income R of the typical day peak regulation auxiliary service market EVA And the response deviation loss of the typical day peak regulation auxiliary service market The four parts are: The expressions of the parts are as follows: In the formula, represents the total cost of energy storage configuration; p represents the discount rate; n represents the expected life of energy storage; C P represents the unit power cost of energy storage batteries and supporting equipment; C E represents the unit capacity cost of energy storage batteries; P ESS,max represents the rated power of the energy storage configuration; E ESS,max represents the rated capacity of the energy storage configuration; represents the electricity purchase cost of the electric vehicle aggregator in the electricity market; represents the actual regulation power of the electric vehicle aggregator within a day; represents the electricity purchase price of the electric vehicle aggregator in the electricity market at time period t; represents the unit time length; R EVA represents the pre-market winning income of the typical day peak shaving auxiliary service market; K t represents the market coefficient of the off-peak period; P t bid represents the pre-market winning power of the electric vehicle aggregator in the peak shaving auxiliary service market at time period t; P t base represents the baseline load of the peak shaving power calculated by the North China source network energy storage platform according to the test results at time period t; represents the peak shaving off-peak price of the peak shaving auxiliary service market; T peak represents the peak shaving period; represents the response deviation loss of the typical day peak shaving auxiliary service market.

4. The column and constraint generation algorithm based optimization method for electric vehicle peak shaving service energy storage according to claim 2, wherein, The step S12 comprises the following steps: The electric vehicle types are divided into two categories: the first category of electric vehicles is fixed power charging, and the second category of electric vehicles is orderly charging; For the first category of electric vehicles, the control mode of vehicle i is as follows: wherein, is the actual charging power of the first type of electric vehicle i at time period t; P i rated is the rated charging power of the electric vehicle i; is the electric quantity of the first type of electric vehicle i at time period t; η i denotes the charging efficiency of the electric vehicle i; E i,or denotes the initial electric quantity of the electric vehicle i; E i,ex denotes the off-grid expected electric quantity of the electric vehicle i; Δt denotes a unit time length; For the second category of electric vehicles, the control mode of vehicle i is as follows: wherein, denotes the day-ahead regulated charging power of the second type of electric vehicle i at time period t; denotes the day-ahead regulated energy of the second type of electric vehicle i at time period t;E i,or denotes the initial energy of the electric vehicle i;E i,ex denotes the off-grid desired energy of the electric vehicle i; and Δt denotes a unit time duration. The first category of electric vehicles and the second category of electric vehicles can be combined into one category, and the electric vehicles are uniformly modeled as follows: In the formula, and respectively represent the day-ahead regulation power and electricity of the electric vehicle; represents the day-ahead ordered charging reservation information of the electric vehicle, which is a 0-1 integer variable, 1 represents the first type of electric vehicle, and 0 represents the second type of electric vehicle; represents the actual charging power of the electric vehicle i as the first type of electric vehicle at the t period; and Δt represents a unit time length.

5. The column and constraint generation algorithm based optimization method for electric vehicle peak shaving service energy storage according to claim 2, wherein, The step S13 comprises the following steps: The intra-day regulation constraint of the electric vehicle aggregator is divided into two parts, namely the intra-day regulation constraint of the electric vehicle and the intra-day regulation constraint of the energy storage; The intra-day regulation constraint of the electric vehicle is basically the same as the day-ahead regulation constraint, except for the constraint changes caused by the introduction of uncertain variables, which are as follows: wherein, denotes the actual regulated charging power of the second type of electric vehicle i at time period t during the day; P i rated is the rated charging power of the electric vehicle i; E i,or denotes the initial state of charge of the electric vehicle i; E i,ex denotes the off-grid desired state of charge of the electric vehicle i; is the state of charge of the second type of electric vehicle i at time period t during the day; E and are the daily regulated power and state of charge of the electric vehicle, respectively; is the daily ordered charging contract uncertain variable of the electric vehicle, with value 1 indicating the first type of electric vehicle and 0 representing the second type of electric vehicle; is the actual charging power of the electric vehicle i at time period t during the day when it is of the first type; Δt denotes the unit time duration; the daily regulation constraint of the energy storage is as follows: In the formula, P ESS,max represents the rated power of the energy storage configuration; E ESS,max represents the rated capacity of the energy storage configuration; P t ch , P t dis respectively represent the energy storage charging power and discharging power; η ch , η dis respectively represent the energy storage charging efficiency and discharging efficiency; is the real-time power of the energy storage in a day at the t period; and Δt represents a unit time length.

6. The column and constraint generation algorithm based optimization method for electric vehicle peak shaving service energy storage according to claim 2, wherein, The step S14 comprises the following steps: The winning power of the electric vehicle aggregator in the day-ahead satisfies the following constraint: In the formula, the first row indicates that the day-ahead regulated power of the electric vehicle aggregator is composed of the sum of the actual power of all internal electric vehicles and the electric vehicle aggregator with energy storage; the second row indicates that the bidding power is not less than the baseline load power in the peak shaving period; P t bid represents the bidding power of the electric vehicle aggregator in the day-ahead peak shaving auxiliary service market at the t period; N EV represents the total number of electric vehicles inside the electric vehicle aggregator; is the day-ahead regulated power of the electric vehicle; P t base represents the baseline load of the peak shaving power calculated by the North China source network energy storage platform according to the test results at the t period; T peak represents the peak shaving period.

7. The column and constraint generation algorithm based method for optimization of electric vehicle peak shaving service energy storage according to claim 2, wherein, The step S15 comprises: The actual regulation power of the electric vehicle aggregator satisfies the following constraint: In the formula, the first row indicates that the actual power of the electric vehicle aggregator is composed of the sum of the actual power of all internal electric vehicles and energy storage; the second row indicates that the actual power in the peak shaving period is not less than the baseline load power and is not greater than the day-ahead winning power; represents the actual regulation power of the electric vehicle aggregator within a day; N EV represents the total number of electric vehicles inside the electric vehicle aggregator; is the day-ahead regulation power of the electric vehicle; P t ch , P t dis respectively represent the charging power and discharging power of the energy storage; P t base represents the baseline load of the North China source network load storage platform in the t period according to the test results approved by calculation; P t bid represents the day-ahead winning power of the electric vehicle aggregator in the peak shaving auxiliary service market in the t period; T peak represents the peak shaving period.

8. The column and constraint generation algorithm based optimization method for electric vehicle peak shaving service energy storage according to claim 1, wherein, The model is converted into a two-stage robust optimization model, comprising the following steps: Step S31: constructing an uncertainty set of the two-stage robust optimization model; The robust optimization method is used to improve the model, and a related box-type uncertainty set is constructed: wherein, denotes the electricity purchase price of the aggregator in the electricity market at time period t; denotes the clearing price of the regulation market; denotes the predicted electricity market price, the regulation market clearing price; is a binary variable indicating whether the price deviates at time period t, taking value 1 if it deviates; energy , Γ peak denotes the maximum number of time periods in which the price deviates; β energy , β peak denotes the degree of price deviation; is a binary variable indicating whether the aggregator complies with the contract, taking value 1 if it complies; N EV denotes the number of electric vehicles that the aggregator can regulate; β EV denotes the maximum proportion of electric vehicles that do not comply with the contract; Step S32: linearizing part of the constraints; The rated power formula P of the configuration energy storage is ESS,max Linearize the formula as follows, Introducing an intermediate variable The rated capacity E of the configuration energy storage is ESS,max Linearized as the following formula, Step S33: reconstructing the model into a two-stage robust optimization model; The formula for minimizing the total daily cost reconstruction into a two-stage robust optimization model is transformed into a two-stage robust optimization; the two-stage robust optimization model is In the formula, y is a continuous variable in the first-stage optimization problem; x is a continuous variable in the second-stage optimization problem; and u is an uncertain scenario variable; The specific expression is as follows:

9. The column and constraint generation algorithm based optimization method for electric vehicle peak shaving service energy storage according to claim 1, wherein, The two-stage robust optimization model is solved based on a column and constraint generation algorithm, comprising: The two-stage robust optimization model is decomposed into a main problem and a sub-problem, and the influence of the uncertain variable is not considered in the main problem, which is solved as a deterministic optimization; On the basis of the deterministic optimization solution, relatively adverse scenarios and corresponding sub-problem decision variables and constraint conditions are added, so that the upper and lower bounds of the objective function are continuously improved until the algorithm converges.

10. The column and constraint generation algorithm based method for optimization of electric vehicle peak shaving service energy storage according to claim 1, wherein, The result analysis and sensitivity analysis are performed to obtain the peak shaving and storage scheme of the aggregator, including: inputting the peak shaving market, the electricity price predicted by the electricity market, the charging data of the electric vehicles in the aggregator, etc., solving by using a column and constraint generation algorithm, performing sensitivity analysis, and obtaining the optimal scheme of the energy storage configuration of the aggregator in the peak shaving auxiliary service market by setting appropriate robust parameters.

Images