A multi-objective bi-level optimization method for electric vehicle load considering peak load shifting

Abstract

This invention belongs to the field of electric vehicle load optimization technology, specifically disclosing a multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling, including the following steps: establishing a multi-objective bi-level optimization model for electric vehicle load; through... KKT Condition and maximum M The method transforms the bi-level optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective linear problem; further, it transforms the single-level multi-objective linear problem into a single-level single-objective problem; and finally, it solves the single-level single-objective problem to obtain... Parteo Frontier; at Parteo The ideal solution is determined at the forefront and serves as the final result of the original two-level multi-objective optimization problem, achieving multi-objective two-level optimization of electric vehicle load. This invention solves the problem of existing two-level optimization operation models that lack consideration of comprehensive indicators such as peak-valley difference and load fluctuation on the distribution network side, resulting in the upper-level model in the two-level model being a single-objective optimization problem. This invention can reduce peak-valley difference and load fluctuation while ensuring the benefits of electric vehicle users.

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle load optimization technology, specifically relating to the design of a multi-objective bi-level optimization method for electric vehicle load that considers peak shaving and valley filling. Background Technology

[0002] In recent years, the power system has faced enormous pressure to balance supply and demand due to factors such as high temperatures. Meanwhile, the distribution network user side possesses a vast amount of adjustable flexible load resources, including electric vehicles, relying on demand-side response (DSC). Demand response , DR Methods such as peak shaving and valley filling, as well as the absorption of new energy sources, can promote the distribution network. Therefore, tapping the adjustable potential on the user side and promoting the participation of electric vehicle loads in source-load interaction is an important development direction for solving the supply and demand balance of new distribution systems.

[0003] Electric vehicle (EV) charging load is increasing year by year with the use of EVs, accounting for up to 12% of the power grid. Optimizing EV charging time can effectively assist in peak shaving and valley filling of the power grid. By managing the charging process of EV charging stations, the peak-valley difference of the power grid is optimized while meeting user charging needs. In the interaction between EV load and the power grid, the needs on both the supply and demand sides usually need to be considered. The first approach is to consider economic benefits, constructing a multi-objective model of EVs that incorporates low-carbon and economic considerations, and using intelligent algorithms to improve the economic efficiency of power stations. The second approach is to consider the safety and stability of the power grid. To reduce the impact of EV charging on the power grid, a multi-objective function is established, with the goal of minimizing the peak-valley difference of the total power grid load curve on the grid side and minimizing the total charging cost on the load side. For the above multi-objective optimization problems, genetic algorithms and particle swarm optimization algorithms are often used to transform the multi-objectives into objective functions. For objectives with different dimensions, normalization and weighted sums are usually used. However, the weights between the objectives are handled rather vaguely, resulting in the optimization results being influenced by the weights and having a certain degree of randomness. In considering multi-objective problems where the interests of supply and demand conflict, there are also ways to achieve this. Pareto Non-dominated solutions yield optimized results.

[0004] In summary, while existing research on demand-side response for flexible loads in distribution systems has constructed a two-layer optimized operation model that takes into account the interests of both the power grid and users, it lacks consideration of comprehensive indicators such as peak-to-valley difference and load fluctuation on the distribution network side. As a result, the upper layer model in the two-layer model is mostly a single-objective optimization problem. Summary of the Invention

[0005] The purpose of this invention is to address the problem that existing two-layer optimization operation models for flexible loads in power distribution systems lack consideration of comprehensive indicators such as peak-valley difference and load fluctuation on the distribution network side, resulting in the upper-layer model in the two-layer model being mostly a single-objective optimization. This invention proposes a multi-objective two-layer optimization method for electric vehicle loads that considers peak shaving and valley filling.

[0006] The technical solution of this invention is: a multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling, comprising the following steps:

[0007] S1. Establish an electric vehicle load model, and based on the electric vehicle load model, establish a multi-objective bi-level optimization model for electric vehicle load;

[0008] S2. Through KKT The condition is used to transform the bi-level multi-objective optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem;

[0009] S3. Utilize the maximum M The method transforms a single-layer multi-objective nonlinear optimization problem into a single-layer multi-objective linear problem;

[0010] S4. Establish an equivalent multi-objective optimization model, standardize each objective function, and transform the single-layer multi-objective linear problem into a single-layer single-objective problem containing the weight coefficients of each objective.

[0011] S5. Solve the single-layer, single-objective problem to obtain... Parteo cutting edge;

[0012] S6. According to Parteo The ideal solution is determined at the forefront and used as the final result of the two-level multi-objective optimization problem to achieve multi-objective two-level optimization of electric vehicle load.

[0013] The beneficial effects of this invention are:

[0014] This invention considers the benefits on both the power grid and the supply and demand sides of electric vehicle users. In the power grid side objectives, in addition to the revenue of distribution network operators, two new indicators, load fluctuation and load peak-valley difference, are introduced. By establishing three indicators, namely power grid side revenue, load fluctuation, and load peak-valley difference, the optimal operating solution is obtained using the Pareto frontier solution. This can reduce peak-valley difference and load fluctuation while ensuring the revenue of electric vehicle users.

[0015] Preferably, the electric vehicle load model in step S1 is constructed using the Monte Carlo method, specifically including the following formulas:

[0016]

[0017] in, Let represent the probability density function at the start of charging of an electric vehicle. express Expectations express standard deviation This indicates the start time of charging for the electric vehicle. This represents the probability density function of the daily mileage of an electric vehicle. for Expectations; express standard deviation Indicates the daily mileage of an electric vehicle. Indicates the initial state of charge. This indicates the state of charge at the end of the electric vehicle's journey. Indicates the distance traveled by the electric vehicle. Indicates the maximum driving distance of an electric vehicle. This indicates the time required to fully charge an electric vehicle. This indicates the state of charge of the electric vehicle as it begins charging. Indicates the capacity of electric vehicles. This indicates the charging efficiency of electric vehicles. This indicates the charging power of an electric vehicle. express time The combined power of Taiwan's electric vehicles; express Time of the first The charging power of an electric vehicle.

[0018] The beneficial effects of the above preferred solution are:

[0019] By using the above-mentioned optimized scheme, the load of the electric vehicle fleet is modeled, and the response capability of the load aggregator is improved by utilizing the electricity consumption behavior characteristics of electric vehicle loads.

[0020] Preferably, the multi-objective bi-level optimization model for electric vehicle load in step S1 includes an upper-level model and a lower-level model; the upper-level model is a multi-objective optimization problem; and the lower-level model is a single-objective optimization problem.

[0021] Preferably, the objective function of the upper-level model is:

[0022]

[0023] in, This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator. This indicates the revenue from selling electricity to users. This represents the revenue from purchasing and selling electricity from the external power grid. This represents expenditure on purchasing load demand response resources for electric vehicles. This indicates the operation and maintenance costs of wind power and solar power. Indicates the electrical load after demand response. Indicates the total time period. and Indicates a unit of time period;

[0024] The constraints of the upper-level model include power balance constraints and power purchase and sale constraints for distribution network operators; the formula for the power balance constraint is:

[0025] ;

[0026] The formula for the power purchase and sale constraints of the distribution network operator is:

[0027]

[0028] in, This indicates the predicted output power of wind power. This indicates the predicted output power of photovoltaic power. This indicates the amount of electricity purchased from the external power grid. This indicates the electric vehicle load after demand response. This indicates the amount of electricity sold to the external power grid. Indicates the base load. Indicates the maximum amount of electricity that can be purchased. Indicates the maximum electricity sales volume. Representing variables of 0 and 1, .

[0029] Preferably, the objective function of the lower-level model is:

[0030]

[0031] in, This represents the profit of the lower-level load aggregator. This represents expenditure on purchasing load demand response resources for electric vehicles. express LA Demand response volume purchased from users, This indicates the number of electric vehicles that have been moved out of service. This indicates the number of electric vehicles entering the market. This indicates the transfer electricity price for electric vehicles;

[0032] The constraints of the lower-level model are:

[0033]

[0034] in, For electric vehicle load in t The upper limit of the transfer amount at any given time. This represents the state variable indicating the departure of the electric vehicle. This indicates that the electric vehicle has transitioned to a state variable. , , , , and They represent t Lagrange inequality multiplier constraints at any time Indicates the total time period.

[0035] Preferably, step S2 includes the following formula:

[0036] Based on the lower-level model, the Lagrangian function is constructed, resulting in:

[0037]

[0038] in, Represent the Lagrange function, and Denote the set of Lagrange inequality multipliers constraint. express DSO Towards LA The electricity price for demand response to electric vehicle loads purchased by the unit;

[0039] For the Lagrange function, using KKT The conditions transform the objective function and constraints of the lower-level model into constraints of the upper-level model, resulting in complementary relaxation conditions. This completes the transformation of the bi-level optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem.

[0040] The formula for the complementary relaxation condition is:

[0041]

[0042] in, Representing the Lagrange function The transfer volume of electric vehicles The partial derivatives of .

[0043] Preferably, step S3 specifically involves: utilizing the maximum M The method transforms the complementary relaxation conditions into linear constraints, resulting in a single-layer multi-objective linear problem;

[0044] The formula for the linear constraint is:

[0045]

[0046]

[0047]

[0048]

[0049]

[0050] in, , , , and Both represent variables of 0 and 1. It represents a very large number.

[0051] Preferably, the formula for the multi-objective optimization model in step S4 is:

[0052]

[0053] in, This represents a multi-objective optimization model. , and Let each represent the objective function of the multi-objective optimization model. This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator.

[0054] Preferably, the formula for standardizing the objective function in step S4 is:

[0055]

[0056] in, Let represent the standardized objective function. This represents the objective function of a multi-objective optimization model. Describe the objective function The minimum value, Describe the objective function The maximum value.

[0057] Preferably, the formula for the single-layer single-objective problem in step S4 is:

[0058]

[0059] in, This represents a multi-objective optimization model; Let represent the standardized objective function. ; , and This represents the weight coefficient of each objective function in a single-layer, single-objective problem. Attached Figure Description

[0060] Figure 1 The diagram shows a flowchart of a multi-objective bi-level optimization method for electric vehicle load that considers peak shaving and valley filling. Detailed Implementation

[0061] Exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary and are intended to illustrate the principles and spirit of the invention, and are not intended to limit the scope of the invention.

[0062] like Figure 1 As shown, a multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling includes the following steps:

[0063] S1. Establish an electric vehicle load model, and based on the electric vehicle load model, establish a multi-objective bi-level optimization model for electric vehicle load;

[0064] S2. Through KKT The condition is used to transform the bi-level multi-objective optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem;

[0065] S3. Utilize the maximum M The method transforms a single-layer multi-objective nonlinear optimization problem into a single-layer multi-objective linear problem;

[0066] S4. Establish an equivalent multi-objective optimization model, standardize each objective function, and transform the single-layer multi-objective linear problem into a single-layer single-objective problem containing the weight coefficients of each objective.

[0067] S5. Solve the single-layer, single-objective problem to obtain... Parteo cutting edge;

[0068] S6. According to Parteo The ideal solution is determined at the forefront and used as the final result of the two-level multi-objective optimization problem to achieve multi-objective two-level optimization of electric vehicle load.

[0069] In this embodiment, the electric vehicle load model in step S1 is constructed using the Monte Carlo method, specifically including the following formulas:

[0070]

[0071] in, Let represent the probability density function at the start of charging of an electric vehicle. express Expectations express standard deviation This indicates the start time of charging for the electric vehicle. This represents the probability density function of the daily mileage of an electric vehicle. for Expectations; express standard deviation Indicates the daily mileage of an electric vehicle. Representing the initial state of charge, we have =1, This indicates the state of charge at the end of the electric vehicle's journey. Indicates the distance traveled by the electric vehicle. Indicates the maximum driving distance of an electric vehicle. This indicates the time required to fully charge an electric vehicle. This indicates the state of charge of the electric vehicle as it begins charging. Indicates the capacity of electric vehicles. This indicates the charging efficiency of electric vehicles. This indicates the charging power of an electric vehicle. express time The combined power of Taiwan's electric vehicles; express Time of the first The charging power of an electric vehicle.

[0072] In this embodiment, the two-layer optimization model of the electric vehicle (EV) load distribution network includes the distribution network operator, load aggregator, and EV load users. The load aggregator signs agreements with the users who can participate in the optimization and controls the EV charging load by connecting to EV charging piles. The multi-objective two-layer optimization model of EV load described in step S1 includes an upper-layer model and a lower-layer model.

[0073] The upper-level model is the distribution network operator ( Distribution system operator , DSO The requirement is a multi-objective optimization problem. The upper-level model mainly considers... DSO The company's own profit from electricity purchase and sale, as well as the peak-to-valley difference and load fluctuation of the power grid; in the upper-level model DSO The power output curves for the wind and solar power sources under its jurisdiction can be obtained based on day-ahead forecasts; DSO As the leader of the upper-level model, it purchases the response volume of electric vehicle load from the load aggregator, with the goal of... DSO It maximizes returns, minimizes peak-to-valley differences, and minimizes load fluctuations.

[0074] The lower-level model is the electric vehicle load aggregator ( Load aggregator ,LA This is a single-objective optimization problem. The lower-level model... LA Increase the enthusiasm of electric vehicle load users to participate in demand response through measures such as electricity pricing and incentives. LA As a follower of the two-level optimization model, the objective function is to maximize the profit obtained by optimizing the electricity consumption behavior of flexible loads. In the established two-level optimization model, the revenue of both the upper and lower levels is related to the amount of demand response resources, that is, the objective values ​​of the upper and lower level models influence each other.

[0075] In this embodiment, the objective function of the upper-level model is:

[0076]

[0077] in, This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator. This indicates the revenue from selling electricity to users. This represents the revenue from purchasing and selling electricity from the external power grid. This represents expenditure on purchasing load demand response resources for electric vehicles. This indicates the operation and maintenance costs of wind power and solar power. Indicates the electrical load after demand response. Indicates the total time period, with =24h, and Indicates a unit of time period;

[0078] Revenue from selling electricity to users includes:

[0079]

[0080] in, This represents the electrical load following demand response, including electric vehicle load and base load. This indicates the electricity price charged by the power distribution network operator to users.

[0081] Representing the revenue from purchasing and selling electricity from the external power grid, we have:

[0082]

[0083] in, The amount of electricity purchased from the external power grid; The amount of electricity sold to the external power grid; The price for electricity purchased from external power grids; The price of electricity sold to the external power grid;

[0084] Expenditure on purchasing load demand response resources for electric vehicles includes:

[0085]

[0086] in, Demand response to electric vehicle load; for DSO Towards LA The electricity price for demand response to electric vehicle loads purchased by the unit;

[0087] The operation and maintenance costs of wind power and solar power are represented by the following:

[0088]

[0089] in, This indicates the predicted output power of wind power. This indicates the previously predicted output power of the photovoltaic system. This represents the unit operation and maintenance cost of wind power. This represents the unit operation and maintenance cost of a photovoltaic system;

[0090] The constraints of the upper-level model include power balance constraints and power purchase and sale constraints for distribution network operators. For the power balance constraint, the sum of the output power of wind power and photovoltaic power and the power purchased from the grid by the distribution network operator should equal the sum of all loads. Therefore, the formula for the power balance constraint is:

[0091] ;

[0092] The formula for the power purchase and sale constraints of the distribution network operator is:

[0093]

[0094] in, This indicates the predicted output power of wind power. This indicates the predicted output power of photovoltaic power. This indicates the amount of electricity purchased from the external power grid. This indicates the electric vehicle load after demand response. This indicates the amount of electricity sold to the external power grid. Indicates the base load. Indicates the maximum amount of electricity that can be purchased. Indicates the maximum electricity sales volume. Representing variables of 0 and 1, .

[0095] In this embodiment, the objective function of the lower-level model is:

[0096]

[0097] in, This represents the profit of the lower-level load aggregator. This represents expenditure on purchasing load demand response resources for electric vehicles. express LA Demand response volume purchased from users, This indicates the number of electric vehicles that have been moved out of service. This indicates the number of electric vehicles entering the market. This indicates the transfer electricity price for electric vehicles;

[0098] The constraints of the lower-level model are:

[0099]

[0100] in, For electric vehicle load in t The upper limit of the transfer amount at any given time. This represents the state variable indicating the departure of the electric vehicle. The variables representing the state transition of an electric vehicle are all 0 or 1. =1 indicates the load transfer amount of electric vehicles. =0 indicates that the electric vehicle load is not transferred; =1 indicates that the load of the electric vehicle has been transferred out. =0 indicates that the electric vehicle load is not transferred out. , , , , and They represent t Lagrange inequality multiplier constraints at any time Indicates the total time period.

[0101] In this embodiment, step S2 includes the following formula:

[0102] Based on the lower-level model, the Lagrangian function is constructed, resulting in:

[0103]

[0104] in, Represent the Lagrange function, and Denote the set of Lagrange inequality multipliers constraint. express DSO Towards LA The electricity price for demand response to electric vehicle loads purchased by the unit;

[0105] For the Lagrange function, usingKKT The conditions transform the objective function and constraints of the lower-level model into constraints of the upper-level model, resulting in complementary relaxation conditions. This completes the transformation of the bi-level optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem.

[0106] The formula for the complementary relaxation condition is:

[0107]

[0108] in, Representing the Lagrange function The transfer volume of electric vehicles The partial derivatives of .

[0109] In this embodiment, step S3 specifically involves: utilizing the maximum M The method transforms the complementary relaxation conditions into linear constraints, resulting in a single-layer multi-objective linear problem;

[0110] The formula for the linear constraint is:

[0111]

[0112]

[0113]

[0114]

[0115]

[0116] in, , , , and Both represent variables of 0 and 1. To represent a very large number, in this embodiment, we take... =1E8.

[0117] In this embodiment, the formula for the multi-objective optimization model in step S4 is:

[0118]

[0119] in, This represents a multi-objective optimization model. , and Let each represent the objective function of the multi-objective optimization model. This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator.

[0120] In this embodiment, the formula for standardizing the objective function in step S4 is:

[0121]

[0122] in, Let represent the standardized objective function. This represents the objective function of a multi-objective optimization model. Describe the objective function The minimum value, Describe the objective function The maximum value.

[0123] In this embodiment, the formula for the single-layer single-objective problem in step S4 is:

[0124]

[0125] in, This represents a multi-objective optimization model; Let represent the standardized objective function. ; , and This represents the weight coefficient of each objective function in a single-layer, single-objective problem.

[0126] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling, characterized in that, Includes the following steps: S1. Establish an electric vehicle load model, and based on the electric vehicle load model, establish a multi-objective bi-level optimization model for electric vehicle load; S2. Through KKT The condition is used to transform the bi-level multi-objective optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem; S3. Utilize the maximum M The method transforms a single-layer multi-objective nonlinear optimization problem into a single-layer multi-objective linear problem; S4. Establish an equivalent multi-objective optimization model, standardize each objective function, and transform the single-layer multi-objective linear problem into a single-layer single-objective problem containing the weight coefficients of each objective. S5. Solve the single-layer, single-objective problem to obtain... Parteo cutting edge; S6. According to Parteo The ideal solution is determined at the forefront and used as the final result of the bi-level multi-objective optimization problem to achieve multi-objective bi-level optimization of electric vehicle load. The electric vehicle load model described in step S1 is constructed using the Monte Carlo method, specifically including the following formulas: in, Let represent the probability density function at the start of charging of an electric vehicle. express Expectations express standard deviation This indicates the start time of charging for the electric vehicle. This represents the probability density function of the daily mileage of an electric vehicle. for Expectations; express standard deviation Indicates the daily mileage of electric vehicles. Indicates the initial state of charge. This indicates the state of charge at the end of the electric vehicle's journey. Indicates the distance traveled by the electric vehicle. Indicates the maximum driving distance of an electric vehicle. This indicates the time required to fully charge an electric vehicle. This indicates the state of charge of the electric vehicle as it begins charging. Indicates the capacity of electric vehicles. This indicates the charging efficiency of electric vehicles. This indicates the charging power of an electric vehicle. express time The combined power of Taiwan's electric vehicles; express Time of the first The charging power of an electric vehicle; The multi-objective bi-level optimization model for electric vehicle load described in step S1 includes an upper-level model and a lower-level model; the upper-level model is a multi-objective optimization problem; and the lower-level model is a single-objective optimization problem. The objective function of the upper-level model is: in, This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator. This indicates the revenue from selling electricity to users. This represents the revenue from purchasing and selling electricity from the external power grid. This represents expenditure on purchasing load demand response resources for electric vehicles. This indicates the operation and maintenance costs of wind power and solar power. Indicates the electrical load after demand response. Indicates the total time period. and Indicates a unit of time period; The constraints of the upper-level model include power balance constraints and power purchase and sale constraints for distribution network operators; the formula for the power balance constraint is: ； The formula for the power purchase and sale constraints of the distribution network operator is: in, This indicates the predicted output power of wind power. This indicates the predicted output power of photovoltaic power. This indicates the amount of electricity purchased from the external power grid. This indicates the electric vehicle load after demand response. This indicates the amount of electricity sold to the external power grid. Indicates the base load. Indicates the maximum amount of electricity that can be purchased. Indicates the maximum electricity sales volume. Representing variables of 0 and 1, ; The objective function of the lower-level model is: in, This represents the profit of the lower-level load aggregator. This represents expenditure on purchasing load demand response resources for electric vehicles. express LA Demand response volume purchased from users, This indicates the number of electric vehicles that have been moved out of service. This indicates the number of electric vehicles entering the market. This indicates the transfer electricity price for electric vehicles; The constraints of the lower-level model are: in, For electric vehicle load in t The upper limit of the transfer amount at any given time. This represents the state variable indicating the departure of the electric vehicle. This indicates that the electric vehicle has transitioned to a state variable. , , , , and They represent t Lagrange inequality multiplier constraints at any time Indicates the total time period.

2. The multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling as described in claim 1, characterized in that, Step S2 includes the following formula: Based on the lower-level model, the Lagrangian function is constructed, resulting in: in, Represent the Lagrange function, and Denote the set of Lagrange inequality multipliers constraint. express DSO Towards LA The electricity price for demand response to electric vehicle loads purchased by the unit; For the Lagrange function, using KKT The conditions transform the objective function and constraints of the lower-level model into constraints of the upper-level model, resulting in complementary relaxation conditions. This completes the transformation of the bi-level optimization problem of the multi-objective bi-level optimization model into a single-level multi-objective nonlinear optimization problem. The formula for the complementary relaxation condition is: in, Representing the Lagrange function The transfer volume of electric vehicles The partial derivatives of .

3. The multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling as described in claim 2, characterized in that, Specifically, step S3 involves: utilizing the maximum... M The method transforms the complementary relaxation conditions into linear constraints, resulting in a single-layer multi-objective linear problem; The formula for the linear constraint is: in, , , , and Both represent variables of 0 and 1. It represents a very large number.

4. The multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling as described in claim 1, characterized in that, The formula for the multi-objective optimization model in step S4 is: in, This represents a multi-objective optimization model. , and Let each represent the objective function of the multi-objective optimization model. This represents the profit of the power distribution network operator. This indicates the peak-to-valley load difference for distribution network operators. This indicates the load fluctuation of the distribution network operator.

5. The multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling according to claim 1, characterized in that, The formula for standardizing the objective function in step S4 is: in, This represents the standardized objective function. This represents the objective function of a multi-objective optimization model. Describe the objective function The minimum value, Describe the objective function The maximum value.

6. The multi-objective bi-level optimization method for electric vehicle load considering peak shaving and valley filling as described in claim 1, characterized in that, The formula for the single-layer, single-objective problem described in step S4 is: in, This represents a multi-objective optimization model; This represents the standardized objective function. ; , and This represents the weight coefficient of each objective function in a single-layer, single-objective problem.

Images