Multi-objective scheduling method for virtual power plant considering uncertainty and user priority

Abstract

The application discloses a kind of considering uncertainty and user priority's virtual power plant multi-objective scheduling method, belong to power system operation and control technical field.For the problem of significant source and load uncertainty, multi-agent interest conflict and user demand neglect in virtual power plant scheduling, the application constructs equipment operation mode selection model based on dynamic priority, uses point estimation method to process random variable, and solves multi-objective problem by combining utopia pursuit method.The specific steps include data acquisition and uncertainty modeling, priority-based operation mode division, random multi-objective optimization model construction, multi-objective normalization processing and strategy output.The method guarantees user rigid demand through priority mechanism, efficiently handles source and load uncertainty, balances the interests of operators, user cost, battery life and other parties, improves new energy consumption and grid stability, and has engineering practicability and robustness.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, and in particular to a virtual power plant multi-objective scheduling method that takes into account uncertainty and user priority. Background Technology

[0002] Virtual power plants (VPPs), as a management model that can efficiently aggregate distributed photovoltaics, energy storage, electric vehicles (EVs), and flexible loads, are gradually becoming a research hotspot. However, current VPP dispatching technologies still face many serious challenges: First, the conflicting interests among the multiple stakeholders are difficult to reconcile. VPP operators pursue profit maximization, while users pursue energy cost minimization and comfort maximization. These two goals are inherently contradictory, and traditional single-goal optimization methods are difficult to balance the demands of all parties.

[0003] Secondly, there are significant uncertainties on both the power source and load sides. The intermittency of photovoltaic power output, coupled with the randomness of electric vehicle user arrival / departure times and initial state of charge, poses a great risk to day-ahead dispatching plans.

[0004] Furthermore, existing scheduling strategies often overlook the individual differences in the needs of electric vehicle owners regarding parking duration and target battery level, which can easily lead to decreased user satisfaction and even fail to meet emergency travel needs.

[0005] Finally, the Monte Carlo simulation method, which is widely used to handle uncertainty, suffers from excessive computational complexity, making it difficult to meet the timeliness requirements of scheduling decisions in practical engineering applications.

[0006] Therefore, there is an urgent need to propose a new virtual power plant scheduling method that can effectively handle source-load uncertainty, coordinate the interests of multiple stakeholders, and fully consider the differentiated needs and priorities of users. Summary of the Invention

[0007] The purpose of this invention is to provide a multi-objective scheduling method for virtual power plants that considers uncertainty and user priorities. This invention achieves efficient and robust operation of virtual power plants by constructing a priority-based operation mode selection mechanism, using point estimation to process random variables, and combining the Utopian pursuit method to solve multi-objective problems.

[0008] To achieve the above objectives, this invention provides a multi-objective scheduling method for virtual power plants that considers uncertainty and user priorities, comprising the following steps: S1. Data Acquisition and Uncertainty Modeling Foundation: Acquire historical operating data, real-time status, and technical parameters of the virtual power plant (VPP) and its aggregated distributed energy resources; the distributed energy resources include photovoltaic power plants (PV), electric vehicle charging stations (CS), battery swapping stations (BSS), and user-side controllable loads (CL); at the same time, acquire the statistical characteristics of input variables with uncertainties, the uncertain input variables including at least the arrival / departure time of electric vehicle users, initial state of charge, and base load demand. S2. Construct a device operation mode selection model based on dynamic priority; for electric vehicles (EVs), calculate the real-time priority factor based on their current state of charge, remaining parking time, and expected target battery level. According to this priority factor The real-time incentive electricity price is used as a guiding signal to dynamically determine whether each electric vehicle is in a forced charging state or a flexible scheduling state, and accordingly determine its operation mode as disordered charging, orderly charging G2V or orderly discharging V2G; for the battery of the battery swapping station, combined with the current photovoltaic irradiance prediction and the peak and valley time information of the power grid, it is determined whether it works in the charging storage mode G2B or the discharging support mode B2G. S3. Construct a stochastic multi-objective optimization scheduling model; use the point estimation method to process the uncertain input variables in step S1, and generate several weighted deterministic scenarios to simulate randomness; on this basis, establish a multi-objective optimization system aimed at minimizing the expected value of the objective function under each scenario; the multi-objective optimization system includes at least: maximizing the net profit of VPP operators, minimizing the investment and operation and maintenance costs of photovoltaic power plants, minimizing the total energy payment costs of demand response participants, minimizing the combined costs of battery aging and user dissatisfaction, and minimizing the cost of virtual power plants purchasing electricity from the main grid; S4. Normalization of the multi-objective model: The multi-objective optimization model constructed in step S3 is processed by the Utopian pursuit method. The multi-objective optimization problem is transformed into a single-objective optimization problem that seeks the compromise solution closest to the ideal point by calculating the ideal point and the ideal point of each sub-objective. S5. Model Solving and Strategy Output: Under the premise of satisfying system operation constraints, the transformed single-objective model is solved, and the day-ahead optimized scheduling plan is output. The plan includes the incentive electricity price strategy for each time period, the photovoltaic power output plan, the charging and discharging power instructions for each EV, the power exchange plan of the battery swapping station, and the reduction or shift of controllable load.

[0009] Preferably, in step S2, the formula for calculating the priority factor of the electric vehicle is: ; in, for Time of the first The node Vehicle priority factor, for Time of the first The node The vehicle's state of charge. For the target state of charge, The expected departure time for the electric vehicle. The shortest charging time required to reach the target state of charge.

[0010] Preferably, in step S2, the EV mode determination logic based on the priority factor is as follows: The priority determination threshold is set to 1; when the calculated priority factor... When the electric vehicle is in a forced charging state, it enters a disordered charging mode, charging at its maximum allowed power and not participating in the virtual power plant's optimized scheduling; when the calculated priority factor... When the electric vehicle is in a flexible scheduling state, it is determined to participate in the optimized scheduling. If the incentive price is lower than the preset price threshold, it is guided to enter the orderly charging G2V mode. If the incentive price is higher than the preset price threshold and its battery charge state meets the discharge conditions, it is guided to enter the orderly discharging V2G mode.

[0011] Preferably, in step S3, a point estimation method is used to handle uncertainty, specifically including: for the system's... Given a random input variable, three deterministic focal points and their corresponding probability weights are constructed for each random variable using the statistical moments of its probability density function. The focal points of different variables are combined to generate a set of deterministic scenarios for simulating uncertainty, and the comprehensive probability weight of each scenario is calculated. These scenarios are used to evaluate the deterministic model multiple times to optimize the expected value of the objective function.

[0012] Preferably, in step S3, the VPP operator's net profit objective function is expressed as: ; in, For VPP in Time Node Incentive pricing for selling or purchasing electricity from the distribution network; The net power injected into the distribution network is represented by positive electricity sales and negative electricity purchases. For the set of nodes within the virtual power plant, For the set of scheduling periods, This represents the length of the scheduling period.

[0013] Preferably, in step S3, the objective function for the combined cost of battery aging and user dissatisfaction includes the cost of battery aging and the cost of user dissatisfaction; the cost of battery aging is a nonlinear cycle life loss cost calculated based on the depth of discharge of electric vehicle batteries and battery swapping station batteries during the scheduling process; the cost of user dissatisfaction is a virtual cost calculated for controllable load users based on the amount of their load being shifted or reduced, and its formula is: ; in, For the first Cost of user dissatisfaction per controllable load node As a price factor, This is the load shift amount. The discomfort coefficient, This is the reference power.

[0014] Preferably, in step S3, the objective function for the investment and operation and maintenance costs of the photovoltaic power plant is expressed as: = in, The daily investment and operation and maintenance costs of photovoltaic power plants, The unit capacity investment cost For photovoltaic installed capacity, The discount rate is... Lifespan in years For unit operation and maintenance costs, This refers to the set of nodes in a photovoltaic power station.

[0015] Preferably, in step S4, the Utopian pursuit method specifically includes the following steps: calculating the optimal solution for each sub-objective under the single-objective optimization condition as the ideal point, and the worst solution as the nano-point; using the formula: ; The values ​​of each sub-objective function are normalized to dimensionless values ​​in the interval [0,1]. The final single-objective optimization function is constructed to minimize the weighted sum of the Euclidean distances between each normalized objective value and the ideal point, and its expression is: ; in, For the first The normalized function value of each sub-objective For the first Sub-goals in decision variables The function value below, For the first The ideal point for each sub-goal For the first The worst solution for each sub-objective These are the weighting coefficients for each sub-objective.

[0016] Preferably, in step S5, the constraints that the model solution needs to satisfy include: (1) Power balance constraints of each node within the virtual power plant: ; (2) State of charge (SoC) iteration constraints for electric vehicles and battery swapping stations: ; (3) Incentive price constraint: stipulate the internal incentive price set by the virtual power plant. Electricity prices must be in the external market. It fluctuates within a certain range, that is, it satisfies: ; in and These are the preset discount and markup coefficients.

[0017] Preferably, in step S2, the method for determining the operating mode of the battery BSS in the battery swapping station is as follows: setting a photovoltaic output threshold and peak and off-peak electricity price periods; when the photovoltaic output is higher than the threshold during the day, controlling the battery swapping station to be in charging and storage mode G2B to absorb excess photovoltaic power generation; at night or during peak grid electricity price periods, controlling the battery swapping station to be in discharging and support mode B2G to supply power to the microgrid or load; at the same time, the number of backup batteries in the battery swapping station must meet the minimum battery swapping demand constraint.

[0018] Therefore, the virtual power plant multi-objective scheduling method of the present invention, which considers uncertainty and user priority based on the above structure, has the following beneficial effects: (1) The dynamic priority mechanism of the present invention can accurately identify and prioritize the rigid needs of electric vehicle users who urgently need to charge, thus avoiding the problem of insufficient user power that may be caused by traditional scheduling.

[0019] (2) The present invention uses point estimation method to replace Monte Carlo simulation with huge computational load. While ensuring the accuracy of model prediction, it significantly reduces computational complexity and improves the engineering practicality of scheduling strategy.

[0020] (3) This invention uses the Utopian tracking method to handle multi-objective conflicts. It does not require subjective setting of weights and can automatically find the optimal solution that takes into account the operator's profit, user cost, equipment life and grid interaction.

[0021] (4) This invention guides flexible resources such as electric vehicles and battery swapping stations to participate in peak shaving and valley filling through reasonable incentive electricity prices and dispatch strategies, effectively absorbs surplus photovoltaic power, reduces the dependence of virtual power plants on the main power grid, and improves the economic efficiency and stability of the distribution network.

[0022] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the overall process of a virtual power plant multi-objective scheduling method that considers uncertainty and user priority according to the present invention. Figure 2 This is a schematic diagram of the direct control architecture of a virtual power plant, which is a multi-objective scheduling method for virtual power plants that takes into account uncertainty and user priority according to the present invention. Figure 3 This is a schematic diagram illustrating the multi-objective optimization principle based on the Utopian pursuit method of the present invention, which considers uncertainty and user priority in a multi-objective scheduling method for virtual power plants. Figure 4 Box plot of electric vehicle state of charge variation under different priorities in a virtual power plant multi-objective scheduling method that considers uncertainty and user priority according to the present invention; Figure 5 This diagram illustrates the reduction and shifting results of controllable load under day-ahead scheduling in a virtual power plant multi-objective scheduling method that considers uncertainty and user priority, according to the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0026] Example like Figure 1 This invention provides a multi-objective scheduling method for virtual power plants that considers uncertainty and user priorities, as detailed below: Step S1, Data Acquisition: To support the subsequent classification of equipment operation modes and the construction of multi-objective optimization models, it is necessary to acquire various basic data required for virtual power plant scheduling in advance. This includes the following steps: S11. Obtain the basic demand curves for photovoltaic module parameters, battery capacity and maximum charging and discharging power of the battery swapping station, and controllable load.

[0027] S12. Obtain external environment forecast data, including photovoltaic power output forecast and day-ahead market electricity price forecast.

[0028] S13. Obtain the behavioral statistics of electric vehicle (EV) users. Assuming that the arrival time, departure time and initial state of charge of the EV follow a truncated normal distribution, obtain their mean and standard deviation.

[0029] Step S2: Construct a device operation mode classification mechanism based on dynamic priority to evaluate the real-time response capability and classify the operation modes of connected electric vehicles (e.g., Figure 2 (As shown). The specific process is as follows: S21, in each scheduling period For the connection in the first The v-th electric vehicle at each charging station is based on its current state of charge. User's expected departure time and expected target power Calculate its real-time priority factor The formula for calculating this priority factor is: ; in, To achieve the charging time required for the target SoC, For battery capacity, This is the maximum charging power. The formula reflects the urgency between the user's remaining parking time and the required charging time (e.g., ...). Figure 4 (As shown).

[0030] S22. Determine the EV's operating mode based on priority factors. Set the priority determination threshold to 1. When the calculated priority factor is reached, the EV is determined to be in a forced charging state, entering an unordered charging mode, charging at its maximum allowed power and not participating in the virtual power plant's optimized scheduling. When the calculated priority factor is reached, the EV is determined to be in a flexible scheduling state, participating in optimized scheduling. At this time, guidance is provided based on the real-time incentive price: if the incentive price is lower than the preset price threshold, it is guided to enter the ordered charging (G2V) mode; if the incentive price is higher than the preset price threshold and its battery SoC meets the discharge conditions, it is guided to enter the ordered discharging (V2G) mode.

[0031] S23. At that time, the electric vehicle was determined to be in a flexible state and entered the coordinated optimization mode. Based on the real-time incentive electricity price signal for the current period, if the electricity price is low, it is classified as orderly charging (G2V) mode; if the electricity price is high and the battery SoC allows, it is classified as orderly discharging (V2G) mode. Simultaneously, based on photovoltaic output forecasts and electricity price information, the charging and discharging mode (G2B or B2G) of the battery at the battery swapping station is determined.

[0032] Step S3: Construct a stochastic multi-objective optimization model considering dual uncertainties, comprehensively evaluate the operator's economic efficiency, user experience, and grid interactivity, and generate a day-ahead dispatch optimization scheme that takes into account the interests of all parties. This specifically includes the following steps: S31. The (2m+1)-point estimation method (PEM) is used to handle the uncertain input variables in S1. For the m random input variables in the system, the first few moments of their probability density functions are used to construct three deterministic focal points and their corresponding weights for each random variable. The focal points of different variables are combined to generate several weighted deterministic scenarios, thereby approximating the distribution characteristics of the original random variables. The comprehensive probability weight of each scenario is calculated, and the deterministic model is evaluated multiple times using these scenarios to optimize the expected value of the objective function.

[0033] S32. Establish a multi-objective optimization function system containing five conflicting sub-objectives, such as... Figure 5 As shown.

[0034] (1) Objective function Maximize the net profit of VPP operators.

[0035] ; in, This is the incentive price for VPPs to sell electricity to the distribution network. This refers to the net power injected into the distribution network.

[0036] (2) Objective function Minimize the daily investment and operation and maintenance costs of photovoltaic power plants.

[0037] (3) Objective function Minimize the total energy payment costs of demand response participants (CS and CL).

[0038] (4) Objective function Minimize overall loss costs, which include battery aging costs and user dissatisfaction costs.

[0039] Battery aging cost is a nonlinear cycle life loss cost calculated based on the depth of discharge (DoD) of electric vehicle batteries and battery swapping station batteries during the dispatching process.

[0040] User dissatisfaction cost is a virtual cost calculated for users with controllable loads, based on the amount of load that is shifted or reduced. The formula is as follows: ; in, This is the load shift amount. The discomfort coefficient, It is a price factor.

[0041] (5) Objective function Minimize the cost for the virtual power plant to purchase electricity from the main grid, such as Figure 3 As shown.

[0042] S33. The Utopia-tracking method is used to solve the multi-objective optimization model. First, the Utopia point (optimal solution) and the Natopia point (worst solution) for each of the five sub-objectives under single-objective optimization are calculated. These two points are then used to normalize the functions of each sub-objective, constructing a single-objective optimization problem with the final objective of minimizing the sum of the Euclidean distances between each normalized objective value and the Utopia point. This leads to the optimal compromise scheduling scheme. The final normalized single-objective optimization function expression is: ; in, For the first Sub-goals in decision variables The function value below, The weighting coefficients for each objective (in this embodiment, we take...) ).

[0043] Furthermore, the mathematical expression for calculating the electric vehicle priority factor in step S21 is as follows: ; in, The shortest charging time required for the electric vehicle to reach the target battery level. This represents the total scheduling cycle. This formula reflects the urgency between the user's remaining parking time and the required charging time.

[0044] Furthermore, the specific expressions for each sub-objective of the multi-objective optimization function system constructed in step S32 are as follows: (1) Objective function F1: Maximize the net profit of VPP operators in, For nodes exist The net power constantly injected into the distribution network. This is the incentive electricity price within the VPP.

[0045] (2) Objective function F4: Minimize the cost of battery aging and user discomfort. This objective consists of two parts: .

[0046] Among them, battery aging cost The calculation primarily considers the cycling losses caused by V2G / B2G operations, based on the depth of discharge (DoD): ; In the formula, For battery replacement costs, This is the cycle lifetime function related to DoD.

[0047] User discomfort cost Primarily targeting the translation operation of controllable loads, an exponential function model is used: ; In the formula $ This is the load shift amount. The user discomfort coefficient.

[0048] Furthermore, the optimization model in step S3 also includes necessary operational constraints: (1) Power balance constraint: For any node and time period, the inflow power equals the outflow power.

[0049] (2) Energy storage SoC constraints: Preventing battery overcharging and over-discharging, for example for EVs: ; (3) Incentive Price Constraint: The internal incentive price set by the VPP should fluctuate within a certain percentage range of the market price to ensure economic rationality. ; in, These are the preset electricity price discount and surcharge coefficients.

[0050] 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A multi-objective scheduling method for virtual power plants considering uncertainty and user priorities, characterized in that, Includes the following steps: S1. Data Acquisition and Uncertainty Modeling Foundation: Acquire historical operating data, real-time status, and technical parameters of the virtual power plant (VPP) and its aggregated distributed energy resources; the distributed energy resources include photovoltaic power plants (PV), electric vehicle charging stations (CS), battery swapping stations (BSS), and user-side controllable loads (CL); at the same time, acquire the statistical characteristics of input variables with uncertainties, the uncertain input variables including at least the arrival / departure time of electric vehicle users, initial state of charge, and base load demand. S2. Construct a device operation mode selection model based on dynamic priority; for electric vehicles (EVs), calculate the real-time priority factor based on their current state of charge, remaining parking time, and expected target battery level. According to this priority factor The real-time incentive electricity price is used as a guiding signal to dynamically determine whether each electric vehicle is in a forced charging state or a flexible scheduling state, and accordingly determine its operation mode as disordered charging, orderly charging G2V or orderly discharging V2G; for the battery of the battery swapping station, combined with the current photovoltaic irradiance prediction and the peak and valley time information of the power grid, it is determined whether it operates in the charging storage mode G2B or the discharging support mode B2G. S3. Construct a stochastic multi-objective optimization scheduling model; The point estimation method is used to process the uncertain input variables in step S1, and several weighted deterministic scenarios are generated to simulate randomness. Based on this, a multi-objective optimization system is established with the aim of minimizing the expected value of the objective function under each scenario; The multi-objective optimization system includes at least: maximizing the net profit of VPP operators, minimizing the investment and operation and maintenance costs of photovoltaic power plants, minimizing the total energy payment costs of demand response participants, minimizing the combined costs of battery aging and user dissatisfaction, and minimizing the cost of virtual power plants purchasing electricity from the main grid; S4. Normalization of the multi-objective model: The multi-objective optimization model constructed in step S3 is processed by the Utopian pursuit method. The multi-objective optimization problem is transformed into a single-objective optimization problem that seeks the compromise solution closest to the ideal point by calculating the ideal point and the ideal point of each sub-objective. S5. Model Solving and Strategy Output: Under the premise of satisfying system operation constraints, the transformed single-objective model is solved, and the day-ahead optimized scheduling plan is output. The plan includes the incentive electricity price strategy for each time period, the photovoltaic power output plan, the charging and discharging power instructions for each EV, the power exchange plan of the battery swapping station, and the reduction or shift of controllable load.

2. The virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S2, the formula for calculating the priority factor of electric vehicles is: ； in, for Time of the first The node Vehicle priority factor, for Time of the first The node The vehicle's state of charge. For the target state of charge, The expected departure time for the electric vehicle. The shortest charging time required to reach the target state of charge.

3. The virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S2, the EV mode determination logic based on the priority factor is as follows: The priority determination threshold is set to 1; when the calculated priority factor... When the electric vehicle is in a forced charging state, it enters a disordered charging mode, charging at its maximum allowed power and not participating in the virtual power plant's optimized scheduling; when the calculated priority factor... When the electric vehicle is in a flexible scheduling state, it is determined to participate in the optimized scheduling. If the incentive price is lower than the preset price threshold, it is guided to enter the orderly charging G2V mode. If the incentive price is higher than the preset price threshold and its battery charge state meets the discharge conditions, it is guided to enter the orderly discharging V2G mode.

4. The virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S3, the point estimation method is used to handle uncertainty, specifically including: for the system's... Given a random input variable, three deterministic focal points and their corresponding probability weights are constructed for each random variable using the statistical moments of its probability density function. The focal points of different variables are combined to generate a set of deterministic scenarios for simulating uncertainty, and the comprehensive probability weight of each scenario is calculated. These scenarios are used to evaluate the deterministic model multiple times to optimize the expected value of the objective function.

5. A virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S3, the objective function for the VPP operator's net profit is expressed as: ； in, For VPP in Time Node Incentive pricing for selling or purchasing electricity from the distribution network. The net power injected into the distribution network is represented by positive electricity sales and negative electricity purchases. For the set of nodes within the virtual power plant, For the set of scheduling periods, This represents the length of the scheduling period.

6. The virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S3, the comprehensive cost objective function for battery aging and user dissatisfaction includes both battery aging cost and user dissatisfaction cost. The battery aging cost is a nonlinear cycle life loss cost calculated based on the depth of discharge of electric vehicle batteries and battery swapping station batteries during the scheduling process. The user dissatisfaction cost is a virtual cost calculated for controllable load users based on the amount of load shifted or reduced, and its formula is: ； in, For the first Cost of user dissatisfaction per controllable load node As a price factor, This is the load shift amount. The discomfort coefficient, This is the reference power.

7. A virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S3, the objective function for the investment and operation and maintenance costs of the photovoltaic power plant is expressed as: = in, The daily investment and operation and maintenance costs of photovoltaic power plants, The unit capacity investment cost For photovoltaic installed capacity, The discount rate is... Lifespan in years For unit operation and maintenance costs, A collection of photovoltaic power plant nodes. For nodes exist Photovoltaic power output at all times.

8. A virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S4, the Utopian pursuit method specifically includes the following steps: calculating the optimal solution for each sub-objective under the single-objective optimization condition as the ideal point, and the worst solution as the nano-point; using the formula: ； The values ​​of each sub-objective function are normalized to dimensionless values ​​in the interval [0,1]. The final single-objective optimization function is constructed to minimize the weighted sum of the Euclidean distances between each normalized objective value and the ideal point, and its expression is: ； in, For the first The normalized function value of each sub-objective For the first Sub-goals in decision variables The function value below, For the first The ideal point for each sub-goal For the first The worst solution for each sub-objective These are the weighting coefficients for each sub-objective.

9. A virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S5, the constraints that the model solution needs to satisfy include: (1) Power balance constraints of each node within the virtual power plant: ； in, For nodes exist Solar power output at all times For nodes exist Constantly exchanging power with the main power grid, For nodes exist Discharge power at any given time For nodes exist The load demand at any time, For nodes exist The charging power at any given time; (2) State of charge (SoC) iteration constraints for electric vehicles and battery swapping stations: ； in, In order to be in The state of charge of the battery at any given time. In order to be in The state of charge of the battery at any given time. For charging efficiency, In order to be in Charging power at any time For time step, For discharge efficiency, Let be the discharge power at time t. This refers to the battery's rated capacity. (3) Incentive price constraint: stipulate the internal incentive price set by the virtual power plant. Electricity prices must be in the external market. It fluctuates within a certain range, that is, it satisfies: ; in and These are the preset discount and markup coefficients.

10. A virtual power plant multi-objective scheduling method considering uncertainty and user priority according to claim 1, characterized in that, In step S2, the specific method for determining the operating mode of the battery BSS in the battery swapping station is as follows: set the photovoltaic output threshold and the peak and off-peak electricity price periods; when the photovoltaic output is higher than the threshold during the day, control the battery swapping station to be in the charging and storage mode G2B to absorb the excess photovoltaic power generation; at night or during peak grid electricity price periods, control the battery swapping station to be in the discharging and support mode B2G to supply power to the microgrid or load; at the same time, the number of backup batteries in the battery swapping station must meet the minimum battery swapping demand constraint.

Images