Virtual power plant two-stage optimization method considering flexible load and renewable energy

Abstract

The invention discloses a virtual power plant two-stage optimization method considering flexible load and renewable energy, which comprises the following steps of: firstly, determining an internal structure of a virtual power plant and a rule of participating in market transaction, and then establishing a day-ahead scheduling optimization model by taking benefit maximization of the virtual power plant as a target; solving the optimal scheduling model to obtain the flexible load and the output of the renewable energy source; and finally, establishing a real-time transaction optimization model according to the flexible load and the output of the renewable energy source, and performing power distribution on the flexible load and the renewable energy source to obtain an optimal scheduling scheme. According to the method, the utilization rate of renewable energy sources can be effectively improved, and the income of the virtual power plant is improved while user requirements are guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of virtual power plant scheduling, and in particular to a two-stage optimization method for virtual power plants considering flexible loads and renewable energy. Background Art

[0002] As one of the three major electric energy indicators, frequency stability is of great significance to industrial production and residents' lives. However, with the continuous expansion of renewable energy, frequency regulation is now greatly challenged by the increasing randomness and intermittency of more and more renewable power generation units. Therefore, it is necessary to explore new frequency regulation methods to assist traditional units in frequency regulation of power systems.

[0003] Developing the potential of urban flexible loads to participate in power system frequency regulation has become a good technical direction. As mentioned above, electric vehicles in smart buildings, as power system loads, can enter the V2G mode to transmit electricity to the grid: air-conditioning loads also show good power response capabilities, which can reduce the impact on human comfort according to thermal inertia. Specifically, flexible loads in smart buildings are usually characterized by disordered operation, dispersion, and small power. Through the aggregation of virtual power plants (VPPs), flexible loads in smart buildings, rooftop photovoltaics, and energy storage can be jointly dispatched. With the help of virtual power plants, flexible loads can participate in the frequency auxiliary services of the power system and reduce the frequency regulation pressure of the power system. However, the existing virtual power plant dispatch ignores the impact of flexible loads and renewable energy, resulting in greater frequency regulation pressure. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a two-stage optimization method for virtual power plants taking into account flexible loads and renewable energy in order to improve the utilization rate of renewable energy and the revenue of virtual power plants.

[0005] Technical solution: The two-stage optimization method of a virtual power plant considering flexible load and renewable energy according to the present invention comprises the following steps:

[0006] Determine the internal structure of the virtual power plant and the rules for participating in market transactions, wherein the components of the internal structure of the virtual power plant include flexible loads and renewable energy, the flexible loads include energy storage batteries, air conditioning loads and electric vehicles, and the renewable energy includes rooftop photovoltaics;

[0007] A day-ahead dispatch optimization model is established with the goal of maximizing the benefits of the virtual power plant, and the output of flexible loads and renewable energy is obtained by solving the optimization dispatch model; the benefits of the virtual power plant include the benefits of the electricity market and the benefits of the frequency regulation market;

[0008] A real-time trading optimization model is established according to the output of flexible loads and renewable energy, and the power of flexible loads and renewable energy is allocated to obtain their optimal scheduling plan.

[0009] Furthermore, the rules for the virtual power plant to participate in market transactions include:

[0010] In the day-ahead phase, virtual power plants bid in the electricity market and frequency regulation market by predicting their own output levels;

[0011] In the real-time stage, the grid operator calculates the real-time output of the virtual power plant by combining the bidding capacity of the electricity market and the frequency regulation market and issues a regulation signal;

[0012] After receiving the regulation signal, the virtual power plant will decompose the regulation signal into each component unit; each component unit will adjust the output, and the market will charge a penalty for the part that fails to be executed.

[0013] Furthermore, the objective function of the day-ahead scheduling optimization model is:

[0014]

[0015] Where T is the scheduling period, For electricity market revenue, For FM market revenue;

[0016]

[0017]

[0018] in, and is the predicted electricity selling price and electricity purchasing price from the power grid in time period t; and It is the amount of electricity sold and purchased by the virtual power plant when trading with the power grid during time period t; Δt is a time period; is the price of the day-ahead reserve regulation capacity predicted by the virtual power plant based on historical data during period t; It is the reserve regulation capacity traded between the virtual power plant and the power grid during time period t.

[0019] Furthermore, the power of the virtual power plant participating in the electricity market is the sum of the power of each component unit:

[0020]

[0021]

[0022] Among them, P E , It is the actual output of virtual power plants, photovoltaics, energy storage batteries, electric vehicles and air conditioning loads participating in the electricity market; among them, Greater than zero indicates that the energy storage battery is discharged, while Greater than zero indicates EV charging;

[0023] The frequency regulation capabilities of virtual power plants participating in the frequency regulation market are as follows:

[0024]

[0025]

[0026]

[0027] In the formula, is the maximum value of photovoltaic output power; is the maximum charging and discharging power of batteries and electric vehicles; is the maximum and minimum operating power of the air conditioning load; It is the ability to scale up and down that can be provided by a virtual power plant.

[0028] Furthermore, the constraints of the day-ahead dispatch optimization model include photovoltaic power generation operation constraints, energy storage operation constraints, air conditioning load operation constraints, electric vehicle operation constraints and photovoltaic forecast robustness constraints.

[0029] Furthermore, the objective function of the real-time transaction optimization model is:

[0030]

[0031] Among them, α f ,α E is the penalty factor. If the real-time capacity is higher than the day-ahead predicted capacity, (.) takes the value of +, otherwise it takes the value of -; Ψ f,t ,Ψ E,t is the deviation between the frequency regulation market and the electricity market; t E is the market price of electricity.

[0032] Furthermore, power distribution is performed according to the energy of the energy storage battery, air conditioning virtual load and electric vehicle load:

[0033] when hour,

[0034]

[0035] when hour,

[0036]

[0037] in, is the actual battery output; the actual output of the air conditioning load and the output of the electric vehicle cluster It can be calculated in the same way; when one of the battery, air conditioning load and electric vehicle clusters cannot fully respond, the missing part is made up by the other members.

[0038] The two-stage optimization system of a virtual power plant considering flexible load and renewable energy according to the present invention comprises:

[0039] A virtual power plant information collection unit, used to determine the internal structure of the virtual power plant and the rules for participating in market transactions, wherein the components of the internal structure of the virtual power plant include flexible loads and renewable energy, wherein the flexible loads include energy storage batteries, air conditioning loads and electric vehicles, and the renewable energy includes rooftop photovoltaics;

[0040] The day-ahead dispatch optimization unit is used to establish a day-ahead dispatch optimization model with the goal of maximizing the benefits of the virtual power plant, and solve the optimization dispatch model to obtain the output of the flexible load and renewable energy; the benefits of the virtual power plant include the benefits of the electricity market and the benefits of the frequency regulation market;

[0041] The real-time transaction optimization unit is used to establish a real-time transaction optimization model according to the output of flexible loads and renewable energy, allocate power to flexible loads and renewable energy, and obtain their optimal scheduling plan.

[0042] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the two-stage optimization method for virtual power plants considering flexible loads and renewable energy is implemented.

[0043] The computer-readable storage medium described in the present invention stores a computer program, and when the computer program is executed by a processor, it implements the two-stage optimization method of the virtual power plant considering flexible loads and renewable energy.

[0044] Beneficial effects: Compared with the prior art, the advantages of the present invention are: the present invention takes into account the good power response capabilities of flexible loads and renewable energy, and effectively alleviates the frequency regulation pressure of the power system by jointly scheduling them; the two-stage optimization model of the virtual power plant of this patent effectively improves the utilization rate of renewable energy, and increases the income of the virtual power plant while ensuring user needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the two-stage optimization method of the virtual power plant of the present invention;

[0046] Figure 2 A schematic diagram of the internal structure of a virtual power plant according to an embodiment of the present invention;

[0047] Figure 3 This is a diagram of the revenue analysis results of the virtual power plant according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0049] like Figure 1 As shown, the two-stage optimization method of virtual power plant considering flexible load and renewable energy includes the following steps.

[0050] S1, determine the internal structure of the virtual power plant for flexible loads and renewable energy:

[0051] Please see attached Figure 2 The internal structure of the virtual power plant considering flexible loads and renewable energy includes four units: rooftop photovoltaics, energy storage batteries, air-conditioning loads and electric vehicles.

[0052] A centralized controller is set up in the virtual power plant to collect the status information of all distributed units, including photovoltaic output, battery charging status, room temperature, electric vehicle SOC, etc. The centralized controller processes this information to calculate and predict the energy demand and regulation capacity of the virtual power plant and bid for it. At the same time, after receiving the regulation signal from the power grid, the centralized controller will issue control instructions, set the output of each unit, and coordinate internal resources. Each unit uses a local controller to receive the control signal and upload the status information and response information to the centralized controller.

[0053] S2 determines the rules for virtual power plants that take into account flexible loads and renewable energy to participate in market transactions.

[0054] In the day-ahead stage, the virtual power plant bids in the electricity market and frequency regulation market by predicting its own output level.

[0055] In the real-time stage, the grid operator calculates the real-time output of the virtual power plant by combining the bidding capacity of the electricity market and the frequency regulation market and issues a control signal.

[0056] After receiving the regulation signal, the virtual power plant will decompose the regulation signal into each component unit; each component unit will adjust the output, and the part that fails to be executed will be charged a penalty by the market.

[0057] S3, establish the objective function of the optimal scheduling model for virtual power plants to conduct day-ahead transactions.

[0058] The day-ahead objective function of the virtual power plant is to maximize revenue, which is expressed as follows:

[0059]

[0060] Where: T is the dispatching period, and the day-ahead stage is 24 hours; the total revenue is divided into two parts, namely the revenue of the virtual power plant participating in the electricity market at time t and the revenue of participating in the frequency regulation market.

[0061] The specific description of the revenue of each part of the virtual power plant is as follows:

[0062] 1) Electricity market revenue

[0063]

[0064] In the formula, and is the predicted electricity selling price and electricity purchasing price from the power grid in time period t; and It is the amount of electricity sold and purchased when the virtual power plant trades with the power grid within the time period t; Δt is a time period, which is 1h in the day-ahead stage.

[0065] 2) FM market revenue

[0066]

[0067] In the formula, is the price of the day-ahead reserve regulation capacity predicted by the virtual power plant based on historical data during period t; It is the reserve regulation capacity traded between the virtual power plant and the power grid during time period t.

[0068] In terms of the backup capacity provision of virtual power plants: photovoltaics provide backup capacity by abandoning light, and electric vehicles and energy storage batteries provide backup capacity by orderly charging through reasonable arrangement of charging plans.

[0069] The power of the virtual power plant participating in the electricity market is the sum of the power of each component unit, that is:

[0070]

[0071] Where P E , It is the actual output of virtual power plants, photovoltaics, energy storage batteries, electric vehicles and air conditioning loads participating in the electricity market; among them, Greater than zero indicates that the energy storage battery is discharged, while Greater than zero indicates EV charging.

[0072] The frequency regulation capabilities of virtual power plants participating in the frequency regulation market are as follows:

[0073]

[0074] In the formula, is the maximum value of photovoltaic output power; is the maximum charging and discharging power of batteries and electric vehicles; is the maximum and minimum operating power of the air conditioning load; It is the ability to scale up and down that can be provided by a virtual power plant.

[0075] S4, establish the constraints of the optimal scheduling model for virtual power plants to conduct day-ahead transactions.

[0076] The constraints of the optimal scheduling model for virtual power plants to conduct day-ahead transactions include photovoltaic power generation operation constraints, energy storage operation constraints, air conditioning load operation constraints, electric vehicle operation constraints, and photovoltaic forecast robustness constraints.

[0077] (1) PV power generation operation constraints

[0078] The photovoltaic power generation operation constraints of the optimal dispatch model for virtual power plants to conduct day-ahead transactions should satisfy that the actual output power of photovoltaic power generation is less than the maximum output power:

[0079]

[0080] (2) Energy storage operation constraints

[0081] The energy storage operation constraints of the optimal dispatch model for virtual power plants to conduct day-ahead transactions should meet the following requirements:

[0082]

[0083] In the formula, is the charge state of the energy storage battery at time t; It is the charging efficiency and discharging efficiency of the energy storage battery; are the minimum and maximum SOC of the battery; it is guaranteed that the energy storage battery has the same SOC at the beginning and end of the day, thus ensuring the continuity of regulation due to equation (13).

[0084] (3) Air conditioning load operation constraints

[0085] A virtual battery model of air conditioning load is established based on factors such as indoor and outdoor temperature and building characteristics: when the room temperature is the upper limit of temperature comfort, it means that the SOC of the virtual battery is zero; when the room temperature is the lower limit of temperature comfort, it means that the SOC of the virtual battery is full;

[0086] The first-order equivalent thermal parameters of the building are as follows:

[0087]

[0088] Where, T o is the outside temperature; T i is room temperature; Q AC is the cooling capacity of the air conditioner; C a is the equivalent specific heat capacity; R1 is the equivalent thermal impedance.

[0089] The electrical and thermal conversion model of the air conditioner is:

[0090]

[0091] Where k1, k2, l1, and l2 are the coefficients of the linear model.

[0092] Then, at time t, the capacity of the virtual battery of the air conditioning load is:

[0093]

[0094] In the formula, is the virtual battery capacity of the air conditioning load at time t; is the outdoor temperature at time t; It is the upper and lower limits of the set indoor comfort temperature zone.

[0095] In addition, the charging state of the air conditioning energy storage is defined as:

[0096]

[0097] Where, T i It is room temperature.

[0098] The constraints of air conditioning load in day-ahead scheduling can be expressed as:

[0099]

[0100] (4) Electric vehicle operation constraints

[0101] In the optimal scheduling model, it is assumed that the electric vehicles are connected to the grid at a fixed time. In order to reduce the computational workload of the upper-level scheduling and protect user privacy, a virtual battery model of the electric vehicle cluster can be established.

[0102]

[0103]

[0104] Where Ne is the number of electric vehicles.

[0105] For each EV, the power level is constrained to ensure that it can complete the charging target within the grid connection time. The energy change caused by the arrival and departure of EVs can be expressed as:

[0106]

[0107] Where W i,t is a Boolean variable representing the state of the i-th electric vehicle connected to the grid at time t.

[0108] Therefore, the change in energy generated by the entire EV cluster due to the arrival and departure of EVs is:

[0109]

[0110] The power constraint of electric vehicles can be expressed as:

[0111]

[0112] (5) Robustness constraints of photovoltaic forecasting

[0113] When the battery's SOC is high, the system as a whole is more responsive to upward signals and less responsive to downward signals; at this time, the predicted value of photovoltaics can be improved and the bidding ability of virtual power plants in the power market in the next day can be improved. The part of photovoltaic output that is insufficient in a day can be borne by energy storage, which can reduce the battery's SOC; similarly, when the battery's SOC is low, the photovoltaic forecast can be reduced so that excess power can be charged to energy storage in real time and the battery's SOC can be increased; by doing so, we can better keep the battery's SOC at 50%, reduce market penalties, and better ensure the continuous dispatch of the battery, that is, meet:

[0114]

[0115] In the formula, is the predicted power of photovoltaics; α is the robust prediction factor of photovoltaics.

[0116] S5, solve the day-ahead model to obtain the predicted output of each unit in the virtual power plant.

[0117] The predicted output of each unit in the virtual power plant refers to the power of each unit in the virtual power plant, that is, P E , It is the actual output of virtual power plants, photovoltaics, energy storage batteries, electric vehicles and air-conditioning loads participating in the electricity market.

[0118] S6, establishes an optimization model for virtual power plants to conduct real-time transactions based on predicted output.

[0119] In the real-time stage, energy storage batteries, electric vehicle virtual batteries, and air conditioning load virtual batteries can be aggregated into virtual batteries of virtual power plants;

[0120] In the electricity market, if the photovoltaic power generation exceeds the photovoltaic power generation in the day-ahead electricity market, the surplus is used as the frequency regulation market response power; otherwise, the shortfall is made up by the virtual battery of the virtual power plant;

[0121] In the frequency regulation market, first determine whether the PV frequency regulation market response power exceeds the regulation signal. If so, the excess power will be stored in the virtual battery of the virtual power plant; if not, the virtual battery of the virtual power plant will make up for it;

[0122] The output of the battery, the EV virtual battery, and the air conditioning load virtual battery is determined by the ratio of the energies.

[0123] The objective function of virtual power plant optimization in real time is to minimize the penalty:

[0124]

[0125] In the formula, α f ,α E is the penalty factor. If the real-time capacity is higher than the day-ahead forecast capacity, (.) takes a positive sign, and vice versa, it takes a negative sign; Ψ f,t ,Ψ E,t is the deviation between the frequency regulation market and the electricity market; t E is the market price of electricity.

[0126] In the real-time phase, the power constraints and capacity constraints are the same as those in the day-ahead phase, and the power deviation is calculated as follows:

[0127]

[0128] In the formula, is the power of the virtual power plant in dealing with the electricity market and frequency regulation market; Δ t It is a regulation signal for grid operators; It is the common response of energy storage batteries, air conditioning loads and electric vehicles.

[0129] Power distribution is performed based on the energy of energy storage batteries, air conditioning virtual loads, and electric vehicle loads:

[0130] when hour,

[0131]

[0132] when hour,

[0133]

[0134] In the formula, is the actual battery output; the actual output of the air conditioning load and the output of the electric vehicle cluster It can be calculated in the same way; when one of the battery, air conditioning load and electric vehicle clusters cannot fully respond, the missing part is made up by the other members.

[0135] S7, solve the established model to obtain the optimal scheduling plan, that is, the output plan of each unit, which is the command value of the output power.

[0136] Please see attached Figure 3 When the constraints are met, real-time scheduling is usually performed according to the predicted instructions of the day-ahead scheduling. The virtual power plant day-ahead optimal scheduling model constructed by the present invention is a mixed integer linear programming model, which is solved by calling the GUROBI solver in MATLAB R2021a. The data of photovoltaic, battery, air conditioning load and electric vehicle are taken from the real data of smart buildings. The bidding results of virtual power plants in the day-ahead power market and frequency regulation market are shown in Figure 2. Figure 3 (a) and Figure 3 As shown in (b). Virtual power plants can effectively participate in the electricity market and frequency regulation market to obtain higher profits while ensuring customer needs. When the virtual power plant only participates in the electricity market, the profit is -2251 yuan and the fine is 3093 yuan, while when it participates in the combined electricity market and frequency regulation market, the profit is 20088 yuan and the fine is 2292 yuan; the comparison results show that the participation of virtual power plants in the electricity market and frequency regulation market can effectively improve their benefits.

[0137] The two-stage optimization system of a virtual power plant considering flexible load and renewable energy according to the present invention comprises:

[0138] A virtual power plant information collection unit, used to determine the internal structure of the virtual power plant and the rules for participating in market transactions, wherein the components of the internal structure of the virtual power plant include flexible loads and renewable energy, wherein the flexible loads include energy storage batteries, air conditioning loads and electric vehicles, and the renewable energy includes rooftop photovoltaics;

[0139] The day-ahead dispatch optimization unit is used to establish a day-ahead dispatch optimization model with the goal of maximizing the benefits of the virtual power plant, and solve the optimization dispatch model to obtain the output of the flexible load and renewable energy; the benefits of the virtual power plant include the benefits of the electricity market and the benefits of the frequency regulation market;

[0140] The real-time transaction optimization unit is used to establish a real-time transaction optimization model according to the output of flexible loads and renewable energy, allocate power to flexible loads and renewable energy, and obtain their optimal scheduling plan.

[0141] The electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the two-stage optimization method for virtual power plants considering flexible loads and renewable energy is implemented.

[0142] The computer-readable storage medium described in the present invention stores a computer program, and when the computer program is executed by a processor, it implements the two-stage optimization method of the virtual power plant considering flexible loads and renewable energy.

[0143] The computer-readable storage media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory, or any other medium that can be used to store the desired program code in the form of instructions or data structures and that can be accessed by a computer.

[0144] The processor is used to execute the computer program stored in the memory to implement each step of the method involved in the above embodiment.

Claims

1. A two-stage optimization method for virtual power plants considering flexible loads and renewable energy, characterized in that: The steps include: Determine the internal structure of the virtual power plant and the rules for participating in market transactions, wherein the components of the internal structure of the virtual power plant include flexible loads and renewable energy, the flexible loads include energy storage batteries, air conditioning loads and electric vehicles, and the renewable energy includes rooftop photovoltaics; A day-ahead dispatch optimization model is established with the goal of maximizing the benefits of the virtual power plant, and the output of flexible loads and renewable energy is obtained by solving the optimization dispatch model; the benefits of the virtual power plant include the benefits of the electricity market and the benefits of the frequency regulation market; A real-time trading optimization model is established according to the output of flexible loads and renewable energy, and the power of flexible loads and renewable energy is allocated to obtain their optimal scheduling plan.

2. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 1 is characterized in that: The rules for the virtual power plant to participate in market transactions include: In the day-ahead phase, virtual power plants bid in the electricity market and frequency regulation market by predicting their own output levels; In the real-time stage, the grid operator calculates the real-time output of the virtual power plant by combining the bidding capacity of the electricity market and the frequency regulation market and issues a regulation signal; After receiving the regulation signal, the virtual power plant will decompose the regulation signal into each component unit; each component unit will adjust the output, and the market will charge a penalty for the part that fails to be executed.

3. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 1 is characterized in that: The objective function of the day-ahead scheduling optimization model is: Where T is the scheduling period, For electricity market revenue, For FM market revenue; in, and is the predicted electricity selling price and electricity purchasing price from the power grid in time period t; and It is the amount of electricity sold and purchased by the virtual power plant when trading with the power grid during time period t; Δt is a time period; is the price of the day-ahead reserve regulation capacity predicted by the virtual power plant based on historical data during period t; It is the reserve regulation capacity traded between the virtual power plant and the power grid during time period t.

4. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 3 is characterized in that: The power of a virtual power plant participating in the electricity market is the sum of the power of each component unit: in, It is the actual output of virtual power plants, photovoltaics, energy storage batteries, electric vehicles and air conditioning loads participating in the electricity market; among them, Greater than zero indicates that the energy storage battery is discharged, while Greater than zero indicates EV charging; The frequency regulation capabilities of virtual power plants participating in the frequency regulation market are as follows: In the formula, is the maximum value of photovoltaic output power; is the maximum charging and discharging power of batteries and electric vehicles; is the maximum and minimum operating power of the air conditioning load; It is the ability to scale up and down that can be provided by a virtual power plant.

5. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 1 is characterized in that: The constraints of the day-ahead dispatch optimization model include photovoltaic power generation operation constraints, energy storage operation constraints, air conditioning load operation constraints, electric vehicle operation constraints and photovoltaic forecast robustness constraints.

6. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 1 is characterized in that: The objective function of the real-time transaction optimization model is: Among them, α f ,α E is the penalty factor. If the real-time capacity is higher than the day-ahead predicted capacity, (.) takes the value of +, otherwise it takes the value of -; Ψ f,t ,Ψ E,t is the deviation between the frequency regulation market and the electricity market; t E is the market price of electricity.

7. The two-stage optimization method for virtual power plants considering flexible loads and renewable energy according to claim 1 is characterized in that: Power distribution is performed based on the energy of energy storage batteries, air conditioning virtual loads, and electric vehicle loads: when hour, when hour, in, is the actual battery output; the actual output of the air conditioning load and the output of the electric vehicle cluster It can be calculated in the same way; when one of the battery, air conditioning load and electric vehicle clusters cannot fully respond, the missing part is made up by the other members.

8. A two-stage optimization system for virtual power plants considering flexible loads and renewable energy, characterized in that: include: A virtual power plant information collection unit, used to determine the internal structure of the virtual power plant and the rules for participating in market transactions, wherein the components of the internal structure of the virtual power plant include flexible loads and renewable energy, wherein the flexible loads include energy storage batteries, air conditioning loads and electric vehicles, and the renewable energy includes rooftop photovoltaics; The day-ahead dispatch optimization unit is used to establish a day-ahead dispatch optimization model with the goal of maximizing the benefits of the virtual power plant, and solve the optimization dispatch model to obtain the output of the flexible load and renewable energy; the benefits of the virtual power plant include the benefits of the electricity market and the benefits of the frequency regulation market; The real-time transaction optimization unit is used to establish a real-time transaction optimization model according to the output of flexible loads and renewable energy, allocate power to flexible loads and renewable energy, and obtain their optimal scheduling plan.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into a processor, it implements the two-stage optimization method for a virtual power plant considering flexible loads and renewable energy according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the two-stage optimization method of a virtual power plant considering flexible loads and renewable energy according to any one of claims 1 to 7.

Images