A method, device and medium for distributed energy participating in virtual power plant dispatching

Abstract

The application provides a method for distributed energy to participate in virtual power plant dispatching, comprising the following steps: S1: acquiring the types and quantities of all distributed energy in the virtual power plant, and determining the adjustable capacity and adjustable cost cost of each distributed energy; S2: the virtual power plant receives power grid dispatching demand, and establishes an internal balanced dispatching model of the virtual power plant; S3: a dynamic optimization algorithm is used to calculate the dispatching strategies of different types of distributed energy in the virtual power plant; S4: the dispatching strategies of different types of distributed energy calculated in step S3 are brought into the internal balanced dispatching model of the virtual power plant established in step S2, so that all distributed energy in the virtual power plant is dispatched. Multiple types of distributed power generation resources in the virtual power plant are all involved in collaborative dispatching.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of virtual power plant dispatching, and particularly relates to a method, device and medium for balancing dispatching of a virtual power plant with participation of multiple types of distributed energy. BACKGROUND

[0002] In recent years, the development focus of power systems in various countries is shifting to clean and decentralized types. Distributed energy has the characteristics of reasonable energy utilization, small loss, less pollution, flexible operation, and good system economy, and can effectively promote the low-carbon and environmentally friendly operation of power systems.

[0003] The following technical problems exist in the actual operation of distributed energy: (1) The layout of new distributed power generation resources has the characteristics of dispersion and small scale, and it is difficult to participate in power system dispatching and power market when connected to the grid, and it cannot fully play its active role;

[0004] (2) Some users are based on distributed power generation technology, rely on small power stations of dispersed resources or users, realize the integrated operation of internal power supply and load microgrid, and access the main grid or independently autonomous operation through coordinated control with the main grid. Some methods now aggregate distributed power into virtual power plants, and consider the virtual power plant as a whole power supply to participate in grid operation and dispatching control. However, there is a lack of a method for balancing aggregation of various types of distributed power generation resources in the virtual power plant.

[0005] Therefore, there is an urgent need for a method for distributed energy to participate in virtual power plant dispatching, so that multiple types of distributed power generation resources in the virtual power plant participate in collaborative dispatching. SUMMARY

[0006] The purpose of the present application is to provide a method for distributed energy to participate in virtual power plant dispatching, to solve the problem of internal balancing dispatching of the virtual power plant.

[0007] The technical solution of the present application: a method for distributed energy to participate in virtual power plant dispatching, comprising the following steps:

[0008] S1: obtaining the types and quantities of all distributed energy in the virtual power plant, and determining the adjustable capacity P and adjustable cost cost of each distributed energy; si

[0009] S2: the virtual power plant receives the grid dispatching demand, and establishes an internal balancing dispatching model of the virtual power plant;

[0010] S3: using a dynamic optimization algorithm to calculate the dispatching strategy of different types of distributed energy in the virtual power plant;

[0011] ​S4: the scheduling strategy of different kinds of distributed energy calculated in step S3 is brought into the internal balanced scheduling model of the virtual power plant established in S2, so as to schedule all the distributed energy in the virtual power plant;

[0012] The dynamic optimization algorithm in S3 is a breadth uniform algorithm, and the specific steps include the following:

[0013] S3.1: input initialization parameters, initial population size N, minimum size NM, merging distance r', core particle set KL size N', maximum iteration number TL, TL is initially set to an integer of 500-1000, particle mass range [m1, m2], which is initially set to [1, 25];

[0014] Wherein, any one particle is represented as:

[0015] pa i ={P1,...P c ...P C}

[0016] In the formula, P c is the power emitted by the cth resource, P C is the power emitted by the cth resource, pa i is the ith particle, and the total number is N;

[0017] According to the constraint condition, pa i ={P1,...P c ...P C}, 0<P c ≤P si , an initial particle is randomly generated in the solution space, a two-layer particle, and two particles are randomly generated in a sphere with a radius of 3r' centered on the initial particle. Similarly, a three-layer particle is further generated with the two-layer particle as the center. Finally, an initial population S with a size of N is generated, each particle in the population is assigned a random value in the particle mass range [m1, m2], and the initial iteration count t is set to 1;

[0018] S3.2: calculate the dominance relationship of particles in the initial population S based on the Pareto relationship, create a core particle set KL, and copy the non-inferior solution to the core particle set KL;

[0019] S3.3: traverse all particles and calculate the Pareto dominance relationship between them,

[0020] If there is a dominance relationship between particles pa i and pa j , select the particle g* with the greatest attraction from the core particle set KL, and the dominated particles pa i and pa j, move towards the core solution and the dominated solution according to the following formula, where g* is the core solution;

[0021]

[0022] where: pa i (t) is the vector representation of particle i at time t, pa j (t) is the vector representation of particle j at time t, pa g* (t) is the vector representation of particle g* at time t, Ri g* is the Euclidean distance between particle i and particle g*;

[0023] Otherwise, select the most attractive g* from the core particle set KL, the non-dominated particle pa i and pa j , move according to the following formula;

[0024]

[0025] where α is a random value between 0 and 1, G * is the mass gravitational coefficient, R ij is the Euclidean distance between particles, m is the mass of the particle, and θ is a random position, ensuring that the solution particle tries to explore more space in the movement; m i is the mass of particle i, m j is the mass of particle j, m g* is the mass of g* particle;

[0026] S3.4: Calculate the distance between particles, if the distance between particles is close enough, less than the merging distance, and the number of particles is greater than the minimum size NM, the particles are merged, and the mass of the merged particles is updated according to the following formula;

[0027] m new = m i + m j

[0028] where m new is the mass of the new particle after merging, m i is the mass of particle i, m j is the mass of particle j, otherwise, proceed to the next step;

[0029] S3.5: Update the particle position, calculate the dominance relationship of particles in the group based on the Pareto relationship, and update the core particle set according to the non-inferior solution sorting;

[0030] S3.6: If the number of iterations is less than the maximum number of iterations TL, return to step 3, otherwise, end the iteration and proceed to the next step;

[0031] S3.7: output the solution in the core particle set, and the output solution is a candidate solution of the optimal solution.

[0032] Preferably, the overall objective function of the internal equilibrium scheduling model in S2 is:

[0033]

[0034] wherein F cost represents the overall cost of the participation of each distributed energy in the virtual power plant in regulation; c is the total number of distributed energies in the virtual power plant, which is a fixed constant, and cost i is the cost of the distributed power generation resource i, which is a constant, and cost i > 0.

[0035] The overall objective function is the minimum value of the overall cost: min(F cost ).

[0036] Preferably, the internal equilibrium scheduling model further comprises another objective function: the maximum output power of the green distributed energy of the virtual power plant, i.e.

[0037]

[0038] wherein P clean is the total output of the green distributed energy of the virtual power plant, n is the number of green distributed energies of the virtual power plant, is the output power of the jth green distributed energy.

[0039] Preferably, the constraints calculated by the internal equilibrium scheduling model include:

[0040] 1) system power balance constraint:

[0041]

[0042] c is the total number of distributed energies in the virtual power plant, which is a fixed constant, and P c is the power output by the cth resource; and p out < p s , wherein p s is the adjustable capacity of the virtual power plant, which is the maximum value that can be output externally by the virtual power plant, and is a limited constant.

[0043] 2) distributed power generation resource output constraint: P c needs to be within the range of its adjustable power, i.e.

[0044] 0 < P c ≤ P si

[0045] wherein P siThe maximum value of each distributed power source is adjusted to the maximum rated value in a safe operation state.

[0046] Preferably, one optimal solution candidate solution in S4 is taken into the objective function, i.e., the minimum sum of the total cost and the maximum green distributed energy output power, to obtain the final scheduling strategy.

[0047] The application further provides a computer device, including a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the method for participating in virtual power plant scheduling by the distributed energy source.

[0048] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor to implement the method for participating in virtual power plant scheduling by the distributed energy source.

[0049] Compared with the prior art, the application has the following beneficial effects:

[0050] (1) The various types of distributed power sources in the virtual power plant are all involved in collaborative scheduling, and in the controllable capacity range, the distributed power sources are relatively balanced by calculating the regulation cost of each distributed power source.

[0051] (2) The benefits of the green distributed power source are fully brought into play, and the method is simple and easy to expand, and the scheduling cost can be effectively controlled.

[0052] (3) In the dynamic optimization calculation, the mass of the particle can improve the mutual attraction, the mass of the particle is combined through the optimization mechanism, the mass of the aggregated particle is changed, the moving speed is improved, and the convergence speed of the algorithm is accelerated; the particle swarm is initialized in layers, the particles are more widely and uniformly distributed in the solution space, and good basic conditions are provided for fast exploration of non-inferior solutions; through the addition of the particle combination mechanism, the number of particles is reduced after combination under the condition that the particles meet the combination condition, so that the number of algorithm cycles is reduced, and the calculation speed is further accelerated.

[0053] (4) The optimal solution obtained by the algorithm is multiple, and the appropriate power combination can be selected according to the actual weather and the operation state of the equipment, and the flexibility is higher.

[0054] The application can help to collaboratively schedule the various types of distributed power sources in the virtual power plant, fully bring the benefits of each distributed power source into play, and effectively control the scheduling cost. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 The flowchart of the application;

[0056] Figure 2 Flowchart of dynamic optimization calculation. DETAILED DESCRIPTION

[0057] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0058] As attached Figure 1 As shown, a method for distributed energy to participate in virtual power plant scheduling includes the following steps:

[0059] S1: Obtain the types and quantities of all distributed energy resources in the virtual power plant and determine the adjustable capacity P of each distributed energy resource. si and adjustable cost cost; where P si is the adjustable maximum value of each distributed power source, taking the maximum rated value under the safe operating state, cost is the cost of distributed power generation resources, its value is a constant, cost>0; confirm the distributed energy within the aggregation range of the virtual power plant, the types of distributed energy can be distributed photovoltaic, wind power, electric vehicles, controllable loads, energy storage systems, etc.

[0060] S2: The virtual power plant receives grid dispatching requirements and establishes an internal equilibrium dispatching model for the virtual power plant;

[0061] S3: Using dynamic optimization algorithms, calculate the scheduling strategies for different types of distributed energy within the virtual power plant;

[0062] S4: The scheduling strategies of different types of distributed energy calculated in step S3 are brought into the internal balanced scheduling model of the virtual power plant established in S2, so as to schedule all distributed energy in the virtual power plant, so that the virtual power plant can efficiently participate in the unified scheduling of the power grid.

[0063] According to the grid dispatch instruction, that is, the power demand, the power p provided by the virtual power plant to the grid system is determined out .p out Less than the maximum adjustable capacity p of the virtual power plant s .p s is the maximum value that can be output from the virtual power plant, which is a limiting constant. Those skilled in the art can calculate it based on the specific system construction situation.

[0064] Power p supplied to the grid system out It is composed of various distributed power sources and is expressed as

[0065]

[0066] where c is the total number of all resources in the virtual power plant, which is a fixed constant. P i is the power output of the i-th resource.

[0067] The control center of the virtual power plant allocates power dispatching according to the operating cost of each distributed power resource, its adjustable capacity and green energy power.

[0068] According to the adjustable power of each distributed power resource and its adjustment cost, the total objective function of the virtual power plant is composed of

[0069]

[0070] where F cost represents the total cost of each distributed energy participating in the adjustment of the virtual power plant; cost i is the cost of the i-th distributed power resource, which is usually constant within a certain period of time, and cost i > 0. This data can be calculated by those skilled in the art according to the specific system construction situation.

[0071] Therefore, one of the objective functions of the internal scheduling optimization of the method is:

[0072] min(F cost )

[0073] At the same time, in order to fully exert the benefits of green distributed power and effectively control the dispatching cost, another objective function is introduced, which is the maximum output power of the green distributed energy of the virtual power plant, i.e.

[0074]

[0075] where n is the number of green distributed energy of the virtual power plant, is the output power of the j-th green distributed energy, such as photovoltaic and wind energy. P clean is the total output of the green distributed energy of the virtual power plant.

[0076] In the running process, the following constraints need to be followed:

[0077] 1) System power balance constraint, all distributed power resources must meet the system requirements for virtual power plant power in the dispatching scheme solving process, i.e.

[0078]

[0079] 2) Distributed power resource output constraint, the output of each distributed power resource needs to be within its adjustable power range, i.e.

[0080] 0 < P i ≤ P si

[0081] P si is the maximum adjustable value of each distributed power supply, which is usually the maximum rated value in the safe operation state.

[0082] Referring to Figure 2 , the balanced power generation strategy solving process of the distributed power supply inside the virtual power plant:

[0083] S3.1: input initialization parameters, population size initial size N, minimum size NM, merging distance r', core particle set KL size N', N, NM, r', N' such numerical values have no universal preference, and need to be adjusted according to specific data conditions each time, the maximum iteration number TL, which is usually adjusted according to the data size, TL is initially set to an integer between 500 and 1000, particle mass range [m1, m2], mainly used to differentiate particles, initially set to [1, 25];

[0084] wherein any one particle is represented as: pa i = {P1,...P c ...P C} that is, the power of all distributed energy sources inside the virtual power plant;

[0085] In the formula, P c is the power emitted by the cth resource, P C is the power emitted by the cth resource, pa i is the ith particle, and the total number is N; that is, different power combinations are randomly generated within the constraint range;

[0086] According to the constraint condition, pa i = {P1,...P c ...P C}, 0 < P c ≤ P si , an initial particle is randomly generated in the solution space, and two particles are randomly generated in a sphere with a radius of 3 times r' centered on the initial particle. Similarly, three-layer particles are further generated with the two-layer particles as the center. Finally, an initial population S with a size of N is generated, each particle in the population is assigned a random value within the particle mass range [m1, m2], and the initialization iteration count t is set to 1.

[0087] S3.2: calculate the dominance relationship of the particles in the initial population S based on the Pareto relationship, create a core particle set KL, and copy the non-inferior solution to the core particle set KL.

[0088] S3.3: Traverse all particles and calculate the Pareto dominance relationship between them.

[0089] If the particle pa i and pa j There is a dominance relationship between them. From the core particle set KL, the particle g* with the largest gravitational force is selected, and the dominated particle pa i and pa j , according to the following formula, move towards the core solution and the dominant solution, where g* is taken as the core solution;

[0090]

[0091] Where: pa i (t) is the vector representation of particle i at time t, pa j (t) is the vector representation of particle j at time t, pa g* (t) is the vector representation of particle g* at time t, Ri g* is the Euclidean distance between particle i and particle g*;

[0092] Otherwise, select g* with the largest gravitational force from the core particle set KL, and the non-dominated particle pa i and pa j , move according to the following formula;

[0093]

[0094] Among them, α is a random value (0, 1), G * is the mass gravitational coefficient, R ij is the Euclidean distance between particles, m is the mass of the particle, θ is the random position, which ensures that the solution particle tries to explore more space during movement; mi is the mass of particle i, mj is the mass of particle j, mg* is the mass of particle g*, and Rig* is the Euclidean distance between particle i and particle g*;

[0095] S3.4: Calculate the distance between particles. If the distance between particles is close enough to be smaller than the merging distance and the number of particles is greater than the minimum size NM, the particles are merged. Update the mass of the merged particles according to the following formula;

[0096] m new =m i +m j

[0097] Among them, m new is the mass of the new particle after merging, m i is the mass of particle i, m j is the mass of particle j, otherwise, proceed to the next step;

[0098] S3.5: update the position of particles, calculate the dominance relationship of particles in the group based on the Pareto relationship, and update the core particle set according to the non-inferior solution ranking;

[0099] S3.6: if the iteration number is less than the maximum iteration number TL, return to step three, otherwise, end the iteration and perform the next step;

[0100] S3.7: finally form a set of optimal solutions (Pareto surface), output the solution in the core particle set as the candidate solution of the optimal solution. Each particle in the core particle set represents the power output of each distributed energy source respectively; that is, different power combinations are obtained.

[0101] The optimal solution obtained by the algorithm is multiple, and the appropriate power combination can be selected according to the actual weather and the running state of the equipment, which is more flexible; for example, when the weather is not good, a particle with low photovoltaic power is selected;

[0102] In S4, an optimal candidate solution is selected according to the actual situation (weather, equipment running state) and brought into the objective function, that is, the minimum total cost and the maximum output power of green distributed energy, to obtain the final scheduling strategy. Since the virtual power plant is not a real power plant, not all power generation is transmitted to the power grid. They contain multiple types, and even adjustable loads, and the power of the virtual power plant is generally given to the power grid on demand, and some are even used internally. For example, the self-provided power plant of some large steel plants will inevitably change the cost if they change their power consumption for the power grid, so a target function needs to be set.

[0103] Finally, the adjustable capacity of each distributed power source is solved, and unified participation in the grid dispatching is carried out.

[0104] Taking a virtual power plant in a certain city as an example, the virtual power plant includes one microgrid, two distributed power sources and one controllable load cluster, and the method for the distributed energy sources to participate in the virtual power plant dispatching includes the following specific steps:

[0105] Firstly, the adjustable capacity P si of each distributed energy source is determined, and the adjustable capacity of the distributed energy sources in the virtual power plant is shown in the following table:

[0106] resource microgrid photovoltaic small gas generation controllable load adjustable capacity / mw 60 40 28 25

[0107] Among them, the adjustable cost cost of each resource is as shown in the following table

[0108] resource microgrid photovoltaic small gas generation controllable load cost 500 150 260 320

[0109] In the next scheduling period, the power grid system needs the virtual power plant to provide capacity p out = 100 MW.

[0110] According to the calculation steps of the method for participating in the virtual power plant dispatch by the distributed energy:

[0111] Input initialization parameters, population size initial size N=50, minimum size NM=25, merging distance r'=5, core particle set KL size N'=15, maximum iteration number TL=1000, particle mass range [10, 25].

[0112] 1) According to the constraint condition, an initial particle is randomly generated in the solution space, and two particles are randomly generated in the spherical surface with a radius of 3r' centered on the initial particle. Similarly, three particles are further generated centered on the two particles, and finally, an initial population S with a size of N is generated. Each particle in the population is assigned a random value in the particle mass range [10, 25], and the initial iteration count t is set to 1.

[0113] 2) Calculate the dominance relationship of particles in the S group based on the Pareto relationship, create a core set KL, and copy the non-inferior solution to the core particle solution set KL.

[0114] 3) Traverse all particles and calculate their Pareto dominance relationship. If there is a dominance relationship between particles p i and p j , select the particle with the greatest gravitational force g* from the set KL, and the dominated particles p i and p j , according to the following formula, move towards the core solution g* and the dominated solution.

[0115]

[0116] Otherwise, select the particle with the greatest gravitational force g* from the set KL, and the non-dominated particle p i and p j , according to the following formula, move to move.

[0117]

[0118] Where α is a random value between 0 and 1, G * is the mass gravitational coefficient, R ij is the Euclidean distance between particles, m is the particle mass, and θ is a random position, which ensures that the solution particle tries to explore more space during movement.

[0119] 4) Calculate the distance between particles. When the distance between particles is close enough and the number of particles is greater than the minimum size NM, the particles are merged, and the mass of the merged particles is updated according to the following formula.

[0120] m new = m i +m j

[0121] 5) Update the position of particles, and calculate the dominance relationship of particles in the group based on the Pareto relationship. At the same time, update the core particle set according to the non-inferior solution ranking.

[0122] 6) Return to step 3 until the number of iterations TL is reached.

[0123] 7) Output the solution in the core particle set, and obtain the final scheduling strategy according to the objective function.

[0124] Finally, the adjustment strategy of each distributed power supply is determined as shown in the following table.

[0125] p out ]]> microgrid photovoltaic small gas generation controllable load 100 mw 38 34 15 13

[0126] The present application makes various types of distributed power resources in the virtual power plant participate in collaborative scheduling, and in the controllable capacity range, the adjustment and control cost of each distributed power resource is calculated to make the distributed power resources be adjusted and controlled relatively evenly; the benefits of green distributed power supply are fully brought into play, and the method is simple and easy to expand, and can effectively control the scheduling cost; when dynamic optimization calculation is performed, the quality of particles can improve their mutual attraction, the merging mechanism of the quality of optimized particles is used to change the quality of gathered particles, improve the moving speed, and accelerate the convergence speed of the algorithm; the hierarchical initialization distribution of the optimized particle swarm is used to make the particles more widely and uniformly distributed in the solution space, and provide good basic conditions for fast exploration of non-inferior solutions; by adding the particle merging mechanism, the number of particles is reduced after merging under the condition that the particles meet the merging condition, so that the number of algorithm loop is reduced, and the calculation speed is further accelerated.

[0127] The present application is not limited to the above-mentioned embodiments, and based on the technical solutions disclosed in the present application, those skilled in the art can make some substitutions and deformations to some technical features according to the disclosed technical content without creative labor, and these substitutions and deformations are all within the protection scope of the present application.

Claims

1. A method for distributed energy to participate in virtual power plant scheduling, characterized by: The following steps are involved: S1: Obtain the types and quantities of all distributed energy resources in the virtual power plant, and determine the adjustable capacity Psi and adjustable cost cost of each distributed energy resource; S2: The virtual power plant receives grid dispatching requirements and establishes an internal equilibrium dispatching model for the virtual power plant; S3: Using dynamic optimization algorithms, calculate the scheduling strategies for different types of distributed energy within the virtual power plant; S4: The scheduling strategies of different types of distributed energy resources calculated in step S3 are introduced into the internal balanced scheduling model of the virtual power plant established in step S2, thereby scheduling all distributed energy resources in the virtual power plant; The dynamic optimization algorithm in S3 is the breadth uniform algorithm. The specific steps include: the following: S3.1: Input initialization parameters, population size initial size N, minimum size NM, merging distance r', core particle set KL size N', maximum number of iterations TL, TL is initially set to an integer between 500 and 1000, particle mass range [m1, m2], initially set to [1, 25]; Among them, any particle is represented as: on i ={P1,...P c ...P C } Where, P c is the power emitted by the cth resource, P C is the power generated by the Cth resource, pa i is the i-th particle, and its total number is N; According to the constraints, pa i ={P1,...P c ...P C }, must satisfy 0 <P c ≤P si , randomly generate an initialization particle in the solution space, a second-layer particle, with the initial particle as the center, randomly generate two particles in a sphere with a radius of 3 times r'. Similarly, three-layer self-particles are further generated with the second-layer particle as the center. Finally, an initial swarm S of size N is generated. Each particle in the swarm is assigned a random value in the particle mass range [m1, m2]. Set the initial iteration count t = 1; S3.2: Calculate the dominance relationship of particles in the initial population S based on the Pareto relation, create a new core particle set KL, and copy the non-inferior solution to the core particle set KL; S3.3: Traverse all particles and calculate the Pareto dominance relationship between them. If the particle pa i and pa j There is a dominance relationship between them. From the core particle set KL, the particle g* with the largest gravitational force is selected, and the dominated particle pa i and pa j , according to the following formula, move towards the core solution and the dominant solution, where g* is taken as the core solution; Where: pa i (t) is the vector representation of particle i at time t, pa j (t) is the vector representation of particle j at time t, is the vector representation of particle g* at time t, is the Euclidean distance between particle i and particle g*; Otherwise, select g* with the largest gravitational force from the core particle set KL, and the non-dominated particle pa i and pa j , move according to the following formula; Among them, α is a random value (0, 1), G * is the mass gravitational coefficient, R ij is the Euclidean distance between particles, m is the particle mass, and θ is the random position, which ensures that the solution particles try to explore more space while moving; m i is the mass of particle i, m j is the mass of particle j, is the mass of the g* particle; S3.4: Calculate the distance between particles. If the distance between particles is close enough to be smaller than the merging distance and the number of particles is greater than the minimum size NM, the particles are merged. Update the mass of the merged particles according to the following formula; m new =m i +m j Among them, m new is the mass of the new particle after merging, m i is the mass of particle i, m j is the mass of particle j, otherwise, proceed to the next step; S3.5: Update the particle positions, calculate the dominance relationship of particles in the group based on the Pareto relationship, and update the core particle set based on the non-inferior solution sorting; S3.6: If the number of iterations is less than the maximum number of iterations TL, return to step 3; otherwise, end the iteration and proceed to the next step. S3.7: Output the solution in the core particle set, and the output solution is a candidate solution for the optimal solution.

2. The method for distributed energy resources to participate in virtual power plant scheduling according to claim 1, characterized in that: The overall objective function of the internal equilibrium scheduling model in S2 is: Among them, F cost represents the total cost of each distributed energy participating in the regulation of the virtual power plant; c is the total number of distributed energy within the virtual power plant, which is a fixed constant, cost i is the cost of distributed generation resource i, which is a constant, and cost i >0; The overall objective function is the minimum value of the overall cost: min(F cost ).

3. The method for distributed energy resources to participate in virtual power plant scheduling according to claim 2, characterized in that: The internal equilibrium scheduling model also includes another objective function: the green distributed energy output power of the virtual power plant is maximized, that is, Among them, P clean is the total output of green distributed energy from the virtual power plant, n is the number of green distributed energy from the virtual power plant, is the output power of the jth green distributed energy.

4. The method for distributed energy resources to participate in virtual power plant scheduling according to claim 3 is characterized in that: The constraints calculated by the internal balanced scheduling model include: 1) System power balance constraints: c is the total number of distributed energy resources within the virtual power plant, which is a fixed constant. c is the power emitted by the cth resource; and satisfies p out <p s , the p s is the adjustable capacity of the virtual power plant, is the maximum value that can be output from the virtual power plant, and is a limiting constant; 2) Distributed power generation resource output constraints: P c It must be within the range of its dispatchable power, that is: 0<P c ≤P si Among them, P si The maximum adjustable value of each distributed power source is the maximum rated value under safe operating conditions.

5. The method for distributed energy resources to participate in virtual power plant scheduling according to claim 4 is characterized in that: In S4, an optimal candidate solution is selected and brought into the objective function, i.e., the minimum overall cost and the maximum output power of green distributed energy, to obtain the final scheduling strategy.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it is based on the method for distributed energy to participate in virtual power plant scheduling according to any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it is based on the method for distributed energy to participate in virtual power plant scheduling according to any one of claims 1 to 5.

Images