A virtual power plant cloud-edge collaborative acceleration method based on perturbation function

Abstract

The application provides a virtual power plant cloud edge collaborative acceleration method based on a disturbance function, comprising the following steps: constructing a centralized optimal scheduling model of a virtual power plant according to power supply capacity coupling constraints of a power distribution network; adopting a cloud edge collaborative algorithm to decompose the centralized optimal scheduling model of the virtual power plant into decentralized optimal scheduling models of the virtual power plant and each distributed energy subsystem, wherein the virtual power plant and each distributed energy subsystem realize decomposition and coordination by exchanging boundary information; and destroying the symmetry of specific parameters in the decentralized optimal scheduling models within a preset error range by adopting a disturbance function subject to a Gaussian distribution, so as to accelerate the convergence speed of the cloud edge collaborative algorithm. The method provided by the application provides theoretical guidance for improving the cloud edge collaborative efficiency of the virtual power plant.

Description

Technical Field

[0001] This invention belongs to the field of distributed energy technology. Background Technology

[0002] To build a clean, low-carbon, flexible, and efficient energy system, flexible resources with multiple time scales, such as distributed power sources and electric vehicles, have become crucial in addressing the energy transition. However, due to the influence of time and space factors, their large-scale integration also poses challenges to the stable operation of the distribution network. Since centralized control of various flexible resources results in low computational efficiency and high communication pressure, decentralized management is necessary. Cloud-edge collaborative computing is a distributed computing architecture. However, oscillations may occur during the iteration process of cloud-edge collaborative computing algorithms, significantly reducing computational efficiency. Therefore, how to accelerate the convergence speed of cloud-edge collaborative computing remains to be studied.

[0003] One of the main causes of oscillations is that parameter symmetry can cause the optimal solutions to the same subproblems to fall in the same polar direction, making it difficult for the solver to select the appropriate solution and thus leading to oscillations. Therefore, breaking parameter symmetry while ensuring the optimal solution remains unchanged, and accelerating the cloud-edge collaboration of virtual power plants, is an urgent problem to be solved. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the purpose of this invention is to propose a virtual power plant cloud-edge collaboration acceleration method based on a perturbation function, which can be used to accelerate the cloud-edge collaboration speed of virtual power plants.

[0006] To achieve the above objectives, a first aspect of the present invention proposes a cloud-edge collaborative acceleration method for virtual power plants based on a perturbation function, comprising:

[0007] A centralized optimization scheduling model for virtual power plants is constructed based on the coupling constraints of distribution network supply capacity.

[0008] The cloud-edge collaborative algorithm is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem. The virtual power plant and each distributed energy subsystem achieve decomposition and coordination by exchanging boundary information.

[0009] By employing a perturbation function that follows a Gaussian distribution to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range, the convergence speed of the cloud-edge collaborative algorithm is accelerated.

[0010] In addition, the virtual power plant cloud-edge collaborative acceleration method based on a perturbation function according to the above embodiments of the present invention may also have the following additional technical features:

[0011] Furthermore, in one embodiment of the present invention, the step of constructing a centralized optimization scheduling model for virtual power plants based on the coupling constraints of distribution network supply capacity includes:

[0012] Construct the centralized optimization scheduling model for the virtual power plant:

[0013]

[0014] Where P represents the net load vector set of the distributed energy subsystem, P = [p s,1 ,p s,2 ,p s,3 ,…,p s,N N is the number of energy subsystems; T is the deadline; T s It is the start time; C t This indicates the real-time electricity price for each time period; It is the net load of the i-th energy subsystem during time period t; It represents the energy demand of energy subsystem i during time period t; It is the total charging power of the charging station corresponding to energy subsystem i during time period t; It is the total discharge power of the charging station corresponding to energy subsystem i during time period t; It is the actual power of the distributed photovoltaic cells configured in energy subsystem i during time period t;

[0015] Consider the coupling constraints of the power distribution network supply capacity:

[0016]

[0017] CAP represents the power supply capacity of the power distribution network.

[0018] Furthermore, in one embodiment of the present invention, the step of decomposing the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm includes:

[0019] The objective function of the virtual power plant is expressed as:

[0020]

[0021] After dualization and simplification, we get:

[0022]

[0023] Where D(Λ) is the expression for the dual function of the Lagrange function; It is the minimum actual cost of energy subsystem i during time period t.

[0024] Furthermore, in one embodiment of the present invention, the step of decomposing the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm includes:

[0025] The objective function of each energy subsystem is expressed as:

[0026]

[0027] in, Z is the minimum actual cost of energy subsystem i during time period t; i It is the set of constraints related to energy subsystem i;

[0028] The photovoltaic output constraints of each energy subsystem are as follows:

[0029]

[0030] in, This is the maximum power generation of distributed photovoltaic systems;

[0031] The time constraints for electric vehicles entering and exiting stations in each of the energy subsystems are as follows:

[0032]

[0033] Among them, t s / e These are the entry and exit times of electric vehicles; t1 and t2 specify the time range for electric vehicles entering and exiting the station; μ s / e σ is the average time for electric vehicles to enter / exit the station; s / e It is the standard deviation of the time for electric vehicles to enter and exit the station;

[0034] The state of charge constraints for electric vehicles in each energy subsystem are as follows:

[0035]

[0036] Among them, SOC v (t) represents the state of charge of the electric vehicle at time t; For the arrival time of the EV, η is the SOC value at the arrival time; EV For the battery's charge and discharge efficiency; For EV departure time; SOCI is the minimum SOC value allowed at departure time; This represents the upper limit of the state of charge;

[0037] The charging and discharging power constraints of electric vehicles in each energy subsystem are as follows:

[0038]

[0039] in, and These are the maximum charging and discharging power of electric vehicles; and These are the charging and discharging power of electric vehicles;

[0040] The power constraints of the charging stations in each energy subsystem are as follows:

[0041]

[0042] Where, Φ EV ={1,2,...,N EV} represents the set of EVs, N EV The number of EVs.

[0043] The charging / discharging power limits for the charging stations in each energy subsystem are as follows:

[0044]

[0045] in, This indicates the maximum allowable power limit for charging and discharging at the charging station.

[0046] Furthermore, in one embodiment of the present invention, the virtual power plant and each distributed energy subsystem achieve decomposition and coordination by exchanging boundary information, including:

[0047] Initialize the Lagrange multipliers and iteration step size parameters;

[0048] By optimizing and solving the sub-problems derived from the decomposition of each energy subsystem, the output of each distributed energy source can be obtained.

[0049] Information exchange takes place between the virtual power plant and each energy subsystem, and the information includes the output status of each distributed energy source.

[0050] Based on the information, the updated objective function is obtained by updating the iteration step size parameter and Lagrange multiplier through a virtual power plant.

[0051] The updated objective function is determined based on the convergence condition. If the convergence condition is met, the algorithm terminates; otherwise, the iteration continues.

[0052] Furthermore, in one embodiment of the present invention, the step of accelerating the convergence speed of the cloud-edge collaborative algorithm by using a perturbation function that follows a Gaussian distribution to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range includes:

[0053] Select the parameters of the perturbation function to be applied;

[0054] The upper limit of the perturbation range that guarantees the optimal solution remains unchanged is obtained through mathematical derivation, and the perturbation amplitude is selected within the range of values. The parameters of the perturbation function to be perturbed are then perturbed according to the selected perturbation amplitude.

[0055] The perturbated parameters are then input into the models of each distributed energy subsystem, and the iterative solution process begins.

[0056] To achieve the above objectives, a second aspect of the present invention proposes a virtual power plant cloud-edge collaborative acceleration device based on a perturbation function, comprising the following modules:

[0057] The module is used to construct a centralized optimization scheduling model for virtual power plants based on the coupling constraints of the distribution network supply capacity.

[0058] The decomposition module is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm. The virtual power plant and each distributed energy subsystem achieve decomposition coordination by exchanging boundary information.

[0059] The perturbation module is used to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range by employing a perturbation function that follows a Gaussian distribution, thereby accelerating the convergence speed of the cloud-edge collaborative algorithm.

[0060] To achieve the above objectives, a third aspect of the present invention provides a computer device, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a virtual power plant cloud-edge collaborative acceleration method based on a perturbation function as described above.

[0061] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements a virtual power plant cloud-edge collaborative acceleration method based on a perturbation function as described above.

[0062] The proposed virtual power plant cloud-edge collaborative acceleration method based on perturbation function in this invention constructs a centralized optimization scheduling model for virtual power plants. Considering the coupling constraints of distribution network supply capacity, the Lagrange relaxation method is used to decompose and coordinate the distributed energy subsystems in the distribution network. The virtual power plant and each energy subsystem perform self-optimization and self-decision-making by exchanging boundary information. Finally, a perturbation function following a Gaussian distribution is used to destroy parameter symmetry within the allowable error range, thereby accelerating the convergence speed of cloud-edge collaborative computing and providing theoretical guidance for improving the cloud-edge collaborative efficiency of virtual power plants. Attached Figure Description

[0063] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0064] Figure 1 This is a flowchart illustrating a cloud-edge collaborative acceleration method for virtual power plants based on a perturbation function, provided in an embodiment of the present invention.

[0065] Figure 2 This is a schematic diagram of a virtual power plant cloud-edge collaborative process based on a perturbation function, provided as an embodiment of the present invention.

[0066] Figure 3 This is a schematic diagram of a cloud-edge collaborative acceleration device for a virtual power plant based on a perturbation function, provided in an embodiment of the present invention. Detailed Implementation

[0067] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0068] The following describes an embodiment of the present invention, a cloud-edge collaborative acceleration method for virtual power plants based on perturbation functions, with reference to the accompanying drawings.

[0069] Example 1

[0070] Figure 1 This is a flowchart illustrating a cloud-edge collaborative acceleration method for virtual power plants based on a perturbation function, provided in an embodiment of the present invention.

[0071] like Figure 1 As shown, the virtual power plant cloud-edge collaborative acceleration method based on perturbation function includes the following steps:

[0072] S101: Construct a centralized optimization scheduling model for virtual power plants based on the coupling constraints of distribution network supply capacity;

[0073] S102: The cloud-edge collaborative algorithm is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem. The virtual power plant and each distributed energy subsystem achieve decomposition and coordination by exchanging boundary information.

[0074] S103: By using a perturbation function that follows a Gaussian distribution to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range, the convergence speed of the cloud-edge collaborative algorithm is accelerated.

[0075] Furthermore, in one embodiment of the present invention, a centralized optimization scheduling model for virtual power plants is constructed based on the coupling constraints of distribution network supply capacity, including:

[0076] Constructing a centralized optimization scheduling model for virtual power plants:

[0077]

[0078] Where P represents the net load vector set of the distributed energy subsystem, P = [p s,1 ,p s,2 ,p s,3 ,…,p s,N N is the number of energy subsystems; T is the deadline; T s It is the start time; C t This indicates the real-time electricity price for each time period; It is the net load of the i-th energy subsystem during time period t; It represents the energy demand of energy subsystem i during time period t; It is the total charging power of the charging station corresponding to energy subsystem i during time period t; It is the total discharge power of the charging station corresponding to energy subsystem i during time period t; It is the actual power of the distributed photovoltaic cells configured in energy subsystem i during time period t;

[0079] Considering the coupling constraints of distribution network supply capacity:

[0080]

[0081] CAP stands for Distribution Network Supply Capacity.

[0082] Furthermore, in one embodiment of the present invention, a cloud-edge collaborative algorithm is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem, including:

[0083] The objective function of the virtual power plant is expressed as:

[0084]

[0085] After dualization and simplification, we get:

[0086]

[0087] Where D(Λ) is the expression for the dual function of the Lagrange function; It is the minimum actual cost of energy subsystem i during time period t.

[0088] Furthermore, in one embodiment of the present invention, a cloud-edge collaborative algorithm is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem, including:

[0089] The objective function of each energy subsystem is expressed as:

[0090]

[0091] in, Z is the minimum actual cost of energy subsystem i during time period t; i It is the set of constraints related to energy subsystem i;

[0092] The photovoltaic output constraints in each energy subsystem are as follows:

[0093]

[0094] in, This is the maximum power generation of distributed photovoltaic systems;

[0095] The time constraints for electric vehicles entering and exiting stations in each energy subsystem are as follows:

[0096]

[0097] Among them, t s / e These are the entry and exit times of electric vehicles; t1 and t2 specify the time range for electric vehicles entering and exiting the station; μ s / e σ is the average time for electric vehicles to enter / exit the station; s / e It is the standard deviation of the time for electric vehicles to enter and exit the station;

[0098] The state of charge constraints for electric vehicles in each energy subsystem are as follows:

[0099]

[0100] Among them, SOC v (t) represents the state of charge of the electric vehicle at time t; For the arrival time of the EV, η is the SOC value at the arrival time; EV For the battery's charge and discharge efficiency; For EV departure time; SOCI is the minimum SOC value allowed at departure time; This represents the upper limit of the state of charge;

[0101] The charging and discharging power constraints for electric vehicles in each energy subsystem are as follows:

[0102]

[0103] in, and These are the maximum charging and discharging power of electric vehicles; and These are the charging and discharging power of electric vehicles;

[0104] The power constraints of charging stations in each energy subsystem are as follows:

[0105]

[0106] Where, Φ EV ={1,2,...,N EV} represents the set of EVs, N EV The number of EVs.

[0107] The charging / discharging power limits for the charging stations in each energy subsystem are as follows:

[0108]

[0109] in, This indicates the maximum allowable power limit for charging and discharging at the charging station.

[0110] Furthermore, in one embodiment of the present invention, the virtual power plant and each distributed energy subsystem achieve decomposition and coordination by exchanging boundary information, including:

[0111] Initialize the Lagrange multipliers and iteration step size parameters;

[0112] By optimizing and solving the sub-problems derived from the decomposition of each energy subsystem, the output of each distributed energy source can be obtained.

[0113] Information is exchanged between the virtual power plant and various energy subsystems, including the output status of each distributed energy source;

[0114] Based on the information, the updated objective function is obtained by updating the iteration step size parameters and Lagrange multipliers using a virtual power plant.

[0115] The updated objective function is determined based on the convergence condition. If the convergence condition is met, the algorithm terminates; otherwise, the iteration continues.

[0116] Furthermore, in one embodiment of the present invention, by employing a perturbation function that follows a Gaussian distribution to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range, the convergence speed of the cloud-edge collaborative algorithm is accelerated, including:

[0117] Select the parameters of the perturbation function to be applied;

[0118] The upper limit of the perturbation range that guarantees the optimal solution remains unchanged is obtained through mathematical derivation, and the perturbation amplitude is selected within the range of values. The parameters of the perturbation function to be perturbed are then perturbed according to the selected perturbation amplitude.

[0119] The perturbated parameters are then input into the models of each distributed energy subsystem, and the iterative solution process begins.

[0120] Example 2

[0121] like Figure 2 The diagram shown is a schematic of a virtual power plant cloud-edge collaborative process based on a perturbation function proposed in an embodiment of the present invention.

[0122] See Figure 2 The process includes:

[0123] 1) Taking into account the physical connections between various flexible resources, a centralized optimization scheduling model for virtual power plants containing multiple flexible resources is established, taking into account the coupling constraints of distribution network supply capacity.

[0124] 1-1) Construct the overall system objective function that includes resources with varying flexibility, as shown in the following expression:

[0125]

[0126] Where P represents the net load vector set of the distributed energy subsystem, P = [p s,1 ,p s,2 ,p s,3 ,…,p s,N N is the number of energy subsystems; T is the deadline; T s It is the start time; C t This indicates the real-time electricity price for each time period; Let be the net load of the i-th energy subsystem during time period t, and its specific expression is as follows:

[0127]

[0128] in, It is the net load of the i-th energy subsystem during time period t; It represents the energy demand of energy subsystem i during time period t; It is the total charging power of the charging station corresponding to energy subsystem i during time period t; It is the total discharge power of the charging station corresponding to energy subsystem i during time period t; It is the actual power of the distributed photovoltaic cells configured in energy subsystem i during time period t.

[0129] 1-2) Couple various flexible resources in the distribution network together as a coupling constraint on the distribution network supply capacity, as shown in the following expression:

[0130]

[0131] Where N is the number of energy subsystems; T is the cutoff time; and CAP is the supply capacity of the distribution network.

[0132] 2) The Lagrange relaxation method is used to decompose and coordinate the centralized optimization scheduling model of the virtual power plant to obtain the virtual power plant operation model.

[0133] The objective function for constructing the virtual power plant operation model is expressed as follows:

[0134]

[0135] Where, λ t Let be the Lagrange multiplier at time t.

[0136] After dualization and simplification, we get:

[0137]

[0138] Where D(Λ) is the expression for the dual function of the Lagrange function; It is the minimum actual cost of energy subsystem i during time period t.

[0139] 3) The Lagrange relaxation algorithm is used to decompose and coordinate the centralized optimization scheduling model of the virtual power plant to obtain the operation model of each energy subsystem.

[0140] 3-1) The objective function for constructing the energy subsystem operation model is as follows:

[0141]

[0142] in, Z is the minimum actual cost of energy subsystem i during time period t; i It is a set of constraints related to energy subsystem i.

[0143] 3-2) Construct the constraints of the energy subsystem, the expressions of which are as follows:

[0144] 3-2-1) Photovoltaic output constraints:

[0145]

[0146] in, This is the maximum power generation of distributed photovoltaic systems.

[0147] 3-2-2) Time constraints for electric vehicles entering and exiting the station:

[0148]

[0149] Among them, ts / e These are the entry and exit times of electric vehicles; t1 and t2 specify the time range for electric vehicles entering and exiting the station; μ s / e σ is the average time for electric vehicles to enter / exit the station; s / e It is the standard deviation of the time for electric vehicles to enter and exit the station.

[0150] 3-2-3) Charge state constraints for electric vehicles:

[0151]

[0152] Among them, SOC v (t) represents the state of charge of the electric vehicle at time t; For the arrival time of the EV, η is the SOC value at the arrival time; EV For the battery's charge and discharge efficiency; For EV departure time; SOCI is the minimum SOC value allowed at departure time; This represents the upper limit of the state of charge.

[0153] 3-2-4) Electric vehicle charging and discharging power constraints:

[0154]

[0155] in, and These are the maximum charging and discharging power of electric vehicles; and These are the charging and discharging power of the electric vehicle.

[0156] 3-2-5) Charging station power constraints:

[0157]

[0158] Where, Φ EV ={1,2,...,N EV} represents the set of EVs, N EV The number of EVs.

[0159] 3-2-6) Charging / discharging power limitations of charging stations:

[0160]

[0161] in, This indicates the maximum allowable power limit for charging and discharging at the charging station.

[0162] 4) Solve the coordination and scheduling model of the virtual power plant and various energy subsystems using the Lagrange relaxation method, including:

[0163] 4-1) Initialize the Lagrange multipliers and iteration step size parameters;

[0164] 4-2) Each energy subsystem optimizes and solves the sub-problems derived from the decomposition to obtain the output of each distributed energy source;

[0165] 4-3) The virtual power plant interacts with various energy subsystems;

[0166] 4-4) Virtual power plant update iteration step size and Lagrange multipliers;

[0167] 4-5) Determine the convergence condition. If the convergence condition is met, terminate the algorithm. If the convergence condition is not met, continue iterating and return to step 4-2).

[0168] 5) Employing a Gaussian-distributed perturbation function to disrupt parameter symmetry within the allowable error range accelerates algorithm convergence, including:

[0169] 5-1) Select the parameters of the perturbation function;

[0170] 5-2) Calculate the upper limit of the perturbation range that guarantees the optimal solution remains unchanged based on mathematical derivation, and select the perturbation amplitude within the range of values;

[0171] 5-3) Substitute the perturbed parameters into the models of each energy subsystem and begin iteratively solving according to 4).

[0172] By applying the Lagrange relaxation method to the centralized optimization scheduling model of virtual power plants, the decomposition and coordination between virtual power plants and various distributed energy subsystems are finally realized. Self-optimization and self-decision-making are carried out by exchanging boundary information. Finally, within the allowable error range, a perturbation function following a Gaussian distribution is used to destroy the parameter symmetry in the model, thereby accelerating the convergence speed of cloud-edge collaborative computing and providing theoretical guidance for improving the cloud-edge collaborative efficiency of virtual power plants.

[0173] The proposed virtual power plant cloud-edge collaborative acceleration method based on perturbation function in this invention constructs a centralized optimization scheduling model for virtual power plants. Considering the coupling constraints of distribution network supply capacity, the Lagrange relaxation method is used to decompose and coordinate the distributed energy subsystems in the distribution network. The virtual power plant and each energy subsystem perform self-optimization and self-decision-making by exchanging boundary information. Finally, a perturbation function following a Gaussian distribution is used to destroy parameter symmetry within the allowable error range, thereby accelerating the convergence speed of cloud-edge collaborative computing and providing theoretical guidance for improving the cloud-edge collaborative efficiency of virtual power plants.

[0174] To achieve the above embodiments, the present invention also proposes a virtual power plant cloud-edge collaborative acceleration device based on a perturbation function.

[0175] Figure 3This is a schematic diagram of a virtual power plant cloud-edge collaborative acceleration device based on a perturbation function, provided as an embodiment of the present invention.

[0176] like Figure 3 As shown, the virtual power plant cloud-edge collaborative acceleration device based on perturbation function includes: a construction module 100, a decomposition module 200, and a perturbation module 300, wherein,

[0177] The module is used to construct a centralized optimization scheduling model for virtual power plants based on the coupling constraints of the distribution network supply capacity.

[0178] The decomposition module is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm. The virtual power plant and each distributed energy subsystem achieve decomposition coordination by exchanging boundary information.

[0179] The perturbation module is used to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range by employing a perturbation function that follows a Gaussian distribution, thereby accelerating the convergence speed of the cloud-edge collaborative algorithm.

[0180] To achieve the above objectives, a third aspect of the present invention provides a computer device, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the virtual power plant cloud-edge collaborative acceleration method based on a perturbation function as described above.

[0181] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the virtual power plant cloud-edge collaborative acceleration method based on a perturbation function as described above.

[0182] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0183] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0184] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A cloud-edge collaborative acceleration method for virtual power plants based on perturbation functions, characterized in that, Includes the following steps: A centralized optimization scheduling model for virtual power plants is constructed based on the coupling constraints of distribution network supply capacity. The cloud-edge collaborative algorithm is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem. The virtual power plant and each distributed energy subsystem achieve decomposition and coordination by exchanging boundary information. By employing a perturbation function that follows a Gaussian distribution to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range, the convergence speed of the cloud-edge collaborative algorithm is accelerated. Specifically, the cloud-edge collaborative algorithm decomposes the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model for the virtual power plant and each distributed energy subsystem, including: The objective function of the virtual power plant is expressed as: ， After dualization and simplification, we get: ， in, Let be the expression for the dual function of the Lagrange function; It is the energy subsystem i in The minimum actual cost for a given period; The objective function of each energy subsystem is expressed as: ， in, It is the energy subsystem i in The minimum actual cost for a given period; It is a set of constraints related to energy subsystem i.

2. The method according to claim 1, characterized in that, The construction of a centralized optimization scheduling model for virtual power plants based on the coupling constraints of distribution network supply capacity includes: Construct the centralized optimization scheduling model for the virtual power plant: ， in, This represents the set of net load vectors for the distributed energy subsystem. N is the number of energy subsystems; T is the deadline. It is the beginning moment; This indicates the real-time electricity price for each time period; It is the net load of the i-th energy subsystem during time period t; It represents the energy demand of energy subsystem i during time period t; It is the total charging power of the charging station corresponding to energy subsystem i during time period t; It is the total discharge power of the charging station corresponding to energy subsystem i during time period t; The distributed photovoltaic cells configured in energy subsystem i are in Actual power during the time period; Consider the coupling constraints of the power distribution network supply capacity: in, This refers to the power supply capacity of the power distribution network.

3. The method according to claim 1, characterized in that, The method of decomposing the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm also includes: The photovoltaic output constraints of each energy subsystem are as follows: ， in, This is the maximum power generation of distributed photovoltaic systems; The time constraints for electric vehicles entering and exiting stations in each of the energy subsystems are as follows: in, This refers to the entry and exit times of electric vehicles at the station; and The time range for electric vehicles to enter and exit the station has been clarified; It is the average time for electric vehicles to enter / exit the station; It is the standard deviation of the time for electric vehicles to enter and exit the station; The state of charge constraints for electric vehicles in each energy subsystem are as follows: in, It represents the state of charge of the electric vehicle at time t; For the arrival time of the EV, The SOC value at the arrival time; For the battery's charge and discharge efficiency; For the departure time of the EV; It is the minimum SOC value allowed when leaving the station; This represents the upper limit of the state of charge; The charging and discharging power constraints of electric vehicles in each energy subsystem are as follows: in, and These are the maximum charging and discharging power of electric vehicles; and These are the charging and discharging power of electric vehicles; The power constraints of the charging stations in each energy subsystem are as follows: in, Represents the set of EVs. The number of EVs; The charging / discharging power limits for the charging stations in each energy subsystem are as follows: in, This indicates the maximum allowable power limit for charging and discharging at the charging station.

4. The method according to claim 1, characterized in that, The virtual power plant and each distributed energy subsystem achieve decomposition and coordination through the exchange of boundary information, including: Initialize the Lagrange multipliers and iteration step size parameters; By optimizing and solving the sub-problems derived from the decomposition of each energy subsystem, the output of each distributed energy source can be obtained. Information exchange takes place between the virtual power plant and each energy subsystem, and the information includes the output status of each distributed energy source. Based on the information, the updated objective function is obtained by updating the iteration step size parameter and Lagrange multiplier through a virtual power plant. The updated objective function is determined based on the convergence condition. If the convergence condition is met, the algorithm terminates; otherwise, the iteration continues.

5. The method according to claim 1, characterized in that, The method of accelerating the convergence speed of the cloud-edge collaborative algorithm by using a Gaussian-distributed perturbation function to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range includes: Select the parameters of the perturbation function to be applied; The upper limit of the perturbation range that guarantees the optimal solution remains unchanged is obtained through mathematical derivation, and the perturbation amplitude is selected within the range of values. The parameters of the perturbation function to be perturbed are then perturbed according to the selected perturbation amplitude. The perturbated parameters are then input into the models of each distributed energy subsystem, and the iterative solution process begins.

6. A virtual power plant cloud-edge collaborative acceleration device based on a perturbation function, characterized in that, Includes the following modules: The module is used to construct a centralized optimization scheduling model for virtual power plants based on the coupling constraints of the distribution network supply capacity. The decomposition module is used to decompose the centralized optimization scheduling model of the virtual power plant into a decentralized optimization scheduling model of the virtual power plant and each distributed energy subsystem using a cloud-edge collaborative algorithm. The virtual power plant and each distributed energy subsystem achieve decomposition coordination by exchanging boundary information. The perturbation module is used to disrupt the symmetry of specific parameters in the distributed optimization scheduling model within a preset error range by employing a perturbation function that follows a Gaussian distribution, thereby accelerating the convergence speed of the cloud-edge collaborative algorithm. The decomposition module is specifically used for: The objective function of the virtual power plant is expressed as: ， After dualization and simplification, we get: ， in, Let be the expression for the dual function of the Lagrange function; It is the energy subsystem i in The minimum actual cost for a given period; The objective function of each energy subsystem is expressed as: ， in, It is the energy subsystem i in The minimum actual cost for a given period; It is a set of constraints related to energy subsystem i.

7. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the virtual power plant cloud-edge collaborative acceleration method based on any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the virtual power plant cloud-edge collaborative acceleration method based on any one of claims 1-5.

Images