Virtual power plant optimization scheduling method and system

Abstract

The invention relates to a virtual power plant optimization scheduling method and system. The method comprises the following steps: establishing an operation model of a multi-type adjustable resource main body; constructing a master-slave game model of the aggregator and the resource main body, wherein the master-slave game model comprises an upper-layer model taking the aggregator as a leader and a lower-layer model taking the resource main body as a follower; solving a master-slave game model of the aggregator and the resource main body to obtain an optimal compensation price strategy and an optimal response power strategy, and constructing an optimal power scheduling plan according to the optimal response power strategy; the optimal power scheduling plan is used for being issued to each adjustable resource main body for execution so as to realize optimal scheduling of the virtual power plant. According to the method, the participation enthusiasm of the resource main body can be stimulated, the user participation degree can be improved, the scheduling instruction execution rate can be improved, the model accuracy can be improved, the cooperative capability of heterogeneous resources can be improved, and the overall operation economy and stability of the virtual power plant can be improved.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and control technology, specifically relating to a virtual power plant optimization scheduling method and system, and particularly an optimization scheduling method and system that considers the two-way game relationship between aggregators and multiple types of adjustable resource entities. Background Technology

[0002] With the deepening of energy transition, the penetration rate of distributed energy in distribution networks is constantly increasing. Virtual Power Plants (VPPs), as an important technological means to aggregate distributed resources and participate in grid dispatch and market transactions, have received widespread attention. The core participants in virtual power plants include aggregators and various adjustable resource entities, including but not limited to distributed photovoltaics, energy storage, and flexible loads.

[0003] Currently, most virtual power plant optimization scheduling methods are centralized, where aggregators act as decision-making centers, issuing scheduling instructions to resource entities with the goal of minimizing their own costs or maximizing their profits. This method has significant drawbacks: First, it ignores the autonomy of resource entities: treating resource entities as passive recipients of instructions ignores their decision-making power and benefit demands as independent stakeholders, resulting in low user participation and difficulty in executing scheduling instructions.

[0004] Second, the model is out of touch with reality: In the electricity market environment, aggregators and resource entities are essentially in a game of interests. Centralized optimization cannot capture this inherent market mechanism, resulting in insufficient model accuracy.

[0005] Third, there is insufficient handling of resource heterogeneity: Distributed photovoltaic, energy storage, air conditioning load, electric vehicles and other resources differ greatly in their operating characteristics, constraints and regulation capabilities. Existing methods are difficult to effectively handle the collaborative optimization problem of these heterogeneous resources within a unified framework.

[0006] Therefore, there is an urgent need for a new method for virtual power plant scheduling that can effectively characterize the interactive game relationship between aggregators and resource entities, and can coordinate and optimize various heterogeneous resources. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a virtual power plant optimization scheduling method and system. By establishing a two-way game model, it minimizes the operating costs of aggregators while ensuring reasonable returns for resource stakeholders. This stimulates the participation of resource stakeholders, increases user engagement, improves the execution rate of scheduling instructions, enhances model accuracy, and improves the collaborative capabilities of heterogeneous resources, thereby improving the overall operational economy and stability of the virtual power plant.

[0008] The present invention adopts the following technical solution.

[0009] A first aspect of the present invention provides a virtual power plant optimized scheduling method, comprising: Establish an operational model for multiple types of adjustable resource entities; A master-slave game model is constructed between aggregators and resource entities, comprising an upper-level model with aggregators as leaders and a lower-level model with resource entities as followers. In the upper-level model, the aggregator aims to minimize total operating costs and makes decisions under pre-defined constraints, resulting in a compensation price strategy issued to various resource entities. In the lower-level model, each resource entity makes decisions based on an established operating model, aiming to maximize its own revenue, resulting in its own response power strategy. Resource entities decide on their power response strategies based on the compensation price strategies, and these power response strategies influence the aggregator's formulation of compensation price strategies, leading to a game between resource entities and aggregators. Solve the master-slave game model between aggregators and resource entities to obtain the optimal compensation price strategy and the optimal response power strategy. Construct the optimal power scheduling plan based on the optimal response power strategy. The optimal power scheduling plan is then distributed to each adjustable resource entity for execution to achieve optimized scheduling of the virtual power plant.

[0010] Optionally, the objective function for the aggregator can be constructed as follows, with the goal of minimizing total operating costs:

[0011] in, This represents the total operating cost of the aggregator. This indicates the price at which electricity is purchased from the higher-level power grid. This represents the amount of electricity purchased from the upper-level power grid at time t. This indicates the compensation price issued by the aggregator to resource entity i. This represents the response power of resource subject i. This represents the cost incurred when the aggregator's total power deviates from the grid command.

[0012] Optionally, the objective function for each resource entity can be constructed with the goal of maximizing the benefits of each resource entity:

[0013] in, This represents the revenue of resource entity i. This represents the response power of resource entity i, including photovoltaic reduction, energy storage regulation power, air conditioning regulation power, and EV regulation power. This represents the cost corresponding to resource entity i.

[0014] Optionally, the master-slave game model for solving the aggregator and resource subject includes: The upper-level particle swarm initialization involves encoding the aggregator's compensation price strategy into particle positions and randomly initializing the particle swarm. For each price strategy represented by a particle, a genetic algorithm is called to solve the problem. With the goal of maximizing the revenue of each resource subject, the optimal response power strategy for all resource subjects under the given price strategy is obtained. The total operating cost corresponding to this pricing strategy is calculated by combining the optimal response power strategies of all resource entities. The particle's velocity and position are updated based on the total operating cost corresponding to the pricing strategy to update the pricing strategy; Repeat the above steps until the particle swarm optimization algorithm converges, and obtain the equilibrium solution of the master-slave game model.

[0015] Optionally, the inertia weight is updated based on the difference between the aggregator's total operating cost at the current moment and the total operating cost at the previous moment. This updated inertia weight is then used in conjunction with the velocity of the new particles, specifically:

[0016] in, Indicates the inertial weight at time t. This represents the maximum value of the inertia weight. This represents the minimum inertial weight. This represents the difference between the total operating cost of the upper-level aggregator at time t and the total operating cost at the previous time. This represents the total operating cost of the upper-level aggregator during the initial iteration. Indicates the scheduling fluctuation coefficient. This represents the adjustment factor, used to control the degree to which the inertia weight changes with total operating costs.

[0017] Optionally, the particle velocity can be updated based on the resource response power, specifically:

[0018] in, This represents the velocity of particle i at time t+1. This represents the velocity of particle i at time t. Indicates the inertial weight at time t. This represents the best-performing price strategy for particle i at time t. This represents the best-performing price strategy in the entire population at time t. This represents the price strategy of particle i at time t. Represents the local learning factor. Represents the global learning factor. , All are random numbers in the range of 0 to 1. This represents the response rate deviation correction factor. This represents the difference between the actual resource response rate and the target resource response rate. The response rate is an indicator for evaluating the actual response effect and is used to reflect whether the total response rate of all resource entities in the virtual power plant has reached the total response rate expected by the aggregator.

[0019] Optionally, resource response sensitivity can be calculated based on the sensitivity of the resource subject's response power to changes in compensation price, and the interval between upper-level and lower-level iterations can be set based on resource response sensitivity to perform hierarchical iteration.

[0020] Optionally, after the upper-level particles are updated, the genetic algorithm is used to solve the response power strategy in this iteration based on the rate of change of the compensation price.

[0021] A second aspect of the present invention provides a virtual power plant optimization scheduling system to implement the above-described virtual power plant optimization scheduling method, the system comprising: The resource creation and management module is used to establish operational models for multiple types of adjustable resource entities; The two-way game optimization calculation module includes a master-slave game construction module and a solution module; among them, A master-slave game theory module is constructed to build a master-slave game model between aggregators and resource entities. This model includes an upper-level model with the aggregator as the leader and a lower-level model with resource entities as followers. In the upper-level model, the aggregator aims to minimize total operating costs and makes decisions under pre-defined constraints, resulting in a compensation price strategy issued to various resource entities. In the lower-level model, each resource entity makes decisions based on an established operating model, aiming to maximize its own revenue, resulting in its own response power strategy. Resource entities then determine their power response strategies based on the compensation price strategies, which in turn influence the aggregator's compensation price strategy, leading to a game between the resource entities and the aggregator. The solution module is used to solve the master-slave game model between aggregators and resource entities to obtain the optimal compensation price strategy and the optimal response power strategy. Based on the optimal response power strategy, the optimal power scheduling plan is constructed. The optimal power scheduling plan is then distributed to each adjustable resource entity for execution to achieve optimized scheduling of the virtual power plant.

[0022] A third aspect of the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded onto the processor, implements the above-described virtual power plant optimization scheduling method.

[0023] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described virtual power plant optimization scheduling method.

[0024] Compared with the prior art, the beneficial effects of the present invention include at least the following: This invention uses the Stackelberg game model to better reflect the actual trading mechanism of the electricity market, and the optimization results are more practically instructive. It also takes into account the interests of aggregators and resource entities, achieving a reasonable distribution of benefits and effectively stimulating the enthusiasm of distributed resources to participate in grid regulation.

[0025] This invention utilizes a lower-level parallel genetic algorithm to effectively handle the significant differences in physical constraints and economic characteristics of resources such as photovoltaics, energy storage, air conditioning, and electric vehicles, thereby enhancing the processing capability of heterogeneous resources and achieving "personalized" management and collaborative optimization of resources.

[0026] The hybrid PSO-GA algorithm designed in this invention combines the global search capability of PSO with the advantages of GA in handling complex constraints. It can efficiently and stably solve complex bi-level optimization problems, ensuring the practical potential of the method. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a schematic diagram of a hybrid PSO-GA algorithm solution process provided in an embodiment of the present invention; Figure 2 This is a schematic block diagram of a virtual power plant optimization scheduling system provided in an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0029] S1: Establish an operational model for multiple types of adjustable resource entities.

[0030] The adjustable resources include at least: distributed photovoltaic (PV), distributed energy storage system (ESS), building air conditioning load (HVAC), and electric vehicle (EV). A power model and operational constraints are established for each type of resource. S1 specifically includes: S1.1: Establish a distributed photovoltaic model, photovoltaic output It depends mainly on the light intensity and can be regarded as a prediction of a given load, with the adjustable amount being the amount of reduction.

[0031] S1.2: Establish a distributed energy storage model, including charging power constraints, discharging power constraints, timing constraints of state of charge, and mutual exclusion constraints for charging and discharging. Specifically, this is expressed as follows:

[0032] in, This represents the state of charge at time t. This represents the state of charge at time t+1. Indicates charging efficiency. Indicates discharge efficiency. Indicates the rated capacity. Represents state variables, This represents the discharge power at time t. This represents the charging power at time t. Indicates the maximum charging power. This indicates the charging state at time t. This indicates the discharge state at time t. Indicates the minimum charge. This indicates the maximum charge.

[0033] S1.3: Establish a building air conditioning load model. Based on the equivalent thermal parameter model, establish the relationship between indoor temperature and air conditioning power, and use the user-set temperature comfort range. As a constraint, it is specifically expressed as follows:

[0034] in, This represents the indoor temperature at time t+1. This represents the indoor temperature at time t. This represents the outdoor temperature at time t. This indicates the minimum indoor temperature value. This indicates the maximum indoor temperature value. This represents the air conditioner power at time t.

[0035] S1.4: Establish an electric vehicle model, based on user travel habits, and establish constraints such as access time, disconnection time, initial SOC (State of Charge, battery remaining percentage), and expected SOC upon departure, thereby determining the adjustable power upper and lower limits within the VPP scheduling cycle. Specifically:

[0036] in, This indicates the remaining battery power of the electric vehicle at the time the user connects to the charging station and begins charging. This indicates the user's offline time, i.e., the time spent away from the charging station. This represents the energy required during the scheduling period. Indicates the rated capacity of the electric vehicle battery. This represents the remaining scheduling time. One scheduling cycle is the offline time minus the access time. The remaining scheduling time is the offline time minus any scheduling moment. This indicates the charging efficiency of electric vehicles. This represents the energy that has been charged at time t. This indicates the lower limit of the adjustable power of an electric vehicle. This indicates the adjustable power limit of an electric vehicle, specifically the maximum charging power allowed by the vehicle's physical hardware and safety regulations. The maximum charging power limited by the physical hardware is the maximum rated power. Safety regulations include reducing the maximum charging power to protect the battery when the State of Charge (SOC) exceeds the SOC threshold. It will decrease as SOC increases.

[0037] In this embodiment, a more accurate model is established to characterize the differences in operating characteristics, constraints, and regulation capabilities of resources such as distributed photovoltaics, energy storage, air conditioning loads, and electric vehicles, thereby enabling more efficient handling of the collaborative optimization problem of these heterogeneous resources within a unified framework.

[0038] S2: Construct a Stackelberg master-slave game model between aggregators and resource entities. This model consists of two layers: an upper-level model with aggregators as leaders and a lower-level model with resource entities as followers. Specifically, it includes: S2.1: In the upper-level model, the aggregator aims to minimize total operating costs and makes decisions under pre-defined constraints, resulting in the aggregator's decision variables. These variables include the power purchased from the upper-level grid and the power purchased from various resource entities. The released compensation price strategy.

[0039] Total operating costs include: the cost of purchasing electricity from the upper-level grid, compensation payments to resource entities, and costs incurred when the aggregated total power deviates from grid commands.

[0040] Construct the aggregator's objective function with the goal of minimizing total operating costs:

[0041] in, This represents the total operating cost of the aggregator. This indicates the price at which electricity is purchased from the higher-level power grid. This represents the amount of electricity purchased from the upper-level power grid at time t. This indicates the compensation price issued by the aggregator to resource entity i. This represents the response power of resource subject i. This represents the cost incurred when the aggregator's total power deviates from the grid command.

[0042] Constraints include, but are not limited to, power balance constraints and market pricing constraints. Power balance constraints are expressed by the following formula:

[0043] in, This represents the amount of electricity purchased from the upper-level power grid at time t. Indicates photovoltaic power generation capacity. Indicates the power of the electrical load. This represents the response power of resource subject i.

[0044] S2.2: In the lower-level model, each resource subject makes decisions based on the operating model established in S1 with the goal of maximizing the benefits of each resource subject, and obtains the response power strategy of each resource subject.

[0045] The resource entity's revenue comes from the compensation it receives in response to the aggregator's scheduling instructions, while its costs consist of its own operating costs and penalty costs.

[0046] Construct objective functions for each resource entity with the goal of maximizing their respective gains:

[0047] in, This represents the revenue of resource entity i. This represents the response power of resource entity i, including photovoltaic reduction, energy storage regulation power, air conditioning regulation power, and EV regulation power. This represents the cost corresponding to resource entity i. The cost includes operating costs and penalty costs. Penalty costs include penalty fees for air conditioning temperatures deviating from user comfort levels and penalty fees for electric vehicles causing inconvenience due to charging.

[0048] S2.3: Upper-level aggregators formulate compensation price strategies. To influence the response behavior of lower-level resource entities Lower-level resource entities follow compensation price strategies. The optimal power response strategy is determined, which in turn affects the cost of the upper layers. This process is iterated until the Stackelberg equilibrium is reached, that is, given the other party's strategy, neither party can increase its revenue by unilaterally changing its own strategy.

[0049] In this embodiment, by effectively characterizing the interactive game relationship between aggregators and resource entities, the inherent market mechanism of the electricity market environment can be more accurately depicted, improving the accuracy of the model, enhancing the decision-making power and benefit demands of resource entities as independent stakeholders, and increasing user participation.

[0050] S3: Solve the master-slave game model constructed in S2 to obtain the optimal response power strategy of the lower-level resource entity. Optimal compensation price strategy for upper-level aggregators An optimal power scheduling plan is constructed based on the optimal response power strategy.

[0051] Furthermore, to address the aforementioned complex bi-level optimization problem, a PSO-GA (Particle Swarm Optimization-Genetic Algorithm) algorithm is designed for solving it, as follows: Figure 1 As shown. The specific steps are as follows: S3.1: Initialize the upper-level PSO, setting the aggregator's compensation price strategy. The particle position is encoded and the particle swarm is randomly initialized.

[0052] S3.2: Lower-level GA solution. For the price strategy represented by each particle, a genetic algorithm is called to solve the lower-level optimization problem. Specifically, the GA calculates the response power strategy for each resource subject. Encoded as chromosomes, and aiming to maximize the first fitness value, selection, crossover, and mutation operations are performed under the constraints in S1 to ultimately obtain the optimal response power strategy for all resource subjects under this price strategy. The first fitness value is the resource subject's benefit. That is, the objective function of the resource subject.

[0053] Furthermore, resource response sensitivity is calculated based on the sensitivity of the resource subject's response power to changes in compensation price. The interval between upper and lower iterations is then set according to this resource response sensitivity to perform hierarchical iteration, specifically: Optionally, when the resource response sensitivity is greater than or equal to the first sensitivity threshold, the response to changes in the compensation price is drastic, and the solution is recalculated after each upper-level iteration; when the first sensitivity threshold is less than the resource response sensitivity and less than the second sensitivity threshold, the solution can be recalculated every N upper-level iterations; when the resource response sensitivity is less than the third sensitivity threshold, the solution can be recalculated every M upper-level iterations, where M is greater than N and both are positive integers.

[0054] In this embodiment, while ensuring the optimization accuracy of the SPO algorithm, redundant calculations in the lower layer iteration can be significantly reduced, improving the overall solution efficiency. At the same time, it can adapt to the response characteristics of different resource subjects and ensure scheduling feasibility.

[0055] Optionally, after the upper-level particles are updated, the genetic algorithm is used to solve the response power strategy in this iteration based on the rate of change of the compensation price.

[0056] Specifically, after the upper-level particles are updated, the price change rate is calculated first. For high-sensitivity resources, the GA solution is called directly to solve the response strategy. For medium or low-sensitivity resources, it is determined whether the price change exceeds the price threshold. If it does not exceed the threshold, the historical best power response strategy is used. If it does exceed the threshold, the GA solution is called to solve the response strategy.

[0057] In this embodiment, redundant computations in the lower-level GA can be reduced, the time for a single PSO iteration can be shortened, ineffective adjustments and local optima can be avoided, and the collaborative solution rate between the upper and lower layers can be improved.

[0058] Resource response sensitivity is calculated using the following formula:

[0059] in, Indicates resource response sensitivity. This represents the rate of change of power in the response of resource subject i. This represents the rate of change in the compensation price for resource entity i.

[0060] S3.3: Upper-layer fitness evaluation, which evaluates the optimal response power strategy returned by the lower layer. Substitute the values ​​into the objective function of the upper layer to calculate the minimum total operating cost corresponding to this pricing strategy. The reciprocal of this value is used as the second fitness value of the PSO particle; the lower the cost, the higher the second fitness.

[0061] S3.4: PSO Update and Iteration. The PSO algorithm updates the velocity and position of particles based on the second fitness value, i.e., updates the price strategy.

[0062] Furthermore, the inertia weight is updated based on the difference between the total operating cost of the upper-layer aggregator (feedback from VPP scheduling) and the total operating cost at the previous moment, and the particle velocity is updated based on the updated inertia weight. Specifically:

[0063] in, Indicates the inertial weight at time t. This represents the maximum value of the inertia weight. This represents the minimum inertial weight. This represents the difference between the total operating cost of the upper-level aggregator at time t and the total operating cost at the previous time. This represents the total operating cost of the upper-level aggregator during the initial iteration. This represents the dispatch fluctuation coefficient, which is dynamically adjusted based on the grid electricity price volatility. It is increased when grid electricity price fluctuations exceed a preset threshold (e.g., set to 1.2), and decreased when grid electricity price fluctuations are less than the preset threshold (e.g., set to 0.8). The value of k is used to control the degree to which the inertia weight changes with the deviation. It can be set to 2 to 5. The larger the value of k, the more significant the impact of the deviation on the inertia weight. The exponential smoothing property avoids weight jumps and improves the stability of the iteration.

[0064] Understandably, when the total cost deviation is large, the search direction deviates from the optimal direction. Increasing this value enhances the global search capability of particles; the smaller the total cost deviation, the closer the particle is to the optimal region. Reduce and enhance local fine-grained search to avoid getting trapped in local optima.

[0065] Furthermore, the particle velocity is updated in conjunction with the resource response power, specifically as follows:

[0066] in, This represents the velocity of particle i at time t+1. This represents the velocity of particle i at time t. Indicates the inertial weight at time t. Let represent the optimal position of particle i and the best-performing price strategy at time t, which is the price strategy that minimizes the aggregator's total cost and achieves the required resource response. This represents the best-performing price strategy in the entire population at time t. This represents the price strategy of the particle at time t. This represents the local learning factor, used to control the intensity of a particle's learning of its own historical best strategy. This represents the global learning factor, used to control the intensity of a particle's learning of the population's globally optimal strategy. , All numbers are random numbers in the range of 0 to 1, to avoid all particles searching in the same direction and getting trapped in local optima. This represents the response rate deviation correction factor, used to prioritize ensuring response compliance in cost reduction during price adjustments. This represents the difference between the actual resource response rate and the target resource response rate. The response rate is an indicator for evaluating the actual response effect and is used to reflect whether the total response rate of all resource entities within the VPP has reached the aggregator's expectations. It can be calculated by dividing the actual total response power of all resource entities at time t by the target total response power.

[0067] It is understandable that a low response rate indicates that the current compensation price is insufficient to incentivize resource providers to offer adequate response power. The price needs to be increased. If the response rate is too low, it indicates that the compensation price is too high, and the price can be lowered to reduce costs. This prevents the upper management from excessively lowering the compensation price in pursuit of low costs, which could lead to insufficient resource response rates. The goal is to ensure that the optimization results simultaneously meet the requirements of cost optimization and scheduling feasibility.

[0068] S3.5: Loop and Termination: Repeat steps S3.2-S3.4 until the PSO algorithm converges. The solution obtained at this point is the Stackelberg equilibrium solution, which includes the optimal compensation price strategy and the optimal power scheduling plan.

[0069] In this embodiment, a hybrid approach combining the upper-level PSO algorithm and the lower-level GA algorithm is used to solve the aforementioned bi-level optimization problem. The decision variables in the upper level are continuous, allowing for continuous optimization by PSO, which converges very quickly. The lower-level decisions involve multiple resource individuals within the VPP, while also satisfying strong constraints such as physical operation. The chromosome encoding and crossover / mutation operations of the GA algorithm better handle this complexity. This invention fully leverages the advantages of both algorithms, achieving faster upper-level convergence. PSO's guiding mechanism, including learning from historical and global bests, quickly concentrates the entire upper-level population on promising regions once a promising area is found, avoiding extensive random exploration in later stages like with GA, thus saving significant computational resources. The lower-level GA provides robustness and feasibility guarantees. PSO lacks a good guiding mechanism when dealing with complex constraints and infeasible solutions. For lower-level problems with complex structures and numerous constraints, GA, through its crossover and mutation operations, is more likely to generate feasible solutions. Furthermore, GA's population diversity mechanism makes it more robust in exploring the feasible solution space. For each upper-level decision, the lower-level GA can provide a response that, while not necessarily absolutely optimal, is certainly feasible, thus ensuring the feasibility of the entire two-level solution.

[0070] S4: The optimal power scheduling plan obtained from S3 is distributed to each adjustable resource entity for execution, thereby achieving optimized operation of the virtual power plant.

[0071] Combination Figure 2 As shown, Embodiment 2 of the present invention provides a virtual power plant optimization scheduling system, which runs the virtual power plant optimization scheduling method as described in Embodiment 1. The system includes: The resource creation and management module is used to establish operational models for multiple types of adjustable resource entities; The two-way game optimization calculation module includes a master-slave game construction module and a solution module; among them, A master-slave game theory module is constructed to build a master-slave game model between aggregators and resource entities. This model includes an upper-level model with the aggregator as the leader and a lower-level model with resource entities as followers. In the upper-level model, the aggregator aims to minimize total operating costs and makes decisions under pre-defined constraints, resulting in a compensation price strategy issued to various resource entities. In the lower-level model, each resource entity makes decisions based on an established operating model, aiming to maximize its own revenue, resulting in its own response power strategy. Resource entities then determine their power response strategies based on the compensation price strategies, which in turn influence the aggregator's compensation price strategy, leading to a game between the resource entities and the aggregator. The solution module is used to solve the master-slave game model between aggregators and resource entities to obtain the optimal compensation price strategy and the optimal response power strategy. Based on the optimal response power strategy, the optimal power scheduling plan is constructed. The optimal power scheduling plan is then distributed to each adjustable resource entity for execution to achieve optimized scheduling of the virtual power plant.

[0072] The resource modeling and management module is also used to store and manage the model parameters and operational constraints of the four types of adjustable resource entities established in S1.

[0073] It is understandable that the bidirectional game optimization calculation module is the core calculation module, which incorporates the hybrid PSO-GA algorithm to execute S2 and S3, solve the Stackelberg game model, and output the optimal scheduling plan.

[0074] The system also includes: The data acquisition and communication module is used to collect real-time data on grid electricity prices, load forecasts, photovoltaic output forecasts, and status information of various resource entities, including but not limited to SOC, indoor temperature, and EV access status, and to issue optimized scheduling instructions to resource entities.

[0075] The human-computer interaction and visualization module provides a system interface for operators to set optimization parameters, start calculation tasks, monitor system operation status, view optimization results (such as cost, revenue, power curves, etc.) and generate reports.

[0076] The system data bus and database serve as the system data hub to enable data sharing among modules and store historical data and calculation results.

[0077] The above modules interact and work together via the system bus.

[0078] Regarding the system in the above embodiments, the specific manner in which each unit performs operations has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0079] Embodiment 3 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it implements the virtual power plant optimization scheduling method described in Embodiment 1.

[0080] Embodiment 4 of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the virtual power plant optimization scheduling method according to Embodiment 1.

[0081] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0082] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0083] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0084] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0085] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A virtual power plant optimized scheduling method, characterized in that, include: Establish an operational model for multiple types of adjustable resource entities; construct a master-slave game model between aggregators and resource entities, including an upper-level model with aggregators as leaders and a lower-level model with resource entities as followers; in the upper-level model, aggregators aim to minimize total operating costs and make decisions under preset constraints to obtain compensation price strategies issued to various resource entities; in the lower-level model, each resource entity makes decisions based on the established operational model to maximize its own revenue and obtains its own response power strategy. Resource entities decide on power response strategies based on compensation price strategies, which in turn influence aggregators to formulate compensation price strategies. Resource entities and aggregators engage in a game. By solving the master-slave game model between aggregators and resource entities, the optimal compensation price strategy and the optimal response power strategy are obtained. Based on the optimal response power strategy, an optimal power scheduling plan is constructed. The optimal power scheduling plan is then distributed to each adjustable resource entity for execution, thereby achieving optimized scheduling of the virtual power plant.

2. The virtual power plant optimization scheduling method according to claim 1, characterized in that: The objective function for the aggregator is constructed as follows, with the goal of minimizing total operating costs: in, This represents the total operating cost of the aggregator. This indicates the price at which electricity is purchased from the higher-level power grid. This represents the amount of electricity purchased from the upper-level power grid at time t. This indicates the compensation price issued by the aggregator to resource entity i. This represents the response power of resource subject i. This represents the cost incurred when the aggregator's total power deviates from the grid command.

3. The virtual power plant optimization scheduling method according to claim 1, characterized in that: Construct objective functions for each resource entity with the goal of maximizing their respective gains: in, This represents the revenue of resource entity i. This represents the response power of resource entity i, including photovoltaic reduction, energy storage regulation power, air conditioning regulation power, and EV regulation power. This represents the cost corresponding to resource entity i.

4. The virtual power plant optimization scheduling method according to claim 1, characterized in that: Solving the master-slave game model between aggregators and resource entities includes: The upper-level particle swarm initialization involves encoding the aggregator's compensation price strategy into particle positions and randomly initializing the particle swarm. For each price strategy represented by a particle, a genetic algorithm is called to solve the problem. With the goal of maximizing the revenue of each resource subject, the optimal response power strategy for all resource subjects under the given price strategy is obtained. The total operating cost corresponding to this pricing strategy is calculated by combining the optimal response power strategies of all resource entities. The particle's velocity and position are updated based on the total operating cost corresponding to the pricing strategy to update the pricing strategy; Repeat the above steps until the particle swarm optimization algorithm converges, and obtain the equilibrium solution of the master-slave game model.

5. The virtual power plant optimization scheduling method according to claim 4, characterized in that: The inertia weight is updated based on the difference between the aggregator's total operating cost at the current moment and the total operating cost at the previous moment. This updated inertia weight is then used in conjunction with the velocity of the new particles, specifically: in, Indicates the inertial weight at time t. This represents the maximum value of the inertia weight. This represents the minimum inertial weight. This represents the difference between the total operating cost of the upper-level aggregator at time t and the total operating cost at the previous time. This represents the total operating cost of the upper-level aggregator during the initial iteration. Indicates the scheduling fluctuation coefficient. This represents the adjustment factor, used to control the degree to which the inertia weight changes with total operating costs.

6. The virtual power plant optimization scheduling method according to claim 4, characterized in that: The particle velocity is updated based on the resource response power, specifically as follows: in, This represents the velocity of particle i at time t+1. This represents the velocity of particle i at time t. Indicates the inertial weight at time t. This represents the best-performing price strategy for particle i at time t. This represents the best-performing price strategy in the entire population at time t. This represents the price strategy of particle i at time t. Represents the local learning factor. Represents the global learning factor. , All are random numbers in the range of 0 to 1. This represents the response rate deviation correction factor. This represents the difference between the actual resource response rate and the target resource response rate. The response rate is an indicator for evaluating the actual response effect and is used to reflect whether the total response rate of all resource entities in the virtual power plant has reached the total response rate expected by the aggregator.

7. The virtual power plant optimization scheduling method according to claim 4, characterized in that: The resource response sensitivity is calculated based on the sensitivity of the resource subject's response power to changes in the compensation price. The interval between upper and lower iterations is set according to the resource response sensitivity to perform hierarchical iteration.

8. The virtual power plant optimization scheduling method according to claim 4, characterized in that: After the upper-level particles are updated, the rate of change of the compensation price determines whether to call the genetic algorithm to solve the response power strategy in this round of iteration.

9. A virtual power plant optimization scheduling system, implementing the virtual power plant optimization scheduling method according to any one of claims 1-8, characterized in that, The system includes: The resource creation and management module is used to establish operational models for multiple types of adjustable resource entities; The two-way game optimization calculation module includes a master-slave game construction module and a solution module; among them, A master-slave game theory module is constructed to build a master-slave game model between aggregators and resource entities. This model includes an upper-level model with the aggregator as the leader and a lower-level model with resource entities as followers. In the upper-level model, the aggregator aims to minimize total operating costs and makes decisions under pre-defined constraints, resulting in a compensation price strategy issued to various resource entities. In the lower-level model, each resource entity makes decisions based on an established operating model, aiming to maximize its own revenue, resulting in its own response power strategy. Resource entities then determine their power response strategies based on their compensation price strategies, which in turn influence the aggregator's compensation price strategy, leading to a game between the resource entities and the aggregator. The solution module is used to solve the master-slave game model between aggregators and resource entities to obtain the optimal compensation price strategy and the optimal response power strategy. Based on the optimal response power strategy, the optimal power scheduling plan is constructed. The optimal power scheduling plan is then distributed to each adjustable resource entity for execution to achieve optimized scheduling of the virtual power plant.

10. An electronic device, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the virtual power plant optimization scheduling method according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the virtual power plant optimization scheduling method according to any one of claims 1-8.

Images