Multi-agent collaborative virtual power plant multi-target hierarchical optimization method and related equipment

Abstract

The invention belongs to the field of power system automation control, and discloses a multi-agent collaborative virtual power plant multi-target hierarchical optimization method and related equipment, a power grid system is divided into a dominant layer, a coordination layer and an equipment layer, and a virtual power plant multi-target hierarchical control architecture of a multi-agent system is obtained; a virtual power plant multi-target hierarchical control architecture based on a multi-agent system constructs a multi-target hierarchical optimization model, wherein the multi-target hierarchical optimization model comprises an equipment layer agent model and a coordination layer agent model; solving the coordination layer agent model based on the actual equipment operation parameters uploaded by the equipment layer agent model to obtain an optimal solution, and outputting the optimal solution as a control strategy so as to realize virtual power plant multi-target hierarchical optimization according to the control strategy; by adopting the method, not only is the comprehensive benefit of the virtual power plant improved, but also the information interaction and cooperative control capability among the agents is enhanced, so that the overall performance of the virtual power plant is remarkably improved, and the stability and reliability of a power system are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system automation control, and specifically relates to a multi-objective hierarchical optimization method for a multi-agent collaborative virtual power plant and related equipment. Background Art

[0002] As the global demand for clean energy continues to grow, distributed energy resources (DERs) such as wind power generation and photovoltaic power generation are increasingly accounting for an increasing proportion of the power system. Distributed energy resources are characterized by intermittency, volatility, and dispersion. After being connected to the traditional power system in large numbers, they have brought many challenges to the stability, reliability, and power quality of the power grid. For example, the output power of wind power generation is greatly affected by changes in wind speed, and photovoltaic power generation depends on the intensity and duration of light. These uncertainties make it difficult to maintain a balance between power supply and demand.

[0003] As a new form of power system organization, Virtual Power Plant (VPP) integrates distributed energy, energy storage equipment, controllable loads and other resources through advanced information technology and intelligent control methods to achieve coordinated and optimized operation, thereby improving the flexibility and reliability of the power system. However, the current control strategy of virtual power plants still has shortcomings in optimization, which is specifically reflected in the following aspects: most existing control optimization methods only focus on a single goal, such as simply pursuing the minimization of power generation costs, while ignoring other important goals such as environmental benefits and network losses, and cannot maximize the comprehensive benefits of virtual power plants. At the same time, in terms of multi-agent collaboration, the information interaction and coordination mechanism between agents is not perfect, resulting in the inability of virtual power plants to give full play to their advantages in resource integration.

[0004] It can be seen that the existing multi-agent collaborative virtual power plant optimization methods mostly adopt a single-objective optimization control method, which leads to poor collaborative control capabilities among various resources, reduces the overall performance and comprehensive benefits of the virtual power plant, and cannot guarantee the stability and reliability of the power system. Summary of the invention

[0005] The present invention provides a multi-objective hierarchical optimization method and related equipment for a multi-agent collaborative virtual power plant, so as to solve the technical problem that the existing multi-agent collaborative virtual power plant optimization methods mostly adopt a single-objective optimization control mode, resulting in poor collaborative control capabilities among various resources, reducing the overall performance and comprehensive benefits of the virtual power plant, and failing to ensure the stability and reliability of the power system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, comprising: The power grid system is divided into the leading layer, the coordination layer and the equipment layer, and the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system is obtained; A multi-objective hierarchical optimization model is constructed based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The coordination layer proxy model is solved based on the actual operating parameters of the equipment uploaded by the device layer proxy model to obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0007] A further improvement of the present invention is that the power grid system is divided into a leading layer, a coordination layer and a device layer to obtain a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, including: The power grid is divided into the leading layer, the coordination layer and the equipment layer; The leading layer includes the distribution network agent and the heating network agent, which are responsible for interacting with the outside world and issuing strategies to the coordination layer; The coordination layer is composed of multiple regional virtual power plant control center agents, which are responsible for managing local regional resources and exchanging information with the leading layer and the equipment layer; The equipment layer includes energy storage agents, electric load agents, electric vehicle agents, gas turbine agents, fuel cell agents, photovoltaic agents, and wind power agents, which are responsible for monitoring equipment status and adjusting power according to instructions; Each agent in the leading layer, coordination layer and equipment layer is divided into two layers of control structure; the upper layer control receives the operating environment information through the sensor, maps the scene after processing by the event processing distributor, negotiates and cooperates with other agents through the communication system to achieve the optimal goal, selects decisions and formulates action responses in the decision maker, and acts on the virtual power plant through the effector; the lower layer control includes voltage control function and frequency control function to ensure supply and demand balance and safe and stable operation; Build a multi-objective hierarchical control architecture for a virtual power plant with a multi-agent system.

[0008] A further improvement of the present invention is that the device layer proxy model includes a wind power generation model, a controllable distributed power supply model and an energy storage device model; wherein, The specific expression of the wind power generation model is as follows:

[0009] Where P WT Represents the output active power of the wind turbine; P rrepresents the rated power of the wind turbine; v represents the wind speed; v ci Indicates the cut-in wind speed; v co Indicates the cut-out wind speed; v r represents the rated wind speed; k1 represents the first coefficient of the relationship between power and wind speed; k2 represents the second coefficient of the relationship between power and wind speed; The controllable distributed power supply model satisfies the following relationship:

[0010] Where P j t represents the output power of the jth controllable distributed generation at time t; P j t-1 represents the output power of the jth controllable distributed generation at time t-1; P j，min represents the minimum output power of the jth controllable distributed power source; P j，max represents the maximum output power of the jth controllable distributed power source; UR i represents the ramp rate of the jth controllable distributed power source; DR j represents the ramp rate of the jth controllable distributed power source; The specific expression of the energy storage device model is as follows:

[0011] In the energy storage device model, the power constraints and energy constraints of charging and discharging are:

[0012] In the formula, SOC i t represents the state of charge of the i-th energy storage device at time t; SOC i t+1 represents the charge state of the i-th energy storage device at time t+1; P t SB,i Cap represents the charging and discharging power of the i-th energy storage device at time t; i represents the capacity of the i-th energy storage device; Indicates the lower limit of the charge state of the i-th energy storage device; Indicates the upper limit of the state of charge of the i-th energy storage device; represents the charge state of the i-th energy storage device at time t+T; It means that at the end of a scheduling cycle T, the charge state of the energy storage device is restored to the state at the beginning of the cycle.

[0013] A further improvement of the present invention is that the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as the upper layer optimization objectives and minimization of network loss as the lower layer optimization objective; wherein: Taking economic cost minimization as the upper optimization goal, the specific expression is as follows:

[0014] In the formula, F1 represents the objective function of minimizing economic cost; t represents different moments, T represents the time set included in the entire scheduling cycle; i represents the index of the power generation equipment; S G Represents the collection of power generation equipment in a virtual power plant, including micro gas turbines, fuel cells and controllable distributed power sources; represents the operation and maintenance cost of the i-th power generation equipment at time t; represents the start / stop status of the i-th power generation equipment at time t; represents the operation and maintenance cost coefficient per unit power of the i-th power generation equipment; represents the output power of the i-th power generation equipment at time t; represents the investment cost of the i-th power generation equipment at time t; represents the investment return rate of the i-th power generation equipment; represents the depreciation period of the i-th power generation equipment; represents the unit capacity investment cost of the i-th power generation equipment; k is the annual utilization hours of the power generation equipment; Taking environmental cost minimization as the upper-level optimization goal, the specific expression is as follows:

[0015] In the formula, F2 represents the environmental cost minimization objective function; j represents the index of the pollutant type, and J is the set of pollutant types; represents the coefficient of the jth pollutant emitted per unit power of the i-th power generation equipment; α represents the environmental impact factor of the j-th pollutant; Taking network loss minimization as the lower-level optimization goal, the specific expression is as follows:

[0016] Where F3 represents the network loss minimization objective function; i, j represents the index of the grid node, i, j∈N, N is the set of grid nodes; G ij represents the conductance between node i and node j; U i represents the voltage amplitude of node i; U j represents the voltage amplitude of node j; θ ij Represents the voltage phase angle difference between node i and node j.

[0017] A further improvement of the present invention is that the coordination layer agent model uses the internal power balance constraint and network security constraint of the virtual power plant as constraint conditions; wherein the network security constraint includes voltage constraint and line capacity constraint; The internal power balance constraint of the virtual power plant is specifically expressed as follows:

[0018] The voltage constraint is specifically expressed as follows:

[0019] The line capacity constraint is specifically expressed as follows:

[0020] In the formula, represents the output power of the i-th power generation equipment, N G Represents a collection of power generation equipment in a virtual power plant; represents the charging and discharging power of the jth energy storage device, represents the collection of energy storage devices; P P Indicates the power purchased by the power plant; P S represents the power sold by the virtual power plant; D represents the load demand of the virtual power plant; P loss Represents the network loss power within the virtual power plant; U min，i Indicates the minimum allowable voltage amplitude of node i; U max，i represents the maximum allowable voltage amplitude of node i; S i represents the apparent power of line i; S min，i represents the minimum permissible apparent power of line i; S max，i Represents the maximum allowable apparent power of line i.

[0021] A further improvement of the present invention is that solving the coordination layer proxy model based on the actual operation parameters of the device uploaded by the device layer proxy model includes: The membership function is used to fuzzy each objective function in the coordination layer agent model, and the multi-objective problem is transformed into a single-objective nonlinear optimization problem based on maximizing satisfaction. The actual operating parameters of the equipment uploaded by the device layer proxy model are used to solve the single-objective nonlinear optimization problem based on maximizing satisfaction and obtain the optimal solution.

[0022] A further improvement of the present invention is that the membership function The specific expression is as follows:

[0023]

[0024] The transformed coordination layer agent model is specifically expressed as follows:

[0025]

[0026] In the formula, , They represent curve characteristic parameters and allowable deviation respectively; , All represent other values ​​of the objective function; represents the adjusted allowable deviation; β represents the minimum value of the membership function of each objective function under the row range constraint.

[0027] In a second aspect, the present invention provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization system, comprising: An architecture building module is used to divide the power grid system into a dominant layer, a coordination layer, and an equipment layer, and obtain a multi-objective hierarchical control architecture of a virtual power plant with a multi-agent system; An optimization model building module is used to build a multi-objective hierarchical optimization model based on a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as upper-layer optimization objectives, and minimization of network loss as lower-layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The model solving module is used to solve the coordination layer proxy model based on the actual operating parameters of the equipment uploaded by the device layer proxy model, obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0028] In a third aspect, the present invention provides a device, comprising: Memory for storing computer programs; A processor is used to implement the steps of the above-mentioned multi-agent collaborative virtual power plant multi-objective hierarchical optimization method when executing the computer program.

[0029] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, is used to implement the steps of the above-mentioned multi-agent collaborative virtual power plant multi-objective hierarchical optimization method.

[0030] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method. By constructing a multi-agent system virtual power plant multi-objective hierarchical control architecture of the leading layer, the coordination layer and the equipment layer, the multi-level and refined management of distributed energy, energy storage equipment, controllable loads and other resources in the virtual power plant is realized; through the actual operating parameters uploaded by the equipment layer agent model, the coordination layer agent model can comprehensively consider the three upper and lower layer optimization goals of minimizing economic costs, minimizing environmental costs and minimizing network losses, and solve the optimal control strategy under the constraints of power balance and network security inside the virtual power plant. The use of this method not only overcomes the limitations of the single-objective optimization of the existing control strategy and improves the comprehensive benefits of the virtual power plant, but also enhances the information interaction and collaborative control capabilities between the agents, thereby significantly improving the overall performance of the virtual power plant and ensuring the stability and reliability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A block diagram of a multi-objective hierarchical control architecture of a multi-agent system provided by an embodiment of the present invention; Figure 2 A schematic diagram of an agent unit structure in a multi-objective hierarchical optimization model provided by an embodiment of the present invention; Figure 3 A schematic diagram of the group output during multi-objective optimization provided by an embodiment of the present invention; Figure 4 A schematic diagram showing a comparison of network losses between single-objective and multi-objective optimization provided by an embodiment of the present invention; Figure 5 A flowchart of a multi-objective hierarchical optimization method for a virtual power plant with multi-agent collaboration provided by an embodiment of the present invention; Figure 6 A structural schematic diagram of a multi-agent collaborative virtual power plant multi-objective hierarchical optimization system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to further understand the content of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0033] As described in the background technology, the current control strategy of virtual power plants still has deficiencies in optimization, which are specifically reflected in the following aspects: most existing control optimization methods only focus on a single goal, such as simply pursuing the minimization of power generation costs, while ignoring other important goals such as environmental benefits and network losses, and are unable to maximize the comprehensive benefits of virtual power plants. At the same time, in terms of multi-agent collaboration, the information interaction and coordination mechanism between agents is not perfect, resulting in the virtual power plant being unable to give full play to its advantages in resource integration.

[0034] Example 1 In order to solve the above problems, this embodiment provides a multi-objective hierarchical optimization method for a virtual power plant with multi-agent collaboration. This method constructs a hierarchical control architecture and introduces a multi-agent collaboration mechanism to achieve optimized operation of the virtual power plant under multiple objectives, improve the resource utilization efficiency of the virtual power plant, reduce operating costs, reduce environmental pollution, and enhance the stability and reliability of the power system.

[0035] This embodiment provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, including: The power grid system is divided into the leading layer, the coordination layer and the equipment layer, and the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system is obtained; A multi-objective hierarchical optimization model is constructed based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The coordination layer proxy model is solved based on the actual operating parameters of the equipment uploaded by the device layer proxy model to obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0036] The optimization method provided in this embodiment is further explained below in conjunction with the accompanying drawings: This embodiment provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, which is as follows: Step 1: Divide the power grid system into the dominant layer, coordination layer and equipment layer to obtain the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system.

[0037] like Figure 1 As shown, the power grid is divided into a dominant layer, a coordination layer and an equipment layer; a multi-objective hierarchical control architecture of a virtual power plant with a multi-agent system is constructed.

[0038] Before building a multi-target hierarchical control architecture for a virtual power plant of a multi-agent system, it is also necessary to design a multi-level heterogeneous communication architecture for the virtual power plant. In view of the characteristics of the virtual power plant, such as rich types of constituent entities, flexible access forms, and large differences in business requirements, this embodiment provides support for building a communication network hierarchical control solution that meets the differentiated business requirements of the virtual power plant based on the multi-level heterogeneous communication solution for the virtual power plant based on the secure interconnection of the enterprise private network and the public network. Specifically, it includes: Basic functions and information interaction of each level of virtual power plant: The platform layer of virtual power plant is divided into production control area and management information area. The production control area includes production real-time control area and production non-real-time control area. The management information area includes production management area and non-production real-time information area. Each area has different functions and exchanges information with each other. At the same time, the platform layer collects and interacts with the Internet through the secure access area, and third-party load aggregators can exchange information with the virtual power plant through the Internet.

[0039] Basic functions and information interaction of each security zone in the virtual power plant platform layer: The security access zone realizes data interaction between the virtual power plant platform layer and the external network under the premise of ensuring security. The edge layer deploys edge servers to perform preliminary processing and storage of data, and uploads the data to the platform layer after aggregation, and receives the policies issued by the platform layer. The access layer adopts EPON (EPON, full name Ethernet Passive Optical Network), industrial Ethernet, WLAN, power wireless private network, 4G / 5G, PLC and other communication technologies to realize the access of business terminal layer equipment. The business terminal layer includes relay protection devices, electric energy metering and collection devices, smart meters, distribution network terminal collection devices, office systems, sensors, etc., as well as distributed photovoltaic, wind turbines, smart distribution networks, industrial loads, civil loads, charging piles, energy storage stations and other virtual power plant business scenario equipment.

[0040] The multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system includes the leading layer, the coordination layer and the equipment layer. The leading layer includes the distribution network agent and the heating network agent, which are responsible for interacting with the outside and issuing strategies to the coordination layer. The coordination layer is composed of multiple regional virtual power plant control center agents, which are responsible for managing local regional resources and exchanging information with the leading layer and the equipment layer; The equipment layer includes energy storage agents, electric load agents, electric vehicle agents, gas turbine agents, fuel cell agents, photovoltaic agents, and wind power agents, which are responsible for monitoring equipment status and adjusting power according to instructions; Each agent in the leading layer, coordination layer and equipment layer is divided into two layers of control structure; the upper layer control receives the operating environment information through the sensor, maps the scene after processing by the event processing distributor, negotiates and cooperates with other agents through the communication system to achieve the optimal goal, selects decisions and formulates action responses in the decision maker, and acts on the virtual power plant through the effector; the lower layer control includes voltage control function and frequency control function to ensure supply and demand balance and safe and stable operation; like Figure 2As shown in the figure, in the multi-objective hierarchical control architecture system of the virtual power plant of the multi-agent system, each agent is divided into two layers of control structure. In the upper-level control, the agent receives the operating environment information of the virtual power plant through the sensor, and processes the perceived raw data in time through the event processing distributor and maps it to a scenario. In this scenario, the agent negotiates and cooperates with other agents through the communication system to achieve the optimal goal, and selects the decision when the optimal goal is selected in the decision maker. Accordingly, appropriate action responses are formulated with the support of existing knowledge or rules of the functional module, and finally act on the virtual power plant through the effector; the lower-level control includes functions such as voltage control and frequency control to ensure supply and demand balance and safe and stable operation, and realize the on-site control of the agent.

[0041] Step 2: Construct a multi-objective hierarchical optimization model based on the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system.

[0042] The multi-objective hierarchical optimization model specifically includes a device layer agent model and a coordination layer agent model; Among them, the device layer agent model includes wind power generation model, controllable distributed power supply model and energy storage device model; Wind power generation model: Wind power generation uses wind turbines to convert wind energy into electrical energy, and its output power changes with wind speed. When the wind speed distribution is known, according to the operating principle of wind power generation, there is the following functional relationship between the wind turbine output active power and wind speed:

[0043] Where P WT Represents the output active power of the wind turbine; P r represents the rated power of the wind turbine; v represents the wind speed; v ci Indicates the cut-in wind speed; v co Indicates the cut-out wind speed; v r represents the rated wind speed; k1 represents the first coefficient of the relationship between power and wind speed; k2 represents the second coefficient of the relationship between power and wind speed; Controllable distributed power model: The controllable distributed power in the virtual power plant should meet the output power constraints and ramp rate constraints.

[0044] The output power constraint is ; The climbing rate constraint is ; Where P j t represents the output power of the jth controllable distributed generation at time t; P j t-1 represents the output power of the jth controllable distributed generation at time t-1; P j，min represents the minimum output power of the jth controllable distributed power source; Pj，max represents the maximum output power of the jth controllable distributed power source; UR i represents the ramp rate of the jth controllable distributed power source; DR j represents the ramp rate of the jth controllable distributed power source; Energy storage device model: Energy storage devices can control charging and discharging more quickly to follow changes in load through energy storage controllers. To protect batteries and extend their service life, virtual power plants cannot be overcharged or over-discharged. Their energy storage model is:

[0045] The power constraints and energy constraints for charging and discharging are:

[0046] In the formula, SOC i t represents the state of charge of the i-th energy storage device at time t; SOC i t+1 represents the charge state of the i-th energy storage device at time t+1; P t SB,i Cap represents the charging and discharging power of the i-th energy storage device at time t; i represents the capacity of the i-th energy storage device; Indicates the lower limit of the charge state of the i-th energy storage device; Indicates the upper limit of the state of charge of the i-th energy storage device; represents the charge state of the i-th energy storage device at time t+T; It means that at the end of a scheduling cycle T, the charge state of the energy storage device is restored to the state at the beginning of the cycle.

[0047] The specific steps for building the coordination layer proxy model are as follows: Objective function construction: As a coordination layer agent, the virtual power plant control center should be able to construct the objective function of this layer according to the incentive signal sent by the distribution network layer, and dispatch the equipment layer agent to achieve regional autonomy. Consider the upper-level goals such as minimizing economic costs, low carbon and environmental protection, and the lower-level goals such as minimizing network losses.

[0048] The coordination layer agent model takes minimizing economic costs and environmental costs as the upper optimization goals, and minimizing network losses as the lower optimization goal; among them: Taking economic cost minimization as the upper optimization goal, the specific expression is as follows:

[0049] In the formula, F1 represents the objective function of minimizing economic cost; t represents different moments, T represents the time set included in the entire scheduling cycle; i represents the index of the power generation equipment; S G Represents the collection of power generation equipment in a virtual power plant, including micro gas turbines, fuel cells and controllable distributed power sources; represents the operation and maintenance cost of the i-th power generation equipment at time t; represents the start / stop status of the i-th power generation equipment at time t; represents the operation and maintenance cost coefficient per unit power of the i-th power generation equipment; represents the output power of the i-th power generation equipment at time t; represents the investment cost of the i-th power generation equipment at time t; represents the investment return rate of the i-th power generation equipment; represents the depreciation period of the i-th power generation equipment; represents the unit capacity investment cost of the i-th power generation equipment; k is the annual utilization hours of the power generation equipment; Taking environmental cost minimization as the upper-level optimization goal, the specific expression is as follows:

[0050] In the formula, F2 represents the environmental cost minimization objective function; j represents the index of the pollutant type, and J is the set of pollutant types; represents the coefficient of the jth pollutant emitted per unit power of the i-th power generation equipment; α represents the environmental impact factor of the j-th pollutant; Taking network loss minimization as the lower-level optimization goal, the specific expression is as follows:

[0051] Where F3 represents the network loss minimization objective function; i, j represents the index of the grid node, i, j∈N, N is the set of grid nodes; G ij represents the conductance between node i and node j; U i represents the voltage amplitude of node i; U j represents the voltage amplitude of node j; θ ij Represents the voltage phase angle difference between node i and node j.

[0052] The coordination layer agent model takes the internal power balance constraints and network security constraints of the virtual power plant as constraints; among them, the network security constraints include voltage constraints and line capacity constraints.

[0053] The internal power balance constraint of the virtual power plant is specifically expressed as follows:

[0054] The voltage constraint is specifically expressed as follows:

[0055] The line capacity constraint is specifically expressed as follows:

[0056] The specific steps of multi-objective optimization are as follows: Determine the membership function: Since the three objectives under consideration are of different natures and have different dimensions and orders of magnitude, there is no clear boundary between the optimal solution of each sub-objective and the optimal solution of multi-objective optimization, which is not conducive to decision makers to select the optimal solution. Therefore, in this embodiment, fuzzy mathematics is used to process each objective function to normalize its value. The objective functions considered are all minimization objective functions, that is, the smaller the value of the objective function, the greater the satisfaction, and a relatively small membership function should be selected. In order to meet the robustness and ensure that the multi-objective function is differentiable within the definition domain, the inverse sigmoid function that is continuously differentiable in the definition domain is used as the membership function. The principle that the values ​​of the semi-descending linear and inverse sigmoid functions are equal at μ at 0.5 and 0.9 is used to determine the parameter value and obtain the membership function of the objective function. for:

[0057]

[0058] In the formula, , They represent curve characteristic parameters and allowable deviation respectively; , All represent other values ​​of the objective function; Indicates the adjusted tolerance.

[0059] Transformation of multi-objective problems: Each objective function is fuzzy processed through the membership function, and the multi-objective problem is transformed into a single-objective nonlinear optimization problem based on maximizing satisfaction.

[0060] Let β be the minimum value of the membership function of each objective function under the operating range constraint, that is, ; The maximum and minimum function models based on satisfaction are:

[0061] For all sub-goals, satisfaction The higher it is, the higher the quality of the corresponding solution is, and the optimal solution can be obtained based on it.

[0062] Step 3: The actual operating parameters of the equipment uploaded by the device layer proxy model are solved by the coordination layer proxy model to obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0063] Here, the device layer agent uploads the collected device operation status information to the coordination layer agent in real time, and the coordination layer agent uploads the comprehensive information in the area to the leading layer agent. The leading layer agent issues strategies to the coordination layer agent according to the instructions of the external power grid and heating system, and the coordination layer agent then sends specific control instructions to the device layer agent.

[0064] Each agent makes decisions based on its own goals and the information it receives. The device layer agent optimizes the operation mode of the equipment based on the local device status and coordination layer instructions; the coordination layer agent formulates the resource scheduling strategy within the region based on the resource status within the region and the instructions of the dominant layer; the dominant layer agent coordinates the operation of each region from a global perspective to achieve the overall optimization of the virtual power plant.

[0065] The specific solution steps of the coordination layer agent model are as follows: First, initialization parameters: set the weight coefficients of each objective function, the parameters of the membership function, and the relevant parameters of the multi-agent system.

[0066] Second, calculate the objective function value: According to the data uploaded by the device layer and the set parameters, calculate the values ​​of the economic cost minimization objective function F1, the environmental cost minimization objective function F2 and the network loss minimization objective function F3.

[0067] Third, determine the membership function value: According to the objective function value, calculate the membership function of each objective function .

[0068] Fourth, transform into a single-objective optimization problem: transform the multi-objective problem into a single-objective nonlinear optimization problem based on maximizing satisfaction.

[0069] Fifth, solve the single-objective optimization problem: Use intelligent optimization algorithms (such as genetic algorithms, particle swarm optimization algorithms, etc.) to solve the single-objective optimization problem and obtain the optimal solution, so as to determine the optimal operation strategy for each equipment in the virtual power plant, which is used as the control strategy output to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0070] In the following, theoretical simulation is used to verify the rationality and applicability of the multi-objective hierarchical optimization method of the multi-agent collaborative virtual power plant provided in this embodiment.

[0071] First, a virtual power plant system is formed by connecting wind turbines, fuel cells, batteries, electric vehicles, photovoltaics, etc. into the model. The operating parameters of large-scale distributed energy are shown in Table 1, where the capacity of the energy storage device is 900kWh, the range of the state of charge SoC is [0.3, 0.9], and the initial energy is 450 kWh.

[0072] Table 1 shows the operating parameters of distributed power supply

[0073] First, according to the objective functions and constraints in the coordination layer proxy model provided in this embodiment, a genetic algorithm is used to solve the single-objective optimization problem, and the results of the single-objective optimization and the calculated membership function parameters are brought into the formula to solve the maximum and minimum functions based on satisfaction. The optimization results under each optimization objective are shown in Table 2: Table 2 shows the multi-objective optimization results.

[0074] It can be seen from Table 2 that when the goal is to minimize the power generation cost of the virtual power plant, the stability of the network and the impact of pollutants on the environment during the operation of the virtual power plant are not taken into account, the environmental cost is 38.95% higher than the minimum environmental cost, the environmental benefit is poor, and the network loss is 0.79% higher than the case of minimum network loss; when the goal is low-carbon and environmental protection optimization, the power generation cost is 6.53% higher than the minimum power generation cost, and the network loss is 0.54% higher than the case of minimum network loss; when considering the safety goal taking network loss as an example, the whole network can achieve safe and stable operation, but the power generation cost is 4.91% higher than the minimum power generation cost, and the environmental cost is 22.51% higher than the minimum environmental cost, and the environmental benefit is not ideal.

[0075] Combination Figure 3 As shown, Figure 3 The output of the dispatchable unit agent of the virtual power plant during multi-objective optimization. Since the battery charging and discharging cost is low, the battery is called first to meet the dispatching requirements. When the leading agent gives the real-time electricity price, the electricity price is low and the load is light in the 1-7 period. The battery in the virtual power plant stores the excess power of wind turbines, photovoltaics, micro gas turbines, etc.; the electricity price is high in the 11-15 period, and the battery with low power generation cost is called first to discharge to make up for the power shortage and enable the virtual power plant to participate in the power market as a producer; in the 19-24 period, due to the significant reduction in the output of the photovoltaic power generation unit and the increase in load, the battery alone cannot meet the internal power balance of the virtual power plant. The active power shortage is provided by the battery, micro gas turbine, fuel cell and the purchase of electricity through the power market. At this time, the virtual power plant participates in the power market as an external power consumer.

[0076] like Figure 4 As shown, Figure 4This is a comparison chart of the network loss of virtual power plants during single-objective and multi-objective optimization. With multi-objective optimization, the network loss of virtual power plants is slightly higher than the optimal situation during single-objective optimization, but the compromise of the optimization results of each single objective is considered as a whole. Combined with Table 2, the multi-objective optimization is 1.24%, 8.51%, and 0.22% higher than the minimum economic cost, minimum environmental cost, and minimum network loss, respectively. Multi-objective optimization takes into account the economy, environmental protection, safety and stability of virtual power plant operation, and has greatly improved the environmental benefits compared to only considering economic or safety goals. It can better realize the comprehensive utilization of large-scale distributed energy under the background of low carbon.

[0077] It can be seen that this embodiment provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, which has the following advantages over existing control measures: First, improve comprehensive benefits: Through multi-objective optimization, the present invention realizes the coordinated optimization of virtual power plants in terms of economic costs, environmental benefits and network losses. Compared with traditional single-objective control strategies, it can significantly reduce power generation costs, reduce pollutant emissions, reduce network losses, and improve the overall comprehensive benefits of virtual power plants.

[0078] Second, enhance system stability: The multi-agent collaboration mechanism enables the various resources within the virtual power plant to be quickly and effectively coordinated and controlled based on real-time information, effectively responding to the intermittency and volatility of distributed energy, and enhancing the stability and reliability of the power system.

[0079] Third, improve market competitiveness: The method of the present invention enables the virtual power plant to better adapt to changes in the electricity market, flexibly adjust power output and resource allocation according to market price signals and grid dispatch instructions, and improve market competitiveness and profitability.

[0080] Fourth, promote the consumption of clean energy: through multi-agent collaboration and multi-objective hierarchical control, it is possible to better integrate distributed energy resources, improve the utilization efficiency of clean energy, and promote the large-scale consumption of clean energy.

[0081] Example 2 like Figure 5 As shown, this embodiment provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, including the following steps: The power grid system is divided into the leading layer, the coordination layer and the equipment layer, and the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system is obtained; A multi-objective hierarchical optimization model is constructed based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The coordination layer proxy model is solved based on the actual operating parameters of the equipment uploaded by the device layer proxy model to obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0082] The power grid system is divided into a leading layer, a coordination layer and a device layer, and a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system is obtained, including: The power grid is divided into the leading layer, the coordination layer and the equipment layer; The leading layer includes the distribution network agent and the heating network agent, which are responsible for interacting with the outside world and issuing strategies to the coordination layer; The coordination layer is composed of multiple regional virtual power plant control center agents, which are responsible for managing local regional resources and exchanging information with the leading layer and the equipment layer; The equipment layer includes energy storage agents, electric load agents, electric vehicle agents, gas turbine agents, fuel cell agents, photovoltaic agents, and wind power agents, which are responsible for monitoring equipment status and adjusting power according to instructions; Each agent in the leading layer, coordination layer and equipment layer is divided into two layers of control structure; the upper layer control receives the operating environment information through the sensor, maps the scene after processing by the event processing distributor, negotiates and cooperates with other agents through the communication system to achieve the optimal goal, selects decisions and formulates action responses in the decision maker, and acts on the virtual power plant through the effector; the lower layer control includes voltage control function and frequency control function to ensure supply and demand balance and safe and stable operation; Build a multi-objective hierarchical control architecture for a virtual power plant with a multi-agent system.

[0083] Specifically, the device layer proxy model includes a wind power generation model, a controllable distributed power supply model and an energy storage device model; wherein, The specific expression of the wind power generation model is as follows:

[0084] Where P WT Represents the output active power of the wind turbine; P r represents the rated power of the wind turbine; v represents the wind speed; v ci Indicates the cut-in wind speed; v co Indicates the cut-out wind speed; v rrepresents the rated wind speed; k1 represents the first coefficient of the relationship between power and wind speed; k2 represents the second coefficient of the relationship between power and wind speed; The controllable distributed power supply model satisfies the following relationship:

[0085] Where P j t represents the output power of the jth controllable distributed generation at time t; P j t-1 represents the output power of the jth controllable distributed generation at time t-1; P j，min represents the minimum output power of the jth controllable distributed power source; P j，max represents the maximum output power of the jth controllable distributed power source; UR i represents the ramp rate of the jth controllable distributed power source; DR j represents the ramp rate of the jth controllable distributed power source; The specific expression of the energy storage device model is as follows:

[0086] In the energy storage device model, the power constraints and energy constraints of charging and discharging are:

[0087] In the formula, SOC i t represents the state of charge of the i-th energy storage device at time t; SOC i t+1 represents the charge state of the i-th energy storage device at time t+1; P t SB,i Cap represents the charging and discharging power of the i-th energy storage device at time t; i represents the capacity of the i-th energy storage device; Indicates the lower limit of the charge state of the i-th energy storage device; Indicates the upper limit of the state of charge of the i-th energy storage device; represents the charge state of the i-th energy storage device at time t+T; It means that at the end of a scheduling cycle T, the charge state of the energy storage device is restored to the state at the beginning of the cycle.

[0088] In particular, the coordination layer agent model takes minimization of economic cost and environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; wherein: Taking economic cost minimization as the upper optimization goal, the specific expression is as follows:

[0089] In the formula, F1 represents the objective function of minimizing economic cost; t represents different moments, T represents the time set included in the entire scheduling cycle; i represents the index of the power generation equipment; S G Represents the collection of power generation equipment in a virtual power plant, including micro gas turbines, fuel cells and controllable distributed power sources; represents the operation and maintenance cost of the i-th power generation equipment at time t; represents the start / stop status of the i-th power generation equipment at time t; represents the operation and maintenance cost coefficient per unit power of the i-th power generation equipment; represents the output power of the i-th power generation equipment at time t; represents the investment cost of the i-th power generation equipment at time t; represents the investment return rate of the i-th power generation equipment; represents the depreciation period of the i-th power generation equipment; represents the unit capacity investment cost of the i-th power generation equipment; k is the annual utilization hours of the power generation equipment; Taking environmental cost minimization as the upper-level optimization goal, the specific expression is as follows:

[0090] In the formula, F2 represents the environmental cost minimization objective function; j represents the index of the pollutant type, and J is the set of pollutant types; represents the coefficient of the jth pollutant emitted per unit power of the i-th power generation equipment; α represents the environmental impact factor of the j-th pollutant; Taking network loss minimization as the lower-level optimization goal, the specific expression is as follows:

[0091] Where F3 represents the network loss minimization objective function; i, j represents the index of the grid node, i, j∈N, N is the set of grid nodes; G ij represents the conductance between node i and node j; U i represents the voltage amplitude of node i; U j represents the voltage amplitude of node j; θ ij Represents the voltage phase angle difference between node i and node j.

[0092] Here, the coordination layer agent model takes the internal power balance constraint and network security constraint of the virtual power plant as constraint conditions; wherein the network security constraint includes voltage constraint and line capacity constraint; The internal power balance constraint of the virtual power plant is specifically expressed as follows:

[0093] The voltage constraint is specifically expressed as follows:

[0094] The line capacity constraint is specifically expressed as follows:

[0095] In the formula, represents the output power of the i-th power generation equipment, N G Represents a collection of power generation equipment in a virtual power plant; represents the charging and discharging power of the jth energy storage device, represents the collection of energy storage devices; P P Indicates the power purchased by the power plant; P S represents the power sold by the virtual power plant; D represents the load demand of the virtual power plant; P loss Represents the network loss power within the virtual power plant; U min，i Indicates the minimum allowable voltage amplitude of node i; U max，i represents the maximum allowable voltage amplitude of node i; S i represents the apparent power of line i; S min，i represents the minimum permissible apparent power of line i; S max，i Represents the maximum allowable apparent power of line i.

[0096] Here, solving the coordination layer proxy model based on the actual operation parameters of the device uploaded by the device layer proxy model includes: The membership function is used to fuzzy each objective function in the coordination layer agent model, and the multi-objective problem is transformed into a single-objective nonlinear optimization problem based on maximizing satisfaction. The actual operating parameters of the equipment uploaded by the device layer proxy model are used to solve the single-objective nonlinear optimization problem based on maximizing satisfaction and obtain the optimal solution.

[0097] In this embodiment, the membership function The specific expression is as follows:

[0098]

[0099] The transformed coordination layer agent model is specifically expressed as follows:

[0100]

[0101] In the formula, , They represent curve characteristic parameters and allowable deviation respectively; , All represent other values ​​of the objective function; represents the adjusted allowable deviation; β represents the minimum value of the membership function of each objective function under the row range constraint.

[0102] like Figure 6 As shown, the present embodiment also provides a multi-agent collaborative virtual power plant multi-objective hierarchical optimization system, including: an architecture building module, used to divide the power grid system into a dominant layer, a coordination layer and an equipment layer, and obtain a multi-objective hierarchical control architecture of a virtual power plant of the multi-agent system; an optimization model building module, used to build a multi-objective hierarchical optimization model based on the multi-objective hierarchical control architecture of a virtual power plant of the multi-agent system, the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic costs and minimization of environmental costs as upper-layer optimization objectives, and minimization of network losses as lower-layer optimization objectives; and takes internal power balance constraints and network security constraints of the virtual power plant as constraints; a model solving module, used to solve the coordination layer agent model based on the actual operating parameters of the equipment uploaded by the equipment layer agent model, obtain the optimal solution and output the optimal solution as a control strategy, so as to realize multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0103] The present invention also provides a device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the multi-agent collaborative virtual power plant multi-objective hierarchical optimization method when executing the computer program.

[0104] When the processor executes the computer program, the steps of multi-objective hierarchical optimization of the virtual power plant with multi-agent collaboration are implemented, for example: dividing the power grid system into a dominant layer, a coordination layer and an equipment layer to obtain a multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system; constructing a multi-objective hierarchical optimization model based on the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system, the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic costs and minimization of environmental costs as upper-layer optimization goals, and minimization of network losses as lower-layer optimization goals; and takes internal power balance constraints and network security constraints of the virtual power plant as constraints; solving the coordination layer agent model based on the actual operating parameters of the equipment uploaded by the equipment layer agent model, obtaining the optimal solution and outputting the optimal solution as the control strategy, so as to realize multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0105] Alternatively, when the processor executes the computer program, the functions of each module in the above-mentioned system are implemented, for example: an architecture building module, which is used to divide the power grid system into a dominant layer, a coordination layer and an equipment layer to obtain a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system; an optimization model building module, which is used to build a multi-objective hierarchical optimization model based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, and the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic costs and minimization of environmental costs as upper-level optimization goals, and minimization of network losses as lower-level optimization goals; and takes internal power balance constraints and network security constraints of the virtual power plant as constraints; a model solving module, which is used to solve the coordination layer agent model based on the actual operating parameters of the equipment uploaded by the equipment layer agent model, obtain the optimal solution and output the optimal solution as a control strategy, so as to realize multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0106] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing preset functions, and the instruction segments are used to describe the execution process of the computer program in the multi-agent collaborative virtual power plant multi-objective hierarchical optimization device. For example, the computer program can be divided into an architecture building module, an optimization model building module and a model solving module; the specific functions of each module are as follows: an architecture building module, used to divide the power grid system into a dominant layer, a coordination layer and an equipment layer, and obtain a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system; an optimization model building module, used to construct a multi-objective hierarchical optimization model based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, and the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic costs and minimization of environmental costs as upper-level optimization goals, and minimization of network losses as lower-level optimization goals; and takes internal power balance constraints and network security constraints of the virtual power plant as constraints; a model solving module, used to solve the coordination layer agent model based on the actual operating parameters of the equipment uploaded by the equipment layer agent model, obtain the optimal solution and output the optimal solution as a control strategy, so as to realize multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

[0107] The multi-agent collaborative virtual power plant multi-objective hierarchical optimization device can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The multi-agent collaborative virtual power plant multi-objective hierarchical optimization device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above is an example of a multi-agent collaborative virtual power plant multi-objective hierarchical optimization device, and does not constitute a limitation on the multi-agent collaborative virtual power plant multi-objective hierarchical optimization device, and may include more components than the above, or a combination of certain components, or different components. For example, the multi-agent collaborative virtual power plant multi-objective hierarchical optimization device may also include input and output devices, network access devices, buses, etc.

[0108] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The processor is the control center of the multi-objective hierarchical optimization of the multi-agent collaborative virtual power plant, and uses various interfaces and lines to connect various parts of the multi-objective hierarchical optimization equipment of the multi-agent collaborative virtual power plant.

[0109] The memory can be used to store the computer program and / or module, and the processor implements various functions of the multi-agent collaborative virtual power plant multi-objective hierarchical optimization device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory.

[0110] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0111] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the multi-objective hierarchical optimization method of a multi-agent collaborative virtual power plant.

[0112] The modules / units integrated in the multi-agent collaborative virtual power plant multi-objective hierarchical optimization system can be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as independent products.

[0113] Based on such understanding, the present invention implements all or part of the processes in the multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the multi-agent collaborative virtual power plant multi-objective hierarchical optimization method can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or preset intermediate form, etc.

[0114] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0115] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0116] The above embodiment is only one of the implementation methods that can realize the technical solution of the present invention. The scope of protection claimed by the present invention is not limited only to this embodiment, but also includes changes, replacements and other implementation methods that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A multi-agent collaborative virtual power plant multi-objective hierarchical optimization method, characterized in that: include: The power grid system is divided into the leading layer, the coordination layer and the equipment layer, and the multi-objective hierarchical control architecture of the virtual power plant of the multi-agent system is obtained; A multi-objective hierarchical optimization model is constructed based on the multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The coordination layer proxy model is solved based on the actual operating parameters of the equipment uploaded by the device layer proxy model to obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

2. The multi-objective hierarchical optimization method of a virtual power plant with multi-agent collaboration according to claim 1 is characterized in that: The power grid system is divided into a leading layer, a coordination layer and a device layer, and a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system is obtained, including: The power grid is divided into the leading layer, the coordination layer and the equipment layer; The leading layer includes the distribution network agent and the heating network agent, which are responsible for interacting with the outside world and issuing strategies to the coordination layer; The coordination layer is composed of multiple regional virtual power plant control center agents, which are responsible for managing local regional resources and exchanging information with the leading layer and the equipment layer; The equipment layer includes energy storage agents, electric load agents, electric vehicle agents, gas turbine agents, fuel cell agents, photovoltaic agents, and wind power agents, which are responsible for monitoring equipment status and adjusting power according to instructions; Each agent in the leading layer, coordination layer and equipment layer is divided into two layers of control structure; the upper layer control receives the operating environment information through the sensor, maps the scene after processing by the event processing distributor, negotiates and cooperates with other agents through the communication system to achieve the optimal goal, selects decisions and formulates action responses in the decision maker, and acts on the virtual power plant through the effector; the lower layer control includes voltage control function and frequency control function to ensure supply and demand balance and safe and stable operation; Build a multi-objective hierarchical control architecture for a virtual power plant with a multi-agent system.

3. The multi-objective hierarchical optimization method of a virtual power plant with multi-agent collaboration according to claim 1 is characterized in that: The device layer proxy model includes a wind power generation model, a controllable distributed power supply model and an energy storage device model; wherein, The specific expression of the wind power generation model is as follows: ； ； ； Where P WT Represents the output active power of the wind turbine; P r represents the rated power of the wind turbine; v represents the wind speed; v ci Indicates the cut-in wind speed; v co Indicates the cut-out wind speed; v r represents the rated wind speed; k1 represents the first coefficient of the relationship between power and wind speed; k2 represents the second coefficient of the relationship between power and wind speed; The controllable distributed power supply model satisfies the following relationship: ； ； ； Where P j t represents the output power of the jth controllable distributed generation at time t; P j t-1 represents the output power of the jth controllable distributed generation at time t-1; P j，min represents the minimum output power of the jth controllable distributed power source; P j，max represents the maximum output power of the jth controllable distributed power source; UR i represents the ramp rate of the jth controllable distributed power source; DR j represents the ramp rate of the jth controllable distributed power source; The specific expression of the energy storage device model is as follows: ； In the energy storage device model, the power constraints and energy constraints of charging and discharging are: ； ； ； In the formula, SOC i t represents the state of charge of the i-th energy storage device at time t; SOC i t+1 represents the charge state of the i-th energy storage device at time t+1; P t SB,i Cap represents the charging and discharging power of the i-th energy storage device at time t; i represents the capacity of the i-th energy storage device; Indicates the lower limit of the charge state of the i-th energy storage device; Indicates the upper limit of the state of charge of the i-th energy storage device; represents the charge state of the i-th energy storage device at time t+T; It means that at the end of a scheduling cycle T, the charge state of the energy storage device is restored to the state at the beginning of the cycle.

4. The multi-objective hierarchical optimization method of a virtual power plant with multi-agent collaboration according to claim 1 is characterized in that: The coordination layer agent model takes minimization of economic cost and environmental cost as the upper layer optimization objectives, and minimization of network loss as the lower layer optimization objective; wherein: Taking economic cost minimization as the upper optimization goal, the specific expression is as follows: ； ； ； In the formula, F1 represents the objective function of minimizing economic cost; t represents different moments, T represents the time set included in the entire scheduling cycle; i represents the index of the power generation equipment; S G Represents the collection of power generation equipment in a virtual power plant, including micro gas turbines, fuel cells and controllable distributed power sources; represents the operation and maintenance cost of the i-th power generation equipment at time t; represents the start / stop status of the i-th power generation equipment at time t; represents the operation and maintenance cost coefficient per unit power of the i-th power generation equipment; represents the output power of the i-th power generation equipment at time t; represents the investment cost of the i-th power generation equipment at time t; represents the investment return rate of the i-th power generation equipment; represents the depreciation period of the i-th power generation equipment; represents the unit capacity investment cost of the i-th power generation equipment; k is the annual utilization hours of the power generation equipment; Taking environmental cost minimization as the upper-level optimization goal, the specific expression is as follows: ； In the formula, F2 represents the environmental cost minimization objective function; j represents the index of the pollutant type, and J is the set of pollutant types; represents the coefficient of the jth pollutant emitted per unit power of the i-th power generation equipment; α represents the environmental impact factor of the j-th pollutant; Taking network loss minimization as the lower-level optimization goal, the specific expression is as follows: ； Where F3 represents the network loss minimization objective function; i, j represents the index of the grid node, i, j∈N, N is the set of grid nodes; G ij represents the conductance between node i and node j; U i represents the voltage amplitude of node i; U j represents the voltage amplitude of node j; θ ij Represents the voltage phase angle difference between node i and node j.

5. The multi-objective hierarchical optimization method of a virtual power plant with multi-agent collaboration according to claim 4 is characterized in that: The coordination layer agent model uses the internal power balance constraint and network security constraint of the virtual power plant as constraint conditions; wherein the network security constraint includes voltage constraint and line capacity constraint; The internal power balance constraint of the virtual power plant is specifically expressed as follows: ； The voltage constraint is specifically expressed as follows: ； The line capacity constraint is specifically expressed as follows: ； In the formula, represents the output power of the i-th power generation equipment, N G Represents a collection of power generation equipment in a virtual power plant; represents the charging and discharging power of the jth energy storage device, represents the collection of energy storage devices; P P Indicates the power purchased by the power plant; P S represents the power sold by the virtual power plant; D represents the load demand of the virtual power plant; P loss Represents the network loss power within the virtual power plant; U min，i Indicates the minimum allowable voltage amplitude of node i; U max，i represents the maximum allowable voltage amplitude of node i; S i represents the apparent power of line i; S min，i represents the minimum permissible apparent power of line i; S max，i Represents the maximum allowable apparent power of line i.

6. A multi-agent collaborative virtual power plant multi-objective hierarchical optimization method according to claim 5, characterized in that: The method of solving the coordination layer proxy model based on the actual operation parameters of the device uploaded by the device layer proxy model includes: The membership function is used to fuzzy each objective function in the coordination layer agent model, and the multi-objective problem is transformed into a single-objective nonlinear optimization problem based on maximizing satisfaction. The actual operating parameters of the equipment uploaded by the device layer proxy model are used to solve the single-objective nonlinear optimization problem based on maximizing satisfaction and obtain the optimal solution.

7. The multi-objective hierarchical optimization method of a virtual power plant with multi-agent collaboration according to claim 6 is characterized in that: The membership function The specific expression is as follows: ； ； ； ； The transformed coordination layer agent model is specifically expressed as follows: ； ； In the formula, , They represent curve characteristic parameters and allowable deviation respectively; , All represent other values ​​of the objective function; represents the adjusted allowable deviation; β represents the minimum value of the membership function of each objective function under the row range constraint.

8. A multi-agent collaborative virtual power plant multi-objective hierarchical optimization system, characterized in that: include: An architecture building module is used to divide the power grid system into a dominant layer, a coordination layer, and an equipment layer, and obtain a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system; An optimization model building module is used to build a multi-objective hierarchical optimization model based on a multi-objective hierarchical control architecture of a virtual power plant of a multi-agent system, wherein the multi-objective hierarchical optimization model includes an equipment layer agent model and a coordination layer agent model; wherein the coordination layer agent model takes minimization of economic cost and minimization of environmental cost as upper-layer optimization objectives, and minimization of network loss as lower-layer optimization objective; and takes internal power balance constraints and network security constraints of the virtual power plant as constraint conditions; The model solving module is used to solve the coordination layer proxy model based on the actual operating parameters of the equipment uploaded by the device layer proxy model, obtain the optimal solution and output the optimal solution as the control strategy to achieve multi-objective hierarchical optimization of the virtual power plant according to the control strategy.

9. A device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the multi-objective hierarchical optimization method of a multi-agent collaborative virtual power plant as described in any one of claims 1-7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it is used to implement the steps of the multi-objective hierarchical optimization method of a multi-agent collaborative virtual power plant as described in any one of claims 1-7.

Images