Comprehensive energy system operation optimization method based on digital twinning and multi-agent

Abstract

The invention relates to an integrated energy system operation optimization method based on digital twinning and multiple agents. The method is suitable for the technical field of comprehensive energy system operation regulation and control. According to the technical scheme, the method comprises the steps of constructing a digital twin model of each energy main body based on an input and output relationship of each energy main body in a target system, and establishing a digital twin model of the target system based on the digital twin model of each energy main body; based on the input, the output and the individual benefit target of each energy main body, constructing an agent model of each energy main body; establishing a virtual verification sand table based on the target system, the digital twin model of each energy main body and the intelligent agent model of each energy main body; and based on the virtual verification sand table, the maximum total benefit of the system is taken as an optimization target, an optimal operation strategy is determined by adopting a heuristic optimization algorithm, and the total benefit of the system is determined based on the individual benefit target value of each energy main body.

Description

Technical Field

[0001] The present invention relates to an operation optimization method for an integrated energy system based on digital twin and multi - agent, which is applicable to the field of operation regulation of integrated energy systems. Background Art

[0002] An integrated energy system can integrate various forms of energy such as heat, cold, electricity, etc. within a region, realize the coordinated optimization and efficient complementarity of various forms of energy flows, meet the multi - aspect energy consumption demands of users within the region for electricity, heat, cold, etc., and improve the overall energy utilization efficiency. Achieving the operation optimization of the integrated energy system is of great significance for improving the accommodation capacity of renewable energy sources such as wind power and photovoltaic power generation, and can effectively promote the economic and reliable operation of the energy system.

[0003] The integrated energy system has characteristics such as large scale, complex structure, and numerous scheduling variables, making it difficult to develop operation optimization strategies, and the actual project operation effects are often not good. Digital twin technology provides a new perspective for realizing the cognition and regulation of the integrated energy system. By constructing a high - fidelity virtual mirror of the integrated energy system in the digital space, it can finely depict the dynamic change process of the integrated energy system in the time and space dimensions, thus helping decision - makers achieve system optimization regulation.

[0004] Through the retrieval of existing patents, including an integrated energy control system based on digital twin (Chinese invention patent, publication number CN117350892A), an integrated energy control method and system based on digital twin (Chinese invention patent, publication number CN\n116014715A), a method for operating an integrated energy system in a park based on digital twin (Chinese invention patent, publication number CN114625022A), a scheduling management method for an integrated energy system based on digital twin (Chinese invention patent, publication number CN115994674A), etc., the above - mentioned patent literatures have all realized the energy management and operation scheduling of the integrated energy system based on digital twin technology. However, in the actual operation process of the integrated energy system, it involves multiple energy entities such as sources, loads, grids, and storages, and different energy entities have different interest goals. How to comprehensively consider the interest demands of various energy entities to achieve the global operation optimization of the integrated energy system still needs further research. Summary of the Invention

[0005] The technical problem to be solved by the present invention is: in view of the above - mentioned problems, to provide an operation optimization method for an integrated energy system based on digital twin and multi - agent.

[0006] The technical solution adopted by the present invention is: an operation optimization method for an integrated energy system based on digital twin and multi - agent, including:

[0007] Based on the input-output relationships of each energy entity in the target system, construct the digital twin models of each energy entity, and based on the digital twin models of each energy entity, establish the digital twin model of the target system;

[0008] Based on the inputs, outputs, and individual interest objectives of each energy entity, construct the agent models of each energy entity;

[0009] Based on the digital twin models of the target system and each energy entity, as well as the agent models of each energy entity, establish a virtual verification sand table;

[0010] Based on the virtual verification sand table, with the maximization of the overall system benefit as the optimization objective, use a heuristic optimization algorithm to determine the optimal operation strategy, where the overall system benefit is determined based on the individual interest objective values of each energy entity;

[0011] The energy entity includes a production capacity entity and an energy consumption entity. The agent model of the production capacity entity can be used to send the input parameter value of the production capacity entity in the operation strategy into the digital twin model of the production capacity entity, and determine the individual interest objective value of the production capacity entity based on the output result of the digital twin model;

[0012] The energy consumption entity can be used to send the input parameter value of the energy consumption entity into the digital twin model of the energy consumption entity, and determine the individual interest objective value of the energy consumption entity based on the output result of the digital twin model;

[0013] The input parameter value of the energy consumption entity is determined based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity.

[0014] The constructing the digital twin models of each energy entity based on the input-output relationships of each energy entity in the target system, and establishing the digital twin model of the target system based on the digital twin models of each energy entity includes:

[0015] Determine the input parameters and output parameters of each energy entity in the target system;

[0016] Describe the mapping relationship between the input parameters and output parameters, and establish the digital twin models of each energy entity;

[0017] According to the topological structure of the target system, determine the input-output relationships between the digital twin models of each energy entity, and establish the digital twin model of the target system.

[0018] The agent model of each energy entity includes a knowledge set, an action set, and a goal set;

[0019] Among them, the knowledge set contains the self - characteristic parameters of the energy entity; the action set contains actions for the energy entity, including extracting the input parameters corresponding to the energy entity and inputting the input parameters into the corresponding digital twin model of the energy entity; the target set contains a target value calculation model, and this calculation model calculates the individual interest target value based on the self - characteristic parameters of the energy entity and the output results of the digital twin model.

[0020] Based on the virtual verification sand table, with the maximum overall system benefit as the optimization goal, a heuristic optimization algorithm is used to determine the optimal operation strategy, including:

[0021] Obtain the preset operation strategy set, and each operation strategy individual in the set contains the input parameter values of each energy - producing entity in the target system;

[0022] The intelligent agent model of the energy - producing entity extracts the input parameter value of this energy - producing entity from the operation strategy individual and inputs the input parameter value into the digital twin model of this energy - producing entity to obtain the output result;

[0023] Based on the digital twin model of the target system and the output results of the digital twin models of each energy - producing entity, determine the input parameter values of each energy - consuming entity;

[0024] Input the input parameter values of the energy - consuming entity into the digital twin model of this energy - consuming entity to obtain the output result;

[0025] The intelligent agent model of the energy entity determines the individual interest target value of this energy - producing entity based on the output result of this energy entity's digital twin model;

[0026] Based on the individual interest target values of each energy - producing entity in the target system, determine the overall system benefit of the target system;

[0027] Calculate the fitness function of the operation strategy individual based on the overall system benefit, and perform iterative calculations based on the fitness function to determine the optimal operation strategy.

[0028] The overall system benefit is determined based on the individual interest target values of each energy entity, including:

[0029]

[0030] Among them, N represents the number of energy entities in the target system, f i represents the individual interest target value of energy entity i, and f ies represents the overall system benefit of the target system.

[0031] An integrated energy system operation optimization device based on digital twin and multi - intelligent agents, including:

[0032] A twin model construction module, which is used to construct digital twin models of each energy entity based on the input-output relationships of each energy entity in the target system, and establish a digital twin model of the target system based on the digital twin models of each energy entity;

[0033] An agent construction module, which is used to construct agent models of each energy entity based on the inputs, outputs and individual interest objectives of each energy entity;

[0034] A verification sand table establishment module, which is used to establish a virtual verification sand table based on the digital twin models of the target system and each energy entity, and the agent models of each energy entity;

[0035] A strategy optimization module, which is used to determine the optimal operation strategy based on the virtual verification sand table with the maximum overall system benefit as the optimization goal, and adopt a heuristic optimization algorithm, where the overall system benefit is determined based on the individual interest objective values of each energy entity;

[0036] The energy entity includes a production capacity entity and an energy consumption entity. The agent model of the production capacity entity can be used to send the input parameter value of the production capacity entity in the operation strategy into the digital twin model of the production capacity entity, and determine the individual interest objective value of the production capacity entity based on the output result of the digital twin model;

[0037] The energy consumption entity can be used to send the input parameter value of the energy consumption entity into the digital twin model of the energy consumption entity, and determine the individual interest objective value of the energy consumption entity based on the output result of the digital twin model;

[0038] The input parameter value of the energy consumption entity is determined based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity.

[0039] A storage medium, on which a computer program executable by a processor is stored. When the computer program is executed, the steps of the comprehensive energy system operation optimization method based on digital twin and multi-agent are implemented.

[0040] An integrated energy system operation optimization device, which has a memory and a processor. A computer program executable by the processor is stored on the memory. When the computer program is executed, the steps of the comprehensive energy system operation optimization method based on digital twin and multi-agent are implemented.

[0041] The beneficial effects of the present invention are as follows: The present invention uses digital twin and multi-agent technologies to establish a virtual rehearsal sand table of the target integrated energy system, conducts simulations through digital twin models, calculates the individual interest objective values of each energy entity in the system based on the simulation results through agent models, and calculates the overall system benefit. The present invention can finely verify the actual operation effects of different operation strategies, and provide a scientific and objective theoretical basis for the operation scheduling of the system.

[0042] The present invention uses a heuristic optimization algorithm to achieve global optimization of the optimal operation plan of the integrated energy system, which can quickly and accurately search for the optimal operation strategy from a large number of alternative operation strategy combinations, has high optimization accuracy and calculation efficiency, and effectively improves the economy of the integrated energy system operation. Description of the Drawings

[0043] Figure 1 It is a schematic diagram of the integrated energy system in the embodiment of the present invention.

[0044] Figure 2 It is a flowchart of the optimal operation method of the integrated energy system in the embodiment of the present invention.

[0045] Figure 3 It is a result comparison of the overall benefits of the integrated energy system in the embodiment of the present invention. Detailed Embodiments

[0046] To better understand the technical solution of the present application, the embodiments of the present application will be described in detail below with reference to the drawings.

[0047] It should be clear that the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0048] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms of "a", "the" and "said" used in the embodiments of the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0049] Figure 1 It is a schematic diagram of the structure of the target integrated energy system (hereinafter referred to as the target system) in this embodiment. The target system is composed of a power subsystem and a thermal subsystem. On the power side, a combined heat and power unit, a thermal power generation unit and the superior power grid jointly supply power to the end electrical load; on the thermal side, the waste heat generated by the combined heat and power unit during power generation and the gas-fired electric boiler jointly supply heat to the end thermal load.

[0050] As Figure 2 shown, this embodiment provides an integrated energy system operation optimization method based on digital twin and multi-agent, which specifically includes the following steps:

[0051] S100. Based on the input-output relationships of the energy entities in the target system, construct digital twin models of the energy entities, and based on the digital twin models of the energy entities, establish a digital twin model of the target system.

[0052] In this embodiment, the energy entities include energy production entities and energy consumption entities. Among them, the energy production entities include thermal power generation units, combined heat and power units, gas boilers, and the superior power grid, and the energy consumption entities include end loads.

[0053] S110. Determine the input parameters and output parameters of each energy entity in the target system as shown in the following table:

[0054]

[0055] S120. Describe the mapping relationship between the input parameters and output parameters using mathematical equations, and establish the digital twin models of each energy entity, specifically including the following models:

[0056] The digital twin model of the thermal power generation unit is shown as follows:

[0057]

[0058] Among them, the superscript t represents time, and a pgu -e pgu is the power coefficient, which are taken as -14.5, 77.9, -145.3, 136.2, and 142.8 respectively in this embodiment.

[0059] The digital twin model of the combined heat and power unit is shown as follows:

[0060]

[0061] Among them, λ gas is the calorific value of natural gas, η chp is the power generation efficiency of the unit, and η loss is the heat loss rate of the unit, which are taken as 9.7 kWh / m 3 0.35, and 0.1 respectively in this embodiment.

[0062] The digital twin model of the gas boiler is shown as follows:

[0063]

[0064] Among them, η gb is the heat production efficiency of the electric boiler, which is taken as 0.75 in this embodiment.

[0065] In this embodiment, selling electricity to the superior power grid is not allowed. Therefore, there are no input parameters in the digital twin model of the superior power grid, and the electric power P pu is directly selected as the output.

[0066] The digital twin model of the end load is shown as follows:

[0067]

[0068] Among them, a P , b P are coefficients of electricity consumption comfort, and a H , b H are coefficients of heat consumption comfort. In this embodiment, 0.26, 0.002, 0.15, and 0.002 are taken respectively.

[0069] It should be particularly noted that in this embodiment, each energy entity uses a basic steady-state mathematical model for mathematical modeling. In practice, complex dynamic models, etc. can also be used to simulate the mapping relationship between the input and output of the energy entity.

[0070] S130. According to the topological structure of the target system, determine the input-output relationship between the digital twin models of each energy entity, and establish the digital twin model of the target system.

[0071] In this embodiment, the sum of the output P pu of the superior power grid, the output P pgu of the thermal power generation unit, and the output P chp of the combined heat and power unit is used as the input P el of the end load, and the sum of the output H gb of the gas boiler and the output H chp of the combined heat and power unit is used as the input H hl of the end load, so as to realize the association between the models of each energy entity and form the digital twin model of the entire system.

[0072] S200. Based on the input, output, and individual interest goals of each energy entity, construct the agent model of each energy entity.

[0073] The agent model of each energy entity in the target system consists of three parts: knowledge, action, and objective, as shown in the following formula:

[0074] Agent = <Knowledge, Action, Objective>

[0075] Among them, Knowledge is a set of knowledge, used to represent the inherent attribute characteristics of the agent, including the self-characteristic parameters of the corresponding energy entity; Action is a set of actions, used to represent the energy consumption and production behaviors that the agent can execute, including extracting the input parameters corresponding to the energy entity and inputting the input parameters into the digital twin model of the corresponding energy entity; Objective is a set of goals, used to represent the individual interest goals that the agent hopes to achieve, and has a goal value calculation model, which calculates the individual interest goal value based on the self-characteristic parameters of the energy entity and the output results of the digital twin model.

[0076] In this embodiment, the agent model of the production capacity entity can be used to send the input parameter values of the production capacity entity in the operation strategy into the digital twin model of the production capacity entity, and determine the individual interest target value of the production capacity entity based on the output result of the digital twin model.

[0077] In this example, the energy-consuming entity can be used to send the input parameter values of the energy-consuming entity into the digital twin model of the energy-consuming entity, and determine the individual interest target value of the energy-consuming entity based on the output result of the digital twin model; the input parameter values of the energy-consuming entity are determined based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity.

[0078] This embodiment specifically includes the following agent models:

[0079] The agent model of the thermal power generating unit is shown as follows:

[0080]

[0081] Among them, the knowledge set is the cost coefficients α1-α3 of the thermal power generating unit; the action set is the fuel input amount F of the thermal power generating unit pgu ; the target set is the total operating cost of the thermal power generating unit within the scheduling time period 1-T, which is described by a quadratic polynomial function in this embodiment.

[0082] The agent model of the combined heat and power unit is shown as follows:

[0083]

[0084] Among them, the knowledge set is the cost coefficients β1-β6 of the combined heat and power unit; the action set is the fuel input amount F of the combined heat and power unit chp ; the target set is the total operating cost of the combined heat and power unit within the scheduling time period 1-T, which is described by a multivariate quadratic polynomial function in this embodiment.

[0085] The agent model of the superior power grid is shown as follows:

[0086]

[0087] Among them, the knowledge set is the grid power purchase price ξ e ; the action set is the power purchase amount P from the power grid pu ; the target set is the power purchase cost of purchasing electricity from the power grid within the scheduling time period 1-T.

[0088] The agent model of the gas boiler is shown as follows:

[0089]

[0090] Among them, the knowledge set is the cost coefficients γ1 and γ2 of the gas boiler; the action set is the fuel input F of the gas boiler gb ; the target set is the operating cost of the gas boiler within the scheduling time period 1 - T

[0091] The agent model of the end - load is shown in the following formula

[0092]

[0093] Among them, the knowledge set is the satisfaction coefficients ω P 、ω H of the end - use electricity and heat; in this embodiment, the active participation of the end - load in the demand - side response is not considered, and the actual load supply is jointly determined by the other agents, so the action set is empty; the target set is the overall energy - use satisfaction of the end - load within the scheduling time period 1 - T. The lower the energy - use satisfaction of the end - user, the smaller this value, and the lower the overall benefit of the system, so as to punish the control effect of this operation strategy

[0094] S300. Based on the digital twin models of the target system and each energy entity, as well as the agent models of each energy entity, establish a virtual verification sand table

[0095] S400. Based on the virtual verification sand table, with the maximum overall benefit of the system as the optimization goal, use a heuristic optimization algorithm to determine the optimal operation strategy, where the overall benefit of the system is determined based on the individual benefit target values of each energy entity

[0096] S410. Obtain the preset operation strategy set, which contains multiple operation strategy individuals, and each operation strategy individual contains the input parameter values of each production capacity entity in the target system at each moment within the preset time period

[0097] In this embodiment, the form of the operation strategy individual is shown in the following formula

[0098]

[0099] Among them, the input parameter values of the thermal power generation unit, the combined heat and power unit, the superior power grid, and the gas boiler at each moment are successively and where the superscripts 1, 2,..., T represent moments

[0100] S420. The agent model of the production capacity entity extracts the input parameter values of this production capacity entity from the operation strategy individual, and inputs the input parameter values into the digital twin model of this production capacity entity to obtain the output result

[0101] In this embodiment, the output of the digital twin model of the thermal power generation unit is The output of the combined heat and power unit model is and The output of the upper-level power grid model is consistent with the action set of the corresponding intelligent agent model, that is, The output of the gas boiler model is

[0102] S430. Based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity, determine the input parameter values of each energy-consuming entity.

[0103] In this embodiment, the input of the terminal load digital twin model is as follows:

[0104]

[0105] S440: Input the input parameter values of the energy-consuming subject into the digital twin model of the energy-consuming subject to obtain an output result.

[0106] S450. The intelligent agent model of each energy entity in the target system determines the individual benefit target value of the production capacity entity based on the output result of the digital twin model of the energy entity.

[0107] The intelligent agent model of each energy entity uses its target value calculation model to calculate the individual interest target value of each energy entity based on its own characteristic parameters and the output results of the digital twin model.

[0108] S460. Determine the overall system benefit of the target system based on the individual benefit target values of each production capacity entity in the target system.

[0109] In this embodiment, the individual benefit target values of each energy subject intelligent agent model are summed up to calculate the overall system benefit of the target system under the individual operation strategy, as shown in the following formula:

[0110]

[0111] Where N represents the number of energy entities in the target system, f i represents the individual interest target value of energy subject i, f ies Represents the overall system benefit of the target system.

[0112] S470: Calculate the fitness function of the individual operation strategies based on the overall benefits of the system, and perform iterative calculations based on the fitness function to determine the optimal operation strategy.

[0113] The following example uses the particle swarm algorithm to illustrate the optimization of the operation strategy. The specific steps are as follows:

[0114] A1: Set the particle swarm size Size = 100, inertia weight ω = 0.8, acceleration factor c1 = c2 = 2, and the upper limit of the number of iterations T max= 150 and counter t = 0;

[0115] A2: Encode the operation strategy as a position vector, and randomly initialize the position vector x of each particle within the operation strategy search space i and velocity vector v i , as shown in the following formula:

[0116]

[0117] v i (0) = [v i,1 , v i,2 ,..., v i,4T

[0118] A3: Input the operation strategy encoded by each particle's position vector into the virtual sandbox composed of digital twins and multi - agents, calculate the overall benefit target value of the system under this operation strategy, and calculate the fitness function of this particle accordingly, as shown in the following formula:

[0119] fiteness i = - f ies

[0120] where fiteness i is the fitness function of particle i;

[0121] A4: Update the velocity vector v i (t + 1) of each particle, as shown in the following formula:

[0122] v i (t + 1) = ω·v i (t) + c1·r1·[p i (t) - x i (t)] + c2·r2·[g(t) - x i (t)]

[0123] where r1 and r2 are random numbers in the interval [0, 1], p i is the position vector corresponding to the minimum fitness function of particle i during the entire iteration process, and g is the position vector corresponding to the minimum fitness function of all particles during the entire iteration process;

[0124] A5: Update the position vector x i (t + 1) of each particle, as shown in the following formula:

[0125] x i,j (t + 1) = x i,j (t) + v i,j (t + 1)

[0126] ​After that, set t = t + 1;

[0127] A6: Determine whether the iterative process converges, as shown in the following formula:

[0128] t≥T max or |g(t) - g(t - 1)| ≤ ε

[0129] where ε is the tolerance threshold, and in this embodiment, ε is taken as 0.1. If the convergence condition is met, the algorithm stops and outputs the operation strategy represented by g(t). Otherwise, return to step A3.

[0130] To verify the optimization performance of the method of the present invention, a traditional heat - based power generation operation method is set for comparison. In this method, the combined heat and power unit preferentially satisfies the supply of heat load, and the insufficient power load is supplied by the thermal power generation unit and the superior power grid. The energy storage device charges during off - peak hours and discharges during peak hours. The total optimization time length T is set to 24, and the optimization interval is set to 1 h.

[0131] The results of the overall benefits of the integrated energy system in the embodiments under the two methods are compared as Figure 3 shown. In the traditional heat - based power generation method, the combined heat and power unit follows the fluctuations of the heat load to supply the heat load. Most of the insufficient power load is supplied by the superior power grid and the thermal power generation unit. At the same time, the charging and discharging logic of the energy storage device is fixed, and the end - user load does not participate in the demand - side response, resulting in a lack of energy coordination and complementarity during periods with high power load and low heat load, causing large energy waste and low total system revenue. In contrast, the method of the present invention can search for the optimal operation strategy of the system by means of game - theoretic optimization in the virtual rehearsal sandbox. The daily overall revenue of the system has increased from $4040.0 to $4859.8, an increase of 20.3%.

[0132] This embodiment also provides an integrated energy system operation optimization device based on digital twin and multi - agent, including:

[0133] A twin model construction module, configured to construct a digital twin model of each energy entity based on the input - output relationship of each energy entity in the target system, and establish a digital twin model of the target system based on the digital twin models of each energy entity;

[0134] An agent construction module, configured to construct an agent model of each energy entity based on the input, output, and individual interest objectives of each energy entity;

[0135] A verification sandbox establishment module, configured to establish a virtual verification sandbox based on the digital twin models of the target system and each energy entity, and the agent models of each energy entity;

[0136] A strategy optimization module, which is used to determine the optimal operation strategy based on a virtual verification sand table with the maximum overall system benefit as the optimization goal, and the overall system benefit is determined based on the individual benefit target values of each energy entity.

[0137] This embodiment also provides a storage medium, on which a computer program executable by a processor is stored. When the computer program is executed, the steps of the above-mentioned integrated energy system operation optimization method based on digital twin and multi-agent are realized.

[0138] This embodiment also provides an integrated energy system operation optimization device, which has a memory and a processor. A computer program executable by the processor is stored on the memory. When the computer program is executed, the steps of the above-mentioned integrated energy system operation optimization method based on digital twin and multi-agent are realized.

[0139] The above embodiments are only a preferred solution of the present invention, but they are not intended to limit the present invention. Those of ordinary skill in the relevant technical field can still make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all technical solutions obtained by adopting equivalent replacement or equivalent transformation fall within the protection scope of the present invention.

Claims

1. An operation optimization method for an integrated energy system based on digital twin and multi-agent, characterized in that Including: Based on the input-output relationships of each energy entity in the target system, construct digital twin models for each energy entity, and based on the digital twin models of each energy entity, establish a digital twin model of the target system; Based on the inputs, outputs, and individual interest objectives of each energy entity, construct an agent model for each energy entity; Based on the digital twin models of the target system and each energy entity, as well as the agent models of each energy entity, establish a virtual verification sand table; Based on the virtual verification sand table, with the maximization of the overall system benefit as the optimization objective, use a heuristic optimization algorithm to determine the optimal operation strategy, where the overall system benefit is determined based on the individual interest objective values of each energy entity; The energy entity includes a production capacity entity and an energy consumption entity. The agent model of the production capacity entity can be used to send the input parameter value of the production capacity entity in the operation strategy into the digital twin model of the production capacity entity, and determine the individual interest objective value of the production capacity entity based on the output result of the digital twin model; The energy consumption entity can be used to send the input parameter value of the energy consumption entity into the digital twin model of the energy consumption entity, and determine the individual interest objective value of the energy consumption entity based on the output result of the digital twin model; The input parameter value of the energy consumption entity is determined based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity.

2. The operation optimization method of the integrated energy system based on digital twin and multi-agent according to claim 1, wherein The constructing digital twin models for each energy entity based on the input-output relationships of each energy entity in the target system, and establishing a digital twin model of the target system based on the digital twin models of each energy entity includes: Determine the input parameters and output parameters of each energy entity in the target system; Describe the mapping relationship between the input parameters and output parameters, and establish digital twin models for each energy entity; According to the topological structure of the target system, determine the input-output relationships between the digital twin models of each energy entity, and establish a digital twin model of the target system.

3. The operation optimization method of the integrated energy system based on digital twin and multi-agent according to claim 1, wherein: The agent model of each energy entity includes a knowledge set, an action set, and a target set; Among them, the knowledge set contains the self-characteristic parameters of the energy entity; the action set contains actions for the energy entity, including extracting the input parameters corresponding to the energy entity and inputting the input parameters into the corresponding digital twin model of the energy entity; the target set contains a target value calculation model, and this calculation model calculates the individual interest objective value based on the self-characteristic parameters of the energy entity and the output result of the digital twin model.

4. The operation optimization method of the integrated energy system based on digital twin and multi-agent according to claim 1, characterized in that The using the heuristic optimization algorithm to determine the optimal operation strategy based on the virtual verification sand table with the maximization of the overall system benefit as the optimization objective includes: Obtain a preset set of operation strategies, and each operation strategy individual in the set contains the input parameter values of each production capacity entity in the target system; The agent model of the production capacity entity extracts the input parameter value of the production capacity entity from the operation strategy individual, and inputs the input parameter value into the digital twin model of the production capacity entity to obtain an output result; Based on the digital twin model of the target system and the output results of the digital twin models of each production capacity entity, determine the input parameter values of each energy consumption entity; Input the input parameter value of the energy consumption entity into the digital twin model of the energy consumption entity to obtain an output result; Based on the output results of the digital twin model of the energy entity, the agent model of the energy entity determines the individual interest target value of the energy production entity; Based on the individual interest target values of each energy production entity in the target system, the overall system interest of the target system is determined; Based on the overall system interest, the fitness function of the operation strategy individual is calculated, and iterative calculation is performed based on the fitness function to determine the optimal operation strategy.

5. The operation optimization method of the integrated energy system based on digital twin and multi-agent according to claim 1, characterized in that The overall system interest is determined based on the individual interest target values of each energy entity, including: Among them, N represents the number of energy entities in the target system, and f i represents the individual interest target value of energy entity i, and f ies represents the overall system interest of the target system.

6. An integrated energy system operation optimization device based on digital twin and multi-agent, characterized in that, Including: The digital twin model construction module is used to construct the digital twin models of each energy entity based on the input-output relationships of each energy entity in the target system, and establish the digital twin model of the target system based on the digital twin models of each energy entity; The agent construction module is used to construct the agent models of each energy entity based on the inputs, outputs, and individual interest targets of each energy entity; The verification sand table establishment module is used to establish a virtual verification sand table based on the digital twin models of the target system and each energy entity, as well as the agent models of each energy entity; The strategy optimization module is used to determine the optimal operation strategy based on the virtual verification sand table with the maximization of the overall system interest as the optimization goal, where the overall system interest is determined based on the individual interest target values of each energy entity; The energy entity includes an energy production entity and an energy consumption entity. The agent model of the energy production entity can be used to send the input parameter values of the energy production entity in the operation strategy into the digital twin model of the energy production entity, and determine the individual interest target value of the energy production entity based on the output results of the digital twin model; The energy consumption entity can be used to send the input parameter values of the energy consumption entity into the digital twin model of the energy consumption entity, and determine the individual interest target value of the energy consumption entity based on the output results of the digital twin model; The input parameter values of the energy consumption entity are determined based on the digital twin model of the target system and the output results of the digital twin models of each energy production entity.

7. A storage medium having stored thereon a computer program executable by a processor, characterized in that: When the calculation and program are executed, the steps of the integrated energy system operation optimization method based on digital twin and multi-agent as described in any one of claims 1 to 5 are implemented.

8. An integrated energy system operation optimization device, having a memory and a processor, and a computer program capable of being executed by the processor is stored on the memory, characterized in that: When the calculation and program are executed, the steps of the integrated energy system operation optimization method based on digital twin and multi-agent as described in any one of claims 1 to 5 are implemented.

Images