A cost optimization method for integrated energy systems

Abstract

This invention discloses a cost optimization method for integrated energy systems, comprising: S1, constructing a multi-objective optimization scheduling model for the integrated energy system based on the requirements of minimizing economic operating costs, minimizing carbon emissions, and maximizing efficiency; S2, extending the traditional single-variable LAP adjustment method and proposing a multivariate LAP adjustment model; S3, extending nodal marginal electricity prices and convex hull pricing to the integrated energy system, and convexizing the original system model; S4, performing simulation examples in the integrated energy system to solve the constructed multi-objective optimization scheduling model, and further solving for nodal energy prices in the integrated energy system based on the convex hull pricing method. This invention solves the problem that existing integrated energy systems cannot achieve minimizing economic operating costs, minimizing carbon emissions, and maximizing efficiency during the optimization scheduling process.

Description

Technical Field

[0001] This invention belongs to the field of energy system cost optimization technology, and more specifically, relates to a cost optimization method for integrated energy systems. Background Technology

[0002] In recent years, with the diversification of energy structures and the introduction of the "dual carbon" target, Integrated Energy Systems (IES), as an important means to achieve efficient energy utilization and energy conservation and emission reduction, have received increasing attention for their optimal scheduling and pricing. Traditional scheduling methods based on a single economic objective can no longer meet the development needs of modern energy systems. There is an urgent need to construct a multi-objective optimal scheduling model that comprehensively considers economics, environmental protection, and energy efficiency, and combines it with an effective pricing mechanism to guide the optimal allocation of resources within the system.

[0003] Existing integrated energy pricing models primarily employ two approaches. The first involves centralizing the sales rights of multiple energy sources under a single entity, uniformly allocating sales across different energy types. This model's advantage lies in its ability to effectively achieve complementary utilization among multiple energy sources. However, while this approach, which relies on a decision-maker with overall control over the production and consumption of multiple energy sources, is feasible for localized integrated energy market optimization, a globally centralized and optimized unified pricing model is overly idealistic and suffers from several problems: First, it requires disruptive institutional reforms, necessitating the establishment of a unified "integrated energy dispatch center," which incurs high costs. Second, the market-wide centralized clearing and optimization model is overly complex, resulting in high market operating costs, low transparency, significant technical difficulties, and challenges in regulation. Third, centralized optimization requires detailed information from market participants, increasing transaction costs and creating the risk of information asymmetry and reduced market efficiency. The second approach involves independent energy operators, managed indirectly through an energy market model. In this energy-sharing model among multiple integrated energy systems, multi-energy transactions exhibit multi-entity characteristics, rendering the original centralized energy management system inapplicable. Many scholars have attempted to address energy pricing in multi-energy, multi-entity integrated energy systems using distributed energy management methods based on decomposition theory. The centralized optimization problem is decomposed into a series of subproblems belonging to controllable resources, and a master problem is introduced to guide the subproblems to find the optimal solution through multiple iterations. However, this type of algorithm has poor robustness in unstable situations such as communication delays. It also pays little attention to the possibility that each subject may pursue its own interests to maximize its own interests, resulting in an inefficient equilibrium state of disorderly competition, which will limit the potential of the system.

[0004] The shortcoming of existing technologies lies in the fact that current integrated energy systems lack the ability to minimize economic operating costs, carbon emissions, and other factors during the optimal scheduling process. Efficiency-maximizing scheduling methods. Summary of the Invention

[0005] The existing integrated energy system cannot achieve the minimization of economic operating costs and carbon emissions during the optimization and scheduling process. To address the issue of maximizing efficiency, this invention provides a cost optimization method for integrated energy systems.

[0006] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0007] A cost optimization method for integrated energy systems includes the following steps:

[0008] S1, based on minimizing economic operating costs, minimizing carbon emissions, and... The most efficient requirement is to construct a multi-objective optimization scheduling model for the integrated energy system;

[0009] S2. Extend the traditional univariate LAP adjustment method and propose a multivariate LAP adjustment model;

[0010] S3. Extend the nodal marginal electricity price and convex hull pricing method to the integrated energy system and convexize the original system model.

[0011] S4. Perform simulation examples in the integrated energy system to solve the multi-objective optimization scheduling model of the integrated energy system, and further solve the node energy price in the integrated energy system based on the convex hull pricing method.

[0012] Furthermore, the detailed steps of step S1 include:

[0013] S101. Construct an economic operating cost objective function for an integrated energy system. Further consider the adjustment costs for the uncertainty of renewable energy and cooling, heating and power load demand. Set the objective function to minimize the expected adjustment cost under the worst-case distribution within the fuzzy set.

[0014] S102. Calculate the carbon emissions generated within the optimized operation system of the integrated energy system.

[0015] S103, Construction The objective function for maximizing efficiency is transformed into a linear programming problem.

[0016] Furthermore, the detailed steps of step S2 include:

[0017] S201. Preset that electric energy storage, thermal energy storage, micro gas turbine, waste heat boiler, gas boiler, absorption chiller, etc. meet the normal operating conditions and carry out the first stage of operation constraints;

[0018] S202. The traditional univariate LAP adjustment method is extended and a multivariate LAP adjustment model is proposed.

[0019] S203, Adjust and constrain the quantities of gas turbines, waste heat boilers and gas boilers, electric chillers and absorption chillers, multi-energy storage, and interactions with the upper-level power grid.

[0020] Furthermore, the detailed steps of step S3 include:

[0021] S301, the non-convex feasible region (∪) of the unit's start-up and shutdown states. n∈ψ Ω) is transformed into its convex hull conv(∪) n∈ψ Ω), to convexize the original system model;

[0022] S302. Relax all 0-1 variables representing the unit's operating status to the interval [0,1], making them continuous variables.

[0023] Furthermore, the detailed steps of step S4 include:

[0024] S401. Obtain network parameters in the natural gas system, parameters in the thermal network, and relevant parameters of gas turbines, gas boilers, electric boilers, and energy storage devices in the integrated energy system. The simulation example is programmed using YALMIP on the MATLAB platform and solved using the Gurobi solver.

[0025] S402. The AE-GAN model is used to further generate prediction error data for each nondeterministic variable in the integrated energy system. The correlation between the actual prediction error scenarios and the generated scenarios will be calculated. Then, the t-distributed random neighborhood embedding (t-SNE) method and principal component analysis (PCA) are used for dimensionality reduction to obtain the correlation between the prediction error data of the actual nondeterministic variables and the generated data.

[0026] S403. Based on the system parameters set above, solve the AE-GAN-based integrated energy system partial brook opportunity optimization model constructed in this paper.

[0027] S404. Based on the convex hull pricing method, further solve the nodal energy prices in the integrated energy system, namely nodal electricity price, nodal gas price, nodal heat price and nodal cooling price.

[0028] Furthermore, in step S101, the objective function for the economic operating cost of the integrated energy system is as follows:

[0029]

[0030] In the formula, F1 represents the economic operating cost of the integrated energy system, F2 represents the cost of purchasing electricity from the external power grid, and F3 represents the cost of purchasing natural gas from the natural gas market. This represents the unit cost of purchasing electricity from an external power grid. This indicates the amount of electricity purchased from the external power grid. This represents the unit cost of selling electricity to the external power grid (i.e., the grid connection price). This refers to the amount of electricity sold to the external power grid.

[0031] The operating cost of the integrated energy system should be:

[0032]

[0033] In the second phase (real-time phase), the non-determinism of renewable energy output and electrical, thermal, and cooling loads is fully considered. To avoid corresponding forecast errors (e.g., Random disturbances in the integrated energy system cause further adjustments to the power / natural gas ratio of flexible resources (operational rescheduling), thus incurring adjustment costs.

[0034] Furthermore, in step S102, the following is introduced: Efficiency serves as an indicator for evaluating the utilization rate of resources in an energy system.

[0035] The objective function for maximizing efficiency is as follows:

[0036]

[0037] In the formula, ψ e ψ q and ψ c These represent the energy quality coefficients for electrical energy, thermal energy, and cold energy, respectively.

[0038] By employing the Charnes-Cooper transformation, this problem is converted into a linear programming problem. Considering the operational states of each component in the integrated energy system, the flow of various energy sources, such as cooling, heating, and electricity, can be optimized, ultimately improving the energy efficiency of the integrated energy system.

[0039] Further, in step S202, given the corresponding random variable and The corresponding prediction error and It can be represented as:

[0040]

[0041] In the formula, and These represent the actual data for wind power output, photovoltaic power output, electrical load, heat load, and cooling load, respectively.

[0042] The traditional univariate LAP adjustment method is extended, and a multivariate LAP adjustment model is proposed as follows:

[0043]

[0044] therefore, and These represent all adjustment rates (variables) corresponding to different prediction errors. Furthermore, They represent the decision variables respectively. The adjustment results are shown. The model will be extended by combining conventional constraints and multi-energy flexibility constraints, thus completing the sub-Bruker optimization model of the integrated energy system.

[0045] Furthermore, in step S302, an approximate description method for the convex hull, namely the integer relaxation pricing method, is introduced. This method relaxes all 0-1 variables representing the unit's operating state to the interval [0,1], making them continuous variables.

[0046] Further, in step S401, the distribution network in this integrated energy system is an improved IEEE-9 node power system, the heating / cooling network is based on an improved Bali 6-node heating system, and the natural gas network uses a 7-node system. The organic coupling of these three energy systems forms the integrated energy node system described in this paper. Based on the annual average load demand forecast data and related parameters of the integrated energy system, and according to the electricity trading price, programming is performed using YALMIP on the MATLAB platform, and the Gurobi solver is called for solution.

[0047] Furthermore, in step S403, an ocean hunting algorithm is used to handle the multi-objective optimization problem. The scheduling method based on the multi-objective genetic optimization algorithm increases the operating cost by 7.7% and carbon emissions by 8.0% compared to the method presented in this paper. Efficiency decreased by 9.7%; the scheduling method based on multi-objective particle swarm optimization algorithm increased the operating cost by 3.5% compared to the method in this paper, but reduced carbon emissions by 2.5%. Efficiency decreased by 9.3%; the scheduling method based on decomposition-based multi-objective evolutionary algorithms resulted in a 6.4% increase in operating cost and a 3.0% increase in carbon emissions compared to the method presented in this paper. Efficiency decreased by 8.3%. Therefore, the optimal trade-off solution proposed in this paper based on the multi-objective marine predator algorithm suffers from lower operating costs, lower carbon emissions, and lower efficiency. It has a greater advantage in three aspects, including efficiency.

[0048] For the nondeterministic factors in the integrated energy system, a comparative analysis will be conducted with robust optimization and stochastic optimization models.

[0049] Compared with the prior art, the present invention has the following advantages:

[0050] By constructing the model and drawing an analogy to the relevant definition of the nodal marginal electricity price package pricing method, the model is extended to the integrated energy system. Then, through numerical simulation, the effectiveness of the model in reducing economic operating costs is verified.

[0051] This model constructs a multi-objective optimization scheduling model based on distributed bars, simultaneously considering economic operating costs, carbon emissions, and... Compared to single-objective optimization models, the three efficiency objectives provide a more comprehensive balance between various economic and environmental indicators of system operation. Furthermore, by extending the traditional univariate LAP adjustment method, a multivariate LAP adjustment model is proposed, which can more flexibly address the uncertainties of renewable energy and load demand, and reduce adjustment costs caused by forecast errors.

[0052] When calculating the carbon emissions generated by the optimized operation of an integrated energy system, the model uses it as an optimization objective. By adjusting the operating status of various devices and energy allocation within the system, it minimizes carbon emissions. The model can flexibly address the uncertainties of renewable energy and load demand, reducing additional carbon emissions caused by prediction errors through strategy adjustments. It considers the flow and conversion of various energy sources, including electricity, heat, and cooling, within the integrated energy system, achieving multi-energy complementarity through optimized scheduling and improving overall energy efficiency. Compared to traditional energy systems, this model has significant advantages in reducing carbon emissions and contributes to promoting the green transformation of energy systems. Attached Figure Description

[0053] Figure 1 This is an overall flowchart of a method for optimizing the price of an integrated energy system according to an embodiment of the present invention;

[0054] Figure 2 This is a schematic diagram of the integrated energy system nodes in an embodiment of the present invention;

[0055] Figure 3 This is a comparison chart of the cumulative probability distribution of real data and generated data in an embodiment of the present invention. Detailed Implementation

[0056] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0057] like Figure 1 As shown, this embodiment provides a cost optimization method for integrated energy systems.

[0058] S1, based on minimizing economic operating costs, minimizing carbon emissions, and... The most efficient requirement is to construct a multi-objective optimization scheduling model for the integrated energy system;

[0059] S2. Extend the traditional univariate LAP adjustment method and propose a multivariate LAP adjustment model;

[0060] S3. Extend the nodal marginal electricity price and convex hull pricing method to the integrated energy system and convexize the original system model.

[0061] S4. Perform simulation examples in the integrated energy system to solve the multi-objective optimization scheduling model of the integrated energy system, and further solve the node energy price in the integrated energy system based on the convex hull pricing method.

[0062] The detailed steps of step S1 include:

[0063] S101. Construct an economic operating cost objective function for an integrated energy system. Further consider the adjustment costs for the uncertainty of renewable energy and cooling, heating and power load demand. Set the objective function to minimize the expected adjustment cost under the worst-case distribution within the fuzzy set.

[0064] S102. Calculate the carbon emissions generated within the optimized operation system of the integrated energy system.

[0065] S103, Construction The objective function for maximizing efficiency is transformed into a linear programming problem.

[0066] The detailed steps of step S2 include:

[0067] S201. Preset that electric energy storage, thermal energy storage, micro gas turbine, waste heat boiler, gas boiler, absorption chiller, etc. meet the normal operating conditions and carry out the first stage of operation constraints;

[0068] S202. The traditional univariate LAP adjustment method is extended and a multivariate LAP adjustment model is proposed.

[0069] S203, Adjust and constrain the quantities of gas turbines, waste heat boilers and gas boilers, electric chillers and absorption chillers, multi-energy storage, and interactions with the upper-level power grid.

[0070] The detailed steps of step S3 include:

[0071] S301, the non-convex feasible region (∪) of the unit's start-up and shutdown states. n∈ψ Ω) is transformed into its convex hull conv(∪) n∈ψ Ω), to convexize the original system model;

[0072] S302. Relax all 0-1 variables representing the unit's operating status to the interval [0,1], making them continuous variables.

[0073] The detailed steps of step S4 include:

[0074] S401. Obtain network parameters in the natural gas system, parameters in the thermal network, and relevant parameters of gas turbines, gas boilers, electric boilers, and energy storage devices in the integrated energy system. The simulation example is programmed using YALMIP on the MATLAB platform and solved using the Gurobi solver.

[0075] S402. The AE-GAN model is used to further generate prediction error data for each nondeterministic variable in the integrated energy system. The correlation between the actual prediction error scenarios and the generated scenarios will be calculated. Then, the t-distributed random neighborhood embedding (t-SNE) method and principal component analysis (PCA) are used for dimensionality reduction to obtain the correlation between the prediction error data of the actual nondeterministic variables and the generated data.

[0076] S403. Based on the system parameters set above, solve the AE-GAN-based integrated energy system partial brook opportunity optimization model constructed in this paper.

[0077] S404. Based on the convex hull pricing method, further solve the nodal energy prices in the integrated energy system, namely nodal electricity price, nodal gas price, nodal heat price and nodal cooling price.

[0078] In step S101, the objective function for the economic operating cost of the integrated energy system is as follows:

[0079]

[0080] In the formula, F EC F1 represents the economic operating cost of the integrated energy system, F2 represents the cost of purchasing electricity from the external power grid, and F3 represents the cost of purchasing natural gas from the natural gas market. This represents the unit cost of purchasing electricity from an external power grid. This indicates the amount of electricity purchased from the external power grid. This represents the unit cost of selling electricity to the external power grid (i.e., the grid connection price). This refers to the amount of electricity sold to the external power grid.

[0081] The operating cost of the integrated energy system should be:

[0082]

[0083] In the second phase (real-time phase), the non-determinism of renewable energy output and electrical, thermal, and cooling loads is fully considered. To avoid corresponding forecast errors (e.g., Random disturbances in the integrated energy system cause further adjustments to the power / natural gas ratio of flexible resources (operational rescheduling), thus incurring adjustment costs.

[0084] In step S102, introduce Efficiency serves as an indicator for evaluating the utilization rate of resources in an energy system.

[0085] The objective function for maximizing efficiency is as follows:

[0086]

[0087] In the formula, ψ e ψ q and ψ c These represent the energy quality coefficients for electrical energy, thermal energy, and cold energy, respectively.

[0088] By employing the Charnes-Cooper transformation, this problem is converted into a linear programming problem. Considering the operational states of each component in the integrated energy system, the flow of various energy sources, such as cooling, heating, and electricity, can be optimized, ultimately improving the energy efficiency of the integrated energy system.

[0089] In step S202, the corresponding random variable is given. and The corresponding prediction error and It can be represented as:

[0090]

[0091] In the formula, and These represent the actual data for wind power output, photovoltaic power output, electrical load, heat load, and cooling load, respectively.

[0092] The traditional univariate LAP adjustment method is extended, and a multivariate LAP adjustment model is proposed as follows:

[0093]

[0094]

[0095] therefore, and These represent all adjustment rates corresponding to different prediction errors. Furthermore, They represent the decision variables respectively. The adjustment results are shown. The model will be extended by combining conventional constraints and multi-energy flexibility constraints, thus completing the sub-Bruker optimization model of the integrated energy system.

[0096] In step S302, an approximate description method for the convex hull, namely the integer relaxation pricing method, is introduced. This method relaxes all 0-1 variables representing the unit's operating state to the interval [0,1], making them continuous variables.

[0097] In step S401, the distribution network in this integrated energy system is an improved IEEE-9 node power system, the heating / cooling network is based on an improved Bali 6-node heating system, and the natural gas network uses a 7-node system. The organic coupling of these three energy systems forms the integrated energy node system described in this paper. Based on the annual average load demand forecast data and related parameters of the integrated energy system, and according to the electricity trading price, the system is programmed using YALMIP on the MATLAB platform, and the Gurobi solver is called for solution.

[0098] In step S403, an ocean hunting algorithm is used to handle the multi-objective optimization problem. The scheduling method based on the multi-objective genetic optimization algorithm increases the operating cost by 7.7% and carbon emissions by 8.0% compared to the method presented in this paper. Efficiency decreased by 9.7%; the scheduling method based on multi-objective particle swarm optimization algorithm increased the operating cost by 3.5% compared to the method in this paper, but reduced carbon emissions by 2.5%. Efficiency decreased by 9.3%; the scheduling method based on decomposition-based multi-objective evolutionary algorithms resulted in a 6.4% increase in operating cost and a 3.0% increase in carbon emissions compared to the method presented in this paper. Efficiency decreased by 8.3%. Therefore, the optimal trade-off solution proposed in this paper based on the multi-objective marine predator algorithm suffers from lower operating costs, lower carbon emissions, and lower efficiency. It has a greater advantage in three aspects, including efficiency.

[0099] For the nondeterministic factors in the integrated energy system, a comparative analysis will be conducted with robust optimization and stochastic optimization models.

[0100] Compared with the prior art, the present invention has the following advantages:

[0101] By constructing the model and drawing an analogy to the relevant definition of the nodal marginal electricity price package pricing method, the model is extended to the integrated energy system. Then, through numerical simulation, the effectiveness of the model in reducing economic operating costs is verified.

[0102] This model constructs a multi-objective optimization scheduling model based on distributed bars, simultaneously considering economic operating costs, carbon emissions, and... Compared to single-objective optimization models, the three efficiency objectives provide a more comprehensive balance between various economic and environmental indicators of system operation. Furthermore, by extending the traditional univariate LAP adjustment method, a multivariate LAP adjustment model is proposed, which can more flexibly address the uncertainties of renewable energy and load demand, and reduce adjustment costs caused by forecast errors.

[0103] When calculating the carbon emissions generated by the optimized operation of an integrated energy system, the model uses it as an optimization objective. By adjusting the operating status of various devices and energy allocation within the system, it minimizes carbon emissions. The model can flexibly address the uncertainties of renewable energy and load demand, reducing additional carbon emissions caused by prediction errors through strategy adjustments. It considers the flow and conversion of various energy sources, including electricity, heat, and cooling, within the integrated energy system, achieving multi-energy complementarity through optimized scheduling and improving overall energy efficiency. Compared to traditional energy systems, this model has significant advantages in reducing carbon emissions and contributes to promoting the green transformation of energy systems.

[0104] The above provides a detailed description of a cost optimization method for integrated energy systems provided in this application. The specific embodiments are described only to aid in understanding the method and its core concepts. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the scope of protection of the claims.

Claims

1. A cost optimization method for an integrated energy system, characterized in that, Comprising the steps of: S1, according to the minimum economic operation cost, the minimum carbon emission and The most efficient demand builds a distributed robust multi-objective optimization scheduling model of comprehensive energy system S2, extending the traditional single variable LAP adjustment method, and putting forward a multi-element LAP adjustment model; S3, extending the node marginal price and the convex hull pricing method to the integrated energy system, and convexifying the original system model; S4, performing example simulation in the integrated energy system, solving the multi-objective optimization scheduling model of the integrated energy system, and further solving the node energy price in the integrated energy system based on the convex hull pricing method; The detailed steps of step S1 include: S101, constructing an economic operation cost objective function of the integrated energy system, and further adjusting the cost of renewable energy and cold and heat load demand uncertainty, setting the objective function as minimizing the expected adjustment cost under the worst case distribution in the fuzzy set; S102, calculating the carbon emissions generated in the optimized operation system of the integrated energy system; S103, constructing maximizing the efficiency objective function and transforming it into a linear programming problem; In step S102, introducing Efficiency as an indicator to assess the resource utilization of the energy system; The efficiency maximization objective function is as follows: where ψ e , ψ q and ψ c represent the energy quality coefficients of electrical, thermal and cold energy, respectively. Through Charnes-Cooper transformation, it is converted into a linear programming problem; The detailed steps of step S2 include: S201, presetting that the electric energy storage, thermal energy storage, micro gas turbine, waste heat boiler, gas boiler and absorption refrigeration machine meet the conventional operation conditions, and performing the first stage operation constraint; S202, extending the traditional single variable LAP adjustment method, and putting forward a multi-element LAP adjustment model; S203, adjusting the constraint of the gas turbine, waste heat boiler, gas boiler, electric refrigeration machine, absorption refrigeration machine, multi-element energy storage and upper grid interaction; In step S202, given the corresponding random variables and the corresponding prediction error and is expressed as: wherein, and respectively represent the real data of wind power output, photovoltaic output and electric load, thermal load and cold load. The traditional single variable LAP adjustment method is extended, and a multi-element LAP adjustment model is as follows: Thus, and respectively represent all adjustment rates corresponding to different prediction errors, in addition, respectively represent adjustment results of the decision variable .

2. The cost optimization method for an integrated energy system according to claim 1, wherein, The detailed steps of step S3 include: S301, the non-convex feasible region (∪) of the unit's start-up and shutdown states. n∈ψ Ω) is transformed into its convex hull conv(∪) n∈ψ Ω), to convexize the original system model; S302, relaxing all 0-1 variables representing the unit operation state to the interval [0, 1] to make them continuous variables.

3. The cost optimization method for an integrated energy system according to claim 2, wherein, The detailed steps of step S4 include: S401, obtaining the related parameters of the gas system, the thermal network, the gas turbine, the gas boiler, the electric boiler and the energy storage equipment in the integrated energy system, programming on the MATLAB platform using YALMIP, and calling the Gurobi solver for solving; S402, using the AE-GAN model to further generate prediction error data of each non-deterministic variable in the integrated energy system, calculating the correlation between the true prediction error scenario and the generated scenario, and using the t-distributed stochastic neighbor embedding (t-SNE) method and principal component analysis (PCA) for dimensionality reduction processing to obtain the correlation between the true non-deterministic variable prediction error data and the generated data; S403, based on the parameters set in step S401 in the natural gas system, solving the integrated energy system distribution robust opportunity optimization model based on AE-GAN constructed in this paper; S404, further solving the node energy price in the integrated energy system based on the convex hull pricing method.

4. The cost optimization method for an integrated energy system according to claim 3, wherein, In step S101, the economic operation cost objective function of the integrated energy system is as follows: In the formula, F EC represents the economic operation cost of the integrated energy system, F1 represents the electricity purchase cost of the integrated energy system from the external power grid, and F2 represents the gas purchase cost of the integrated energy system from the natural gas market; represents the unit cost of purchasing electricity from the external power grid, represents the amount of electricity purchased from the external power grid, represents the unit cost of selling electricity to the external power grid, represents the amount of electricity sold to the external power grid; The operation cost of the integrated energy system should be: In the second stage, the non-deterministic of renewable energy output and electric load, thermal load, cold load are fully considered, represents the adjustment cost.

5. The cost optimization method for an integrated energy system according to claim 4, wherein, In step S302, an approximate description method for the convex hull is introduced, which relaxes all 0-1 variables representing the unit operation state into the interval [0, 1] to make them continuous variables.

6. The cost optimization method for an integrated energy system according to claim 5, wherein, In step S401, the distribution network in the integrated energy system is an improved IEEE-9 node power system, the heating / cooling network is based on an improved Bali 6 node heating system, and the natural gas network adopts a 7 node system; the organic coupling of the above three energy systems forms the integrated energy node system in this paper. According to the annual average load demand prediction data and related parameters in the integrated energy system, the transaction electricity price is used to program on the MATLAB platform using YALMIP, and the Gurobi solver is called for solving.

Images