Integrated Energy System Operation Method Based on Spectral Normalization Generative Adversarial Network

Abstract

The present invention discloses an operation method for an integrated energy system based on a spectral normalization generative adversarial network, which includes obtaining a typical scenario set of uncertainty variables, fusing the scenario information of uncertainty variables, using the 1-norm constraint and the box constraint to formulate a robust uncertainty set, and constructing a three-stage robust optimization scheduling model with the lowest operating cost as the objective; transforming the three-stage robust optimization scheduling model into a single-layer mixed integer model and using a solver to solve it to obtain a scheduling plan. The scheduling plan of the present invention can reduce the impact of multiple uncertainties such as the electricity price of the superior power grid and the source-load output on the economy and stability of the integrated energy system, and decision-makers can balance the economic efficiency and conservatism of the IES by selecting appropriate uncertainty parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of hybrid energy systems, and in particular to an operation method for an integrated energy system based on a spectral normalization generative adversarial network. Background Technique

[0002] With the gradual depletion of fossil energy and the increasingly serious environmental pollution, the efficient utilization of energy has become the focus of research by experts and scholars in the energy field. The integrated energy system (IES) couples four types of energy, namely cold, heat, electricity, and gas, to provide users with various load demands such as cold, heat, and electricity, which can achieve complementary utilization of energy, reduce the system operation cost, improve the energy utilization efficiency, and reduce pollutant emissions. For an integrated energy system containing renewable energy, the existence of multiple uncertainties such as the randomness, intermittency of power generation, and the volatility of various loads makes it impossible to accurately predict the power generation and load, and thus it is impossible to formulate an accurate scheduling plan for the system, which has a great impact on the stability and economy of the operation of the integrated energy system.

[0003] A large number of literatures have considered the uncertainty of the source and load in the research on the optimal scheduling of integrated energy systems and achieved rich research results. For example, in the literature such as Zhang Jinhui, Wang Xu, Jiang Chuanwen, etc., "Optimal Scheduling Method for a Multi-Network Coupled Integrated Energy System Considering the Uncertainty of Traffic Flow" [J]. Power System Technology, 2019, 43(9): 3081-3089, when constructing the optimal scheduling model of the integrated energy system, the problem of energy supply and demand imbalance caused by the spatio-temporal uncertainty of the operation of new energy vehicles was considered.

[0004] The above-mentioned literatures have provided many valuable research methods for the uncertainty problems in the operation of integrated energy systems and achieved rich research results. However, in the existing research results, generally only the uncertainty of a single source and load in the system is considered or only one method is used to deal with multiple uncertainties, while ignoring the respective characteristics of the uncertainties such as the power generation of renewable energy and the load prediction error in the system, which has a certain one-sidedness. In fact, when there is randomness in the power generation of renewable energy, its probability distribution function is determined, so it is suitable to use the stochastic method to deal with it; there is an error in load prediction, and its probability distribution and membership relationship are difficult to determine, so it is suitable to use the interval method or the robust optimization method to deal with it. Compared with the interval method, the robust optimization method generally focuses on extreme cases and requires satisfying the constraints in the worst case, and the obtained optimization results are often too conservative. Summary of the Invention

[0005] The purpose of the present invention is to provide an operation method for an integrated energy system based on a spectral normalization generative adversarial network to solve the problems raised in the above background technique.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An integrated energy system operation method based on a spectral normalization generative adversarial network, comprising:

[0008] Obtain a typical scenario set of uncertainty variables, fuse the scenario information of uncertainty variables, use 1-norm constraint and box constraint to formulate a robust uncertainty set, and construct a three-stage robust optimization scheduling model with the goal of minimizing the operating cost;

[0009] Convert the three-stage robust optimization scheduling model into a single-layer mixed integer model, and use a solver to solve it to obtain a scheduling plan.

[0010] Further, obtaining a typical scenario set of uncertainty variables, fusing the scenario information of uncertainty variables, and using 1-norm constraint and box constraint to formulate a robust uncertainty set specifically includes: introducing the weight matrix of the discriminator in spectral norm normalized WGAN-GP to constrain the discrimination performance, and realizing global 1-Lipschitz constraint by calculating the maximum singular value of the weight matrix of each layer of the discriminator.

[0011] Further, the objective function of WGAN-GP is as follows:

[0012] (1)

[0013] Assume the input renewable energy data set X , the discriminator includes discriminators , , , , , and ;

[0014] , for any input X there is:

[0015] (2)

[0016] Mark the set of weights and biases of each layer of the discriminator :

[0017] (3)

[0018] Then we get:

[0019] (4)

[0020] According to we get:

[0021] (5)

[0022] Normalize the weights as follows:

[0023] (6)

[0024] For any X :

[0025] (7)

[0026] Calculate the maximum singular value of the weight matrix for each layer using inequality (7);

[0027] where is the spectral norm of is the unit distance, is the weight for the th layer discrimination, is the bias for the th layer discrimination, is the total number of discriminator layers, is the weight of the discriminator output, is the dimension of the discriminator output for the th layer, is the weight for the th layer discrimination, is the dimension of the discriminator output for the th layer, is the weight for the th layer discrimination, is the spectral norm of is the spectral norm of is the spectral norm of is the Wasserstein distance between the real and generated scenarios needs to satisfy 1-Lipschitz continuity, is the real scenario distribution pattern, is the random noise distribution probability, is the probability of the generated scenario, is the discriminator output, is the expected value of the input noise.

[0028] Furthermore, constructing a three-stage robust optimal scheduling model with the lowest operating cost as the goal specifically includes a first-stage scheduling model, a second-stage scheduling model, and a third-stage scheduling model.

[0029] Furthermore, in the first-stage scheduling model, the objective function is to minimize the switching cost of 0-1 variables , including the power exchange status between the integrated energy system and the superior power grid and the charge and discharge status of energy storage devices. The objective function expression is:

[0030] (8)

[0031] The constraint conditions are:

[0032] (9)

[0033] (10)

[0034] In the formula, TE is the scheduling period, is the cost coefficient of 0-1 variable switching, is the depreciation cost coefficient of the charge and discharge of energy storage devices, , are the 0-1 variables for procurement and sales at the t moment respectively, , are the 0-1 variables for the charge and discharge status of energy storage devices at the t moment respectively.

[0035] Furthermore, in the second-stage scheduling model, the objective function is to minimize the power exchange cost between the integrated energy system and the superior power grid , and the objective function expression is:

[0036] (11)

[0037] The constraint conditions are:

[0038] (12)

[0039] (13)

[0040] In the formula, , are the power purchase price and power sales price of the power exchange between the integrated energy system and the superior power grid at the t moment respectively, , are the power purchase power and power sales power exchanged between the integrated energy system and the superior power grid at the t moment respectively, , are respectively the maximum values of the purchased power and the sold power exchanged between the integrated energy system and the superior power grid.

[0041] Furthermore, in the third-stage scheduling model, the objective function is to minimize the operating cost of the integrated energy system , specifically including the gas purchase cost , the operating cost of the energy storage device and the carbon trading cost , , and the expressions are respectively:

[0042] (14)

[0043] (15)

[0044] (16)

[0045] The constraint conditions include the CHP-CCS-P2G system coupling model constraint conditions, the carbon trading constraint conditions, the energy storage device constraint conditions, the gas boiler model constraint conditions, and the integrated energy system power balance constraint conditions;

[0046] In the formula, is the gas price for the integrated energy system to purchase gas from the superior power grid, , , are respectively at t the natural gas consumed by the combined heat and power unit, the gas boiler, and the power-to-hydrogen unit at the moment.

[0047] Furthermore, the expression of the CHP-CCS-P2G system coupling model constraint conditions is as follows:

[0048] (17)

[0049] (18)

[0050] (19)

[0051] (20)

[0052] (21)

[0053] (22)

[0054] (23)

[0055] (24)

[0056] (25)

[0057] (26)

[0058] (27)

[0059] (28)

[0060] In the formula, is the calorific value of natural gas, , are the power generation and heat production efficiencies of the combined heat and power unit respectively, , are respectively at t the output electric power and heat power of the combined heat and power unit at the moment, , are respectively the maximum values of the output electric power and heat power of the combined heat and power unit, is the carbon emission conversion coefficient of the combined heat and power unit, is at t the carbon emission value of the combined heat and power unit at the moment, is at t the amount of CO 2 captured by the carbon capture unit at the moment; is the power consumption coefficient for the carbon capture unit to capture CO 2 , is at t the electric power consumed by the carbon capture unit to capture CO 2 at the moment, is the conversion coefficient of the amount of CO 2 required for the power-to-gas unit to produce natural gas and the power consumption, is at t the value of CO 2 consumed by the power-to-gas unit at the moment, is at t the electric power consumed by the power-to-gas unit at the moment, is the efficiency of the power-to-gas unit to be converted into natural gas, is at t the net output of the CHP-CCS-P2G system for coordinated operation at the moment.

[0061] Further, after CCS-P2G absorption and conversion, the actual carbon emissions of the CHP-CCS-P2G system and the total carbon emissions of the microgrid are expressed as the following formula:

[0062] (29)

[0063] (30)

[0064] Wherein, and are the carbon emission coefficients of the gas boiler and the upstream power grid, is the electric power purchased by the integrated energy system from the upstream power grid.

[0065] Furthermore, the expression of the carbon trading constraint condition is as follows:

[0066] (31)

[0067] (32)

[0068] Wherein, is the electro-thermal conversion coefficient, and are the carbon emission coefficients of the gas unit and the coal-fired unit respectively. The actual carbon quota of the integrated energy system participating in the carbon trading market , represents the carbon quota sold by the integrated energy system (which can also be said to be the free quota allocated to the integrated energy system).

[0069] Furthermore, the expression of the energy storage device constraint condition is as follows:

[0070] (33)

[0071] (34)

[0072] (35)

[0073] (36)

[0074] Wherein, and are respectively the capacities of the energy storage device at t and t- time 1. and are respectively the charging and discharging powers of the energy storage device at t time. The energy storage device includes ESS and HSD. and are respectively the charging and discharging efficiencies of the energy storage device. and are respectively the upper and lower limits of the capacity of the energy storage device. and are respectively the maximum values of the charging and discharging powers of the energy storage device.

[0075] Furthermore, the constraint condition expression of the gas boiler model is as follows:

[0076] (37)

[0077] (38)

[0078] In the formula, is the thermal power output of the gas boiler, is the operating efficiency of the gas boiler, is the upper limit of the thermal power output of the gas boiler.

[0079] Furthermore, the power balance constraint of the integrated energy system includes the following electric power balance expression:

[0080] (39)

[0081] And the thermal power balance, the expression is as follows:

[0082] (40)

[0083] In the formula, is the output of the wind turbine at t time, 、 are the charging and discharging powers of the electric energy storage unit at t time respectively, is the electric load, 、 are the charging and discharging powers of the thermal energy storage unit at t time respectively, is the thermal load at t time.

[0084] Furthermore, by adopting a simplification method combining double-layer CCG, Karush-Kuhn-Tucker conditions, and the big M method, the three-stage robust optimization scheduling model is transformed into a single-layer mixed-integer model, and a solver is used for solving. The obtained scheduling scheme specifically includes: obtaining the three-stage robust optimization scheduling model through multiple uncertainty processing based on the polyhedral uncertainty set, decomposing and reconstructing the three-stage robust optimization scheduling model to obtain a single-layer mixed-integer model, and then solving it through the GUROBI solver to obtain the scheduling scheme.

[0085] Furthermore, the expression of the three-stage robust optimization scheduling model obtained through multiple uncertainty processing based on the polyhedral uncertainty set is as follows:

[0086] (41)

[0087] (42)

[0088] In the formula, is the uncertainty set of the power interaction price between the second-stage integrated energy system and the superior power grid, is the uncertainty set of the power sources and loads in the dispatching of the third-stage integrated energy system, and are the uncertainty variables of electricity price and power sources and loads respectively, and are the expected values of the uncertainty variables of electricity price and power sources and loads respectively, and are the maximum and minimum values of the electricity price uncertainty variable respectively, and are respectively the maximum and minimum values of the 0-1 variable of the uncertain change of and are the maximum and minimum values of the power sources and loads uncertainty variable respectively, and are respectively the maximum and minimum values of the 0-1 variable of the uncertain change of and are respectively and the uncertainty control parameters of is the electrical load, is the thermal load at t , and are respectively the electricity purchase price and the electricity selling price of the power exchange between the integrated energy system and the superior power grid at t , is the output of the wind turbine at t .

[0089] Furthermore, the full matrix formula of the three-stage robust optimal dispatching model is as follows:

[0090] (43)

[0091] (44)

[0092] In the formula, TE is the dispatching period, and the coefficient matrices of the first-stage constraint conditions and and , the coefficient matrices of the second-stage constraint conditions and and and , the coefficient matrices of the third-stage constraint conditions and and and , , , , the sets of decision variables for the first, second, and third stages are respectively , , , , are respectively the 0-1 variables for procurement and sales at the t moment, , are respectively the 0-1 variables for the charging and discharging states of the energy storage device at the t moment, , are respectively the purchased and sold electric powers exchanged between the integrated energy system and the superior power grid at the t moment, , are respectively the electric power and thermal power output by the combined heat and power unit at the t moment, is the electric power consumed by the carbon capture unit to capture CO t at the 2 moment, is the electric power consumed by the power-to-hydrogen unit at the t moment, , are respectively the charging and discharging powers of the energy storage device at the t moment, is the thermal power output of the gas boiler.

[0093] Furthermore, the decomposition and reconstruction of the three-stage robust optimization scheduling model to obtain a single-layer mixed-integer model specifically includes: using the nested CCG algorithm inside and outside to decompose and reconstruct the three-stage robust optimization scheduling model, decomposed into the master problem MP and the sub-problem SP, and the expressions are as follows:

[0094] (45)

[0095] (46)

[0096] After decomposition by the CCG algorithm, using the KKT conditions to transform the multi-layer nested structure of the three-stage robust optimization scheduling model into a single-layer MILP structure, thus obtaining a single-layer mixed-integer model. The specific transformation process is as follows:

[0097] The expression form of MP is (47)

[0098] The expression form of SP is (48)

[0099] The main problem objective of the sub-problems after decomposition is to solve the trading power between the integrated energy system and the superior power grid under the worst trading electricity price, and the sub-problem objective of the sub-problems is to solve the unit scheduling plan under the worst source-load scenario of the integrated energy system. The expressions are as follows:

[0100] (49)

[0101] (50)

[0102] In the formula, is the worst solution to obtain the uncertainty elements from the sub-problems, and the auxiliary variable , the auxiliary variable .

[0103] Furthermore, it is then solved by the GUROBI solver, and the obtained scheduling plan specifically includes:

[0104] 1) Initialize the outer-layer CCG algorithm problem. The initialized upper and lower bounds of the outer layer are , , initialize the iteration number , and determine the convergence termination condition threshold ;

[0105] 2) Solve the main problem of the outer-layer CCG algorithm to obtain the optimal solution , update the lower bound value of the outer-layer problem , and check whether the convergence condition is satisfied, where . If the convergence condition is satisfied, terminate and the solution is completed; otherwise, continue to solve;

[0106] 3) Solve the inner-layer CCG algorithm problem. Initialize the inner-layer CCG algorithm problem. The initialized upper and lower bounds are , , initialize the iteration number , and determine the convergence termination condition threshold ;

[0107] 4) Solve the main problem of the inner-layer CCG algorithm to obtain the optimal solution , and update the lower bound value of the inner-layer problem ;

[0108] 5) Solve the sub-problem of the inner-layer CCG algorithm to obtain the optimal solution , and update the upper bound value of the inner-layer problem , and check whether the convergence condition is satisfied. If the convergence condition is satisfied, terminate the inner-layer problem and jump to 3); otherwise 6) continue to solve;

[0109] 6) Add the variables and constraints in the second and third stages to the main problem of the outer-layer CCG algorithm, increment the initialization iteration count by one, and return to 2).

[0110] To achieve the above object, the present invention also provides the following technical solutions:

[0111] An integrated energy system operating device based on a spectral normalization generative adversarial network, comprising:

[0112] A construction module, configured to obtain a typical scenario set of uncertainty variables, fuse the scenario information of uncertainty variables, use 1-norm constraint and box constraint to formulate a robust uncertainty set, and construct a three-stage robust optimization scheduling model with the lowest operating cost as the objective;

[0113] A scheduling module, configured to use a simplification method combining a two-layer CCG algorithm, Karush-Kuhn-Tucker conditions, and the Big-M method to transform the three-stage robust optimization scheduling model into a single-layer mixed-integer model, and use a solver to solve it to obtain a scheduling plan.

[0114] To achieve the above object, the present invention also provides the following technical solutions:

[0115] A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described in any one of the above when executing the computer program.

[0116] To achieve the above object, the present invention also provides the following technical solutions:

[0117] A computer-readable storage medium, on which a computer program is stored, and the computer program implements the steps of the method described in any one of the above when executed by a processor.

[0118] Compared with the prior art, the beneficial effects of the present invention are:

[0119] In the present invention, in order to reduce the impact of multiple uncertainties of the upper-layer grid electricity price and source (negative) load on the economic efficiency and security of the IES, a three-stage robust scheduling method is proposed. To manage multiple uncertainties, a spectral normalization generative adversarial network (SNWGAN) and K-means++ clustering are used to generate initial scenarios of uncertainty variables. Based on these initial scenarios, a robust polyhedral uncertainty set is constructed using 1-norm and box constraints, and a three-stage robust optimization scheduling model with the objective of minimizing the operating cost is developed. Then, the model is simplified using a combination of two-layer column and constraint generation (CCG), Karush-Kuhn-Tucker (KKT) conditions, and the Big-M method, and directly solved using the commercial solver GUROBI. The main conclusions are as follows:

[0120] 1) Scenario generation and reduction: Use the SNWGAN and K-means++ clustering algorithms to generate and reduce the initial scenarios of uncertain variables. This method effectively utilizes the historical dataset to generate scenarios that are similar to but not exactly the same as the historical data.

[0121] 2) Enhanced robustness: Compared with traditional deterministic optimization models and other robust optimization (RO) models, the proposed three-stage RO model enhances the system's ability to mitigate uncertainties and is more in line with the actual scheduling scenario. In addition, the adopted algorithm can solve the optimal model solution in a relatively short time (22.81 seconds).

[0122] 3) Sensitivity analysis: Through the sensitivity analysis of uncertain parameters, it is found that as the uncertain parameters increase, the total system optimization cost rises and the robustness of the scheduling plan is enhanced. Decision-makers can balance the economic efficiency and conservatism of the IES by selecting appropriate uncertain parameters. Brief Description of the Drawings

[0123] Figure 1 It is the flowchart of the method of the present invention.

[0124] Figure 2 It is the structure diagram of the IES.

[0125] Figure 3 It is the process diagram of the double-layer CCG algorithm solution of the present invention.

[0126] Figure 4 It is the diagram for evaluating the generated scenarios by using the Wasserstein distance of the present invention.

[0127] Figure 5 It is the diagram for evaluating the generated scenarios by using the Fréchet inception distance of the present invention.

[0128] Figure 6 It is the diagram for evaluating the generated scenarios by using the t-SNE distribution of the present invention.

[0129] Figure 7 It is the diagram for calculating the similarity degree between each cluster by using the Davies-Bouldin index coefficient of the present invention.

[0130] Figure 8 It is the diagram of the power output optimization result of the IES of the present invention.

[0131] Figure 9 It is the diagram of the heat output optimization result of the IES of the present invention.

[0132] Figure 10 It is the diagram of the electricity purchase price during the worst-case scenario analysis of the present invention.

[0133] Figure 11This is the electricity price diagram for the worst scenario analysis of the present invention.

[0134] Figure 12 This is the worst output and electricity price distribution diagram for the worst scenario of source load in the first stage of the present invention.

[0135] Figure 13 This is the worst output and electricity price distribution diagram for the worst scenario of source load in the second stage of the present invention.

[0136] Figure 14 This is the worst output and electricity price distribution diagram for the worst scenario of source load in the third stage of the present invention.

[0137] Figure 15 This is a diagram showing the influence of the uncertain parameters of the present invention on the total cost of optimizing the energy base.

[0138] Figure 16 This is a diagram of the equipment model of the method of the present invention.

[0139] Figure 17 It is a diagram of the internal structure of the computer device of the present invention. DETAILED DESCRIPTION

[0140] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0141] See also Figures 1 to 17 , the present invention provides a technical solution:

[0142] In order to reduce the impact of multiple uncertainties such as the electricity price of the upper power grid and the output of sources and loads on the economy and stability of the integrated energy system, a scenario generation method based on spectral normalization generative adversarial network is proposed. First, a typical scenario set of uncertain variables is obtained; second, the scenario information of uncertain variables is integrated, and the 1-norm constraint and box constraint are used to formulate a robust uncertainty set, and a three-stage robust optimization scheduling model with the lowest operating cost as the goal is constructed. Then, a simplification method combining double-layer CCG, Karush-Kuhn-Tucker conditions, and big M method is used to transform the three-stage robust model into a single-layer mixed integer model, and the commercial solver GUROBI is used to solve it. Finally, the effectiveness of the proposed model and method is verified through example simulation. CCG: column and constraint algorithm, Karush-Kuhn-Tucker: Karush Kuhn Tucker.

[0143] 1. Generation of Typical Scenarios Based on SNWGAN and K-means++ Clustering

[0144] 1.1 SNWGAN

[0145] WGAN-GP is one of the most important derivative models of GAN. It adopts the Gradient Penalty (GP) gradient penalty term to make the Wasserstein (earth mover's distance) distance satisfy the Lipschitz continuity condition. Lipschitz continuity guarantees the existence of the optimal discriminant function and the unique Nash equilibrium, which is crucial for the stable training of the model.

[0146] WGAN-GP consists of two neural networks. One is the generator, G (), and the other is the discriminator, D (); G Taking random noise as input, it is used to mine the internal features of real sample data, so as to generate a large number of new data samples similar to the real data; D Taking real data and generated data as input, the purpose is to distinguish real data and generated data as much as possible. The objective function of WGAN-GP is shown as follows.

[0147] (1)

[0148] However, the GP term introduced by WGAN-GP can only make the Wasserstein distance approximately satisfy the Lipschitz continuity, and it is difficult to achieve global constraints. In the present invention, the spectral norm is introduced to normalize the weight matrix in the discriminant network, further constraining its discriminant performance.

[0149] Assume the input is X , and the discriminator expression , for any input X there is:[[]]

[0150] (2)

[0151] Overall, mark the discriminator weights and biases:

[0152] (3)

[0153] Then we can get:

[0154] (4)

[0155] According to we can get:

[0156] (5)

[0157] Normalize the weights:

[0158] (6)

[0159] In summary, for any X:

[0160] (7)

[0161] By using the above inequality, the global 1-Lipschitz constraint can be achieved by only calculating the largest singular value of the weight matrix of each layer.

[0162] Let be the spectral norm of be the weights of each layer of the discriminator, be the biases of each layer of the discriminator, be the l-th layer of the discriminator; L is the total number of layers of the discriminator; be the output degree matrix of the discriminator, be the Wasserstein distance between the two distributions of the real scenario and the generated scenario needs to satisfy 1-Lipschitz continuity, be the real scenario xn be the distribution probability of be the random noise be the distribution probability of be the distribution probability of the generated scenario.

[0163] 2. Structure of the integrated energy system and the three-stage RO scheduling model

[0164] 2.1 Structure of the integrated energy system

[0165] The integrated energy system constructed by the present invention mainly includes a combined heat and power (CHP) unit, a power to gas (P2G) unit, a carbon capture and storage (CCS) unit, a wind turbine (WT), a gas boiler (GB), an electrical energy storage (ESS), a heat energy storage (HSD), a Superior Grid (SG), a Natural Gas Network (NSN), and an electrical load (EL), a heat load (HL), as specifically shown in Figure 2 as follows.

[0166] 2.2 Three-stage RO scheduling model of the integrated energy system

[0167] 2.2.1 First-stage scheduling model

[0168] 1) Objective function

[0169] In the first stage, we take the minimization of the 0-1 variable state switching cost as the objective function, including the power exchange state between the IES and the superior power grid and the charge and discharge states of energy storage devices. The expression of the objective function is as follows:

[0170] (8)

[0171] 2) Constraint conditions

[0172] (9)

[0173] (10)

[0174] 2.2.2 Second-stage scheduling model

[0175] 1) Objective function

[0176] In the second stage, we take the minimization of the power exchange cost between the IES and the superior power grid as the objective function. The expression of the objective function is as follows:

[0177] (11)

[0178] 、 are respectively t the power purchase price and the power selling price of the power exchange between the IES and the superior power grid at time 、 are respectively t the power purchase power and the power selling power exchanged between the IES and the superior power grid at time

[0179] 2) Constraint conditions

[0180] (12)

[0181] (13)

[0182] 2.2.3 Third-stage scheduling model

[0183] 1) Objective function

[0184] In the third stage, the objective function is to minimize the operating cost of the IES, specifically including the gas purchase cost 、the energy storage operating cost and the carbon trading cost ( For the specific calculation process of

[0185] (14)

[0186] (15)

[0187] (16)

[0188] In the formula, is the gas price for IES to purchase gas from the superior power grid; is the depreciation cost coefficient of energy storage charging and discharging;

[0189] 2) Constraint conditions

[0190] (a) Constraint conditions of the CHP-CCS-P2G system coupling model

[0191] (17)

[0192] (18)

[0193] (19)

[0194] (20)

[0195] (21)

[0196] (22)

[0197] (23)

[0198] (24)

[0199] (25)

[0200] (26)

[0201] (27)

[0202] (28)

[0203] In the formula, is the natural gas consumed by the CHP unit; is the calorific value of natural gas; 、 are the power generation and heat production efficiencies of the CHP unit respectively; 、 are the electric power and heat power output by the CHP unit respectively; 、 are the maximum values of the electric power and heat power output by the CHP unit respectively; is the carbon emission conversion coefficient of the CHP unit; is the carbon emission value of the CHP unit; is the value of CCS capturing CO 2 ; is the power consumption coefficient of the CCS system for capturing CO 2 ; is the electric power consumed by CCS for capturing CO 2 ; is the conversion coefficient of the amount of CO 2 required for P2G to generate natural gas and the power consumption; is the value of CO 2 consumed by P2G; is the electric power consumed by P2G; is the efficiency of P2G converting to natural gas; is the net output of the coordinated operation of the CHP-CCS-P2G system.

[0204] After being absorbed and converted by CCS-P2G, the actual carbon emissions of the CHP-CCS-P2G and the total carbon emissions of the microgrid can be expressed as:

[0205] (29)

[0206] (30)

[0207] In the formula, , are the carbon emission coefficients of the GB and the superior power grid; is the electric power purchased by the integrated energy system from the superior power grid (assuming that all the electric energy from the power grid comes from coal-fired units).

[0208] (b) Carbon trading constraint conditions

[0209] Currently, the allocation mechanism of domestic carbon trading is mostly free allocation. The initial carbon emission quota model of the IES is as follows:

[0210] (31)

[0211] In the formula, is the free quota allocated to the integrated energy system; , are the carbon emission coefficients of the gas-fired unit and the coal-fired unit respectively.

[0212] Thus, the actual carbon quota of the integrated energy system participating in the carbon trading market , a positive value represents that the integrated energy system purchases carbon quotas, and a negative value represents that the integrated energy system sells carbon quotas.

[0213] (32)

[0214] (c)Energy storage device constraints

[0215] (33)

[0216] (34)

[0217] (35)

[0218] (36)

[0219] (d)Gas boiler model and its constraints

[0220] (37)

[0221] (38)

[0222] (e)IES power balance constraints

[0223] The IES power balance includes the electric power balance and the thermal power balance, and the constraints are formulas (39) and (40) respectively.

[0224] (39)

[0225] (40)

[0226] 3. Multi - uncertainty RO modeling and model solution

[0227] 3.1 Handling multi - uncertainty based on polyhedral uncertainty sets

[0228] Considering the uncertainties of the electricity price of the superior power grid and the IES source - load, polyhedral uncertainty sets are used to describe their fluctuation ranges. For the uncertainty set of the electricity price of the power interaction between the IES and the superior power grid in the second stage, the modeling is as follows:

[0229] (41)

[0230] For the uncertainty set of the source - load in the IES scheduling in the third stage, the modeling is as follows:

[0231] (42)

[0232] 3.2 Full - matrix formula of the three - stage RO model

[0233] For the convenience of understanding in the following text, the full - matrix formula of the above model is as follows:

[0234] (43)

[0235] (44)

[0236] In the formula, 、 、 is the coefficient matrix of the first-stage constraint conditions, 、 、 、 is the coefficient matrix of the second-stage constraint conditions, 、 、 、 、 、 、 is the coefficient matrix of the third-stage constraint conditions, 、 、 are the sets of decision variables for the three stages respectively.

[0237] 3.3 RO Model Decomposition and Reconstruction

[0238] The three-stage RO model constructed in the present invention belongs to a non-deterministic polynomial (NP)-hard problem and cannot be directly solved. To address the computational challenge, we use the inner-outer nested CCG algorithm to decompose and reconstruct the original problem into a master problem (MP) and a subproblem (SP) as follows:

[0239] (45)

[0240] (46)

[0241] In the formula, 、 are auxiliary variables.

[0242] After decomposition by the CCG algorithm, the original problem is transformed into a nested convex optimization problem. The original multi-layer nested structure can be transformed into a single-layer MILP problem using the KKT conditions and can be directly solved using the GUROBI solver. The specific transformation process is as shown in Eqs. (47)-(52).

[0243] The master problem is as shown in Eq. (47):

[0244] (47)

[0245] is the worst-case solution for obtaining the uncertain elements from the subproblem.

[0246] The expression form of the sub - problem is shown in Equation (48).

[0247] (48)

[0248] The sub - problem after CCG transformation is still in a nested structure and cannot be directly solved. Therefore, the CCG algorithm is used to decompose the inner - layer sub - problem again:

[0249] The main - problem objective of the decomposed sub - problem is to solve the transaction power between the IES and the superior power grid under the worst - case trading electricity price, and the sub - problem objective of the sub - problem is to solve the unit scheduling plan under the worst - case IES source - load scenario, as shown in (49) and (50).

[0250] (49)

[0251] (50)

[0252] 3.4 RO Model Solving Process

[0253] The RO model after reconstruction and decomposition can be solved by the GUROBI solver. The specific solving steps are as follows:

[0254] 1) Initialize the outer - layer CCG algorithm problem. Initialize the upper and lower bounds as 、 respectively, initialize the iteration number , and determine the convergence termination condition threshold .

[0255] 2) Solve the outer - layer CCG main problem to obtain the optimal solution , and update the lower - bound value of the outer - layer problem . Check whether the convergence condition is satisfied, where . If the convergence condition is satisfied, terminate and the solving ends; otherwise, continue to solve.

[0256] 3) Solve the inner - layer CCG problem. Initialize the inner - layer CCG problem. Initialize the upper and lower bounds as 、 respectively, initialize the iteration number , and determine the convergence termination condition threshold .

[0257] 4) Solve the inner - layer CCG main problem to obtain the optimal solution , and update the lower - bound value of the inner - layer problem .

[0258] 5) Solve the inner - layer CCG sub - problem to obtain the optimal solution , update the upper bound value of the inner-layer problem , check whether the convergence condition is satisfied , if the convergence condition is satisfied, terminate the inner-layer problem and jump to 3); otherwise 6), then continue to solve.

[0259] 6) Add the variables and corresponding constraint conditions of the second and third stages to the outer-layer CCG master problem, and let , return to 2).

[0260] 4. Case study

[0261] 4.1 Case setup

[0262] This case study uses a demonstration project of a comprehensive energy system in North China as the simulation object. The parameters of the system units are shown in Table 1, the predicted electricity prices are shown in Table 2, and the operating coefficients of the units in this invention are from relevant literature. The data of the SNWGAN model comes from the measured historical data of the source and load in the park for 2 years (from January 1, 2022 to December 31, 2023). The time interval of the data is 1 h, and 80% of the output data is used as the training set, and 20% of the data is the test set. In the case study, the SNWGAN model is built using the deep learning framework Pytorch, and the number of iterations is set to 20,000 times. The computer configuration is an Intel Core i7 processor with a main frequency of 1.8 GHz and 16 GB of memory.

[0263] Table 1 System parameters and coefficients

[0264]

[0265] Table 2 Predicted electricity prices

[0266]

[0267] 4.2 SNWGAN performance verification

[0268] To verify the quality of the scenarios generated by SNWGAN, this invention uses the Wasserstein distance, the Frechet inception distance, and t-SNE to evaluate the generated scenarios.

[0269] (1) Wasserstein distance: The goal of model training is to make the loss as small as possible, that is, to minimize the Wasserstein distance between the true distribution and the generated distribution.

[0270] (2) Frechet inception distance: Calculate the distance between the true sample distribution and the generated sample distribution, and then measure the quality of the generated samples. The lower the score, the closer the two distributions are.

[0271] (3) t-SNE distribution: It can be used for the visualization of high-dimensional data. It converts the similarity between data points into joint probabilities and attempts to minimize the KL divergence between the joint probabilities of the low-dimensional embedding and the high-dimensional data.

[0272] The results of these three evaluation metrics are as Figures 4 to 6 shown. It can be seen that the generated scenario distribution is very close to the real scenario distribution, verifying the accuracy of the scenarios generated by SNWGAN.

[0273] 4.3 Clustering Results

[0274] The Davies-Bouldin Index (DBI) coefficient calculates the similarity between each cluster and obtains the average value of all similarities to measure the quality of the clustering effect. The farther the distance between clusters, the smaller the DBI value, and the better the clustering result at this time. From Figure 7 it can be seen that for the WT output, the optimal number of clusters is 3; for EL, the optimal number of clusters is 2; for HL, the optimal number of clusters is 2. The typical scenarios of WT, EL, and HL calculated using probability weighting are shown in Figure 7 the initial expected values.

[0275] 4.4 IES Optimal Scheduling Scheme

[0276] The three-stage robust optimization iteration process is shown in Table 3. According to Table 3, under the hardware conditions of the present invention, the model converges after 2 outer layer iterations, the algorithm running time is 22.81 s, and the solution speed meets the requirements of day-ahead scheduling.

[0277] Table 3 Inner and outer nested CCG iteration solution process

[0278]

[0279] After the iteration converges, the electricity and heat distribution maps are as Figures 8 to 9 shown. It can be found in Figure 8 that the wind power output in IES plays a great role, indicating a high level of wind power consumption during the scheduling period. During the night time period [3:00 - 6:00], the electricity load demand is low, the wind power output is high, the CHP unit output decreases significantly, and the energy storage is charged to store the excess electric energy. While during the peak electricity load period [16:00:21:00], due to the "electricity determined by heat" operation mode, the electric power output of the CHP unit is restricted, and the energy supply of IES cannot meet the electricity load demand, and the insufficient part is supplemented by the discharge of the energy storage.

[0280] As Figure 9 shown, the heat load is mainly provided by CHP, heat energy storage, and GB. When the heat energy exceeds the heat load demand, the excess heat energy can be stored and released when the heat energy is insufficient.

[0281] 4.5 Uncertainty Analysis

[0282] 4.5.1 Worst-Case Scenario Analysis

[0283] Figures 10 to 11 This is the worst-case scenario for the electricity price in the second stage. It can be seen that there is a certain error between the actual electricity purchase price obtained in the robust optimization of the second stage and the electricity purchase price predicted in advance. From the worst electricity purchase price, it can be seen that the electricity price in the period from 20:00 to 03:00 is at the upper limit of the robust interval. The electricity purchase price in these periods is higher than the generated electricity purchase price scenarios, and the actual electricity purchase cost of the IES is higher than the predicted electricity purchase cost, which reflects that the IES operator has increased the operating cost to resist the impact of uncertainty. Similarly, there are also some errors between the actual electricity selling price in the second-stage optimization and the electricity selling price predicted in advance. The electricity selling prices from 12:00 to 13:00 and from 19:00 to 21:00 are at the lower limit of the robust interval, indicating that the actual electricity selling price is lower than the predicted electricity selling price at that time, because the IES reduces the electricity selling income at that time to prevent uncertainty risks.

[0284] Figures 12 to 14 This is the worst-case scenario for the power source and load in the third stage. It can be seen that the worst-case output of the wind power in the third stage is similar to the distribution of the electricity selling price. In most periods, it is at the lower extreme point of the fluctuation interval, and the actual output has decreased compared with the expected value. The system needs to purchase additional electric energy from the outside or increase the output of the unit to fill the difference between the actual output and the predicted output. Therefore, the robustness of this scenario is the strongest. The actual power of the multi-energy load is higher than the expected value, which is just the opposite of the distribution results of the wind power and the electricity selling price, but the impact is the same, that is, to resist the impact of the robustness in the fluctuation period, the system needs to purchase additional electric energy from the outside or increase the output of the unit.

[0285] 4.5.2 Comparison of Uncertain Optimization Methods

[0286] To verify the advantages of the three-stage adjustable robust optimization method of the present invention in dealing with uncertain problems, the present invention discusses 4 different uncertain modeling methods. Using the control variable method, the uncertain variables are regarded as experimental variables, and the remaining conditions are the same as the initial settings of the example, specifically as follows:

[0287] Scenario 1: The three-stage robust optimization considering the electricity price of the superior power grid and the uncertainty of the power source and load proposed by the present invention.

[0288] Scenario 2: The three-stage robust optimization of the robust model considering the uncertainty of the electricity price of the superior power grid.

[0289] Scenario 3: The three-stage robust optimization of the robust model considering the uncertainty of the power source and load.

[0290] Scenario 4: Using the method of deterministic optimization, the predicted values of the electricity price of the superior power grid and the source-load power are used as the optimization configuration data.

[0291] The operating costs of the IES under each scenario are shown in Table 4.

[0292] Table 4 Operating cost results of the IES under each scenario

[0293]

[0294] As can be seen from Table 4:

[0295] Compared with Scenario 1, the total cost in Scenario 2 is reduced by 15.47%. Among them, the gas purchase cost and the carbon trading cost are significantly reduced, by 13.5% and 32.71% respectively. This is because the uncertainty of the source-load in the IES is ignored, resulting in a significant decrease in the output of the CHP unit, reducing the gas purchase and carbon trading costs. Although Scenario 2 improves the economy of the IES operation, the formulated scheduling plan has a low adaptability to the actual source-load output, exaggerating the system's ability to absorb new energy.

[0296] Compared with Scenario 1, the change range of the total cost in Scenario 3 is not large. Among them, the power exchange cost is reduced by 120.9584 yuan, the gas purchase cost is reduced by 19 yuan, and the carbon trading cost is reduced by 0.2318 yuan. It can be seen that considering the uncertainty of the electricity price of the superior power grid will increase the power interaction cost between the IES and the superior power grid to offset the uncertainty of the electricity price, and has little impact on other costs in the system.

[0297] Compared with Scenario 1, in Scenario 4, neither of the two uncertainties is considered, resulting in a 18.1% reduction in the total system operating cost. Among them, the carbon trading cost is reduced by 33.74%, the power exchange cost is reduced by 19.89%, and the gas purchase cost is reduced by 15.15%. The economy of the system is improved significantly. However, this does not mean that deterministic optimization is superior to uncertain optimization. This is because the uncertainty risk and the rescheduling cost brought by the prediction errors of the electricity price of the superior power grid and the source-load output are not considered in deterministic optimization, but in fact, the uncertainty risk is underestimated, which will lead to higher actual scheduling costs.

[0298] 4.6 Sensitivity analysis of uncertain parameters

[0299] In the robust optimization model, the magnitude of the uncertain parameter represents the total duration of the uncertainty occurring during the entire scheduling period. Therefore, it reflects the degree of interference of the uncertainty on the robust optimization model. Figure 15 Illustrates the impact of the magnitude of the uncertain parameter on the total optimization cost of the IES.

[0300] Figure 15It is shown that as the uncertainty of electricity price and source load increases, the system operating cost also rises, indicating enhanced robustness. This positive correlation is due to the fact that larger uncertainty parameters mean that more uncertainty scenarios of trading prices and wind / solar power output are considered in the optimization process. Therefore, during the investment stage, the system continuously increases the installed capacity of storage, and during the scheduling process, the control cost is increased to mitigate various uncertainties. This leads to an increase in the investment cost of the device and an increase in the operating cost, thus resulting in an overall increase in the total cost. The analysis shows that decision-makers can effectively balance the economic efficiency and conservatism of the IES by selecting appropriate uncertainty parameters.

[0301] 4 Conclusions

[0302] To mitigate the impact of various uncertainties of the upper-layer grid electricity price and source load on the economic efficiency and security of the IES, a three-stage robust scheduling method is proposed. To manage multiple uncertainties, the spectral normalization generative adversarial network (SNWGAN) and K-means++ clustering are used to generate the initial scenarios of uncertainty variables. Based on these initial scenarios, a robust polyhedral uncertainty set is constructed using the 1-norm and box constraints, and a three-stage robust optimization scheduling model aiming to minimize the operating cost is developed. Then, the model is simplified using a combination of the two-layer column and constraint generation (CCG), Karush-Kuhn-Tucker (KKT) conditions, and the Big-M method, and directly solved using the commercial solver GUROBI. The main conclusions are as follows:

[0303] 1) Scenario generation and reduction: The SNWGAN and K-means++ clustering algorithms are used to generate and reduce the initial scenarios of uncertainty variables. This method effectively utilizes the historical data set to generate scenarios that are similar but not exactly the same as the historical data.

[0304] 2) Enhanced robustness: Compared with the traditional deterministic optimization model and other robust optimization (RO) models, the proposed three-stage RO model enhances the system's ability to mitigate uncertainties and is more in line with the actual scheduling scenario. In addition, the adopted algorithm solves the optimal model solution within 22.81 seconds.

[0305] 3) Sensitivity analysis: Through the sensitivity analysis of uncertainty parameters, it is found that as the uncertainty parameters increase, the total system optimization cost rises, and the robustness of the scheduling plan is enhanced. Decision-makers can balance the economic efficiency and conservatism of the IES by selecting appropriate uncertainty parameters.

[0306] As Figure 16 shown, the operating equipment of the integrated energy system based on the spectral normalization generative adversarial network includes a construction module and a scheduling module;

[0307] The construction module is used to obtain a set of typical scenarios of uncertainty variables, fuse the scenario information of uncertainty variables, adopt 1-norm constraint and box constraint to formulate a robust uncertainty set, and construct a three-stage robust optimization scheduling model with the goal of minimizing the operating cost;

[0308] The scheduling module is used to transform the three-stage robust optimization scheduling model into a single-layer mixed-integer model by using a simplification method that combines the double-layer CCG algorithm, Karush-Kuhn-Tucker conditions, and the big M method, and use a solver to solve it to obtain a scheduling plan.

[0309] As Figure 17 shown, the operating equipment of the integrated energy system based on the spectral normalization generative adversarial network can adopt various types of computing devices such as desktop computers, notebooks, palmtop computers, and cloud servers. This device consists of multiple components such as a processor and a memory, and is not limited thereto, and may include other components or different combinations. The device examples provided here are only for illustrative purposes and are not restrictive.

[0310] The processor can be various types of devices, such as CPU, DSP, ASIC, FPGA, etc., and can also include programmable logic devices, discrete gate, transistor logic devices, and discrete hardware components. The memory is used to store computer programs and modules. The processor runs or executes the computer program modules stored in the memory and calls the data stored in the memory, so as to realize various functions of the operation of the integrated energy system based on the spectral normalization generative adversarial network.

[0311] The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash device, or other volatile solid-state storage devices.

[0312] The present invention also provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for operating an integrated energy system based on a spectral normalization generative adversarial network are implemented.

[0313] When the operation method of the integrated energy system based on the spectral normalization generative adversarial network is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device.

[0314] Based on such an understanding, all or part of the processes in the above-mentioned operation method of the integrated energy system based on the spectral normalization generative adversarial network of the present invention can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned operation method of the integrated energy system based on the spectral normalization generative adversarial network can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or preset intermediate form, etc.

[0315] A computer-readable storage medium can refer to any entity or device that can carry computer program code, including recording media, USB flash drives, mobile hard disks, magnetic disks, optical disks, computer memories, read-only memories (ROMs), random access memories (RAMs), software distribution media, etc.

[0316] It should be noted that the content included 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, the computer-readable storage medium does not include electrical carrier signals and telecommunication signals.

[0317] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for operating an integrated energy system based on a spectral normalization generative adversarial network, characterized in that: include: Obtain a typical scenario set of uncertain variables, integrate the scenario information of uncertain variables, use 1-norm constraints and box constraints to formulate a robust uncertainty set, and build a three-stage robust optimization scheduling model for operating costs; The three-stage robust optimization scheduling model is transformed into a single-layer mixed integer model and solved using a solver to obtain a scheduling solution. The three-stage robust optimization scheduling model is converted into a single-layer mixed integer model, and the model is solved by the solver to obtain a scheduling solution. Specifically, the three-stage robust optimization scheduling model is obtained by processing multiple uncertainties based on polyhedral uncertainty sets, and the three-stage robust optimization scheduling model is decomposed and reconstructed to obtain a single-layer mixed integer model, which is then solved by the GUROBI solver to obtain a scheduling solution. The expression of the three-stage robust optimization scheduling model obtained by multiple uncertainty processing based on polyhedral uncertainty sets is as follows: （41） （42） In the formula, is the uncertainty set of the power interaction price between the integrated energy system and the upper power grid in the second stage, is the uncertainty set of sources and loads in the third stage of integrated energy system scheduling, , are electricity price and source load uncertainty variables, and are the expected values ​​of electricity price and source load uncertainty variables, , are the maximum and minimum values ​​of the uncertainty variable of electricity price, , They are The maximum and minimum values ​​of the uncertain 0-1 variable, , are the maximum and minimum values ​​of the source load uncertainty variables, , They are The maximum and minimum values ​​of the uncertain 0-1 variable, , They are and The uncertainty control parameter, is the electrical load, For t When heat load, , Respectively in t The purchase and sale prices of electricity exchanged between the integrated energy system and the upper power grid at all times, For t The output of the wind turbine at any moment; The full matrix formula of the three-stage robust optimization scheduling model is as follows: （43） （44） Where TE is the scheduling period, and the coefficient matrix of the first stage constraint condition is , , , the coefficient matrix of the second stage constraints , , , , the coefficient matrix of the third stage constraints , , , , , , , the sets of decision variables in the first, second and third stages are , , , , Respectively in t Purchase and sales at the moment 0-1 variable, , Respectively in t The charging and discharging status of the energy storage device at the moment 0-1 variable, , Respectively in t The power purchased and sold by the integrated energy system and the upper power grid at all times. , Respectively in t The electrical power and thermal power output of the cogeneration unit at any given moment, For t The electrical power consumed by the carbon capture unit to capture CO2 at any given moment, For t The electric power consumed by the hydrogen production unit at the moment, , Respectively in t The charging and discharging power of the energy storage device at all times, is the thermal power output of the gas boiler; The three-stage robust optimization scheduling model is decomposed and reconstructed to obtain a single-layer mixed integer model, which specifically includes: using the inner and outer nested CCG algorithm to decompose and reconstruct the three-stage robust optimization scheduling model into the main problem MP and the sub-problem SP, and the expression is as follows: （45） （46） After decomposition by the CCG algorithm, the multi-layer nested structure of the three-stage robust optimization scheduling model is transformed into a single-layer MILP structure using the KKT condition, thereby obtaining a single-layer mixed integer model. The specific transformation process is as follows: MP is expressed as (47) SP is expressed as (48) The main problem goal of the decomposed sub-problem is to solve the transaction power between the integrated energy system and the upper power grid under the worst transaction electricity price. The sub-problem goal of the sub-problem is to solve the unit dispatch plan under the worst integrated energy system source-load scenario. The expression is as follows: (49) (50) In the formula, is the worst solution to the uncertainty element obtained from the subproblem, and the auxiliary variable , auxiliary variables .

2. The method according to claim 1, characterized in that A typical scene set of uncertain variables is obtained, the scene information of uncertain variables is integrated, and the 1-norm constraint and box constraint are used to formulate a robust uncertainty set. Specifically, the spectral norm is introduced to normalize the weight matrix of the discriminator in WGAN-GP to constrain the discriminant performance, and the global 1-Lipschitz constraint is realized by calculating the singular values ​​of the weight matrix of each layer of the discriminator.

3. The method according to claim 2, characterized in that The objective function of WGAN-GP is as follows: （1） Assume the input renewable energy data set X , the discriminator includes the discriminator , , , , , and ; , for the input X have: （2） The set of weights and biases of each layer of the label discriminator : （3） get: （4） according to have to: （5） Weight Normalize: （6） For any X : （7） Use inequality (7) to calculate the singular values ​​of the weight matrix of each layer; in, for The spectral norm of is the unit distance, yes No. The weight of layer discrimination, yes No. The deviation of layer discrimination, is the total number of discriminator layers, yes The weight of the discriminant output, yes No. The dimension of the layer discriminant output, yes No. The weight of layer discrimination, yes No. The dimension of the layer discriminant output, yes No. The weight of layer discrimination, yes The spectral norm of yes The spectral norm of yes The spectral norm of is the Wasserstein distance between the real scene and the generated scene. To satisfy 1-Lipschitz continuity, For real scene The distribution law of is random noise The distribution probability of is the probability of generating a scenario, is the output of the discriminator, is the expected value of the input noise.

4. The method according to claim 1, characterized in that The three-stage robust optimization scheduling model for constructing operating costs specifically includes the first-stage scheduling model, the second-stage scheduling model and the third-stage scheduling model.

5. The method according to claim 4, characterized in that In the first stage scheduling model, the 0-1 variable switching cost is used as the objective function , including the power exchange status between the integrated energy system and the upper power grid, the charging and discharging status of the energy storage equipment, and the objective function The expression is: （8） The constraints are: （9） （10） Where TE is the scheduling period, is the cost coefficient for 0-1 variable switching, is the depreciation cost coefficient of charging and discharging energy storage equipment, , Respectively in t Purchase and sales at the moment 0-1 variable, , Respectively in t The charging and discharging status of the energy storage device at each moment is a 0-1 variable.

6. The method according to claim 5, characterized in that In the second stage dispatch model, the power exchange cost between the integrated energy system and the upper power grid is taken as the objective function. , the objective function The expression is: （11） The constraints are: （12） （13） In the formula, , Respectively in t The purchase and sale prices of electricity exchanged between the integrated energy system and the upper power grid at all times, , Respectively in t The power purchased and sold by the integrated energy system and the upper power grid at all times. , They are respectively the maximum values ​​of the purchased power and sold power exchanged between the integrated energy system and the upper-level power grid.

7. The method according to claim 6, characterized in that In the third stage dispatch model, the operating cost of the integrated energy system is used as the objective function , including gas purchase cost , Energy storage equipment operating costs Carbon trading costs , , and The expressions are: （14） （15） （16） The constraints include CHP-CCS-P2G system coupling model constraints, carbon trading constraints, energy storage equipment constraints, gas boiler model constraints, and integrated energy system power balance constraints; In the formula, The price of gas purchased from the upper grid for the integrated energy system. , , Respectively in t The natural gas consumed by the cogeneration unit, gas boiler and power hydrogen production unit at all times, , Respectively in t The charging and discharging power of the energy storage device at all times.

8. The method according to claim 7, characterized in that The constraint expressions of the CHP-CCS-P2G system coupling model are as follows: （17） （18） （19） （20） （21） （22） （23） （24） （25） （26） （27） （28） In the formula, is the calorific value of natural gas, , are the power generation and heat generation efficiencies of the cogeneration unit, , Respectively in t The electrical power and thermal power output of the cogeneration unit at any given moment, , are the maximum values ​​of the electrical power and thermal power output of the cogeneration unit, is the carbon emission conversion coefficient of the cogeneration unit, For t The carbon emission value of the cogeneration unit at each moment, For t The value of CO2 captured by the carbon capture unit at the moment; The power consumption factor for capturing CO2 in the carbon capture unit, For t The electrical power consumed by the carbon capture unit to capture CO2 at any given moment, The conversion factor between the amount of CO2 required to generate natural gas for the power-to-hydrogen unit and the power consumption, For t The value of CO2 consumed by the power hydrogen production unit at the moment, For t The electric power consumed by the hydrogen production unit at the moment, The efficiency of the power-to-hydrogen unit to convert natural gas, For t Net output of CHP-CCS-P2G system in coordinated operation at all times.

9. The method according to claim 8, characterized in that After CCS-P2G absorption and conversion, the actual carbon emissions of the CHP-CCS-P2G system and the total carbon emissions of the microgrid It is expressed as the following formula: （29） （30） In the formula, , is the carbon emission coefficient of the gas boiler and the upper power grid, The electric power purchased from the upper power grid for the integrated energy system. is the carbon emission value of the cogeneration unit, Natural gas consumed by gas boilers and power hydrogen production units, It is the purchased power exchanged between the integrated energy system and the upper-level power grid.

10. The method according to claim 9, characterized in that The carbon trading constraint expression is as follows: （31） （32） In the formula, is the electrothermal conversion coefficient, , The carbon emission coefficients of gas-fired units and coal-fired units, respectively, and the carbon quota of the integrated energy system in the carbon trading market , Selling carbon allowances on behalf of integrated energy systems, For t The thermal power output of the gas boiler at all times.

11. The method according to claim 10, characterized in that The energy storage device constraint expression is as follows: （33） （34） （35） （36） In the formula, , Respectively in t and t- The capacity of the energy storage device at 1 moment, , are the charging and discharging efficiency of the energy storage device, , are the upper and lower limits of the energy storage device capacity, , are the maximum values ​​of charging and discharging power of the energy storage device respectively.

12. The method according to claim 11, characterized in that The constraint expressions of the gas boiler model are as follows: （37） （38） In the formula, is the thermal power output of the gas boiler, For gas boiler operating efficiency, It is the upper limit of the thermal power output of the gas boiler.

13. The method according to claim 12, characterized in that The power balance constraint of the integrated energy system includes the electric power balance expression as follows: （39） And the thermal power balance, the expression is as follows: （40） In the formula, For t The output of the wind turbine at any moment, , Respectively in t The charging and discharging power of the energy storage unit at each moment, is the electrical load, , Respectively in t The charging and discharging power of the thermal energy storage unit at all times, For t When heat load.

14. The method according to claim 1, wherein: Then the GUROBI solver is used to solve the problem, and the scheduling solution includes: 1) Initialize the outer CCG algorithm problem, and initialize the upper and lower bounds of the outer layer as , , initialize the number of iterations , determine the convergence termination condition threshold ; 2) Solve the main problem of the outer CCG algorithm and obtain the optimal solution , update the lower bound of the outer problem , check whether the convergence condition is met ,in , if the convergence condition is met, the solution is terminated; otherwise, the solution continues; 3) Solve the inner CCG algorithm problem and initialize the inner CCG algorithm problem. The upper and lower bounds are initialized as , , initialize the number of iterations , determine the convergence termination condition threshold ; 4) Solve the main problem of the inner CCG algorithm and obtain the optimal solution , update the lower bound of the inner problem ; 5) Solve the inner CCG algorithm sub-problem and obtain the optimal solution , update the upper bound of the inner problem , check whether the convergence condition is met , if the convergence condition is met, terminate the inner problem and jump to 3); otherwise 6) continue to solve; 6) Add the variables and constraints of the second and third stages to the outer CCG algorithm main problem, and initialize the number of iterations Add one and return to 2).

15. Integrated energy system operation equipment based on spectral normalization generative adversarial network, characterized in that: include: The construction module is used to obtain a typical scenario set of uncertain variables, integrate the scenario information of uncertain variables, use 1-norm constraints and box constraints to formulate a robust uncertainty set, and build a three-stage robust optimization scheduling model for operating costs; The scheduling module converts the three-stage robust optimization scheduling model into a single-layer mixed integer model and uses the solver to solve it to obtain the scheduling plan; The three-stage robust optimization scheduling model is converted into a single-layer mixed integer model, and the model is solved by the solver to obtain a scheduling solution. Specifically, the three-stage robust optimization scheduling model is obtained by processing multiple uncertainties based on polyhedral uncertainty sets, and the three-stage robust optimization scheduling model is decomposed and reconstructed to obtain a single-layer mixed integer model, which is then solved by the GUROBI solver to obtain a scheduling solution. The expression of the three-stage robust optimization scheduling model obtained by multiple uncertainty processing based on polyhedral uncertainty sets is as follows: （41） （42） In the formula, is the uncertainty set of the power interaction price between the integrated energy system and the upper power grid in the second stage, is the uncertainty set of sources and loads in the third stage of integrated energy system scheduling, , are electricity price and source load uncertainty variables, and are the expected values ​​of electricity price and source load uncertainty variables, , are the maximum and minimum values ​​of the uncertainty variable of electricity price, , They are The maximum and minimum values ​​of the uncertain 0-1 variable, , are the maximum and minimum values ​​of the source load uncertainty variables, , They are The maximum and minimum values ​​of the uncertain 0-1 variable, , They are and The uncertainty control parameter, is the electrical load, For t When heat load, , Respectively in t The purchase and sale prices of electricity exchanged between the integrated energy system and the upper power grid at all times, For t The output of the wind turbine at any moment; The full matrix formula of the three-stage robust optimization scheduling model is as follows: （43） （44） Where TE is the scheduling period, and the coefficient matrix of the first stage constraint condition is , , , the coefficient matrix of the second stage constraints , , , , the coefficient matrix of the third stage constraints , , , , , , , the sets of decision variables in the first, second and third stages are , , , , Respectively in t Purchase and sales at the moment 0-1 variable, , Respectively in t The charging and discharging status of the energy storage device at the moment 0-1 variable, , Respectively in t The power purchased and sold by the integrated energy system and the upper power grid at all times. , Respectively in t The electrical power and thermal power output of the cogeneration unit at any given moment, For t The electrical power consumed by the carbon capture unit to capture CO2 at any given moment, For t The electric power consumed by the hydrogen production unit at the moment, , Respectively in t The charging and discharging power of the energy storage device at all times, is the thermal power output of the gas boiler; The three-stage robust optimization scheduling model is decomposed and reconstructed to obtain a single-layer mixed integer model, which specifically includes: using the inner and outer nested CCG algorithm to decompose and reconstruct the three-stage robust optimization scheduling model into the main problem MP and the sub-problem SP, and the expression is as follows: （45） （46） After decomposition by the CCG algorithm, the multi-layer nested structure of the three-stage robust optimization scheduling model is transformed into a single-layer MILP structure using the KKT condition, thereby obtaining a single-layer mixed integer model. The specific transformation process is as follows: MP is expressed as (47) SP is expressed as (48) The main problem goal of the decomposed sub-problem is to solve the transaction power between the integrated energy system and the upper power grid under the worst transaction electricity price. The sub-problem goal of the sub-problem is to solve the unit dispatch plan under the worst integrated energy system source-load scenario. The expression is as follows: (49) (50) In the formula, is the worst solution to the uncertainty element obtained from the subproblem, and the auxiliary variable , auxiliary variables .

16. The device according to claim 15, characterized in that A typical scene set of uncertain variables is obtained, the scene information of uncertain variables is integrated, and the 1-norm constraint and box constraint are used to formulate a robust uncertainty set. Specifically, the spectral norm is introduced to normalize the weight matrix of the discriminator in WGAN-GP to constrain the discriminant performance, and the global 1-Lipschitz constraint is realized by calculating the singular values ​​of the weight matrix of each layer of the discriminator.

17. The device according to claim 16, characterized in that The objective function of WGAN-GP is as follows: （1） Assume the input renewable energy data set X , the discriminator includes the discriminator , , , , , and ; , for the input X have: （2） The set of weights and biases of each layer of the label discriminator : （3） get: （4） according to have to: （5） Weight Normalize: （6） For any X : （7） Use inequality (7) to calculate the singular values ​​of the weight matrix of each layer; in, for The spectral norm of is the unit distance, yes No. The weight of layer discrimination, yes No. The deviation of layer discrimination, is the total number of discriminator layers, yes The weight of the discriminant output, yes No. The dimension of the layer discriminant output, yes No. The weight of layer discrimination, yes No. The dimension of the layer discriminant output, yes No. The weight of layer discrimination, yes The spectral norm of yes The spectral norm of yes The spectral norm of is the Wasserstein distance between the real scene and the generated scene. To satisfy 1-Lipschitz continuity, For real scene The distribution law of is random noise The distribution probability of is the probability of generating a scenario, is the output of the discriminator, is the expected value of the input noise.

18. The device according to claim 15, characterized in that The three-stage robust optimization scheduling model for constructing operating costs specifically includes the first-stage scheduling model, the second-stage scheduling model and the third-stage scheduling model.

19. The device according to claim 15, characterized in that In the first stage scheduling model, the 0-1 variable switching cost is used as the objective function , including the power exchange status between the integrated energy system and the upper power grid, the charging and discharging status of the energy storage equipment, and the objective function The expression is: （8） The constraints are: （9） （10） Where TE is the scheduling period, is the cost coefficient for 0-1 variable switching, is the depreciation cost coefficient of charging and discharging energy storage equipment, , Respectively in t Purchase and sales at the moment 0-1 variable, , Respectively in t The charging and discharging status of the energy storage device at each moment is a 0-1 variable.

20. The device according to claim 19, characterized in that In the second stage dispatch model, the power exchange cost between the integrated energy system and the upper power grid is taken as the objective function. , the objective function The expression is: （11） The constraints are: （12） （13） In the formula, , Respectively in t The purchase and sale prices of electricity exchanged between the integrated energy system and the upper power grid at all times, , Respectively in t The power purchased and sold by the integrated energy system and the upper power grid at all times. , They are respectively the maximum values ​​of the purchased power and sold power exchanged between the integrated energy system and the upper-level power grid.

21. The device according to claim 20, characterized in that In the third stage dispatch model, the operating cost of the integrated energy system is used as the objective function , including gas purchase cost , Energy storage equipment operating costs Carbon trading costs , , and The expressions are: （14） （15） （16） The constraints include CHP-CCS-P2G system coupling model constraints, carbon trading constraints, energy storage equipment constraints, gas boiler model constraints, and integrated energy system power balance constraints; In the formula, The price of gas purchased from the upper grid for the integrated energy system. , , Respectively in t The natural gas consumed by the cogeneration unit, gas boiler and power hydrogen production unit at all times, , Respectively in t The charging and discharging power of the energy storage device at all times.

22. The device according to claim 21, characterized in that The constraint expressions of the CHP-CCS-P2G system coupling model are as follows: （17） （18） （19） （20） （21） （22） （23） （24） （25） （26） （27） （28） In the formula, is the calorific value of natural gas, , are the power generation and heat generation efficiencies of the cogeneration unit, , Respectively in t The electrical power and thermal power output of the cogeneration unit at any given moment, , are the maximum values ​​of the electrical power and thermal power output of the cogeneration unit, is the carbon emission conversion coefficient of the cogeneration unit, For t The carbon emission value of the cogeneration unit at each moment, For t The value of CO2 captured by the carbon capture unit at the moment; The power consumption factor for capturing CO2 in the carbon capture unit, For t The electrical power consumed by the carbon capture unit to capture CO2 at any given moment, The conversion factor between the amount of CO2 required to generate natural gas for the power-to-hydrogen unit and the power consumption, For t The value of CO2 consumed by the power hydrogen production unit at the moment, For t The electric power consumed by the hydrogen production unit at the moment, The efficiency of the power-to-hydrogen unit to convert natural gas, For t Net output of CHP-CCS-P2G system in coordinated operation at all times.

23. The device according to claim 22, characterized in that After CCS-P2G absorption and conversion, the actual carbon emissions of the CHP-CCS-P2G system and the total carbon emissions of the microgrid It is expressed as the following formula: （29） （30） In the formula, , is the carbon emission coefficient of the gas boiler and the upper power grid, The electric power purchased from the upper power grid for the integrated energy system. is the carbon emission value of the cogeneration unit, Natural gas consumed by gas boilers and power hydrogen production units, It is the purchased power exchanged between the integrated energy system and the upper-level power grid.

24. The device according to claim 23, characterized in that The carbon trading constraint expression is as follows: （31） （32） In the formula, is the electrothermal conversion coefficient, , The carbon emission coefficients of gas-fired units and coal-fired units, respectively, and the carbon quota of the integrated energy system in the carbon trading market , Selling carbon allowances on behalf of integrated energy systems, For t The thermal power output of the gas boiler at all times.

25. The device according to claim 24, characterized in that The energy storage device constraint expression is as follows: （33） （34） （35） （36） In the formula, , Respectively in t and t- The capacity of the energy storage device at 1 moment, , are the charging and discharging efficiency of the energy storage device, , are the upper and lower limits of the energy storage device capacity, , are the maximum values ​​of charging and discharging power of the energy storage device respectively.

26. The device according to claim 25, characterized in that The constraint expressions of the gas boiler model are as follows: （37） （38） In the formula, is the thermal power output of the gas boiler, For gas boiler operating efficiency, It is the upper limit of the thermal power output of the gas boiler.

27. The device according to claim 26, characterized in that The power balance constraint of the integrated energy system includes the electric power balance expression as follows: （39） And the thermal power balance, the expression is as follows: （40） In the formula, For t The output of the wind turbine at any moment, , Respectively in t The charging and discharging power of the energy storage unit at each moment, is the electrical load, , Respectively in t The charging and discharging power of the thermal energy storage unit at all times, For t When heat load.

28. The device according to claim 27, characterized in that Then the GUROBI solver is used to solve the problem, and the scheduling solution includes: 1) Initialize the outer CCG algorithm problem, and initialize the upper and lower bounds of the outer layer as , , initialize the number of iterations , determine the convergence termination condition threshold ; 2) Solve the main problem of the outer CCG algorithm and obtain the optimal solution , update the lower bound of the outer problem , check whether the convergence condition is met ,in , if the convergence condition is met, the solution is terminated; otherwise, the solution continues; 3) Solve the inner CCG algorithm problem and initialize the inner CCG algorithm problem. The upper and lower bounds are initialized as , , initialize the number of iterations , determine the convergence termination condition threshold ; 4) Solve the main problem of the inner CCG algorithm and obtain the optimal solution , update the lower bound of the inner problem ; 5) Solve the inner CCG algorithm sub-problem and obtain the optimal solution , update the upper bound of the inner problem , check whether the convergence condition is met , if the convergence condition is met, terminate the inner problem and jump to 3); otherwise 6) continue to solve; 6) Add the variables and constraints of the second and third stages to the outer CCG algorithm main problem, and initialize the number of iterations Add one and return to 2).

29. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method according to any one of claims 1 to 14 when executing the computer program.

30. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 14 are implemented.

Images