Multi-region hydrogen-containing integrated energy system collaborative operation optimization method under uncertain environment

Abstract

The application discloses a multi-region hydrogen-containing comprehensive energy system collaborative operation optimization method under an uncertain environment, belongs to the cross field of energy systems and artificial intelligence, and comprises the following steps: establishing a multi-region zero-carbon hydrogen-containing comprehensive energy system collaborative operation optimization model under an uncertain environment considering electricity-heat-cold transaction; decomposing the collaborative operation optimization model into a regional in-worst-scenario operation cost minimization sub-problem and a multi-region collaborative operation optimization sub-problem; solving the regional optimization sub-problem based on a robust optimization algorithm to obtain worst-scenario information and corresponding independent operation cost of each region; solving the inter-regional collaborative operation optimization sub-problem based on a distributed optimization algorithm to obtain optimal energy transaction volume, an optimal scheduling strategy and a cooperative operation cost of each region; and distributing the total revenue of multi-region collaborative operation based on a non-linear energy mapping method. The method can reduce the operation cost of each regional system.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of energy systems and artificial intelligence, specifically a method for optimizing the collaborative operation of multi-region hydrogen-containing integrated energy systems under uncertain environments. Background Technology

[0002] To achieve net-zero carbon emissions, it is imperative to increase the utilization of renewable energy. However, renewable energy is intermittent and uncertain, making it difficult to achieve real-time matching of energy supply and demand without large-scale economic storage. To overcome this challenge, the introduction of hydrogen energy storage systems holds significant promise because they possess numerous advantages not found in other energy storage systems (such as large-capacity storage, long discharge time, zero pollution, ease of transportation, and independence from specific geographical conditions), enabling cross-seasonal energy storage and cross-regional energy distribution. Moreover, by coordinating the operation of hydrogen energy storage systems with other types of storage systems, system energy efficiency can be effectively improved. Against this backdrop, the operational optimization of integrated hydrogen energy systems has attracted widespread attention.

[0003] Existing research has proposed several methods for optimizing the operation of hydrogen-containing integrated energy systems, such as Lyapunov optimization, deep reinforcement learning, model predictive control, and robust optimization. To further improve the reliability and economy of system operation, some studies have conducted collaborative operation optimization of multi-regional hydrogen-containing integrated energy systems. However, existing research still has the following limitations: (1) the forms of energy trading between regions are too singular, mainly relying on electricity interconnection, which limits the flexibility of multi-energy complementarity; (2) existing models focus more on grid-connected systems and pay insufficient attention to off-grid zero-carbon systems; (3) there is a general lack of effective cross-regional benefit distribution mechanisms, making it difficult to ensure the sustainability of cooperation. Therefore, it is urgent to study the operation optimization and benefit distribution mechanism of multi-regional hydrogen-containing zero-carbon integrated energy systems that take into account electricity, heat, and cooling.

[0004] The above research faces the following challenges: (1) there are high-dimensional uncertainties related to renewable energy output and load; (2) there are hydrogen-containing integrated energy system groups belonging to different stakeholders; (3) there are spatial coupling constraints between energy transactions, and the energy storage level of the energy storage system will introduce time coupling constraints. In response to the above challenges, existing stochastic optimization methods rely on precise probability distributions to generate a large number of scenarios, which are computationally burdensome and rely on strong assumptions; centralized optimization has shortcomings such as poor scalability, weak privacy protection capabilities, and high consumption of communication and computing resources, making it difficult to solve optimization problems under multiple stakeholders; in order to overcome the shortcomings of existing methods, this invention proposes a multi-region hydrogen-containing zero-carbon integrated energy system operation optimization and benefit distribution mechanism based on distributed robust optimization and nonlinear energy mapping method, which takes into account electricity-heat-cold transactions. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an optimization method for the coordinated operation of multi-region hydrogen-containing integrated energy systems under uncertain environments. The aim is to achieve coordinated operation and reasonable distribution of benefits for multi-region hydrogen-containing zero-carbon integrated energy systems that take into account electricity-heat-cooling transactions, thereby reducing the system operating costs of integrated energy systems in each region.

[0006] To achieve the above objectives, the present invention employs the following technical solution:

[0007] This invention provides a method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions, including:

[0008] Step (1) Establish a collaborative operation optimization model for a multi-regional zero-carbon hydrogen-containing integrated energy system that takes into account electricity-heat-cooling transactions under uncertain conditions;

[0009] Step (2) decomposes the multi-region zero-carbon hydrogen integrated energy system collaborative operation optimization model into a sub-problem of minimizing operating costs under the worst scenario in the region and a sub-problem of collaborative operation optimization between multiple regions considering electricity-heat-cooling transactions;

[0010] Step (3) Based on the two-stage robust optimization framework and column and constraint generation algorithm, solve the subproblem of minimizing the operating cost under the worst scenario in the regional hydrogen-containing integrated energy system, and obtain the worst scenario information and corresponding independent operating cost of each region;

[0011] Step (4) Obtain the worst scenario information for each region in Step (3), and solve the cooperative operation optimization sub-problem of the multi-region hydrogen-containing integrated energy system, including electricity-heat-cold transactions, based on the alternating direction multiplier method of the enhanced adaptive prediction-correction mechanism, to obtain the optimal energy transaction volume between each region, the optimal scheduling strategy within each region, and the cooperative operation cost.

[0012] Step (5) obtains the independent operating cost of each region in step (3), the optimal energy trading volume and cooperative operating cost of each region in step (4), quantifies the contribution of different regions in energy sharing based on the nonlinear energy mapping method, and then allocates the total revenue of multi-regional collaborative operation.

[0013] Furthermore, the collaborative operation optimization model of the multi-regional zero-carbon hydrogen integrated energy system considering electricity-heat-cooling transactions under uncertain conditions established in step (1) includes decision variables, constraints, and objective functions.

[0014] Decision variables include Time slot region Purchase of hydrogen , Time slot region Actual output power of photovoltaic , Time slot region Electrolytic cell input power , Time slot region fuel cell output power , Time slot region Electric boiler input power , Time slot region Ground source heat pump heating input power , Time slot region Ground source heat pump cooling input power , Time slot region Excess heat energy from fuel cells , Time slot region Adsorption chiller input heat energy , Time slot region Hot water tank input heat power , Time slot region Hot water tank output heat power , Time slot region Cold water tank input cooling power , Time slot region Cold water tank output cooling power , Time slot region Electricity load demand , Time slot region Heat load demand , Time slot region Cooling load demand , Time slot region Maximum power generation of photovoltaic and Time slot region With the region Interactive electrical power Thermal power and cooling power .

[0015] The constraints include:

[0016] ,

[0017] ,

[0018] ，

[0019] ，

[0020] ，

[0021] ，

[0022] ，

[0023] ，

[0024] ，

[0025] ，

[0026] ，

[0027] ，

[0028] ，

[0029] ，

[0030] ，

[0031] ，

[0032] ，

[0033] ，

[0034] ，

[0035] ，

[0036] ，

[0037] ，

[0038] ，

[0039] ，

[0040] ，

[0041] ，

[0042] ，

[0043] ，

[0044] ，

[0045] ，

[0046] ，

[0047] ，

[0048] ，

[0049] ，

[0050] ，

[0051] ，

[0052] ，

[0053] ，

[0054] ，

[0055] ,

[0056] ,

[0057] Where: Formula (1) represents the photovoltaic operation constraints, for Time slot region The maximum photovoltaic power generation; Formulas (2) to (8) represent the operating constraints of the hydrogen energy storage system. for Time slot region Hydrogen tank storage capacity, and They are respectively regions Electrolyte-to-hydrogen conversion efficiency and fuel cell-to-hydrogen conversion efficiency Indicates the time slot length. For the region Maximum storage capacity of hydrogen tank and They are respectively regions Maximum input power of electrolyzer and maximum output power of fuel cell For the region Maximum hydrogen purchase volume for Time slot region The heat generated by the fuel cell and They are respectively regions The electro-thermal conversion efficiency and heat recovery efficiency of the fuel cell; Equations (9) to (13) represent the operating constraints of the hot water tank. for Time slot region The storage capacity of the hot water tank and They are respectively regions The input and output efficiency of the hot water tank. For the region Maximum storage capacity of hot water tank and They are respectively regions The maximum input heat power and maximum output heat power of the hot water tank; Formulas (14) to (18) represent the operating constraints of the cold water tank. for Time slot region The storage capacity of the cold water tank and They are respectively regions The cold water tank's input and output power efficiency. For the region Maximum storage capacity of cold water tank and They are respectively regions The maximum input cooling power and maximum output cooling power of the cold water tank; Formulas (19) to (20) represent the operating constraints of the electric boiler. for Time slot region Electric boilers output heat energy. For the region Electric boiler electro-thermal conversion efficiency For the region Maximum input power of electric boiler; Formulas (21) to (24) represent the operating constraints of ground source heat pump. and They are respectively Time slot region Ground source heat pumps output both heat and cold energy. and They are respectively regions The electro-thermal conversion efficiency and electro-cooling conversion efficiency of ground source heat pumps For the region The maximum input power of the ground source heat pump; formulas (25) to (29) represent the operating constraints of the geothermal well. for Time slot region Geothermal well storage capacity For the region Maximum storage capacity of geothermal wells; For the region The maximum value of the thermal energy injected into the geothermal well; Formulas (30) to (32) represent the operating constraints of the adsorption chiller. and They are respectively Time slot region The cooling energy output and power required by an adsorption chiller. and They are respectively regions The heat-to-cold conversion efficiency and the cold-to-electricity conversion efficiency of adsorption chillers for Time slot region The maximum heat absorption of the adsorption chiller; Formulas (33) to (38) represent the energy balance constraints. , as well as They are respectively Time slot region Losses of electrical energy, thermal energy, and cold energy; Equations (39) to (41) represent interactive coupling constraints. For the region With the region Maximum electrical power of the interaction; For the region With the region Maximum thermal power of the interaction; For the region With the region Maximum cooling power of the interaction;

[0058] The objective function includes:

[0059] ,

[0060] ,

[0061] ,

[0062] in: Indicates the number of collaborative operation regions; Indicates the region Uncertain variables; Indicates the region The set of internal operational decision variables; This represents the set of energy trading volumes between regions; Indicates the region Decisions regarding internal operations and uncertain variables The operating cost function; This indicates information regarding inter-regional energy trading volume. The interaction cost function; , , as well as Representing the period inner area The costs of photovoltaic power reduction, hydrogen energy system operation, hydrogen purchase costs, and energy demand loss. This represents a discount factor; Represents a scheduling cycle inner area Revenue from energy transactions with other regions Represents a scheduling cycle inner area Grid access fees incurred from energy interactions with other regions.

[0063] Furthermore, the subproblem of minimizing operating costs under the worst-case scenario within the aforementioned regional hydrogen-containing integrated energy system includes decision variables, constraints, and an objective function.

[0064] The set of decision variables in the operating cost minimization problem includes:

[0065] ,

[0066] ,

[0067] ,

[0068] in: , Representing regions The set of outer and inner optimization variables in a two-stage robust optimization framework; Indicates the region A set of uncertain variables , , , They are respectively Time slot region Forecast values ​​for maximum photovoltaic power generation, electricity load demand, heat load demand, and cooling load demand. , , , They are respectively Time slot region Deviations in maximum photovoltaic power generation, electrical load demand, heat load demand, and cooling load demand. , , , These are binary variables corresponding to photovoltaic power generation, electrical load, heat load, and cooling load, respectively. , , , These represent the uncertainties corresponding to photovoltaic power generation, electrical load, heat load, and cooling load, respectively. for Time slot region Operating indicators of the electrolytic cell for Time slot region The start-up indicator of the electrolytic cell. for Time slot region The shut-off sign of the electrolytic cell. for Time slot region Operating indicators of fuel cells for Time slot region The start-up indicator for fuel cells, for Time slot region Fuel cell shutdown indicator; , , , , , , These are the auxiliary variables corresponding to heat generation by ground source heat pump, cold generation by ground source heat pump, waste heat absorption by geothermal well, heat absorption by hot tank, heat release by hot tank, cold absorption by cold tank, and cold release by cold tank, respectively. , , , , , , , , , , , , , , , , , , , , , , , Same as defined in step (1);

[0069] The objective function and constraints in the subproblem of minimizing operating costs include:

[0070] ,

[0071] ,

[0072] ,

[0073] ,

[0074] ,

[0075] ,

[0076] ,

[0077] ,

[0078] ,

[0079] ,

[0080] ,

[0081] ,

[0082] ,

[0083] ,

[0084] ,

[0085] ,

[0086] ,

[0087] ,

[0088] ,

[0089] ,

[0090] ,

[0091] ,

[0092] ,

[0093] in: and Together constituting the region The set of internal operational decisions, namely ; Indicates the next scheduling cycle under the worst-case scenario. inner area Independent operating costs; , , , , , Same as defined in step (1); This represents the cost reduction coefficient for photovoltaics. and These represent two coefficients respectively; , and These represent the costs of operating, starting up, and shutting down the electrolyzer and fuel cell, respectively. , and These are indicators representing the operation, start-up, and shutdown of the electrolyzer and fuel cell, respectively. This represents a collection of electrolyzers and fuel cell equipment; The formula represents the price of hydrogen purchase; formulas (54) to (66) and (70) represent the constraints introduced by auxiliary variables; formulas (67) to (69) represent the system... Energy self-balancing constraint under no-trading conditions.

[0094] Furthermore, the optimization sub-problem of the coordinated operation of the multi-regional hydrogen-containing integrated energy system, including electricity-heat-cooling transactions, is as follows:

[0095] ,

[0096] ,

[0097] ,

[0098] ,

[0099] in: express The comprehensive economic cost of a regional integrated energy system under the worst-case scenario Represents a scheduling cycle inner area Total operating costs , Same as defined in step (1); , , They represent in time Region and The electrical, thermal, and cooling power of regional interaction; , , Representing regions The unit price paid for the interaction of electrical power, thermal power, and cooling power; when electrical power Thermal power Cooling power All are greater than 0, indicating that the region To the region Provide energy; otherwise, obtain energy from region j. , , These represent the grid connection fees for units of electrical energy, heat energy, and cold energy, respectively.

[0100] Furthermore, the steps for solving the subproblem of minimizing operating costs under the worst-case scenario in a regional hydrogen-containing integrated energy system based on the two-stage robust optimization framework and column and constraint generation algorithm are as follows:

[0101] (5.1) The problem of minimizing the operating cost under the worst scenario in the proposed regional hydrogen-containing integrated energy system is transformed into the main problem and sub-problems of a two-stage robust optimization model;

[0102] (5.2) Based on strong duality theory, the subproblems of the two-stage robust optimization model are transformed into their dual problems;

[0103] (5.3) Based on the column and constraint generation algorithm, the two-stage robust optimization model is solved iteratively to obtain the worst-case scenario. and corresponding optimal operating cost ;

[0104] (5.4) The worst-case scenario for each region Inputting the data into a collaborative operation optimization model that considers electricity-heat-cooling transactions among multi-regional hydrogen-containing integrated energy systems, the independent optimal operating costs under the worst-case scenario for each region are determined. Input into the benefit distribution model.

[0105] Furthermore, the compact form of the main problem of the two-stage robust optimization model is as follows:

[0106] ,

[0107] ,

[0108] ,

[0109] ,

[0110] ,

[0111] ,

[0112] ,

[0113] ,

[0114] ,

[0115] ,

[0116] ,

[0117] ,

[0118] ,

[0119] in: and In formula (48) The corresponding coefficient column vector sum The corresponding coefficient column vector; , , , , , , , , , , , These are the coefficient matrices of the variables under the corresponding constraints; , , , , , , , , , , , These are constant column vectors; formula (77) represents the expression in step (2) regarding... The inequality constraints include formulas (63) to (66); formula (78) represents the inequality constraints in step (2) regarding... Equality constraints; Formula (79) represents the equation constraint in step (2) regarding The equality constraints include formulas (8) to (9), formula (14), formula (19), formula (21) to (22), formula (25), and formula (30) to (31); formula (80) represents the equality constraints in step (2) regarding... The inequality constraints include formula (28); formula (81) represents the inequality constraints in step (2) regarding and The equality constraints include formula (2); formula (82) represents the equation constraint in step (2) regarding and The inequality constraints include formulas (54) to (62); formula (83) represents the inequality constraints in step (2) regarding... and The equality constraints include formulas (67) to (69); formula (84) represents the equality constraints in step (2) regarding... and The inequality constraints include formulas (1), (36), and (38); formula (85) is the one in step (2). The upper and lower limit constraints include formulas (3) to (5), formulas (10) to (12), formulas (15) to (17), formula (20), formulas (23) to (24) and formulas (26) to (27); formula (86) is step (2). The upper and lower limits are constrained by formulas (6) and (70). This represents the maximum number of iterations. Auxiliary variables introduced; For the first The region obtained in the second iteration Optimal strategy for subproblems; For the first The region obtained by solving the subproblem in the next iteration The worst-case scenario for new energy output and load power.

[0120] The compact form of the subproblem of the two-stage robust optimization model is as follows:

[0121] ,

[0122] ,

[0123] in: Indicates a given set Time zone Optimize variables feasible domain, , , , , , , , Let represent the dual variables corresponding to each constraint in the minimization problem of the second stage. Same as defined in step (2).

[0124] In a given Minimizing the inner layer of the subproblem is a linear problem, which, according to strong duality theory and correspondence, can be transformed into... Form, and with the outer layer Problem merging, the dual problem of the subproblem proposed in (5.2) is as follows:

[0125] ,

[0126] ,

[0127] ,

[0128] Formula (91) contains bilinearity and When the dual problem reaches its maximum value, the uncertain variable... The value of should be the boundary of the fluctuation range described by the source-load uncertainty set. In the hydrogen-containing integrated energy system studied in this paper, the operating cost of the system is higher when the photovoltaic output reaches the minimum value of the range and the load power reaches the maximum value of the range, which is more in line with the definition of the "worst-case scenario". Therefore, all extreme points of the uncertainty set can be characterized by binary variables and linear constraints, and the Big-M method can be used to transform the bilinear terms into easily solvable linear terms. Source-load uncertainty set Rewrite it in the following form:

[0129] ,

[0130] bilinear terms and Rewritten as:

[0131] ,

[0132] ,

[0133] in: Source load prediction data matrix; This is the source load prediction deviation matrix; and This is a continuous auxiliary variable that is introduced.

[0134] The relevant constraints of the Big-M method are:

[0135] ,

[0136] in: It is a sufficiently large positive real number; For binary variables, a value of 1 indicates that the uncertain variable for the corresponding time period has reached the boundary of the interval.

[0137] Therefore, the objective function in the dual form of the subproblem can be expressed as:

[0138] ,

[0139] Furthermore, the steps for iteratively solving the two-stage robust optimization model based on the column and constraint generation algorithm are as follows:

[0140] (7.1) Initialize, set the region The power output and load prediction curves of new energy sources are set as the initial scenario. The upper bound of the initial optimization problem is The lower bound is Number of iterations ;

[0141] (7.2) Substitute the output and load power of the new energy units under the initialization scenario into the main problem and solve it to obtain the optimal solution of the main problem. At the same time, update the lower bound. ;

[0142] (7.3) Fixed Substitute the subproblems into the subproblems and solve them to obtain the optimal solution to the subproblems. and harsh scenes and update the upper bound. ;

[0143] (7.4) The convergence threshold of the given algorithm is ,like Then stop iterating and return the optimal solution. and Otherwise, add variables. and the following constraints:

[0144] ,

[0145] Furthermore, the steps for solving the sub-problem of coordinated operation optimization of a multi-region hydrogen-containing integrated energy system based on the alternating direction multiplier method with an enhanced adaptive prediction-correction mechanism are as follows:

[0146] (8.1) Obtain the worst-case scenario information for each region in step 5.3. ;

[0147] (8.2) Based on the enhanced adaptive prediction-correction mechanism and the alternating direction multiplier method, the optimal energy trading volume is obtained by distributed solution of the sub-problem of coordinated operation optimization of multi-region hydrogen-containing integrated energy system. Optimal Energy Trading Volume II Optimal energy trading volume three Optimal scheduling strategy for each region during collaborative runtime and operating costs ;

[0148] (8.3) The optimal energy trading volume Optimal Energy Trading Volume II Optimal energy trading volume three Operating costs of each region during collaborative runtime Input into the benefit redistribution model, output the optimal energy trading volume for each region. Optimal Energy Trading Volume II Optimal energy trading volume three and optimal scheduling strategy .

[0149] Furthermore, the steps for the distributed solution of the multi-region hydrogen-containing integrated energy system collaborative operation optimization sub-problem based on the alternating direction multiplier method with enhanced adaptive prediction-correction mechanism are as follows:

[0150] The objective function of the optimization problem for the coordinated operation of a multi-regional hydrogen-containing integrated energy system is to minimize the total cost of the cooperative alliance. Due to the existence of dual-variable coupling in the problem, auxiliary variables are introduced for decoupling before solving the problem:

[0151] ,

[0152] in: Types of energy trading; For the region Expectations and Regions Energy trading volume; For the region Expectations and Regions Energy trading volume;

[0153] By region For example, the augmented Lagrangian function of the objective function in cooperative runtime. as follows:

[0154] ,

[0155] in: and These represent the Lagrange multiplier and the penalty factor, respectively.

[0156] The specific iterative steps of the algorithm are as follows:

[0157] (9.1) Each region is accessed via the Lagrange function. Calculate the expected energy trading volume as the forecast value to obtain the predicted energy trading volume. , and Lagrange multipliers ;

[0158] ,

[0159] (9.2) Sharing the first The next iteration predicts energy trading volume. , and Lagrange multipliers The expected energy trading volume is obtained based on the correction steps. , and Lagrange multipliers as follows:

[0160] ,

[0161] in: This is a correction factor, whose function is to correct the relationship between the predicted value and the previous iteration value;

[0162] (9.3) Calculate the original residuals according to formulas (103) to (104). and dual residual ;

[0163] ,

[0164] ,

[0165] (9.4) Determine the convergence status of the algorithm;

[0166] ,

[0167] in: and This is the iterative convergence accuracy value; if it satisfies formula (105), then the algorithm converges, and the output is the optimal energy trading volume. Costs of each region during collaborative runtime ; , as well as These represent the original residuals of the trading volumes of electrical energy, thermal energy, and cold energy, respectively. , as well as These represent the dual residuals of the trading volumes of electrical energy, thermal energy, and cold energy, respectively.

[0168] (9.5) If formula (105) is not satisfied, return to step (9.1) to continue iterating and adaptively update the penalty factor. as follows:

[0169] ,

[0170] in: This is an empirical coefficient.

[0171] Furthermore, the steps for benefit allocation based on the nonlinear energy mapping method are as follows:

[0172] When systems cooperate in a game, giving energy to other systems or receiving energy from other systems is considered a contribution to the alliance. Generally, the price of energy purchased from a supplier is higher than the price sold by the supplier; therefore, the contribution of providing energy to other systems is considered greater than the contribution of receiving the same amount of energy. The specific calculation process for quantifying the degree of contribution is described below:

[0173] (10.1) Obtain the independent operating costs of each region in step 5.3. Optimal energy trading volume in step 8.3 , , Operating costs of each region during collaborative runtime First, calculate the energy that each region contributes to the whole when participating in the cooperation. and the energy obtained from the total , Indicates the number of collaborative operation regions; Indicates the scheduling period. Indicates time slot, and Region:

[0174] ,

[0175] ,

[0176] (10.2) Then, the contribution of each region is quantified according to the degree of participation in energy exchange, and the natural constant is selected. Quantifying the contribution of an exponential function mathematical model with base as the base :

[0177] ,

[0178] in, This represents the maximum energy supply in each region; This represents the maximum energy gain in each region;

[0179] (10.3) Finally, output the final cost for each region. As shown below:

[0180] .

[0181] In summary, the technical solution of this invention can achieve the scheduling and control of a multi-regional zero-carbon hydrogen-containing integrated energy system under the consideration of four uncertain factors: photovoltaic, electricity load demand, heat load demand, and cooling load demand. Furthermore, it can distribute the benefits after the system's coordinated operation based on the contribution of each region to energy sharing. Compared with existing technologies, it achieves the following beneficial effects:

[0182] (1) Compared with the independent operation mode of each region, the multi-region zero-carbon hydrogen integrated energy system collaborative operation optimization model for uncertain environment considering electricity-heat-cold trading proposed in this invention can identify the worst scenario in each region and optimize scheduling decisions, thereby improving the system's risk resistance; rationally allocate energy, reduce system load loss, and improve the system's energy supply reliability; reduce system operating costs and improve the economic benefits of the entire energy network.

[0183] (2) Compared with existing algorithms, the present invention can handle high uncertainty and multi-agent collaborative problems at the same time. The proposed method does not require manual selection of appropriate penalty factors. The addition of the prediction correction mechanism can accelerate the convergence of the algorithm and achieve results closer to centralized solutions in fewer iterations. Attached Figure Description

[0184] Figure 1 This is a flowchart of the method for optimizing the coordinated operation of a multi-region hydrogen-containing integrated energy system under uncertain conditions proposed in this invention;

[0185] Figure 2 This is a diagram illustrating the convergence process of the algorithm proposed in this invention. Figure 2 In diagram (a), the cost values ​​for each region are distributed and iterated. Figure 2(b) shows the iterative process of the total cost value of the cooperative alliance. Detailed Implementation

[0186] The present invention will now be described in further detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for illustrating the technical solutions of the present invention more clearly, and should not be construed as limiting the scope of protection of the present invention.

[0187] Example: This example uses three integrated hydrogen energy systems as cases for analysis. In this example, the electrical load, heat load, cooling load, and photovoltaic power generation data are all selected from typical daily data of a certain region. This example uses the Matlab 2021a compilation environment and the Cplex solver called through the Yalmip toolbox to perform distributed solving of the model.

[0188] To verify the feasibility and effectiveness of the method of the present invention, five comparative schemes were introduced:

[0189] Option 1 operates independently in each region. Within the scope of uncertainty factors, 1000 scenarios are randomly selected to minimize the operating cost of each region. The scenario corresponding to the maximum operating cost after optimization in each region is selected.

[0190] Option 2 allows each region to operate independently, employing a robust optimization algorithm to minimize the operating cost of each region under the worst-case scenario.

[0191] Option 3 involves collaborative operation across different regions, employing a centralized robust optimization algorithm to minimize the cost of multi-regional collaborative operation under the worst-case scenario.

[0192] Option 4 involves collaborative operation across regions, employing a distributed robust optimization algorithm based on the alternating multiplier method with predictive correction (PCB-ADMM) to minimize the cost of multi-region collaborative operation under the worst-case scenario.

[0193] Option 5 involves collaborative operation across regions, employing a distributed robust optimization algorithm based on the traditional Alternating Multiplier Method (ADMM) to minimize the cost of multi-regional collaborative operation under the worst-case scenario.

[0194] The feasibility of the proposed method is verified from two aspects: the effectiveness in identifying the worst-case scenario and the convergence of the proposed algorithm. As shown in Table 1, compared to Scheme 1, the optimized cost value of the proposed method for the worst-case scenario identified in each region is higher than the cost value when randomly selecting 1000 scenarios in each region, thus verifying the effectiveness of the proposed method in identifying the worst-case scenario. Figure 2 It can be seen that when the A-PCB-ADMM algorithm iterates to 73 times, the cost functions of each region converge, thus verifying the effectiveness of the original problem decomposition. The use of distributed algorithms can make the variables in the coupled regions infinitely close, that is, the original computational pressure can be alleviated by decomposing the original problem.

[0195] Table 1 - Comparison of Severe Scenery in Different Regions

[0196]

[0197] The effectiveness of the method of this invention was verified by comparing the independent operation mode and the cooperative operation mode. As shown in Table 2, the total cost under the cooperative operation mode proposed by the method of this invention is RMB 91,501.34, while the total cost under the independent operation mode of each region is RMB 99,403.60, representing a 7.95% reduction in total cost. Under the cooperative operation mode, the load demand loss costs for Region 1 and Region 3 are RMB 14,071.17 and RMB 20,855.47, respectively, while under the independent operation mode of each region, the load demand loss costs for Region 1 and Region 3 are RMB 8,363.67 and RMB 8,194.60, respectively, representing reductions of 40.55% and 60.7%. It is evident that the method of this invention can reduce system load loss, improve system power supply reliability, reduce system operating costs, and improve the economic efficiency of the entire energy network. Under the cooperative operation mode, the costs of Region 2 before profit distribution were higher than those under the independent operation mode, at RMB 19,538.16 and RMB 14,830.535 respectively. After profit redistribution, the costs of each region decreased by 4.53%, 29.67%, and 3.84% respectively compared to the independent operation mode. This demonstrates that the method of this invention can reasonably allocate profits, incentivize system participation in cooperation, and enhance the stability of the alliance.

[0198] Table 2 - Comparison of Independent Operation Mode and Cooperative Operation Mode

[0199]

[0200] The superiority of the proposed method is verified by comparing the performance of distributed algorithms. As shown in Table 3, the traditional ADMM algorithm and the PCB-ADMM algorithm converge in 213 and 206 iterations, respectively, while the proposed method converges in the 73rd iteration. This demonstrates that the proposed method achieves results closer to centralized solutions in fewer iterations, thus improving the efficiency of practical applications.

[0201] Table 3 - Comparison of Performance Metrics of Distributed Algorithms

[0202]

Claims

1. A method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions, characterized in that, Includes the following steps: Step 1: Establish a collaborative operation optimization model for a multi-regional zero-carbon hydrogen-containing integrated energy system considering electricity-heat-cooling transactions under uncertain conditions; Step 2: Decompose the multi-regional zero-carbon hydrogen integrated energy system collaborative operation optimization model into a sub-problem of minimizing operating costs under the worst-case scenario within the region and a multi-regional collaborative operation optimization sub-problem considering electricity-heat-cooling transactions; Step 3: Based on the two-stage robust optimization framework and column and constraint generation algorithm, solve the subproblem of minimizing the operating cost under the worst scenario in the regional hydrogen-containing integrated energy system, and obtain the worst scenario information and corresponding independent operating cost of each region; The steps in step 3, which propose a two-stage robust optimization framework and a column and constraint generation algorithm to solve the subproblem of minimizing operating costs under the worst-case scenario in the regional hydrogen-containing integrated energy system, are as follows: (5.1) The problem of minimizing the operating cost under the worst scenario in the proposed regional hydrogen-containing integrated energy system is transformed into the main problem and sub-problems of a two-stage robust optimization model; (5.2) Based on strong duality theory, the subproblems of the two-stage robust optimization model are transformed into their dual problems; (5.3) Based on the column and constraint generation algorithm, the two-stage robust optimization model is solved iteratively to obtain the worst-case scenario. and corresponding optimal operating cost ; (5.4) The worst-case scenario for each region Inputting the data into a collaborative operation optimization model that considers electricity-heat-cooling transactions among multi-regional hydrogen-containing integrated energy systems, the independent optimal operating costs under the worst-case scenario for each region are determined. Input into the benefit distribution model; Step 4: Obtain the worst-case scenario information for each region in Step 3, and solve the cooperative operation optimization sub-problem of multi-region hydrogen-containing integrated energy system and electricity-heat-cold trading based on the alternating direction multiplier method of the enhanced adaptive prediction-correction mechanism, so as to obtain the optimal energy trading volume between regions, the optimal scheduling strategy within each region, and the cooperative operation cost. The steps in step 4 for solving the sub-problem of coordinated operation optimization of a multi-region hydrogen-containing integrated energy system based on the alternating direction multiplier method with an enhanced adaptive prediction-correction mechanism are as follows: (8.1) Obtain the worst-case scenario information for each region in step (5.3). ; (8.2) Based on the enhanced adaptive prediction-correction mechanism and the alternating direction multiplier method, the optimal energy trading volume is obtained by distributed solution of the sub-problem of coordinated operation optimization of multi-region hydrogen-containing integrated energy system. Optimal Energy Trading Volume II Optimal energy trading volume three Optimal scheduling strategy for each region during collaborative runtime and operating costs ; (8.3) The optimal energy trading volume Optimal Energy Trading Volume II Optimal energy trading volume three Operating costs of each region during collaborative runtime Input into the benefit distribution model, output the optimal energy trading volume. Optimal Energy Trading Volume II Optimal energy trading volume three and optimal scheduling strategy ; Step 5: Obtain the independent operating costs of each region in Step 3, the optimal energy trading volume and cooperative operating costs of each region in Step 4, quantify the contribution of different regions in energy sharing based on the nonlinear energy mapping method, and then allocate the total revenue of multi-regional collaborative operation.

2. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 1, characterized in that, The collaborative operation optimization model of the multi-regional zero-carbon hydrogen-containing integrated energy system considering electricity-heat-cooling transactions under uncertain conditions established in step 1 includes decision variables, constraints, and objective functions; Decision variables include Time slot region Purchase of hydrogen , Time slot region Actual output power of photovoltaic , Time slot region Electrolytic cell input power , Time slot region fuel cell output power , Time slot region Electric boiler input power , Time slot region Ground source heat pump heating input power , Time slot region Ground source heat pump cooling input power , Time slot region Excess heat energy from fuel cells , Time slot region Adsorption chiller input heat energy , Time slot region Hot water tank input heat power , Time slot region Hot water tank output heat power , Time slot region Cold water tank input cooling power , Time slot region Cold water tank output cooling power , Time slot region Electricity load demand , Time slot region Heat load demand , Time slot region Cooling load demand , Time slot region Maximum power generation of photovoltaic and Time slot region With the region Interactive electrical power Thermal power and cooling power ; The constraints include: (1)， (2)， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， Where: Formula (1) represents the photovoltaic operation constraints, for Time slot region The maximum photovoltaic power generation; Formulas (2) to (8) represent the operating constraints of the hydrogen energy storage system. for Time slot region Hydrogen tank storage capacity, and They are respectively regions Electrolyte-to-hydrogen conversion efficiency and fuel cell-to-hydrogen conversion efficiency Indicates the time slot length. For the region Maximum storage capacity of hydrogen tank and They are respectively regions Maximum input power of electrolyzer and maximum output power of fuel cell For the region Maximum hydrogen purchase volume for Time slot region The heat generated by the fuel cell and They are respectively regions The electro-thermal conversion efficiency and heat recovery efficiency of the fuel cell; Equations (9) to (13) represent the operating constraints of the hot water tank. for Time slot region The storage capacity of the hot water tank and They are respectively regions The input and output efficiency of the hot water tank. For the region Maximum storage capacity of hot water tank and They are respectively regions The maximum input heat power and maximum output heat power of the hot water tank; Formulas (14) to (18) represent the operating constraints of the cold water tank. for Time slot region The storage capacity of the cold water tank and They are respectively regions The cold water tank's input and output power efficiency. For the region Maximum storage capacity of cold water tank and They are respectively regions The maximum input cooling power and maximum output cooling power of the cold water tank; Formulas (19) to (20) represent the operating constraints of the electric boiler. for Time slot region Electric boilers output heat energy. For the region Electric boiler electro-thermal conversion efficiency For the region Maximum input power of electric boiler; Formulas (21) to (24) represent the operating constraints of ground source heat pump. and They are respectively Time slot region Ground source heat pumps output both heat and cold energy. and They are respectively regions The electro-thermal conversion efficiency and electro-cooling conversion efficiency of ground source heat pumps For the region The maximum input power of the ground source heat pump; formulas (25) to (29) represent the operating constraints of the geothermal well. for Time slot region Geothermal well storage capacity For the region Maximum storage capacity of geothermal wells; For the region The maximum value of the thermal energy injected into the geothermal well; Formulas (30) to (32) represent the operating constraints of the adsorption chiller. and They are respectively Time slot region The cooling energy output and power required by an adsorption chiller. and They are respectively regions The heat-to-cold conversion efficiency and the cold-to-electricity conversion efficiency of adsorption chillers for Time slot region The maximum heat absorption of the adsorption chiller; Formulas (33) to (38) represent the energy balance constraints. , as well as They are respectively Time slot region Losses of electrical energy, thermal energy, and cold energy; Equations (39) to (41) represent interactive coupling constraints. For the region With the region Maximum electrical power of the interaction; For the region With the region Maximum thermal power of the interaction; For the region With the region Maximum cooling power of the interaction; The objective function includes: ， ， ， in: Indicates the number of collaborative operation regions; Indicates the region Uncertain variables; Indicates the region The set of internal operational decision variables; This represents the set of energy trading volumes between regions; Indicates the region Decisions regarding internal operations and uncertain variables The operating cost function; This indicates information regarding inter-regional energy trading volume. The interaction cost function; , , as well as Representing the period inner area The costs of photovoltaic power reduction, hydrogen energy system operation, hydrogen purchase costs, and energy demand loss. This represents a discount factor; Represents a scheduling cycle inner area Revenue from energy transactions with other regions Represents a scheduling cycle inner area Grid access fees incurred from energy interactions with other regions.

3. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 2, characterized in that, Step 2 involves the subproblem of minimizing operating costs under the worst-case scenario within the regional hydrogen-containing integrated energy system, which includes decision variables, objective function, and constraints. The set of decision variables in the subproblem of minimizing operating costs includes: ， ， ， in: , Representing regions The set of outer and inner optimization variables in a two-stage robust optimization framework; Indicates the region A set of uncertain variables , , , They are respectively Time slot region Forecast values ​​for maximum photovoltaic power generation, electricity load demand, heat load demand, and cooling load demand. , , , They are respectively Time slot region Deviations in maximum photovoltaic power generation, electrical load demand, heat load demand, and cooling load demand. , , , These are binary variables corresponding to photovoltaic power generation, electrical load, heat load, and cooling load, respectively. , , , These represent the uncertainties corresponding to photovoltaic power generation, electrical load, heat load, and cooling load, respectively. for Time slot region Operating indicators of the electrolytic cell for Time slot region The start-up indicator of the electrolytic cell. for Time slot region The shut-off sign of the electrolytic cell. for Time slot region Operating indicators of fuel cells for Time slot region The start-up indicator for fuel cells, for Time slot region Fuel cell shutdown indicator; , , , , , , These are the auxiliary variables corresponding to heat generation by ground source heat pump, cold generation by ground source heat pump, waste heat absorption by geothermal well, heat absorption by hot tank, heat release by hot tank, cold absorption by cold tank, and cold release by cold tank, respectively. Indicates the region Uncertain variables, for Time slot region The amount of hydrogen purchased, for Time slot region Electricity load demand for Time slot region Heat load demand, for Time slot region Cooling load demand for Time slot region Maximum power generation of photovoltaic power for Time slot region Actual output power of photovoltaics for Time slot region Electrolytic cell input power, for Time slot region Fuel cell output power, for Time slot region Electric boiler input power, for Time slot region Ground source heat pump heating input power, for Time slot region Ground source heat pump cooling input power, for Time slot region Fuel cells generate excess heat energy. for Time slot region Geothermal well storage capacity for Time slot region Adsorption chillers input heat energy, for Time slot region Hydrogen tank storage capacity, for Time slot region The hot water tank input heat power, for Time slot region The hot water tank outputs heat power. for Time slot region The storage capacity of the hot water tank for Time slot region The cold water tank input cooling power, for Time slot region The cold water tank outputs cooling power. for Time slot region The storage capacity of the cold water tank , , They are respectively Time slot region Losses of electrical energy, thermal energy, and cold energy; The objective function and constraints in the subproblem of minimizing operating costs include: ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， ， in: and Together constituting the region The set of internal operational decisions, namely ; Indicates the next scheduling cycle under the worst-case scenario. inner area Independent operating costs; Indicates the region Decisions regarding internal operations and uncertain variables The running cost function, , , , Representing the period inner area The costs of photovoltaic power reduction, hydrogen energy system operation, hydrogen purchase costs, and energy demand loss. This represents a discount factor; This represents the cost reduction coefficient for photovoltaics. and Indicates two coefficients; , and These represent the costs of operating, starting up, and shutting down the electrolyzer and fuel cell, respectively. , and These are indicators representing the operation, start-up, and shutdown of the electrolyzer and fuel cell, respectively. This represents a collection of electrolyzers and fuel cell equipment; The formula represents the price of hydrogen purchase; formulas (54) to (66) and (70) represent the constraints introduced by auxiliary variables; formulas (67) to (69) represent the region. Energy self-balancing constraint under no-trade conditions and They are respectively regions Maximum input power of electrolyzer and maximum output power of fuel cell For the region Maximum input power of ground source heat pump and They are respectively regions The maximum input heat power and maximum output heat power of the hot water tank. and They are respectively regions The maximum input cooling power and maximum output cooling power of the cold water tank. For the region The maximum value of thermal energy injected into a geothermal well. for Time slot region The power required for cooling energy in an adsorption chiller. for Time slot region Electric boilers output heat energy. for Time slot region Ground source heat pumps output heat energy. for Time slot region The heat generated by the fuel cell.

4. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 3, characterized in that, The optimization sub-problem for the coordinated operation of multi-regional hydrogen-containing integrated energy systems, including electricity-heat-cooling transactions, established in step 2 is as follows: ， ， in: express The comprehensive economic cost of a regional integrated energy system under the worst-case scenario Represents a scheduling cycle inner area Total operating costs Represents a scheduling cycle inner area Revenue from energy transactions with other regions Represents a scheduling cycle inner area Grid access fees incurred from energy interactions with other regions.

5. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 4, characterized in that, The compact form of the main problem of the two-stage robust optimization model proposed in step (5.1) is as follows: ， ， ， ， ， ， ， ， ， ， ， ， ， in: , Representing regions The outer and inner sets of optimization variables in a two-stage robust optimization framework. For uncertain variables, and In formula (48) The corresponding coefficient column vector sum The corresponding coefficient column vector; , , , , , , , , , , , These are the coefficient matrices of the variables under the corresponding constraints; , , , , , , , , , , , These are constant column vectors; formula (75) represents the expression regarding... Inequality constraints; Formula (76) expresses the inequality constraints regarding Equality constraints; Formula (77) represents the expression regarding Inequality constraints; Formula (78) expresses the inequality constraints regarding Equality constraints; Formula (79) represents the expression regarding and Equality constraints; Formula (80) represents the expression regarding and Inequality constraints; Formula (81) expresses the inequality constraints regarding and Equality constraints; Formula (82) represents the expression regarding and Inequality constraints; Formula (83) is The upper and lower limits are constrained; formula (84) is Upper and lower limit constraints; This represents the maximum number of iterations. Auxiliary variables introduced; For the first The region obtained in the second iteration Optimal strategy for subproblems; For the first The region obtained by solving the subproblem in the next iteration The worst-case scenario for renewable energy output and load power; The compact form of the subproblem of the proposed two-stage robust optimization model is as follows: ， ， in: Indicates the region A set of uncertain variables Indicates a given set Time zone Optimize variables feasible domain, , , , , , , , Let each of these represent the dual variables corresponding to each constraint in the minimization problem of the second stage; The dual problem of the subproblem proposed in step (5.2) is as follows: ， 。 6. The method for optimizing the coordinated operation of a multi-region hydrogen-containing integrated energy system under uncertain conditions according to claim 5, characterized in that, The steps in step (5.3) for iteratively solving the two-stage robust optimization model based on the column and constraint generation algorithm are as follows: (7.1) Initialize, set the region The power output and load prediction curves of new energy sources are set as the initial scenario. The upper bound of the initial optimization problem is The lower bound is Number of iterations ; (7.2) Substitute the output and load power of the new energy units under the initialization scenario into the main problem and solve it to obtain the optimal solution of the main problem. At the same time, update the lower bound. ; (7.3) Fixed By substituting the subproblem into its dual problem and solving it, we can obtain the optimal solution to the subproblem. and harsh scenes and update the upper bound. ; (7.4) The convergence threshold of the given algorithm is ,like Then stop iterating and return the optimal solution. and Otherwise, add variables. and the following constraints: 。 7. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 6, characterized in that, The steps in step (8.2) for distributed solution of the multi-region hydrogen-containing integrated energy system collaborative operation optimization sub-problem based on the alternating direction multiplier method with enhanced adaptive prediction-correction mechanism are as follows: Introducing auxiliary variables for decoupling: ， in: Types of energy trading; For the region Expectations and Regions Energy trading volume; For the region Expectations and Regions Energy trading volume; By region For example, the augmented Lagrangian function of the objective function in cooperative runtime. as follows: ， in: and These are Lagrange multipliers and penalty factors, respectively. Indicates the number of collaborative operation regions; T represents the scheduling period; The specific iterative steps of the algorithm are as follows: (9.1) Each region is accessed via the Lagrange function. Calculate the expected energy trading volume as the forecast value to obtain the predicted energy trading volume. , and Lagrange multipliers ; (9.2) Sharing the first The next iteration predicts energy trading volume. , and Lagrange multipliers The expected energy trading volume is obtained based on the correction steps. , and Lagrange multipliers as follows: in: This is a correction factor, whose function is to correct the relationship between the predicted value and the previous iteration value; (9.3) Calculate the original residuals according to formulas (96) to (97). and dual residual ; ， ， (9.4) Determine the convergence status of the algorithm; in: and It is the iterative convergence accuracy value; if it satisfies formula (98), then the algorithm converges, and the optimal energy trading volume in the output is... Costs of each region during collaborative runtime ; , as well as These represent the original residuals of the trading volumes of electrical energy, thermal energy, and cold energy, respectively. , as well as These represent the dual residuals of the trading volumes of electrical energy, thermal energy, and cold energy, respectively. (9.5) If formula (98) is not satisfied, return to step (9.1) to continue iterating and adaptively update the penalty factor. as follows: ， in: This is an empirical coefficient.

8. The method for optimizing the coordinated operation of a multi-regional hydrogen-containing integrated energy system under uncertain conditions according to claim 7, characterized in that, The steps for benefit allocation based on the nonlinear energy mapping method in step 5 are as follows: (10.1) Obtain the independent operating costs of each region in step (5.3). The optimal energy trading volume in step (8.3) Optimal Energy Trading Volume II Optimal energy trading volume three Operating costs of each region during collaborative runtime First, calculate the energy that each region contributes to the whole when participating in the cooperation. and the energy obtained from the total , Indicates the number of collaborative operation regions; Indicates the scheduling period. Indicates time slot, and Region: ， ， (10.2) Then, the contribution of each region is quantified according to the degree of participation in energy exchange, and the natural constant is selected. Quantifying the contribution of an exponential function mathematical model with base as the base : ， in, This represents the maximum energy supply in each region; This represents the maximum energy gain in each region; (10.3) Finally, output the final cost for each region. As shown below: 。

Images