A Method and System for Optimizing the Operation of Hydrogen-Electric Coordinated Distribution Network Based on Carbon-Electric Coupling Constraints

By constructing a two-layer optimization model and inter-layer iterative solution, the problems of insufficient consideration of hydrogen-electric coupling characteristics and inadequate integration of carbon-electric constraints were solved, thus realizing efficient new energy consumption and low-carbon operation of the distribution network.

CN123073933APending Publication Date: 2026-11-10STATE GRID ZHEJIANG ELECTRIC POWER CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611284926.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-24
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing power distribution network optimization methods fail to fully consider the temporal coupling characteristics of hydrogen power equipment and the integration of carbon and electricity constraints when faced with a high proportion of new energy and hydrogen energy equipment access, resulting in low new energy consumption rate, excessive carbon emissions and low dispatch efficiency.

Method used

A two-layer optimization model based on carbon-electric coupling constraints is constructed, consisting of an upper distribution network coordination layer and a lower hydrogen-electric microgrid autonomous layer. By using data-driven prediction of source and load data and initializing parameters, the optimization model is solved through inter-layer iteration to achieve multi-objective optimization of carbon-electricity and minimization of microgrid operating costs.

Benefits of technology

It improves the engineering practicality of new energy consumption rate, reduces carbon emissions and scheduling calculation, realizes collaborative decision-making of global optimization and local autonomy, and solves the problems of difficulty in balancing multiple objectives and low scheduling efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123073933A_ABST
    Figure CN123073933A_ABST
Patent Text Reader

Abstract

This invention relates to the field of power system control technology, and particularly to a method and system for optimizing the operation of a hydrogen-electricity coordinated distribution network based on carbon-electricity coupling constraints. The method includes: initializing hydrogen-electricity microgrid parameters, distribution network parameters, and multi-objective weight parameters. A two-layer optimization model is constructed, comprising an upper distribution network coordination layer and a lower hydrogen-electricity microgrid autonomous layer. The upper distribution network coordination layer determines the total interactive power of the microgrid group based on the optimal carbon-electricity multi-objective, while the lower hydrogen-electricity microgrid autonomous layer optimizes the electrolyzer power, fuel cell power, hydrogen storage capacity, and single microgrid interactive power by minimizing microgrid operating costs. The two-layer optimization model is solved iteratively between layers, and the optimized operating parameters of each device and the interactive power for each time period are output based on the final iteration results. This approach solves the technical problem of insufficient multi-objective coordinated scheduling efficiency of the distribution network due to inadequate consideration of hydrogen-electricity coupling characteristics and insufficient integration of carbon-electricity constraints in existing technologies.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power system control technology, and in particular to a method and system for optimizing the operation of a hydrogen-electric coordinated distribution network based on carbon-electric coupling constraints. Background Technology

[0002] The penetration rate of new energy sources, represented by distributed photovoltaic and wind power, continues to climb, with renewable energy accounting for over 50% in some industrial parks. On the other hand, hydrogen energy, with its clean and efficient energy storage characteristics, is seeing related equipment such as electrolyzers, fuel cells, and hydrogen storage tanks gradually integrated into the power distribution network, forming a closed-loop energy system of "hydrogen production-hydrogen storage-hydrogen power generation," also known as "hydrogen-electricity co-generation microgrids," which has become an important component of the new power distribution network. However, the high proportion of new energy and hydrogen energy equipment integrated into the network makes the operation characteristics of the power distribution network extremely complex, posing a serious challenge to existing optimization methods. In the field of power distribution network optimization and scheduling, existing technologies have established a certain foundation. For example, existing technology (authorization announcement number CN118412851B) discloses a hierarchical distributed scheduling method for multi-microgrid power distribution systems under carbon flow tracing. This method constructs a two-layer scheduling model including carbon trading constraints and uses the objective cascade analysis method for iterative solution, aiming to achieve coordinated low-carbon economic operation of the power distribution network and multi-microgrid systems. This method introduces the idea of ​​hierarchical distributed optimization, which to some extent solves the problem of multi-entity coordinated scheduling. However, this method primarily focuses on modeling and scheduling equipment such as gas turbines, carbon capture, and power-to-gas converters. Its underlying microgrid model does not specifically model the unique temporal coupling characteristics of devices like electrolyzers and fuel cells in hydrogen-electricity synergy scenarios. For example, hydrogen production from electrolyzers needs to closely follow fluctuations in photovoltaic output, while fuel cell power generation needs to be linked to the hydrogen release process from the storage tank. This strong temporal coupling relationship between electricity and hydrogen energy is not fully considered in this existing technology, making it difficult to achieve energy self-balancing within the microgrid when dealing with hydrogen-electricity synergy microgrids with a high proportion of new energy sources, thus limiting the effective absorption of new energy.

[0003] This demonstrates that existing power distribution network optimization methods generally suffer from insufficient adaptability to the new scenario of multi-level interaction between "source-load-storage-hydrogen". Traditional models fail to deeply integrate the dynamic coupling characteristics of hydrogen power equipment with low-carbon constraints such as carbon quotas and carbon trading costs, resulting in optimization objectives often limited to a single economic goal, which is difficult to meet the requirements of regional low-carbon operation. Furthermore, when multiple hydrogen-power microgrids are connected, the lack of an effective hierarchical decision-making mechanism of "microgrid autonomy - distribution network coordination" easily leads to power conflicts between microgrids or the failure of global objectives, making it difficult to achieve optimal coordination among multiple objectives such as renewable energy consumption, carbon reduction efficiency, power supply reliability, and investment returns. In addition, traditional global centralized optimization models have high computational complexity, making it difficult to meet the engineering requirements of real-time dispatching of power distribution networks. Therefore, existing technologies, due to insufficient consideration of the coupling characteristics of hydrogen and electricity, result in low renewable energy consumption rates, excessive carbon emissions, and low dispatching efficiency. Summary of the Invention

[0004] To address the aforementioned shortcomings or deficiencies, this invention provides a method and system for optimizing the operation of a hydrogen-electric coordinated distribution network based on carbon-electric coupling constraints. This method can solve the technical problem that existing technologies suffer from insufficient consideration of hydrogen-electric coupling characteristics and inadequate integration of carbon-electric constraints, leading to insufficient efficiency in multi-objective coordinated scheduling of the distribution network.

[0005] This invention provides a method for optimizing the operation of a hydrogen-electricity coordinated distribution network based on carbon-electricity coupling constraints. The method includes: acquiring scenario variables and predicting source-load data using a data-driven approach; initializing hydrogen-electricity microgrid parameters, distribution network parameters, and multi-objective weight parameters; constructing a two-layer optimization model comprising an upper-layer distribution network coordination layer and a lower-layer hydrogen-electricity microgrid autonomous layer; the upper-layer distribution network coordination layer determining the total interactive power of the microgrid group based on carbon-electricity multi-objective optimization; and the lower-layer hydrogen-electricity microgrid autonomous layer optimizing the electrolyzer power, fuel cell power, hydrogen storage capacity, and single-microgrid interactive power by minimizing microgrid operating costs. The two-layer optimization model is solved iteratively between layers. Responding to the condition that the inter-layer objective deviation is less than a preset threshold and the number of iterations does not exceed a preset upper limit, the optimized schemes for the operating parameters of each device and the interactive power for each time period are output based on the final iteration results.

[0006] According to a second aspect, this invention provides a hydrogen-electricity coordinated distribution network operation optimization system based on carbon-electric coupling constraints, which relies on a distribution network energy management platform. The system includes a processing unit and a storage unit. The storage unit stores executable instructions, and the processing unit integrates a data-driven prediction module, a two-layer optimization solution module, and an inter-layer iterative coordination module. When the executable instructions are invoked by the processing unit, they drive the system to execute any of the hydrogen-electricity coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in this invention's embodiments.

[0007] According to a third aspect, the present invention provides an electronic device comprising: at least one processing unit; and a storage unit communicatively connected to the at least one processing unit. The storage unit stores executable instructions that can be invoked by the at least one processing unit. When the executable instructions are executed by the at least one processing unit, the at least one processing unit causes it to execute any of the hydrogen-electric coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in the embodiments of the present invention.

[0008] According to another aspect of the present invention, a medium for storing executable instructions is provided, wherein the computer instructions are used to cause a computer to execute any of the hydrogen-electric coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in the embodiments of the present invention.

[0009] The present invention provides a method for optimizing the operation of a hydrogen-electricity coordinated distribution network based on carbon-electricity coupling constraints. This method is achieved through three core steps: source-load prediction and parameter initialization, construction of a two-layer optimization model, and inter-layer iterative solution. Specifically, it acquires scenario variables and predicts source-load data based on a data-driven method, initializes hydrogen-electricity microgrid parameters, distribution network parameters, and multi-objective weight parameters, providing an accurate input foundation and flexible weight configuration capabilities for subsequent optimization. A two-layer optimization model is constructed, comprising an upper-layer distribution network coordination layer and a lower-layer hydrogen-electricity microgrid autonomous layer. The upper layer determines the total interactive power of the microgrid group based on carbon-electricity multi-objective optimization, while the lower layer optimizes the electrolyzer power, fuel cell power, hydrogen storage capacity, and single microgrid interactive power by minimizing microgrid operating costs. This achieves deep integration of carbon constraints and electric power scheduling, as well as time-series coupling optimization of hydrogen-electricity equipment. Through inter-layer iterative solution of the two-layer optimization model, in response to inter-layer objective deviations being less than a preset threshold and the number of iterations not exceeding a preset upper limit, the optimization schemes for the operating parameters of each device and the interactive power of each time period are output based on the final iteration results, establishing a collaborative decision-making mechanism that considers both global optimization and local autonomy.

[0010] In this technical solution, the present invention addresses the problem of insufficient multi-objective coordinated scheduling efficiency of power distribution networks caused by inadequate consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints, as described in the background technology. It introduces carbon-electric multi-objective optimization and constructs carbon emission constraints at the upper layer, and introduces time-series coupling optimization of electrolyzers, fuel cells, and hydrogen storage tanks at the lower layer. This deeply couples low-carbon constraints such as carbon quotas and carbon trading costs with power scheduling, solving the fundamental defect of existing technologies that cannot simultaneously adapt to hydrogen-electric coordinated scenarios and the integration of carbon-electric constraints. To address the difficulties in balancing multiple objectives and low scheduling efficiency caused by this defect, a two-layer architecture of "upper-layer global coordination - lower-layer microgrid autonomy" and an inter-layer iterative mechanism are established. First, the upper layer determines the total interactive power of the microgrid group, and then the lower layer optimizes the output of internal equipment in each microgrid. Through iterative coordination, global objectives and local interests tend to align, thereby achieving the unity of optimal multi-objective coordinated carbon-electricity and self-balancing hydrogen-electricity energy in the decision-making logic. Therefore, the technical solution of the present invention solves the technical problem that the existing technology has insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints, resulting in insufficient efficiency of multi-objective coordinated scheduling of distribution networks, and improves the new energy absorption rate, carbon reduction efficiency and engineering practicality of scheduling calculation. Attached Figure Description

[0011] Figure 1 This is a flowchart of a hydrogen-electric coordinated distribution network operation optimization method based on carbon-electric coupling constraints according to an embodiment of the present invention; Figures 2a-2c A time-series diagram illustrating the specific power output of each microgrid device and its energy interaction with the distribution network during a typical test day in another embodiment of the present invention. Figure 3 A schematic diagram of the 24-hour optimized interactive power timing output by the upper distribution network coordination layer in another embodiment of the present invention; Figure 4 This is a block diagram of an electronic device used to implement embodiments of the present invention. Detailed Implementation

[0012] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0013] During the development of this invention, researchers, through extensive experiments and data analysis, discovered a strong correlation between the temporal coupling characteristics of hydrogen-electric equipment and carbon-electric constraints: the hydrogen production power of the electrolyzer needs to closely follow the fluctuations in photovoltaic output, the power generation power of the fuel cell needs to be linked with the hydrogen release process of the hydrogen storage tank, and carbon emissions are positively correlated with the power purchased by the microgrid from the distribution network. These three factors jointly determine the operating efficiency and low-carbon level of the distribution network. Based on this relationship, this invention innovatively proposes this technical solution, utilizing a data-driven method to predict source-load data. It constructs a two-layer optimization model consisting of an upper distribution network collaboration layer and a lower hydrogen-electric microgrid autonomous layer, combining an inter-layer iterative solution mechanism, and employing collaborative optimization that minimizes carbon-electric multi-objective optimization and microgrid operating costs, embodying the core concept of "layered autonomy + global collaboration."

[0014] Specifically, the invention team discovered through experiments that traditional power distribution network optimization methods have technical flaws: they fail to consider the temporal coupling characteristics of hydrogen-powered devices and the integration of carbon and electricity constraints is insufficient, resulting in low renewable energy absorption rates, excessive carbon emissions, and long scheduling calculation times. However, the hydrogen-electricity coordinated power distribution network operation optimization method based on carbon-electricity coupling constraints proposed in this invention can incorporate carbon costs, reliability costs, and benefits into a unified objective function, thereby improving the efficiency of multi-objective coordinated optimization. By determining the total interactive power of the microgrid group at the upper level and optimizing the output of each microgrid device at the lower level, it is possible to achieve a balance between hydrogen-electric energy self-balancing and global carbon emission reduction, and the scheduling calculation time can meet real-time engineering requirements.

[0015] Therefore, this invention provides a method for optimizing the operation of a hydrogen-electric coordinated distribution network based on carbon-electric coupling constraints, according to the first aspect. This method can be applied to a system for optimizing the operation of a hydrogen-electric coordinated distribution network based on carbon-electric coupling constraints (hereinafter referred to as the "system"). This system can be deployed on a distribution network energy management platform to complete source-load prediction, two-layer optimization solutions, and inter-layer iterative coordination, thereby achieving safe, economical, and low-carbon operation of the distribution network. Specifically, this system can be deployed in various hardware environments, including but not limited to: industrial servers, cloud computing nodes, and edge computing gateways. Figure 1 As shown, the method may include: Step S110: Obtain scenario variables and predict source load data based on data-driven methods, and initialize hydrogen power microgrid parameters, distribution network parameters and multi-objective weight parameters.

[0016] Scenario variables refer to the set of external factors affecting photovoltaic output and electrical load, including meteorological data (such as sunlight intensity, temperature, and wind speed), date type (weekday or holiday), and equipment status (such as electrolyzer start / stop indicators and hydrogen storage tank pressure levels). Source-load data refers to the time series of predicted photovoltaic output and predicted electrical load, used to characterize the microgrid's power generation capacity and electricity demand over a future period. Hydrogen-electricity microgrid parameters refer to the technical specifications related to the operation of hydrogen energy equipment, including the rated power of the electrolyzer, the rated power of the fuel cell, and the upper limit of the hydrogen storage tank capacity. Distribution network parameters refer to network-level constraint data such as line transmission capacity, carbon emission quotas, and carbon emission factors for electricity purchases. Multi-objective weighting parameters refer to weighting coefficients used to balance the importance of carbon costs, reliability costs, and benefits.

[0017] Specifically, the system relies on a power distribution network energy management platform. Its data acquisition module can obtain hourly sunshine intensity and wind speed data for the next 24 hours from a weather forecast service interface, read the current start-up and shutdown status of the electrolyzer and fuel cell from the microgrid monitoring system, and extract load curves of the same date type from a historical database. The system inputs these scenario variables into a pre-trained Long Short-Term Memory (LSTM) network model and outputs hourly photovoltaic power output and electricity load prediction values ​​for the next 24 hours. Simultaneously, the system reads the hydrogen-powered microgrid parameters (e.g., maximum electrolyzer power 500 kW, maximum fuel cell power 300 kW, hydrogen storage tank capacity 2000 cubic meters), power distribution network parameters (e.g., line transmission capacity 10 MW, carbon emission quota of 5 tons of carbon dioxide per day), and multi-objective weight parameters (e.g., carbon cost weight 0.4, reliability cost weight 0.3, revenue weight 0.3) from the configuration file to complete initialization.

[0018] For example, in a practical application in an industrial park, the system acquires weather forecast data for July 15, 2025 (peak solar intensity 800 watts per square meter, average wind speed 4 meters per second), with the date type being a weekday. The equipment status shows the electrolyzer is available and the fuel cell is on standby. The LSTM model predicts that the peak photovoltaic output for that day will be 1.2 MW and the peak electrical load will be 2.5 MW. The initial parameters for the hydrogen-powered microgrid are: electrolyzer power range 0 to 500 kW, fuel cell power range 0 to 300 kW, and initial hydrogen storage capacity of 500 cubic meters; the distribution network parameters are: line transmission capacity 8 MW, carbon emission quota 4 tons of carbon dioxide; and the multi-objective weighting parameters are: carbon cost 0.4, reliability cost 0.3, and revenue 0.3.

[0019] Step S120: Construct a two-layer optimization model that includes an upper distribution network coordination layer and a lower hydrogen-electric microgrid autonomous layer. The upper distribution network coordination layer determines the total interactive power of the microgrid group based on carbon-electricity multi-objective optimization, while the lower hydrogen-electric microgrid autonomous layer optimizes the electrolyzer power, fuel cell power, hydrogen storage capacity, and single microgrid interactive power by minimizing the microgrid operating cost.

[0020] The upper-level distribution network coordination layer refers to the optimization level that makes decisions from a global perspective of the distribution network. Its optimization objective is multi-objective optimality of carbon electricity, that is, simultaneously minimizing carbon costs, reliability costs, and maximizing distribution network revenue. The decision variable is the total interactive power of the microgrid group (i.e., the algebraic sum of the power exchanged between all microgrids and the distribution network). The lower-level hydrogen-electricity microgrid autonomous layer refers to the optimization level that makes decisions from a local perspective of a single microgrid. Its optimization objective is to minimize the operating cost of the microgrid. The decision variables include electrolyzer power (electric power consumed in hydrogen production), fuel cell power (electric power output from power generation), hydrogen storage capacity (hydrogen volume in the hydrogen storage tank), and the interactive power between a single microgrid and the distribution network (power purchased or sold). The two-level optimization model refers to a mathematical programming model in which the upper and lower levels are interconnected through coupling variables (interactive power) and solved alternately.

[0021] Specifically, the system can solve the mixed-integer linear programming model by calling commercial solvers (such as Gurobi and CPLEX) or open-source solvers (such as SCIP). At the upper layer, the system establishes an optimization problem with the goal of minimizing the weighted sum of carbon and electricity multi-objectives. Constraints include line transmission power constraints (total interactive power of the microgrid group does not exceed line transmission capacity) and carbon emission constraints (incorporating excess carbon emission costs into the objective function). At the lower layer, the system establishes an optimization problem with the goal of minimizing the microgrid operating cost (electricity purchase cost minus electricity sales revenue plus equipment operation and maintenance costs). Constraints include equipment output constraints (electrolyzer power between 0 and rated maximum, fuel cell power between 0 and rated maximum), hydrogen storage constraints (hydrogen storage between minimum safe storage and maximum capacity), and power balance constraints (PV output plus fuel cell output plus purchased power equals electrical load plus electrolyzer power plus electricity sales power).

[0022] For example, assume the distribution network connects to three hydrogen-powered microgrids. The upper-level model sets the carbon cost at 100 yuan per ton of carbon dioxide, the reliability cost coefficient at 0.5 yuan per kilowatt-hour of power shortage, and the revenue as the electricity sales price multiplied by the electricity sales volume. In the lower-level model, microgrid 1 has a photovoltaic capacity of 2 MW, an electrolyzer rated power of 500 kW, a fuel cell rated power of 300 kW, and a hydrogen storage tank capacity of 2000 cubic meters. The system calculates the optimal total interactive power of the upper-level microgrid (electricity purchase direction) using the MILP solver as 2.5 MW. The optimized result for lower-level microgrid 1 is: electrolyzer power 400 kW, fuel cell power 200 kW, hydrogen storage capacity increased from 500 cubic meters to 600 cubic meters, and single-microgrid interactive power of 0.8 MW (electricity purchase).

[0023] Step S130: Solve the two-layer optimization model through inter-layer iteration. In response to the inter-layer target deviation being less than a preset threshold and the number of iterations not exceeding a preset upper limit, output the optimization scheme of the operating parameters of each device and the interaction power of each time period based on the final iteration results.

[0024] The inter-layer objective deviation refers to the relative change in the sum of the upper-layer and lower-layer objective function values ​​between two adjacent iterations, used to measure whether the solution tends to stabilize. A preset threshold is a small positive number (e.g., 0.01); convergence is considered achieved when the inter-layer objective deviation is less than this value. A preset upper limit is the maximum allowed number of iterations (e.g., 50) to prevent infinite loops. The final iteration result refers to the set of optimal solutions for both the upper and lower layers after the last iteration. The operating parameters of each device refer to the power or storage capacity settings of the electrolyzer, fuel cell, and hydrogen storage tank for each time period. The optimization scheme for the interactive power in each time period refers to the power purchase or sales curves of the microgrid and distribution network for each time period.

[0025] Specifically, the system operates through the following iterative process: First, the upper layer solves for the total interactive power of the microgrid group and distributes it to each microgrid according to a preset decomposition rule (e.g., based on the proportion of photovoltaic installed capacity of each microgrid), obtaining the adjusted interactive power for each microgrid. Then, each microgrid uses this adjusted interactive power as an additional constraint to resolve the lower-level model, updating its interactive power requirements and feeding them back to the upper layer. The upper layer updates the total interactive power based on the feedback and resolves, repeating this process. After each iteration, the inter-layer target deviation is calculated. If the deviation is less than a preset threshold and the number of iterations has not exceeded the upper limit, the iteration stops and the current result is output; otherwise, the penalty function multiplier is updated and the next iteration begins.

[0026] For example, in the above three microgrid cases, the system sets a preset threshold of 0.01 and a preset upper limit of 30 iterations. After the first iteration, the inter-layer target deviation is 0.25, and after the fifth iteration, the deviation drops to 0.008, satisfying the convergence condition. The system outputs the final results: the electrolyzer power of microgrid 1 is 450 kW from 9:00 to 16:00 (utilizing peak photovoltaic power), the fuel cell power is 250 kW from 18:00 to 22:00 (peak evening load), and the hydrogen storage capacity increases to 700 cubic meters during the day and decreases to 400 cubic meters at night; the interaction power curves between each microgrid and the distribution network show that microgrid 1 sells 0.3 MW of electricity at noon and purchases 0.6 MW of electricity at night. This optimization scheme can be directly sent to the microgrid controller for execution.

[0027] This embodiment provides a method for optimizing the operation of a hydrogen-electricity coordinated distribution network based on carbon-electricity coupling constraints. This method is achieved through three core steps: source-load prediction and parameter initialization, construction of a two-layer optimization model, and inter-layer iterative solution. Specifically, it acquires scenario variables and predicts source-load data based on a data-driven method, initializing hydrogen-electricity microgrid parameters, distribution network parameters, and multi-objective weight parameters, providing an accurate input foundation and flexible weight configuration capabilities for subsequent optimization. A two-layer optimization model is constructed, comprising an upper-layer distribution network coordination layer and a lower-layer hydrogen-electricity microgrid autonomous layer. The upper layer determines the total interactive power of the microgrid group based on carbon-electricity multi-objective optimization, while the lower layer optimizes the electrolyzer power, fuel cell power, hydrogen storage capacity, and single microgrid interactive power by minimizing microgrid operating costs. This achieves deep integration of carbon constraints and electric power scheduling, as well as time-series coupling optimization of hydrogen-electricity equipment. Through inter-layer iterative solution of the two-layer optimization model, in response to inter-layer objective deviations being less than a preset threshold and the number of iterations not exceeding a preset upper limit, the optimization schemes for the operating parameters of each device and the interactive power of each time period are output based on the final iteration results, establishing a collaborative decision-making mechanism that considers both global optimization and local autonomy.

[0028] In this embodiment, the problem of insufficient multi-objective coordinated scheduling efficiency of power distribution networks caused by inadequate consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints, as described in the background technology, is addressed by introducing carbon-electric multi-objective optimization and constructing carbon emission constraints at the upper layer, and introducing time-series coupling optimization of electrolyzers, fuel cells, and hydrogen storage tanks at the lower layer. This deeply couples low-carbon constraints such as carbon quotas and carbon trading costs with power scheduling, solving the fundamental defect of existing technologies that cannot simultaneously adapt to hydrogen-electric coordinated scenarios and the integration of carbon-electric constraints. To address the multi-objective balancing difficulties and low scheduling efficiency caused by this defect, a two-layer architecture of "upper-layer global coordination - lower-layer microgrid autonomy" and an inter-layer iterative mechanism are established. First, the upper layer determines the total interactive power of the microgrid group, and then the lower layer optimizes the output of equipment within each microgrid. Through iterative coordination, the global objectives and local interests tend to be consistent, thereby achieving the unity of optimal carbon-electric multi-objective coordination and hydrogen-electric energy self-balancing in the decision-making logic. Therefore, the technical solution of this embodiment solves the technical problem that the existing technology has insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints, resulting in insufficient efficiency of multi-objective coordinated scheduling of distribution networks, and improves the new energy absorption rate, carbon reduction efficiency and engineering practicality of scheduling calculation.

[0029] To verify the effectiveness of this method, a typical summer day in an industrial park was selected as the test scenario. Figures 2a-2c The solution results of the lower-level optimization model in this scenario are shown. Specifically, Figures 2a-2c These three charts are time-series diagrams showing the specific power output of each microgrid device and its energy interaction with the distribution network on a typical test day, after the optimization solution of the lower-level hydrogen-electric microgrid autonomous layer. They visually demonstrate the actual operational effectiveness of the hydrogen-electric coupling system in mitigating new energy fluctuations and achieving energy time-shifting.

[0030] Specifically, Figure 2a The graph shows the power output of the lower-level microgrid equipment. The horizontal axis represents 24 hours in a day, and the vertical axis represents the active power of the equipment, measured in kilowatts (kW). The graph illustrates the dynamic changes in electrolyzer power and fuel cell power. It can be observed that the electrolyzer power exhibits a highly correlated fluctuation with photovoltaic (PV) output. During peak PV output periods (e.g., 6:00 AM to 7:00 PM), it operates continuously, adjusting its power according to PV output fluctuations to convert surplus electrical energy into chemical energy for storage. Fuel cell power, on the other hand, is mainly distributed during nighttime peak load periods and parts of the daytime, used to supply power to the microgrid when renewable energy output is insufficient. The two exhibit good temporal complementarity. For example, during the low load period from 0:00 AM to 5:00 AM, both types of equipment operate at low power or are shut down; while from 9:00 AM to 11:00 AM, when PV output reaches its peak, the electrolyzer power subsequently climbs to a high level of approximately 150 to 200 kW for efficient hydrogen production.

[0031] Figure 2bThis diagram illustrates the power interaction between the microgrid and the distribution network. The horizontal axis represents time, and the vertical axis represents the interaction power, measured in kW. Positive values ​​represent the microgrid purchasing electricity from the distribution network, while negative values ​​represent the microgrid selling electricity to the distribution network. This diagram illustrates how the microgrid adjusts its power exchange with the main grid to coordinate with the regulation of its internal hydrogen energy storage system. For example, during periods such as 0:00 to 5:00 AM and 8:00 to 11:00 PM, the microgrid is in a purchasing state, absorbing electricity from the main grid to meet its base load demand. Conversely, during peak solar power generation periods from 11:00 AM to 2:00 PM, the microgrid feeds excess electricity back to the distribution network, achieving local consumption of renewable energy.

[0032] Figure 2c The graph shows the state of the hydrogen storage tank, with time on the horizontal axis and the amount of hydrogen stored in the tank on the vertical axis, in kilograms (kg). This graph clearly illustrates the throughput process of the hydrogen storage system as an energy buffer. Data shows that the hydrogen storage capacity accumulates steadily in a stepped or sloping manner during the day when the electrolyzer is operating at high power, while it decreases in a stepped manner in the evening and at night when the fuel cell starts operating. For example, the hydrogen storage capacity is maintained at a base level of approximately 100 kg in the early morning, and as hydrogen production continues during the day, it accumulates to a peak of nearly 200 kg by the evening, then is gradually consumed during nighttime power generation, until it rises again the following morning. The data flow from these three graphs corroborates each other, forming a direct proof of the solution results of the lower-level optimization model. By precisely controlling the start-up, shutdown, and ramp-up rates of the electrolyzer and fuel cell, the system not only achieves power balance within the microgrid but also... Figure 2b The interactive power adjustment effectively reduced the power flow pressure on the distribution network, verifying the technical advantages of this scheme in improving the efficiency of multi-objective coordinated scheduling of the distribution network.

[0033] In another embodiment, Figure 3 The diagram shows the 24-hour optimized interactive power time series output by the upper-level distribution network coordination layer after the two-layer collaborative optimization model has completed its iterative solution. It details the dynamic net power exchange between the distribution network and the upper-level large power grid (or adjacent grid areas) under the constraints of carbon-electricity multi-objective optimization, intuitively reflecting the operational characteristics of this scheme in smoothing regional load fluctuations and achieving peak shaving and valley filling at the global level.

[0034] Specifically, Figure 3The horizontal axis represents 24 hours in a day, and the vertical axis represents the power exchange, measured in kW. The values ​​shown in the chart are all extremely negative (approximately between -199.9999972 kW and -199.9999888 kW). In the conventional definition of power dispatching systems, a negative sign indicates that the distribution network is supplying electricity to the external grid (i.e., net grid connection). Therefore, this chart reveals that during the optimization period, the new energy and hydrogen-powered microgrid systems within this distribution network area are in a surplus state, continuously feeding green electricity to the main grid. The slight fluctuations in the curve represent the fine-tuning of power exchange by the optimization algorithm at different times: for example, during the low load period from 0:00 to 5:00 AM, the power exchange slightly increases (the value decreases slightly, i.e., the grid connection power decreases slightly), which is usually to reduce the external power supply to meet the basic electricity demand of the local microgrid; while at noon and in the afternoon, as the photovoltaic output increases, the power exchange reaches its lowest point (the value reaches its minimum, i.e., the grid connection power reaches its maximum), reflecting the system's maximization of the absorption and utilization of new energy. For example, at time period 0 (0:00), the optimized interactive power is approximately -199.999998 kW, indicating that the distribution network is transmitting power to the external grid at a rate close to 200 kW. As time progresses to time period 4 (4:00 AM), the power curve reaches a peak of approximately -199.9999974 kW, meaning that the power transmitted to the external grid has slightly decreased. The power then continues to decline, reaching its trough around time period 18 (6:00 PM) at approximately -199.999986 kW, indicating that the distribution network's net power transmission reaches its maximum at this time. This is closely related to the combined effect of the large-scale generation of renewable energy during the day and the discharge of energy storage within the microgrid supporting local loads. This upper-level optimized interactive power curve is the final execution command for the energy management of the entire distribution network. The system sends the smooth power curve that meets the line transmission capacity constraints to the main transformer or active power control system of the distribution network. This can effectively avoid the drastic power fluctuations that may occur in traditional dispatching, ensure the frequency stability of the external power grid, and improve the economy and low carbon emissions of the distribution network through precise power planning.

[0035] In some embodiments, the steps for constructing a two-layer optimization model include: A collaborative upper-level distribution network layer is constructed, with the goal of multi-objective optimization of carbon electricity and the constraints of line transmission power and carbon emission.

[0036] Among them, carbon-electricity multi-objective optimization refers to the weighted minimization of three components: carbon cost, reliability cost, and distribution network revenue. Carbon cost is calculated by multiplying carbon emissions by the carbon price; reliability cost is calculated by multiplying the power shortage amount by the unit power shortage loss coefficient; and distribution network revenue is obtained by subtracting electricity purchase expenditure from electricity sales revenue. Line transmission power constraint means that the absolute value of the total interactive power of the microgrid group must not exceed the upper limit of the line transmission capacity to prevent line overload. Carbon emission constraint means incorporating the excess cost of carbon emissions (emissions exceeding the carbon quota multiplied by the excess penalty unit price) into the objective function to achieve a soft constraint on carbon emissions.

[0037] Specifically, the system can define the objective function in the upper-level optimization model as a weighted sum of carbon cost, reliability cost, and negative revenue (i.e., cost form), with the weights determined by multi-objective weight parameters. Constraints include: the total interactive power of the microgrid group is less than or equal to the upper limit of the line transmission capacity and greater than or equal to the lower limit (usually a negative value, indicating reverse power transmission); carbon emissions are obtained by multiplying the purchased power of each microgrid by the carbon emission factor of the purchased power, and the excess portion is multiplied by the penalty price and added to the objective function. For example, a distribution network has a line transmission capacity upper limit of 10 MW, a carbon quota of 5 tons of CO2 per day, a carbon price of 100 yuan per ton, and an excess penalty of 150 yuan per ton. The system sets the carbon cost weight to 0.4, the reliability cost weight to 0.3, and the revenue weight to 0.3. The upper-level model solves for a total interactive power of 2.5 MW (in the direction of power purchase) and carbon emissions of 3.2 tons, which does not exceed the quota, therefore there is no excess cost.

[0038] A lower-level autonomous layer for hydrogen-electric microgrids is constructed, with the goal of optimizing microgrid operating costs, and constrained by equipment output constraints, hydrogen storage capacity constraints, and power balance constraints.

[0039] Microgrid operating cost optimization refers to minimizing the total operating cost of a single microgrid, including the cost of purchasing electricity from the distribution network minus the revenue from selling electricity to the distribution network, plus the operation and maintenance costs of the electrolyzer and fuel cell. Equipment output constraints mean that the power of both the electrolyzer and fuel cell must operate between their rated minimum and maximum values. Hydrogen storage constraints mean that the volume of hydrogen in the hydrogen storage tank must be maintained between the minimum safe storage capacity and the maximum capacity. Power balance constraints mean that in each time period, the sum of photovoltaic output, fuel cell output, and purchased electricity equals the sum of electrical load, electrolyzer power, and sold electricity.

[0040] Specifically, the system can establish an independent lower-level optimization model for each microgrid. The objective function is the sum of electricity purchase cost minus electricity sales revenue plus equipment operation and maintenance costs for each time period. Constraints include: electrolyzer power between 0 and its rated maximum value, fuel cell power between 0 and its rated maximum value, hydrogen storage between the minimum safe storage (e.g., 10% of capacity) and the maximum capacity, and a power balance equation. Furthermore, a hydrogen storage time-series constraint needs to be introduced, i.e., the hydrogen storage in the current time period equals the hydrogen storage in the previous time period plus the hydrogen production from the electrolyzer minus the hydrogen consumption from the fuel cell. For example, a microgrid has a photovoltaic installed capacity of 2 MW, an electrolyzer rated power of 500 kW, a fuel cell rated power of 300 kW, a hydrogen storage tank capacity of 2000 cubic meters, and a minimum safe storage of 200 cubic meters. At a certain time period, the photovoltaic output is 1.2 MW, the electrical load is 1.5 MW, the electricity purchase price is 0.5 yuan per kilowatt-hour, the electricity sales price is 0.3 yuan per kilowatt-hour, the electrolyzer operation and maintenance cost is 0.02 yuan per kilowatt-hour, and the fuel cell operation and maintenance cost is 0.05 yuan per kilowatt-hour. The optimization result of the lower-level model is: electrolyzer power 400 kW (utilizing excess photovoltaic power), fuel cell shutdown, hydrogen storage capacity increases from 500 cubic meters to 560 cubic meters, electricity purchase power is 0.3 MW (to make up for the remaining gap), and the total operating cost is 150 yuan.

[0041] The coupling variable between the upper distribution network coordination layer and the lower hydrogen-electric microgrid autonomous layer is set as interactive power, and the inter-layer iterative convergence criterion and penalty function multiplier update rule are configured.

[0042] In this context, coupling variables refer to the shared decision variables connecting the upper and lower optimization models; in this case, it is the interaction power between each microgrid and the distribution network. The inter-layer iterative convergence criterion is used to determine whether the iterations have reached consensus. It is typically defined as the relative change in the sum of the upper and lower objective function values ​​in two adjacent iterations being less than a preset threshold. The penalty function multiplier update rule refers to adjusting the multipliers of the penalty term based on the current deviation when the iteration has not converged, in order to force the solutions of the upper and lower layers to gradually approach each other. Commonly used methods include the augmented Lagrange multiplier method or the alternating direction method of multipliers (ADMM).

[0043] Specifically, at the start of each iteration, the upper layer allocates the current optimal total interactive power to each microgrid according to a preset decomposition rule (e.g., based on the proportion of photovoltaic installed capacity), as an additional constraint on the lower-level model. After each microgrid solves the lower-level model, it returns its respective interactive power requirement. The upper layer updates the total interactive power based on the feedback and solves again. After each iteration, the inter-layer target deviation is calculated. If the deviation is less than a preset threshold (e.g., 0.01) and the number of iterations does not exceed a preset upper limit (e.g., 50 times), the iteration stops; otherwise, the multiplier is adjusted according to the penalty function multiplier update rule, for example, the multiplier is increased to the original multiplier plus the step size multiplied by the current deviation, and then the next iteration begins. For example, for a system with 3 microgrids, the preset threshold is 0.01, and the preset upper limit is 30 times. After the first iteration, the inter-layer target deviation is 0.35, the initial value of the multiplier is 0, and after the update, the multiplier becomes 0.1. After the 5th iteration, the deviation drops to 0.009, satisfying the convergence condition, and the iteration terminates. Ultimately, the interaction power of each microgrid is consistent with the value allocated by the upper layer, achieving consistency between global and local objectives.

[0044] Therefore, according to the above implementation method, the system can effectively solve the problem of insufficient multi-objective coordinated scheduling efficiency of the distribution network caused by insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints by constructing a two-layer optimization model of upper-layer global collaboration and lower-layer local autonomy, and coordinate coupling variables by means of inter-layer iteration mechanism, thereby improving the new energy consumption rate and carbon reduction efficiency.

[0045] In some embodiments, the step of solving the bi-level optimization model through inter-level iteration includes: The total interactive power of the microgrid group output by the upper distribution network coordination layer is obtained and distributed to each microgrid according to the preset decomposition rules to obtain the adjusted interactive power of each microgrid.

[0046] The total interactive power of the microgrid group refers to the algebraic sum of the power exchanged between all microgrids and the distribution network, obtained from the upper-level optimization model. A positive value indicates that the distribution network supplies power to the microgrid, while a negative value indicates that the microgrid supplies power to the distribution network. The preset decomposition rules refer to the calculation criteria for reasonably allocating the total interactive power to each microgrid. Common rules include allocation based on the proportion of photovoltaic installed capacity of the microgrid or allocation based on the carbon emission reduction contribution of the microgrid. The adjusted interactive power refers to the target interactive power value obtained by each microgrid according to the decomposition rules, which serves as an additional constraint for the lower-level optimization model.

[0047] Specifically, after the system completes the upper-level solution, it reads the total interactive power value and then calculates the interactive power share that each microgrid should bear based on the percentage of its photovoltaic installed capacity to the total installed capacity of all microgrids. If a microgrid has a larger photovoltaic installed capacity, it is allocated a larger share of electricity purchase or sales quota, and vice versa. The system uses this share as the interactive power constraint boundary for that microgrid in the lower-level model. For example, a distribution network connects three microgrids with photovoltaic installed capacities of 2 MW, 3 MW, and 5 MW, respectively, for a total installed capacity of 10 MW. The total interactive power output from the upper level is 2.5 MW (electricity purchase direction). Based on the installed capacity ratio, microgrid 1 is allocated 0.5 MW (2 divided by 10 multiplied by 2.5), microgrid 2 is allocated 0.75 MW, and microgrid 3 is allocated 1.25 MW. Each microgrid uses this as the adjusted interactive power.

[0048] Each microgrid uses the adjusted interactive power as an additional constraint to resolve the lower-level hydrogen-electricity microgrid autonomous layer, update the interactive power requirements of each microgrid, and feed them back to the upper-level distribution network coordination layer.

[0049] In this context, additional constraints refer to new interactive power equality or inequality constraints added to the existing constraints at the lower level, requiring that the deviation between the actual interactive power of the microgrid and the adjusted interactive power be within a certain range. Interactive power demand refers to the power value that the microgrid expects to exchange with the distribution network, obtained after solving the lower-level model; this may differ from the adjusted interactive power. Feedback refers to the process of summarizing the new interactive power demands of each microgrid and sending them to the upper-level model, allowing the upper-level model to adjust the total interactive power in the next iteration.

[0050] Specifically, the system adds a constraint to the lower-level optimization model of each microgrid: the absolute value of the difference between the microgrid's interactive power and the adjusted interactive power does not exceed a preset slack variable (e.g., 0.1 MW). Then, the solver is called again to solve the lower-level model. After the solution is completed, the interactive power values ​​obtained from the optimization of each microgrid are extracted, summarized into a new microgrid group interactive power vector, and uploaded to the upper-level model. For example, the adjusted interactive power of microgrid 1 is 0.5 MW, but after considering its own photovoltaic output and load, the lower-level model finds that the optimal interactive power is 0.45 MW (due to reduced electricity purchase) because of its higher photovoltaic output. The optimal interactive power of microgrid 2 is 0.8 MW, and that of microgrid 3 is 1.3 MW. The system feeds these three values ​​back to the upper level.

[0051] Calculate the inter-layer target deviation. When the inter-layer target deviation meets the preset convergence criterion, terminate the iteration and output the final result; otherwise, update the penalty function multiplier and proceed to the next iteration.

[0052] The inter-layer objective deviation refers to the relative change in the sum of the upper-layer and lower-layer objective function values ​​between two adjacent iterations. It is calculated as the absolute value of the difference between the objective sum of the current iteration and the previous iteration, divided by the objective sum of the previous iteration. The preset convergence criterion is a very small positive threshold (e.g., 0.01); the algorithm is considered convergent when the deviation is less than this threshold. The penalty function multiplier is a penalty parameter in the augmented Lagrange multiplier method, used to gradually tighten the violation cost of coupling constraints during iteration, prompting the solutions of upper and lower layers to converge. The update rule typically adds the multiplier to the step size multiplied by the current violation amount of the coupling constraint.

[0053] Specifically, after each iteration, the system saves the current upper-layer target value F_up and the sum of all lower-layer target values ​​F_down, calculating the total S = F_up + F_down. This is then compared to the sum S_old from the previous iteration, calculating the deviation |S - S_old| / |S_old|. If the deviation is less than a preset threshold (e.g., 0.01) and the number of iterations does not exceed a preset upper limit (e.g., 50), the iteration stops, and the current device operating parameters and interaction power scheme for each microgrid are output. Otherwise, the penalty function multiplier is updated according to the rule: new multiplier = old multiplier + step size × (deviation between current interaction power and adjusted interaction power), and the next iteration begins. For example, after the first iteration, the upper-layer target value F_up = 5000 yuan, the sum of lower-layer target values ​​F_down = 3200 yuan, and the total S = 8200 yuan. After the second iteration, F_up = 5100 yuan, F_down = 3150 yuan, and the total S = 8250 yuan. The deviation |8250-8200| / 8200 = 0.0061, which is less than the threshold of 0.01, therefore the iteration terminates. The system outputs the final electrolyzer power, fuel cell power, hydrogen storage capacity, and the power interaction scheme for each time period.

[0054] Therefore, according to the above implementation method, the system can gradually coordinate the optimization results of the upper and lower layers through the inter-layer iteration mechanism, ensuring the consistency between the global carbon-electric multi-objective optimization and the minimization of local microgrid operating costs, thereby effectively solving the problem of insufficient multi-objective collaborative scheduling efficiency of the distribution network caused by insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints.

[0055] In some embodiments, the step of outputting the optimized schemes for the operating parameters of each device and the interaction power of each time period based on the final iteration results includes: Obtain time-series data on the electrolyzer power, fuel cell power, and hydrogen storage capacity of each microgrid in the final iteration results.

[0056] The final iteration result refers to the optimal solution set of the upper and lower layers after the last iteration, containing the state variable values ​​of each microgrid in all time periods. Electrolyzer power time-series data refers to the numerical sequence of active power consumed by the electrolyzer in each time period, in kW. Fuel cell power time-series data refers to the numerical sequence of active power output by the fuel cell in each time period, in kW. Hydrogen storage time-series data refers to the numerical sequence of hydrogen volume stored in the hydrogen storage tank at the end of each time period, in cubic meters. ).

[0057] Specifically, after iterative convergence, the system extracts the electrolyzer power array, fuel cell power array, and hydrogen storage array for each microgrid across the entire optimization time domain (e.g., 24 time periods) from the solution results of the lower-level optimization model. This data is temporarily stored in memory in a structured table format, with each record containing the microgrid number, time period number, and corresponding power or storage value. For example, in the aforementioned 3-microgrid case, the system extracts the electrolyzer power time-series data for microgrid 1 as follows: Unit: kilowatt; Fuel cell power time-series data are as follows: Unit: kilowatt; Time-series data for hydrogen storage is as follows: The unit is cubic meters.

[0058] Obtain the time-series data of the interactive power between each microgrid and the distribution network at each time period.

[0059] The interactive power time-series data refers to the sequence of active power values ​​exchanged between each microgrid and the distribution network in each time period, with the unit being kW. Positive values ​​indicate that the microgrid purchases electricity from the distribution network, while negative values ​​indicate that the microgrid sells electricity to the distribution network.

[0060] Specifically, the system extracts the interaction power array for each microgrid across all time periods from the solution results of the lower-level optimization model. This array is stored synchronously with the electrolyzer power, fuel cell power, and hydrogen storage data for easy subsequent integration. For example, the time-series interaction power data for microgrid 1 is as follows: The unit is kilowatt. A positive value indicates electricity purchase, and a negative value indicates electricity sale. It can be seen that microgrid 1 sells electricity to the distribution network during the midday peak photovoltaic output period (10:00 to 13:00) and purchases electricity during the morning and evening peak load periods.

[0061] The time series data is integrated into an optimization scheme and then output.

[0062] The optimization scheme refers to a complete set of scheduling instructions that includes the operating parameters and interactive power of all devices in the microgrid. It is usually presented in a standard data format (such as a comma-separated CSV file or a JSON object) and can be directly sent to the microgrid controller for execution.

[0063] Specifically, the system aligns the time-series data of electrolyzer power, fuel cell power, hydrogen storage, and interaction power of each microgrid according to the microgrid number and time period number, merging them into a multidimensional data table. Then, the system calls the output interface to convert this data table into a preset communication protocol format (e.g., Modbus TCP message or IEC 61850 dataset) and forwards it to the local controller of each microgrid through the distribution network energy management platform. Simultaneously, the system saves a backup in the local log for post-event analysis and auditing. For example, the system generates a CSV file containing 3 microgrids and 24 time periods, with columns named: microgrid number, time period, electrolyzer power (kW), fuel cell power (kW), hydrogen storage (m³), and interaction power (kW). The file size is approximately 5 kilobytes. This file is transmitted to each microgrid controller via the network, and the controller adjusts the start-up, shutdown, and output of the electrolyzers and fuel cells accordingly, and controls the grid connection switches with the distribution network. The entire output process takes less than 1 second, meeting real-time scheduling requirements.

[0064] Therefore, according to the above implementation method, the system can transform the global optimal solution obtained by inter-layer iterative solution into directly executable equipment scheduling instructions, realize the unity of carbon-electric multi-objective optimization and hydrogen-electric energy self-balancing, and thus effectively solve the problem of insufficient multi-objective collaborative scheduling efficiency of distribution network caused by insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints.

[0065] In some embodiments, an upper-level distribution network coordination layer is constructed, with carbon-electricity multi-objective optimization as the objective and line transmission power constraints and carbon emission constraints as the constraints, including: The sub-items for multi-objective optimization of carbon electricity are identified, including carbon cost, reliability cost, and benefits.

[0066] Carbon cost refers to the total carbon emissions generated by the microgrid network from electricity purchases within the optimization time domain, multiplied by the carbon price, expressed in yuan. Reliability cost refers to the economic loss caused by load reduction due to insufficient power supply, calculated by multiplying the amount of power shortage by the unit power shortage loss coefficient, expressed in yuan. Revenue refers to the revenue from the distribution network selling electricity to the microgrid minus the expenditure on purchasing electricity from the microgrid, expressed in yuan.

[0067] Specifically, the system calculates the total carbon emissions by multiplying the purchased power of each microgrid by its carbon emission factor (the number of kilograms of carbon dioxide emitted per unit kilowatt-hour), and then multiplying this by the carbon price (yuan per ton of carbon dioxide) to obtain the carbon cost. The reliability cost is calculated by multiplying the preset power shortage loss assessment rate (yuan per kilowatt-hour) by the predicted power shortage. Revenue is calculated based on the time-of-use electricity purchase price and the electricity sales price. For example, a distribution network connecting three microgrids has a total purchased power of 60 megawatt-hours over 24 hours. With a carbon emission factor of 0.5 kilograms per kilowatt-hour, the total carbon emissions are 30 tons. At a carbon price of 100 yuan per ton, the carbon cost is 3000 yuan. Reliability cost: With a predicted power shortage of 2 megawatt-hours and a power shortage loss assessment rate of 500 yuan per megawatt-hour, the reliability cost is 1000 yuan. Revenue: Electricity sales revenue of 8000 yuan, electricity purchase expenditure of 5000 yuan, and net revenue of 3000 yuan. The three sub-items are denoted as C_carbon=3000 yuan, C_reliability=1000 yuan, and Profit=3000 yuan respectively.

[0068] Establish line transmission power constraints to limit the total interactive power of the microgrid group within the transmission capacity range.

[0069] The total interactive power of the microgrid group refers to the algebraic sum of the power exchanged between all microgrids and the distribution network. A positive value indicates that the distribution network supplies power to the microgrid, and a negative value indicates that the microgrid supplies power to the distribution network. The unit is kilowatts. The transmission capacity range refers to the absolute value of the maximum active power that the distribution network lines are allowed to pass through, including the upper limit of the forward transmission capacity and the upper limit of the reverse transmission capacity. The unit is kilowatts.

[0070] Specifically, the system adds an inequality constraint to the upper-level optimization model: the total interactive power of the microgrid group is less than or equal to the upper limit of the forward transmission capacity and greater than or equal to the upper limit of the reverse transmission capacity (usually a negative value). This constraint ensures that the actual transmission power of the line does not exceed the thermal stability limit at any given time. For example, the upper limit of the forward transmission capacity of a distribution network line is 10 MW, and the upper limit of the reverse transmission capacity is -5 MW (allowing the microgrid to send a maximum of 5 MW of power to the distribution network). When solving the upper-level model, the total interactive power of the microgrid group is limited to... Within the kilowatt range. If the calculated total interactive power of the microgrid group is 12 megawatts at a certain time period, it will be constrained and forcibly adjusted to 10 megawatts.

[0071] Construct carbon emission constraints and incorporate the excess cost of carbon emissions into the objective function.

[0072] The carbon emission excess cost refers to the additional cost calculated by multiplying the portion of the microgrid's actual carbon emissions that exceed its carbon allowance by the excess penalty price, expressed in yuan. A carbon allowance refers to the total permitted carbon dioxide emissions allocated to the distribution network by the government or regulatory agency, expressed in tons. The excess penalty price is the fine payable for each ton of excess carbon dioxide emitted, expressed in yuan per ton.

[0073] Specifically, the system adds a term to the upper-level objective function: max(0, actual carbon emissions - carbon quota) × excess penalty price. This soft constraint allows for a certain degree of over-emission, but suppresses excessive emissions through economic penalties, thus achieving a multi-objective trade-off between carbon and electricity. For example, if the carbon quota is 5 tons of CO2 per day, and the actual carbon emissions are 6.2 tons, exceeding the quota by 1.2 tons, with an excess penalty price of 150 yuan per ton, then the cost of exceeding the carbon emission quota is 180 yuan. This 180 yuan is added to the objective function and participates in the weighted summation along with carbon costs, reliability costs, and benefits, affecting the optimization result.

[0074] Therefore, according to the above implementation method, the system can effectively solve the problem of insufficient multi-objective coordinated scheduling efficiency of distribution network caused by insufficient integration of carbon and electricity constraints by clearly defining the carbon and electricity multi-objective components, line transmission power constraints and carbon emission excess cost penalty mechanism, and deeply coupling carbon constraints with power scheduling.

[0075] In some embodiments, a lower-level autonomous layer of the hydrogen-electric microgrid is constructed, with the goal of optimizing the microgrid's operating costs, and constrained by equipment output constraints, hydrogen storage capacity constraints, and power balance constraints, including: The breakdown of microgrid operating costs to be optimized includes electricity purchase costs, electricity sales revenue, and equipment operation and maintenance costs.

[0076] The electricity purchase cost refers to the fee paid by the microgrid to purchase electricity from the distribution network, calculated by multiplying the purchased power for each time period by the corresponding time-of-use electricity price and then summing the results. The unit is yuan. Electricity sales revenue refers to the income obtained by the microgrid from selling electricity to the distribution network, calculated by multiplying the sold power for each time period by the corresponding time-of-use electricity price and then summing the results. The unit is yuan. Equipment operation and maintenance cost refers to the maintenance and wear and tear costs incurred during the operation of the electrolyzer and fuel cell, typically proportional to the equipment output, and is also measured in yuan.

[0077] Specifically, the system calculates the electricity purchase cost for each time period by multiplying the purchased power by the purchased price, and the electricity sales revenue for that time period by multiplying the sold power by the sold price. It also calculates two maintenance costs by multiplying the electrolyzer power by the unit maintenance cost coefficient and the fuel cell power by the unit maintenance cost coefficient, respectively. Finally, the system sums up the electricity purchase costs for all time periods, subtracts the electricity sales revenue for all time periods, and adds the equipment maintenance costs for all time periods to obtain the total operating cost of the microgrid. For example, a microgrid purchases a total of 12 MWh of electricity within 24 hours at an average purchase price of 0.5 yuan per kWh, resulting in a purchase cost of 6000 yuan; it sells a total of 3 MWh of electricity at an average selling price of 0.3 yuan per kWh, generating revenue of 900 yuan; the electrolyzer operates for a cumulative total of 80 MWh, with a maintenance cost coefficient of 0.02 yuan per kWh, resulting in a maintenance cost of 1600 yuan; and the fuel cell operates for a cumulative total of 30 MWh, with a maintenance cost coefficient of 0.05 yuan per kWh, resulting in a maintenance cost of 1500 yuan. The total operating cost is 6000 - 900 + 1600 + 1500 = 8200 yuan.

[0078] Establish constraints on equipment output and hydrogen storage capacity, limiting the power of the electrolyzer and fuel cell to within their respective output ranges, and limiting the hydrogen storage capacity of the hydrogen storage tank to within a safe range.

[0079] The output range refers to the interval between the minimum and maximum power that the electrolyzer and fuel cell are allowed to operate at. Typically, the minimum power is 0 (allowing for shutdown), and the maximum power is the rated power of the equipment. The safety range refers to the lower and upper limits that the hydrogen volume in the hydrogen storage tank must maintain. The lower limit is usually 10% of the total capacity (to prevent complete venting that could damage the equipment), and the upper limit is the total capacity (to prevent overpressure).

[0080] Specifically, the system adds inequality constraints to each microgrid: the electrolyzer power is greater than or equal to 0 and less than or equal to the electrolyzer's rated power; the fuel cell power is greater than or equal to 0 and less than or equal to the fuel cell's rated power; and the hydrogen storage capacity is greater than or equal to the minimum safe storage capacity and less than or equal to the maximum capacity of the hydrogen storage tank. These constraints ensure that the equipment does not operate beyond its physical limits. For example, a microgrid has an electrolyzer rated power of 500 kW, a fuel cell rated power of 300 kW, a hydrogen storage tank capacity of 2000 cubic meters, and a minimum safe storage capacity of 200 cubic meters. The constraints are: the electrolyzer power is between 0 and 500 kW, the fuel cell power is between 0 and 300 kW, and the hydrogen storage capacity is between 200 and 2000 cubic meters. If, at a certain time, the lower-level model attempts to set the electrolyzer power to 550 kW, it will be forced to adjust to 500 kW by the constraints.

[0081] We construct power balance constraints and hydrogen storage time-series constraints to correlate the changes in hydrogen storage and power at different time periods.

[0082] The power balance constraint means that the active power within the microgrid must be conserved within each time period: photovoltaic output plus fuel cell output plus purchased power equals electrical load plus electrolyzer power plus sold power. The hydrogen storage time-series constraint means that the change in the amount of hydrogen in the storage tank is determined by the hydrogen production rate of the electrolyzer and the hydrogen consumption rate of the fuel cell: the hydrogen storage amount in the current time period equals the hydrogen storage amount in the previous time period plus the hydrogen production rate of the electrolyzer minus the hydrogen consumption rate of the fuel cell, where the hydrogen production rate is proportional to the electrolyzer power and the hydrogen consumption rate is proportional to the fuel cell power.

[0083] Specifically, the system establishes an equality constraint for each time period: photovoltaic power P_pv plus fuel cell power P_fc plus purchased electricity P_buy equals electrical load P_load plus electrolyzer power P_el plus sold electricity P_sell. Simultaneously, a recursive relationship for hydrogen storage is established: hydrogen storage H(t) equals H(t-1) plus eta_el multiplied by P_el(t) multiplied by Delta_t minus eta_fc multiplied by P_fc(t) multiplied by Delta_t, where eta_el is the electrolyzer hydrogen production efficiency (cubic meters per kilowatt-hour), eta_fc is the fuel cell hydrogen consumption efficiency (cubic meters per kilowatt-hour), and Delta_t is the time period length (hours). For example, in a microgrid at time t=10 (10 AM), the photovoltaic output P_pv=800 kW, the electrical load P_load=600 kW, the electrolyzer power P_el=400 kW, the fuel cell power P_fc=0, the power purchased P_buy=200 kW, and the power sold P_sell=0. The power balance is: 800 + 0 + 200 = 600 + 400 + 0, which holds true. In the previous time period, the hydrogen storage H(9) = 500 cubic meters, the electrolyzer hydrogen production efficiency eta_el = 0.5 cubic meters per kilowatt-hour, the time period length was 1 hour, and the hydrogen production was 0.5 multiplied by 400 multiplied by 1 equaled 200 cubic meters; the fuel cell hydrogen consumption was 0. In the current time period, the hydrogen storage H(10) = 500 + 200 - 0 equals 700 cubic meters, which meets the hydrogen storage constraint (200 to 2000 cubic meters).

[0084] Therefore, according to the above implementation method, the system can construct an operating cost optimization model that includes electricity purchase cost, electricity sales revenue and equipment operation and maintenance cost, and combine it with equipment output constraints, hydrogen storage constraints, power balance constraints and hydrogen storage time constraints to achieve time-series coupling optimization of hydrogen and electricity energy within the microgrid, thereby effectively solving the problem of low new energy consumption rate caused by insufficient consideration of hydrogen-electric coupling characteristics.

[0085] In some embodiments, the step of initializing the parameters of the hydrogen-electric microgrid includes: The optimized time domain is set to a preset period length, and the preset period is divided into equally spaced time periods, with each time period corresponding to a set of source-load data and time-of-use electricity price data.

[0086] The optimization time domain refers to the time range covered by the distribution network operation optimization, typically taking one day (24 hours) as a complete scheduling cycle. The preset cycle length refers to the total duration of the optimization time domain, in hours. Equal interval time slots refer to several time segments into which the optimization time domain is evenly divided, each segment having the same length, typically 1 hour. Source-load data refers to the predicted photovoltaic output and electrical load for each time slot, in kilowatts. Time-of-use pricing data refers to the price at which the distribution network sells electricity to the microgrid (purchase price) and purchases electricity from the microgrid (sales price) for each time slot, in yuan per kilowatt-hour.

[0087] Specifically, the system reads the optimized time domain length as 24 hours and the time step as 1 hour from the configuration file, thus dividing the entire day into 24 equally spaced time periods, numbered from 1 to 24. Considering that distribution network dispatch instructions are usually issued on an hourly basis, and that fluctuations in photovoltaic output and load have significant characteristics on an hourly scale, a 1-hour time step is selected. Then, the system stores the 24 photovoltaic output values ​​and 24 electrical load values ​​predicted in step S110 into a data table according to the corresponding time periods, and simultaneously obtains the purchase price and sales price of electricity for the 24 time periods from the electricity market interface, forming a complete time period data matrix. For example, the optimized time domain for the distribution network of an industrial park is set to 0:00 to 24:00 on July 15, 2025, a total of 24 hours, with a time step of 1 hour. During time period 1 (0:00~1:00), the photovoltaic output is 0 kW, the electrical load is 300 kW, the purchase price of electricity is 0.3 yuan per kilowatt-hour, and the sales price of electricity is 0.2 yuan per kilowatt-hour. During time period 13 (12:00~13:00), the photovoltaic output is 1200 kW, the electrical load is 800 kW, the purchase price of electricity is 0.6 yuan per kilowatt-hour, and the sales price of electricity is 0.4 yuan per kilowatt-hour. The system stores this data in a two-dimensional array, with the row index being the time period number, and the columns including photovoltaic output, electrical load, purchase price of electricity, and sales price of electricity.

[0088] Configure preset decomposition rules, which can be based on the proportion of microgrid photovoltaic installed capacity or on the contribution of carbon emission reduction.

[0089] The pre-defined decomposition rules refer to the mathematical criteria used by the upper-level distribution network coordination layer to allocate the total interactive power of the microgrid group to each microgrid. Allocation based on the proportion of microgrid photovoltaic installed capacity means allocating power according to the percentage of each microgrid's rated photovoltaic panel power to the total installed capacity; the larger the installed capacity, the greater the share of total interactive power it bears. Allocation based on carbon emission reduction contribution means allocating power according to each microgrid's relative contribution to carbon emission reduction over a past period; microgrids with greater emission reduction contributions receive more favorable interactive power quotas (e.g., lower electricity purchase quotas or higher electricity sales quotas).

[0090] Specifically, during the initialization phase, the system reads the photovoltaic (PV) installed capacity data of each microgrid, calculates the percentage of each microgrid's installed capacity relative to the total installed capacity, and uses this as the weight vector for the default decomposition rule. If the user selects carbon emission reduction contribution allocation, the system extracts the carbon emission reduction (relative to the baseline emission) of each microgrid from the historical database for the most recent week, calculates the percentage of each microgrid's emission reduction relative to the total emission reduction, and uses this as the decomposition weight. For example, if the PV installed capacities of three microgrids are 2 MW, 3 MW, and 5 MW, respectively, with a total installed capacity of 10 MW, the weights based on the installed capacity ratio are 0.2, 0.3, and 0.5. If the total upper-layer interactive power in a certain iteration is 2.5 MW (purchased electricity), then microgrid 1 is allocated 0.5 MW, microgrid 2 is allocated 0.75 MW, and microgrid 3 is allocated 1.25 MW. If carbon emission reduction contribution is used for allocation, assuming microgrid 1 reduced emissions by 2 tons last week, microgrid 2 reduced emissions by 3 tons, and microgrid 3 reduced emissions by 5 tons, for a total reduction of 10 tons, with weights of 0.2, 0.3, and 0.5 respectively, the allocation result will be the same. However, in practice, the two rules may lead to different weight distributions, and users can choose flexibly according to policy guidance.

[0091] We constructed a hydrogen storage time series constraint to correlate the changes in hydrogen storage capacity at different time periods with the hydrogen production capacity of the electrolyzer and the hydrogen consumption of the fuel cell.

[0092] The hydrogen storage time series constraint refers to the recursive formula describing the evolution of hydrogen volume in the storage tank over time. Electrolyzer hydrogen production capacity refers to the output of hydrogen converted from electrical energy per unit time, measured in cubic meters per hour, and is directly proportional to the electrolyzer power and hydrogen production efficiency. Fuel cell hydrogen consumption capacity refers to the amount of hydrogen consumed by the fuel cell to generate electricity per unit time, measured in cubic meters per hour, and is directly proportional to the fuel cell power and hydrogen consumption efficiency. The change in hydrogen storage capacity equals the hydrogen production capacity minus the hydrogen consumption capacity multiplied by the time period.

[0093] Specifically, the system establishes the following recursive relationship for each microgrid: the hydrogen storage at the end of the current time period equals the hydrogen storage at the end of the previous time period plus the hydrogen production of the electrolyzer during this time period minus the hydrogen consumption of the fuel cell during this time period. Here, hydrogen production equals the electrolyzer power multiplied by the hydrogen production efficiency coefficient multiplied by the time period length (in hours), and hydrogen consumption equals the fuel cell power multiplied by the hydrogen consumption efficiency coefficient multiplied by the time period length. This constraint ensures that changes in hydrogen storage are strictly coupled with equipment output in time. For example, a microgrid has an electrolyzer hydrogen production efficiency of 0.5 cubic meters per kilowatt-hour, a fuel cell hydrogen consumption efficiency of 0.4 cubic meters per kilowatt-hour, and a time period length of 1 hour. At the end of the previous time period, the hydrogen storage was 500 cubic meters. In this time period, the electrolyzer power is 400 kilowatts, and the fuel cell power is 200 kilowatts. Then, the hydrogen production is 0.5 multiplied by 400 multiplied by 1 equals 200 cubic meters, and the hydrogen consumption is 0.4 multiplied by 200 multiplied by 1 equals 80 cubic meters. At the end of this period, the hydrogen storage capacity will be 500 + 200 - 80 = 620 cubic meters. This value must fall within the safe range of hydrogen storage constraints (e.g., 200 to 2000 cubic meters).

[0094] Therefore, according to the above implementation method, the system can lay a solid data foundation and constraint framework for solving the subsequent two-layer optimization model by reasonably setting the optimized time domain and time period division, flexibly configuring the interactive power decomposition rules, and accurately constructing the hydrogen storage time sequence coupling relationship. This effectively solves the problem of insufficient multi-objective collaborative scheduling efficiency of the distribution network caused by insufficient consideration of hydrogen-electric coupling characteristics and insufficient integration of carbon-electric constraints.

[0095] According to a second aspect, this invention provides a hydrogen-electricity coordinated distribution network operation optimization system based on carbon-electric coupling constraints, which relies on a distribution network energy management platform. The system includes a processing unit and a storage unit. The storage unit stores executable instructions, and the processing unit integrates a data-driven prediction module, a two-layer optimization solution module, and an inter-layer iterative coordination module. When the executable instructions are invoked by the processing unit, they drive the system to execute any of the hydrogen-electricity coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in this invention's embodiments.

[0096] According to a third aspect, the present invention provides an electronic device comprising: at least one processing unit; and a storage unit communicatively connected to the at least one processing unit. The storage unit stores executable instructions that can be invoked by the at least one processing unit. When the executable instructions are executed by the at least one processing unit, the at least one processing unit causes it to execute any of the hydrogen-electric coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in the embodiments of the present invention.

[0097] According to another aspect of the present invention, a medium for storing executable instructions is provided, wherein the computer instructions are used to cause a computer to execute any of the hydrogen-electric coordinated distribution network operation optimization methods based on carbon-electric coupling constraints in the embodiments of the present invention.

[0098] Figure 4 A schematic block diagram of an electronic device 600 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0099] like Figure 4 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0100] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and communication unit 609, such as network card, modem, wireless transceiver, etc. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0101] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as a method for optimizing the operation of a hydrogen-electric co-generation distribution network based on carbon-electric coupling constraints. For example, in some embodiments, a method for optimizing the operation of a hydrogen-electric co-generation distribution network based on carbon-electric coupling constraints can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the method for optimizing the operation of a hydrogen-electric co-generation distribution network based on carbon-electric coupling constraints described above can be performed. Alternatively, in other embodiments, the computing unit 601 may be configured, by any other suitable means (e.g., by means of firmware), to perform a hydrogen-electric coordinated distribution network operation optimization method based on carbon-electric coupling constraints.

[0102] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0103] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0104] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0106] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0107] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0108] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0109] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for optimizing the operation of a hydrogen-electric coordinated distribution network based on carbon-electric coupling constraints, characterized in that, include: Acquire scenario variables and predict source-load data based on data-driven methods, and initialize hydrogen-electric microgrid parameters, distribution network parameters, and multi-objective weight parameters; A two-layer optimization model is constructed, consisting of an upper distribution network coordination layer and a lower hydrogen-electric microgrid autonomous layer. The upper distribution network coordination layer determines the total interactive power of the microgrid group based on carbon-electric multi-objective optimization, while the lower hydrogen-electric microgrid autonomous layer optimizes the electrolyzer power, fuel cell power, hydrogen storage capacity, and single microgrid interactive power by minimizing the microgrid operating cost. The two-layer optimization model is solved by inter-layer iteration. In response to the inter-layer target deviation being less than a preset threshold and the number of iterations not exceeding a preset upper limit, the optimization schemes for the operating parameters of each device and the interaction power of each time period are output based on the final iteration results.

2. The method according to claim 1, characterized in that, The construction steps of the two-layer optimization model include: The upper-level distribution network coordination layer is constructed with carbon-electricity multi-objective optimization as the objective and line transmission power constraints and carbon emission constraints as the constraints. The lower-level hydrogen-electric microgrid autonomous layer is constructed with the goal of optimizing the microgrid operating cost and with constraints on equipment output, hydrogen storage capacity, and power balance. The coupling variable between the upper distribution network coordination layer and the lower hydrogen-electric microgrid autonomous layer is set as interactive power, and the inter-layer iterative convergence criterion and penalty function multiplier update rule are configured.

3. The method according to claim 2, characterized in that, The step of solving the two-layer optimization model through inter-layer iteration includes: The total interactive power of the microgrid group output by the upper distribution network coordination layer is obtained and distributed to each microgrid according to the preset decomposition rules to obtain the adjusted interactive power of each microgrid. Each microgrid uses the adjusted interactive power as an additional constraint to resolve the lower-level hydrogen-electricity microgrid autonomous layer, updates the interactive power requirements of each microgrid, and feeds them back to the upper-level distribution network coordination layer. Calculate the inter-layer target deviation. When the inter-layer target deviation meets the preset convergence criterion, terminate the iteration and output the final result; otherwise, update the penalty function multiplier and proceed to the next iteration.

4. The method according to claim 3, characterized in that, The steps for outputting the optimized schemes for the operating parameters of each device and the interaction power of each time period based on the final iteration results include: Obtain time-series data of electrolyzer power, fuel cell power, and hydrogen storage capacity for each microgrid in the final iteration results; Obtain the time-series data of the interactive power between each microgrid and the distribution network at each time period; The time series data is integrated into an optimization scheme and then output.

5. The method according to claim 2, characterized in that, The construction of the upper-level distribution network coordination layer, with carbon-electricity multi-objective optimization as the objective and line transmission power constraints and carbon emission constraints as the constraints, includes: The sub-items of the carbon-electric multi-objective optimization are determined, including carbon cost, reliability cost, and benefits; Construct the line transmission power constraint to limit the total interactive power of the microgrid group within the transmission capacity range; The carbon emission constraints are constructed by incorporating the excess cost of carbon emissions into the objective function.

6. The method according to claim 2, characterized in that, The construction of the lower-level hydrogen-electric microgrid autonomous layer aims to optimize microgrid operating costs and is constrained by equipment output constraints, hydrogen storage capacity constraints, and power balance constraints, including: The components of the microgrid operation cost optimization are determined, including electricity purchase cost, electricity sales revenue, and equipment operation and maintenance cost; Establish constraints on equipment output and hydrogen storage capacity, limiting the power of the electrolyzer and fuel cell to within their respective output ranges, and limiting the hydrogen storage capacity of the hydrogen storage tank to within a safe range. We construct power balance constraints and hydrogen storage time-series constraints to correlate the changes in hydrogen storage and power at different time periods.

7. The method according to claim 2, characterized in that, The steps for initializing the parameters of the hydrogen-electric microgrid include: The optimized time domain is set to a preset period length, and the preset period is divided into equally spaced time periods, with each time period corresponding to a set of source load data and time-of-use electricity price data. Configure preset decomposition rules, which are based on the proportion of microgrid photovoltaic installed capacity or the proportion based on carbon emission reduction contribution. We constructed a hydrogen storage time series constraint to correlate the changes in hydrogen storage capacity at different time periods with the hydrogen production capacity of the electrolyzer and the hydrogen consumption of the fuel cell.

8. A hydrogen-electricity coordinated distribution network operation optimization system based on carbon-electric coupling constraints, characterized in that, The system relies on the power distribution network energy management platform; The system includes a processing unit and a storage unit. The storage unit stores executable instructions. The processing unit integrates a data-driven prediction module, a two-layer optimization solution module, and an inter-layer iterative coordination module. When the executable instructions are invoked by the processing unit, they drive the system to execute the method as described in any one of claims 1 to 7.

9. An electronic device, characterized in that, include: At least one processing unit; and a storage unit communicatively connected to the at least one processing unit; The storage unit stores executable instructions that can be invoked by the at least one processing unit. When the executable instructions are executed by the at least one processing unit, the at least one processing unit causes the at least one processing unit to perform the method as described in any one of claims 1 to 7.

10. A medium for storing executable instructions, characterized in that, When the executable instruction is invoked by the processing unit, it causes the processing unit to perform the method as described in any one of claims 1 to 7.