Double-layer capacity optimization configuration method for multi-scene hydrogen load comprehensive energy system

Abstract

The invention relates to the technical field of hydrogen load capacity configuration, in particular to a multi-scene hydrogen load comprehensive energy system double-layer capacity optimal configuration method, which comprises the following steps of: firstly, constructing a multi-scene model containing strong, weak, discontinuous and non-hydrogen loads based on the time sequence characteristics of hydrogen loads in an industrial park; then a comprehensive energy system mathematical model covering wind and light power generation, hydrogen energy conversion, multi-type energy storage and cogeneration units is established; then, constructing a double-layer capacity optimization configuration model of which the upper layer takes the whole life cycle cost minimization and the lower layer takes the operation cost minimization and the system stability optimization as targets; and finally, solving a Pareto leading edge by adopting a non-dominated sorting genetic algorithm with an elitist strategy, making a decision by utilizing an approximate ideal solution sorting method, and screening out a capacity configuration scheme with optimal comprehensive performance. The problem that a traditional configuration scheme is difficult to adapt to multi-scene hydrogen load fluctuation is effectively solved, and the economical efficiency, the stability and the renewable energy consumption capacity of the system are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of hydrogen load capacity configuration technology, and specifically to a two-layer capacity optimization configuration method for a multi-scenario integrated energy system with hydrogen load. Background Technology

[0002] With the urgent need for global low-carbon development strategies and energy system reform, building a clean energy-dominated power system has become a key path. However, wind and solar power are constrained by natural conditions, and their power generation exhibits random fluctuations, posing challenges to the stability and reliability of energy supply. Energy Equivalent Systems (IES), as an important component and carrier of the new energy system, differ from traditional energy systems by maximizing the integration of various resources and appropriately optimizing and coordinating the production, storage, transportation, conversion, and consumption processes of different types of energy systems. Through multi-energy complementarity including electricity, heat, cooling, and gas, IES offers advantages in energy cascade utilization and multi-energy coupling complementarity, effectively improving energy utilization efficiency and renewable energy absorption rates. Among them, High-Speed ​​Storage Systems (HSS), with their high energy density, cross-seasonal storage capabilities, and clean, zero-carbon characteristics, have become a core technology for integrating fluctuating new energy sources such as wind and solar power and enhancing system flexibility, and their applications in industry, transportation, and other fields are becoming increasingly widespread.

[0003] Energy system (IES) capacity configuration is crucial for the efficient and coordinated utilization of multiple energy sources, improving economic efficiency and environmental benefits. Researchers typically select the lowest cost or maximum benefit as the optimization objective for different application scenarios, conducting capacity configuration calculations for various energy devices, including thermal power, hydrogen energy, and energy storage. Research focuses on multi-objective optimization, system dynamics, and equipment combination configuration to address the issues of output fluctuations and load uncertainties in renewable energy.

[0004] However, hydrogen load exhibits significant temporal differences across various scenarios, making traditional single-capacity configuration schemes ill-suited to diverse needs, leading to decreased system economy and stability. Existing research largely focuses on single hydrogen load scenarios, lacking adaptive optimization schemes for multiple scenarios. Algorithms and frameworks for solving multi-objective optimization models rarely comprehensively consider objectives such as balancing equipment lifecycle costs, operating costs, and system stability. Therefore, a capacity configuration method that can accommodate hydrogen loads across multiple scenarios is needed. Summary of the Invention

[0005] The purpose of this invention is to provide a two-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system. By constructing a multi-scenario hydrogen load model and a two-layer optimization architecture, and using the NSGA-II and TOPSIS algorithms, the optimal capacity configuration of the system is achieved under the balance of economy and stability, effectively improving the system's adaptability to diverse hydrogen loads and energy utilization efficiency.

[0006] To achieve the above-mentioned technical objectives and effects, the present invention is implemented through the following technical solution: A two-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system includes the following steps: S1: Construct a multi-scenario hydrogen load model: Based on the specific physical application scenarios of hydrogen load in industrial parks in industrial production and transportation, four typical scenarios are set up: strong hydrogen load, weak hydrogen load, intermittent hydrogen load and no hydrogen load. The hydrogen load demand curve is quantitatively determined based on the time-series characteristics of hydrogen load in each scenario. The time-series characteristics include bimodal curves, pulse fluctuations and nighttime peaks. S2: Establish a comprehensive energy system mathematical model: Construct a comprehensive energy system architecture including wind and solar power generation equipment models, hydrogen energy conversion equipment models, multi-type energy storage equipment models, and traditional cogeneration unit models, clarify the power conversion relationships and operating constraints between each piece of equipment. The hydrogen energy conversion equipment model includes an electric hydrogen production process, a hydrogen-to-electricity and heat production process, and a hydrogen-to-methane production process. The energy storage equipment model includes a hydrogen storage process, electric energy storage, thermal energy storage, and gas energy storage. S3: Construct a two-layer capacity optimization configuration model: Establish an optimization configuration model with upper and lower layer feedback structures. The upper layer optimization takes minimizing the total life cycle cost of equipment as the single objective, and the decision variables are the installed capacity of wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, methanation units, and cogeneration units. The lower layer optimization takes minimizing the operating energy purchase and sale cost and optimizing system stability as multiple objectives. Based on the equipment capacity configuration scheme given by the upper layer, simulate the typical 24-hour operation scheduling process and output quantitative indicators of operating cost and stability. S4: Optimization and Decision-Making: A non-dominated sorting genetic algorithm with an elite strategy is used to solve the upper-level multi-objective optimization problem, generating a Pareto front solution set that represents the balance between economy and stability. Then, the approximation ideal solution sorting method is used to make multi-objective decisions on the solutions in the Pareto front solution set, and the capacity configuration scheme with the best overall performance is selected.

[0007] Furthermore, step S1 specifically includes: The strong hydrogen load corresponds to the ammonia production process, which has a typical bimodal curve characteristic. During the nighttime off-peak period, it operates at the minimum safe operating load, and the load transition is smooth and in line with the actual equipment adjustment capacity. The weak hydrogen load corresponds to the steel chemical industry scenario, including annealing furnace protective gas and continuous casting cooling applications. Compared with the ammonia production process, the total load is significantly reduced and exhibits pulse-like fluctuation characteristics, reflecting the intermittent high demand of the application scenario, while maintaining a bi-peak structure, with the most basic hydrogen consumption maintained during the nighttime valley period. The intermittent hydrogen load corresponds to the application of hydrogen refueling stations, which have obvious intermittent peak and trough periods of hydrogen load. At night, it exhibits explosive load characteristics to centrally handle the hydrogen refueling needs of logistics vehicles and buses, while during the day, it manifests as fragmented needs for short-term, low-volume refueling by private cars. The equipment operation must follow the forced cooling protection mechanism after high-power operation.

[0008] Furthermore, step S2 specifically includes: Construct models for wind and solar power generation equipment, namely, mathematical models of wind turbine power and mathematical models of photovoltaic power. The mathematical model for wind turbine power (WT) is as follows: In the formula: for Wind speed at any given moment; For a given Output power at any given time (kW); Pr is the rated output power; To cut in wind speed; Rated wind speed; To cut off the wind speed.

[0009] The mathematical model for photovoltaic (PV) generator power is as follows: In the formula: For photovoltaic power stations The actual output power at any given time; Let be the solar radiation intensity received by the photovoltaic cell area at time t; and let A be the photovoltaic cell area. The nominal efficiency is given under standard test conditions; β is the temperature coefficient. For photovoltaic cells in The operating temperature at any given time; Th represents the ambient temperature.

[0010] The hydrogen energy conversion equipment model includes an electro-hydrogen production process, a hydrogen-to-electricity and heat production process, and a hydrogen-to-methane production process; Hydrogen production via electrolysis: In the formula, Let be the hydrogen production power at time t; Let t be the electrical power input to the electrolytic cell. The energy conversion efficiency of the electrolyzer for hydrogen production; This refers to the amount of hydrogen that can be produced per kilowatt-hour of electricity in the electrolyzer. This represents the upper limit of the electrical power of the electrolytic cell; This is the rated capacity of the electrolytic cell, used to limit its ramp rate.

[0011] Hydrogen generation electrothermal process: In the formula, Let be the hydrogen power consumed by the hydrogen fuel cell at time t; denoted as η, where η is the energy conversion efficiency of the hydrogen fuel cell from hydrogen to electricity; r is the amount of hydrogen consumed by the hydrogen fuel cell to generate one kilowatt-hour of electricity. / Let t be the electrical power and thermal power output of the hydrogen fuel cell at time t, and let the adjustable electrothermal ratio be controlled in a coordinated manner. This represents the upper limit of hydrogen consumption power in a hydrogen fuel cell; This is the rated capacity of the hydrogen fuel cell, used to limit its ramp rate. / These are logical variables representing the start-up and shutdown states of the electrolyzer and the hydrogen fuel cell device at time t. =1 indicates that the cell is producing hydrogen, while the other indicates that the cell is consuming hydrogen. If both are 0, it means that neither the cell nor the cell is in operation.

[0012] Hydrogen to methane production process: In the formula, Let be the hydrogen power consumed by the hydrogen fuel cell at time t; denoted as η, where η is the energy conversion efficiency of the hydrogen fuel cell from hydrogen to electricity; r is the amount of hydrogen consumed by the hydrogen fuel cell to generate one kilowatt-hour of electricity. / Let t be the electrical power and thermal power output of the hydrogen fuel cell at time t, and let the adjustable electrothermal ratio be controlled in a coordinated manner. This represents the upper limit of hydrogen consumption power in a hydrogen fuel cell; This is the rated capacity of the hydrogen fuel cell, used to limit its ramp rate. / These are logical variables representing the start-up and shutdown states of the electrolyzer and the hydrogen fuel cell device at time t. =1 indicates that the cell is producing hydrogen, while the other indicates that the cell is consuming hydrogen. If both are 0, it means that neither the cell nor the cell is in operation.

[0013] The models of various types of energy storage devices include hydrogen storage, electrical energy storage, thermal energy storage, and gas energy storage; In the hydrogen storage stage, high-pressure hydrogen storage tanks and low-pressure hydrogen storage tanks are installed: In the formula, / / / The hydrogen filling and discharging efficiencies at time t are the low-pressure hydrogen storage tank and the high-pressure hydrogen storage tank, respectively. / / / These represent the maximum hydrogen charging / discharging efficiencies for SH and HSH, respectively. / / / These are the logical variables for hydrogen charging and discharging at time t. A value of 1 indicates that the low-pressure hydrogen storage tank is in hydrogen storage mode; otherwise, it is in hydrogen release mode. =1 indicates that the high-pressure hydrogen storage tank is in the hydrogen storage state, and vice versa; SH(t) is the hydrogen storage state in the low-pressure hydrogen storage tank at time t; HSH(t) is the hydrogen storage state in the high-pressure hydrogen storage tank at time t. / / / These are the hydrogen filling and discharging efficiency coefficients of the hydrogen storage tank; / This refers to the rated capacity of the hydrogen storage tank. / The hydrogen storage ratio in the hydrogen storage tank at time t; / These are the upper and lower limits for the hydrogen storage ratio in low-pressure hydrogen storage tanks. / These are the upper and lower limits for the hydrogen storage ratio in high-pressure hydrogen storage tanks.

[0014] Electric energy storage: In the formula, Let t be the energy storage capacity at time t; τ be the self-discharge rate of the energy storage. / These represent the charging power and discharging power of the energy storage at time t, respectively. / These represent the charging efficiency and discharging efficiency of energy storage, respectively. The rated capacity is the battery's rated capacity; the SOC is the instantaneous indicator of the battery's current remaining power. / These are the lower and upper limits for the charging state, respectively; / These are the logical variables representing the charging and discharging of energy storage at time t. =1 indicates that it is in the charging state, and the opposite indicates that it is in the discharging state; This represents the upper limit of charging and discharging power. / and represent the charging and discharging power of the battery at time t, respectively.

[0015] Thermal energy storage: In the formula, SQ(t) is the thermal energy storage capacity at time t; This refers to the heat loss rate; / These represent the heat storage and heat release power at time t, respectively. / These are the heat storage efficiency and the heat release efficiency, respectively. This refers to the rated capacity of the thermal storage tank. This is an instantaneous indicator of the current remaining heat in the thermal storage tank; / These are the lower and upper limits of the thermal storage state, respectively; / Let the logical variables characterize energy storage and heat release at time t, respectively. =1 indicates that it is in the state of heat storage, and the opposite indicates that it is in the state of heat release. This represents the upper limit of the heat storage and release power. / , respectively, represent the heat storage and release power of the thermal storage tank at time t.

[0016] Gas storage: Let be the gas storage capacity at time t; This refers to the gas consumption loss rate; / These represent the gas storage and gas release power of the energy storage system at time t, respectively. / These are the gas storage and venting efficiencies, respectively. This refers to the rated capacity of the gas storage tank. This is an instantaneous indicator of the current remaining gas volume in the gas storage tank; / These are the lower and upper limits of the gas storage state, respectively; / These are the logical variables representing energy storage and gas release at time t. =1 indicates that it is in the gas storage state, and the opposite indicates that it is in the gas release state. This is the upper limit of the gas storage and release capacity; / These represent the gas storage and release power of the gas storage tank at time t.

[0017] Construct a model of a traditional combined heat and power unit using natural gas as feedstock; In the formula, Let t be the natural gas consumed by the combined heat and power unit; Let t be the amount of natural gas the system purchases from the natural gas network at time t; This is the upper limit of the gas consumption power of a combined heat and power (CHP) unit; To improve the energy conversion efficiency of gas-to-electricity conversion in combined heat and power (CHP) units; The amount of natural gas consumed to generate 1 kWh of electricity for a combined heat and power (CHP) unit; / The figures represent the power generation and heat generation efficiencies of the combined heat and power (CHP) unit at time t, respectively, and the dynamic power-to-heat ratio is used during operation; the waste heat recovery heating efficiency is 95%. The rated capacity of the combined heat and power (CHP) unit is used to limit its ramp rate.

[0018] Furthermore, step S3 specifically includes: Based on a comprehensive energy system mathematical model, a two-layer capacity optimization configuration model is constructed: the upper-layer optimization aims to minimize the total life-cycle cost of equipment, with the decision variables being the capacity of wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, and combined heat and power units; the lower-layer optimization aims to minimize the cost of purchasing and selling energy and optimize system stability. Based on the equipment capacity given in the upper layer, a typical 24-hour operation scheduling is simulated, and operating cost and stability indicators are output. Details are as follows: The economic objective of the upper-level optimization model is mainly the total lifecycle cost of its main equipment, specifically including the initial investment cost C of the system. d Operation and maintenance costs and failure costs C O&M And equipment disposal costs C s The expression for the economic objective function of the upper-level model is: In the formula, Total lifecycle cost of the equipment; , , , , , , These are the average daily costs over the entire lifecycle of wind power, photovoltaic power, electrolyzers, hydrogen storage tanks, hydrogen fuel cells, methane reactors, and combined heat and power equipment. In the calculation, failure costs and operation and maintenance costs are combined into operation and maintenance costs. The rated capacity of each device; Construction cost per unit capacity for each piece of equipment; The cost of operation and maintenance per unit capacity; The disposal cost per unit capacity; The equipment's service life is represented by r, which is the annual interest rate, typically 0.03.

[0019] The lower-level optimization aims to minimize the cost of purchasing and selling energy and optimize system stability. The economic objective of the lower-level optimization model during system operation primarily involves the costs of purchasing, selling, and discarding energy, specifically including the system's energy purchase cost, energy sale cost, and energy discard cost. The expression for the economic objective function of the lower-level scheduling model is: In the formula, The total cost of purchasing and selling abandoned energy for system scheduling; Energy purchase costs for system scheduling; For system scheduling energy sales costs; The cost of energy curtailment during system scheduling; / These are the costs of purchasing electricity and gas, respectively. / Let t represent the amount of electricity and gas purchased. / The unit price for purchasing electricity and gas; / These are the costs of selling heat and electricity, respectively. The amount of heat sold at time t; This refers to the unit price for hot water; / / / The on-grid electricity price for electricity sold by different devices; / The costs of abandoning electricity and hydrogen, respectively; / These represent the maximum power output that the wind power and photovoltaic installations can generate at time t, respectively. / , respectively, represent the amount of wind and solar power that can be absorbed during the scheduling at time t; ε is the penalty factor for wind and solar power curtailment; The amount of hydrogen discarded at time t; The unit price for purchasing and selling hydrogen.

[0020] System stability is also a crucial indicator in system scheduling. In this invention, stability is represented by grid loss Pw, and the expression for the stability objective function of the lower-level scheduling model is: In the formula This refers to the power grid loss per hour; This is the stability characterization function.

[0021] The constraints of the integrated energy system model include the supply and demand balance constraints of four major energy sources: electricity, heat, gas, and hydrogen.

[0022] Electric power balance constraints: Thermal power balance constraint: Gas power balance constraints: Hydrogen power balance constraint: Furthermore, step S4 specifically includes: The NSGA-II algorithm is used to solve the upper-level optimization problem and generate the Pareto front solution set. The TOPSIS method is then used to make multi-objective decisions on the solutions in the Pareto front solution set and select the capacity configuration scheme with the best overall performance.

[0023] Multi-objective optimization problems involve mutual constraints and exclusions, differences in the dimensions of various objective functions, and the inability to obtain optimal solutions simultaneously. Therefore, multi-objective optimization problems generally yield only one set of optimal solutions. Within this set, it is impossible to further compare the optimal solutions among the other solutions. Such solutions are called non-dominated solutions or Pareto optimal solutions.

[0024] The specific steps of the NSGA-II optimization algorithm are as follows: (1) Fast Non-Dominated Sort. The NSGA-II algorithm uses a fast non-dominated sorting mechanism to classify individuals in the population. This process first calculates the dominance degree nx of each individual x (i.e., the number of individuals that dominate the solution), and then divides the population into several levels according to the dominance relationship: the first level consists of individuals with Pareto front nx=0, the second level contains solutions that are dominated by the first level but not dominated by each other, and so on, until all individuals are classified.

[0025] (2) Crowding Calculation. The distance between individuals during evolution reflects their distribution density in the solution space. Using the crowding index can reduce the impact of human intervention and achieve objective comparison among individuals in the population. This mechanism ensures that the Pareto optimal solution set is uniformly distributed in the target space, thereby maintaining the diversity of optimal solution selection.

[0026] (3) Tournament Decision-Making. When selecting parent individuals to participate in crossover / mutation, this paper uses a binary tournament method. Based on the individual's fitness function value, two individuals are randomly selected to compare their fitness, and the individual with the best fitness is retained. This process is repeated iteratively until the number of selected individuals matches the population size.

[0027] (4) Crossover and mutation.

[0028] 1) Simulated binary crossover: Offspring are generated among individuals with real-number encodings to maintain solution diversity. Two individuals are selected from any iterative population and designated as... and The j-th gene location of its individual is used and This indicates that new individuals are generated through crossover. and The calculation formula is as follows: The crossover probability density function is given by the following equation: 2) Polynomial Mutation: This involves introducing random perturbations to avoid local optima. A mutation rate parameter is added, and individuals generated after crossover undergo a gene mutation to obtain new individuals in the population. Let the individual... Represented as an individual The j-th gene, , These are the upper and lower bounds, respectively. From which mutated individuals are obtained... The process is as follows: randomly generate a random number. ∈(0,1), calculate ,individual ,judge Check if the boundary is exceeded; if so, take the boundary value.

[0029] In the formula, It is a self-defined non-negative real number called the variation distribution index. The size of the value affects the degree of variation; the larger the value, the smaller the difference between the mutated value and the original value.

[0030] (5) Elite Retention Strategy In each generation, individuals from the parent and child generations are merged. Through non-dominated sorting and crowding calculation, individuals with high fitness are selected to generate a new population, completing the iterative optimization process. This method ensures that excellent individuals are not lost and improves the convergence speed and solution quality of the algorithm.

[0031] TOPSIS-based scheme decision-making. The solution obtained by the above two-level optimization algorithm is a Pareto-like solution set composed of multiple non-dominated solutions, each of which can be considered an optimal solution. Therefore, in actual construction scenarios, investors need to select the solution that best meets the actual needs from all the obtained non-dominated solutions as the final equipment configuration scheme, which leads to a multi-objective decision-making problem. This invention uses the TOPSIS method to determine the configuration scheme, which selects the best option based on its proximity to the positive and negative ideal solutions. The implementation steps are as follows: (1) Construct a standardized matrix. For the case with n schemes and n×m indicators, each indicator is positiveized and normalized to form a standardized matrix.

[0032] (2) Determine the positive ideal solution Z+ and the negative ideal solution Z-. The positive and negative ideal solutions are respectively composed of the maximum and minimum values ​​of each column in Z.

[0033] (3) Calculate the distance between each scheme and the positive and negative ideal solutions, and calculate the closeness. .

[0034] In the formula, The weight of the m-th indicator; 0 ≤ ≤1, The larger the value, the better the overall performance of the solution.

[0035] The TOPSIS decision-making method has significant advantages. It is based on raw data analysis and has no limit on the number of decision objectives or sample size. It is suitable for regional IES capacity optimization, can objectively evaluate candidate solutions, and ultimately obtain the optimal planning solution.

[0036] The optimal prediction results are embedded into a mixed integer linear programming model, and the optimized scheduling results are simulated and solved. The constraints of the MILP model include comprehensive energy system constraints, including internal electrical, thermal, and hydrogen power balance constraints, carbon capture constraints, and equipment output constraints.

[0037] Furthermore, the optimal capacity configuration scheme ultimately output by the method specifically includes quantitative values ​​for the number of wind turbine generators installed, the number of photovoltaic generators installed, the rated capacity of the electrolyzer, the rated capacity of the hydrogen fuel cell, the rated capacity of the hydrogen storage tank, the rated capacity of the methanation unit, and the rated capacity of the combined heat and power unit, providing direct design basis for the actual construction of the integrated energy system in the industrial park.

[0038] The beneficial effects of this invention are: This invention effectively solves the industry challenge of traditional single-capacity configuration schemes being unable to adapt to diverse energy demands by constructing a refined multi-scenario hydrogen load model. Targeting the different application characteristics of hydrogen load in industrial parks across industrial production and transportation, four typical scenarios are set up: strong hydrogen load, weak hydrogen load, intermittent hydrogen load, and no hydrogen load. The hydrogen load demand curve is quantified based on the temporal characteristics of each scenario (such as bimodal curves, pulse fluctuations, and nighttime peaks). This differentiated modeling approach allows the system capacity configuration to accurately respond to dynamic demand changes in different scenarios. When facing the intermittent peak loads unique to hydrogen refueling stations, the system configures hydrogen storage tanks of appropriate capacity and introduces a forced cooling protection mechanism, satisfying both the explosive demand for concentrated hydrogen refueling of logistics vehicles at night and ensuring the safe operation of the equipment. This refined modeling based on scenario characteristics fundamentally improves the system's adaptability to load fluctuations and avoids energy waste or supply shortages caused by unreasonable configuration.

[0039] The integrated energy system mathematical model constructed in this invention realizes the synergistic optimization of multiple energy flows and the cascade utilization of energy. The system architecture fully integrates wind and solar power generation equipment, hydrogen energy conversion equipment (including multiple conversion stages such as electro-hydrogen production, hydrogen-to-electricity and heat production, and hydrogen-to-methane production), and various types of energy storage equipment. Through clearly defined power conversion relationships and operational constraints, it forms an organic whole of complementary electricity, heat, hydrogen, and gas. By using hydrogen energy as a flexible energy carrier, the inherent randomness and volatility of wind and solar power generation are effectively mitigated. During periods of surplus wind and solar power generation, the system converts electrical energy into hydrogen energy for storage via an electrolyzer; when power generation is insufficient, the stored hydrogen energy can be converted into the required forms of electricity, heat, etc., through fuel cells or methanation devices. This multi-energy coupling conversion mechanism not only significantly improves the local absorption capacity of renewable energy but also enhances overall energy efficiency through the cascade utilization of energy, providing a reliable path for the system to achieve low-carbon operation.

[0040] This invention's two-layer optimization configuration model achieves synergy between long-term planning and short-term operation in its decision-making mechanism. The upper-layer optimization aims to minimize the total lifecycle cost of equipment, comprehensively considering equipment investment, operation and maintenance, and disposal costs, thus constructing an economically sustainable capacity configuration scheme for the system. The lower-layer optimization focuses on the energy purchase and sales costs and stability of the system's daily operation. By simulating typical 24-hour operation scheduling, it accurately assesses the economic operating range and safe and stable boundaries under a given capacity configuration. Through a feedback mechanism between the upper and lower layers, equipment capacity planning and system operation strategies are closely linked, effectively overcoming the dilemma of balancing economy and stability in traditional methods. In particular, by introducing a non-dominated sorting genetic algorithm to generate a Pareto front solution set, and then combining it with an approximation-ideal-solution sorting method for multi-objective decision-making, the final selected capacity configuration scheme achieves an optimal balance among multiple mutually constraining objectives.

[0041] This invention, through mechanisms such as fast non-dominated sorting, crowding calculation, and elite retention, ensures both the diversity of the optimal solution set and improves the algorithm's convergence performance. The subsequent approximation of the ideal solution sorting decision method objectively evaluates candidate schemes based on the original data, effectively avoiding the influence of subjective preferences on the decision results. The final output includes specific quantitative configuration parameters such as the number of wind turbine generators, photovoltaic generators, and the rated capacity of electrolyzers, providing clear technical guidance for the actual engineering construction of industrial parks. This complete technical route from optimization modeling to decision output significantly reduces the engineering trial-and-error costs caused by improper configuration and enhances the operability of the solution in practice.

[0042] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a diagram illustrating the hydrogen load demand in various scenarios as described in this embodiment of the invention. Figure 2 This is the overall energy flow diagram of a two-layer capacity optimization configuration method considering hydrogen load fluctuations in multiple scenarios, as described in an embodiment of the present invention. Figure 3 The basic flow of the NSGA-II algorithm described in this embodiment of the invention; Figure 4 This is a flowchart illustrating the solution process for the dual-layer capacity optimization configuration as described in an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Example 1 This embodiment describes a two-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system, which addresses the multi-scenario hydrogen load demand, such as... Figure 1 As shown, this embodiment combines the application scenarios of hydrogen energy in industries such as industry and transportation, and sets up four different scenarios of hydrogen load demand, which correspond to strong hydrogen load (ammonia production process), weak hydrogen load (steel and chemical industry), intermittent hydrogen load (hydrogen refueling station) and control group with no hydrogen load.

[0047] Ammonia production process: It has a typical bimodal curve, with the lowest safe operating load during the nighttime trough, and the actual equipment adjustment capacity for smooth transition load; Steel and chemical industries: Typical hydrogen energy applications include annealing furnace protective gas and continuous casting cooling. Compared with ammonia plants, the total amount will be significantly reduced and may have pulse fluctuations, reflecting the intermittent high demand of application scenarios. It also has a bimodal structure, maintaining the most basic hydrogen consumption during the nighttime valley period.

[0048] In the transportation sector, the most obvious characteristic of hydrogen load at hydrogen refueling stations is the intermittent peak and off-peak periods. At night, there is a surge in load, which is concentrated on handling hydrogen refueling for logistics vehicles and buses. During the day, the fragmented demand is mainly for short-term, low-volume refueling for private cars. For the equipment, a protection mechanism that requires forced cooling time is also needed after each high-power operation.

[0049] No hydrogen load scenario: This includes only electrical and thermal loads, used to compare and analyze the economic differences in hydrogen energy equipment configuration.

[0050] Reference Figure 2 In this embodiment, the system integrates energy production, transmission, consumption, and storage, comprising four core components: energy production, conversion devices, energy storage equipment, and end-user consumption. The system utilizes renewable energy for power generation, achieves efficient utilization through multi-form energy conversion, regulates supply and demand balance through energy storage, and constructs an energy conversion matrix using multi-energy flow coupling devices. This system achieves complementary and synergistic multi-energy integration of electricity, heat, and hydrogen, effectively matching the diverse energy needs of end-users.

[0051] The specific energy flow process is as follows: wind and solar power units supply electricity to the electrical load, and to meet the hydrogen load, hydrogen energy coupling units operate simultaneously. Through the coupling conversion between electricity and hydrogen, the electrolyzer converts some of the electrical energy into hydrogen. After meeting the hydrogen load supply, excess hydrogen is stored in a hydrogen storage tank for later use. When there is electrical redundancy generated by wind and solar power units, both the battery unit and the electrolyzer can absorb the electricity. When the power generation of wind and solar power units does not fully meet the energy demand of the electrical load, hydrogen fuel cells can use excess hydrogen to generate electricity, and gas turbines can also use gas purchased from the gas grid or converted from hydrogen to gas by the methanation unit, as well as natural gas in the storage tank for combined heat and power (CHP). Both hydrogen fuel cells and gas turbines are gas-to-electricity-heat devices, and the heat generated during their operation is stored or sold during peak heat load periods after meeting the heat load.

[0052] Based on the needs of energy production, transmission, consumption, and storage, a comprehensive mathematical model of the energy system is established, including energy production, conversion devices, energy storage equipment, and end-use consumption. This includes models for wind and solar power generation equipment, hydrogen energy conversion equipment (electrolyzer, hydrogen fuel cell, methane reactor), energy storage equipment (hydrogen storage tank, battery, thermal storage tank, gas storage device), and a traditional combined heat and power (CHP) unit, clearly defining the power conversion relationships and constraints of each device.

[0053] Reference Figure 3In another embodiment, the integrated energy system optimization employs a two-level programming method. The upper-level model, based on multi-energy load demand, determines the optimal installed capacity combination of various equipment within the regional energy system. It focuses on multi-energy flow balance constraints and energy coupling relationships, while also evaluating the total lifecycle cost; equipment capacity is positively correlated with total cost. The lower-level model, based on the capacity parameters determined by the upper level, performs dynamic optimization scheduling with the goals of typical daily energy trading and operational stability. It calculates equipment output hourly, feeds back the operating status to the upper level, evaluates the economic and reliability indicators of the capacity configuration scheme, and achieves iterative updates to the configuration scheme. When convergence conditions are met, a Pareto-like front curve is constructed, and the TOPSIS multi-criteria decision-making method is used to determine the optimal scheme.

[0054] A two-layer capacity optimization configuration model is constructed: the upper layer optimization aims to minimize the total life cycle cost of equipment, and the decision variables are the capacity of wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, and cogeneration units; the lower layer optimization aims to minimize the operating energy purchase and sale cost and optimize system stability. Based on the equipment capacity given in the upper layer, the model simulates the 24-hour operation scheduling of a typical day and outputs the operating cost and stability indicators. The economic objective of the upper-level optimization model is mainly the total lifecycle cost of its main equipment, specifically including the initial investment cost C of the system. d Operation and maintenance costs and failure costs C O&M And equipment disposal costs C s The smaller the objective function, the better the system's economy. The expression for the economic objective function of the upper-level model is: In the formula, Total lifecycle cost of the equipment; , , , , , , These are the average daily costs over the entire lifecycle of wind power, photovoltaic power, electrolyzers, hydrogen storage tanks, hydrogen fuel cells, methane reactors, and combined heat and power equipment. In the calculation, failure costs and operation and maintenance costs are combined into operation and maintenance costs. The rated capacity of each device; Construction cost per unit capacity for each piece of equipment; The cost of operation and maintenance per unit capacity; The disposal cost per unit capacity; denoted as the equipment's service life; r is the annual interest rate, taken as 0.03.

[0055] The lower-level optimization aims to minimize the cost of purchasing and selling energy and optimize system stability. The economic objective of the lower-level optimization model during system operation primarily involves the costs of purchasing, selling, and discarding energy, specifically including the system's energy purchase cost, energy sale cost, and energy discard cost. The expression for the economic objective function of the lower-level scheduling model is: In the formula, The total cost of purchasing and selling abandoned energy for system scheduling; Energy purchase costs for system scheduling; For system scheduling energy sales costs; The cost of energy curtailment during system scheduling; / These are the costs of purchasing electricity and gas, respectively. / Let t represent the amount of electricity and gas purchased. / The unit price for purchasing electricity and gas; / These are the costs of selling heat and electricity, respectively. The amount of heat sold at time t; This refers to the unit price for hot water; / / / The on-grid electricity price for electricity sold by different devices; / The costs of abandoning electricity and hydrogen, respectively; / These represent the maximum power output that the wind power and photovoltaic installations can generate at time t, respectively. / , respectively, represent the amount of wind and solar power that can be absorbed during the scheduling at time t; ε is the penalty factor for wind and solar power curtailment; The amount of hydrogen discarded at time t; The unit price for purchasing and selling hydrogen.

[0056] System stability is also a crucial indicator in system scheduling. In this invention, stability is represented by grid loss Pw, and the expression for the stability objective function of the lower-level scheduling model is: In the formula This refers to the power grid loss per hour; This is the stability characterization function.

[0057] The upper-level optimization model aims to determine the optimal capacity configuration of the system equipment. The decision variables primarily include the number and scale of power generation and hydrogen storage system units. Specifically, the number of renewable energy units is limited to 100, the upper limit of the capacity of each energy conversion device is 1000kW, and the upper limit of the hydrogen storage tank capacity is 2000kW. The lower-level optimization model aims to optimize the typical daily operating strategy. The decision variables are the equipment output levels at different times. The basic data required for the model include: purchased electricity and gas quantities, energy market price parameters, and equipment technical specifications. Load characteristic data is derived from statistical analysis of historical operating data. The lower-level optimization scheduling uses the GUROBI commercial solver, suitable for large-scale mixed integer programming. The YALMIP toolbox is used for modeling to reduce the difficulty of GUROBI modeling and ensure accurate solutions.

[0058] Reference Figure 4 In another embodiment, the optimization is based on the NSGA-II algorithm. NSGA-II is a multi-objective optimization algorithm based on Pareto optimal solutions. Its optimization process includes: fast non-dominated sorting, which prioritizes retaining Pareto optimal solutions in the population sorting; crowding calculation, which selects uniformly distributed solutions among individuals with the same dominance front; and genetic operations including selection operators, crossover mutation, and elite retention strategies.

[0059] The specific steps of the NSGA-II optimization algorithm are as follows: (1) Fast Non-Dominated Sort. The NSGA-II algorithm uses a fast non-dominated sort mechanism to classify individuals in the population. This process first calculates the dominance degree nx of each individual x (i.e., the number of individuals that dominate the solution), and then divides the population into several levels according to the dominance relationship: the first level consists of individuals with Pareto front nx=0, the second level includes solutions that are dominated by the first level but not dominated by each other, and so on, until all individuals are classified.

[0060] (2) Crowding Calculation. The distance between individuals during evolution reflects their distribution density in the solution space. Using the crowding index can reduce the impact of human intervention and achieve objective comparison among individuals in the population. This mechanism ensures that the Pareto optimal solution set is uniformly distributed in the target space, thereby maintaining the diversity of optimal solution selection.

[0061] (3) Tournament Decision-Making. When selecting parent individuals to participate in crossover / mutation, this paper uses a binary tournament method. Based on the individual's fitness function value, two individuals are randomly selected to compare their fitness, and the individual with the best fitness is retained. This process is repeated iteratively until the number of selected individuals matches the population size.

[0062] (4) Crossover and mutation.

[0063] 1) Simulated binary crossover: Offspring are generated among individuals with real-number encodings to maintain solution diversity. Two individuals are selected from any iterative population and designated as... and The j-th gene location of its individual is used and This indicates that new individuals are generated through crossover. and The calculation formula is as follows: The crossover probability density function is given by the following equation: 2) Polynomial Mutation: This involves introducing random perturbations to avoid local optima. A mutation rate parameter is added, and individuals generated after crossover undergo a gene mutation to obtain new individuals in the population. Let the individual... Represented as an individual The j-th gene, , These are the upper and lower bounds, respectively. From which mutated individuals are obtained... The process is as follows: randomly generate a random number. ∈(0,1), calculate ,individual ,judge Check if it exceeds the boundary; if it does, take the boundary value.

[0064] In the formula, It is a self-defined non-negative real number called the variation distribution index. The size of the value affects the degree of variation; the larger the value, the smaller the difference between the mutated value and the original value.

[0065] (5) Elite Retention Strategy In each generation, individuals from the parent and child generations are merged. Through non-dominated sorting and crowding calculation, individuals with high fitness are selected to generate a new population, completing the iterative optimization process. This method ensures that excellent individuals are not lost and improves the convergence speed and solution quality of the algorithm.

[0066] TOPSIS-based Scheme Decision Making. The solution obtained by the aforementioned two-level optimization algorithm is a Pareto-like solution set composed of multiple non-dominated solutions, each of which can be considered optimal. Therefore, in practical construction scenarios, investors need to select the solution that best meets their actual needs from all the obtained non-dominated solutions as the final equipment configuration scheme, which leads to a multi-objective decision problem. This invention uses the TOPSIS method to determine the configuration scheme, which selects the best option based on its proximity to the positive and negative ideal solutions. It is based on raw data analysis, with no limit on the number of decision objectives and sample size, and can objectively evaluate candidate schemes to ultimately obtain the optimal planning scheme.

[0067] Table 1 Optimal Population Size Configuration Results <![CDATA[ V WT (Taiwan) <![CDATA[ V PV (Taiwan) <![CDATA[ V EL (kW)]]> <![CDATA[ V HFC (kW)]]> <![CDATA[ V SH (kW)]]> <![CDATA[ V MR (kW)]]> <![CDATA[ V CHP (kW)]]> LCC (yuan) COST (RMB) Scenario 1 8.9 3.8 337 388 1204 510 503 5136 -2705 Scenario 2 6.8 4.3 271 414 1309 310 516 4913 -2226 Scenario 3 6.4 4.0 372 284 1309 519 278 4858 -3287 Scenario 4 7.8 3.0 330 263 1433 300 245 4628 -2850 This embodiment utilizes the NSGA-II multi-objective optimization algorithm. Based on the aforementioned system parameters and boundary conditions, the system capacity configuration solution is obtained using MATLAB software. The system has 7 decision variables and 3 objective functions. To balance offspring diversity and the accuracy of the optimization results, the initial population size is set to 100 and the number of iterations to 200. Simulations were performed on four different hydrogen load scenarios, yielding a total of 4*100 non-dominated optimal solutions.

[0068] The optimal population for scheme decision-making using the TOPSIS method yielded the optimal capacities of electrolyzers, hydrogen fuel cells, methane reactors, cogeneration units, and hydrogen storage tanks under four scenarios, as well as the optimal number of wind and solar power units, as shown in Table 1. According to the examples, with constant electrical and thermal loads, the smaller the hydrogen load, the smaller the capacity of the hydrogen fuel cells and methane reactors, and the configuration of hydrogen fuel cells and methane reactors is a dynamic, inverse relationship. Since the system only has one power-consuming end besides the electrical load—the electrolyzer—the capacity of the electrolyzer does not change significantly. As the hydrogen load decreases, to flexibly utilize the characteristics of hydrogen energy as a flexible source load, the capacity of the low-pressure hydrogen storage tank will increase, and the capacity configuration of the traditional cogeneration system will relatively decrease.

[0069] The average daily lifecycle cost of the equipment is between 4,500 and 5,200 yuan, indicating a relatively balanced capacity configuration. If a larger hydrogen load demand needs to be met, the equipment configuration cost will increase. Regarding the purchase and sale of waste energy, to meet the pulse-like fluctuations of hydrogen load in Scenario 3, the configuration will be optimized for maximum absorption in a pulse-like manner. Therefore, in areas with stable loads, more energy can be sold as waste energy, balancing more of the cost. With reasonable scheduling, the cost per typical day is between 1,600 and 2,700 yuan, which is moderate, and the equipment exhibits strong operational stability.

[0070] In summary, this invention proposes a two-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system. First, based on the temporal characteristics of hydrogen load in industrial parks, a multi-scenario model is constructed, encompassing strong, weak, intermittent, and no hydrogen load scenarios. Then, a comprehensive energy system mathematical model is established, covering wind and solar power generation, hydrogen energy conversion, various types of energy storage, and combined heat and power (CHP) units. Subsequently, a two-layer capacity optimization configuration model is constructed, with the upper layer aiming to minimize the entire lifecycle cost and the lower layer aiming to minimize operating costs and optimize system stability. Finally, a non-dominated sorting genetic algorithm with an elitist strategy is used to solve the Pareto front, and a sorting method for approximating ideal solutions is used for decision-making to select the capacity configuration scheme with optimal overall performance. This invention effectively solves the problem that traditional configuration schemes are difficult to adapt to hydrogen load fluctuations in multiple scenarios, significantly improving the system's economy, stability, and renewable energy absorption capacity.

[0071] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A two-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system, characterized in that, Includes the following steps: S1: Construct a multi-scenario hydrogen load model: Based on the specific physical application scenarios of hydrogen load in industrial parks in industrial production and transportation, four typical scenarios are set up: strong hydrogen load, weak hydrogen load, intermittent hydrogen load and no hydrogen load. The hydrogen load demand curve is quantitatively determined based on the time-series characteristics of hydrogen load in each scenario. The time-series characteristics include bimodal curves, pulse fluctuations and nighttime peaks. S2: Establish a comprehensive energy system mathematical model: Construct a comprehensive energy system architecture including wind and solar power generation equipment models, hydrogen energy conversion equipment models, multi-type energy storage equipment models, and traditional cogeneration unit models, clarify the power conversion relationships and operating constraints between each piece of equipment. The hydrogen energy conversion equipment model includes an electric hydrogen production process, a hydrogen-to-electricity and heat production process, and a hydrogen-to-methane production process. The energy storage equipment model includes a hydrogen storage process, electric energy storage, thermal energy storage, and gas energy storage. S3: Construct a two-layer capacity optimization configuration model: Establish an optimization configuration model with upper and lower layer feedback structures. The upper layer optimization takes minimizing the total life cycle cost of equipment as the single objective, and the decision variables are the installed capacity of wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, methanation units, and cogeneration units. The lower layer optimization takes minimizing the operating energy purchase and sale cost and optimizing system stability as multiple objectives. Based on the equipment capacity configuration scheme given by the upper layer, simulate the typical 24-hour operation scheduling process and output quantitative indicators of operating cost and stability. S4: Optimization and Decision-Making: A non-dominated sorting genetic algorithm with an elite strategy is used to solve the upper-level multi-objective optimization problem, generating a Pareto front solution set that represents the balance between economy and stability. Then, the approximation ideal solution sorting method is used to make multi-objective decisions on the solutions in the Pareto front solution set, and the capacity configuration scheme with the best overall performance is selected.

2. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: Step S1 specifically includes: The strong hydrogen load corresponds to the ammonia production process, which has a typical bimodal curve characteristic. During the nighttime off-peak period, it operates at the minimum safe operating load, and the load transition is smooth and in line with the actual equipment adjustment capacity. The weak hydrogen load corresponds to the steel chemical industry scenario, including annealing furnace protective gas and continuous casting cooling applications. Compared with the ammonia production process, the total load is significantly reduced and exhibits pulse-like fluctuation characteristics, reflecting the intermittent high demand of the application scenario, while maintaining a bi-peak structure, with the most basic hydrogen consumption maintained during the nighttime valley period. The intermittent hydrogen load corresponds to the application of hydrogen refueling stations, which have obvious intermittent peak and trough periods of hydrogen load. At night, it exhibits explosive load characteristics to centrally handle the hydrogen refueling needs of logistics vehicles and buses, while during the day, it manifests as fragmented needs for short-term, low-volume refueling by private cars. The equipment operation must follow the forced cooling protection mechanism after high-power operation.

3. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: Step S2 specifically includes: Construct models for wind and solar power generation equipment, namely, mathematical models of wind turbine power and mathematical models of photovoltaic power; The mathematical model for wind turbine power is as follows: In the formula: for Wind speed at any given moment; In a given Output power at any given time (kW); Pr is the rated output power; To cut in wind speed; Rated wind speed; To cut off the wind speed; The mathematical model for photovoltaic generator power is as follows: In the formula: For photovoltaic power stations The actual output power at any given time; Let be the solar radiation intensity received by the photovoltaic cell area at time t; and let A be the photovoltaic cell area. The nominal efficiency is given under standard test conditions; β is the temperature coefficient. For photovoltaic cells in The operating temperature at any given time; Th represents the ambient temperature. The hydrogen energy conversion equipment model includes an electro-hydrogen production process, a hydrogen-to-electricity and heat production process, and a hydrogen-to-methane production process; Hydrogen production via electrolysis: In the formula, Let be the hydrogen production power at time t; Let t be the electrical power input to the electrolytic cell; The energy conversion efficiency of the electrolyzer for hydrogen production; This refers to the amount of hydrogen that can be produced per kilowatt-hour of electricity in the electrolyzer. This represents the upper limit of the electrical power of the electrolytic cell; The rated capacity of the electrolytic cell is used to limit its ramp rate. Hydrogen generation electrothermal process: In the formula, Let be the hydrogen power consumed by the hydrogen fuel cell at time t; denoted as η, where η is the energy conversion efficiency of the hydrogen fuel cell from hydrogen to electricity; r is the amount of hydrogen consumed by the hydrogen fuel cell to generate one kilowatt-hour of electricity. / Let t be the electrical power and thermal power output of the hydrogen fuel cell at time t, and let the adjustable electrothermal ratio be controlled in a coordinated manner. This represents the upper limit of hydrogen consumption power in a hydrogen fuel cell; The rated capacity of the hydrogen fuel cell is used to limit its ramp rate. / These are logical variables representing the start-up and shutdown states of the electrolyzer and the hydrogen fuel cell device at time t. =1 indicates that the cell is producing hydrogen in the electrolyzer, and the opposite indicates that the cell is consuming hydrogen in the fuel cell. If both are 0, it means that neither the electrolyzer nor the fuel cell is in operation. Hydrogen to methane production process: In the formula, Let be the hydrogen power consumed by the hydrogen fuel cell at time t; denoted as η, where η is the energy conversion efficiency of the hydrogen fuel cell from hydrogen to electricity; r is the amount of hydrogen consumed by the hydrogen fuel cell to generate one kilowatt-hour of electricity. / Let t be the electrical power and thermal power output of the hydrogen fuel cell at time t, and let the adjustable electrothermal ratio be controlled in a coordinated manner. This represents the upper limit of hydrogen consumption power in a hydrogen fuel cell; The rated capacity of the hydrogen fuel cell is used to limit its ramp rate. / These are logical variables representing the start-up and shutdown states of the electrolyzer and the hydrogen fuel cell device at time t. =1 indicates that the cell is producing hydrogen in the electrolyzer, and the opposite indicates that the cell is consuming hydrogen in the fuel cell. If both are 0, it means that neither the electrolyzer nor the fuel cell is in operation. The models of various types of energy storage devices include hydrogen storage, electrical energy storage, thermal energy storage, and gas energy storage; In the hydrogen storage stage, high-pressure hydrogen storage tanks and low-pressure hydrogen storage tanks are installed: In the formula, / / / The hydrogen filling and discharging efficiencies at time t are the low-pressure hydrogen storage tank and the high-pressure hydrogen storage tank, respectively. / / / These represent the maximum hydrogen charging / discharging efficiencies for SH and HSH, respectively. / / / These are the logical variables for hydrogen charging and discharging at time t. A value of 1 indicates that the low-pressure hydrogen storage tank is in hydrogen storage mode; otherwise, it is in hydrogen release mode. =1 indicates that the high-pressure hydrogen storage tank is in the hydrogen storage state, and vice versa; SH(t) is the hydrogen storage state in the low-pressure hydrogen storage tank at time t; HSH(t) is the hydrogen storage state in the high-pressure hydrogen storage tank at time t. / / / These are the hydrogen filling and discharging efficiency coefficients of the hydrogen storage tank; / This refers to the rated capacity of the hydrogen storage tank. / The hydrogen storage ratio in the hydrogen storage tank at time t; / These represent the upper and lower limits of the hydrogen storage ratio in low-pressure hydrogen storage tanks. / These are the upper and lower limits for the hydrogen storage ratio in high-pressure hydrogen storage tanks; Electric energy storage: In the formula, Let t be the energy storage capacity at time t; τ be the self-discharge rate of the energy storage. / These represent the charging power and discharging power of the energy storage at time t, respectively. / These are the charging efficiency and discharging efficiency of energy storage, respectively. The rated capacity is the battery's rated capacity; the SOC is the instantaneous indicator of the battery's current remaining power. / These are the lower and upper limits for the charging state, respectively; / These are the logical variables representing the charging and discharging of energy storage at time t. =1 indicates that it is in the charging state, and the opposite indicates that it is in the discharging state; This represents the upper limit of charging and discharging power. / , respectively, represent the charging and discharging power of the battery at time t; Thermal energy storage: In the formula, SQ(t) is the thermal energy storage capacity at time t; This refers to the heat loss rate; / These represent the heat storage and heat release power at time t, respectively. / These are the heat storage efficiency and the heat release efficiency, respectively. This refers to the rated capacity of the thermal storage tank. This is an instantaneous indicator of the current remaining heat in the thermal storage tank. / These are the lower and upper limits of the thermal storage state, respectively; / Let the logical variables characterize energy storage and heat release at time t, respectively. =1 indicates that it is in the state of heat storage, and the opposite indicates that it is in the state of heat release. This represents the upper limit of the heat storage and release power. / , respectively, represent the heat storage and release power of the thermal storage tank at time t; Gas storage: Let be the gas storage capacity at time t; This refers to the gas consumption loss rate; / These represent the gas storage and gas release power of the energy storage system at time t, respectively. / These are the gas storage and venting efficiencies, respectively. This refers to the rated capacity of the gas storage tank; This is an instantaneous indicator of the current remaining gas volume in the gas storage tank; / These are the lower and upper limits of the gas storage state, respectively; / These are the logical variables representing energy storage and gas release at time t. =1 indicates that it is in the gas storage state, and the opposite indicates that it is in the gas release state. This is the upper limit of the gas storage and release capacity; / These represent the gas storage and release power of the gas storage tank at time t; Construct a model of a traditional combined heat and power unit using natural gas as feedstock; In the formula, Let t be the natural gas consumed by the combined heat and power unit; Let t be the amount of natural gas the system purchases from the natural gas network at time t; This is the upper limit of the gas consumption power of a combined heat and power (CHP) unit; To improve the energy conversion efficiency of gas-to-electricity conversion in combined heat and power (CHP) units; The amount of natural gas consumed to generate 1 kWh of electricity for a combined heat and power (CHP) unit; / The figures represent the power generation and heat generation efficiencies of the combined heat and power (CHP) unit at time t, respectively, and the dynamic power-to-heat ratio is used during operation; the waste heat recovery heating efficiency is 95%. The rated capacity of the combined heat and power (CHP) unit is used to limit its ramp rate.

4. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: Step S3 specifically includes: Combining a mathematical model of an integrated energy system, a two-layer capacity optimization configuration model is constructed: the upper-layer optimization aims to minimize the total life-cycle cost of equipment, with decision variables being the capacity of wind and solar turbines, electrolyzers, hydrogen fuel cells, hydrogen storage tanks, and combined heat and power units; the lower-layer optimization aims to minimize the cost of operating energy purchases and sales and optimize system stability. Based on the equipment capacity given in the upper layer, it simulates a typical 24-hour operation scheduling, outputting operating cost and stability indicators; including: The economic objective of the upper-level optimization model is mainly the total lifecycle cost of its main equipment, specifically including the initial investment cost C of the system. d Operation and maintenance costs and failure costs C O&M And equipment disposal costs C s The expression for the economic objective function of the upper-level model is: In the formula, Total lifecycle cost of the equipment; , , , , , , These are the average daily costs over the entire lifecycle of wind power, photovoltaic power, electrolyzers, hydrogen storage tanks, hydrogen fuel cells, methane reactors, and combined heat and power equipment. In the calculation, failure costs and operation and maintenance costs are combined into operation and maintenance costs. The rated capacity of each device; Construction cost per unit capacity for each piece of equipment; The cost of operation and maintenance per unit capacity; The disposal cost per unit capacity; The useful life of the equipment; r is the annual interest rate, taken as 0.03; The lower-level optimization aims to minimize the cost of purchasing and selling energy and optimize system stability. The economic objective of the lower-level optimization model during system operation primarily involves the costs of purchasing, selling, and discarding energy, specifically including the system's energy purchase cost, energy sale cost, and energy discard cost. The expression for the economic objective function of the lower-level scheduling model is as follows: In the formula, The total cost of purchasing and selling abandoned energy for system scheduling; Energy purchase costs for system scheduling; For system scheduling energy sales costs; The cost of energy curtailment during system scheduling; / These are the costs of purchasing electricity and gas, respectively. / Let t be the amount of electricity and gas purchased. / The unit price for purchasing electricity and gas; / These are the costs of selling heat and electricity, respectively. The amount of heat sold at time t; This refers to the unit price for hot water; / / / The on-grid electricity price for electricity sold by different devices; / The costs of abandoning electricity and hydrogen, respectively; / These represent the maximum power output that the wind power and photovoltaic installations can generate at time t, respectively. / , respectively, represent the amount of wind and solar power that can be absorbed during the scheduling at time t; ε is the penalty factor for wind and solar power curtailment; The amount of hydrogen discarded at time t; The unit price for purchasing and selling hydrogen; System stability is also one of the important indicators in system scheduling; the stability of this invention is represented by grid loss Pw, and the expression of the stability objective function of the lower-level scheduling model is: In the formula This refers to the power grid loss per hour; It is a stability characterization function; The constraints of the integrated energy system model include the supply and demand balance constraints of the four major energy sources: electricity, heat, gas, and hydrogen. Electric power balance constraints: Thermal power balance constraint: Gas power balance constraints: Hydrogen power balance constraint: 。 5. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: The NSGA-II optimization algorithm in step S4 includes the following steps: S4.1.1: Fast Non-Dominated Sort; The NSGA-II algorithm uses a fast non-dominated sorting mechanism to classify individuals in the population. This process first calculates the dominance degree nx of each individual x, and then divides the population into several levels according to the dominance relationship: the first level consists of individuals with Pareto front nx=0, the second level contains solutions that are dominated by the first level but not dominated by each other, and so on, until all individuals are classified. S4.1.2: Crowding degree calculation; the distance between individuals during evolution reflects their distribution density in the solution space; using the crowding degree index can reduce the impact of human intervention and achieve objective comparison between individuals in the population; ensure that the Pareto optimal solution set is evenly distributed in the target space, thereby maintaining the diversity of optimal solution selection; S4.1.4: Tournament decision-making; When selecting parent individuals to participate in crossover / mutation, this paper uses the binary tournament method for decision-making; based on the fitness function value of an individual, two individuals are randomly selected to compare their fitness, and the individual with the best fitness is retained; this process is repeated cyclically until the number of selected individuals is consistent with the population size. S4.1.5: Simulated binary crossover: Offspring are generated among individuals with real-number encodings to maintain solution diversity; two individuals are selected from any iterative population and designated as... and The j-th gene location of its individual is used and This means that new individuals are generated through crossover. and The calculation formula is as follows: The crossover probability density function is given by the following formula: S4.1.6: Polynomial Mutation: A random perturbation is introduced to avoid local optima. A mutation rate parameter is added, and individuals generated after crossover undergo a gene mutation to obtain new individuals in the population. Let the individual... Represented as an individual The j-th gene, , These are the upper and lower bounds, respectively; from which the mutated individuals are obtained. The process is as follows: randomly generate a random number. ∈(0,1), calculate ,individual ,judge Check if it exceeds the boundary; if it does, take the boundary value. In the formula, It is a self-defined non-negative real number called the variation distribution index; The magnitude of the value affects the degree of variation; the larger the value, the smaller the difference between the mutated value and the original value. S4.1.7: Elite Preservation Strategy: In each generation, individuals from the parent and offspring generations are merged. Through non-dominated sorting and crowding calculation, individuals with high fitness are selected to generate a new population, thus completing the population iterative optimization process.

6. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: The TOPSIS method in step S4 includes the following steps: S4.2.1: Construct a standardized matrix; For the case with n schemes and n×m indicators, the indicators are positiveized and normalized to form a standardized matrix; S4.2.2: Determine the positive ideal solution Z+ and the negative ideal solution Z-; the positive and negative ideal solutions are respectively composed of the maximum and minimum values ​​of each column in Z; S4.2.3: Calculate the distance between each solution and the positive and negative ideal solutions, and calculate the closeness. ; In the formula, The weight of the m-th indicator; 0 ≤ ≤1, The larger the value, the better the overall performance of the solution.

7. The dual-layer capacity optimization configuration method for a multi-scenario hydrogen load integrated energy system as described in claim 1, characterized in that: The optimal capacity configuration scheme ultimately output by the method includes quantitative values ​​for the number of wind turbine generators, the number of photovoltaic generators, the rated capacity of electrolyzers, the rated capacity of hydrogen fuel cells, the rated capacity of hydrogen storage tanks, the rated capacity of methanation units, and the rated capacity of cogeneration units, providing direct design basis for the actual construction of integrated energy systems in industrial parks.

Images