A monthly target guided electric hydrogen energy storage multi-scenario rolling regulation method

Abstract

This invention discloses a monthly target-guided multi-scenario rolling control method for hybrid electric-hydrogen energy storage. In the monthly aggregation optimization stage, K-medoids clustering is applied to multi-year historical source-load data to generate the total monthly aggregated source-load energy, which is used as input to solve for the monthly net hydrogen storage target, thus capturing more representative seasonal fluctuation characteristics. In the intra-month rolling scheduling stage, a residual moving block bootstrapping method is used to generate multi-scenario source-load prediction data. A multi-scenario stochastic model predictive control method considering extreme events is established, embedding the monthly net hydrogen storage target into hourly rolling optimization constraints. The scheduling strategy of the first execution segment is solved and executed in a rolling manner to address short-term uncertainties. This method fully leverages the synergistic control advantages of hybrid electric-hydrogen energy storage across time scales, effectively solving the system's short-term uncertainties and long-term supply-demand mismatch problems, improving the system's economic efficiency and robustness.

Description

Technical Field

[0001] This invention relates to the field of energy planning and operation scheduling, specifically a rolling control method for multi-scenario electric hydrogen energy storage guided by monthly targets. Background Technology

[0002] With the increasing penetration rate of renewable energy, the operation of modern energy systems is undergoing profound changes. Although renewable energy sources such as photovoltaics can effectively reduce carbon emissions, their output is characterized by significant intermittency, volatility, and randomness, which can easily lead to supply-demand mismatches in the system over short timescales, thus posing challenges to the safe and stable operation of the system.

[0003] Meanwhile, photovoltaic power output is affected by natural conditions such as solar radiation, exhibiting significant seasonal variations. Energy demand, including electrical and thermal loads, also fluctuates periodically due to temperature and seasonal factors. The coupling effect between the source and load sides at different time scales means that the integrated energy system faces not only short-term intraday balance issues but also long-term supply-demand mismatches spanning months and even seasons.

[0004] To address the aforementioned multi-timescale mismatch issues, energy storage technologies are widely used in integrated energy systems. Hydrogen energy storage, with its long storage cycle and suitability for long-term energy transfer, can be used to alleviate seasonal supply-demand mismatches; while electrical energy storage offers fast response and flexible adjustment, making it suitable for handling intraday high-frequency fluctuations. Therefore, combining hydrogen and electrical energy storage to construct a hybrid energy storage system leverages the complementary advantages of different energy storage methods across time scales.

[0005] However, existing methods for scheduling hybrid energy storage systems still have certain shortcomings. Most methods focus on optimization within a single time scale, failing to fully realize the coordinated operation between long-cycle hydrogen energy storage and short-cycle electrical energy storage, thus hindering the realization of the comprehensive regulation advantages of hybrid energy storage systems across time scales. Furthermore, existing methods are typically based on single-year or deterministic data for modeling, making it difficult to comprehensively reflect the system's seasonal supply and demand characteristics. The robustness of system scheduling also needs improvement in the face of extreme events such as sudden drops in photovoltaic output and sudden load surges. Therefore, it is necessary to propose a monthly target-guided multi-scenario rolling control method for hydrogen-electric energy storage to improve the system's economic efficiency and robustness. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by proposing a monthly target-guided rolling control method for multi-scenario electric hydrogen energy storage. By combining monthly aggregate optimization with intra-month rolling scheduling, an optimized scheduling method for hybrid energy storage systems considering extreme events is constructed, achieving cross-timescale collaborative control and improving the economic efficiency and robustness of system operation.

[0007] To achieve the above objectives, the technical solution specifically adopted by the present invention is as follows:

[0008] A monthly target-guided multi-scenario rolling control method for electric hydrogen energy storage, preferably comprising a monthly aggregation optimization stage and an intra-month rolling scheduling stage, with the following specific steps:

[0009] Step S100, Monthly Aggregation Optimization Stage: Obtain historical source-load data of the electricity-heat-hydrogen integrated energy system over many years and divide it by month. Use the K-medoids clustering algorithm to extract typical daily scenarios and corresponding weights for each month. Sum the energy of each time period within the typical day and aggregate it according to the weights to calculate the total monthly aggregated source-load energy for each month. Using the total monthly aggregated source-load energy as input, construct a monthly aggregation optimization model with the goal of minimizing annual operating costs. Solve the model under the conditions of energy supply and demand balance and equipment operation constraints to output the monthly net hydrogen storage target for each month.

[0010] Step S200, Intra-monthly Rolling Scheduling Phase: For the current month, obtain baseline source-load data, and generate disturbance scenarios using the residual moving block bootstrap method based on historical prediction residuals to construct multi-scenario source-load prediction data that includes baseline predictions and extreme event scenarios; use a multi-scenario stochastic model prediction control method for intra-monthly rolling scheduling. In each round of rolling scheduling, multiple source-load scenarios including baseline predictions and extreme events are comprehensively considered, and the monthly net hydrogen storage target is embedded as a constraint in the rolling window; after each optimization solution, only the first execution segment strategy in the optimal scheduling sequence is executed, and then the system state and the remaining monthly net hydrogen storage target are updated, and the rolling window is pushed forward to repeat the above optimization process until the end of the current month.

[0011] Preferably, the monthly aggregated source-load total energy includes the monthly aggregated photovoltaic energy, the monthly aggregated electrical energy, the monthly aggregated thermal energy, and the monthly aggregated hydrogen energy.

[0012] Preferably, in step S100, the method for calculating the total monthly aggregated energy of source and load is as follows: using the month as an index, first obtain the photovoltaic power generation and electrical load power of each typical day of the corresponding month at each time, and calculate the photovoltaic daily energy and electrical load daily energy of each typical day of the corresponding month respectively; then multiply the photovoltaic daily energy and electrical load daily energy of each typical day by the proportional weight of the corresponding typical day and sum them, and then multiply by the number of days in the month to obtain the total monthly aggregated energy of photovoltaic and the total monthly aggregated energy of electrical load of the month respectively, and the sum of the proportional weights corresponding to the typical days of each month is 1; the total monthly aggregated energy of heat load and hydrogen load of each month is calculated according to the above calculation method for the total monthly aggregated energy of photovoltaic and electrical load.

[0013] Preferably, the monthly aggregate optimization model aims to minimize the annual operating cost by subtracting the system's electricity sales revenue for each month of the year from the sum of the system's energy purchase cost, operation and maintenance cost, environmental cost, and penalty cost for each month. The system energy purchase cost is calculated as the product of the total electricity purchase amount and the electricity price for the corresponding month, plus the product of the total heat purchase amount and the heat price for the corresponding month. The operation and maintenance cost is calculated as the product of the electrolyzer operation and maintenance cost coefficient and the total electricity consumption of the electrolyzer in the corresponding month, plus the product of the compressor operation and maintenance cost coefficient and the total electricity consumption of the compressor in the corresponding month, plus the fuel cell operation and maintenance cost coefficient and the fuel cell electricity generation in the corresponding month. The total cost is the sum of the product of the total amount, the product of the electric boiler operation and maintenance cost coefficient and the total electricity consumption of the electric boiler in the corresponding month, and the product of the waste heat recovery device operation and maintenance cost coefficient and the total heat recovered by the waste heat recovery device in the corresponding month; the environmental cost is the environmental cost coefficient multiplied by the product of the electricity purchase carbon emission factor and the total electricity purchase amount in the corresponding month, and the product of the heat purchase carbon emission factor and the total heat purchase amount in the corresponding month; the penalty cost is the sum of the product of the electricity curtailment penalty cost coefficient and the total electricity curtailment amount in the corresponding month, and the product of the hydrogen curtailment penalty cost coefficient and the total hydrogen curtailment amount in the corresponding month; the electricity sales revenue is the product of the electricity sales price and the total electricity sales amount in the corresponding month.

[0014] Preferably, the constraints of the monthly aggregation optimization model include electrical network constraints, thermal network constraints, and hydrogen network constraints.

[0015] Preferably, the power network constraint includes a power balance constraint, which is the sum of the total photovoltaic power generation, the total fuel cell power generation, and the total system power purchase for the corresponding month. This sum is equal to the sum of the total electrical load, the total power consumption of the electrolyzer, the total power consumption of the compressor, the total power consumption of the electric boiler, the total power sales, and the total power curtailment for the corresponding month. Furthermore, the total system power purchase does not exceed the maximum system power purchase, and each power indicator is the cumulative amount within the total hours of the corresponding month.

[0016] Preferably, the thermal network constraints include thermal energy balance constraints, electric boiler constraints, and waste heat recovery device constraints.

[0017] The heat balance constraint is the sum of the total heat purchased, the total heat generated by the electric boiler, the total heat recovered by the electrolytic cell, and the total heat recovered by the fuel cell in the corresponding month, which is equal to the sum of the total heat load and the total heat abandoned in the corresponding month, and the total heat purchased does not exceed the maximum heat purchase of the system.

[0018] The electric boiler constraint is the total heat output of the electric boiler in the corresponding month, which is equal to the product of the electric boiler's heat output efficiency and the total power consumption of the electric boiler, and the total heat output of the electric boiler does not exceed the product of the electric boiler's maximum heat output power and the total hours in the corresponding month.

[0019] The constraint of the waste heat recovery device is that the total amount of heat recovered by the waste heat recovery device in the corresponding month is equal to the heat recovery efficiency of the waste heat recovery device multiplied by the sum of the total heat generated by the electrolyzer and the total heat generated by the fuel cell, and the total amount of heat recovered by the waste heat recovery device does not exceed the product of the maximum heat recovery power of the waste heat recovery device and the total number of hours in the corresponding month.

[0020] Preferably, the hydrogen network constraints include hydrogen energy balance constraints, electrolyzer constraints, compressor constraints, fuel cell constraints, and hydrogen storage tank constraints.

[0021] The hydrogen energy balance constraint is the total amount of hydrogen produced by the electrolyzer in the corresponding month, which is equal to the sum of the total hydrogen consumption of the fuel cell, the total amount of hydrogen added to the hydrogen storage tank, the total amount of hydrogen released from the hydrogen storage tank, and the total amount of hydrogen discarded.

[0022] The constraints of the electrolyzer are: the total amount of hydrogen produced by the electrolyzer in the corresponding month, which is equal to the product of the electrolyzer's hydrogen production efficiency and the total power consumption of the electrolyzer, divided by the lower calorific value of hydrogen; and the total power consumption of the electrolyzer does not exceed the product of the electrolyzer's maximum power consumption and the total hours of the corresponding month, and is not less than the product of the electrolyzer's minimum power consumption and the total hours of the corresponding month. The total heat produced by the electrolyzer in the corresponding month is equal to 1 minus the product of the difference in the electrolyzer's hydrogen production efficiency and the total power consumption of the electrolyzer.

[0023] The compressor constraint is that the actual operating parameters of the compressor satisfy the matching relationship between the compressor efficiency and the compressor compression ratio, where the compressor compression ratio is the ratio of the hydrogen pressure after compression to the hydrogen pressure before compression.

[0024] The fuel cell constraint is that the total power output of the fuel cell in the corresponding month is equal to the product of the fuel cell power output efficiency and the total hydrogen consumption of the fuel cell, multiplied by the lower calorific value of hydrogen. The total power output of the fuel cell does not exceed the product of the maximum power output of the fuel cell and the total hours of the corresponding month. The total heat output of the fuel cell in the corresponding month is equal to 1 minus the product of the difference in fuel cell power output efficiency and the total hydrogen consumption of the fuel cell, multiplied by the lower calorific value of hydrogen.

[0025] The hydrogen storage tank constraint is the actual hydrogen storage volume of the hydrogen storage tank in the corresponding month, which is equal to the initial hydrogen storage volume of the hydrogen storage tank plus the monthly net hydrogen storage target. When all months of scheduling are completed, the actual hydrogen storage volume of the hydrogen storage tank at the end of the year is equal to the initial state hydrogen storage volume of the hydrogen storage tank at the beginning of the year.

[0026] Preferably, in step S200, the method for constructing multi-scenario source load prediction data is as follows: using the scenario as an index, the source load prediction data of each scenario is the sum of the baseline prediction data and the perturbation sequence. The perturbation sequence is generated based on the historical residual set using the residual moving block bootstrap method. During the generation process, a fixed block length is set to perform block sampling on the historical residual set.

[0027] Preferably, the rolling scheduling phase within the month aims to minimize the operating cost of the rolling window, which means minimizing the sum of the products of the probability of each scenario and the operating cost of the corresponding scenario within the rolling window time series.

[0028] Preferably, the constraints of the monthly rolling scheduling phase are hourly time scale constraints, including power balance constraints, power storage state constraints and hydrogen storage state constraints.

[0029] The power balance constraint is the sum of photovoltaic power generation, fuel cell power generation, and system power purchase at a certain moment in each scenario, which is equal to the sum of the differences between the power load, power consumption of the electrolyzer, power consumption of the compressor, power consumption of the electric boiler, power sales, power abandonment, power charging of the electric storage, and power discharging of the electric storage at that moment.

[0030] The energy storage state constraint is the state of charge (SBC) of the energy storage system at a certain moment under each scenario. It is equal to the product of the SBC and the system's self-loss efficiency at the previous moment, plus the product of the energy storage charging power and charging efficiency at the current moment, minus the ratio of the energy storage discharging power and discharging efficiency at the current moment. The SBC is between the maximum and minimum values. The energy storage charging power and discharging power do not exceed their respective maximum power, and the binary variables of the charging state and discharging state are not both 1. The SBC of the energy storage system is matched with the battery capacity.

[0031] The hydrogen storage state constraint is the actual hydrogen storage capacity of the hydrogen storage tank at a certain moment in each scenario, which is equal to the actual hydrogen storage capacity of the hydrogen storage tank at the previous moment, plus the hydrogen charging rate at this moment, and minus the hydrogen discharging rate at this moment. The actual hydrogen storage capacity of the hydrogen storage tank is between the maximum and minimum hydrogen storage values. The hydrogen charging rate and the hydrogen discharging rate do not exceed their respective maximum rates, and the binary variables of the hydrogen charging state and the hydrogen discharging state are not both 1.

[0032] Preferably, the rolling solution and state update in the intra-month rolling scheduling phase includes the following steps:

[0033] Step 1, calculate the target hydrogen storage capacity of the window: multiply the remaining net hydrogen storage target capacity of the current month by the ratio of the current rolling window step size to the remaining total time step size of the month to obtain the target hydrogen storage capacity allocated to the current rolling window. The remaining net hydrogen storage target capacity of the month at the time of the first optimization is the net hydrogen storage target of the month output in step S100.

[0034] Step 2, construct target tracking and consistency solution: Set tracking constraints so that the expected value of the hydrogen storage change in each scenario within the prediction window is equal to the hydrogen storage change corresponding to the target hydrogen storage in the window; during the solution process, the scheduling control variables of each scenario are forced to remain consistent in the first execution segment of the rolling window. The scheduling control variables include the charging and discharging power of electric energy storage, the charging and discharging rate of hydrogen energy storage, and related secondary variables.

[0035] Step 3, Closed-loop state update: After the first execution segment is completed, the actual state of charge of the electrical storage and the hydrogen storage capacity of the hydrogen tank are used as the initial boundary conditions for the next rolling cycle; at the same time, the remaining net hydrogen storage target for the current month is updated according to the actual change in hydrogen storage, the rolling window is pushed forward and the above optimization process is repeated until the scheduling of the current month ends. The initial value of the electrical storage state of charge is matched with the step size of the rolling window execution segment, and the initial value of the hydrogen storage tank is the hydrogen storage capacity after the previous round of rolling optimization.

[0036] This invention has the following characteristics and beneficial effects:

[0037] This invention overcomes the limitations of traditional methods that separate long-term energy storage planning from short-term operation scheduling by designing two tightly coupled stages: "monthly aggregation optimization" and "intra-monthly rolling scheduling." At the monthly level, a monthly net hydrogen storage target is set with the goal of minimizing annual operating costs, guiding hydrogen storage to store hydrogen in surplus months and release hydrogen in shortage months, thus achieving cross-seasonal energy transfer (see [link]). Figure 6 (Annual hydrogen storage scheduling results); at the day-ahead hourly level, multi-scenario stochastic model predictive control is used to mitigate intraday source-load fluctuations by leveraging the rapid response capability of electrical energy storage. Monthly targets serve as constraints for rolling optimization, dynamically decomposing long-term macro planning into each short-term scheduling window, thus organically synergizing the "seasonal shift" advantage of hydrogen energy storage with the "volatility mitigation" advantage of electrical energy storage. Figure 5 and Figure 7 The energy balance results on monthly and intraday scales are presented respectively, verifying that the proposed method can simultaneously ensure the supply and demand balance of the system on both long-term and short-term time scales.

[0038] Existing methods often model based on data from a single year or typical days, making it difficult to comprehensively reflect the fluctuations in source load due to seasonal variations and annual randomness. This invention innovatively employs the K-medoids clustering algorithm to process multi-year historical source load data during the monthly aggregation optimization stage, extracting representative typical day scenarios and weights for each month, and then calculating the total monthly aggregated source load energy. This process effectively captures seasonal fluctuations such as "high in summer and low in winter" photovoltaic output and "high in winter and low in summer" electric heating load, making the established monthly net hydrogen storage target more closely aligned with the actual long-term supply and demand situation. Based on this, optimization is performed with the goal of minimizing annual operating costs, avoiding energy misallocation or waste caused by insufficient representation of seasonal characteristics, and significantly improving the system's economic efficiency throughout its entire lifecycle.

[0039] To address the issues of large prediction errors and frequent extreme events in daytime scheduling, this invention introduces a multi-scenario generation technology based on the residual moving block bootstrapping method during the monthly rolling scheduling phase. This technology utilizes historical prediction residuals to construct disturbance sequences, enabling the generation of source load prediction data that includes baseline predictions and various extreme fluctuation scenarios (such as...). Figure 4 As shown, this approach overcomes the shortcomings of traditional deterministic prediction or simple scenario generation methods in terms of insufficient coverage of extreme cases. Based on this, a multi-scenario stochastic model predictive control method is adopted, incorporating the probability-weighted cost of each scenario into the optimization objective and setting consistency constraints to ensure that the scheduling strategy of each scenario remains consistent in the first execution segment. When prediction deviations or sudden extreme events occur during actual operation, the system can dynamically adjust subsequent scheduling by rolling the solution, executing only the first segment of the strategy, and updating the state and target margin in real time (closed-loop state update), thus avoiding system imbalance under extreme conditions due to reliance on a single scenario. Figure 7 This demonstrates that, under extreme disturbance scenarios, the energy storage system can quickly respond to charging and discharging commands and maintain short-term power balance, proving that the proposed method has strong robustness and anti-interference capabilities.

[0040] The objective function of this invention explicitly incorporates carbon emission costs (carbon emission factors for purchasing electricity and heat) and penalties for wasting electricity and hydrogen. This proactively reduces the system's carbon emission level and minimizes the waste of renewable energy during optimized scheduling. Simultaneously, by recovering heat from the electrolyzer and fuel cell through a waste heat recovery device to meet heat load requirements, the cascade utilization of electricity, heat, and hydrogen energy flows is achieved, further improving overall energy efficiency. Furthermore, the hydrogen storage tank state constraints and target tracking mechanism set in the rolling scheduling allow the hydrogen storage charging and discharging strategy to conform to monthly macro-level targets while also being flexibly adjusted based on daily conditions. This avoids rigid, one-size-fits-all scheduling and enhances the system's flexibility in handling complex operating conditions.

[0041] In summary, this invention, with "monthly target guidance + multi-scenario rolling optimization" as its core, comprehensively solves the problem of cross-timescale supply and demand mismatch in integrated energy systems with a high proportion of renewable energy grid connection from three dimensions: time-scale coordination, seasonal pattern mining, and robust control of extreme events. It has achieved significant technological progress in improving the system's economy, robustness, and environmental friendliness. Attached Figure Description

[0042] Figure 1 This is a flowchart of the monthly target-guided multi-scenario rolling control method for electric hydrogen energy storage proposed in this invention;

[0043] Figure 2 This is a diagram of the integrated electric-thermal-hydrogen energy system proposed in this invention;

[0044] Figure 3 This represents the total monthly aggregated source load energy for each month.

[0045] Figure 4 This includes multi-scenario source-load prediction data that includes baseline predictions and extreme events;

[0046] Figure 5 This is a graph showing the energy balance results during the monthly aggregation optimization phase.

[0047] Figure 6 This is a graph showing the annual hydrogen storage scheduling results based on the monthly net hydrogen storage target;

[0048] Figure 7 This is a diagram showing the energy balance results for one day during the monthly rolling scheduling phase.

[0049] Figure 8 This is a cost comparison chart of rolling scheduling results under extreme operating conditions. Detailed Implementation

[0050] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0051] A method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets, such as Figure 1 As shown, it includes a monthly aggregation optimization phase and an intra-month rolling scheduling phase. The specific steps are as follows:

[0052] Step S100: Monthly aggregation optimization phase.

[0053] S101. Obtain historical source-load data of the integrated electric-heat-hydrogen energy system over many years and divide it by month. Use the K-medoids clustering algorithm to extract typical daily scenes and corresponding weights for each month.

[0054] Among them, the typical day scenario is a representative time series data used to reflect the daily fluctuation characteristics of the source load in the month. The method of obtaining the typical day scenario is to pre-set the clustering parameter K for the multi-year historical source load data of each month, and then use the clustering algorithm to perform cluster analysis to divide it into K sets with similar fluctuation characteristics, that is, K typical days.

[0055] S102. Sum the energy of each time period during a typical day and aggregate it according to the weights mentioned above to calculate the total monthly aggregated source load energy for each month.

[0056] The monthly aggregated energy of the source load includes the monthly aggregated energy of photovoltaic load, the monthly aggregated energy of electrical load, the monthly aggregated energy of thermal load, and the monthly aggregated energy of hydrogen load.

[0057] The method for calculating the total monthly aggregated energy of photovoltaic power generation and electrical load is as follows: Using the month as an index, first obtain the photovoltaic power generation and electrical load power at each time point on each typical day of the corresponding month, and calculate the daily photovoltaic energy and daily electrical load energy for each typical day of the corresponding month. Then, multiply the daily photovoltaic energy and daily electrical load energy of each typical day by their respective proportional weights, sum them, and multiply by the number of days in the month to obtain the total monthly aggregated energy of photovoltaic power generation and the total monthly aggregated energy of electrical load for that month. The sum of the proportional weights corresponding to the typical days of each month is 1. The total monthly aggregated energy of heat load and hydrogen load for each month is calculated using the same method as above for calculating the total monthly aggregated energy of photovoltaic power generation and electrical load, as shown in the following expressions:

[0058]

[0059] In the formula, Indexed by month; For time step; and Months typical day exist Photovoltaic power generation and electrical load power at any given time; and Months typical day Daily energy consumption from photovoltaic and electrical loads; and Months Monthly aggregated total energy from photovoltaic and electrical loads; For months The number of days included; For months Typical daily set; For months typical day The corresponding proportional weights, and satisfying In addition, the heat load for each month and hydrogen load The calculation method for monthly aggregated total energy is the same as that for photovoltaic and electrical loads.

[0060] S103. Using the total energy of monthly polymerization source load as input, construct a monthly polymerization optimization model with the goal of minimizing annual operating costs. Solve the model under the conditions of energy supply and demand balance and equipment operation constraints, and output the monthly net hydrogen storage target for each month.

[0061] In this embodiment, the monthly aggregate optimization model aims to minimize the annual operating cost by subtracting the sum of the system's energy purchase cost, operation and maintenance cost, environmental cost, and penalty cost for each month of the year from the system's electricity sales revenue for each month. The system energy purchase cost is calculated as the product of the total electricity purchase amount and the electricity price for the corresponding month, plus the product of the total heat purchase amount and the heat price for the corresponding month. The operation and maintenance cost is calculated as the product of the electrolyzer operation and maintenance cost coefficient and the total electricity consumption of the electrolyzer for the corresponding month, plus the product of the compressor operation and maintenance cost coefficient and the total electricity consumption of the compressor for the corresponding month, plus the product of the fuel cell operation and maintenance cost coefficient and the total electricity generated by the fuel cell for the corresponding month. The calculation method is as follows: The sum of the product of the operating cost coefficient of the electric boiler and the total electricity consumption of the electric boiler in the corresponding month, and the sum of the product of the operating cost coefficient of the waste heat recovery device and the total heat recovered by the waste heat recovery device in the corresponding month; the environmental cost is the environmental cost coefficient multiplied by the product of the carbon emission factor for electricity purchase and the total electricity purchase in the corresponding month, and the sum of the product of the carbon emission factor for purchased heat and the total heat purchase in the corresponding month; the penalty cost is the product of the penalty cost coefficient for abandoned electricity and the total amount of abandoned electricity in the corresponding month, and the sum of the penalty cost coefficient for abandoned hydrogen and the total amount of abandoned hydrogen in the corresponding month; the electricity sales revenue is the product of the electricity sales price and the total amount of electricity sold in the corresponding month. The specific model is as follows:

[0062] Objective function:

[0063]

[0064] In the formula, A collection of months throughout the year; , , , and The first The monthly system's energy purchase cost, operation and maintenance cost, environmental cost, penalty cost, and electricity sales revenue; , and These are the system's electricity purchase price, heat purchase price, and electricity sales price, respectively. and The first The total monthly electricity and heat purchases by the system; , , , and These are the operation and maintenance cost coefficients for electrolyzers, compressors, fuel cells, electric boilers, and waste heat recovery devices, respectively. , , , The first Total monthly power consumption of electrolyzers, compressors, electric boilers, and total power generation of fuel cells; For the first The total amount of heat recovered by the heat recovery device over the past month; Indicates the environmental cost coefficient; and These are the carbon emission factors for purchasing electricity and heat, respectively. and These are the penalty cost coefficients for abandoned electricity and abandoned hydrogen, respectively. and The first Total amount of electricity and hydrogen wasted by the monthly system; For the first Total monthly electricity sales volume of the system.

[0065] Furthermore, constraints are introduced into the monthly aggregation optimization model, including electrical network constraints, thermal network constraints, and hydrogen network constraints.

[0066] Specifically, the electrical network constraint is the power balance constraint.

[0067] Power balance constraints:

[0068] In the formula, This represents the system's maximum power purchase capacity. For months Total hours.

[0069] Thermal network constraints include thermal balance constraints, electric boiler constraints, and waste heat recovery device constraints.

[0070] Thermal energy balance constraints:

[0071]

[0072] In the formula, This represents the total heat output of electric boilers in a given month. , These represent the total monthly heat recovery from the electrolyzer and fuel cell, respectively. For months Total amount of heat wasted; This represents the maximum heat capacity that the system can purchase.

[0073] Electric boiler constraints:

[0074]

[0075] In the formula, The heat production efficiency of the electric boiler; This is the maximum heat output of the electric boiler.

[0076] Constraints of waste heat recovery devices:

[0077]

[0078] In the formula, The heat recovery efficiency of the waste heat recovery device; , Months Total heat production from electrolyzers and fuel cells; This represents the maximum recoverable heat power of the waste heat recovery device.

[0079] Hydrogen network constraints include hydrogen energy balance constraints, electrolyzer constraints, compressor constraints, fuel cell constraints, and hydrogen storage tank constraints.

[0080] Hydrogen energy balance constraints:

[0081]

[0082] In the formula, For months Total hydrogen production from the electrolyzer; For months Total hydrogen consumption of fuel cells; , Months Total amount of hydrogen added to and released from the hydrogen storage tank.

[0083] Electrolytic cell constraints:

[0084]

[0085] In the formula, Hydrogen production efficiency of the electrolyzer; Hydrogen has a low calorific value; , These represent the maximum and minimum power consumption of the electrolytic cell, respectively.

[0086] Compressor constraints:

[0087]

[0088] In the formula, For compressor efficiency; This represents the compression ratio of the compressor.

[0089] Fuel cell constraints:

[0090]

[0091] In the formula, To improve the power generation efficiency of fuel cells; This represents the maximum power output of the fuel cell.

[0092] Hydrogen storage tank constraints:

[0093]

[0094] In the formula, For months The actual hydrogen storage capacity of the hydrogen storage tank; This represents the initial hydrogen storage capacity of the hydrogen storage tank. The target is for monthly net hydrogen storage.

[0095] Then, the monthly aggregation optimization model is solved with the goal of minimizing annual operating costs, and the monthly net hydrogen storage target for each month is output.

[0096] Step S200: Intra-month rolling scheduling phase.

[0097] S201. For the current month, obtain the baseline source load data, and generate disturbance scenarios using the residual moving block bootstrap method based on historical forecast residuals, so as to construct multi-scenario source load forecast data that includes baseline forecasts and extreme event scenarios.

[0098] It should be noted that the baseline source load data is directly obtainable normal historical source load data. Subsequent scenario-generated source load data includes predicted data referenced to historical baseline data and extreme predicted data deviating from the baseline (such as sudden PV drops, load surges, etc.). The method for constructing multi-scenario source load prediction data is as follows: using scenarios as indexes, the source load prediction data for each scenario... The sum of the baseline predicted data and the perturbation sequence is given. The perturbation sequence is generated using the residual moving block bootstrap method based on the historical residual set. During the generation process, a fixed block length is set to sample the historical residual set in blocks, as shown in the following expression:

[0099]

[0100] In the formula, For scene indexing; ; This serves as the baseline forecast data; The perturbation sequence is obtained based on the residual moving block bootstrapping method; For historical residuals; The length is the block length.

[0101] It should be noted that the baseline prediction data is based on the baseline source load data, which is the normal prediction value without considering extreme cases, and some small perturbations are added to the normal historical source load data to distinguish it from the normal historical data. Therefore, the baseline prediction data can be obtained through conventional technical means, and will not be described in detail in this embodiment.

[0102] S202. A multi-scenario stochastic model predictive control method is adopted for monthly rolling scheduling. In each round of rolling scheduling, multiple source-load scenarios including baseline predictions and extreme events are comprehensively considered, and the monthly net hydrogen storage target is embedded as a constraint condition into the rolling window.

[0103] The rolling scheduling phase within the month aims to minimize the operating cost of the rolling window. This objective is to minimize the sum of the products of the probability of each scenario and the operating cost of the corresponding scenario within the rolling window time series, as expressed below:

[0104] In the formula, For the corresponding scene index; This represents the probability of the corresponding scenario; This is the time series of the current scrolling window.

[0105] Furthermore, the monthly rolling scheduling phase is subject to constraints on an hourly timescale, and new constraints are set for power balance, electrical energy storage, and hydrogen energy storage status.

[0106] The power balance constraint is the sum of photovoltaic power generation, fuel cell power generation, and system power purchase at a certain moment in each scenario. It is equal to the sum of the differences between the electrical load power, electrolyzer power consumption, compressor power consumption, electric boiler power consumption, electricity sales power, abandoned power, and the charging and discharging power of the energy storage at that moment. The expression is as follows:

[0107] In the formula, and Scenes Down The charging and discharging power of the instantaneous energy storage system.

[0108] The state of charge (SOC) constraint for energy storage is the state of charge of the energy storage system at a certain moment under each scenario. It is equal to the product of the SOC and the system's self-loss efficiency at the previous moment, plus the product of the charging power and charging efficiency at the current moment, minus the ratio of the discharging power and discharging efficiency at the current moment. The SOC lies between its maximum and minimum values. Both the charging and discharging power of the energy storage system do not exceed their respective maximum power, and the binary variables for charging and discharging are not simultaneously equal to 1. The SOC of the energy storage system is matched with the battery capacity, as expressed below:

[0109]

[0110] In the formula, For the scene Down The state of charge of the energy storage system at any given time; For the system's self-loss efficiency; and These are the system's charging and discharging efficiencies, respectively. This refers to the capacity of the battery. and These are the maximum and minimum values ​​of the state of charge, respectively; and Scenes The two variables representing the charging and discharging states of the system; and These are the maximum values ​​of the system's charging and discharging power, respectively.

[0111] The hydrogen storage state constraint is the actual hydrogen storage capacity of the hydrogen storage tank at a certain moment under each scenario. This capacity equals the actual hydrogen storage capacity of the hydrogen storage tank at the previous moment, plus the hydrogen charging rate at this moment, and minus the hydrogen discharging rate at this moment. Furthermore, the actual hydrogen storage capacity of the hydrogen storage tank is between its maximum and minimum values. The hydrogen charging rate and hydrogen discharging rate do not exceed their respective maximum rates, and the binary variables for the hydrogen charging state and the hydrogen discharging state are not simultaneously equal to 1. The expression is as follows:

[0112]

[0113] In the formula, For the scene Down The actual hydrogen storage capacity of the hydrogen storage tank at any given time; and Scenes Down The rate of hydrogen charging and discharging at any given time; and These are the maximum and minimum values ​​for hydrogen storage, respectively. and Scenes The binary variables of the hydrogen charging and discharging states of the lower system; and These represent the maximum values ​​of the system's hydrogen charging and discharging rates, respectively.

[0114] S203. After each optimization solution, only the first execution segment strategy in the optimal scheduling sequence is executed. Then, the system state and the remaining monthly net hydrogen storage target are updated, and the rolling window is pushed forward to repeat the above optimization process until the end of the current month.

[0115] Specifically, it includes the following steps:

[0116] Step 1, Calculate the target hydrogen storage capacity for the window: Multiply the remaining net hydrogen storage target for the current month by the ratio of the current rolling window step size to the remaining total time step size for the month to obtain the target hydrogen storage capacity allocated to the current rolling window. The monthly remaining net hydrogen storage target for the first optimization is the monthly net hydrogen storage target output in step S100, and its expression is:

[0117]

[0118] In the formula, The target hydrogen storage capacity for the current rolling window; This represents the current scroll window step size; This represents the total remaining time step for the month. The monthly remaining net hydrogen storage target (satisfied during the first optimization) ).

[0119] Step 2, Constructing Target Tracking and Consistency Solution: Set tracking constraints so that the expected value of hydrogen storage change in each scenario within the prediction window is equal to the hydrogen storage change corresponding to the target hydrogen storage amount in the window; during the solution process, force the scheduling control variables of each scenario to remain consistent in the first execution segment of the rolling window. The scheduling control variables include the charging and discharging power of electric energy storage, the charging and discharging rate of hydrogen energy storage, and related secondary variables, the expression of which is:

[0120]

[0121] In the formula, This is the execution segment for the scrolling window; These are the variables that need to be controlled in each scenario (such as the charging and discharging power of electric energy storage, the charging and discharging efficiency of hydrogen energy storage, and the corresponding secondary variables).

[0122] Step 3, Closed-Loop State Update: After the first execution segment is completed, the actual state of charge of the electrical storage and the hydrogen storage capacity of the hydrogen tank are used as the initial boundary conditions for the next rolling cycle. Simultaneously, the remaining net hydrogen storage target for the current month is updated based on the actual change in hydrogen storage. The rolling window is then pushed forward, and the above optimization process is repeated until the current month's scheduling ends. The initial value of the electrical storage state of charge matches the step size of the rolling window execution segment, and the initial value of the hydrogen storage tank is the hydrogen storage capacity after the previous round of rolling optimization, expressed as:

[0123]

[0124] In the formula, Set the segment step size for the scrolling window; This represents the initial state of charge of the electrical energy storage. The initial value of the hydrogen storage tank for the current rolling optimization (i.e., the amount of hydrogen stored after the previous round of rolling optimization).

[0125] To verify the effectiveness of the monthly-daily rolling optimization scheduling method considering extreme events proposed in this invention in addressing seasonal and short-term supply-demand mismatches, a typical integrated electricity-heat-hydrogen energy system was selected for simulation testing. The relevant simulation scenarios and scheduling results are illustrated below with reference to the accompanying drawings:

[0126] Figure 2 The physical topology of the integrated energy system employed in this invention is illustrated. This system includes photovoltaic power generation units, electrothermal hydrogen multi-energy loads, and a hybrid energy storage network consisting of seasonal hydrogen energy storage and short-term electrical energy storage, comprising an electrolyzer, hydrogen storage tank, compressor, and fuel cell. This provides the physical basis for achieving coordinated scheduling across time scales.

[0127] Figure 3 The data shows the total monthly aggregated source and load energy obtained by K-medoids clustering during the monthly aggregate optimization phase, which intuitively reflects the seasonal fluctuation characteristics of photovoltaic output and various loads throughout the year based on historical data over many years. Figure 4 The study showcases multi-scenario source-load prediction data generated by the residual moving block bootstrapping method during the monthly rolling scheduling phase. This includes baseline prediction scenarios and extreme event scenarios, which serve as the scenario inputs for constructing the monthly rolling optimized scheduling.

[0128] Figure 5 The energy balance results of the monthly aggregation optimization phase are presented to illustrate the coupling and distribution relationships of various types of energy in the system on a monthly scale. Figure 6 The results of annual hydrogen storage scheduling based on monthly net hydrogen storage targets are shown. The system produces and stores hydrogen during months with surplus photovoltaic power and releases hydrogen energy during months with relative energy shortages, demonstrating that hydrogen storage can achieve cross-seasonal energy transfer and complete monthly target tracking. The annual cost of this hydrogen storage scheduling strategy is US$116,761.19, which is 2.71% lower than the cost of traditional scheduling strategies. Figure 7 The results show the energy balance on a daily scale during the monthly rolling scheduling phase under extreme disturbance scenarios. Under multi-scenario stochastic model predictive control, energy storage smooths out short-term power fluctuations through charge and discharge regulation to maintain the supply and demand balance of the system under short-cycle conditions. Figure 8 The results show the comparison of daily operating costs under extreme conditions on the selected test days. Considering the rolling scheduling strategy for multiple scenarios, the average daily operating cost is $310.51 when dealing with extreme situations, which is a 17.43% reduction compared to the cost of a single scenario.

[0129] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for multi-scenario rolling control of electro-hydrogen energy storage guided by monthly targets, characterized in that, The method includes a monthly aggregation optimization phase and an intra-month rolling scheduling phase, and the specific steps are as follows: Step S100, Monthly Aggregation Optimization Stage: Obtain multi-year historical source-load data of the integrated energy system of electricity-heat-hydrogen and divide it by month. Use K-medoids clustering algorithm to extract typical daily scenarios and corresponding weights for each month. The energy of each typical intraday period is summed and aggregated according to the weights to calculate the total monthly aggregated source load energy for each month. Using the total monthly aggregated source load energy as input, a monthly aggregation optimization model is constructed with the goal of minimizing annual operating costs. The model is solved under the conditions of energy supply and demand balance and equipment operation constraints to output the monthly net hydrogen storage target for each month. The typical daily scenario is a representative time series data used to reflect the daily fluctuation characteristics of source load within the month; Step S200, monthly rolling scheduling stage: For the current month, obtain the baseline source load data, and generate the disturbance scenario based on the historical prediction residual using the residual moving block bootstrap method, so as to construct multi-scenario source load prediction data including baseline prediction and extreme event scenarios; A multi-scenario stochastic model predictive control method is adopted for monthly rolling scheduling. In each round of rolling scheduling, multiple source-load scenarios including baseline predictions and extreme events are comprehensively considered, and the monthly net hydrogen storage target is embedded as a constraint into the rolling window. After each optimization solution, only the first execution segment strategy in the optimal scheduling sequence is executed, and then the system state and the remaining monthly net hydrogen storage target are updated. The rolling window is then pushed forward and the above process is repeated until the end of the current month.

2. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 1, characterized in that, The monthly aggregated source-load total energy includes the monthly aggregated photovoltaic energy, the monthly aggregated electrical energy, the monthly aggregated thermal energy, and the monthly aggregated hydrogen energy.

3. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 2, characterized in that, In step S100, the method for calculating the total monthly aggregated energy of source and load is as follows: using the month as an index, first obtain the photovoltaic power generation and electrical load power of each typical day of the corresponding month at each time, and calculate the photovoltaic daily energy and electrical load daily energy of each typical day of the corresponding month respectively; then multiply the photovoltaic daily energy and electrical load daily energy of each typical day by the proportional weight of the corresponding typical day and sum them, and then multiply by the number of days in the month to obtain the total monthly aggregated energy of photovoltaic and the total monthly aggregated energy of electrical load of the month respectively, and the sum of the proportional weights corresponding to the typical days of each month is 1; the total monthly aggregated energy of heat load and hydrogen load of each month is calculated according to the above calculation method for the total monthly aggregated energy of photovoltaic and electrical load.

4. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 3, characterized in that, The monthly aggregate optimization model aims to minimize the annual operating cost by subtracting the system's electricity sales revenue for each month of the year from the sum of the system's energy purchase cost, operation and maintenance cost, environmental cost, and penalty cost for each month. The system energy purchase cost is calculated as the product of the total electricity purchase amount and the electricity price for the corresponding month, plus the product of the total heat purchase amount and the heat price for the corresponding month. The operation and maintenance cost is calculated as the product of the electrolyzer operation and maintenance cost coefficient and the total electricity consumption of the electrolyzer for the corresponding month, plus the product of the compressor operation and maintenance cost coefficient and the total electricity consumption of the compressor for the corresponding month, plus the fuel cell operation and maintenance cost coefficient and the total electricity generated by the fuel cell for the corresponding month. The environmental cost is the sum of the product of the electricity consumption factor and the total electricity consumption of the electric boiler in the corresponding month, and the product of the operation and maintenance cost factor and the total heat recovered by the waste heat recovery device in the corresponding month; the environmental cost is the environmental cost factor multiplied by the product of the carbon emission factor of electricity purchase and the total electricity purchase in the corresponding month, and the product of the carbon emission factor of heat purchase and the total heat purchase in the corresponding month; the penalty cost is the sum of the product of the electricity abandonment penalty cost factor and the total electricity abandonment in the corresponding month, and the product of the hydrogen abandonment penalty cost factor and the total hydrogen abandonment in the corresponding month; the electricity sales revenue is the product of the electricity sales price and the total electricity sales in the corresponding month.

5. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 4, characterized in that, The constraints of the monthly aggregation optimization model include electrical network constraints, thermal network constraints, and hydrogen network constraints.

6. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 5, characterized in that, The power network constraints include power balance constraints, which are the sum of the total photovoltaic power generation, the total fuel cell power generation, and the total system power purchase for the corresponding month. This sum is equal to the sum of the total electrical load, the total power consumption of the electrolyzer, the total power consumption of the compressor, the total power consumption of the electric boiler, the total power sales, and the total power curtailment for the corresponding month. Furthermore, the total system power purchase does not exceed the maximum system power purchase. All power indicators are the cumulative amounts within the total hours of the corresponding month.

7. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 5, characterized in that, The thermal network constraints include thermal balance constraints, electric boiler constraints, and waste heat recovery device constraints. The heat balance constraint is the sum of the total heat purchased, the total heat generated by the electric boiler, the total heat recovered by the electrolytic cell, and the total heat recovered by the fuel cell in the corresponding month, which is equal to the sum of the total heat load and the total heat abandoned in the corresponding month, and the total heat purchased does not exceed the maximum heat purchase of the system. The electric boiler constraint is the total heat output of the electric boiler in the corresponding month, which is equal to the product of the electric boiler's heat output efficiency and the total power consumption of the electric boiler, and the total heat output of the electric boiler does not exceed the product of the electric boiler's maximum heat output power and the total hours in the corresponding month. The constraint of the waste heat recovery device is that the total amount of heat recovered by the waste heat recovery device in the corresponding month is equal to the heat recovery efficiency of the waste heat recovery device multiplied by the sum of the total heat generated by the electrolyzer and the total heat generated by the fuel cell, and the total amount of heat recovered by the waste heat recovery device does not exceed the product of the maximum heat recovery power of the waste heat recovery device and the total number of hours in the corresponding month.

8. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 5, characterized in that, The hydrogen network constraints include hydrogen energy balance constraints, electrolyzer constraints, compressor constraints, fuel cell constraints, and hydrogen storage tank constraints. The hydrogen energy balance constraint is the total amount of hydrogen produced by the electrolyzer in the corresponding month, which is equal to the sum of the total hydrogen consumption of the fuel cell, the total amount of hydrogen added to the hydrogen storage tank, the total amount of hydrogen released from the hydrogen storage tank, and the total amount of hydrogen discarded. The constraints of the electrolyzer are: the total amount of hydrogen produced by the electrolyzer in the corresponding month, which is equal to the product of the electrolyzer's hydrogen production efficiency and the total power consumption of the electrolyzer, divided by the lower calorific value of hydrogen; and the total power consumption of the electrolyzer does not exceed the product of the electrolyzer's maximum power consumption and the total hours of the corresponding month, and is not less than the product of the electrolyzer's minimum power consumption and the total hours of the corresponding month. The total heat produced by the electrolyzer in the corresponding month is equal to 1 minus the product of the difference in the electrolyzer's hydrogen production efficiency and the total power consumption of the electrolyzer. The compressor constraint is that the actual operating parameters of the compressor satisfy the matching relationship between the compressor efficiency and the compressor compression ratio, where the compressor compression ratio is the ratio of the hydrogen pressure after compression to the hydrogen pressure before compression. The fuel cell constraint is that the total power output of the fuel cell in the corresponding month is equal to the product of the fuel cell power output efficiency and the total hydrogen consumption of the fuel cell, multiplied by the lower calorific value of hydrogen. The total power output of the fuel cell does not exceed the product of the maximum power output of the fuel cell and the total hours of the corresponding month. The total heat output of the fuel cell in the corresponding month is equal to 1 minus the product of the difference in fuel cell power output efficiency and the total hydrogen consumption of the fuel cell, multiplied by the lower calorific value of hydrogen. The hydrogen storage tank constraint is the actual hydrogen storage volume of the hydrogen storage tank in the corresponding month, which is equal to the initial hydrogen storage volume of the hydrogen storage tank plus the monthly net hydrogen storage target. When all months of scheduling are completed, the actual hydrogen storage volume of the hydrogen storage tank at the end of the year is equal to the initial state hydrogen storage volume of the hydrogen storage tank at the beginning of the year.

9. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 1, characterized in that, In step S200, the method for constructing multi-scenario source load prediction data is as follows: using the scenario as an index, the source load prediction data of each scenario is the sum of the baseline prediction data and the perturbation sequence. The perturbation sequence is generated based on the historical residual set using the residual moving block bootstrap method. During the generation process, a fixed block length is set to perform block sampling on the historical residual set.

10. The method for multi-scenario rolling regulation of electric hydrogen energy storage guided by monthly targets according to claim 9, characterized in that, The rolling scheduling phase within the month aims to minimize the operating cost of the rolling window, which means minimizing the sum of the products of the probability of each scenario and the operating cost of the corresponding scenario within the rolling window time series.

11. The method for multi-scenario rolling control of electric hydrogen energy storage guided by monthly targets according to claim 10, characterized in that, The constraints of the monthly rolling scheduling phase are hourly time scale constraints, including power balance constraints, power storage state constraints, and hydrogen storage state constraints. The power balance constraint is the sum of photovoltaic power generation, fuel cell power generation, and system power purchase at a certain moment in each scenario, which is equal to the sum of the differences between the power load, power consumption of the electrolyzer, power consumption of the compressor, power consumption of the electric boiler, power sales, power abandonment, power charging of the electric storage, and power discharging of the electric storage at that moment. The energy storage state constraint is the state of charge (SBC) of the energy storage system at a certain moment under each scenario. It is equal to the product of the SBC and the system's self-loss efficiency at the previous moment, plus the product of the energy storage charging power and charging efficiency at the current moment, minus the ratio of the energy storage discharging power and discharging efficiency at the current moment. The SBC is between the maximum and minimum values. The energy storage charging power and discharging power do not exceed their respective maximum power, and the binary variables of the charging state and discharging state are not both 1. The SBC of the energy storage system is matched with the battery capacity. The hydrogen storage state constraint is the actual hydrogen storage capacity of the hydrogen storage tank at a certain moment in each scenario, which is equal to the actual hydrogen storage capacity of the hydrogen storage tank at the previous moment, plus the hydrogen charging rate at this moment, and minus the hydrogen discharging rate at this moment. The actual hydrogen storage capacity of the hydrogen storage tank is between the maximum and minimum hydrogen storage values. The hydrogen charging rate and the hydrogen discharging rate do not exceed their respective maximum rates, and the binary variables of the hydrogen charging state and the hydrogen discharging state are not both 1.

12. The method for multi-scenario rolling regulation of electro-hydrogen energy storage guided by monthly targets according to claim 11, characterized in that, The rolling solution and state update during the monthly rolling scheduling phase include the following steps: Step 1, calculate the target hydrogen storage capacity of the window: multiply the remaining net hydrogen storage target capacity of the current month by the ratio of the current rolling window step size to the remaining total time step size of the month to obtain the target hydrogen storage capacity allocated to the current rolling window. The remaining net hydrogen storage target capacity of the month at the time of the first optimization is the net hydrogen storage target of the month output in step S100. Step 2, construct target tracking and consistency solution: Set tracking constraints so that the expected value of the hydrogen storage change in each scenario within the prediction window is equal to the hydrogen storage change corresponding to the target hydrogen storage in the window; during the solution process, the scheduling control variables of each scenario are forced to remain consistent in the first execution segment of the rolling window. The scheduling control variables include the charging and discharging power of electric energy storage, the charging and discharging rate of hydrogen energy storage, and related secondary variables. Step 3, Closed-loop state update: After the first execution segment is completed, the actual state of charge of the electrical storage and the hydrogen storage capacity of the hydrogen tank are used as the initial boundary conditions for the next rolling cycle; at the same time, the remaining net hydrogen storage target for the current month is updated according to the actual change in hydrogen storage, the rolling window is pushed forward and the above optimization process is repeated until the scheduling of the current month ends. The initial value of the electrical storage state of charge is matched with the step size of the rolling window execution segment, and the initial value of the hydrogen storage tank is the hydrogen storage capacity after the previous round of rolling optimization.

Images