Multi-time scale scheduling method for hybrid hydrogen production system considering hydrogen load demand response

Abstract

The present application proposes a multi-time scale scheduling method of a hydrogen production system considering hydrogen load demand response, processes source and load prediction data of the hydrogen production system, and simultaneously establishes a hydrogen load demand response model based on price incentive response; subsequently, an electrolytic cell array shift control strategy based on Lagrange optimization is adopted to make the modular hydrogen production cluster fully exert the wide-range operation characteristics of the hydrogen production equipment; finally, a multi-time scale scheduling method of the hydrogen production system considering hydrogen load demand response is proposed, with the minimum total operation cost of the system as the target, so that different electrolytic cells complement each other in terms of economy and dynamic response characteristics, the battery provides power regulation, the super capacitor provides rapid response, and in the day-ahead scheduling, a multi-energy coordinated energy management strategy is formulated to realize reasonable allocation of system energy; the present application can further suppress the influence of source and load fluctuations and improve the stability and flexibility of the hydrogen production system.

Description

Technical Field

[0001] This invention relates to the field of multi-timescale optimization scheduling technology for new energy hydrogen production systems, and in particular to a multi-timescale scheduling method for hybrid hydrogen production systems that takes into account hydrogen load demand response. Background Technology

[0002] Currently, with the increasing penetration of new energy sources such as wind and solar power into the power grid, the fluctuation of source load has a more and more significant impact on the stable operation of the system. Renewable energy hydrogen production, as a promising green hydrogen production method, is particularly important for achieving renewable energy consumption and mitigating source load fluctuations. However, most existing hydrogen production systems use a single, large-capacity alkaline electrolyzer, leading to a difficulty in coupling economic efficiency with dynamic performance. Furthermore, large-capacity single units cannot fully utilize the wide-range characteristics of the electrolyzer, and long-term operation severely impacts the lifespan of the equipment. Therefore, how to combine the advantages of alkaline electrolyzers and proton exchange membrane electrolyzers to build a hydrogen production system that considers the coordinated operation of multiple types of electrolyzer arrays is a key challenge that urgently needs to be addressed in future green hydrogen production. Secondly, with the increasingly widespread demand for hydrogen load, the current approach of considering only the user side of hydrogen energy, which has fixed demand and cannot respond to hydrogen production prices, is no longer suitable for the source-load interaction of future hydrogen production systems. Therefore, it is necessary to fully explore the potential of hydrogen load demand response to improve the flexibility of system peak shaving, valley filling, and operation scheduling. Third, source-load data also exhibits uncertainty. Existing hydrogen production systems, which schedule data based on deterministic models, often produce results that differ significantly from actual performance. Therefore, subsequent optimization of hydrogen production systems must consider the uncertainty of source-load predictions to more accurately reflect prediction errors. Fourth, source-load volatility varies considerably across different time scales. Most existing hydrogen production systems operate on a day-ahead scale, lacking rolling optimization and correction based on real-time ultra-short-term intraday predictions. Furthermore, existing hybrid hydrogen production units, batteries, and supercapacitors lack energy scheduling strategies that consider the energy storage characteristics and state constraints at different scales. Therefore, establishing a multi-time-scale operational optimization strategy for hybrid hydrogen production systems, considering the different response characteristics of the two types of electrolyzers and the operating characteristics of different energy storage systems, is a critical challenge that urgently needs to be addressed.

[0003] Currently, researchers have conducted extensive studies on the optimal scheduling of new energy hydrogen production systems. However, existing studies rarely correct source load forecast data, mostly employing deterministic source load models, which leads to significant errors during system scheduling. Furthermore, the demand for hydrogen energy is increasing, and there is a dual constraint between energy demand and energy prices. A fixed energy price does not stimulate the potential of hydrogen demand. Subsequently, starting from the core equipment of the hydrogen production system, existing studies mostly use a single type of large-capacity unit for hydrogen production, which cannot simultaneously guarantee the economic efficiency and dynamic flexibility of the hydrogen production system. Finally, most existing research on hydrogen production systems remains at the day-ahead level. However, due to the significant differences in wind and solar power volatility across different time scales, and the advantages of different hydrogen production and energy storage devices at different time scales, it is necessary to conduct in-depth research on the multi-time-scale optimal operation of hydrogen production systems.

[0004] To address the aforementioned problems, the technical solution proposed in this invention aims to overcome the shortcomings of existing technologies by providing a multi-timescale scheduling method for hybrid hydrogen production systems that considers hydrogen load demand response. This method further mitigates the impact of source load fluctuations and improves the stability and flexibility of the hydrogen production system. The method described in this invention includes: firstly, processing source load prediction data using kernel density estimation and the Frank-Copula function; then, generating typical source load scenarios using an improved K-means clustering algorithm; and secondly, establishing a price-incentive-based hydrogen load demand response model to further explore the potential of hydrogen load, given that hydrogen load demand gradually increases and hydrogen prices change accordingly. Next, based on the operating characteristics of two types of electrolyzers, a modular rotation strategy for electrolyzers is established. Finally, considering the performance and advantages of the two types of electrolyzer arrays, batteries, and supercapacitors at different timescales, a multi-timescale optimized scheduling strategy for the hybrid hydrogen production system is established with the goal of minimizing the overall system operating cost. Summary of the Invention

[0005] This invention proposes a multi-timescale scheduling method for hybrid hydrogen production systems that takes into account hydrogen load demand response, which can further mitigate the impact of source load fluctuations and improve the stability and flexibility of the hydrogen production system.

[0006] The present invention adopts the following technical solution.

[0007] This invention presents a multi-timescale scheduling method for hybrid hydrogen production systems that considers hydrogen load demand response. To improve source load prediction accuracy and reduce prediction error uncertainty, the method processes source load prediction data using kernel density estimation, the Frank-Copula function, and mean clustering. If hydrogen demand is increasingly constrained by energy prices as demand grows, a price-incentive-based hydrogen load demand response model is also established. Furthermore, if modular hydrogen production clusters of the same capacity can fully utilize the wide operating range of hydrogen production equipment compared to large-capacity single units, this invention employs a Lagrange optimization-based electrolyzer array rotation control strategy to ensure the modular hydrogen production clusters fully utilize their hydrogen production capacity. The equipment exhibits a wide operating range. Finally, considering the complementary advantages of the two types of electrolyzers in terms of economy and dynamic response characteristics, the energy regulation role of batteries under reasonable charge conditions, and the rapid response role of supercapacitors, a multi-timescale scheduling method for the hybrid hydrogen production system, taking into account hydrogen load demand response, is proposed with the goal of minimizing the total system operating cost. This method enables different electrolyzers in the hybrid hydrogen production system to complement each other in terms of economy and dynamic response characteristics, allows batteries to provide energy regulation under reasonable charge conditions, and leverages the rapid response role of supercapacitors. In day-ahead scheduling, an energy management strategy that considers the multi-energy coordination of hybrid electrolyzers, hydrogen storage tanks, batteries, and hydrogen load is formulated to achieve a rational allocation of system energy.

[0008] The hybrid hydrogen production system employs an electro-hydrogen production unit to construct a hybrid hydrogen production array, with the topology as follows: Figure 1 As shown, the hydrogen production topology of the hybrid hydrogen production system includes a new energy supply layer, a green hydrogen production layer, and a hydrogen load demand layer;

[0009] The scheduling method first considers the uncertainty of prediction errors on both the source and load sides to obtain typical source and load scenarios. Simultaneously, a hydrogen load demand response model is established on the demand side. Through price incentives and user demand satisfaction constraints, the hydrogen load demand for each time period is rationally optimized, thereby improving wind and solar utilization and the flexibility of hydrogen load demand response. Secondly, batteries and supercapacitors in the green hydrogen production layer are configured to play an auxiliary regulatory role at different time scales. When new energy fluctuations are severe and hydrogen load demand is high, the batteries and supercapacitors maintain the power supply of the hybrid electrolyzer and the stability of the bus voltage through their own charging and discharging. Finally, the hydrogen production advantages of AEL and PEMEL are differentiated. In this invention, a hybrid electrolyzer array is used in the green hydrogen production layer. During hydrogen production, both types of electrolyzers are divided into small-capacity unit combinations, and each unit is numbered to broaden the power operating range of the electrolyzers. Simultaneously, the numbered units in each type of electrolyzer array are operated using a rotating operation strategy proposed in this invention, with different time periods for work sequencing and power allocation. This ensures uniform operating time and lifespan for each unit. Furthermore, the total power allocation between the alkaline electrolyzer and the proton exchange membrane electrolyzer adopts the hybrid electrolyzer power allocation strategy proposed in this invention and is rationally coordinated with the hydrogen storage tank to ensure hydrogen load supply and new energy consumption.

[0010] The method is based on the hydrogen production topology of a hybrid hydrogen production system. If there are significant differences in power generation and consumption of the source load, different hydrogen production equipment, and different energy storage units at different time scales, a multi-time scale energy management strategy for the hybrid hydrogen production system with coordinated interaction of multiple energy devices is adopted. Specifically, in the day-ahead phase, considering the source load uncertainty model and the hydrogen load demand response model, with a preset duration of 1 hour as the optimization scale and hydrogen load user satisfaction as the constraint, the system operating cost is minimized by coupling the advantages of different hydrogen production equipment and the auxiliary power regulation role of the battery under safe charging state. However, to further reduce the impact of source load errors on system operation... To mitigate the impact of scheduling and avoid frequent starts and stops of alkaline electrolyzers and frequent fluctuations in wind and solar power leading to multiple changes in the power purchase and sale status of the local system, this invention first uses a 15-minute optimization scale and a 4-hour rolling domain for intraday scheduling. By considering the reduceable characteristics of load demand response, the rapid adjustment and frequent start-stop characteristics of PEMEL electrolyzers, and the auxiliary characteristics of batteries, the output deviation between the day and the day is minimized. Finally, this invention further smooths out instantaneous power fluctuations of wind and solar power and improves energy utilization by using a preset short duration of 5 minutes as the time scale, aiming to minimize the output cost of supercapacitors and the output deviation of other equipment.

[0011] The multi-timescale scheduling method for hybrid hydrogen production systems adopts a multi-timescale energy management strategy for hybrid hydrogen production systems based on the coordination of multiple energy devices, including the following steps;

[0012] Step 1: Generate typical source-load scenarios and establish a hydrogen load demand response model;

[0013] Step 2: Establish a day-ahead-scale hybrid hydrogen production system scheduling model based on multi-device coordination; including the system's day-ahead energy management strategy and the system's day-ahead objective function;

[0014] Step 3: Establish a day-scale hybrid hydrogen production system scheduling model based on multi-device coordination;

[0015] Step 4: Establish a real-time scheduling model for a hybrid hydrogen production system based on supercapacitor regulation.

[0016] The output of wind and solar power and the energy consumption of users are uncertain. If the scheduling of the hybrid hydrogen production system is studied across multiple time scales based solely on deterministic models, the results will differ significantly from reality. Secondly, existing studies show that the scheduling results of each source-load scenario are relatively independent, and few studies characterize the uncertainty of source-load prediction errors by combining typical scenarios and their probabilities in the objective function. Therefore, this invention prioritizes addressing the uncertainty of source-load processing and establishes multiple types of typical scenarios, integrating them into the objective function across multiple time scales, before optimizing the system scheduling. In step one, when generating typical source-load scenarios, kernel density estimation, Frank-Copula function, and mean clustering algorithm are used to process the source-load prediction data. For the hydrogen load side, considering that the demand for hydrogen energy is becoming more widespread and has multiple uncertainties, and that the price of hydrogen energy is also increasing with the demand, the peak-valley price difference has changed in many places. The uncertainty of hydrogen demand and hydrogen price and their negative correlation characteristics are combined to construct typical scenarios of hydrogen load and hydrogen price, and then reasonably characterize the uncertainty of the two. The specific formulas are shown in the following formulas (1)-(6). For the source side, the corresponding algorithm is used to incorporate the uncertainty of wind and solar power output and the characteristic that wind and solar power output in the same region often have a negative correlation and complementary relationship. The specific formulas are shown in the following formulas (7)-(12).

[0017] Formula 1;

[0018] Formula 2;

[0019] Formula 3;

[0020] In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the average hydrogen load demand at time j; Let be the average hydrogen load price at time j; The hydrogen load demand at time j on day d; The hydrogen load price is shown on day d at time j; σ is the standard deviation of the sample data.

[0021] Formula 4;

[0022] Formula 5;

[0023] Formula 6;

[0024] In the formula: , respectively, are the marginal distribution functions of hydrogen load demand and hydrogen load price; C is the Copula join function; Let be the joint distribution function of hydrogen load and hydrogen price at time j;

[0025] Formula 7;

[0026] Formula 8;

[0027] Formula 9;

[0028] In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the sample mean of wind power output at time j; Let be the sample mean of photovoltaic power output at time j; The wind power output at time j on day d; The photovoltaic output at time j on day d; σ is the standard deviation of the sample data;

[0029] Formula 10;

[0030] Formula 11;

[0031] Formula 12;

[0032] In the formula: These are the edge distribution functions of wind power output and photovoltaic power output, respectively; C is the Copula connection function. Let be the joint distribution function of wind and solar power output at time j.

[0033] Traditional hydrogen load applications typically adopt a "produce and use as needed" approach without considering hydrogen load demand. However, with the increasing demand for hydrogen energy, cross-regional pipeline or hydrogen transport vehicles have emerged, leading to the increasingly widespread use of hydrogen energy. At the same time, hydrogen load prices will not remain fixed as in traditional energy use scenarios, but will fluctuate according to different time periods and locations. In step one, a hydrogen load demand response model based on price incentives is used. Incentive demand response is used as a resource scheduling tool to participate in the game between hydrogen load demand and hydrogen load price. Specific methods include: when energy prices change or system reliability decreases, after the hydrogen user receives a direct compensation notice from the hydrogen producer to induce a reduction in load or a notice to increase the hydrogen load price, the hydrogen user reduces or transfers the hydrogen load for a certain period, thereby achieving a positive interaction between the energy supply layer and the energy application layer. After considering the IDR model, the system hydrogen demand is divided into basic hydrogen load, curtailable hydrogen load (CL), and shiftable hydrogen load (SL). The hydrogen load demand response model based on IDR is shown in the following formulas (13)-(16).

[0034] Formula 13;

[0035] Formula 14;

[0036] Formula 15;

[0037] Formula 16;

[0038] In the formula: These are the self-elasticity coefficient and cross-elasticity coefficient of hydrogen load price, respectively; These represent the adjustment amount of hydrogen load at time t and the original demand for hydrogen load, respectively. They are respectively The hydrogen price and its change at any given time.

[0039] Day-ahead scheduling is the foundation for optimizing system operation across multiple time scales and also provides a reference for reducing intraday and real-time scheduling errors. This invention fully considers the advantages of different hydrogen production equipment, the auxiliary regulation role of batteries under state-of-charge safety zoning, and the scheduling potential of hydrogen load demand response in day-ahead scheduling. It proposes a day-ahead energy management strategy for a multi-equipment coordinated and interactive hydrogen production system, with a specific flowchart as follows: Figure 3 As shown. In step two, the system's day-ahead energy management strategy is specifically as follows: when the hydrogen production system meets the requirements... When the output of wind and solar power is greater than or equal to the power limit of the hydrogen production system, if the state of charge of the battery at the previous moment has reached the upper limit of the state of charge, both types of hydrogen production systems will operate at full load power, and the excess power will be included in the curtailment of wind and solar power. If the state of charge of the battery at the previous moment has not reached the upper limit, the hydrogen production system will also operate at full load power, and the excess power will be transferred to charge the lithium battery. If the excess power exceeds the power limit of the current period, the charging power will be set as the upper limit of the charging power for that period, and the remainder will be treated as curtailment of wind and solar power. If the excess power does not exceed the power limit of the current period, the battery charging power will be equal to the excess power.

[0040] When the hydrogen production system meets But at the same time, it satisfies When the wind and solar power output is less than the upper limit of the hydrogen production system's power but greater than or equal to the lower limit of hydrogen production power, if the battery's state of charge (SOC) was greater than the safe upper limit in the previous moment, then the wind and solar power will be allocated to the hybrid hydrogen production array for power distribution, thereby producing green hydrogen. If the battery's SOC is between 0.6 and 0.8 and the battery was charging in the previous moment, and to effectively avoid overcharging and over-discharging of the battery, all wind and solar power will be allocated to the hybrid hydrogen production power distribution strategy in this moment. However, if the battery was discharging in the previous moment, it is necessary to determine whether the electrolyzer was operating at the minimum hydrogen production power in the previous moment. If the electrolyzer array was operating at the minimum hydrogen production power in the previous moment... To avoid frequent start-ups and shutdowns of the electrolyzer array and prolonged standby, all power at this moment is allocated to the hybrid hydrogen production array for green hydrogen production. Otherwise, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production. If the battery's state of charge was below 0.6% at the previous moment, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production.

[0041] When the hydrogen production system meets Then, under the premise of ensuring the safe operation of the battery, it is necessary to discharge to assist in supplying hydrogen production power. If the state of charge of the battery at the previous moment is greater than or equal to 0.2, the battery will discharge to supply power for the mixed hydrogen production array power distribution. If the battery is less than 0.2, in order to minimize the waste of wind and solar power, the current wind and solar power should be divided by the single-cell power limit to obtain the minimum number of operable cells. Then, the start-up and shutdown should be allocated according to the characteristics of the two types of electrolyzers. In the allocation process, in order to reduce the number of start-ups and shutdowns of alkaline electrolyzers, the number of alkaline electrolyzers in operation should be prioritized to reduce the number of alkaline electrolyzers that need to be shut down. Proton exchange membrane electrolyzers can be started and stopped in emergencies. Finally, the waste power is calculated.

[0042] Based on the evaluation of existing hybrid hydrogen production power capacity configurations, it is known that the ratio of alkaline electrolyzer capacity to proton exchange membrane electrolyzer capacity in a hydrogen production system is 2.5:1. Therefore, upon entering the hybrid electrolyzer power allocation strategy module, this invention allocates power to the hybrid electrolyzer according to the following strategy, as shown in the flowchart below. Figure 4 As shown, in the power allocation strategy of the hybrid electrolyzer in the hybrid hydrogen production system, when the total hydrogen production power of the hybrid hydrogen production system is greater than or equal to the full-load power of the alkaline electrolyzer, in order to ensure the stable operation of the alkaline electrolyzer at its base load, the alkaline electrolyzer is prioritized to operate at full-load power, and the remaining power is allocated to the proton exchange membrane electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer and less than or equal to the full-load power of the proton exchange membrane electrolyzer, frequent start-ups and shutdowns of the two types of electrolyzers are reduced, and the hydrogen production power needs to be evenly distributed between the proton exchange membrane electrolyzer and the alkaline electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer but greater than the full-load power of the proton exchange membrane electrolyzer, this stage allows for greater flexibility. When using a hybrid hydrogen production array, it is still necessary to check whether the hydrogen production power is greater than or equal to twice the full-load power of the proton exchange membrane electrolyzer. If the hydrogen production power is less than or equal to twice the full-load power of the proton exchange membrane electrolyzer, the full-load power of the proton exchange should be preferentially allocated to the alkaline electrolyzer array, and the remaining power should be allocated to the proton exchange membrane electrolyzer array. The purpose is to avoid the alkaline electrolyzer being in a long-term overload state, while maintaining the operating state of the proton exchange membrane electrolyzer to avoid frequent start-ups and shutdowns. Conversely, if the full-load power is preferentially allocated to the proton exchange membrane, the remaining power should be allocated to the alkaline electrolyzer, thereby reducing the amount of wasted electricity. Finally, the power of the alkaline electrolyzer and the power of the proton exchange membrane electrolyzer are respectively transferred to the array operation module.

[0043] In hydrogen production systems, to further enhance the flexibility of electrolyzers and ensure the consistency of the operating status and lifespan of individual electrolyzers, this invention, considering the operating characteristics of two types of electrolyzer equipment, employs a Lagrange-optimized rotating operation strategy for the electrolyzer array, where the control diagram is as follows: Figure 5 As shown. This invention includes the following control scenarios using a proton exchange membrane electrolyzer, where P... EL The total hydrogen production power of the input proton exchange membrane system needs to be allocated to the array power.

[0044] Scenario 1: Overload Condition

[0045] At this time, all electrolytic cells are operating at full capacity. Even at this time, the requirements could not be met. Therefore, considering the overload characteristics of the electrolytic cells, some electrolytic cells were put into overload mode. The number of electrolytic cells operating in overload mode, m, and the remaining power were determined. The expression is as follows:

[0046] Formula 17;

[0047] Formula 18;

[0048] In the formula: "floor" is the floor function;

[0049] Based on the allocation, the m units waiting to work at the corresponding time will be placed at the highest overload power operation; Assigned to the (m+1)th monomer; the remaining monomers are... run.

[0050] Formula 19;

[0051] In the formula: This represents the operating power of the nth cell at the corresponding time.

[0052] Based on its operating characteristics, the overload rate of a single electrolytic cell should be less than or equal to 0.5.

[0053] Formula 20;

[0054] In the formula: λn is the overload rate of electrolytic cell n; These are the total operating time and overload operating time of electrolytic cell n, respectively.

[0055] Scenario 2: Stable operating conditions This operating condition requires both ensuring balanced operation of each unit and improving overall hydrogen production efficiency. A power allocation strategy based on the Lagrange multiplier method is adopted, expressed as follows:

[0056] Formula 21;

[0057] In the formula: f is the overall efficiency at the corresponding time; The power of the i-th unit actually participating in hydrogen production at the corresponding moment; Let be the input operating power of the i-th unit at the corresponding time.

[0058] Formula 22;

[0059] Formula 23;

[0060] In the formula: The Lagrangian function is constructed; For Lagrange multipliers;

[0061] For constraints; They are respectively For variables Differentiate; φ represents the variables respectively. Differentiate;

[0062] Scenario 3: Low power operation At this time, P EL It is insufficient to support each unit to work in standby mode, but in order to reduce the number of units that are shut down and the downtime, the units are made to work in standby mode to the greatest extent possible.

[0063] Let m be the number of units in operation and the remaining power be... The expression is as follows:

[0064] Formula 24;

[0065] Formula 25;

[0066] To make rational use Under the premise of ensuring the safe operation of the electrolytic cell, The unit is assigned to the first unit to be worked at the corresponding time. The initial assignment for this working condition is as follows:

[0067] Formula 26;

[0068] Outer layer rotation: To avoid the front units being in an overloaded state for a long time and the rear units being in an offline state for a long time, resulting in an imbalance in the working time and operating status of the units within the array, an array rotation optimization strategy is adopted, the specific formula of which is as follows:

[0069] Formula 27;

[0070] In the formula: Let be the ID of the single entity at time t-1 that is at the k-th ID; n is the total number of single entities in the array; mod is the modulo operation, i.e., taking the remainder when k-1 is divided by n. , where q and n are both integers; initially, the self-number is selected;

[0071] After the system completes hydrogen production, to make the energy scheduling of the hydrogen production system more rational, this invention rationally allocates hydrogen energy based on the demand response of the hydrogen load, as shown in the flowchart below. Figure 5As shown, the specific power allocation strategy is as follows: When the hydrogen production of the mixed electrolyzer exceeds the system's hydrogen demand response, the remaining hydrogen production is the difference between the two. To improve energy utilization and avoid energy waste, this invention introduces a hydrogen storage tank to participate in the system's energy management. If the current hydrogen charge state is greater than or equal to 0.9, the remaining hydrogen production is treated as discarded energy. If the current hydrogen charge state is less than 0.9, the remaining hydrogen production can be stored in the hydrogen storage tank, but it needs to be determined whether the hydrogen charging power limit is met. If it exceeds the limit, the hydrogen charging power is set at the upper limit value, and the final discarded energy is obtained from the initial remaining hydrogen. If the hydrogen production is less than the upper limit of the hydrogen storage tank's charging power, and the limit is not exceeded, the remaining hydrogen production can be stored in the hydrogen storage tank. However, when the hydrogen production of the mixed electrolyzer is less than the demand after the hydrogen demand response, if the hydrogen charge of the hydrogen storage tank is greater than or equal to 0.1, the hydrogen storage tank can release hydrogen within a safe limit to meet the demand in order to further meet the load demand. However, if the released hydrogen production exceeds the hydrogen release limit of the hydrogen storage tank, the released hydrogen amount is set as the hydrogen release limit of the hydrogen storage tank, and the hydrogen shortage is the limit plus the mixed hydrogen production production minus the hydrogen load demand. If the hydrogen charge of the hydrogen storage tank is less than 0.1, the mixed hydrogen production production minus the hydrogen demand is the hydrogen shortage of the system.

[0072] In step two, an energy scheduling model for the hydrogen production system with the lowest daily operating cost is constructed by combining day-ahead source load forecast data and energy management strategies. However, this invention considers the impact of source load uncertainty in the scheduling model and incorporates the scenario probabilities of source load uncertainty into the system's day-ahead objective function.

[0073] The calculation formulas for various cost indicators in scenarios with source load uncertainty, and the calculation formula for the total cost considering source load uncertainty, are shown in the following formulas.

[0074] Formula 28;

[0075] Formula 29;

[0076] Formula 30;

[0077] Formula 31;

[0078] Formula 32;

[0079] Formula 33;

[0080] In the formula: The overall operating cost of the system should take into account the uncertainty of source load. This represents the total operating cost in scenario S. Let S be the uncertain probability of the S-th scenario; For the system's operation and maintenance costs; The operating cost of the hybrid electrolytic cell array; The cost of system energy curtailment penalties; To compensate for the cost of hydrogen load demand response; These are the operation and maintenance cost coefficients for batteries, hydrogen storage tanks, photovoltaic systems, and wind turbines, respectively. These represent the operating outputs of the battery, hydrogen storage tank, photovoltaic system, and wind turbine at time t, respectively. The usage costs for AEL and PEM are respectively; The startup and shutdown costs for AEL and PEM are respectively; These are the penalty coefficients for system energy curtailment / deficit and hydrogen curtailment / deficit, respectively. These represent the magnitude of the system's abandoned / deficient electrical energy and abandoned / deficient hydrogen energy, respectively.

[0081] These are the transferable compensation coefficient and the reduction compensation coefficient for hydrogen load, respectively. These represent the hydrogen load that can be transferred at time t and the hydrogen load that can be reduced during time period t, respectively.

[0082] The relevant constraints of the scheduling model must satisfy the balance between electricity and hydrogen power, and all equipment in the hydrogen production system must also operate within safe limits.

[0083] In step three, considering the differences in source load output at different time scales, and to reduce the impact of source load output deviations at different time scales on system operation scheduling, the prediction time domain is shortened to 4 hours and the time scale to 15 minutes based on the day-ahead scheduling, and intraday scheduling is carried out on the system. However, during the scheduling process, this invention fully considers the differences in operating characteristics between the two types of electrolyzers and to avoid frequent start-ups and shutdowns of alkaline electrolyzers. Therefore, the start-up and shutdown status of alkaline electrolyzers remains unchanged at this stage, and only the output is adjusted based on the existing status. Proton exchange membrane electrolyzers can be started, stopped, and their output adjusted according to their operating advantages. At the same time, the batteries can ensure that the state of charge is within a reasonable operating range to assist in the energy regulation of the hydrogen production system. Secondly, under the requirement of prioritizing the fulfillment of load demand, this invention only considers the situation where hydrogen load demand can be reduced at this stage. Finally, the impact of source load uncertainty is considered in the scheduling model, and the scenario probability of source load uncertainty is incorporated into the objective function.

[0084] The calculation expressions for various cost indicators in scenarios with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are as follows:

[0085] Formula 34;

[0086] Formula 35;

[0087] Formula 36;

[0088] Formula 37;

[0089] Formula 38;

[0090] Formula 39;

[0091] Formula 40;

[0092] In the formula: These are the penalty costs for power fluctuation deviations in the electrolyzer, battery, and hydrogen storage tank, respectively.

[0093] The figures represent the power fluctuations of the alkaline electrolyzer, proton exchange membrane electrolyzer, battery, and hydrogen storage tank at time t.

[0094] The relevant constraints of the scheduling model still need to satisfy the balance of electricity and hydrogen power, and each device also needs to operate within the safety limits.

[0095] Entering the real-time phase, this invention considers the characteristics of rapid response and large variation in instantaneous power fluctuations. Frequent changes to the start-up and shutdown states of the battery and proton exchange membrane electrolyzer during the real-time phase would not only severely impact equipment lifespan but also significantly affect system operation and scheduling. Therefore, this invention fully leverages the advantages of supercapacitors—rapid response and good instantaneous smoothing effect—and formulates a real-time scheduling model for supercapacitors based on the variable baseline criterion. In step four, the processing method on the load side during the real-time phase is the same as in the intraday phase. Finally, in the scheduling model, this invention considers the impact of source-load uncertainty and incorporates the scenario probability of source-load uncertainty into the objective function. The relevant constraints of the scheduling model must satisfy the balance between electricity and hydrogen power, and each device must operate within safe limits.

[0096] The calculation expressions for various cost indicators in the scenario with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are shown below:

[0097]

[0098] Formula 41;

[0099]

[0100] Formula 42;

[0101] Formula 43;

[0102]

[0103] Formula 44;

[0104]

[0105] Formula 45;

[0106] Formula 46;

[0107] Formula 47;

[0108]

[0109] Formula 48;

[0110] Formula 49;

[0111] In the formula: These are the real-time power of wind and solar power and the power after smoothing, respectively. These represent the charging power and discharging power of the supercapacitor, respectively. The penalty cost for curtailing wind and solar power in real time; The baseline power of the supercapacitor during time period t is given.

[0112] This invention proposes a multi-timescale scheduling method for a hybrid hydrogen production system that considers hydrogen load demand response. First, to reasonably characterize the uncertainty of source load prediction errors, this invention employs kernel density estimation and the Frank-Copula function to process source load prediction data, and uses mean clustering to generate typical scenarios for the prediction data, thereby effectively avoiding the biases introduced by deterministic source load models in the scheduling results. Second, to fully explore the potential of hydrogen load demand response, this invention establishes a price-incentive-based hydrogen load demand response model, thereby ensuring user satisfaction while further mitigating source load fluctuations and improving the system's energy utilization rate. Third, starting from the core equipment of the hydrogen production system, this invention fully utilizes the respective advantages of alkaline electrolyzers and proton exchange membrane electrolyzers in terms of hydrogen production cost and dynamic performance, and considers the wide range of characteristics of small-capacity single-unit hydrogen production. First, this invention proposes a rotating control strategy for the electrolyzer array and a power allocation strategy for the hybrid electrolyzer, thereby further improving the flexibility and stability of the hydrogen production system. Second, considering the differences in source load fluctuations at different time scales and the differences and advantages of different hydrogen production equipment and energy storage equipment at different time scales, the multi-time scale scheduling method for the hybrid hydrogen production system that takes into account hydrogen load demand response proposed in this invention can fully utilize the advantages and characteristics of different types of electrolyzers and energy storage units. It proposes an energy management strategy for coordinated interaction of multiple devices in the hybrid hydrogen production system, and establishes a day-ahead, intraday, and real-time optimization scheduling model for the system with the goal of minimizing the total system operating cost. This effectively solves the problem of low reliability of the hybrid hydrogen production system operation optimization results caused by source load prediction errors, and solves the problem of wind and solar power being difficult to suppress in short cycles through the rapid response of supercapacitors.

[0113] This invention focuses on hybrid hydrogen production systems and proposes a multi-timescale scheduling method for such systems that considers hydrogen load demand response. The method includes: First, to improve source load prediction accuracy and reduce prediction error uncertainty, this invention uses kernel density estimation, the Frank-Copula function, and mean clustering to process the source load prediction data. Simultaneously, as hydrogen demand increases, its dependence on energy prices becomes increasingly significant; therefore, this invention establishes a hydrogen load demand response model based on price incentives. Subsequently, compared to large-capacity single-unit systems, modular hydrogen production clusters of the same capacity can fully utilize the wide-range operating characteristics of hydrogen production equipment; therefore, this invention proposes a Lagrange optimization-based electrical... The invention proposes a multi-timescale scheduling method for a hybrid hydrogen production system, considering the complementary advantages of the two types of electrolyzers in terms of economy and dynamic response characteristics, the energy regulation role of batteries under reasonable charge states, and the rapid response role of supercapacitors. This method aims to minimize the total system operating cost and takes into account hydrogen load demand response. In day-ahead scheduling, the invention formulates a multi-energy coordination energy management strategy considering hybrid electrolyzers, hydrogen storage tanks, batteries, and hydrogen load, enabling rational allocation of system energy, further improving the system's energy utilization rate, and reducing day-ahead forecasting errors.

[0114] This invention implements intraday and real-time scheduling for a hybrid hydrogen production system. In intraday scheduling, considering the better dynamic response performance of the proton exchange membrane electrolyzer (PEMEE) and the slower dynamic response of the alkaline electrolyzer, and to prevent frequent changes in the pre-determined start-up and shutdown of the alkaline electrolyzer due to intraday wind and solar power fluctuations, the operating state variables of the alkaline electrolyzer remain fixed during intraday scheduling, with output adjustments only made under this state. The PEMEE can adapt to rapid power fluctuations based on its own response advantages. Simultaneously, the battery can flexibly and safely cooperate with the PEMEE in energy regulation, taking into account SOC (State of Charge) zoning. At this scale, a 4-hour rolling time domain is used, with continuous rolling optimization every 15 minutes. In real-time scheduling, a 20-minute rolling time domain is used, with an optimization cycle of 5 minutes. During this stage, the electrolyzer, battery, and system purchase / sales status remain fixed. Only the rapid adjustment effect of the supercapacitor is considered to smooth out real-time wind and solar power fluctuations within the system, further improving the system's energy utilization rate and reducing intraday output deviations.

[0115] The hydrogen production equipment topology proposed in this invention can achieve full-chain optimization of green hydrogen application by using a hydrogen load demand response model, simultaneously linking new energy wind and solar power output, hybrid energy storage combining batteries and supercapacitors, coordinated scheduling of hybrid hydrogen production arrays, and hydrogen storage tanks, thereby realizing "source-storage-production-load" optimization. This aims to provide a valuable reference for the application and development of future wind and solar on-site hydrogen production and consumption sites, regional integrated energy system industrial parks, zero-carbon industrial parks, and green hydrogen refueling stations in the country. Attached Figure Description

[0116] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0117] Appendix Figure 1 This is a schematic diagram of the topology of a hybrid hydrogen production system that takes into account hydrogen load demand response in an embodiment of the present invention;

[0118] Appendix Figure 2 This is a schematic diagram of the day-ahead energy management strategy process in an embodiment of the present invention;

[0119] Appendix Figure 3 This is a schematic diagram of the power distribution process of the hybrid electrolyzer in an embodiment of the present invention;

[0120] Appendix Figure 4 This is a schematic diagram of the power allocation strategy for the electrolytic cell array in an embodiment of the present invention;

[0121] Appendix Figure 5 This is a schematic diagram of the energy management strategy between hydrogen demand and hydrogen production in an embodiment of the present invention. Detailed Implementation

[0122] As shown in the figure, a multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response is presented. To improve source load prediction accuracy and reduce prediction error uncertainty, the method employs kernel density estimation, the Frank-Copula function, and mean clustering algorithms to process the source load prediction data of the hybrid hydrogen production system. If, as hydrogen demand increases, its dependence on energy prices becomes increasingly significant, a hydrogen load demand response model based on price incentives is also established. Furthermore, if, compared to large-capacity single-unit systems, modular hydrogen production clusters of the same capacity can fully utilize the wide operating range of hydrogen production equipment, this invention employs a Lagrange optimization-based electrolyzer array rotation control strategy to ensure the modular hydrogen production clusters fully utilize their capabilities. The wide operating range of hydrogen production equipment is considered. Finally, taking into account the complementary advantages of the two types of electrolyzers in terms of economy and dynamic response characteristics, the power regulation role of batteries under reasonable charge conditions, and the rapid response role of supercapacitors, a multi-timescale scheduling method for hybrid hydrogen production systems is proposed, aiming to minimize the total system operating cost and taking into account hydrogen load demand response. This method enables different electrolyzers in the hybrid hydrogen production system to complement each other in terms of economy and dynamic response characteristics, and leverages the power regulation role of batteries under reasonable charge conditions and the rapid response role of supercapacitors. In day-ahead scheduling, an energy management strategy that considers the multi-energy coordination of hybrid electrolyzers, hydrogen storage tanks, batteries, and hydrogen load is formulated to achieve a rational allocation of system energy.

[0123] The hybrid hydrogen production system employs an electro-hydrogen production unit to construct a hybrid hydrogen production array, with the topology as follows: Figure 1 As shown, the hydrogen production topology of the hybrid hydrogen production system includes a new energy supply layer, a green hydrogen production layer, and a hydrogen load demand layer;

[0124] The scheduling method first considers the uncertainty of prediction errors on both the source and load sides to obtain typical source and load scenarios. Simultaneously, a hydrogen load demand response model is established on the demand side. Through price incentives and user demand satisfaction constraints, the hydrogen load demand for each time period is rationally optimized, thereby improving wind and solar utilization and the flexibility of hydrogen load demand response. Secondly, batteries and supercapacitors in the green hydrogen production layer are configured to play an auxiliary regulatory role at different time scales. When new energy fluctuations are severe and hydrogen load demand is high, the batteries and supercapacitors maintain the power supply of the hybrid electrolyzer and the stability of the bus voltage through their own charging and discharging. Finally, the hydrogen production advantages of AEL and PEMEL are differentiated. In this invention, a hybrid electrolyzer array is used in the green hydrogen production layer. During hydrogen production, both types of electrolyzers are divided into small-capacity unit combinations, and each unit is numbered to broaden the power operating range of the electrolyzers. Simultaneously, the numbered units in each type of electrolyzer array are operated using a rotating operation strategy proposed in this invention, with different time periods for work sequencing and power allocation. This ensures uniform operating time and lifespan for each unit. Furthermore, the total power allocation between the alkaline electrolyzer and the proton exchange membrane electrolyzer adopts the hybrid electrolyzer power allocation strategy proposed in this invention and is rationally coordinated with the hydrogen storage tank to ensure hydrogen load supply and new energy consumption.

[0125] The method is based on the hydrogen production topology of a hybrid hydrogen production system. If there are significant differences in power generation and consumption of the source load, different hydrogen production equipment, and different energy storage units at different time scales, a multi-time scale energy management strategy for the hybrid hydrogen production system with coordinated interaction of multiple energy devices is adopted. Specifically, in the day-ahead phase, considering the source load uncertainty model and the hydrogen load demand response model, with a preset duration of 1 hour as the optimization scale and hydrogen load user satisfaction as the constraint, the system operating cost is minimized by coupling the advantages of different hydrogen production equipment and the auxiliary power regulation role of the battery under safe charging state. However, to further reduce the impact of source load errors on system operation... To mitigate the impact of scheduling and avoid frequent starts and stops of alkaline electrolyzers and frequent fluctuations in wind and solar power leading to multiple changes in the power purchase and sale status of the local system, this invention first uses a 15-minute optimization scale and a 4-hour rolling domain for intraday scheduling. By considering the reduceable characteristics of load demand response, the rapid adjustment and frequent start-stop characteristics of PEMEL electrolyzers, and the auxiliary characteristics of batteries, the output deviation between the day and the day is minimized. Finally, this invention further smooths out instantaneous power fluctuations of wind and solar power and improves energy utilization by using a preset short duration of 5 minutes as the time scale, aiming to minimize the output cost of supercapacitors and the output deviation of other equipment.

[0126] The multi-timescale scheduling method for hybrid hydrogen production systems adopts a multi-timescale energy management strategy for hybrid hydrogen production systems based on the coordination of multiple energy devices, including the following steps;

[0127] Step 1: Generate typical source-load scenarios and establish a hydrogen load demand response model;

[0128] Step 2: Establish a day-ahead-scale hybrid hydrogen production system scheduling model based on multi-device coordination; including the system's day-ahead energy management strategy and the system's day-ahead objective function;

[0129] Step 3: Establish a day-scale hybrid hydrogen production system scheduling model based on multi-device coordination;

[0130] Step 4: Establish a real-time scheduling model for a hybrid hydrogen production system based on supercapacitor regulation.

[0131] The output of wind and solar power and the energy consumption of users are uncertain. If the scheduling of the hybrid hydrogen production system is studied across multiple time scales based solely on deterministic models, the results will differ significantly from reality. Secondly, existing studies show that the scheduling results of each source-load scenario are relatively independent, and few studies characterize the uncertainty of source-load prediction errors by combining typical scenarios and their probabilities in the objective function. Therefore, this invention prioritizes addressing the uncertainty of source-load processing and establishes multiple types of typical scenarios, integrating them into the objective function across multiple time scales, before optimizing the system scheduling. In step one, when generating typical source-load scenarios, kernel density estimation, Frank-Copula function, and mean clustering algorithm are used to process the source-load prediction data. For the hydrogen load side, considering that the demand for hydrogen energy is becoming more widespread and has multiple uncertainties, and that the price of hydrogen energy is also increasing with the demand, the peak-valley price difference has changed in many places. The uncertainty of hydrogen demand and hydrogen price and their negative correlation characteristics are combined to construct typical scenarios of hydrogen load and hydrogen price, and then reasonably characterize the uncertainty of the two. The specific formulas are shown in the following formulas (1)-(6). For the source side, the corresponding algorithm is used to incorporate the uncertainty of wind and solar power output and the characteristic that wind and solar power output in the same region often have a negative correlation and complementary relationship. The specific formulas are shown in the following formulas (7)-(12).

[0132] Formula 1;

[0133] Formula 2;

[0134] Formula 3;

[0135] In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the average hydrogen load demand at time j; Let be the average hydrogen load price at time j; The hydrogen load demand at time j on day d; The hydrogen load price is shown on day d at time j; σ is the standard deviation of the sample data.

[0136] Formula 4;

[0137] Formula 5;

[0138] Formula 6;

[0139] In the formula: , respectively, are the marginal distribution functions of hydrogen load demand and hydrogen load price; C is the Copula join function; Let be the joint distribution function of hydrogen load and hydrogen price at time j;

[0140] Formula 7;

[0141] Formula 8;

[0142] Formula 9;

[0143] In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the sample mean of wind power output at time j; Let be the sample mean of photovoltaic power output at time j; The wind power output at time j on day d; The photovoltaic output at time j on day d; σ is the standard deviation of the sample data;

[0144] Formula 10;

[0145] Formula 11;

[0146] Formula 12;

[0147] In the formula: These are the edge distribution functions of wind power output and photovoltaic power output, respectively; C is the Copula connection function. Let be the joint distribution function of wind and solar power output at time j.

[0148] Traditional hydrogen load applications typically adopt a "produce and use as needed" approach without considering hydrogen load demand. However, with the increasing demand for hydrogen energy, cross-regional pipeline or hydrogen transport vehicles have emerged, leading to the increasingly widespread use of hydrogen energy. At the same time, hydrogen load prices will not remain fixed as in traditional energy use scenarios, but will fluctuate according to different time periods and locations. In step one, a hydrogen load demand response model based on price incentives is used. Incentive demand response is used as a resource scheduling tool to participate in the game between hydrogen load demand and hydrogen load price. Specific methods include: when energy prices change or system reliability decreases, after the hydrogen user receives a direct compensation notice from the hydrogen producer to induce a reduction in load or a notice to increase the hydrogen load price, the hydrogen user reduces or transfers the hydrogen load for a certain period, thereby achieving a positive interaction between the energy supply layer and the energy application layer. After considering the IDR model, the system hydrogen demand is divided into basic hydrogen load, curtailable hydrogen load (CL), and shiftable hydrogen load (SL). The hydrogen load demand response model based on IDR is shown in the following formulas (13)-(16).

[0149] Formula 13;

[0150] Formula 14;

[0151] Formula 15;

[0152] Formula 16;

[0153] In the formula: These are the self-elasticity coefficient and cross-elasticity coefficient of hydrogen load price, respectively; These represent the adjustment amount of hydrogen load at time t and the original demand for hydrogen load, respectively. They are respectively The hydrogen price and its change at any given time.

[0154] Day-ahead scheduling is the foundation for optimizing system operation across multiple time scales and also provides a reference for reducing intraday and real-time scheduling errors. This invention fully considers the advantages of different hydrogen production equipment, the auxiliary regulation role of batteries under state-of-charge safety zoning, and the scheduling potential of hydrogen load demand response in day-ahead scheduling. It proposes a day-ahead energy management strategy for a multi-equipment coordinated and interactive hydrogen production system, with a specific flowchart as follows: Figure 3 As shown. In step two, the system's day-ahead energy management strategy is specifically as follows: when the hydrogen production system meets the requirements... When the output of wind and solar power is greater than or equal to the power limit of the hydrogen production system, if the state of charge of the battery at the previous moment has reached the upper limit of the state of charge, both types of hydrogen production systems will operate at full load power, and the excess power will be included in the curtailment of wind and solar power. If the state of charge of the battery at the previous moment has not reached the upper limit, the hydrogen production system will also operate at full load power, and the excess power will be transferred to charge the lithium battery. If the excess power exceeds the power limit of the current period, the charging power will be set as the upper limit of the charging power for that period, and the remainder will be treated as curtailment of wind and solar power. If the excess power does not exceed the power limit of the current period, the battery charging power will be equal to the excess power.

[0155] When the hydrogen production system meets But at the same time, it satisfies When the wind and solar power output is less than the upper limit of the hydrogen production system's power but greater than or equal to the lower limit of hydrogen production power, if the battery's state of charge (SOC) was greater than the safe upper limit in the previous moment, then the wind and solar power will be allocated to the hybrid hydrogen production array for power distribution, thereby producing green hydrogen. If the battery's SOC is between 0.6 and 0.8 and the battery was charging in the previous moment, and to effectively avoid overcharging and over-discharging of the battery, all wind and solar power will be allocated to the hybrid hydrogen production power distribution strategy in this moment. However, if the battery was discharging in the previous moment, it is necessary to determine whether the electrolyzer was operating at the minimum hydrogen production power in the previous moment. If the electrolyzer array was operating at the minimum hydrogen production power in the previous moment... To avoid frequent start-ups and shutdowns of the electrolyzer array and prolonged standby, all power at this moment is allocated to the hybrid hydrogen production array for green hydrogen production. Otherwise, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production. If the battery's state of charge was below 0.6% at the previous moment, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production.

[0156] When the hydrogen production system meets Then, under the premise of ensuring the safe operation of the battery, it is necessary to discharge to assist in supplying hydrogen production power. If the state of charge of the battery at the previous moment is greater than or equal to 0.2, the battery will discharge to supply power for the mixed hydrogen production array power distribution. If the battery is less than 0.2, in order to minimize the waste of wind and solar power, the current wind and solar power should be divided by the single-cell power limit to obtain the minimum number of operable cells. Then, the start-up and shutdown should be allocated according to the characteristics of the two types of electrolyzers. In the allocation process, in order to reduce the number of start-ups and shutdowns of alkaline electrolyzers, the number of alkaline electrolyzers in operation should be prioritized to reduce the number of alkaline electrolyzers that need to be shut down. Proton exchange membrane electrolyzers can be started and stopped in emergencies. Finally, the waste power is calculated.

[0157] Based on the evaluation of existing hybrid hydrogen production power capacity configurations, it is known that the ratio of alkaline electrolyzer capacity to proton exchange membrane electrolyzer capacity in a hydrogen production system is 2.5:1. Therefore, upon entering the hybrid electrolyzer power allocation strategy module, this invention allocates power to the hybrid electrolyzer according to the following strategy, as shown in the flowchart below. Figure 4 As shown, in the power allocation strategy of the hybrid electrolyzer in the hybrid hydrogen production system, when the total hydrogen production power of the hybrid hydrogen production system is greater than or equal to the full-load power of the alkaline electrolyzer, in order to ensure the stable operation of the alkaline electrolyzer at its base load, the alkaline electrolyzer is prioritized to operate at full-load power, and the remaining power is allocated to the proton exchange membrane electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer and less than or equal to the full-load power of the proton exchange membrane electrolyzer, frequent start-ups and shutdowns of the two types of electrolyzers are reduced, and the hydrogen production power needs to be evenly distributed between the proton exchange membrane electrolyzer and the alkaline electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer but greater than the full-load power of the proton exchange membrane electrolyzer, this stage allows for greater flexibility. When using a hybrid hydrogen production array, it is still necessary to check whether the hydrogen production power is greater than or equal to twice the full-load power of the proton exchange membrane electrolyzer. If the hydrogen production power is less than or equal to twice the full-load power of the proton exchange membrane electrolyzer, the full-load power of the proton exchange should be preferentially allocated to the alkaline electrolyzer array, and the remaining power should be allocated to the proton exchange membrane electrolyzer array. The purpose is to avoid the alkaline electrolyzer being in a long-term overload state, while maintaining the operating state of the proton exchange membrane electrolyzer to avoid frequent start-ups and shutdowns. Conversely, if the full-load power is preferentially allocated to the proton exchange membrane, the remaining power should be allocated to the alkaline electrolyzer, thereby reducing the amount of wasted electricity. Finally, the power of the alkaline electrolyzer and the power of the proton exchange membrane electrolyzer are respectively transferred to the array operation module.

[0158] In hydrogen production systems, to further enhance the flexibility of electrolyzers and ensure the consistency of the operating status and lifespan of individual electrolyzers, this invention, considering the operating characteristics of two types of electrolyzer equipment, employs a Lagrange-optimized rotating operation strategy for the electrolyzer array, where the control diagram is as follows: Figure 5 As shown. This invention includes the following control scenarios using a proton exchange membrane electrolyzer, where P... EL The total hydrogen production power of the input proton exchange membrane system needs to be allocated to the array power.

[0159] Scenario 1: Overload Condition

[0160] At this time, all electrolytic cells are operating at full capacity. Even at this time, the requirements could not be met. Therefore, considering the overload characteristics of the electrolytic cells, some electrolytic cells were put into overload mode. The number of electrolytic cells operating in overload mode, m, and the remaining power were determined. The expression is as follows:

[0161] Formula 17;

[0162] Formula 18;

[0163] In the formula: "floor" is the floor function;

[0164] Based on the allocation, the m units waiting to work at the corresponding time will be placed at the highest overload power operation; Assigned to the (m+1)th monomer; the remaining monomers are... run.

[0165] Formula 19;

[0166] In the formula: This represents the operating power of the nth cell at the corresponding time.

[0167] Based on its operating characteristics, the overload rate of a single electrolytic cell should be less than or equal to 0.5.

[0168] Formula 20;

[0169] In the formula: λn is the overload rate of electrolytic cell n; These are the total operating time and overload operating time of electrolytic cell n, respectively.

[0170] Scenario 2: Stable operating conditions This operating condition requires both ensuring balanced operation of each unit and improving overall hydrogen production efficiency. A power allocation strategy based on the Lagrange multiplier method is adopted, expressed as follows:

[0171] Formula 21;

[0172] In the formula: f is the overall efficiency at the corresponding time; The power of the i-th unit actually participating in hydrogen production at the corresponding moment; Let be the input operating power of the i-th unit at the corresponding time.

[0173] Formula 22;

[0174] Formula 23;

[0175] In the formula: The Lagrangian function is constructed; For Lagrange multipliers;

[0176] For constraints; They are respectively For variables Differentiate; φ represents the variables respectively. Differentiate;

[0177] Scenario 3: Low-power operation At this time, P EL It is insufficient to support each unit to work in standby mode, but in order to reduce the number of units that are shut down and the downtime, the units are made to work in standby mode to the greatest extent possible.

[0178] Let m be the number of units in operation and the remaining power be... The expression is as follows:

[0179] Formula 24;

[0180] Formula 25;

[0181] To make rational use Under the premise of ensuring the safe operation of the electrolytic cell, The unit is assigned to the first unit to be worked at the corresponding time. The initial assignment for this working condition is as follows:

[0182] Formula 26;

[0183] Outer layer rotation: To avoid the leading cells being under overload for extended periods and the trailing cells being under shutdown for extended periods, resulting in an imbalance in the working time and operating status of the cells within the array, an array rotation optimization strategy is adopted. The specific formula is as follows:

[0184] Formula 27;

[0185] In the formula: Let be the ID of the single entity at time t-1 that is at the k-th ID; n is the total number of single entities in the array; mod is the modulo operation, i.e., taking the remainder when k-1 is divided by n. , where q and n are both integers; initially, the self-number is selected;

[0186] After the system completes hydrogen production, to make the energy scheduling of the hydrogen production system more rational, this invention rationally allocates hydrogen energy based on the demand response of the hydrogen load, as shown in the flowchart below. Figure 5As shown, the specific power allocation strategy is as follows: When the hydrogen production of the mixed electrolyzer exceeds the system's hydrogen demand response, the remaining hydrogen production is the difference between the two. To improve energy utilization and avoid energy waste, this invention introduces a hydrogen storage tank to participate in the system's energy management. If the current hydrogen charge state is greater than or equal to 0.9, the remaining hydrogen production is treated as discarded energy. If the current hydrogen charge state is less than 0.9, the remaining hydrogen production can be stored in the hydrogen storage tank, but it needs to be determined whether the hydrogen charging power limit is met. If it exceeds the limit, the hydrogen charging power is set at the upper limit value, and the final discarded energy is obtained from the initial remaining hydrogen. If the hydrogen production is less than the upper limit of the hydrogen storage tank's charging power, and the limit is not exceeded, the remaining hydrogen production can be stored in the hydrogen storage tank. However, when the hydrogen production of the mixed electrolyzer is less than the demand after the hydrogen demand response, if the hydrogen charge of the hydrogen storage tank is greater than or equal to 0.1, the hydrogen storage tank can release hydrogen within a safe limit to meet the demand in order to further meet the load demand. However, if the released hydrogen production exceeds the hydrogen release limit of the hydrogen storage tank, the released hydrogen amount is set as the hydrogen release limit of the hydrogen storage tank, and the hydrogen shortage is the limit plus the mixed hydrogen production production minus the hydrogen load demand. If the hydrogen charge of the hydrogen storage tank is less than 0.1, the mixed hydrogen production production minus the hydrogen demand is the hydrogen shortage of the system.

[0187] In step two, an energy scheduling model for the hydrogen production system with the lowest daily operating cost is constructed by combining day-ahead source load forecast data and energy management strategies. However, this invention considers the impact of source load uncertainty in the scheduling model and incorporates the scenario probabilities of source load uncertainty into the system's day-ahead objective function.

[0188] The calculation formulas for various cost indicators in scenarios with source load uncertainty, and the calculation formula for the total cost considering source load uncertainty, are shown in the following formulas.

[0189] Formula 28;

[0190] Formula 29;

[0191] Formula 30;

[0192] Formula 31;

[0193] Formula 32;

[0194] Formula 33;

[0195] In the formula: The overall operating cost of the system should take into account the uncertainty of source load. This represents the total operating cost in scenario S. Let S be the uncertain probability of the S-th scenario; For the system's operation and maintenance costs; The operating cost of the hybrid electrolytic cell array; The cost of system energy curtailment penalties; To compensate for the cost of hydrogen load demand response; These are the operation and maintenance cost coefficients for batteries, hydrogen storage tanks, photovoltaic systems, and wind turbines, respectively. These represent the operating outputs of the battery, hydrogen storage tank, photovoltaic system, and wind turbine at time t, respectively. The usage costs for AEL and PEM are respectively; The startup and shutdown costs for AEL and PEM are respectively; These are the penalty coefficients for system energy curtailment / deficit and hydrogen curtailment / deficit, respectively. These represent the magnitude of the system's abandoned / deficient electrical energy and abandoned / deficient hydrogen energy, respectively.

[0196] These are the transferable compensation coefficient and the reduction compensation coefficient for hydrogen load, respectively. These represent the hydrogen load that can be transferred at time t and the hydrogen load that can be reduced during time period t, respectively.

[0197] The relevant constraints of the scheduling model must satisfy the balance between electricity and hydrogen power, and all equipment in the hydrogen production system must also operate within safe limits.

[0198] In step three, considering the differences in source load output at different time scales, and to reduce the impact of source load output deviations at different time scales on system operation scheduling, the prediction time domain is shortened to 4 hours and the time scale to 15 minutes based on the day-ahead scheduling, and intraday scheduling is carried out on the system. However, during the scheduling process, this invention fully considers the differences in operating characteristics between the two types of electrolyzers and to avoid frequent start-ups and shutdowns of alkaline electrolyzers. Therefore, the start-up and shutdown status of alkaline electrolyzers remains unchanged at this stage, and only the output is adjusted based on the existing status. Proton exchange membrane electrolyzers can be started, stopped, and their output adjusted according to their operating advantages. At the same time, the batteries can ensure that the state of charge is within a reasonable operating range to assist in the energy regulation of the hydrogen production system. Secondly, under the requirement of prioritizing the fulfillment of load demand, this invention only considers the situation where hydrogen load demand can be reduced at this stage. Finally, the impact of source load uncertainty is considered in the scheduling model, and the scenario probability of source load uncertainty is incorporated into the objective function.

[0199] The calculation expressions for various cost indicators in scenarios with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are as follows:

[0200] Formula 34;

[0201] Formula 35;

[0202] Formula 36;

[0203] Formula 37;

[0204] Formula 38;

[0205] Formula 39;

[0206] Formula 40;

[0207] In the formula: These are the penalty costs for power fluctuation deviations in the electrolyzer, battery, and hydrogen storage tank, respectively.

[0208] The figures represent the power fluctuations of the alkaline electrolyzer, proton exchange membrane electrolyzer, battery, and hydrogen storage tank at time t.

[0209] The relevant constraints of the scheduling model still need to satisfy the balance of electricity and hydrogen power, and each device also needs to operate within the safety limits.

[0210] Entering the real-time phase, this invention considers the characteristics of rapid response and large variation in instantaneous power fluctuations. Frequent changes to the start-up and shutdown states of the battery and proton exchange membrane electrolyzer during the real-time phase would not only severely impact equipment lifespan but also significantly affect system operation and scheduling. Therefore, this invention fully leverages the advantages of supercapacitors—rapid response and good instantaneous smoothing effect—and formulates a real-time scheduling model for supercapacitors based on the variable baseline criterion. In step four, the processing method on the load side during the real-time phase is the same as in the intraday phase. Finally, in the scheduling model, this invention considers the impact of source-load uncertainty and incorporates the scenario probability of source-load uncertainty into the objective function. The relevant constraints of the scheduling model must satisfy the balance between electricity and hydrogen power, and each device must operate within safe limits.

[0211] The calculation expressions for various cost indicators in the scenario with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are shown below:

[0212]

[0213] Formula 41;

[0214]

[0215] Formula 42;

[0216] Formula 43;

[0217]

[0218] Formula 44;

[0219]

[0220] Formula 45;

[0221] Formula 46;

[0222] Formula 47;

[0223]

[0224] Formula 48;

[0225] Formula 49;

[0226] In the formula: These are the real-time power of wind and solar power and the power after smoothing, respectively. These represent the charging power and discharging power of the supercapacitor, respectively. The penalty cost for curtailing wind and solar power in real time; The baseline power of the supercapacitor during time period t is given.

Claims

1. A multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response, characterized in that: The proposed method uses kernel density estimation, Frank-Copula function, and mean clustering algorithm to process source-load prediction data for the hybrid hydrogen production system. Simultaneously, it establishes a hydrogen load demand response model based on price incentives. Subsequently, a Lagrange optimization-based rotating control strategy for the electrolyzer array is adopted to fully leverage the wide-range operating characteristics of the modular hydrogen production cluster. Finally, a multi-timescale scheduling method for the hybrid hydrogen production system, considering hydrogen load demand response and aiming to minimize the total system operating cost, is proposed. This method allows different electrolyzers to complement each other in terms of economic efficiency and dynamic response characteristics, enabling batteries to provide power regulation and supercapacitors to respond quickly. Furthermore, in day-ahead scheduling, a multi-energy coordination energy management strategy considering hybrid electrolyzers, hydrogen storage tanks, batteries, and hydrogen load is formulated to achieve rational energy allocation within the system.

2. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 1, characterized in that: The hybrid hydrogen production system uses an electro-hydrogen production unit to form a hybrid hydrogen production array. The hydrogen production topology of the hybrid hydrogen production system includes a new energy supply layer, a green hydrogen production layer, and a hydrogen load demand layer. The scheduling method first considers the uncertainty of prediction errors on both the source and load sides to obtain typical source and load scenarios. Simultaneously, a hydrogen load demand response model is established on the demand side, optimizing hydrogen load demand for each time period through price incentives and user demand satisfaction constraints. Secondly, batteries and supercapacitors in the green hydrogen production layer are configured to play an auxiliary regulatory role at different time scales. When new energy fluctuations are severe and hydrogen load demand is high, batteries and supercapacitors maintain the power supply of the hybrid electrolyzer and the stability of the bus voltage through their own charging and discharging. Finally, by differentiating the hydrogen production advantages of AEL and PEMEL, in the green hydrogen production layer... The hydrogen production layer employs a hybrid electrolyzer array. During the hydrogen production process, both types of electrolyzers are divided into small-capacity unit combinations, and each unit is numbered to broaden the power operating range of the electrolyzers. Furthermore, a rotating operation strategy is adopted for the numbered units within each type of electrolyzer array, with their operation time and power allocation varying across different time periods to ensure uniform operating time and lifespan for each unit. Additionally, the total power distribution between the alkaline electrolyzer and the proton exchange membrane electrolyzer utilizes a hybrid electrolyzer power allocation strategy, coordinated appropriately with the hydrogen storage tank, to guarantee hydrogen load supply and new energy consumption.

3. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 2, characterized in that: The method is based on the hydrogen production topology of a hybrid hydrogen production system. If there are significant differences in power generation and consumption of the source load, different hydrogen production equipment, and different energy storage units at different time scales, a multi-time scale energy management strategy for the hybrid hydrogen production system with coordinated interaction of multiple energy devices is adopted. Specifically, during the day-ahead phase, with a preset duration as the optimization scale and hydrogen load user satisfaction as the constraint, the system operating cost is minimized by coupling the advantages of different hydrogen production equipment and the auxiliary power regulation role of the battery under safe charging state. By considering the reduceable characteristics of load demand response, the rapid adjustment and frequent start-stop characteristics of the electrolyzer, and the auxiliary characteristics of the battery, the output deviation between the day-ahead and intraday phases is minimized. Finally, with a preset short duration as the time scale, the goal is to minimize the output cost of the supercapacitor and the output deviation of other equipment to further smooth out the instantaneous power fluctuations of wind and solar power and improve energy utilization. The multi-timescale scheduling method for hybrid hydrogen production systems adopts a multi-timescale energy management strategy for hybrid hydrogen production systems based on the coordination of multiple energy devices, including the following steps; Step 1: Generate typical source-load scenarios and establish a hydrogen load demand response model; Step 2: Establish a day-ahead-scale hybrid hydrogen production system scheduling model based on multi-device coordination; including the system's day-ahead energy management strategy and the system's day-ahead objective function; Step 3: Establish a day-scale hybrid hydrogen production system scheduling model based on multi-device coordination; Step 4: Establish a real-time scheduling model for a hybrid hydrogen production system based on supercapacitor regulation.

4. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response as described in claim 3, characterized in that: In step one, when generating typical source-load scenarios, kernel density estimation, Frank-Copula function, and mean clustering algorithm are used to process the source-load prediction data. For the hydrogen load side, the uncertainty of hydrogen demand and hydrogen price and their negative correlation characteristics are combined to construct typical scenarios of hydrogen load and hydrogen price, and then reasonably characterize the uncertainty of the two. The specific formulas are shown in formulas (1) to (6) below. For the source side, the corresponding algorithm is used to incorporate the uncertainty of wind and solar power output and the characteristic that wind and solar power output in the same region often have a negative correlation and complementary relationship. The specific formulas are shown in formulas (7) to (12) below. Official 1; Official 2; Official 3; In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the average hydrogen load demand at time j; Let be the average hydrogen load price at time j; The hydrogen load demand at time j on day d; The hydrogen load price is shown on day d at time j; σ is the standard deviation of the sample data. Official 4; Official 5; Official 6; In the formula: , respectively, are the marginal distribution functions of hydrogen load demand and hydrogen load price; C is the Copula join function; Let be the joint distribution function of hydrogen load and hydrogen price at time j; Official 7; Official 8; Official 9; In the formula: D is the sample size, i.e., the number of days in a year; j is the number of time points; h is the kernel window width; K is the kernel function; Let be the sample mean of wind power output at time j; Let be the sample mean of photovoltaic power output at time j; The wind power output at time j on day d; The photovoltaic output at time j on day d; σ is the standard deviation of the sample data; Official 10; Official 11; Official 12; In the formula: These are the edge distribution functions of wind power output and photovoltaic power output, respectively; C is the Copula connection function. Let be the joint distribution function of wind and solar power output at time j.

5. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 3, characterized in that: In step one, a hydrogen load demand response model based on price incentives is used to participate in the game between hydrogen load demand and hydrogen load price through resource scheduling. The specific methods include: when energy prices change or system reliability decreases, after the hydrogen user receives a direct compensation notice from the hydrogen producer to induce a reduction in load or a notice to increase the hydrogen load price, the hydrogen user reduces or transfers the hydrogen load for a certain period of time, thereby realizing a positive interaction between the energy supply layer and the energy application layer. After considering the IDR model, the system hydrogen demand is divided into basic hydrogen load, reducible hydrogen load and transferable hydrogen load. The hydrogen load demand response model based on IDR is shown in the following formulas (13)-(16). Official 13; Official 14; Official 15; Official 16; In the formula: These are the self-elasticity coefficient and cross-elasticity coefficient of hydrogen load price, respectively; These represent the adjustment amount of hydrogen load at time t and the original demand for hydrogen load, respectively. They are respectively The hydrogen price and its change at any given time.

6. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 3, characterized in that: In step two, the system's day-ahead energy management strategy is specifically as follows: when the hydrogen production system meets the requirements... When the output of wind and solar power is greater than or equal to the power limit of the hydrogen production system, if the state of charge of the battery at the previous moment has reached the upper limit of the state of charge, both types of hydrogen production systems will operate at full load power, and the excess power will be included in the curtailment of wind and solar power. If the state of charge of the battery at the previous moment has not reached the upper limit, the hydrogen production system will also operate at full load power, and the excess power will be transferred to charge the lithium battery. If the excess power exceeds the power limit of the current period, the charging power will be set as the upper limit of the charging power for that period, and the remainder will be treated as curtailment of wind and solar power. If the excess power does not exceed the power limit of the current period, the battery charging power will be equal to the excess power. When the hydrogen production system meets But at the same time, it satisfies When the wind and solar power output is less than the upper limit of the hydrogen production system's power but greater than or equal to the lower limit of hydrogen production power, if the battery's state of charge (SOC) was greater than the safe upper limit in the previous moment, then the wind and solar power will be allocated to the hybrid hydrogen production array for power distribution, thereby producing green hydrogen. If the battery's SOC is between 0.6 and 0.8 and the battery was charging in the previous moment, and to effectively avoid overcharging and over-discharging of the battery, all wind and solar power will be allocated to the hybrid hydrogen production power distribution strategy in this moment. However, if the battery was discharging in the previous moment, it is necessary to determine whether the electrolyzer was operating at the minimum hydrogen production power in the previous moment. If the electrolyzer array was operating at the minimum hydrogen production power in the previous moment... To avoid frequent start-ups and shutdowns of the electrolyzer array and prolonged standby, all power at this moment is allocated to the hybrid hydrogen production array for green hydrogen production. Otherwise, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production. If the battery's state of charge was below 0.6% at the previous moment, the battery is prioritized for charging. If the remaining power is greater than or equal to the lower limit of hydrogen production power, it is allocated to the hybrid hydrogen production array. Otherwise, based on the battery's state of charge, it is discharged to provide auxiliary power for hybrid array hydrogen production. When the hydrogen production system meets Then, under the premise of ensuring the safe operation of the battery, it is necessary to discharge to assist in supplying hydrogen production power. If the state of charge of the battery at the previous moment is greater than or equal to 0.2, the battery will discharge to supply power for the mixed hydrogen production array power distribution. If the battery is less than 0.2, in order to reduce the waste of wind and solar power, the current wind and solar power needs to be divided by the single-cell power limit to obtain the minimum number of operable cells. Then, the start-up and shutdown are allocated according to the characteristics of the two types of electrolyzers. In the allocation process, in order to reduce the number of start-ups and shutdowns of alkaline electrolyzers, the number of alkaline electrolyzers that need to be operated should be prioritized, and the number of alkaline electrolyzers that need to be shut down should be reduced. The proton exchange membrane electrolyzers should be started and stopped in an emergency. Finally, the waste power is calculated.

7. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 6, characterized in that: In the power allocation strategy for the mixed electrolyzers in a hybrid hydrogen production system, when the total hydrogen production power of the hybrid system is greater than or equal to the full-load power of the alkaline electrolyzer, to ensure stable base-load operation of the alkaline electrolyzer, it is prioritized to operate at full-load power, with the remaining power allocated to the proton exchange membrane electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer but less than or equal to the full-load power of the proton exchange membrane electrolyzer, to reduce frequent start-ups and shutdowns of the two types of electrolyzers, the hydrogen production power should be evenly distributed between the proton exchange membrane electrolyzer and the alkaline electrolyzer. When the system's hydrogen production power is less than the full-load power of the alkaline electrolyzer but greater than the full-load power of the proton exchange membrane electrolyzer, this stage allows for more flexible response... When using a hybrid hydrogen production array, it is still necessary to check whether the hydrogen production power is greater than or equal to twice the full-load power of the proton exchange membrane electrolyzer. If the hydrogen production power is less than or equal to twice the full-load power of the proton exchange membrane electrolyzer, the full-load power of the proton exchange should be preferentially allocated to the alkaline electrolyzer array, and the remaining power should be allocated to the proton exchange membrane electrolyzer array. The purpose is to avoid the alkaline electrolyzer being in a long-term overload state, while maintaining the operating state of the proton exchange membrane electrolyzer to avoid frequent start-ups and shutdowns. Conversely, if the full-load power is preferentially allocated to the proton exchange membrane, the remaining power should be allocated to the alkaline electrolyzer, thereby reducing the amount of wasted electricity. Finally, the power of the alkaline electrolyzer and the power of the proton exchange membrane electrolyzer are respectively transferred to the array operation module. In hydrogen production systems, to further enhance the flexibility of electrolyzers and ensure the consistency of operating status and lifespan of individual electrolyzers, a Lagrange-optimized rotating operation strategy for the electrolyzer array is used. This strategy includes the following control scenarios using proton exchange membrane electrolyzers, where P... EL The total hydrogen production power of the input proton exchange membrane system needs to be allocated to the array power. Scenario 1: Overload Condition At this time, all electrolytic cells are operating at full capacity. Even at this time, the requirements could not be met. Therefore, considering the overload characteristics of the electrolytic cells, some electrolytic cells were put into overload mode. The number of electrolytic cells operating in overload mode, m, and the remaining power were determined. The expression is as follows: Official 17; Official 18; In the formula: "floor" is the floor function; Based on the allocation, the m units waiting to work at the corresponding time will be placed at the highest overload power operation; Assigned to the (m+1)th monomer; the remaining monomers are... run. Official 19; In the formula: This represents the operating power of the nth cell at the corresponding time. Based on its operating characteristics, the overload rate of a single electrolytic cell should be less than or equal to 0.

5. Official 20; In the formula: λn is the overload rate of electrolytic cell n; These are the total operating time and overload operating time of electrolytic cell n, respectively. Scenario 2: Stable operating conditions This operating condition requires both ensuring balanced operation of each unit and improving overall hydrogen production efficiency. A power allocation strategy based on the Lagrange multiplier method is adopted, expressed as follows: Official 21; In the formula: f is the overall efficiency at the corresponding time; The power of the i-th unit actually participating in hydrogen production at the corresponding moment; Let be the input operating power of the i-th unit at the corresponding time. Official 22; Official 23; In the formula: The Lagrangian function is constructed; For Lagrange multipliers; For constraints; They are respectively For variables Differentiate; φ represents the variables respectively. Differentiate; Scenario 3: Low power operation At this time, P EL It is insufficient to support each unit to work in standby mode, but in order to reduce the number of units that are shut down and the downtime, the units are made to work in standby mode to the greatest extent possible. Let m be the number of units in operation and the remaining power be... The expression is as follows: Official 24; Official 25; To make rational use Under the premise of ensuring the safe operation of the electrolytic cell, The unit is assigned to the first unit to be worked at the corresponding time. The initial assignment for this working condition is as follows: Official 26; Outer layer rotation: To avoid the leading cells being under overload for extended periods and the trailing cells being under shutdown for extended periods, resulting in an imbalance in the working time and operating status of the cells within the array, an array rotation optimization strategy is adopted. The specific formula is as follows: Official 27; In the formula: Let be the ID of the single entity at time t-1 that is at the k-th ID; n is the total number of single entities in the array; mod is the modulo operation, i.e., taking the remainder when k-1 is divided by n. , where q and n are both integers; initially, the self-number is selected; After the system completes hydrogen production, hydrogen energy is allocated based on the demand response of the hydrogen load. The specific power allocation strategy is as follows: When the hydrogen production of the mixed electrolyzer exceeds the system's hydrogen demand response, the remaining hydrogen production is the difference between the two. A hydrogen storage tank is introduced to participate in the system's energy management. If the current state of hydrogen charge is greater than or equal to 0.9, the current remaining hydrogen production is treated as waste energy. If the current state of hydrogen charge is less than 0.9, the remaining hydrogen production is stored in the hydrogen storage tank, but it is necessary to determine whether the hydrogen charging power limit is met. If the limit is exceeded, the hydrogen charging power is set at the upper limit value, and the final waste energy is the initial remaining amount. If the hydrogen production minus the upper limit of the hydrogen storage tank's charging power does not exceed the limit, the remaining hydrogen production can be stored in the hydrogen storage tank. However, when the hydrogen production of the mixed electrolyzer is less than the demand after the hydrogen demand response, if the hydrogen charge state of the hydrogen storage tank is greater than or equal to 0.1, in order to further meet the load demand, the hydrogen storage tank can release hydrogen within a safe limit to meet the demand. However, if the released hydrogen production exceeds the hydrogen release limit of the hydrogen storage tank, the released hydrogen amount is set as the hydrogen release limit of the hydrogen storage tank, and the hydrogen shortage is the limit plus the mixed hydrogen production production minus the hydrogen load demand. If the hydrogen charge state of the hydrogen storage tank is less than 0.1, then the mixed hydrogen production production minus the hydrogen demand is the hydrogen shortage of the system.

8. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 6, characterized in that: In step two, an energy scheduling model for the hydrogen production system with the lowest daily operating cost is constructed by combining day-ahead source load forecast data and energy management strategies. However, this invention considers the impact of source load uncertainty in the scheduling model and incorporates the scenario probabilities of source load uncertainty into the system's day-ahead objective function. The calculation formulas for various cost indicators in scenarios with source load uncertainty, and the calculation formula for the total cost considering source load uncertainty, are shown in the following formulas. Official 28; Official 29; Official 30; Official 31; Official 32; Official 33; In the formula: The overall operating cost of the system should take into account the uncertainty of source load. This represents the total operating cost in scenario S. Let S be the uncertain probability of the S-th scenario; For the system's operation and maintenance costs; The operating cost of the hybrid electrolytic cell array; The cost of system energy curtailment penalties; To compensate for the cost of hydrogen load demand response; These are the operation and maintenance cost coefficients for batteries, hydrogen storage tanks, photovoltaic systems, and wind turbines, respectively. These represent the operating outputs of the battery, hydrogen storage tank, photovoltaic system, and wind turbine at time t, respectively. The usage costs for AEL and PEM are respectively; The startup and shutdown costs for AEL and PEM are respectively; These are the penalty coefficients for system energy curtailment / deficit and hydrogen curtailment / deficit, respectively. These represent the magnitude of the system's abandoned / deficient electrical energy and abandoned / deficient hydrogen energy, respectively. These are the transferable compensation coefficient and the reduction compensation coefficient for hydrogen load, respectively. These represent the hydrogen load that can be transferred at time t and the hydrogen load that can be reduced during time period t, respectively. The relevant constraints of the scheduling model must satisfy the balance between electricity and hydrogen power, and all equipment in the hydrogen production system must also operate within safe limits.

9. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 6, characterized in that: In step three, considering the differences in source load output at different time scales, and to reduce the impact of source load output deviations at different time scales on system operation scheduling, the predicted time domain and time scale are shortened based on the day-ahead scheduling, and intraday scheduling is implemented for the system. Considering the differences in operating characteristics between the two types of electrolyzers and to avoid frequent start-ups and shutdowns of alkaline electrolyzers, the start-up and shutdown status of alkaline electrolyzers remains unchanged at this stage, with only output adjustments made based on the existing status. Proton exchange membrane electrolyzers are started up, shut down, and have their output adjusted according to their operational advantages, while batteries ensure that the state of charge is within a reasonable operating range to assist in the energy regulation of the hydrogen production system. Secondly, under the requirement of prioritizing load demand, only the reduction of hydrogen load demand is considered at this stage. Finally, the impact of source load uncertainty is considered in the scheduling model, and the probability of source load uncertainty scenarios is incorporated into the objective function. The calculation expressions for various cost indicators in scenarios with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are as follows: Official 34; Official 35; Official 36; Official 37; Official 38; Official 39; Official 40; In the formula: These are the penalty costs for power fluctuation deviations in the electrolyzer, battery, and hydrogen storage tank, respectively. The figures represent the power fluctuations of the alkaline electrolyzer, proton exchange membrane electrolyzer, battery, and hydrogen storage tank at time t. The relevant constraints of the scheduling model still need to satisfy the balance of electricity and hydrogen power, and each device also needs to operate within the safety limits.

10. The multi-timescale scheduling method for a hybrid hydrogen production system considering hydrogen load demand response according to claim 6, characterized in that: In step four, the processing method on the load side during the real-time phase is the same as that during the intraday phase; the relevant constraints of the scheduling model must meet the power balance of electricity and hydrogen, and all equipment in the hybrid hydrogen production system must operate within the safety limits. The calculation expressions for various cost indicators in the scenario with source load uncertainty, and the calculation expression for the total cost considering source load uncertainty, are shown below: Formula 41; Formula 42; Official 43; Formula 44; Formula 45; Official 46; Official 47; Formula 48; Official 49; In the formula: These are the real-time power of wind and solar power and the power after smoothing, respectively. These represent the charging power and discharging power of the supercapacitor, respectively. The penalty cost for curtailing wind and solar power in real time; The baseline power of the supercapacitor during time period t is given.

Images