Load response control method and device for hydrogen production system by electrolysis of water

Abstract

The application discloses a load response control method and device of a water electrolysis hydrogen production system. First, the actual input power of the water electrolysis hydrogen production system in a historical period and meteorological prediction data are obtained, then a power prediction curve in a future time is predicted through a power model, and then a dynamic coupling model is solved based on the power prediction curve, so that the load response speed, temperature stability and pressure stability meet the corresponding optimization target, the temperature control sequence and the pressure control sequence are obtained, and then the working state of the water electrolysis hydrogen production system is adjusted in advance based on the temperature control sequence and the pressure control sequence. Not only can the load response speed be improved, but also the sudden change of temperature and pressure can be avoided, so that the system can quickly and stably respond, and in the scene of renewable energy power fluctuation, the timeliness of the load response of the water electrolysis hydrogen production system and the wide power range operation stability are effectively improved.

Description

Technical Field

[0001] This application relates to the field of water electrolysis technology, specifically to a load response control method and apparatus for a water electrolysis hydrogen production system. Background Technology

[0002] While the installed capacity of renewable energy (wind power and solar power) continues to grow, the output power of wind and solar power is highly volatile and intermittent (affected by changes in wind speed and sunlight), making large-scale grid connection and absorption difficult. Electrolysis of water to produce hydrogen, as a flexible energy storage method, can convert the electrical energy from renewable energy sources into hydrogen energy for storage, becoming one of the key technologies for solving the problem of renewable energy absorption.

[0003] However, existing water electrolysis hydrogen production systems (especially proton exchange membrane electrolyzers and alkaline electrolyzers) have significant technical bottlenecks: First, load response is lagging. The temperature, pressure, and other parameters of the electrolyzer require a certain amount of time to adjust. When wind / solar power changes suddenly, the electrolyzer cannot quickly track load changes, which can easily lead to power curtailment or system overload. Second, the stability of operation over a wide power range is poor. When the electrolyzer operates at low power (e.g., less than 20% of rated power), membrane electrode drying and uneven current density are likely to occur. When operating at high power (e.g., more than 80% of rated power), over-temperature and over-pressure are likely to occur, affecting the stability of operation. Third, load fluctuations lead to system lifespan loss. Frequent power changes can cause drastic fluctuations in temperature and pressure, accelerating membrane electrode aging and electrolyzer corrosion, thus shortening the system's service life.

[0004] In related technologies, improvement schemes for the load response of hydrogen production by water electrolysis are mostly concentrated on proportional-integral-derivative (PID) control or adaptive control, which can achieve passive response to the current power.

[0005] However, the above methods have slow response speed and poor stability for loads with a wide power range. Summary of the Invention

[0006] In view of this, this application provides a load response control method and apparatus for a water electrolysis hydrogen production system, which can not only improve the load response speed, but also avoid sudden changes in temperature and pressure, enabling the system to respond quickly and smoothly. In the scenario of renewable energy power fluctuations, it effectively improves the timeliness of load response and the operational stability of the water electrolysis hydrogen production system over a wide power range.

[0007] To solve the above problems, the technical solution provided in this application is as follows:

[0008] On one hand, embodiments of this application provide a load response control method for a water electrolysis hydrogen production system, the method comprising:

[0009] Acquire historical power data and meteorological forecast data for the water electrolysis hydrogen production system. The historical power data is used to identify the actual input power over a historical period.

[0010] Based on historical power data and meteorological forecast data, a power forecast curve for the future is obtained through a power forecast model.

[0011] Based on the power prediction curve, the dynamic coupling model is solved to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives. The multiple optimization objectives include the optimization objectives corresponding to load response speed, temperature stability and pressure stability. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer.

[0012] Based on temperature and pressure control sequences, the temperature and pressure of the water electrolysis hydrogen production system are pre-adjusted within a preset time before power fluctuations are predicted in the power prediction curve.

[0013] As one possible implementation, the above method also includes:

[0014] A lifespan loss model for a water electrolysis hydrogen production system was constructed, with temperature fluctuation amplitude and pressure shock frequency as the core variables.

[0015] Based on the power prediction curve, the dynamic coupling model is solved to obtain temperature control sequences and pressure control sequences that satisfy multiple optimization objectives, including:

[0016] Based on the power prediction curve, the dynamic coupling model and the lifetime loss model are jointly optimized and solved to obtain temperature control sequence and pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

[0017] As one possible implementation, in the process of solving the dynamic coupling model based on the power prediction curve to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives, the above method also includes:

[0018] When the power change amplitude in the power prediction curve is greater than the first preset threshold, the weighting coefficient corresponding to the load response speed is increased; when the power change amplitude is greater than or equal to the first preset threshold, the weighting coefficient corresponding to the degree of lifespan loss is increased.

[0019] As one possible implementation, the above method also includes:

[0020] Obtain historical meteorological data;

[0021] Construct an initial power prediction model;

[0022] Based on historical meteorological data and historical power data, the initial power prediction model is trained to obtain the power prediction model.

[0023] As one possible implementation, the above method also includes:

[0024] Based on the anode temperature, cathode temperature, internal pressure, and membrane electrode humidity of the electrolyzer, with temperature and pressure as control variables and actual input power as output variable, an initial dynamic coupling model is constructed.

[0025] Based on the actual input power under different temperatures and pressures, the parameters of the dynamic coupling model are adjusted to obtain the dynamic coupling model.

[0026] In another aspect, embodiments of this application provide a load response control device for a water electrolysis hydrogen production system, the device comprising:

[0027] The acquisition unit is used to acquire historical power data and meteorological forecast data of the water electrolysis hydrogen production system. The historical power data is used to identify the actual input power within a historical period.

[0028] The power prediction unit is used to obtain the power prediction curve for the future time period based on historical power data and meteorological forecast data through a power prediction model.

[0029] The optimization solution unit is used to solve the dynamic coupling model based on the power prediction curve to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives. The multiple optimization objectives include the optimization objectives corresponding to load response speed, temperature stability and pressure stability. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer.

[0030] An adjustment unit is used to pre-adjust the temperature and pressure of the water electrolysis hydrogen production system within a preset time before predicting power fluctuations in the power prediction curve, based on a temperature control sequence and a pressure control sequence.

[0031] As one possible implementation, the device also includes a building unit for constructing a lifespan loss model of the water electrolysis hydrogen production system, with temperature fluctuation amplitude and pressure shock number as core variables.

[0032] The optimization solution unit is used to jointly optimize and solve the dynamic coupling model and the lifetime loss model based on the power prediction curve, so as to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

[0033] As one possible implementation, optimizing the solution element is also used for:

[0034] When the power change amplitude in the power prediction curve is greater than the first preset threshold, the weighting coefficient corresponding to the load response speed is increased; when the power change amplitude is greater than or equal to the first preset threshold, the weighting coefficient corresponding to the degree of lifespan loss is increased.

[0035] As one possible implementation, the device also includes a first model training unit for:

[0036] Obtain historical meteorological data;

[0037] Construct an initial power prediction model;

[0038] Based on historical meteorological data and historical power data, the initial power prediction model is trained to obtain the power prediction model.

[0039] As one possible implementation, the device also includes a second model training unit for:

[0040] Based on the anode temperature, cathode temperature, internal pressure, and membrane electrode humidity of the electrolyzer, with temperature and pressure as control variables and actual input power as output variable, an initial dynamic coupling model is constructed.

[0041] Based on the actual input power under different temperatures and pressures, the parameters of the dynamic coupling model are adjusted to obtain the dynamic coupling model.

[0042] In another aspect, this application provides a computer device, which includes a processor and a memory:

[0043] Memory is used to store computer programs;

[0044] The processor is used to execute any of the above methods according to the computer program.

[0045] In another aspect, this application provides a computer-readable storage medium for storing a computer program that, when executed by a computer device, implements the method described in any of the above-mentioned embodiments.

[0046] In another aspect, this application provides a computer program product that includes a computer program, which, when run on a computer device, causes the computer device to perform any of the methods described above.

[0047] As can be seen from the above technical solution, this solution first obtains the actual input power and meteorological forecast data of the water electrolysis hydrogen production system over a historical period. Then, based on the historical power data and meteorological forecast data, it predicts the power forecast curve for the future time period through a power model to obtain the future load change trend. Subsequently, it solves the dynamic coupling model based on the power forecast curve. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer. Based on the power forecast curve, it is possible to obtain the temperature control sequence and pressure control sequence to meet the corresponding optimization objectives of load response speed, temperature stability and pressure stability. Then, based on the temperature control sequence and pressure control sequence, the working state of the water electrolysis hydrogen production system is pre-adjusted, which can not only improve the load response speed, but also avoid sudden changes in temperature and pressure, so that the system can respond quickly and stably. In the scenario of renewable energy power fluctuation, it effectively improves the timeliness of load response and the stability of operation over a wide power range of the water electrolysis hydrogen production system. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A schematic flowchart illustrating a load response control method for a water electrolysis hydrogen production system provided in this application embodiment;

[0050] Figure 2 This is a schematic diagram of a load response control device for a water electrolysis hydrogen production system provided in an embodiment of this application. Detailed Implementation

[0051] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0052] As described in the background section, most improvement schemes for the load response of hydrogen production via water electrolysis focus on PID control or adaptive control, which can only achieve a passive response to the current power and cannot anticipate power fluctuations in renewable energy sources. While some schemes introduce power prediction, they do not fully consider the dynamic coupling relationship between temperature, pressure, and power, resulting in slow response speed and poor stability over a wide power range. Specifically, these drawbacks include the following:

[0053] (1) Insufficient wide power range coverage: It can only operate stably in the range of 30%-90% of the rated power, and cannot achieve full range coverage of 10%-100%. At low power (rated power <30%), the membrane electrode is prone to drying and uneven current density. At high power (rated power >90%), over-temperature and overload are prone to occur. It cannot adapt to scenarios with large fluctuations in wind power / photovoltaic power (such as photovoltaic power approaching 0 at night, or wind power suddenly increasing from 10% to 100% due to gusts).

[0054] (2) The coupling relationship between temperature, pressure and power is not utilized: only temperature and pressure are used as fixed constraints, and no dynamic coupling model of the three is established. Temperature and pressure are not optimized in advance as control variables, which leads to power tracking relying on a single current regulation. The response speed is limited by the lag of the chemical reaction in the electrolyzer (only reaching the minute level), and it is impossible to achieve a fast response of seconds to minutes.

[0055] (3) Single optimization objective: Only power tracking accuracy is considered, without taking into account the stability of temperature and pressure fluctuations, resulting in large temperature and pressure fluctuations when the load changes suddenly (temperature fluctuations can reach ±8℃, and pressure fluctuations can reach ±0.1MPa), which further aggravates the wear and tear of the system life.

[0056] (4) Insufficient coordination between prediction and control: Power prediction is only used for "trend reference" and is not combined with advance adjustment of temperature and pressure parameters (such as when the power is predicted to increase suddenly, the temperature and pressure are not increased in advance to adapt to the high power operation state). It is still "passive tracking" rather than "active prediction", and the response speed is limited.

[0057] To address the aforementioned issues, this application provides a load response control method for a water electrolysis hydrogen production system. First, it acquires the actual input power and meteorological forecast data of the water electrolysis hydrogen production system over a historical period. Then, based on the historical power data and meteorological forecast data, it predicts the power forecast curve for the future time period using a power model, obtaining the future load change trend. Next, it solves a dynamic coupling model based on the power forecast curve. This dynamic coupling model describes the coupling relationship between temperature, pressure, and power during the operation of the electrolyzer. Furthermore, based on the power forecast curve, it obtains the temperature control sequence and pressure control sequence that satisfy the corresponding optimization objectives for load response speed, temperature stability, and pressure stability. Based on these temperature and pressure control sequences, it pre-adjusts the operating state of the water electrolysis hydrogen production system, which not only improves the load response speed but also avoids sudden changes in temperature and pressure, enabling the system to respond quickly and smoothly. In scenarios with fluctuating renewable energy power, this effectively improves the timeliness of the load response and the operational stability over a wide power range of the water electrolysis hydrogen production system.

[0058] The solutions provided in this application relate to the field of water electrolysis technology, and are specifically illustrated through the following embodiments.

[0059] See Figure 1 The diagram shown is a schematic flow chart of a load response control method for a water electrolysis hydrogen production system provided in an embodiment of this application, including steps S101-S104.

[0060] S101: Obtain historical power data and weather forecast data for the water electrolysis hydrogen production system.

[0061] Historical power data is used to identify the actual input power over a historical period, referring to the actual electrical power consumed by the electrolyzer. This power is related to the output of renewable energy sources, such as wind / solar power, on the input side of the water electrolysis hydrogen production system, as well as the system's absorption capacity. Historical power data can reflect the system's actual operating characteristics, degradation trends, and the impact of non-meteorological factors.

[0062] Meteorological forecast data refers to meteorological data predicted for a future period, such as wind speed, light intensity, and ambient temperature. Meteorological data can affect input power. Based on the correlation between meteorological data and the corresponding actual input power, the input power for a future period can be determined by analyzing the mapping relationship. By combining historical power data, the input power can be corrected, achieving accurate prediction of the input power for a future period and obtaining a power prediction curve.

[0063] The "load response" of a water electrolysis hydrogen production system refers to the electrolyzer's ability to track and adapt to fluctuations in input power, that is, its ability to quickly change its operating power to match external demand after receiving power fluctuation signals from the grid or renewable energy sources. It can be characterized by adjustment time, steady-state error, and overshoot.

[0064] S102: Based on historical power data and meteorological forecast data, a power forecast curve for the future is obtained through a power forecast model.

[0065] A power prediction model refers to a model used to predict input power. This application does not impose specific limitations on this model; for example, it can be a model built based on an LSTM neural network or an ARIMA model. Based on the constructed power prediction model, regression prediction can be performed on the power over a future period to obtain a power prediction curve. For example, if the input is power data from the past hour and meteorological forecast data for the next hour, the output will be a power prediction curve for the next 5-60 minutes (time resolution of 1 second).

[0066] As one possible implementation, this application provides a specific implementation of the power prediction model, as detailed in A1-A3.

[0067] A1: Obtain historical meteorological data.

[0068] A2: Construct the initial power prediction model.

[0069] A3: Based on historical meteorological data and historical power data, the initial power prediction model is trained to obtain the power prediction model.

[0070] Historical meteorological data refers to meteorological data over a historical period, including wind speed, light intensity, and ambient temperature.

[0071] For example, historical meteorological data T={T1, T2, ..., Tn} of the location of the water electrolysis hydrogen production system are obtained. Then, an initial power prediction model is constructed based on LSTM, enabling the output power prediction model to model time series features. Subsequently, the initial power prediction model is trained based on historical meteorological data and historical power data W={W1, W2, ..., Wn}, so that the initial power prediction model can model the correlation between meteorological data and actual input power, as well as the variation characteristics of input power. Then, the predicted power Pi is obtained based on Ti and {W1, ..., Wi-1}. The parameters of the initial power prediction model are iteratively adjusted based on the error between the predicted power Pi and Wi. The power prediction model is verified through a test set. When the prediction error does not exceed 8%, the power prediction model is obtained, ensuring the reliability of the prediction results.

[0072] Therefore, by using historical meteorological data and historical power data, the prediction accuracy of power prediction models can be improved, thereby providing accurate data basis for load response control.

[0073] S103: Based on the power prediction curve, the dynamic coupling model is solved to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives.

[0074] Among them, several optimization objectives include optimization objectives corresponding to load response speed, temperature stability, and pressure stability, respectively.

[0075] Load response speed refers to the tracking response speed of the electrolytic cell to fluctuations in input power.

[0076] Temperature stability describes the amplitude of temperature fluctuations; for example, temperature stability can be determined based on upper and lower temperature limits. Pressure stability describes the amplitude of pressure fluctuations; for example, pressure stability can be determined based on upper and lower pressure limits. Temperature and pressure stability are used to reflect the stability of a system.

[0077] The optimization target for load response speed is to control the load response speed within the corresponding first preset range; the optimization target for pressure stability is to control the pressure fluctuation range within the corresponding second preset range; and the optimization target for temperature stability is to control the temperature fluctuation range within the corresponding third preset range.

[0078] During system operation, load response speed, temperature stability, and pressure stability are mutually balancing. This application does not impose specific limitations on the aforementioned three preset ranges. The purpose of multi-objective optimization is to minimize the load response speed and maintain stable temperature and pressure, i.e., to seek a balance between load response speed, temperature stability, and pressure stability. For example, the load response speed can be set to less than 30 seconds, the temperature fluctuation range can be controlled within 3 degrees Celsius, and the pressure fluctuation range can be controlled within 0.05 MPa, thus obtaining multiple optimization objectives.

[0079] The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of an electrolyzer, thus realizing the modeling of the dynamic response performance of the electrolyzer.

[0080] As one possible implementation, this application provides a method for constructing a dynamic coupling model, as detailed in B1-B2.

[0081] B1: Based on the anode temperature, cathode temperature, internal pressure, and membrane electrode humidity of the electrolyzer, with temperature and pressure as control variables and actual input power as the output variable, an initial dynamic coupling model is constructed.

[0082] B2: Based on the actual input power under different temperatures and pressures, the parameters of the dynamic coupling model are adjusted to obtain the dynamic coupling model.

[0083] For example, considering the heat and mass transfer processes inside the electrolytic cell and the electrochemical reaction characteristics of the membrane electrode, a nonlinear state-space model is constructed as the initial dynamic coupling model. The state variables include the anode / cathode temperature, internal pressure, and membrane electrode humidity of the electrolytic cell; the control variables are temperature (adjusted via heating / cooling modules) and pressure (adjusted via inlet / outlet valve groups); the output variable is the actual input power of the electrolytic cell. The model's parameters are adjusted using experimental data (actual input power at different temperatures and pressures) to ensure model accuracy (output error ≤ 3%), resulting in the final dynamic coupling model.

[0084] Therefore, a nonlinear state-space model is constructed to accurately describe the dynamic relationship between the temperature distribution, pressure distribution, membrane electrode humidity and actual input power inside the electrolyzer, so that the dynamic coupling model can accurately characterize the dynamic response relationship between pressure, temperature and power under the working state of the electrolyzer.

[0085] Temperature control sequence refers to a sequence used to control the temperature setpoint of the system, and pressure control sequence refers to a sequence used to control the pressure setpoint of the system.

[0086] It should be noted that, in the embodiments of this application, the temperature control sequence and the pressure control sequence are used to adjust the temperature and pressure to match the power after the fluctuation within a preset time before the power fluctuation. In other words, the temperature and pressure are raised or lowered in advance based on the temperature control sequence and the pressure control sequence, so that when the power fluctuates, the system can quickly and smoothly reach the temperature and pressure that match the predicted power.

[0087] By inputting the power prediction curve into the dynamic coupling model, optimization can be performed to obtain the temperature control sequence and the pressure control sequence. Thus, within the time period corresponding to the power prediction curve, the load response speed, pressure stability, and temperature stability can be taken into account.

[0088] S104: Based on the temperature control sequence and pressure control sequence, the temperature and pressure of the water electrolysis hydrogen production system are pre-adjusted within a preset time before the power fluctuation is predicted in the power prediction curve.

[0089] The preset time refers to the time for pre-adjusting the temperature and pressure. This application does not impose specific limitations on this, such as adjusting the temperature 5 minutes in advance, adjusting the pressure 3 minutes in advance, etc.

[0090] Based on temperature and pressure control sequences, temperature and pressure are controlled within a preset time before power surges, such as 5 minutes or 3 minutes. For example, the temperature and pressure are pre-increased, so that when the power surge occurs, the target temperature and pressure are quickly reached. This enables a fast and stable response to a wide range of power loads (such as 10% to 100% of the rated power).

[0091] Temperature and pressure can be controlled in the following ways: For example, a combination of electric heating and water cooling, equipped with a PID auxiliary controller, can be used to regulate the temperature, ensuring a response time of less than or equal to 2 seconds and a temperature control accuracy error within 0.1 degrees Celsius. Electric regulating valves (one at the inlet and one at the outlet) can be used to control the pressure, ensuring an adjustment accuracy error within 0.01 MPa and a response time of less than or equal to 1 second.

[0092] Furthermore, optimal control for multiple objectives is achieved through rolling optimization. For example, data is collected in real-time at a sampling frequency of 1Hz within the current control cycle, including wind / photovoltaic power (accuracy ±0.5%), electrolytic cell temperature (accuracy ±0.1℃), pressure (accuracy ±0.005MPa), and actual output power (accuracy ±0.5%). In the next control cycle, based on the power data from the current control cycle and the meteorological forecast data for the next control cycle, a power prediction curve for the next control cycle is obtained. Then, based on this power prediction curve, the pressure control sequence and temperature control sequence for the next control cycle are obtained by solving the dynamic coupling model, and these control variables are applied to the actuators for control. Thus, optimal control for multiple objectives is achieved through rolling optimization.

[0093] Therefore, this application first obtains the actual input power and meteorological forecast data of the water electrolysis hydrogen production system over a historical period. Then, based on the historical power data and meteorological forecast data, it predicts the power forecast curve for the future time period through a power model to obtain the future load change trend. Subsequently, it solves the dynamic coupling model based on the power forecast curve. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer. Based on the power forecast curve, it is possible to obtain the temperature control sequence and pressure control sequence to meet the corresponding optimization objectives of load response speed, temperature stability and pressure stability. Then, based on the temperature control sequence and pressure control sequence, the temperature and pressure of the water electrolysis hydrogen production system are pre-adjusted. This not only improves the load response speed, but also avoids sudden changes in temperature and pressure, enabling the system to respond quickly and stably. In the scenario of renewable energy power fluctuations, this effectively improves the timeliness of load response and the stability of operation over a wide power range of the water electrolysis hydrogen production system.

[0094] In related technologies, the response strategies for water electrolysis hydrogen production systems lack a lifespan guarantee mechanism and do not consider the impact of sudden current changes and temperature / pressure fluctuations on membrane electrode aging and electrolyzer corrosion, resulting in rapid system lifespan degradation under long-term operation. Based on this, this application provides an optimization method, specifically as described in C1-C2. C2 is a specific implementation of S103, which involves solving the dynamic coupling model based on the power prediction curve to obtain temperature and pressure control sequences that satisfy multiple optimization objectives.

[0095] C1: Using temperature fluctuation amplitude and pressure shock frequency as core variables, a lifespan loss model for the water electrolysis hydrogen production system is constructed.

[0096] C2: Based on the power prediction curve, the dynamic coupling model and the lifetime loss model are jointly optimized and solved to obtain temperature control sequence and pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

[0097] Temperature fluctuation amplitude describes the range or intensity of change in a water electrolysis hydrogen production system within a working time window. It can be expressed as temperature fluctuation amplitude by subtracting the maximum temperature value from the maximum temperature value. For example, if the temperature rises from 10 degrees Celsius to 15 degrees Celsius at time T1, the temperature fluctuation amplitude is 5 degrees Celsius.

[0098] The number of pressure shocks is used to describe the number of times the pressure of the water electrolysis hydrogen production system changes drastically during a period of operation. This application does not specifically limit this, for example, a pressure fluctuation exceeding a preset threshold can be recorded as a pressure shock.

[0099] The greater the temperature fluctuation or the more frequent the pressure shocks, the faster the membrane electrode aging, the more severe the corrosion of the electrolyzer body, and the greater the life loss value. The life loss model is used to model the degree of life loss of the water electrolysis hydrogen production system.

[0100] For example, working data over a period of time is acquired, and the amplitude of temperature fluctuations and the number of pressure shocks are statistically analyzed. The lifespan loss of each temperature change and pressure shock event is then calculated cumulatively. Specifically, the rainflow counting method is used to extract discrete temperature changes (including amplitude, mean, and number of changes) from the continuous temperature signal, while simultaneously identifying pressure shock events (including amplitude and number of pressure changes). Then, based on the failure mechanisms of various materials in the electrolytic cell (such as the aging degree of the membrane electrode and the corrosion degree of the electrolytic cell body), the lifespan loss caused by a single temperature cycle or a single pressure shock is calculated. A coupling coefficient is introduced to quantify the synergistic damage effect of temperature-pressure synchronous events, thus constructing a lifespan loss model.

[0101] Therefore, by constructing a lifespan loss model based on the temperature fluctuation amplitude and the number of pressure shocks, the degree of lifespan loss of the electrolyzer due to each temperature fluctuation or pressure shock can be modeled, directly linking the temperature fluctuation amplitude and the number of pressure shocks to the degree of system lifespan loss. Furthermore, based on the temperature-pressure-power dynamic coupling model, the lifespan loss model is introduced for multi-objective optimization, enabling multi-dimensional dynamic coupling modeling of temperature, pressure, power, and lifespan. This achieves coordinated management and control of "control parameters - operating status - equipment lifespan," strictly constraining extreme temperature and pressure values ​​while ensuring load response speed, avoiding membrane electrode damage or electrolyzer corrosion caused by extreme fluctuations, achieving a balance between rapid response and lifespan failure, and further improving the stability of the water electrolysis hydrogen production system.

[0102] As one possible implementation, during the execution of S103, the method also includes:

[0103] When the power change amplitude in the power prediction curve is greater than the first preset threshold, the weighting coefficient corresponding to the load response speed is increased; when the power change amplitude is greater than or equal to the first preset threshold, the weighting coefficient corresponding to the degree of lifespan loss is increased.

[0104] The power change amplitude refers to the magnitude of the power fluctuation. This application does not impose specific limitations on this. For example, if the power is predicted to increase from 10% to 20% from the power prediction curve, the corresponding power change amplitude is 10%.

[0105] Weight coefficients are used to identify the relative importance of the objective in the overall optimization process. Increasing the weight coefficient of the objective means that the optimization algorithm is more inclined to find a solution that performs better on that objective.

[0106] This application embodiment can dynamically adjust the priority of multiple objectives by adjusting the weighting coefficients. For example, in the target stage where the predicted power change amplitude is greater than a first preset threshold (such as the low power response stage where wind power increases from 10% of rated power to 90%), the weighting coefficient corresponding to the load response speed is increased, that is, the weighting coefficient for optimizing the model in the direction of faster load response speed is increased, thereby improving the system's ability to respond quickly to sudden load changes. When the power change amplitude is less than or equal to the first preset threshold, the weighting coefficient corresponding to the load response speed is decreased, and the weighting coefficient corresponding to the degree of lifespan loss is increased, so that the model is optimized in the direction of reducing lifespan loss.

[0107] Furthermore, for the low power response stage (such as the predicted increase from 10% of rated power to 20%), the weight of temperature fluctuations is reduced to ensure a smooth power increase and avoid membrane electrode drying. For the high power response stage (such as the predicted increase from 90% of rated power to 100%), the temperature and pressure are strictly constrained to not exceed the rated values ​​to avoid damage from over-temperature and over-pressure.

[0108] Therefore, the optimized model takes into account power tracking accuracy, temperature and pressure stability, and lifespan loss. The priority of multiple objectives can be dynamically adjusted through weighting coefficients, so that the model can be optimized in the expected direction.

[0109] To more clearly describe the load response control method of this water electrolysis hydrogen production system, a system architecture for implementing this method is proposed below, comprising five core modules: a power prediction module, an electrolyzer dynamic coupling model module, a model predictive control algorithm module, an execution regulation module, and a data acquisition and feedback module. Based on these five core modules, a specific implementation scenario is described.

[0110] Take the increase in wind power from 30% to 80% of rated power as an example.

[0111] Step 1: The power prediction module predicts 10 minutes in advance that the power will increase from 30% to 80%, and outputs a power prediction curve.

[0112] Step 2: Using the model predictive control algorithm module and the electrolytic cell dynamic coupling model module, based on the power prediction curve and the dynamic coupling model, solve the optimization function and output the optimal control sequence for temperature and pressure, so that the temperature can be increased 5 minutes in advance (from 60℃ to 70℃) and the pressure can be increased 3 minutes in advance (from 0.3MPa to 0.5MPa).

[0113] Step 3: By executing the adjustment module to adjust the temperature and pressure according to the control sequence, the temperature and pressure are adjusted to the optimal value that matches 80% of the power before the actual power changes abruptly.

[0114] Step 4: When the actual wind power changes abruptly, the electrolytic cell is already in the optimal operating state, with rapid power tracking and a response time (error ≤ 5%) of 5 seconds;

[0115] Step 5: Collect data in real time and feed it back to the control algorithm to fine-tune the temperature and pressure to ensure stable power operation. Temperature fluctuation ≤2℃, pressure fluctuation ≤0.03MPa, and life loss value controlled within the allowable range.

[0116] For low power response (e.g., increasing from 10% of rated power to 20%), the control algorithm ensures a smooth power increase and avoids membrane electrode drying by reducing the weight of temperature fluctuations. For high power response (e.g., increasing from 90% to 100%), the temperature and pressure are strictly constrained to not exceed the rated values ​​to avoid damage from over-temperature and over-pressure.

[0117] The embodiments of this application have the following key technologies:

[0118] (1) The core strategy of “prediction + active control”.

[0119] Breaking through the limitations of "passive response" in related technologies, the power prediction module (predicting the power curve for the next 5-60 minutes with an error of ≤±8%) obtains the load change trend, allowing the electrolyzer to adjust its operating parameters in advance, rather than just passively responding to current power fluctuations. This fundamentally solves the problem of response lag and achieves a rapid response at the second to minute level, which is 10-60 times faster than the existing minute-level response speed.

[0120] (2) Dynamic coupling modeling of temperature-pressure-power-lifetime in all dimensions.

[0121] This is the core foundation of control, unlike related technologies that only focus on control logic for a single parameter. On one hand, it constructs a nonlinear state-space model to accurately describe the dynamic relationship between temperature distribution, pressure distribution, membrane electrode humidity, and actual output power within the electrolyzer; on the other hand, it innovatively incorporates a lifespan loss quantification model, directly linking temperature fluctuation amplitude and pressure shock frequency with system lifespan decay, achieving coordinated management of "control parameters - operating status - equipment lifespan," filling the gap in existing technologies that do not fully incorporate lifespan loss.

[0122] (3) Rapid response and lifetime protection mechanism for load mutation.

[0123] When power fluctuations exceed 10% / second of the rated power, a special response mechanism is triggered: the power tracking weight is prioritized to ensure second-level power tracking; at the same time, the extreme values ​​of temperature (not exceeding ±5℃ of the rated value) and pressure (not exceeding ±10% of the rated value) are strictly constrained through the life loss model to avoid damage to the membrane electrode or corrosion of the electrolytic cell caused by extreme fluctuations, thus achieving a balance between "rapid response" and "life guarantee".

[0124] (4) Wide power range with full coverage adaptability.

[0125] Specifically addressing the issue of related technologies being unable to adapt to extreme low-power and high-power scenarios, this technology achieves stable operation across the entire power range from 10% to 100% of the rated power. At low power, it prevents membrane electrode drying and uneven current density; at high power, it prevents over-temperature and over-pressure. Within this range, the power response time is only 1-30 seconds, temperature fluctuation is ≤3℃, and pressure fluctuation is ≤0.05MPa, adapting to the actual needs of significant power fluctuations in wind and solar power.

[0126] Therefore, the embodiments of this application have the following advantages:

[0127] (1) Wide power range: Achieves full coverage from 10% to 100% of rated power to meet the needs of renewable energy with large fluctuations;

[0128] (2) Fast response: The power response time is 1-30 seconds (adjusted according to the power fluctuation range), which is 10-60 times faster than the existing technology (minute level);

[0129] (3) Stable operation: temperature fluctuation range ≤3℃, pressure fluctuation range ≤0.05MPa, to avoid system oscillation caused by sudden load changes;

[0130] (4) Lifespan guarantee: Through lifespan loss model constraints, the system lifespan is extended by 20%-30% compared to related technologies.

[0131] (5) Proactive early warning: By quantitatively assessing the degree of component degradation, predictive maintenance can be achieved, avoiding sudden downtime and improving the reliability and economy of system operation.

[0132] Based on the above embodiments, this application provides a load response control device for a water electrolysis hydrogen production system, referencing... Figure 2 The diagram shown is a schematic of a load response control device for a water electrolysis hydrogen production system provided in an embodiment of this application. The device 200 includes:

[0133] The acquisition unit 201 is used to acquire historical power data and meteorological forecast data of the water electrolysis hydrogen production system. The historical power data is used to identify the actual input power within a historical period.

[0134] The power prediction unit 202 is used to obtain the power prediction curve for the future time period based on the historical power data and the meteorological prediction data through a power prediction model.

[0135] The optimization solution unit 203 is used to solve the dynamic coupling model based on the power prediction curve to obtain a temperature control sequence and a pressure control sequence that satisfy multiple optimization objectives. The multiple optimization objectives include optimization objectives corresponding to load response speed, temperature stability and pressure stability. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolytic cell.

[0136] The adjustment unit 204 is used to adjust the temperature and pressure of the water electrolysis hydrogen production system in advance within a preset time before predicting power fluctuations in the power prediction curve, based on the temperature control sequence and the pressure control sequence.

[0137] As one possible implementation, the device also includes a construction unit for constructing a lifespan loss model of the water electrolysis hydrogen production system, using temperature fluctuation amplitude and pressure shock count as core variables.

[0138] The optimization solution unit is used to perform joint optimization solution on the dynamic coupling model and the lifetime loss model based on the power prediction curve, so as to obtain a temperature control sequence and a pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

[0139] As one possible implementation, the optimization solving unit is further configured to:

[0140] When the power change amplitude in the power prediction curve is greater than the first preset threshold, the weighting coefficient corresponding to the load response speed is increased; when the power change amplitude is greater than or equal to the first preset threshold, the weighting coefficient corresponding to the degree of lifespan loss is increased.

[0141] As one possible implementation, the device further includes a first model training unit for:

[0142] Obtain historical meteorological data;

[0143] Construct an initial power prediction model;

[0144] Based on the historical meteorological data and the historical power data, the initial power prediction model is trained to obtain the power prediction model.

[0145] As one possible implementation, the device further includes a second model training unit for:

[0146] Based on the anode temperature, cathode temperature, internal pressure, and membrane electrode humidity of the electrolytic cell, with temperature and pressure as control variables and the actual input power as the output variable, an initial dynamic coupling model is constructed.

[0147] Based on the actual input power under different temperatures and pressures, the parameters of the dynamic coupling model are adjusted to obtain the dynamic coupling model.

[0148] Based on the above embodiments, this application provides a computer device, which includes a processor and a memory:

[0149] The memory is used to store computer programs;

[0150] The processor is used to execute the load response control method of the above-described water electrolysis hydrogen production system according to the computer program.

[0151] Based on the above embodiments, this application provides a computer-readable storage medium for storing a computer program, which, when executed by a computer device, implements the load response control method of the above-described water electrolysis hydrogen production system.

[0152] Based on the above embodiments, this application provides a computer program product including a computer program, which, when run on a computer device, causes the computer device to execute the load response control method of the above-described water electrolysis hydrogen production system.

[0153] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0154] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A load response control method for a water electrolysis hydrogen production system, characterized in that, The method includes: Acquire historical power data and meteorological forecast data of the water electrolysis hydrogen production system, wherein the historical power data is used to identify the actual input power over a historical period; Based on the historical power data and the meteorological forecast data, a power forecast curve for the future time period is obtained through a power forecast model. Based on the power prediction curve, the dynamic coupling model is solved to obtain a temperature control sequence and a pressure control sequence that satisfy multiple optimization objectives. The multiple optimization objectives include optimization objectives corresponding to load response speed, temperature stability and pressure stability. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer. Based on the temperature control sequence and the pressure control sequence, the temperature and pressure of the water electrolysis hydrogen production system are pre-adjusted within a preset time before the power fluctuation is predicted in the power prediction curve.

2. The method according to claim 1, characterized in that, The method further includes: A lifespan loss model for the water electrolysis hydrogen production system was constructed using temperature fluctuation amplitude and pressure shock frequency as core variables. The process of solving the dynamic coupling model based on the power prediction curve to obtain temperature control sequences and pressure control sequences that satisfy multiple optimization objectives includes: Based on the power prediction curve, the dynamic coupling model and the lifetime loss model are jointly optimized and solved to obtain temperature control sequence and pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

3. The method according to claim 2, characterized in that, In the process of solving the dynamic coupling model based on the power prediction curve to obtain the temperature control sequence and pressure control sequence that satisfy multiple optimization objectives, the method further includes: When the power change amplitude in the power prediction curve is greater than the first preset threshold, the weighting coefficient corresponding to the load response speed is increased; when the power change amplitude is greater than or equal to the first preset threshold, the weighting coefficient corresponding to the degree of lifespan loss is increased.

4. The method according to claim 1, characterized in that, The method further includes: Obtain historical meteorological data; Construct an initial power prediction model; Based on the historical meteorological data and the historical power data, the initial power prediction model is trained to obtain the power prediction model.

5. The method according to claim 1, characterized in that, The method further includes: Based on the anode temperature, cathode temperature, internal pressure, and membrane electrode humidity of the electrolytic cell, with temperature and pressure as control variables and the actual input power as the output variable, an initial dynamic coupling model is constructed. Based on the actual input power under different temperatures and pressures, the parameters of the dynamic coupling model are adjusted to obtain the dynamic coupling model.

6. A load response control device for a water electrolysis hydrogen production system, characterized in that, The device includes: The acquisition unit is used to acquire historical power data and meteorological forecast data of the water electrolysis hydrogen production system. The historical power data is used to identify the actual input power within a historical period. The power prediction unit is used to obtain the power prediction curve for the future time period based on the historical power data and the meteorological prediction data through the power prediction model. The optimization solution unit is used to solve the dynamic coupling model based on the power prediction curve to obtain a temperature control sequence and a pressure control sequence that satisfy multiple optimization objectives. The multiple optimization objectives include optimization objectives corresponding to load response speed, temperature stability and pressure stability. The dynamic coupling model is used to describe the coupling relationship between temperature, pressure and power during the operation of the electrolyzer. An adjustment unit is used to pre-adjust the temperature and pressure of the water electrolysis hydrogen production system within a preset time before predicting power fluctuations in the power prediction curve, based on the temperature control sequence and the pressure control sequence.

7. The apparatus according to claim 6, characterized in that, The device further includes: The construction unit is used to construct the life loss model of the water electrolysis hydrogen production system with temperature fluctuation amplitude and pressure shock number as core variables. The optimization solution unit is used to perform joint optimization solution on the dynamic coupling model and the lifetime loss model based on the power prediction curve, so as to obtain a temperature control sequence and a pressure control sequence that satisfy multiple optimization objectives and the optimization objectives corresponding to the degree of lifetime loss.

8. A computer device, characterized in that, The computer device includes a processor and memory: The memory is used to store computer programs; The processor is configured to perform the method according to any one of claims 1-5 according to the computer program.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program that, when executed by a computer device, performs the method described in any one of claims 1-5.

10. A computer program product comprising a computer program, characterized in that, When it is run on a computer device, it causes the computer device to perform the method described in any one of claims 1-5.

Images