Method for improving flexibility of hydrogen production system based on multi-source cooperation and adaptive regulation

Abstract

The application discloses a kind of based on multi-source cooperation and adaptive regulation and control hydrogen production system flexibility promotion method, belong to new energy technology field.The method of the present application is executed by digital twin as global decision center, including: obtaining the predicted data of wind and solar power generation power, real-time frequency deviation signal of power grid and the operating state of each electrolytic cell in hydrogen production system;Based on the predicted data of wind and solar power generation power and frequency deviation signal, generate global optimization scheduling instruction with the goal of improving system flexibility and economy;According to global optimization scheduling instruction, cooperatively control electrolytic cell dynamic reconfiguration and impose virtual synchronous generator VSG control on hydrogen production load;Actual operating data after cooperative control is collected and fed back to equipment fatigue loss analysis model and wind and solar output prediction model, for fatigue damage assessment and power prediction in next cycle are dynamically revised.The method of the present application effectively reduces system operating cost while greatly improving the flexibility of new energy hydrogen production system.

Description

Technical Field

[0001] This invention belongs to the field of new energy technology, and in particular relates to a method for improving the flexibility of hydrogen production systems based on multi-source synergy and adaptive regulation. Background Technology

[0002] To achieve the "dual-carbon" goal, hydrogen energy has been established as a national strategic energy source. Among these, utilizing new energy sources such as wind and solar power to produce hydrogen is a key pathway to achieving zero-carbon hydrogen sources. However, the current mainstream water electrolysis hydrogen production technology suffers from insufficient flexibility in the face of the inherent volatility of new energy power generation, which has become a core bottleneck.

[0003] Current mainstream renewable energy hydrogen production technologies, such as alkaline (ALK) water electrolysis and proton exchange membrane (PEM) water electrolysis, have many inherent limitations in terms of flexibility. While ALK water electrolysis offers advantages such as relatively low cost and high technological maturity, its load regulation range is relatively narrow, generally operating stably only within 30%-100% of the rated load. Furthermore, its load regulation rate is slow, often requiring several minutes or even longer to switch from low to high load, making it difficult to quickly adapt to rapid changes in renewable energy power. Although PEM water electrolysis offers a faster response time, operating at high current densities poses significant challenges to equipment stability and durability. Prolonged exposure to frequent load fluctuations accelerates the aging and wear of critical components such as electrodes and membranes, significantly increasing equipment maintenance costs and replacement frequency, thus limiting its widespread application in renewable energy hydrogen production scenarios requiring high flexibility. Furthermore, existing hydrogen production systems lack efficient intelligent control mechanisms for integration with new energy power generation systems. They cannot accurately and dynamically optimize hydrogen production operation strategies based on real-time changes in new energy power, resulting in difficulty in improving the overall system flexibility and energy utilization efficiency.

[0004] Therefore, there is an urgent need for a new technical solution to address this problem. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method for improving the flexibility of a hydrogen production system based on multi-source synergy and adaptive regulation, which is used to solve the technical problem that the water electrolysis hydrogen production device in the existing new energy hydrogen production system cannot efficiently and stably adapt to the wide power fluctuation of new energy power generation due to insufficient flexibility.

[0006] The technical solution adopted in this invention is to provide a method for improving the flexibility of a hydrogen production system based on multi-source synergy and adaptive control. The method is executed by a digital twin as the global decision-making center and includes the following cyclical process:

[0007] Real-time sensing and prediction: Acquire predicted data on wind and solar power generation, real-time frequency deviation signals of the power grid, and the operating status of each electrolyzer in the hydrogen production system;

[0008] Global optimization decision: Based on the predicted power of wind and solar power and the frequency deviation signal, a multi-agent reinforcement learning algorithm is used to generate a global optimization scheduling instruction that includes an electrolyzer combination strategy, virtual synchronous generator (VSG) control parameters, and a power change rate constraint threshold, with the goal of improving system flexibility and economy. The power change rate constraint threshold is dynamically calculated based on the equipment fatigue loss analysis model.

[0009] Cooperative control execution: Based on the electrolyzer combination strategy in the global optimization scheduling instruction, the dynamic on / off state of the solid-state relay matrix is ​​controlled to reconfigure the circuit of the hybrid electrolyzer composed of alkaline electrolyzer (ALK) and proton exchange membrane electrolyzer (PEM); based on the VSG control parameters in the global optimization scheduling instruction, virtual synchronous generator (VSG) control is applied to the hydrogen production load to make the hydrogen production system respond to grid frequency fluctuations.

[0010] Dynamic feedback correction: Collect the actual operating data of the system after executing the collaborative control execution steps, and feed it back to the equipment fatigue loss analysis model and the wind and solar power output prediction model for dynamic correction of fatigue damage assessment and power prediction in the next cycle.

[0011] The electrolyzer combination strategy employs the XGBoost intelligent algorithm to establish a system efficiency optimization model with the goal of maximizing the overall hydrogen production efficiency of the system.

[0012] ;

[0013] ;

[0014] ;

[0015] In the formula, For system efficiency; and These represent the hydrogen production power of the ALK and PEM electrolyzers, respectively. and They provide power from wind and solar respectively; and These are the hydrogen production efficiencies of the PEM and ALK electrolyzers at the i-th time, respectively. Let be the power at time i; and These represent the minimum and maximum power, respectively.

[0016] The dynamic on / off control of the solid-state relay matrix specifically involves: parsing the electrolytic cell combination strategy into switching commands for each relay in the solid-state relay matrix, wherein the switching commands are determined by the required power. Compared with the preset power range threshold [L] k H k The decision is triggered by the following logic:

[0017] ;

[0018] In the formula, For the required power, and This represents the power range corresponding to the k-th solid-state relay.

[0019] The application of virtual synchronous generator (VSG) control to the hydrogen production load is specifically achieved through the following rotor motion equations:

[0020] ;

[0021] ;

[0022] In the formula, J is the virtual inertia, and D is the damping coefficient. This refers to the power grid frequency deviation. This refers to the frequency modulation coefficient; This refers to the output power.

[0023] The values ​​of the virtual inertia J and the damping coefficient D are derived from the global optimization scheduling instruction.

[0024] The dynamic calculation of the power change rate constraint threshold based on the equipment fatigue loss analysis model specifically includes:

[0025] The cumulative fatigue damage was calculated by statistically analyzing the cyclic amplitude and number of power fluctuations using the rainflow counting method. :

[0026] ;

[0027] In the formula, and To statistically analyze the amplitude and number of power fluctuation cycles; In amplitude The number of cycles that cause the device to fail;

[0028] Based on the accumulated fatigue damage, the power change rate constraint threshold is dynamically set:

[0029] ;

[0030] In the formula, The maximum allowable rate of power change. The baseline rate of change is determined by the properties of the electrolytic cell material. This is the critical damage threshold.

[0031] The specific method for correcting the wind and solar power output prediction model in the dynamic feedback correction step is as follows:

[0032] The power grid frequency deviation Introducing a wind and solar power output prediction model, the corrected photovoltaic power output is calculated:

[0033] ;

[0034] In the formula Mapping the VSG frequency regulation coefficient to the derating ratio of the photovoltaic inverter; To adjust the photovoltaic output.

[0035] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement steps of a method for improving the flexibility of a hydrogen production system based on multi-source coordination and adaptive control.

[0036] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for improving the flexibility of a hydrogen production system based on multi-source coordination and adaptive control.

[0037] Through the above design scheme, the present invention can bring the following beneficial effects:

[0038] 1. Dynamic reconfiguration of electrolyzer: A PEM-ALK hybrid electrolyzer architecture is designed, and the circuit is dynamically reconfigured through a solid-state relay matrix to broaden the operating power range and reduce the number of start-ups and shutdowns.

[0039] 2. Virtual Synchronous Generator Control for Hydrogen Production Load: The “Virtual Synchronous Generator (VSG)” technology is adopted, enabling the hydrogen production system to respond quickly to power fluctuations and maintain stable grid operation, just like a traditional generator.

[0040] 3. Wind and solar co-generation planning: By establishing a co-generation model for wind and solar energy, the flexibility and proactive adjustment capabilities of the energy hydrogen production system can be improved, enabling it to better adapt to the volatility of wind and solar resources.

[0041] 4. Equipment flexibility and synergy: Based on the equipment fatigue loss analysis model, the power change rate of the electrolyzer is dynamically controlled to slow down equipment aging while maintaining the same fluctuation level, thus constructing a hybrid electrolysis hydrogen production system that has both rapid response capability and economic efficiency.

[0042] 5. Digital twin global optimization enhances system flexibility: By utilizing a multi-agent reinforcement learning platform, the interaction between wind power, photovoltaic, energy storage and hydrogen production systems is simulated in real time to generate the most flexible operation and scheduling strategy, thereby improving the coordination and adaptability of the entire system.

[0043] 6. The system response speed has been significantly improved, reducing the response delay from the second level to the millisecond level, thereby fully meeting the strict requirements of the primary frequency regulation of the power grid, effectively reducing the system operating cost, and improving the flexibility of the new energy hydrogen production system by 5%. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the process for improving the flexibility of a hydrogen production system based on multi-source synergy and adaptive control, as described in this invention. Detailed Implementation

[0045] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0046] (1) The dynamic adjustment technology for electrolytic cells is mainly used to solve the power regulation problem faced by ALK electrolytic cells and PEM electrolytic cells when operating together. Specific implementation methods include: expanding the load regulation range of the two types of electrolytic cells through a solid-state relay (SSR) switching system to enhance overall flexibility; and using the XGBoost intelligent algorithm to predict the optimal operating state of the electrolytic cells, rationally allocating the load, reducing frequent start-ups and shutdowns, and extending service life. The system efficiency optimization model is based on the current-voltage characteristics and dynamic response equations of the electrolytic cells themselves, thereby achieving more efficient and stable operation. The system efficiency optimization model based on the current-voltage characteristics and dynamic response equations of the electrolytic cells is as follows:

[0047] ;

[0048] In the formula, For system efficiency; and These represent the hydrogen production power of the ALK and PEM electrolyzers, respectively. and They contribute power from wind and solar respectively.

[0049] The threshold-triggered SSR switching strategy is designed as follows:

[0050] ;

[0051] In the formula, For the required power, and This refers to the power range corresponding to the k-th SSR (e.g., SSR1 controls PEM, SSR2 controls ALK).

[0052] The objective function of this optimization model is to maximize the overall hydrogen production efficiency of the system, aiming to improve its operational flexibility.

[0053] ;

[0054] ;

[0055] In the formula, and These are the hydrogen production efficiencies of the PEM and ALK electrolyzers at the i-th time, respectively. Let be the power at time i; and These represent the minimum and maximum power, respectively.

[0056] (2) The virtual synchronous generator control for hydrogen production load aims to achieve dynamic optimization control of a high-performance electro-hydrogen conversion system. Specifically, the switching state of the solid-state relay matrix in (1) is used as the input parameter of the VSG algorithm to dynamically adjust its virtual inertia J and damping coefficient D; the VSG provides virtual inertial support for the system by simulating the rotor motion characteristics of the synchronous generator, and the value range of its virtual inertia J is usually 0.1*10 kg·m². This control method first needs to establish the rotor motion equation of the VSG:

[0057] ;

[0058] ;

[0059] In the formula, J is the virtual inertia, and D is the damping coefficient. This refers to the power grid frequency deviation. Frequency modulation factor (adjustable range 0.5%-5%); This refers to the output power.

[0060] (3) The wind-solar co-generation plan aims to improve the flexibility and active support capability of the energy hydrogen production system. Wind and solar power outputs have certain complementary characteristics in terms of time sequence, but there is still strong volatility. Therefore, it is necessary to construct a wind-solar power output prediction model to smooth power fluctuations and provide more stable input power for the electrolyzer (4). In this process, the frequency deviation signal generated by the VSG control in (2) will be used to... This is incorporated into the wind and solar power output prediction model to achieve dynamic correction of the prediction results. The mathematical expression for photovoltaic power output is as follows:

[0061] ;

[0062] In the formula, Photovoltaic conversion efficiency; The surface area of ​​the photovoltaic cell array is expressed in units of... ; Total solar radiation, in units ; The transmittance of a photovoltaic cell module; The operating temperature of photovoltaic cell modules, in units of .

[0063] The expression for the combined wind and solar power output is:

[0064] ;

[0065] In theory, it is glorious Efforts made at all times; These are the maximum and minimum output values, respectively. For normalized wind power or solar power in Efforts are made at all times.

[0066] Add a frequency compensation term to the wind and solar power output prediction model:

[0067] ;

[0068] In the formula Mapping the VSG frequency regulation coefficient to the derating ratio of the photovoltaic inverter; To adjust the photovoltaic output.

[0069] (4) Equipment flexibility coordination aims to improve the adaptability of system operation while reducing economic losses. The specific implementation method is as follows: the power fluctuation spectrum in (3) is used as the input of the rainflow counting method for fatigue damage analysis; the power change rate is constrained by the dynamic amplitude limiting strategy, thereby reducing equipment fatigue losses. This method calculates the cumulative fatigue damage by statistically analyzing the cyclic amplitude and frequency of power fluctuations, and its formula can be expressed as:

[0070] ;

[0071] In the formula, and To statistically analyze the amplitude and number of power fluctuation cycles; In amplitude The number of cycles that cause the device to fail.

[0072] The dynamic limiting adaptive threshold is:

[0073] ;

[0074] In the formula, The maximum allowable rate of power change. The baseline rate of change is determined by the properties of the electrolytic cell material. This is the critical damage threshold. To implement the fatigue damage index, it is calculated using the rainflow counting method, with a range of 0-1.

[0075] (5) The digital twin serves as the global decision-making unit of the system. Through intelligent optimization algorithms, it integrates real-time data and models from each stage (1) to (5) to achieve dynamic optimization and collaborative control. Specifically, the system uses the SSR matrix state and the efficiency of each unit in (1) as the basis for its dynamic optimization and collaborative control. Taking the power demand of VSG as input, and comprehensively considering factors such as the dynamic power operation boundary, overload duration limit, start-up and shutdown characteristics and efficiency characteristics of the hybrid electric hydrogen production system, the optimal electrolyzer combination strategy is output to improve the instantaneous flexibility of the hydrogen energy system. At the same time, based on the grid frequency deviation and equipment damage parameters in (2), the VSG control parameters are dynamically adjusted to achieve the best balance between response speed and response efficiency. With the help of the wind and solar power output prediction results in (3), the coordinated scheduling of electrolyzers and energy storage is optimized to generate a forward-looking scheduling plan with a 24-hour cycle. Finally, the fatigue damage assessment in (4) is incorporated into the global optimization objective, and the damage weight coefficient in the multi-agent reinforcement learning reward function is dynamically corrected to achieve the optimal system economy while improving the flexibility of the hydrogen production process.

[0076] The proposed optimization method for improving the flexibility of new energy hydrogen production takes the improvement of the flexibility of the electro-hydrogen production system as its core objective. Through a multi-dimensional collaborative optimization method, it systematically improves the operational flexibility of the hydrogen production device from two dimensions: instantaneous flexibility and process flexibility. This enables the device to track the fluctuations in new energy power in real time, efficiently and safely, maximize the ability to absorb fluctuating energy, and ensure stable and efficient operation of the system within a wide load range.

[0077] The implementation of the present invention is not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and are included within the protection scope of the present invention.

Claims

1. A method for improving the flexibility of a hydrogen production system based on multi-source collaboration and adaptive regulation, characterized by: The method is performed by a digital twin as a global decision center, including the following cyclic process: Real-time perception and prediction: obtaining predicted data of wind-solar power generation, real-time frequency deviation signal of power grid, and running state of each electrolyzer in the hydrogen production system; Global optimization decision: based on the predicted data of wind-solar power generation and the frequency deviation signal, using a multi-agent reinforcement learning algorithm to generate a global optimization scheduling instruction containing electrolyzer combination strategy, virtual synchronous generator (VSG) control parameter, and power change rate constraint threshold, with the goal of improving system flexibility and economy, the power change rate constraint threshold is dynamically calculated based on a device fatigue loss analysis model; the device fatigue loss analysis model is used to dynamically calculate the power change rate constraint threshold, specifically including: The cycle amplitude and times of power fluctuation are counted by rain flow counting method, and cumulative fatigue damage is calculated : ； wherein and is the statistical power fluctuation cycle amplitude and number; is the cycle number leading to equipment failure at amplitude ​ Based on the cumulative fatigue damage, the power change rate constraint threshold is dynamically set: ； wherein is the current allowed maximum power rate of change, is the reference rate of change, determined by the cell material properties, is the critical damage threshold; Coordinated control execution: according to the electrolyzer combination strategy in the global optimization scheduling instruction, control the dynamic on-off of the solid-state relay matrix to reconfigure the circuit of the mixed electrolysis stack composed of alkaline electrolyzer (ALK) and proton exchange membrane electrolyzer (PEM); according to the VSG control parameter in the global optimization scheduling instruction, the virtual synchronous generator (VSG) control is applied to the hydrogen production load, so that the hydrogen production system responds to the frequency fluctuation of the power grid; Dynamic feedback correction: collect the actual running data of the system after the coordinated control execution step, and feed it back to the device fatigue loss analysis model and wind-solar output prediction model for dynamic correction of fatigue damage assessment and power prediction in the next cycle.

2. The method for improving flexibility of a hydrogen production system based on multi-source collaboration and adaptive regulation according to claim 1, characterized in that: The electrolyzer combination strategy uses an XGBoost intelligent algorithm to establish a system efficiency optimization model with the goal of maximizing the total hydrogen production efficiency of the system: ； ； ； wherein, is the system efficiency; and and represent the hydrogen production power of the ALK and PEM electrolyzer, respectively; and and represent the wind and solar power output, respectively; and and represent the hydrogen production efficiency of the PEM and ALK electrolyzer at the i-th time instant, respectively; is the power at time i; and and represent the minimum and maximum power, respectively.

3. The method for improving flexibility of a hydrogen production system based on multi-source collaboration and adaptive regulation according to claim 2, characterized in that: The dynamic on-off control of the solid-state relay matrix specifically comprises: resolving the electrolytic cell combination strategy into switch commands for each relay in the solid-state relay matrix, the switch commands being determined by a required power and a preset power interval threshold [L k ,H k ], and the trigger logic being that: ； In the formula, is the required power, and is the power interval corresponding to the kth solid-state relay.

4. The method for improving flexibility of a hydrogen production system based on multi-source collaboration and adaptive regulation according to claim 3, characterized in that: The virtual synchronous generator (VSG) control is implemented through the following rotor motion equation: ； ； where J is a virtual inertia, D is a damping coefficient, is the grid frequency deviation; is the frequency modulation coefficient; is the output power; The values of the virtual inertia J and the damping coefficient D come from the global optimization scheduling instruction.

5. The method for improving flexibility of a hydrogen production system based on multi-source collaboration and adaptive regulation according to claim 4, characterized in that: In the dynamic feedback correction step, the specific way to correct the wind-solar output prediction model is: The grid frequency deviation is calculated A wind-solar power output prediction model is introduced to calculate the corrected photovoltaic output: ； wherein is the frequency modulation coefficient of the VSG mapped to the derating ratio of the PV inverter; is the adjusted PV output.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: The processor executes the program to implement the steps of the method of any one of claims 1 to 5.

7. A computer readable storage medium having stored thereon a computer program, characterized in that: The program is executed by the processor to implement the steps of the method of any one of claims 1 to 5.

Images