Hydrogen production power dynamic allocation method based on abandoned power characteristics

Abstract

The invention relates to a hydrogen production power dynamic allocation method based on abandoned power characteristics. The method comprises the following steps: acquiring an abandoned power sequence of a measured area in real time; carrying out decomposition through an adaptive mode decomposition algorithm based on spectral correlation to obtain a mode component; dividing into a high-frequency component set and a low-frequency component set to obtain a high-frequency component and a steady-state component; treating by using a fast response group divided into a proton exchange membrane electrolytic bath and a slow response group divided into an alkaline electrolytic bath; calculating a dynamic allocation factor based on the sliding time window; and dynamically allocating power to the fast response group and the slow response group according to the obtained dynamic allocation factor. According to the method, the self-adaptive modal decomposition algorithm is used for processing abandoned power data, the problems of under-decomposition and over-decomposition are solved by dynamically adjusting the modal number, and the decomposed high-frequency component and low-frequency component are processed by the fast-response proton exchange membrane electrolytic cell and the slow-response alkaline electrolytic cell respectively; and the distribution of the abandoned power is realized based on the dynamic distribution factor.

Description

Technical Field

[0001] This invention belongs to the field of hydrogen energy production technology, and in particular to a method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power. Background Technology

[0002] With the accelerated global energy transition, the installed capacity of renewable energy sources, represented by wind and solar power, continues to climb. However, these energy sources are characterized by significant intermittency and volatility, easily leading to insufficient grid absorption capacity and resulting in power curtailment. Although the curtailment rate of wind and solar power has decreased in recent years, many regions still have high levels of curtailment. Power curtailment not only wastes energy but also restricts the large-scale development of renewable energy. Meanwhile, hydrogen energy, as a clean and efficient secondary energy carrier, has a natural coupling between its production process's electricity demand and the volatility of power curtailment. Converting curtailed electricity into hydrogen energy through water electrolysis technology can improve the utilization rate of renewable energy and provide low-cost green hydrogen for the hydrogen economy, becoming an important path to solve the power curtailment problem.

[0003] The main technologies for producing hydrogen by electricity are as follows: (1) Alkaline electrolyzer for hydrogen production uses an alkaline solution as the electrolyte and separates the anode and cathode by asbestos or polymer membranes. This technology is mature, low-cost and large-scale, but has slow dynamic response and difficult product separation. (2) Proton exchange membrane electrolyzer for hydrogen production, with perfluorosulfonic acid proton exchange membrane as the core, the membrane simultaneously separates gases and conducts protons, eliminating the need for liquid electrolyte. This technology has fast dynamic response, small size, and high integration, but it is costly and has limited single-cell scale; (3) Solid oxide electrolyzer, which uses solid oxide ceramic as electrolyte. This technology has high energy efficiency and no dependence on precious metals, but it has slow start-up and poor dynamic performance. (4) Anion exchange membrane electrolyzer, which uses anion exchange membrane as electrolyte. This technology has low cost and better dynamic response, but the technology is not mature.

[0004] Currently, the mainstream electro-hydrogen production technologies are mainly alkaline electrolyzers and proton exchange membrane electrolyzers. Power allocation strategies for hydrogen production systems are mostly based on fixed thresholds or simple priority logic, making it difficult to accurately match the dynamic characteristics of surplus power. Traditional methods often directly connect surplus power to the electrolyzer, leading to frequent start-ups and shutdowns or operation in inefficient ranges, reducing equipment lifespan and hydrogen production efficiency. Furthermore, the temporal distribution of surplus power is highly uncertain; for example, wind power output is high at night, while photovoltaic power fluctuates greatly during the day. Optimizing power allocation using these characteristics is a key challenge for improving system economy and stability.

[0005] The current power allocation methods mainly include the following: (1) Fixed power allocation method is simple to implement and does not require complex algorithms and real-time data support. However, this method cannot adapt to the dynamic fluctuations of power abandonment, which can easily lead to power abandonment waste or long-term operation of electrolyzers in the inefficient range, making it difficult to meet the needs of large-scale, high-fluctuation scenarios.

[0006] (2) Rule-based control methods, such as priority allocation and fuzzy logic control, dynamically adjust power through preset rules. Priority allocation can adapt to fast start-up scenarios based on equipment characteristics, while fuzzy logic control is good at handling uncertainty issues and achieving smooth allocation. However, these methods rely too much on experience to design rules, making it difficult to cover all working conditions and lacking robustness.

[0007] (3) Model predictive control, based on dynamic model to predict future operating conditions, can achieve global optimization, can coordinate wind and solar power output, energy storage and hydrogen production power, improve the utilization rate of hydrogen energy system, and support multi-objective optimization. However, its computational complexity is high and it has stringent requirements for real-time data acquisition and hardware performance.

[0008] (4) Multi-objective optimization algorithms, such as NSGA-II and MOPSO, can simultaneously optimize multiple objectives such as economy, efficiency and stability. They are suitable for complex modeling scenarios such as hybrid electrolytic cell systems. However, these algorithms are time-consuming to iterate and are difficult to meet the real-time control requirements.

[0009] (5) The chain distribution strategy achieves smooth power distribution by gradually deploying electrolytic cell stacks, enabling the system to complete power redistribution in a short time and significantly improving reliability. However, it requires additional hardware such as spare stacks, which increases system cost. Moreover, the dynamic response speed is limited and may not be able to adjust in time under extreme fluctuations.

[0010] The above methods all lack analysis of the high-frequency and low-frequency components of the abandoned power characteristics, and do not fully consider the fast response characteristics of proton exchange membrane electrolyzers and the slow response characteristics of alkaline electrolyzers. Summary of the Invention

[0011] The purpose of this invention is to overcome the shortcomings of the prior art and propose a dynamic allocation method for hydrogen production power based on the characteristics of abandoned power. Based on the characteristics of abandoned power, abandoned power is decomposed into high-frequency components and steady-state components. A dynamic allocation factor is introduced to match the response characteristics of proton exchange membrane electrolyzers and alkaline electrolyzers, responding to abandoned power fluctuations in real time, maximizing the use of the characteristics of different hydrogen production technologies, extending equipment life and reducing investment costs.

[0012] The technical problem solved by this invention is achieved through the following technical solution: A method for dynamically allocating hydrogen production power based on the characteristics of abandoned power includes the following steps: Step 1: Obtain the real-time power curtailment sequence of the measured area; Step 2: Decompose the abandoned power sequence using an adaptive mode decomposition algorithm based on spectral correlation to obtain modal components; Step 3: Divide the modal components into a high-frequency component set and a low-frequency component set, and then obtain the high-frequency components and steady-state components; Step 4: Divide the electrolyzer into a fast-response group for proton exchange membrane electrolyzers and a slow-response group for alkaline electrolyzers, and process the high-frequency component and steady-state component respectively. Step 5: Based on the partitioning in Step 4, calculate the dynamic allocation factor using a sliding time window; Step 6: Based on the obtained dynamic allocation factor, dynamically allocate power to the fast response group and the slow response group.

[0013] Furthermore, the specific implementation method of step 1 is as follows: real-time acquisition of the power curtailment sequence of the measured area. P ab ( t ),in t =1,2,..., T , T This represents the total length of the data.

[0014] Furthermore, the specific implementation method of step 2 is as follows: set the initial number of modes. K = K 0, defines the correlation threshold. f, If 0 < φ < 1, execute the traditional VMD decomposition algorithm to obtain the intrinsic mode function sequence:

[0015] in, IMF k Indicates the first k Each intrinsic mode component Calculate the correlation coefficient based on the intrinsic mode function sequence. r :

[0016] in, Represents modal components IMF k The mean, s IMFk Represents modal components IMF k standard deviation For the first k-1 Each intrinsic mode component IMF k-1 The mean, For the first k-1 Each intrinsic mode component For the first k Each intrinsic mode component sIMFk-1 Modal components IMF k-1 Standard deviation; when r ≤ f The number of modes that need to be set for output. K * =K， otherwise K=K +1 until the condition is met.

[0017] Furthermore, the specific implementation method of step 3 is as follows: modal components IMF k Divided into high-frequency component sets H and low-frequency component sets L High-frequency components are obtained. P high ( t ) is a set of high-frequency components H The sum of all internal components, steady-state components P steady ( t () is the low-frequency component set L The sum of the components within.

[0018] Furthermore, the specific implementation method of step 4 is as follows: divide the electrolytic cell into two groups, one group having a rated power of P P The fast-response group of the proton exchange membrane electrolyzer is used to process high-frequency fluctuation components. P high ( t The other group has a rated power of P A The slow-response group of the alkaline electrolyzer is used to handle the steady-state component. P steady ( t ).

[0019] Furthermore, the specific implementation method of step 5 is as follows: based on the sliding time window, quantify the characteristics of power curtailment fluctuations and calculate the allocation factor. α ( t ):

[0020]

[0021] in, s high ( t () represents the standard deviation of the high-frequency components, reflecting the fluctuation range; u high For the window P high (t) mean, lis a smoothing constant, and m is the length of the set time window.

[0022] Furthermore, the specific implementation method of step 6 is as follows: dynamically allocate power to the fast response group. P PEM ( t ) and slow response group P ALK (t If the power output exceeds the rated power of the electrolyzer group, a power limiting process will be implemented to allocate hydrogen production power.

[0023] .

[0024] The advantages and positive effects of this invention are: 1. This invention decomposes the characteristics of abandoned electricity into high-frequency components and steady-state components, introduces a dynamic allocation factor to respond to abandoned electricity fluctuations in real time, and matches the response characteristics of proton exchange membrane electrolyzers and alkaline electrolyzers to maximize the use of the characteristics of different hydrogen production technologies, extend equipment life, reduce investment costs, and is easy to operate with low computational complexity.

[0025] This invention proposes an adaptive variational mode decomposition algorithm, which introduces a mode correlation criterion on the basis of traditional variational mode decomposition to achieve automatic optimization of the number of modes. It can more accurately decompose the curtailed power into high-frequency and low-frequency components, without the need for manual preset of the number of modes, reducing mode aliasing, and has good operability. Attached Figure Description

[0026] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to the accompanying drawings.

[0028] A method for dynamically allocating hydrogen production power based on the characteristics of abandoned power, such as Figure 1 As shown, it includes the following steps: Step 1: Obtain the power curtailment sequence of the measured area in real time.

[0029] The specific implementation method of step 1 is as follows: real-time acquisition of the power curtailment sequence of the measured area. P ab ( t ),in t =1,2,..., T , T This represents the total length of the data.

[0030] Step 2: Decompose the abandoned power sequence using an adaptive mode decomposition algorithm based on spectral correlation to obtain mode components.

[0031] Traditional Variational Mode Decomposition (VMD) requires a preset number of modes, but the actual signal complexity is unknown. This method dynamically adjusts the preset number of modes by analyzing the spectral correlation between adjacent modes, addressing both under-decomposition (too small a preset number leading to mode aliasing) and over-decomposition (too large a preset number introducing spurious components). An initial number of modes is set... K = K 0, defines the correlation threshold. f, If 0 < φ < 1, execute the traditional VMD decomposition algorithm to obtain the intrinsic mode function sequence:

[0032] in, IMF k Indicates the first k Each intrinsic mode component Calculate the correlation coefficient based on the intrinsic mode function sequence. r :

[0033] in, Represents modal components IMF k The mean, s IMFk Represents modal components IMF k standard deviation For the first k-1 Each intrinsic mode component IMF k-1 The mean, For the first k-1 Each intrinsic mode component For the first k Each intrinsic mode component s IMFk-1 Modal components IMF k-1 Standard deviation; when r ≤ f The number of modes that need to be set for output. K * =K， otherwise K=K +1 until the condition is met.

[0034] Step 3: Divide the modal components into a high-frequency component set and a low-frequency component set, and then obtain the high-frequency component and the steady-state component.

[0035] The specific implementation method of step 3 is as follows: high-frequency component set H ={ IMFk}(when k ≤min( K * / 2)), Low-frequency component set L ={ IMF k}(when k >min( K * / 2)), to obtain high frequency components P high ( t ) is a set of high-frequency components H The sum of all internal components, steady-state components P steady ( t () is the low-frequency component set L The sum of the components within.

[0036] Step 4: Divide the electrolyzer into a fast-response group for proton exchange membrane electrolyzers and a slow-response group for alkaline electrolyzers, and process the high-frequency component and steady-state component respectively.

[0037] One group has a rated power of P P The fast-response group of the proton exchange membrane electrolyzer (PEM) is used to process high-frequency fluctuation components. P high ( t The other group has a rated power of P A The slow-response group of the alkaline electrolyzer (ALK) is used to process the steady-state component. P steady ( t ).

[0038] Step 5: Based on the division in Step 4, calculate the dynamic allocation factor based on the sliding time window to provide a basis for the real-time allocation of power between the two response groups.

[0039] Based on the sliding time window (e.g., window size) m =60 minutes), quantify the characteristics of power curtailment fluctuations and calculate allocation factors. α ( t ):

[0040]

[0041] In the formula s high ( t () represents the standard deviation of the high-frequency components, reflecting the fluctuation range;u high For the window P high (t) mean, l The smoothing constant is 0.1 by default; m is the length of the set time window.

[0042] Step 6: Based on the obtained dynamic allocation factor, dynamically allocate power to the fast response group and the slow response group.

[0043] Dynamically allocate power to fast response group P PEM ( t ) and slow response group P ALK (t If the power output exceeds the rated power of the electrolyzer group, a power limiting process will be implemented to allocate hydrogen production power.

[0044]

[0045] It should be emphasized that the embodiments described in this invention are illustrative rather than limiting. Therefore, this invention includes, but is not limited to, the embodiments described in the specific implementation. Any other implementations derived by those skilled in the art based on the technical solutions of this invention are also within the scope of protection of this invention.

Claims

1. A method for dynamic allocation of hydrogen production power based on curtailed power characteristics, characterized in that: Includes the following steps: Step 1: Obtain the real-time power curtailment sequence of the measured area; Step 2: Decompose the abandoned power sequence using an adaptive mode decomposition algorithm based on spectral correlation to obtain modal components; Step 3: Divide the modal components into a high-frequency component set and a low-frequency component set, and then obtain the high-frequency components and steady-state components; Step 4: Divide the electrolyzer into a fast-response group for proton exchange membrane electrolyzers and a slow-response group for alkaline electrolyzers, and process the high-frequency component and steady-state component respectively. Step 5: Based on the partitioning in Step 4, calculate the dynamic allocation factor using a sliding time window; Step 6: Based on the obtained dynamic allocation factor, dynamically allocate power to the fast response group and the slow response group.

2. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 1 is as follows: real-time acquisition of the power curtailment sequence of the measured area. P ab ( t ),in t =1,2,..., T , T This represents the total length of the data.

3. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 2 is as follows: Let the initial number of modes be... K = K 0, defines the correlation threshold. φ, If 0 < φ < 1, execute the traditional VMD decomposition algorithm to obtain the intrinsic mode function sequence: ； in, IMF k Indicates the first k Each intrinsic mode component Calculate the correlation coefficient based on the intrinsic mode function sequence. ρ : ； in, Represents modal components IMF k The mean, σ IMFk Represents modal components IMF k standard deviation For the first k-1 Each intrinsic mode component IMF k-1 The mean, For the first k-1 Each intrinsic mode component For the first k Each intrinsic mode component σ IMFk-1 Modal components IMF k-1 Standard deviation; when ρ ≤ φ The number of modes that need to be set for output. K * =K， otherwise K=K +1 until the condition is met.

4. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 3 is as follows: Modal components... IMF k Divided into high-frequency component sets H and low-frequency component sets L High-frequency components are obtained. P high ( t ) is a set of high-frequency components H The sum of all internal components, steady-state components P steady ( t () is the low-frequency component set L The sum of the components within.

5. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 4 is as follows: divide the electrolytic cell into two groups, one group having a rated power of... P P The fast-response group of the proton exchange membrane electrolyzer is used to process high-frequency fluctuation components. P high ( t The other group has a rated power of P A The slow-response group of the alkaline electrolyzer is used to handle the steady-state component. P steady ( t ).

6. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 5 is as follows: based on the sliding time window, quantify the characteristics of power curtailment fluctuations and calculate the allocation factor. α ( t ): ； ； in, σ high ( t () represents the standard deviation of the high-frequency components, reflecting the fluctuation range; u high For the window P high (t) mean, λ is a smoothing constant, and m is the length of the set time window.

7. The method for dynamic allocation of hydrogen production power based on the characteristics of abandoned power as described in claim 1, characterized in that: The specific implementation method of step 6 is as follows: dynamically allocate power to the fast response group. P PEM ( t ) and slow response group P ALK (t If the power output exceeds the rated power of the electrolyzer group, a power limiting process will be implemented to allocate hydrogen production power. ； 。

Images