A typical working condition generation method for photovoltaic water electrolysis hydrogen production and its application

Abstract

The present invention proposes a method for generating typical working conditions of photovoltaic water electrolysis for hydrogen production. It includes extracting the time-frequency components of the output curve of the photovoltaic power station, decomposing the original output curve into low-frequency components, medium-frequency components, and high-frequency components by using empirical wavelet transform; defining and extracting the characteristic sequence of each time-frequency component based on the working characteristics of the proton exchange membrane electrolyzer; objectively weighting the extracted characteristic sequence and solving the optimal clustering number, and obtaining a typical characteristic quantity sequence based on the weighted fuzzy clustering analysis method; reconstructing the corresponding basic component and fluctuation component according to the characteristic sequence, and finally forming a typical working condition of photovoltaic water electrolysis for hydrogen production that can be tested in the laboratory.

Description

Technical Field

[0001] The present invention relates to the field of hydrogen energy production, and is aimed at analyzing the working conditions of an electrolyzer for producing green hydrogen by photovoltaic electrolysis of water, and in particular, to a method for generating typical working conditions for producing hydrogen by photovoltaic electrolysis of water. Background Art

[0002] With the proposal of the "dual carbon" goal, the proportion of renewable energy generation in the power structure of my country's power system is increasing. However, taking photovoltaic power as an example, the large peak-to-valley difference between day and night output and the output volatility during the daytime make it difficult to connect to the grid.

[0003] Hydrogen production by water electrolysis is a green hydrogen production technology. Hydrogen energy can be stored for a long time and in large capacity, and is well suited to photovoltaic output with obvious seasonal and daily fluctuation characteristics. The "green hydrogen" produced by photovoltaic water electrolysis is an important carrier for achieving green and low-carbon transformation, and will become an important part of the national energy system. Hydrogen energy has a very broad market prospect in the future. With the proposal of the "14th Five-Year Plan for Scientific and Technological Innovation in the Energy Field", my country will build a complete hydrogen energy industry chain integrating "production, storage and application" and actively build a number of related demonstration projects. Hydrogen production by proton exchange membrane water electrolysis has a power response capability of minutes or even seconds, and the operating power range can be from 5% to 150% of the rated power, which is more able to adapt to the strong volatility of photovoltaic output. Therefore, it is widely used in the field of photovoltaic hydrogen production. However, the problem of proton exchange membrane electrolysis hydrogen production is that the cost of electrolyzer materials is high and industrialization is difficult. Under the photovoltaic fluctuation input with a large range of changes and frequent start and stop, the proton exchange membrane electrolyzer will produce performance degradation, such as membrane degradation and catalyst activity reduction, which will lead to a decrease in hydrogen production efficiency.

[0004] The typical working condition of an electrolyzer refers to the working mode of a simulated test electrolyzer under a typical application scenario. However, no relevant standards have been set for the study of the typical working conditions of PEM electrolyzers in isolated hydrogen production scenarios. The study of the typical working conditions of electrolyzers for hydrogen production by electrolysis can provide test input conditions for the adaptability analysis of electrolyzers under laboratory conditions and the exploration of the laws of life performance degradation, thereby constructing an adaptability evaluation system, which is conducive to standardizing relevant industry standards and accelerating the development of the field of hydrogen production from renewable energy. Summary of the invention

[0005] The purpose of the present invention is to provide a method for extracting typical operating conditions of an electrolyzer in a scenario of photovoltaic-connected electrolyzer hydrogen production, so as to solve the problem that there are currently no typical operating conditions for electrolyzer hydrogen production.

[0006] The present invention takes into account the time-frequency characteristics of photovoltaics and the performance attenuation of the electrolyzer under fluctuating input, and obtains the working power characteristics of the electrolyzer in the scenario of producing green hydrogen using photovoltaics in the "Three Norths" region.

[0007] The technical solution adopted by the present invention is as follows:

[0008] Step S1, using empirical wavelet transform to perform modal decomposition on the original photovoltaic output curve P;

[0009] Step S2: for each mode, a dimensionality reduction analysis is performed on the photovoltaic output curve using characteristic indicators, and the decomposed curve is divided into three time periods for further operating condition feature extraction;

[0010] Step S3, using an objective weighting method to assign weights to the characteristic indicator sequence, and then using a weighted fuzzy clustering method to process the characteristic indicator sequence to obtain a typical characteristic indicator sequence;

[0011] Step S4: reconstruct a simplified typical operating curve of hydrogen production by electrolysis according to the characteristic indicator sequence, and construct a typical operating curve containing scenario probabilities according to the proportion of corresponding types of typical characteristic indicators.

[0012] Beneficial Effects

[0013] The present invention proposes a method for generating typical operating conditions of a proton exchange membrane electrolyzer for photovoltaic direct-coupled electrolysis to produce hydrogen, which makes up for the defect that there were no rules for formulating operating conditions related to proton exchange membrane electrolyzers.

[0014] The present invention utilizes empirical wavelet transformation to process the photovoltaic output curve, and can decompose the photovoltaic output curve into a frequency component mode of "basic output + fluctuating output", which is convenient for the subsequent extraction of characteristic indicators; the present invention takes into account the fluctuation characteristics of photovoltaic output and the attenuation characteristics of the electrolyzer, and comprehensively proposes a method for generating working conditions by feature extraction, clustering analysis, and simplified reconstruction. It is verified by examples that the curve reconstructed by this method has good similarity with the original curve, and can well characterize the working mode of the electrolyzer under typical photovoltaic output scenarios; and a complete working condition generation process is proposed, which has good operability and portability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is an overall flow chart of the implementation process of the present invention;

[0016] Figure 2 It is a schematic diagram of decomposing the photovoltaic output curve by the empirical wavelet transform in the present invention;

[0017] Figure 3 It is a schematic diagram of reconstructing the curve according to the characteristics;

[0018] Figure 4 It is a comparison diagram of the original curve and the reconstructed curve corresponding to each type of typical characteristic sequence after weighted fuzzy clustering in the present invention;

[0019] Figure 5 To generate the final simplified schematic diagram of the typical working condition. DETAILED DESCRIPTION

[0020] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] like Figure 1 As shown, it is a flow chart of the entire operation of generating the photovoltaic electrolysis hydrogen production operating condition curve proposed by the present invention.

[0022] Reference Figure 1 The specific steps for generating typical working conditions of photovoltaic electrolysis hydrogen production are as follows:

[0023] Step S1, using empirical wavelet transform to perform modal decomposition on the original photovoltaic output curve;

[0024] Step S101: divide the frequency domain band of the photovoltaic output power curve P by using the locmaxmin method to obtain the support interval of the normalized Fourier spectrum.

[0025] A n =[ω n-1 ,ω n ],n=1,2,…,N (1)

[0026]

[0027] In formula (1), Λ n is the nth frequency band interval after division, ω n is the boundary frequency value of the nth frequency band interval, and N is the determined number of intervals.

[0028] Step S102: Design an empirical wavelet function ψ n (ω) and the empirical scaling function The filter is composed of a filter used to extract the signal components within the divided interval.

[0029]

[0030]

[0031] In formula (3) and formula (4), τ n is the width of the nth interval frequency band buffer after division. The specific expression of the polynomial β is β(x)=x 4 (35-84x+70x 2 -20x 3 ).

[0032] Step S103: Detail coefficient and approximation coefficient The inner product is calculated by using the empirical wavelet function and the empirical scaling function with the original output curve, and the required number of approximation coefficients and detail coefficients are selected according to the filtering requirements to reconstruct the required photovoltaic output curve.

[0033]

[0034]

[0035]

[0036]

[0037] In formula (5) and (6),<x,y> represents the inner product of two signals, F -1 In equations (7) and (8), x*y represents the convolution of two signals.

[0038] Step S104: setting a frequency component threshold δ according to requirements, and dividing each interval in the normalized Fourier spectrum into one of low-frequency, medium-frequency, and high-frequency components according to the threshold.

[0039] The low-frequency component reflects the overall trend of the curve, the mid-frequency component reflects the degree of fluctuation of the curve, and the high-frequency component reflects the error of the force curve and the ripple of the power electronic equipment. The electrolytic stack does not need to bear this part and it is removed by filtering.

[0040] The present invention sets the low-frequency component threshold to 0.05 and the medium-frequency component threshold to 0.1. The decomposed components are Figure 2 exhibit. Figure 2 It shows the schematic diagram of each component after dividing the photovoltaic output curve data of a certain day into intervals according to the local maximum and minimum value (locmaxmin) rule and processing it with the empirical wavelet transform, that is, the result after the operation of steps S101-S104. The low-frequency component conforms to the ideal photovoltaic output curve trend, the medium-frequency component reflects the fluctuation characteristics of photovoltaic output in different periods, and the high-frequency component is removed by filtering and does not participate in hydrogen electrolysis.

[0041] Step S2: for each mode, a dimensionality reduction analysis is performed on the photovoltaic output curve using characteristic indicators, and the decomposed curve is divided into three time periods for further operating condition feature extraction;

[0042] Step S201, analyzing the attenuation characteristics of the electrolyzer. The electrolyzer for producing hydrogen by electrolysis will produce performance attenuation under fluctuating input. The three types of attenuation that mainly affect the life of the electrolytic stack are membrane degradation, catalyst degradation, and porous transport layer oxidation.

[0043] Under different input current densities, fluoride in the proton exchange membrane has different degradation rates under the dual effects of mechanical stress and chemical reaction, and the degradation will lead to a decrease in membrane thickness and membrane conductivity and an increase in the hydrogen and oxygen permeability coefficient.

[0044] When the input current suddenly changes, the anode iridium catalyst will be dissolved, poisoned, migrated, oxidized, etc., causing its effective surface area to decrease. As a result, the exchange current density decreases and the resistance of the activated area increases.

[0045] Under continuous input current, the titanium-containing porous transport layer (PTL) oxidizes, causing an increase in ohmic resistance and an increase in ohmic overvoltage. If an iridium coating is added to the titanium-containing PTL as a protective film, the formation of titanium oxide can be inhibited and the performance of the electrolytic cell can be improved.

[0046] The performance of the electrolytic hydrogen production cell will degrade under extreme current density and sudden input, and the operating condition indicators need to extract the current density and current density fluctuation.

[0047] Step S202: Combine the volatility of photovoltaic power output and the attenuation characteristics of the electrolytic cell to extract the characteristic indicators of the low-frequency and medium-frequency components obtained in step S1.

[0048] First, the photovoltaic output period is decomposed into three typical periods: the morning period is 7:00-11:00; the noon period is 11:00-13:00; the afternoon period is 13:00-17:00, and they are represented by subscripts 1, 2, and 3 respectively.

[0049] Based on the low-frequency component of the photovoltaic output curve, the following two characteristic indicators are defined:

[0050] 1) The average value during the noon period can reflect the average size of the maximum photovoltaic output period and the maximum load level of the electrolyzer.

[0051]

[0052] 2) The ramp rate in the morning and afternoon periods can reflect the speed of change of photovoltaic output and the load change rate of the electrolyzer in a certain period of time.

[0053]

[0054] Based on the intermediate frequency component of the photovoltaic output curve, the following two characteristic indicators are defined:

[0055] 3) The fluctuation amplitudes in the morning, noon and afternoon periods can measure the average peak-to-valley difference of the intermediate frequency component of the photovoltaic output curve and the power fluctuation of the electrolyzer.

[0056]

[0057] 4) The fluctuation frequency in the morning, noon and afternoon periods can measure the number of peaks and valleys in the intermediate frequency component of the photovoltaic output curve and the power fluctuation speed of the electrolyzer.

[0058]

[0059] In formulas (9)-(12), X k | k＝1～9 represents the kth feature sequence extracted, i represents the time period (1 represents morning, 2 represents noon, and 3 represents afternoon), N s Indicates the number of peaks and valleys of the curve within a certain period of time.

[0060] Step S3: Use an objective weighting method to assign weights to the characteristic indicator sequence, and then use a weighted fuzzy clustering method to process the characteristic indicator sequence to obtain a typical characteristic indicator sequence.

[0061] Step S301: Perform feature weight configuration on the feature sequence extracted in step S2 to obtain weight items for weighted fuzzy clustering analysis.

[0062] The objective weighting method is used to weight the characteristic sequence. First, the photovoltaic output characteristic sequence is normalized, and then the indicator variability s is used to j and index conflict R j Evaluate the feature sequence, assign corresponding weights according to the evaluation results, and obtain the weight matrix W j .

[0063]

[0064]

[0065]

[0066]

[0067] In formula (13), Represents the average value of the jth feature index in the data set, s j Represents the standard deviation of the jth feature index in the data set, r ij is the correlation coefficient between index i and index j, C j is the evaluation coefficient of the jth characteristic index.

[0068] Step S302: Determine the optimal number of clusters c according to the cluster validity function.

[0069] The clustering validity function g applicable to fuzzy clustering is given by:

[0070] g(U, V) = wsd-sep (17)

[0071]

[0072]

[0073] In formula (17), wsd represents the degree of clustering of samples within a class, and sep represents the degree of separation between classes.

[0074] In formula (18), the parameters

[0075] In formula (19), the overlap function

[0076] When the value of function g is the smallest, the corresponding number of clusters c is the optimal number of clusters.

[0077] Step S303: clustering the weighted feature sequences using weighted fuzzy clustering analysis.

[0078] The specific clustering operation is to repeatedly iterate and solve the membership function U and the cluster center V until the distance function F reaches the minimum value or the number of iterations reaches the upper limit. The iterative formula of weighted fuzzy clustering analysis is:

[0079]

[0080]

[0081]

[0082] In formula (20), n is the total number of samples, c is the number of clusters, m is the total number of feature indicators, and U ij is the membership function of the i-th sample to the j-th cluster center. The specific calculation formula is shown in formula (21). jk is the jth cluster center corresponding to the kth characteristic index, and the specific calculation formula is shown in formula (22). q is a constant 2. d(X ik , V jk ) is used to calculate X ik With V jk The weighted Euclidean distance between them.

[0083] Step S4: reconstruct a simplified typical operating curve of hydrogen production by electrolysis according to the characteristic indicator sequence, and construct a typical output curve according to the proportion of corresponding types of typical characteristic indicators.

[0084] Step S401: using the maximum output index X1 of each type of typical characteristic sequence, and maintaining the output value in the corresponding working condition period.

[0085] Step S402: using the climbing rate index X2, X3 of each type of typical characteristic sequence, climbing at the rate in the entire operating range. The above two steps form a simplified low-frequency component.

[0086] Step S403, using the fluctuation amplitude indexes X4, X5, X6 and the fluctuation frequency indexes X7, X8, X9 in each output period of each type of typical characteristic sequence, a rectangular wave with the same amplitude and frequency is constructed, that is, a simplified intermediate frequency component.

[0087] Step S404: superimpose the simplified low-frequency component and the medium-frequency component signals to obtain a simplified photovoltaic electrolysis hydrogen production electrolyzer operating condition curve under each typical scenario.

[0088] Figure 3 This is a schematic diagram of the reconstruction part of steps S401-S404. This schematic diagram is a summary of steps S401-S404, that is, using the extracted characteristic indicators as the basis for reconstructing data, the climbing rate and the operating output level are reconstructed into a new low-frequency component, and the fluctuation amplitude and the fluctuation frequency (collectively referred to as the fluctuation rate) are reconstructed into a new high-frequency component. The superposition of the two components is the newly generated simplified photovoltaic electrolysis hydrogen production electrolyzer operating curve.

[0089] Step S405: Calculate the cluster ratio of the photovoltaic output characteristic sequence corresponding to each typical scenario, reconstruct according to the ratio, and generate a photovoltaic electrolysis hydrogen production electrolyzer operating condition curve containing the scenario occurrence probability.

[0090] Step S406: If it is necessary to further meet the requirements for simplifying laboratory tests, the generated curve can be smoothed, a typical power operating point of the electrolytic hydrogen production cell is selected, a smoothing rule is formulated, nearby power output points are classified as typical power operating points, and the generated typical operating curve of electrolytic hydrogen production is further simplified.

[0091] The specific smoothing rules are shown in the following table:

[0092] Table 1 Simplified smoothing rules for power range of working conditions

[0093]

[0094] The three-year output data of a photovoltaic power station in a remote area of ​​the Three Northern Areas is used as the basic data, from which data corresponding to some site maintenance or extremely rare extreme weather conditions are excluded. The rated capacity of the photovoltaic power station is 20MW, and the total rated capacity of the PEM electrolysis stack is 10MW.

[0095] By applying the working condition generation method of the present invention to process the data, three major components can be obtained after decomposition of all output data. After feature extraction and cluster analysis, the corresponding clustered feature sequence is obtained, and then a simplified test working condition curve is reconstructed based on the sequence.

[0096] The clustering results obtained by the clustering method proposed in the present invention are as follows.

[0097] Table 2 Feature sequence of the first clustering results

[0098] X1 X2 X3 X4 X5 Normalized value 0.7252 0.4634 0.6094 0.1832 0.0111 X6 X7 X8 X9 Normalized value 0.0391 0.0169 0.3871 0.2133

[0099] Table 3 Feature sequences of the second clustering results

[0100] X1 X2 X3 X4 X5 Normalized value 0.5566 0.3196 0.7048 0.0582 0.0968 X6 X7 X8 X9 Normalized value 0.1748 0.1695 0.1290 0.1600

[0101] Table 4 Feature sequence of the third clustering result

[0102] X1 X2 X3 X4 X5 Normalized value 0.0846 0.0412 0.9512 0.0444 0.0863 X6 X7 X8 X9 Normalized value 0.0181 0.0339 0.0753 0.0667

[0103] Table 5 Characteristic sequence of the fourth clustering result

[0104] X1 X2 X3 X4 X5 Normalized value 0.7291 0.4458 0.6172 0.2560 0.1524 X6 X7 X8 X9 Normalized value 0.2112 0.1864 0.0860 0.0933

[0105] The first type of clustering result has the largest X1, and the values ​​of X2 and X3 are relatively large; X4 is large and X7 is small; X5 is small and X8 is large; X6 is small and X9 is large; it represents the working curve of the electrolytic cell under a sunny day, which changes load at a uniform rate to an overload condition, and then changes load at a uniform rate to shutdown, with a small degree of fluctuation in between.

[0106] The second type of clustering results has larger values ​​of X1, X2, and X3; smaller values ​​of X4 and X7; smaller values ​​of X5 and X8; and smaller values ​​of X6 and X9; which represents the working curve of the electrolytic cell under small fluctuations on a cloudy day, with low-frequency and high-amplitude components superimposed during the climbing process.

[0107] The third type of clustering results has the smallest X1, very small X2, and the largest X3; X4 and X7 are very small; X5 and X8 are very small; X6 and X9 are very small; this represents the working curve of the electrolytic cell under rain and snow, which is basically in the low-load operating range.

[0108] The fourth type of clustering results has larger values ​​of X1, X2, and X3; larger values ​​of X4 and X7; larger values ​​of X5 and X8; and larger values ​​of X6 and X9; which represents the working curve of the electrolytic cell under large fluctuations on a cloudy day, with high-frequency and low-amplitude components superimposed during the climbing process.

[0109] The comparison between the original output curve corresponding to the characteristic sequence and the curve simplified and reconstructed using the characteristic sequence is shown in the figure below. Figure 4 shown. Figure 4It is a comparison diagram of four types of typical operating curves before and after reconstruction, which shows that the operating curve simplification method of the present invention can correctly restore the overall power size and fluctuation of the original output curve, that is, the simplification method proposed in the present invention has certain effectiveness, and the generated operating curve has the advantage of being easy to implement the test, and can retain the output characteristics of the typical photovoltaic output curve, which is of testing significance.

[0110] The mean absolute percentage error (MAPE) of the simplified typical curve and the original curve is calculated, and the results are shown in the following table.

[0111] Table 6 Average absolute percentage error between simplified operating curve and original curve

[0112]

[0113] In the case where there is no strict alignment in timing, the calculated mean absolute percentage error value is still less than 0.4, which proves the effectiveness of the simplified method proposed in the present invention.

[0114] Figure 5 The final simplified operating curve generated by proportional construction is shown. The operating curve starts from 0min, and the simplified typical operating curve generated by the above reconstruction is generated according to the proportion of the scene within the input range of 5%-150% of the proton exchange membrane electrolyzer, and finally shut down to form a complete operating test. The entire test process lasts 11520min. When a cyclic test is required, the simplified typical operating curve can be repeatedly added during the start and stop process. The figure shows the output range and power changes of the electrolyzer when working under different photovoltaic input scenario ratios.

[0115] The technical solution of the present invention decomposes the photovoltaic output curve according to the output characteristics through empirical wavelet transform, and combines the photovoltaic output characteristics with the working characteristics of the electrolyzer. According to the pattern of characteristic sequence clustering, a method for generating typical working conditions of a photovoltaic electrolysis hydrogen production electrolyzer is proposed, which solves the gap in the relevant generation technology of the test working conditions of the proton exchange membrane water electrolyzer with photovoltaic input, and is of great significance for providing basic input data for the subsequent performance attenuation test of the electrolysis hydrogen production electrolyzer and the in-depth exploration of its performance attenuation characteristics.

[0116] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for generating typical working conditions of photovoltaic water electrolysis hydrogen production, characterized by: The following steps are involved: Step S1, using empirical wavelet transform to perform modal decomposition on the original photovoltaic output curve P; Step S2: for each mode, the photovoltaic output curve is analyzed by using characteristic indicators for dimension reduction, and the decomposed curve is divided into three time periods for operating condition feature extraction; the method of using characteristic indicators for dimension reduction for each mode and dividing the curve into three time periods for operating condition feature extraction includes: Analyze the attenuation characteristics of the electrolyzer. Considering that the electrolytic hydrogen production electrolyzer will have performance attenuation under extreme current density and sudden input, the operating condition indicators need to extract the current density and current density fluctuation; The decomposed curve is divided into three time periods: morning period 7:00-11:00, noon period 11:00-13:00, afternoon period 13:00-17:00, and they are represented by subscripts 1, 2, and 3 respectively; Combining the volatility of the photovoltaic output curve and the attenuation characteristics of the electrolyzer, the low-frequency component of the curve is defined and the average photovoltaic output during the noon period is extracted: Morning and afternoon climbing rates: For the intermediate frequency component of the curve, define and extract the fluctuation amplitude and fluctuation frequency of three time periods: in: ； Step S3, using an objective weighting method to assign weights to the characteristic indicator sequence, and then using a weighted fuzzy clustering method to process the characteristic indicator sequence to obtain a typical characteristic indicator sequence; Step S4: reconstruct a simplified typical operating curve of hydrogen production by electrolysis according to the characteristic indicator sequence, and construct a typical operating curve containing scenario probabilities according to the proportion of corresponding types of typical characteristic indicators.

2. The method for generating typical working conditions of photovoltaic water electrolysis hydrogen production according to claim 1, characterized in that: The step 1 further includes the following: The frequency spectrum of the original photovoltaic output curve P is divided into frequency bands according to the locmaxmin rule to obtain the support interval of the frequency-normalized Fourier spectrum; Using the empirical wavelet function ψ n (ω) and the empirical scaling function Design an empirical wavelet transform filter, the filter is used to extract the signal components within the divided interval; Select the required approximation coefficient according to the filtering requirements and detail coefficient The number of , reconstructs the required photovoltaic output curve; in: 。 3. The method for generating typical working conditions of photovoltaic water electrolysis hydrogen production according to claim 1, characterized in that: The method of using an objective weighting method to configure weights for a characteristic indicator sequence and then using a weighted fuzzy clustering method to obtain a typical characteristic indicator sequence includes: The objective weighting method is used to configure the feature weights of the extracted feature sequence to obtain the weight items of weighted fuzzy clustering analysis; According to the clustering validity function g, the number of clusters corresponding to its minimum value is solved as the optimal number of clusters c; The weighted fuzzy clustering analysis is used to cluster the weighted feature sequence, that is, the membership function U is iteratively solved. ij With cluster center V ij , after iterating until the distance function F(U, V, W) reaches the minimum or the number of iterations meets the requirements, the feature sequence corresponding to the cluster center is obtained, that is, the typical feature sequence; in: 。 4. The method for generating typical working conditions of photovoltaic water electrolysis hydrogen production according to claim 1, characterized in that: The step 4 further includes the following contents: A simplified operating curve of the photovoltaic electrolysis hydrogen production electrolyzer is reconstructed according to the characteristic quantities of the typical characteristic sequence, X1 is used as the peak output of the low-frequency component of the reconstructed curve, X2 and X3 are used as the climbing rate of the low-frequency component of the reconstructed curve, X4, X5, and X6 are used as the rectangular wave amplitude of the intermediate frequency component of the reconstructed curve, and X7, X8, and X9 are used as the rectangular wave frequency of the intermediate frequency component of the reconstructed curve; finally, the low-frequency component and the intermediate frequency component of the reconstructed curve are combined to form the final component; Calculate the cluster ratio of the photovoltaic output characteristic sequence corresponding to each typical scenario, and construct a photovoltaic electrolysis hydrogen production electrolyzer operating curve containing the probability of scenario occurrence based on the ratio; If it is necessary to further meet the requirements for simplifying laboratory tests, the generated curve can be smoothed, the typical power operating point of the electrolytic hydrogen production cell can be selected, smoothing rules can be formulated, and the nearby power output points can be classified as typical power operating points to further simplify the generated typical operating curve of electrolytic hydrogen production.

5. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored program, wherein the program controls the device where the non-volatile storage medium is located to execute the method according to any one of claims 1 to 4 when the program is executed.

6. An electronic device, characterized in that: It comprises a processor and a memory; the memory stores computer-readable instructions, and the processor is used to execute the computer-readable instructions, wherein the computer-readable instructions execute the method described in any one of claims 1 to 4 when executed.

Images