Control method and system of PEM water electrolysis hydrogen production equipment

Abstract

The invention belongs to the technical field of PEM water electrolysis hydrogen production, and provides a control method and system of PEM water electrolysis hydrogen production equipment. The method comprises the following steps: extracting a differential pressure oscillation signal and analyzing to obtain a differential pressure unbalance area; through associated risk modeling of the pressure difference and the permeability, outputting a risk level for predicting hydrogen permeation; executing a differential intervention strategy, extracting effective inhibition characteristics, and constructing a two-dimensional evaluation domain to judge whether hydrogen permeation is completely inhibited or not; and if the hydrogen is not completely inhibited, extracting and screening a hydrogen reverse migration path to obtain a response isolation path and a permeation duration coefficient, and triggering equipment safety isolation and inerting response control according to the permeation duration coefficient. The control system comprises an unbalance analysis module, a penetration analysis module, an inhibition judgment module, a path extraction module and a response control module. According to the invention, risk assessment and efficient safety control of the PEM water electrolysis hydrogen production equipment can be realized, and the safety and stability of equipment operation are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of PEM water electrolysis hydrogen production, and in particular relates to a control method and system for PEM water electrolysis hydrogen production equipment. Background Art

[0002] The proton exchange membrane (PEM) is a core component of PEM water electrolysis hydrogen production equipment. Its performance directly affects electrolysis efficiency and operational safety. During the PEM water electrolysis hydrogen production process, protons migrate from the anode to the cathode through the PEM. The pressure differential between the anode and cathode, as well as hydrogen permeation behavior, must be strictly controlled. Failure to do so can easily lead to membrane damage, reduced efficiency, and even safety accidents.

[0003] Existing technologies have numerous deficiencies in the control of PEM water electrolysis hydrogen production equipment: In differential pressure monitoring and imbalance location, they are unable to distinguish between high-frequency, minute pressure fluctuations at the component level and low-frequency gradient changes at the system level. This makes it difficult to identify the imbalanced area when the hydrogen side pressure drops suddenly, increasing the blindness of troubleshooting. In hydrogen permeation risk assessment, there is a lack of coordinated analysis of local permeation at the component level and overall permeation at the system level. This makes it impossible to accurately calculate the hydrogen permeation flux through physical mechanism models at the membrane electrode scale, and it is difficult to construct a global risk association based on the system-level flow field and pressure distribution, which can easily lead to misjudgment or omission of risks.

[0004] In terms of risk intervention, existing technologies have not formulated differentiated strategies for imbalances within components and systems, the intervention is not targeted enough, and there is a lack of quantitative evaluation methods for the intervention effects, making it difficult to determine whether the suppression is effective; in terms of secondary risk prevention and control, there is insufficient tracking of secondary pressure fluctuations and hydrogen migration paths after intervention, and it is impossible to block the spread of risks in a timely manner; in terms of safety response, most of the control is triggered based on real-time pressure difference or concentration data, and the impact of free diffusion and trap-state adsorption of hydrogen in the material on the persistence of permeation is not considered. It is difficult to achieve a graded response based on the nature of the risk, which restricts the safety and stability of equipment operation.

[0005] To this end, the present invention provides a control method and system for a PEM water electrolysis hydrogen production device. Summary of the Invention

[0006] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] A control method for a PEM water electrolysis hydrogen production device comprises the following steps:

[0009] Collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area;

[0010] Obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform risk modeling on the correlation between pressure difference and permeability, and output the predicted permeation risk level;

[0011] Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective suppression features are extracted from the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective suppression features to determine whether hydrogen permeation has been completely and effectively suppressed.

[0012] If complete and effective suppression is not achieved, the electrolyte flow rate is adjusted in stages, and the pressure difference oscillation signal energy and permeation suppression rate after the staged adjustment are extracted to determine the secondary fluctuation state. If it exists, the hydrogen reverse migration path is extracted, and the hydrogen reverse migration path is screened and processed to obtain the response isolation path.

[0013] As a further technical solution of the present invention: the method of performing the regional analysis is:

[0014] Obtain component-level and system-level differential pressure oscillation signals, perform spatiotemporal correlation analysis, and obtain spatiotemporal correlation coefficients.

[0015] Extract the pressure difference oscillation signal energy corresponding to the component-level pressure difference oscillation signal and the system-level pressure difference oscillation signal, and construct the imbalance area positioning function based on the spatiotemporal correlation coefficient;

[0016] The pressure difference imbalance area where the pressure drops suddenly on the hydrogen side of the PEM electrolyzer is determined based on the imbalance area positioning function.

[0017] As a further technical solution of the present invention: the method of obtaining the component-level pressure difference oscillation signal and the system-level pressure difference oscillation signal is:

[0018] Construct a layered differential pressure monitoring network in the PEM electrolyzer system, dividing the differential pressure data into component-level differential pressure data and system-level differential pressure data;

[0019] The pressure difference data at the component level and the pressure difference data at the system level are subjected to data preprocessing and dynamic feature extraction respectively to obtain the pressure difference oscillation signal at the component level and the pressure difference oscillation signal at the system level.

[0020] As a further technical solution of the present invention, the method of performing the correlation risk modeling of the pressure difference and permeability is:

[0021] Establish a physical mechanism model at the membrane electrode scale to obtain the hydrogen permeation flux at the component level;

[0022] Establish a system-scale physical mechanism model to obtain the system-level hydrogen permeation flux;

[0023] Calculate a risk index based on component-level and system-level hydrogen permeation fluxes, and determine the risk level of the pressure difference imbalance area based on the risk index;

[0024] If the risk level of the pressure difference imbalance area is at a medium or low risk level, a dynamic risk model of pressure difference-permeability is constructed to output the predicted permeability risk level.

[0025] As a further technical solution of the present invention: the method for calculating the risk index is:

[0026] Get the level imbalance factor;

[0027] Obtaining the excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold and the excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold and performing summing processing;

[0028] The summed result is multiplied by the hierarchical imbalance factor to obtain the penetration risk index.

[0029] As a further technical solution of the present invention: the method for obtaining the hierarchical imbalance factor is:

[0030] Obtain the pressure difference imbalance area within the component layer and the pressure difference imbalance area within the system layer corresponding to the high-risk interval, and obtain the pressure difference imbalance area within the high-risk component layer and the pressure difference imbalance area within the high-risk system layer;

[0031] If the pressure imbalance area within the high-risk component layer or the pressure imbalance area within the high-risk system layer has a conduction effect in the electrolyzer, or if there is a pressure imbalance area between adjacent battery packs in the electrolyzer, then the pressure imbalance area within the component layer or the pressure imbalance area within the system layer shall be merged to obtain the total imbalance area within the component layer and the total imbalance area within the system layer;

[0032] The area overlap between the total imbalance area within the component layer and the total imbalance area within the system layer is calculated to obtain the hierarchical imbalance factor.

[0033] As a further technical solution of the present invention: the method for determining whether the hydrogen permeation is completely and effectively suppressed is:

[0034] Obtain the imbalance shrinkage ratio and permeation inhibition rate of all pressure imbalance areas and construct an inhibition efficiency matrix;

[0035] Based on the inhibition efficiency matrix, a two-dimensional inhibition evaluation domain is constructed with the imbalance shrinkage ratio as the horizontal axis and the penetration inhibition rate as the vertical axis. The imbalance shrinkage ratio and penetration inhibition rate of the inhibition efficiency matrix are used as evaluation coordinates.

[0036] An inhibition evaluation boundary line is set, and the evaluation coordinates are mapped to a two-dimensional inhibition evaluation domain based on the inhibition evaluation boundary line. If the evaluation coordinates of the pressure imbalance area are in the third quadrant of the two-dimensional inhibition evaluation domain, it is considered that hydrogen permeation has not been completely and effectively suppressed.

[0037] As a further technical solution of the present invention: the method for obtaining the unbalanced shrinkage ratio and the permeation inhibition rate is:

[0038] If the predicted infiltration risk level is high, different types of differentiated intervention control strategies are formulated based on the type of pressure imbalance area;

[0039] Obtain the absolute change in the area of ​​all pressure difference imbalance regions before and after the intervention control, and perform ratio processing on the absolute change and the area of ​​the pressure difference imbalance region before the intervention control to obtain the imbalance contraction ratio;

[0040] Obtain the hydrogen permeation flux before and after the intervention control in the pressure imbalance area, and construct the permeation concentration field of each pressure imbalance area before and after the intervention;

[0041] Calculate the hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure imbalance area before and after intervention;

[0042] The deviation ratio of the hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure difference imbalance area before and after intervention was calculated to obtain the permeation inhibition rate.

[0043] As a further technical solution of the present invention: the response isolation path is obtained by:

[0044] The peak hydrogen permeation concentration and the pressure gradient of each imbalanced hydrogen reverse migration path are extracted and dimensionlessly processed. The dimensionlessly processed peak hydrogen permeation concentration and pressure gradient are summed to obtain the path characteristic coefficient.

[0045] Obtaining a pressure difference imbalance region where a secondary fluctuation suppression state exists as a pressure difference selection region;

[0046] Obtain the number of associated pressure difference selection zones in the hydrogen reverse migration path of each imbalance type, calculate the ratio of the number of pressure difference selection zones to the total pressure difference imbalance zones, and obtain the regional correlation degree;

[0047] The path characteristic coefficient and the regional correlation degree are calculated to obtain the path selection coefficient;

[0048] Obtaining the path selection coefficients of the hydrogen reverse migration paths of the component-level imbalance type and the system-level imbalance type;

[0049] The hydrogen reverse migration pathway with the highest pathway selectivity coefficient was selected as the response isolation pathway.

[0050] A control system for a PEM water electrolysis hydrogen production device, comprising the following modules:

[0051] Imbalance Analysis Module: This module is used to collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area.

[0052] Permeation analysis module: used to obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform correlation risk modeling of pressure difference and permeability, and output the predicted permeation risk level;

[0053] Inhibition determination module: Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective inhibition features are extracted from the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective inhibition features to determine whether hydrogen permeation has been completely and effectively suppressed.

[0054] Path extraction module: If complete and effective suppression is not achieved, the electrolyte flow rate is adjusted in stages. The energy of the pressure difference oscillation signal and the permeation suppression rate after the staged adjustment are extracted to determine the secondary fluctuation state. If the secondary fluctuation state exists, the hydrogen reverse migration path is extracted and screened to obtain the corresponding isolation path.

[0055] Response control module: used to extract the free state and trap state characteristics of the response isolation path, combine the trap effect to perform permeation persistence analysis to obtain the permeation persistence coefficient, and determine whether to trigger the safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient.

[0056] The beneficial effects of the present invention are as follows:

[0057] 1. By collecting differential pressure data between the cathode and anode of the PEM electrolyzer and establishing a layered differential pressure monitoring network, high-precision micro-differential pressure sensors are used at the component level to capture high-frequency small fluctuations, while conventional differential pressure sensors are used at the system level to capture low-frequency gradient changes. This enables multi-scale and refined monitoring of differential pressure data. Combined with spatiotemporal correlation analysis of the differential pressure oscillation signal and an imbalance region location function, the system can locate the differential pressure imbalance region where the hydrogen side pressure drops suddenly. This overcomes the limitation of traditional overall monitoring that makes it difficult to distinguish between local and system-level pressure anomalies, provides regional guidance for subsequent targeted intervention, and improves the accuracy and efficiency of fault location.

[0058] 2. Component-level micro hydrogen sensors and system-level hydrogen concentration transmitters are used to collect hydrogen permeation concentrations at different levels. The hydrogen permeation flux is calculated by combining physical mechanism models at the membrane electrode and system scales, enabling a coordinated analysis of hydrogen permeation behavior at both the local component and overall system levels. By constructing a risk index and assessing risk levels based on spatial correlation, both local microscopic risks and global systemic risks are considered. A dynamic correlation model predicts risk escalation trends, reducing risk misjudgments or omissions caused by single-scale assessments and providing a timely basis for risk intervention.

[0059] 3. Differentiated intervention strategies are implemented based on risk level and imbalance area type. Local electric field gradients are adjusted for internal component imbalances, and regional voltage distribution is optimized for internal system imbalances, improving the targeted nature of intervention measures. By extracting the imbalance shrinkage ratio and permeation suppression rate as effective suppression features and constructing a two-dimensional evaluation domain to determine the hydrogen permeation suppression effect, the actual impact of intervention measures can be evaluated, making the intervention effects quantifiable and verifiable, and reducing the continued accumulation of risks caused by ineffective interventions.

[0060] 4. When hydrogen permeation is not completely suppressed, the electrolyte flow is adjusted in stages based on the negative correlation adjustment model, and the adjustment amplitude can be dynamically matched according to different suppression effects, so that the flow control is more in line with actual needs; by monitoring the secondary fluctuation state and locating the starting point of hydrogen migration in combination with the spatiotemporal distribution, it is beneficial to timely discover the risk rebound after intervention; and screening the response isolation path based on the path characteristic coefficient and regional correlation provides a clear target for blocking the reverse migration of hydrogen, effectively preventing the further spread of risks; by extracting the free state and trap state characteristics of the response isolation path and calculating the permeation persistence coefficient in combination with the trap effect, the continuous risk of hydrogen permeation can be deeply analyzed; based on the permeation persistence coefficient, the corresponding safety isolation and inerting response control are triggered, and graded measures from enhanced control to emergency shutdown can be flexibly taken according to the degree of risk, which is beneficial to reduce the possibility of safety accidents caused by hydrogen permeation and provide a reliable line of defense for the safe and stable operation of hydrogen production equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The present invention will be further described below with reference to the accompanying drawings.

[0062] Figure 1 This is a flow chart of a control method for a PEM water electrolysis hydrogen production device according to an embodiment of the present invention;

[0063] Figure 2 is a flow chart for determining whether hydrogen permeation has been completely and effectively suppressed according to an embodiment of the present invention;

[0064] Figure 3 This is a control system module architecture diagram of a PEM water electrolysis hydrogen production device described in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0066] Example 1

[0067] See also Figure 1 As shown, a control method for a PEM water electrolysis hydrogen production device according to an embodiment of the present invention includes the following steps:

[0068] Step 1: Collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area;

[0069] The method for collecting the cathode and anode pressure differential data of the PEM electrolyzer and establishing a layered pressure differential monitoring network for the cathode and anode pressure differential data is as follows:

[0070] Preferably, a layered pressure differential monitoring network is constructed in the PEM electrolyzer system to divide the pressure differential data into component layer pressure differential data and system layer pressure differential data;

[0071] It is understandable that the purpose of building a layered pressure differential monitoring network is to:

[0072] Objective 1: Distinguish pressure characteristics at different scales: By dividing differential pressure data into component and system layers, we can specifically capture pressure changes of different frequencies and amplitudes, alleviating the limitation of traditional single-scale monitoring that cannot distinguish between local and system-level pressure anomalies.

[0073] Objective 2: Support imbalance area location: By performing spatiotemporal correlation analysis on component-level and system-level pressure differential oscillation signals, combined with an imbalance area location function, the specific imbalance area where the hydrogen side pressure drops suddenly can be located. This reduces the misjudgment or omission of fault scope caused by traditional overall monitoring, provides clear regional guidance for subsequent differentiated intervention, and improves the accuracy and efficiency of fault location.

[0074] Objective 3: Extracting differential pressure oscillation signals: Based on the dynamic characteristics of hierarchical data, component-level and system-level oscillation signal extraction rules are set separately. This can generate differential pressure oscillation signals at different levels, providing a more detailed data source for subsequent analysis.

[0075] The pressure difference data of the component layer is obtained by deploying component-level monitoring points on both sides of the membrane electrode and the plate flow channel outlet. A high-precision micro-pressure difference sensor is set at each component monitoring point to capture high-frequency small fluctuations in pressure.

[0076] System-level differential pressure data is obtained by deploying system-level monitoring points at the inlet and outlet of adjacent cells in the electrolysis stack. Conventional differential pressure sensors are set at these system-level monitoring points to capture low-frequency gradient changes in pressure.

[0077] It can be understood that the high-precision micro differential pressure sensor has a range of ≤±5kPa, an accuracy of ≥0.1%FS, a sampling frequency of 1KHz, and can capture high-frequency small fluctuations of 10 to 100H;

[0078] The conventional differential pressure sensor has a range of ≤±50kPa, an accuracy of ≥0.5%FS, and a sampling frequency of 100Hz, which is used to capture low-frequency gradient changes of 0.1 to 10Hz;

[0079] Among them, the pressure difference oscillation signal generated by the layered pressure difference monitoring network is extracted, and the pressure drop on the hydrogen side of the PEM electrolyzer is analyzed based on the pressure difference oscillation signal. The pressure difference imbalance area is obtained as follows:

[0080] The pressure difference data at the component level and the pressure difference data at the system level are subjected to data preprocessing and dynamic feature extraction, respectively. Based on the dynamic features, component-level pressure difference oscillation signal extraction rules and system-level pressure difference oscillation signal extraction rules are set;

[0081] It is understandable that for the pressure difference data at the component level, data preprocessing is first performed: high-frequency filtering is used to remove low-frequency interference, sensor noise is eliminated through a smoothing algorithm, and high-frequency pressure fluctuation characteristics of 10-100Hz are retained; dynamic feature extraction is then performed, focusing on extracting the peak amplitude, frequency distribution, duration, and instantaneous change rate characteristics of the pressure fluctuation. Based on these characteristics, the component-level pressure difference oscillation signal extraction rule is set: when the fluctuation frequency is within the range of 10-100Hz, the peak amplitude exceeds 5% of the average pressure value of the component under normal operating conditions, and the duration is ≥0.1 second, it is determined to be a component-level pressure difference oscillation signal;

[0082] For system-level pressure differential data, data preprocessing uses low-frequency filtering (such as a 10Hz low-pass filter) to retain low-frequency trends of 0.1 to 10Hz, and baseline correction is used to eliminate drift caused by factors such as ambient temperature. Dynamic feature extraction focuses on features such as the pressure gradient change rate, the pressure differential difference between adjacent monitoring points, and the fluctuation period. Based on these features, the system-level pressure differential oscillation signal extraction rule is set: when the pressure gradient change rate exceeds 10% of the design value, and the pressure differential difference between adjacent single battery groups exceeds ±2kPa for more than 3 seconds, it is determined to be a system-level pressure differential oscillation signal.

[0083] If the pressure difference data of the component layer meets the component-level pressure difference oscillation signal extraction rule, a component-level pressure difference oscillation signal is generated;

[0084] If the pressure difference data of the system layer meets the system-level pressure difference oscillation signal extraction rule, then a system-level pressure difference oscillation signal is generated;

[0085] Among them, the method of locating the pressure imbalance area where the pressure on the hydrogen side of the PEM electrolyzer drops suddenly is:

[0086] Obtain component-level and system-level differential pressure oscillation signals, perform spatiotemporal correlation analysis, and obtain spatiotemporal correlation coefficients. , i is the number of the high-precision micro differential pressure sensor, j is the number of the conventional differential pressure sensor;

[0087] Those skilled in the art will appreciate that the high-frequency oscillation signal at the component level is time-synchronized with the low-frequency oscillation signal captured by the system-level conventional differential pressure sensor, and the same continuous sampling time window is selected. Within the physical area covered by the system-level sensor, the component-level sensor is located within the physical area covered by the system-level sensor. The Pearson correlation coefficient is then applied only to the two types of signals within the same area to obtain a spatiotemporal correlation coefficient reflecting the abnormal correlation between the pressure difference between the local area and the overall area to which it belongs, providing a basis for locating the imbalanced area.

[0088] By establishing the imbalance region positioning function: ;

[0089] Wherein, i is the number of the high-precision micro differential pressure sensor, and j is the number of the conventional differential pressure sensor;

[0090] is the pressure difference oscillation signal energy monitored by the i-th high-precision micro-pressure difference sensor;

[0091] The energy of the pressure difference oscillation signal monitored by the jth conventional pressure difference sensor;

[0092] As the baseline energy, is the benchmark parameter;

[0093] It can be understood that the pressure differential oscillation signals collected by the i-th high-precision micro-pressure differential sensor and the j-th conventional pressure differential sensor are first subjected to bandpass filtering for noise reduction, and then wavelet decomposition is used to extract the energy of each frequency band, and the energy value of the corresponding sensor monitoring signal is quantified;

[0094] When the electrolyzer is operating stably at rated current and pressure balance, multiple sets of historical data are continuously collected to calculate the energy average of the component-level and system-level pressure difference signals under normal conditions as the baseline energy.

[0095] Determined by calibration experiment: In the test scenario of known pressure difference imbalance area, technicians in this field adjust The value of makes the positioning error of the imbalance area positioning function to the real imbalance area optimal, and determines , to adapt the weight relationship between system-level and component-level signal energy;

[0096] Determine the pressure difference imbalance area where the pressure drops suddenly on the hydrogen side of the PEM electrolyzer based on the imbalance area positioning function;

[0097] It will be understood by those skilled in the art that by solving the sensor number combination (i, j) that maximizes the function value of the local area of ​​the PEM electrolyzer, the local area monitored by the corresponding component-level or system-level sensor is the pressure difference imbalance area where the hydrogen side pressure drops suddenly;

[0098] Step 2: Obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform risk modeling on the correlation between pressure difference and permeability, and output the predicted permeation risk level;

[0099] The method for obtaining the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area is:

[0100] Based on the pressure imbalance area, the pressure imbalance area is matched with the imbalance area type of the PEM electrolyzer to obtain the pressure imbalance area within the component layer and the pressure imbalance area within the system layer;

[0101] If the pressure difference is unbalanced in the component layer, the micro hydrogen sensors arranged at the edge of the component layer will collect the hydrogen concentration values ​​on the anode and cathode sides;

[0102] If it is an imbalance area within the system layer, the hydrogen concentration value is collected based on the hydrogen concentration transmitters installed at the inlet and outlet of the electrolytic stack and the single cell group;

[0103] The method for modeling the associated risk of pressure difference and permeability is as follows:

[0104] S201. Establish a physical mechanism model at the membrane electrode scale to obtain the hydrogen permeation flux at the module level;

[0105] Preferably, based on the hydrogen concentration values ​​of the anode and cathode sides collected by the component layer micro hydrogen sensor, the hydrogen concentration values ​​of the cathode and anode sides are subjected to difference processing, and the result obtained by the difference processing is subjected to ratio processing with the solubility of hydrogen in the membrane to obtain the local partial pressure gradient on both sides of the membrane. ;

[0106] Preferably, by the formula: Obtaining component-level hydrogen permeation flux ;

[0107] in, is the hydrogen permeation flux, is the diffusion coefficient and solubility of hydrogen in the membrane, obtained from basic experiments on membrane electrode materials, R is the thermodynamic constant, and T is the absolute temperature;

[0108] is the membrane electrode thickness, is the hydrogen partial pressure gradient on both sides of the membrane, is the activation energy of hydrogen, is the pressure correction function, which is used to correct the nonlinear effect under high pressure;

[0109] S202. Establish a system-scale physical mechanism model to obtain the system-level hydrogen permeation flux;

[0110] Preferably, by the formula: Obtaining system-level hydrogen permeation flux ;

[0111] in, They are the anode and cathode flow channel pressure difference and design pressure difference of a single battery pack, is the anode and cathode gas flow rate, β is the flow field influence coefficient;

[0112] is the component-level reference flux, which is the flux of all component-level sensors in the same electrolyzer. The average is obtained to represent the typical permeability characteristics of a single component;

[0113] The flow field influence coefficient is obtained through flow field calibration experiments: that is, technicians in this field change the flow channel blockage and flow distribution, measure the deviation of the system-level and component-level permeation, and fit the amplification coefficient of the flow field non-uniformity on the permeation. ;

[0114] Anode and cathode flow channel pressure difference of a single battery pack The design pressure difference is calculated by the difference between the cathode inlet pressure and the anode outlet pressure of a single cell group in the PEM electrolyzer. Provided by electrolyzer design documentation;

[0115] The anode and cathode gas flows are measured by gas mass flow meters;

[0116] S203, calculating a risk index based on the component-level hydrogen permeation flux and the system-level hydrogen permeation flux, and determining a risk level of the pressure difference imbalance area based on the risk index;

[0117] It can be understood that by comparing the component-level hydrogen permeation flux with the preset membrane electrode safety threshold, and by comparing the system-level hydrogen permeation flux with the preset electrolytic stack safety threshold, the excess multiples of the component-level hydrogen permeation flux and the membrane electrode safety threshold, and the excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold are calculated respectively;

[0118] Based on the spatial correlation between the pressure imbalance areas within the component layer and the pressure imbalance areas within the system layer, a risk index is constructed and divided into four different risk index ranges.

[0119] It is understood that the four different risk index ranges include:

[0120] Low-risk zone: There is no overlap between the imbalance areas at the component and system levels, the sum of the excess multiples is less than 1, there is no cross-regional transmission effect, and both local and global risks are minor.

[0121] Medium-low risk range: There is some overlap between the two, with a hierarchical imbalance factor of 0.1-0.3 and a sum of excess multiples of 1-2. There is no obvious transmission, but local risks tend to spread to the system;

[0122] Medium-to-high risk range: The overlap is high, the layer imbalance factor is 0.3-0.6, the sum of the exceedance multiples is 2-3, there is a conduction effect between adjacent battery packs, and local risks are related to the global situation;

[0123] High-risk area: High overlap, layer imbalance factor > 0.6, sum of excess multiples > 3, component and system layer imbalance areas converge and transmit significantly, and the overall risk is severe;

[0124] Among them, the risk index is constructed by combining the spatial correlation between component-level high-risk areas and system-level abnormal areas as follows:

[0125] Obtain the pressure difference imbalance area within the component layer and the pressure difference imbalance area within the system layer corresponding to the high-risk interval, and obtain the pressure difference imbalance area within the high-risk component layer and the pressure difference imbalance area within the high-risk system layer;

[0126] If the pressure imbalance area within the high-risk component layer or the pressure imbalance area within the high-risk system layer has a conduction effect in the electrolyzer;

[0127] That is, if the boundary distance between the pressure difference imbalance areas within the two component layers is within the preset range value, or if there is a pressure difference imbalance area between adjacent battery packs of the electrolyzer, the pressure difference imbalance areas within the component layer or the pressure difference imbalance areas within the system layer are merged and processed separately to obtain the total imbalance area within the component layer and the total imbalance area within the system layer;

[0128] Calculate the area overlap between the total imbalance area within the component layer and the total imbalance area within the system layer to obtain the hierarchical imbalance factor;

[0129] The excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold are summed, and the summed result is multiplied by the hierarchical imbalance factor to obtain the permeation risk index;

[0130] Based on the risk level of the pressure imbalance area, a coordinated assessment of local micro-risk and global system risk is achieved;

[0131] S204: If the risk level of the pressure difference imbalance area is at a medium or low risk level, a pressure difference-permeability dynamic risk model is constructed to output a predicted permeation risk level;

[0132] It will be understood by those skilled in the art that, after a pressure imbalance area with a medium or low risk level is identified, a dynamic correlation model based on time series is constructed by collecting the dynamic pressure difference data and the corresponding hydrogen permeation flux at the component level and system level in the area in real time, combined with the pressure difference-permeability baseline database under historical normal operating conditions;

[0133] The dynamic correlation model of the time series uses the current pressure difference change rate and the growth rate of hydrogen permeation flux as input variables. By calculating the dynamic coupling coefficient used to reflect the immediate impact of the pressure difference change on the permeability, and introducing environmental correction factors such as temperature and humidity, the predicted permeation risk index at different time points is obtained.

[0134] Based on the preset threshold range of the index, for example, based on the critical value of upgrading from historical medium-low risk to medium-high risk, the corresponding predicted penetration risk level is output.

[0135] Example 2

[0136] like Figure 1 As shown, a control method for a PEM water electrolysis hydrogen production device according to an embodiment of the present invention includes the following steps:

[0137] Step 3: Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective suppression features are extracted for the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective suppression features to determine whether hydrogen permeation is completely and effectively suppressed.

[0138] Among them, based on the predicted infiltration risk level, the differentiated intervention strategy is implemented according to the type of pressure imbalance area:

[0139] If the predicted infiltration risk level is high, different types of differentiated intervention control strategies are formulated based on the type of pressure imbalance area;

[0140] Among them, the ways to formulate different types of differentiated intervention control strategies are:

[0141] S301: If the imbalance is within the component, formulate a local intervention strategy to adjust the local electric field gradient of the cathode of the unbalanced single cell;

[0142] Preferably, the upper boundary value of the penetration risk index interval corresponding to the low risk level is obtained;

[0143] By calculating the deviation ratio between the predicted penetration risk and the upper boundary value, the adjustment boundary coefficient of the component internal imbalance is obtained;

[0144] Obtain the electric field control gain of the PEM electrolyzer membrane electrode, multiply the electric field control gain by the adjustment boundary coefficient of the internal imbalance of the component to obtain the intervention adjustment amount of the internal imbalance of the component;

[0145] Obtaining the dominant direction of hydrogen permeation in the pressure difference imbalance region, and adjusting the cathode electric field gradient in the opposite direction of the dominant direction of hydrogen permeation in combination with the intervention adjustment amount;

[0146] S302: If there is an internal imbalance in the system, a regional coordinated intervention strategy is formulated to adjust the local electric field gradient of the cathode of the unbalanced single cell;

[0147] If the imbalance is internal to the system and the voltage difference imbalance area is not limited to a single battery pack, the cathode electric field distribution of the single battery group is correlated and the voltage distribution between adjacent batteries is optimized by building a multivariate prediction model;

[0148] Among them, the method of effectively suppressing the feature extraction of the pressure difference imbalance area after intervention is:

[0149] Obtain the absolute change in the area of ​​all pressure difference imbalance regions before and after the intervention control, and perform ratio processing on the absolute change and the area of ​​the pressure difference imbalance region before the intervention control to obtain the imbalance contraction ratio;

[0150] Obtain the hydrogen permeation flux before and after the intervention control in the pressure imbalance area, and construct the permeation concentration field of each pressure imbalance area before and after the intervention;

[0151] Among them, when constructing the permeation concentration field of the pressure imbalance area before and after the intervention, the hydrogen permeation flux of the corresponding component-level or system-level sensor in the area is first extracted based on the positioning results of the imbalance area. According to the spatial geometry of the imbalance area (dividing the calculation grid, using the monitoring data as the boundary conditions, and solving the equations using numerical methods such as finite element method), the spatial distribution of hydrogen concentration in the area before the intervention is obtained; for the same area after the implementation of the intervention measures, the parameters are repeatedly collected and substituted into the model for solution. Finally, the permeation concentration field before and after the intervention is compared to quantify the regulatory effect of the intervention on the concentration distribution;

[0152] Calculate the hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure imbalance area before and after intervention;

[0153] The deviation ratio of hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure imbalance area before and after intervention was calculated to obtain the permeation inhibition rate;

[0154] The unbalanced shrinkage ratio and permeation inhibition rate were used as effective inhibition characteristics;

[0155] Among them, the method of constructing a two-dimensional evaluation domain based on the effective suppression characteristics and determining whether hydrogen permeation has achieved complete and effective suppression is as follows:

[0156] Obtain the imbalance shrinkage ratio and permeation inhibition rate of all pressure imbalance areas and construct an inhibition efficiency matrix;

[0157] Based on the inhibition efficiency matrix, a two-dimensional inhibition evaluation domain is constructed with the imbalance shrinkage ratio as the horizontal axis and the penetration inhibition rate as the vertical axis. The imbalance shrinkage ratio and penetration inhibition rate of the inhibition efficiency matrix are used as evaluation coordinates.

[0158] Set the suppression evaluation boundary line, and map the evaluation coordinates to the two-dimensional suppression evaluation domain based on the suppression evaluation boundary line, such as Figure 2 As shown in FIG, if the evaluation coordinates of the pressure difference imbalance region are in the third quadrant of the two-dimensional inhibition evaluation domain, it is determined that hydrogen permeation has not been completely and effectively suppressed. Otherwise, there is a possibility that hydrogen permeation is completely and effectively suppressed.

[0159] It should be explained that in the two-dimensional evaluation domain, the imbalance contraction ratio on the horizontal axis reflects the degree of contraction of the pressure imbalance area, with positive values ​​indicating contraction and negative values ​​indicating expansion. The permeation inhibition rate on the vertical axis reflects the reduction in the hydrogen permeation gradient (positive values ​​indicate inhibition and negative values ​​indicate aggravation).

[0160] The third quadrant corresponds to situations where the imbalance contraction ratio is negative (area expansion) and the permeation suppression rate is negative (permeation intensifies), directly identifying scenarios where intervention measures have failed to effectively contain the risk and may even cause it to spread. This can reduce the misinterpretation of local improvement but overall deterioration as effective suppression, ensure timely identification of pressure imbalance areas where suppression has failed, and provide clear targeting for subsequent secondary prevention and control measures such as electrolyte flow grading adjustment and hydrogen reverse migration path tracking, thereby improving the accuracy of control strategies and equipment operation safety.

[0161] Those skilled in the art will understand that when setting the inhibition evaluation boundary line, the baseline distribution of statistical indicators is calculated based on the historical data of the normal and stable operating conditions of the electrolytic cell; then, combined with the operating condition correlation, the upper limit of the allowable fluctuation of the pressure difference and the critical value of the concentration gradient under different currents are fitted to construct the boundary function of the operating conditions; through verification experiments simulating the success or failure of inhibition, the boundary parameters are iteratively corrected to form an inhibition evaluation boundary line that can dynamically adapt to the operating conditions and quantitatively judge whether the inhibition effect meets the standards. For example, the pressure difference fluctuation needs to be lower than the amplitude within the boundary line, and the concentration exceeding the limit area needs to be shrunk to the safe range surrounded by the boundary line.

[0162] Step 4: If complete and effective suppression is not achieved, the permeation suppression rate is fed back in real time and a negative correlation adjustment model is constructed. The electrolyte flow rate is graded and adjusted. The energy of the differential pressure oscillation signal after graded adjustment and the permeation suppression rate are extracted to determine the secondary fluctuation state. If a secondary fluctuation state exists, the starting point of hydrogen migration is located through the spatiotemporal distribution, the hydrogen reverse migration path is extracted, and the hydrogen reverse migration path is screened to obtain a response isolation path.

[0163] Among them, if complete and effective inhibition is not achieved, the method of real-time feedback of the penetration inhibition rate and construction of a negative correlation adjustment model is as follows:

[0164] Obtain the pressure difference imbalance area that has not achieved complete effective suppression, as well as the corresponding imbalance shrinkage ratio and permeation suppression rate;

[0165] A negatively correlated adjustment amplitude model is constructed based on the imbalance shrinkage ratio and permeation suppression rate to adjust the electrolyte flow rate in the pressure imbalance area of ​​the PEM electrolyzer where full suppression is not achieved.

[0166] Among them, the adjustment logic of the adjustment amplitude model is: preferentially adjust the electrolyte flow in the imbalance area through the single-cell local flow control device, such as a micro electromagnetic flow valve, deployed at the inlet of the single-cell flow channel in the component layer.

[0167] The adjustment range is negatively correlated with the permeation inhibition rate: if the permeation inhibition rate is less than 30%, it indicates poor inhibition effect, and the flow rate adjustment range is ±10% to 15% of the base flow rate; if the permeation inhibition rate is 30% or less and less than 60%, it indicates moderate inhibition effect, and the adjustment range is ±5% to 10%;

[0168] The adjustment direction is based on differential pressure fluctuations: If the component-level differential pressure oscillation signal shows that the hydrogen side pressure is continuously low, the hydrogen permeation drive is enhanced and the cathode side electrolyte flow is increased to increase the cathode pressure and reduce the differential pressure;

[0169] If the hydrogen concentration on the anode side increases abnormally, that is, the reverse permeation of hydrogen intensifies, the electrolyte flow on the anode side is increased to flush the permeated hydrogen and reduce the local hydrogen concentration;

[0170] Among them, the method of determining the secondary fluctuation state is:

[0171] Obtaining the pressure difference oscillation signal energy and permeation suppression rate in the pressure difference imbalance area after adjusting the electrolyte flow rate;

[0172] Based on the pressure difference oscillation signal energy and the permeation suppression rate, a comparative analysis is performed to determine whether there is a secondary fluctuation suppression state;

[0173] It can be understood that the energy of the pressure difference oscillation signal before and after the intervention in the pressure difference imbalance area is first extracted through time-frequency analysis, and the permeation inhibition rate is calculated according to the hydrogen permeation flux formula. Then, the energy threshold and inhibition rate target are calibrated based on the stable operating condition data of the electrolyzer, and a correlation criterion is established between the two. For example, when the inhibition rate reaches the target, the energy should decrease synchronously and be lower than the threshold.

[0174] Collect real-time data after intervention and compare the deviation between the current energy and threshold, and the current suppression rate and the target. If there is an abnormal rise in energy exceeding the preset normal threshold and the suppression rate falls back to the target coupling characteristics, it is determined that a secondary fluctuation suppression state exists, that is, the initial suppression fails or the fluctuation re-ignites;

[0175] Among them, there is a way to extract the reverse migration path of hydrogen that suppresses the pressure difference balance:

[0176] Preferably, the method of extracting the hydrogen reverse migration path and screening the hydrogen reverse migration path is:

[0177] S401. Locating the starting point of hydrogen migration based on the spatiotemporal distribution of the component-level pressure difference oscillation signal;

[0178] It is understandable that the component-level high-precision micro-differential pressure sensor will capture local high-frequency pressure fluctuations on the positive and negative sides of the membrane electrode, the plate flow channel outlet, etc. Under normal operating conditions, these fluctuations are within a stable range; when the membrane electrode has local damage, flow channel blockage or seal failure and other abnormalities, the local anode and cathode pressure difference will suddenly increase or decrease sharply in a short period of time, and its change rate and amplitude will be higher than the preset normal fluctuation threshold. The point above the preset normal fluctuation threshold is the pressure difference sudden change point. Since hydrogen migration is often directly related to the local pressure difference, the sudden change point is regarded as the starting point of hydrogen migration, providing initial positioning for tracking the reverse migration path of hydrogen;

[0179] The pressure difference sudden change point of the membrane electrode is obtained based on the spatiotemporal distribution of the module-level pressure difference oscillation signal, which is used as the starting point of hydrogen migration;

[0180] S402, combining the concentration gradient of the micro hydrogen sensor, tracing the microscopic path of hydrogen from the cathode to the anode as the reverse migration path of hydrogen;

[0181] Among them, the reverse migration path of the unbalanced hydrogen in the system is as follows:

[0182] S411, analyzing the pressure difference gradient direction of adjacent battery groups to determine the pressure conduction path across the battery groups;

[0183] Preferably, the inlet and outlet pressure differences of adjacent single battery groups are collected, the gradient change of the pressure difference in the spatial dimension is calculated, and the pressure conduction direction and path of the battery group are fitted according to the law of pressure conduction from high gradient area to low gradient area;

[0184] S412. Locate the macro leakage point based on the hydrogen concentration distribution;

[0185] Preferably, the hydrogen permeation concentration field is called to extract the spatial distribution of hydrogen concentration to locate the hydrogen concentration gradient mutation and the concentration exceeding the limit area, and mark them as macro leakage points;

[0186] S413: Comprehensively analyze the pressure difference conduction and concentration distribution to output the reverse migration path of hydrogen;

[0187] Preferably, the pressure conduction path is associated with the macro leakage point, the leakage source is traced back against the pressure gradient, the pressure difference continuity and concentration anomaly are integrated, and the complete path of hydrogen reverse migration is output;

[0188] The peak hydrogen permeation concentration and the pressure gradient of each imbalanced hydrogen reverse migration path are extracted and dimensionlessly processed. The dimensionlessly processed peak hydrogen permeation concentration and pressure gradient are summed to obtain the path characteristic coefficient.

[0189] Obtaining a pressure difference imbalance region where a secondary fluctuation suppression state exists as a pressure difference selection region;

[0190] Obtain the number of associated pressure difference selection zones in the hydrogen reverse migration path of each imbalance type, calculate the ratio of the number of pressure difference selection zones to the total pressure difference imbalance zones, and obtain the regional correlation degree;

[0191] The path characteristic coefficient and the regional correlation degree are multiplied to obtain the path selection coefficient;

[0192] Obtaining the path selection coefficients of the hydrogen reverse migration paths of the component-level imbalance type and the system-level imbalance type;

[0193] The hydrogen reverse migration pathway with the highest pathway selectivity coefficient was selected as the response isolation pathway.

[0194] It is understood that the purpose of extracting the response isolation path is to:

[0195] Objective 1: Analyze the coupled propagation channels of pressure differential oscillation and hydrogen permeation, locate the root causes of abnormal conduction across the component-system level, clarify the propagation paths of pressure waves and hydrogen permeation in micro- and macro-structures such as membrane electrodes and flow channels, and identify the starting area of ​​imbalance;

[0196] Objective 2: Based on the diffusion coefficient, capture rate, and release rate kinetics within the pathway, derive targeted strategies for fluctuation suppression and leakage blocking. For example, by manipulating the flow field to block pressure conduction or utilizing trap properties to enhance hydrogen capture, intervention plans can be designed at the pathway mechanism level.

[0197] Objective 3: To quantify the path permeation persistence coefficient as a mass transfer response characteristic, predict the risk of abnormal re-ignition after the implementation of suppression measures, predict the possibility of hydrogen permeation rebound or pressure difference oscillation recurrence through path parameters, and assist in evaluating the long-term effectiveness of the suppression effect.

[0198] Step 5: Extract the free state and trap state characteristics of the response isolation path, combine the trap effect to perform permeation persistence analysis to obtain the permeation persistence coefficient, and determine whether to trigger safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient;

[0199] The method for extracting the free state and trap state characteristics of the response isolation path and combining the trap effect to perform penetration persistence analysis is as follows:

[0200] The diffusion coefficient D of hydrogen in the response isolation path is extracted as the free state feature;

[0201] Acquire hydrogen in the response isolation path at the capture rate and release rate As a trap state feature;

[0202] The osmotic persistence equation: Obtain the permeability persistence coefficient Kc;

[0203] in, is the trap density;

[0204] It can be understood that the diffusion coefficient D is extracted by constructing a simulation test cell equivalent to the response isolation path of the electrolyzer, using a fluorescent tracer method to label hydrogen atoms, monitoring the diffusion of the concentration field over time, and fitting the time-space curve of the concentration distribution in combination with Fick's second law;

[0205] Construct a simulation test system physically equivalent to the response isolation path of the electrolyzer; transiently introduce hydrogen into the system, monitor the rising process of hydrogen concentration at the trap site through the micro-area concentration probe, fit the pseudo-first-order adsorption kinetic equation, and inversely deduce the adsorption rate constant capture rate ;

[0206] After stopping the hydrogen supply, the decay curve of the hydrogen concentration in the trap was monitored, the desorption kinetic model was fitted, and the release rate spectrum of hydrogen at different temperatures was analyzed to obtain the release rate. ;

[0207] Among them, the method for determining whether to trigger safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient is:

[0208] Preferably, the critical threshold of the permeation persistence coefficient is calibrated through multi-physics field simulation or accelerated testing based on safety standards such as the hydrogen embrittlement threshold, hydrogen explosion lower limit, and equipment pressure resistance design of key components such as membrane electrodes and seals of hydrogen production equipment, combined with historical failure data;

[0209] Real-time acquisition of response isolation path data to obtain the permeation persistence coefficient Kc: the free hydrogen diffusion coefficient D is extracted using a micro hydrogen sensor and diffusion model, and the trap capture rate is obtained using electrochemical testing. and release rate , the trap density was counted by micro-CT ;

[0210] Among them, the method for determining whether to trigger safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient is:

[0211] The larger the permeability persistence coefficient Kc value, the higher the proportion of free hydrogen, and the greater the risk of continuous permeation; when the permeability persistence coefficient exceeds the critical threshold of the permeability persistence coefficient, the third-level response is immediately triggered;

[0212] Among them, the three-level response includes: PLC controls the gas valve to cut off the cathode / anode circuit for rapid isolation, starts nitrogen purge at 3 times the volume flow rate to complete inerting replacement (monitoring hydrogen concentration <4%vol), links the power module to reduce load and shut down and triggers the sound and light alarm; if the threshold is not reached, the permeation continuity coefficient Kc is dynamically refreshed every 5 seconds, and the trend is predicted in combination with the pressure difference and concentration data, and strategies such as strengthening electrolyte flow control and cathode electric field intervention are suppressed to suppress risk escalation.

[0213] Example 3

[0214] See also Figure 3 As shown, the control system of a PEM water electrolysis hydrogen production device according to an embodiment of the present invention includes the following modules:

[0215] Imbalance Analysis Module: This module is used to collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area.

[0216] Permeation analysis module: used to obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform correlation risk modeling of pressure difference and permeability, and output the predicted permeation risk level;

[0217] Inhibition determination module: Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective inhibition features are extracted from the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective inhibition features to determine whether hydrogen permeation has been completely and effectively suppressed.

[0218] Path extraction module: If complete and effective suppression is not achieved, it is used to provide real-time feedback on the permeation suppression rate and build a negative correlation adjustment model. The electrolyte flow rate is adjusted in stages. The energy of the pressure difference oscillation signal after the staged adjustment and the permeation suppression rate are extracted to determine the secondary fluctuation state. If the secondary fluctuation state exists, the starting point of hydrogen migration is located through the spatiotemporal distribution, and the hydrogen reverse migration path is extracted. The hydrogen reverse migration path is screened and processed to obtain the response isolation path.

[0219] Response control module: used to extract the free state and trap state characteristics of the response isolation path, combine the trap effect to perform permeation persistence analysis to obtain the permeation persistence coefficient, and determine whether to trigger the safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient.

[0220] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for a PEM water electrolysis hydrogen production device, characterized in that: The following steps are involved: Collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area; Obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform risk modeling on the correlation between pressure difference and permeability, and output the predicted permeation risk level; Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective suppression features are extracted from the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective suppression features to determine whether hydrogen permeation has been completely and effectively suppressed. If complete and effective suppression is not achieved, the electrolyte flow rate is adjusted in stages, and the pressure difference oscillation signal energy and permeation suppression rate after the staged adjustment are extracted to determine the secondary fluctuation state. If it exists, the hydrogen reverse migration path is extracted, and the hydrogen reverse migration path is screened and processed to obtain the response isolation path.

2. The control method of a PEM water electrolysis hydrogen production equipment according to claim 1, characterized in that: The regional analysis is performed as follows: Obtain component-level and system-level differential pressure oscillation signals, perform spatiotemporal correlation analysis, and obtain spatiotemporal correlation coefficients. Extract the pressure difference oscillation signal energy corresponding to the component-level pressure difference oscillation signal and the system-level pressure difference oscillation signal, and construct the imbalance area positioning function based on the spatiotemporal correlation coefficient; The pressure difference imbalance area where the pressure drops suddenly on the hydrogen side of the PEM electrolyzer is determined based on the imbalance area positioning function.

3. The control method of a PEM water electrolysis hydrogen production equipment according to claim 2, characterized in that: The method of obtaining the component-level pressure difference oscillation signal and the system-level pressure difference oscillation signal is as follows: Construct a layered differential pressure monitoring network in the PEM electrolyzer system, dividing the differential pressure data into component-level differential pressure data and system-level differential pressure data; The pressure difference data at the component level and the pressure difference data at the system level are subjected to data preprocessing and dynamic feature extraction respectively to obtain the pressure difference oscillation signal at the component level and the pressure difference oscillation signal at the system level.

4. The control method of a PEM water electrolysis hydrogen production equipment according to claim 1, characterized in that: The method for modeling the associated risk of pressure difference and permeability is as follows: Establish a physical mechanism model at the membrane electrode scale to obtain the hydrogen permeation flux at the component level; Establish a system-scale physical mechanism model to obtain the system-level hydrogen permeation flux; Calculating a risk index based on component-level and system-level hydrogen permeation fluxes, and determining the risk level of the pressure difference imbalance area based on the risk index; If the risk level of the pressure difference imbalance area is at a medium or low risk level, a dynamic risk model of pressure difference-permeability is constructed to output the predicted permeability risk level.

5. The control method of a PEM water electrolysis hydrogen production equipment according to claim 4, characterized in that: The risk index is calculated as follows: Get the level imbalance factor; Obtaining the excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold and the excess multiples of the system-level hydrogen permeation flux and the electrolytic stack safety threshold and performing summing processing; The summed result is multiplied by the hierarchical imbalance factor to obtain the penetration risk index.

6. The control method of a PEM water electrolysis hydrogen production equipment according to claim 5, characterized in that: The method for obtaining the said level imbalance factor is: Obtain the pressure difference imbalance area within the component layer and the pressure difference imbalance area within the system layer corresponding to the high-risk interval, and obtain the pressure difference imbalance area within the high-risk component layer and the pressure difference imbalance area within the high-risk system layer; If the pressure imbalance area within the high-risk component layer or the pressure imbalance area within the high-risk system layer has a conduction effect in the electrolyzer, or if there is a pressure imbalance area between adjacent battery packs in the electrolyzer, then the pressure imbalance area within the component layer or the pressure imbalance area within the system layer shall be merged to obtain the total imbalance area within the component layer and the total imbalance area within the system layer; The area overlap between the total imbalance area within the component layer and the total imbalance area within the system layer is calculated to obtain the hierarchical imbalance factor.

7. The control method of a PEM water electrolysis hydrogen production equipment according to claim 1, characterized in that: The method for determining whether the hydrogen permeation is completely and effectively suppressed is: Obtain the imbalance shrinkage ratio and permeation inhibition rate of all pressure imbalance areas and construct an inhibition efficiency matrix; Based on the inhibition efficiency matrix, a two-dimensional inhibition evaluation domain is constructed with the imbalance shrinkage ratio as the horizontal axis and the penetration inhibition rate as the vertical axis. The imbalance shrinkage ratio and penetration inhibition rate of the inhibition efficiency matrix are used as evaluation coordinates. An inhibition evaluation boundary line is set, and the evaluation coordinates are mapped to a two-dimensional inhibition evaluation domain based on the inhibition evaluation boundary line. If the evaluation coordinates of the pressure imbalance area are in the third quadrant of the two-dimensional inhibition evaluation domain, it is considered that hydrogen permeation has not been completely and effectively suppressed.

8. The control method of a PEM water electrolysis hydrogen production equipment according to claim 7, characterized in that: The unbalanced shrinkage ratio and permeation inhibition rate are obtained as follows: If the predicted infiltration risk level is high, different types of differentiated intervention control strategies are formulated based on the type of pressure imbalance area; Obtain the absolute change in the area of ​​all pressure difference imbalance regions before and after the intervention control, and perform ratio processing on the absolute change and the area of ​​the pressure difference imbalance region before the intervention control to obtain the imbalance contraction ratio; Obtain the hydrogen permeation flux before and after the intervention control in the pressure imbalance area, and construct the permeation concentration field of each pressure imbalance area before and after the intervention; Calculate the hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure imbalance area before and after intervention; The deviation ratio of the hydrogen permeation gradient in the dominant direction of hydrogen permeation in the pressure difference imbalance area before and after intervention was calculated to obtain the permeation inhibition rate.

9. The control method of a PEM water electrolysis hydrogen production equipment according to claim 1, characterized in that: The response isolation path is obtained as follows: The peak hydrogen permeation concentration and the pressure gradient of each imbalanced hydrogen reverse migration path are extracted and dimensionlessly processed. The dimensionlessly processed peak hydrogen permeation concentration and pressure gradient are summed to obtain the path characteristic coefficient. Obtaining a pressure difference imbalance region where a secondary fluctuation suppression state exists as a pressure difference selection region; Obtain the number of associated pressure difference selection zones in the hydrogen reverse migration path of each imbalance type, calculate the ratio of the number of pressure difference selection zones to the total pressure difference imbalance zones, and obtain the regional correlation degree; The path characteristic coefficient and the regional correlation degree are calculated to obtain the path selection coefficient; Obtaining the path selection coefficients of the hydrogen reverse migration paths of the component-level imbalance type and the system-level imbalance type; The hydrogen reverse migration pathway with the highest pathway selectivity coefficient was selected as the response isolation pathway.

10. A control system for a PEM water electrolysis hydrogen production device, used to implement the control method for a PEM water electrolysis hydrogen production device according to any one of claims 1 to 9, characterized in that: Includes the following modules: Imbalance Analysis Module: This module is used to collect the cathode and anode pressure differential data of the PEM electrolyzer, establish a layered pressure differential monitoring network for the cathode and anode pressure differential data, extract the pressure differential oscillation signal generated by the layered pressure differential monitoring network, and perform regional analysis on the pressure drop on the hydrogen side of the PEM electrolyzer based on the pressure differential oscillation signal to obtain the pressure differential imbalance area. Permeation analysis module: used to obtain the hydrogen permeation concentration of the PEM electrolyzer membrane electrode in the pressure imbalance area, perform correlation risk modeling of pressure difference and permeability, and output the predicted permeation risk level; Inhibition determination module: Based on the predicted permeation risk level, differentiated intervention strategies are implemented according to the type of pressure imbalance area. Effective inhibition features are extracted from the pressure imbalance area after intervention. A two-dimensional evaluation domain is constructed based on the effective inhibition features to determine whether hydrogen permeation has been completely and effectively suppressed. Path extraction module: If complete and effective suppression is not achieved, the electrolyte flow rate is adjusted in stages. The energy of the pressure difference oscillation signal and the permeation suppression rate after the staged adjustment are extracted to determine the secondary fluctuation state. If the secondary fluctuation state exists, the hydrogen reverse migration path is extracted and screened to obtain the corresponding isolation path. Response control module: used to extract the free state and trap state characteristics of the response isolation path, combine the trap effect to perform permeation persistence analysis to obtain the permeation persistence coefficient, and determine whether to trigger the safety isolation and inerting response control of the hydrogen production equipment based on the permeation persistence coefficient.

Images