Running state monitoring and scheduling optimization method and system for composite membrane electrolytic cell

Abstract

The invention discloses an operation state monitoring and scheduling optimization method and system for a composite membrane electrolytic cell, and particularly relates to the technical field of water electrolysis hydrogen production. According to the method, the health state and the corresponding life calibration coefficient are determined through piecewise linear mapping on the basis of the comprehensive health index, and the degradation rate is calculated in combination with the time sequence health index so as to adjust the life optimization coefficient; meanwhile, the ratio of the mean value of historical operation power to rated power and the nonlinear acceleration effect of quantized power on aging are introduced, finally, the obtained pre-estimated reduction coefficient is substituted into a formula to calculate the residual life, the real-time state and power influence of equipment are fully fused, and the deviation between a prediction result and the actual life is remarkably reduced.

Description

Technical Field

[0001] This invention relates to the field of water electrolysis for hydrogen production technology, and more specifically, to a method and system for monitoring and optimizing the operation status of a composite membrane electrolyzer. Background Technology

[0002] In the field of hydrogen production technology through water electrolysis, composite membrane electrolyzers are the core production equipment. Their operational stability, energy efficiency, and service life directly determine the economy and safety of the hydrogen production system. With the large-scale development of the hydrogen energy industry, composite membrane electrolyzers are gradually being upgraded towards high power and long-cycle operation.

[0003] However, existing operation monitoring and scheduling technologies still have the following shortcomings in practical applications:

[0004] Traditional composite membrane electrolyzer status monitoring relies on a single data source, which cannot fully reflect the health status of the equipment under the coupling of multiple physical fields. This leads to a significant deviation between health assessment and actual equipment status, creating potential safety hazards for subsequent operation and scheduling.

[0005] In terms of lifespan prediction, the dynamic impact of power and equipment status on the aging rate of equipment during actual operation was not considered. On the one hand, the equipment's service life was only used as a basic parameter and was not calibrated in conjunction with real-time health status, resulting in a large deviation between the prediction results and the actual lifespan. On the other hand, the accelerating effect of historical operating power on aging was not quantified and was only corrected by a fixed coefficient, which could not reflect the nonlinear relationship between "power-aging".

[0006] The lack of dynamically adaptable safety operating boundaries means that existing solutions typically set fixed and conservative safety power limits without considering the dynamic changes in the health status of the equipment throughout its entire life cycle. When the equipment is in the early or middle stages of health, the fixed boundaries will limit its production potential and fail to fully leverage the advantages of efficient operation. When the equipment enters the health degradation stage, the original fixed boundaries are difficult to match the actual safety requirements, which may lead to safety accidents such as thermal runaway and membrane damage. This not only fails to guarantee production safety but also reduces the flexibility and economy of equipment operation.

[0007] To address this, a method and system for monitoring and optimizing the operation status of composite membrane electrolyzers are proposed. Summary of the Invention

[0008] To overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and system for monitoring and optimizing the operation status of a composite membrane electrolyzer.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A method for monitoring and optimizing the operation status of a composite membrane electrolyzer includes:

[0011] S1: Construct a multi-source feature dataset that includes electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data;

[0012] S2: Employ deep learning algorithms to fuse multi-source feature datasets and output a comprehensive health index. Using the comprehensive health index and historical operating power as input, combined with the equipment's service life, output the equipment's estimated remaining lifespan.

[0013] S3: Combining the output comprehensive health index and estimated remaining lifespan, dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer under the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window;

[0014] S4: The objective function is to maximize the total profit within the scheduling cycle. The objective function explicitly includes production revenue, energy consumption cost and aging cost. The dynamic operating window is used as the core constraint. Rolling optimization is performed to obtain the optimal power command.

[0015] Specifically, in step S2, a comprehensive health index is output by fusing multi-source feature datasets.

[0016] Electrochemical impedance spectroscopy data were analyzed to extract charge transfer resistance and membrane resistance;

[0017] The infrared thermal imaging data is analyzed to extract the highest temperature in the film region, the temperature difference between the anode and cathode, and the proportion of hot spot area.

[0018] The acoustic emission data is analyzed to extract the peak amplitude of the signal and the proportion of high-frequency events;

[0019] Determine the normal operating values ​​corresponding to each type of data source based on historical normal data;

[0020] The feature parameters extracted from each type of data source are combined with the corresponding normal operating values ​​and weighted to obtain the electrical health index, thermal health index, and mechanical health index.

[0021] After normalizing the electrical health index, thermal health index, and mechanical health index, a weighted fusion process is used to obtain the comprehensive health index of the equipment.

[0022] Specifically, step S2 uses the comprehensive health index and historical operating power as inputs;

[0023] A piecewise linear mapping rule is set between the comprehensive health index and health status to convert the comprehensive health index into health status; where health status includes normal status, moderate deterioration status and severe deterioration status, and each health status corresponds to a lifespan calibration coefficient;

[0024] Extract the comprehensive health index from the x periods prior to the current time point, and denot it as the time-series health index;

[0025] The time-series health index and the comprehensive health index at the current time point are respectively labeled as... According to the formula Calculate the degradation rate ;

[0026] Degradation rate Multiplying this by the preset rate sensitivity coefficient yields the trend adjustment coefficient;

[0027] The sum of the trend adjustment coefficient and the life calibration coefficient is taken as the life optimization coefficient of the equipment.

[0028] Extract the historical operating power of the x cycles before the current time point, and take the average value as the average operating power;

[0029] The ratio is calculated using the average operating power as the numerator and the rated power as the denominator, and is denoted as . ;

[0030] pass Calculate the power correction factor ;in This is the preset power aging factor.

[0031] Specifically, in step S2, the estimated remaining lifespan of the equipment is output based on the equipment's existing lifespan.

[0032] Multiply the lifespan optimization factor by the power aging factor to obtain the estimated reduction factor;

[0033] The estimated remaining lifespan of the equipment is calculated using the formula: Estimated Remaining Lifespan = Theoretical Lifespan of Equipment - Actual Lifespan × Estimated Reduction Factor.

[0034] Specifically, in step S3, the absolute safe power limit of the electrolyzer under the current state is dynamically calculated;

[0035] The remaining service index of the equipment is obtained by calculating the ratio between the estimated remaining life of the equipment as the numerator and the theoretical life of the equipment as the denominator.

[0036] After normalizing the remaining usage index and comprehensive health index of the equipment, the equipment status assessment index is obtained by weighted calculation logic.

[0037] Using the rated power as a benchmark, a piecewise linear mapping rule is set between the state assessment index and the safety adjustment coefficient. After converting the state assessment index into the safety adjustment coefficient, it is multiplied by the benchmark to obtain the absolute safe power limit of the electrolytic cell under the current state.

[0038] Specifically, in step S3, the recommended high-efficiency power range of the electrolyzer under the current state is dynamically calculated;

[0039] With an absolute safe power limit as a constraint, the maximum and minimum power that satisfy Q being higher than a preset reference value are identified within the constraint range, and the range between the maximum and minimum power is used as the recommended high-efficiency power range; where Q = product output / energy consumption.

[0040] Specifically, the objective function in step S4 explicitly includes production revenue, energy consumption costs, and aging costs;

[0041] Production revenue, energy consumption costs, and aging costs are respectively labeled as... , , ;

[0042] The variables are defined as follows:

[0043] = ;in Let be the output at time t. = ×v, Let t be the power at time t, and v be the production coefficient. This refers to the unit price of the product.

[0044] ; The time step length, For electricity price;

[0045] ; This is the preset aging cost coefficient; Indicates the lifetime calibration factor;

[0046] in It is limited to the dynamically running window.

[0047] Specifically, S4 uses the dynamic running window as the core constraint and performs rolling optimization to obtain the optimal power command;

[0048] The objective function is expressed as follows: Where T is the scheduling period and t is the time step within the period; For production revenue, For energy consumption costs, Costs related to aging;

[0049] The optimal power obtained from the solution is output as the optimal power command, and then the process proceeds to the next optimization cycle.

[0050] A system for monitoring and optimizing the operation status of a composite membrane electrolyzer, comprising:

[0051] The data input module is used to input electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data, and integrate them into a multi-source feature dataset;

[0052] The health assessment module is used to fuse various features in a multi-source feature dataset using a deep learning algorithm to output a comprehensive health index. Taking the comprehensive health index and historical operating power as input, combined with the equipment's service life, it outputs the estimated remaining lifespan of the equipment.

[0053] The boundary constraint module is used to combine the output comprehensive health index and estimated remaining life to dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer in the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window.

[0054] The constraint adjustment module is used to maximize the total profit within the scheduling cycle as the objective function. The objective function explicitly includes production revenue, energy consumption cost, and aging cost. The dynamic operating window is used as the core constraint condition, and rolling optimization is performed to obtain the optimal power command.

[0055] The technical effects and advantages of this invention are as follows:

[0056] (1) Based on the comprehensive health index, the health status and corresponding life calibration coefficient are determined by piecewise linear mapping, and the degradation rate is calculated by combining the time-series health index to adjust the life optimization coefficient; at the same time, the ratio of the average historical operating power to the rated power is introduced to quantify the nonlinear acceleration effect of power on aging. Finally, the remaining life is calculated by substituting the obtained estimated reduction coefficient into the formula, which fully integrates the influence of real-time equipment status and power, and significantly reduces the deviation between the prediction results and the actual life.

[0057] (2) By constructing a multi-source feature dataset of electrochemical impedance spectroscopy data, infrared thermal imaging data and acoustic emission data, key parameters such as charge transfer resistance, membrane temperature and signal peak amplitude are extracted and analyzed. After weighted integration and normalization, the three-dimensional health index of electrical, thermal and mechanical properties is obtained. Finally, the comprehensive health index is output through deep learning fusion, which can fully cover the multi-dimensional health status of equipment, greatly improve the matching degree between health assessment and actual equipment status, and provide accurate health basis for subsequent scheduling.

[0058] (3) By combining the comprehensive health index and the estimated remaining life, the state assessment index is first obtained by weighting the remaining use index and the comprehensive health index, and then the safety adjustment coefficient is mapped to dynamically determine the absolute safe power limit. At the same time, the power range with energy efficiency higher than the preset value is identified as the recommended high-efficiency power range. The intersection of the two forms a dynamic operation window. The window can be adjusted in real time according to the health status of the equipment. When the equipment is healthy, the power range is expanded to release the production capacity, and when the equipment is degraded, the range is narrowed to ensure safety, taking into account both operational flexibility and safety.

[0059] (4) This invention aims to maximize the total profit of the scheduling cycle, explicitly incorporates production revenue, energy consumption cost and aging cost into the objective function, and uses the dynamic operating window as the core constraint. It solves the optimal power command through rolling optimization, which not only considers short-term energy consumption cost control, but also avoids excessive equipment wear through aging cost quantification, balances short-term revenue and long-term equipment life, solves the hidden loss problem of traditional scheduling, and improves the overall economic benefits of hydrogen production system. Attached Figure Description

[0060] Figure 1 This is a flowchart of a method for monitoring and optimizing the operating status of a composite membrane electrolyzer according to the present invention;

[0061] Figure 2 This is a schematic diagram of a composite membrane electrolyzer operation status monitoring and scheduling optimization system according to the present invention. Detailed Implementation

[0062] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] Example 1

[0064] like Figure 1 As shown, a method for monitoring and optimizing the operation status of a composite membrane electrolyzer includes:

[0065] Dataset Construction: Construct a multi-source feature dataset including electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data;

[0066] An enhanced sensor array, including an impedance analyzer, an infrared thermal imager, and an acoustic sensor, is deployed in the composite membrane electrolyzer system to acquire high-frequency electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data.

[0067] Internal health assessment: Based on the preprocessed multi-source feature dataset, a deep learning algorithm is used to fuse multi-dimensional features to output a comprehensive health index. The comprehensive health index and historical operating power are used as inputs, combined with the equipment's service life, to output the estimated remaining life of the equipment.

[0068] Data preprocessing: noise points are removed from electrochemical impedance spectroscopy data using the least squares method; environmental interference is filtered out from acoustic emission data using wavelet transform; and infrared data is smoothed using mean filtering. Multi-source data synchronization is achieved based on timestamps (accuracy ≤ 1ms).

[0069] Specifically:

[0070] Electrochemical impedance spectroscopy data were analyzed to extract charge transfer resistance and membrane resistance;

[0071] The infrared thermal imaging data is analyzed to extract the highest temperature in the film region, the temperature difference between the anode and cathode, and the proportion of hot spot area.

[0072] The acoustic emission data is analyzed to extract the peak amplitude of the signal and the proportion of high-frequency events;

[0073] The normal operating value corresponding to each type of data source is determined based on historical normal data; historical normal data refers to the stable operating data of the new equipment. For example, the normal operating value of the membrane resistor is 0.7 Ω·cm².

[0074] The feature parameters extracted from each type of data source are combined with the corresponding normal operating values ​​and weighted to obtain the electrical health index, thermal health index, and mechanical health index.

[0075] Additional notes: The feature parameters proposed for each type of data source are labeled as follows:

[0076] The charge transfer resistance and film resistance are respectively labeled as... ;

[0077] The highest temperature in the membrane region, the temperature difference between the anode and cathode, and the percentage of hot spot area are respectively marked as follows: ;

[0078] The peak amplitude of the signal and the proportion of high-frequency events are respectively labeled as ;

[0079] Mark the normal operating values ​​corresponding to each type of data source:

[0080] The normal charge transfer resistance and normal film resistance are respectively labeled as... ;

[0081] The normal maximum temperature of the membrane region, the normal temperature difference between the anode and cathode, and the normal area percentage of the hotspot are respectively marked as follows: ;

[0082] The peak amplitude of the normal signal and the proportion of normal high-frequency events are respectively labeled as follows: ;

[0083] That is, through formula (1) (2) (3) The characteristic parameters and normal operating values ​​of each type of data source are weighted and integrated to obtain the electrical health index. Thermal health index and medical device health index ,in These are the weighting coefficients for charge transfer resistance and film resistance; The weighting coefficients are the highest temperature in the membrane region, the temperature difference between the anode and cathode, and the proportion of hot spot area. The weighting coefficients for peak signal amplitude and the proportion of high-frequency events; Electrical Health Index Thermal health index and medical device health index The higher the value, the worse the equipment's health.

[0084] An increase or rise in charge transfer resistance and membrane resistance indicates a poorer electrochemical health of the device;

[0085] An increase or expansion of the highest temperature in the membrane area, the temperature difference between the anode and cathode, and the proportion of hot spot area indicates a worse thermal health of the equipment.

[0086] The larger the peak amplitude of the signal and the lower the proportion of high-frequency events, the worse the mechanical health of the equipment.

[0087] After normalizing the electrical health index, thermal health index and mechanical health index, a weighted fusion process is used to obtain the comprehensive health index of the equipment.

[0088] To further explain, the comprehensive health index is obtained by multiplying the electrical health index, thermal health index, and mechanical health index by their respective set electrochemical health weights, thermal health weights, and mechanical health weights, and then summing them up.

[0089] For example, by using historical data statistics (such as 60% of equipment failures originating from electrochemical degradation, 25% from thermal runaway, and 15% from mechanical damage), weights can be assigned to the three-dimensional health index:

[0090] Electrochemical health weight: 0.6, membrane and electrode aging directly determine lifespan;

[0091] Thermal health weight: 0.25, thermal stress accelerates material aging;

[0092] Mechanical health weight: 0.15, structural integrity affects equipment stability;

[0093] The comprehensive health index is a weighted sum of three dimensions: (e.g., electrical health index) =0.6, Thermal Health Index =0.5, Medical Device Health Index =0.4, then the comprehensive health index = 0.6×0.6+0.25×0.5+0.15×0.4=0.555.

[0094] A piecewise linear mapping rule is set between the comprehensive health index and health status to convert the comprehensive health index into health status. The health status includes normal status, moderate deterioration status and severe deterioration status, and each health status corresponds to a lifespan calibration coefficient. The lifespan calibration coefficient is limited to the range of 0.03-0.288, and the lifespan calibration coefficient corresponding to the severe deterioration status is the highest, corresponding to 0.288.

[0095] Additional explanation: This involves setting three sets of index intervals corresponding to the comprehensive health index, with each interval corresponding to a health status. Based on the equipment's entire lifecycle data (the relationship between the complete health index and actual lifespan of new equipment → failure), a piecewise linear mapping is established, for example:

[0096] The overall health index is labeled M;

[0097] when This corresponds to a health status;

[0098] when This corresponds to a moderate degradation state;

[0099] when This corresponds to a severely degraded state.

[0100] Extract the comprehensive health index from the x periods prior to the current time point, and denot it as the time-series health index;

[0101] The time-series health index and the comprehensive health index at the current time point are respectively labeled as... According to the formula Calculate the degradation rate ;

[0102] Additional explanation, if If the value is less than 0, the degradation rate decreases over time, indicating an improvement in the equipment's health (usually due to maintenance or operational optimization). It is 0.

[0103] Degradation rate Multiplying this by the preset rate sensitivity coefficient yields the trend adjustment coefficient;

[0104] The function of the trend adjustment factor: to adjust the degradation rate This translates into an adjustment range for the lifetime calibration coefficient, calibrated based on the "correlation between degradation rate and actual lifetime reduction" in historical data;

[0105] Calibration method: Collect a large number of historical samples (each sample contains degradation rate and actual remaining lifetime), and back-calculate the optimal rate sensitivity coefficient by minimizing the prediction error (such as mean square error).

[0106] For example, if historical data shows that when the replacement rate is 0.002 (i.e., the comprehensive health index increases by 0.048 per day), the actual remaining lifespan is 10% shorter than the result predicted by only the lifespan calibration coefficient, then the trend adjustment coefficient needs to be 0.1. Substituting into the formula, we get 1.1 = 0.002 × trend adjustment coefficient, and solving for the trend adjustment coefficient, we get 50.

[0107] The rate of equipment health degradation directly affects its remaining lifespan: if the health index rises rapidly over the last N cycles (e.g., N=24 hours) (accelerated degradation), the lifespan reduction factor needs to be further increased;

[0108] Suppose the monitoring parameters of a composite membrane electrolyzer are as follows:

[0109] Cycle length: 1 hour, N=24 (last 24 hours); Current comprehensive health index = 0.555; 24 hours ago = 0.5; Rate sensitivity coefficient is 50;

[0110] The rate of degradation is calculated to be approximately 0.0023 (meaning an average increase of 0.0023 per hour, indicating a slow deterioration in health).

[0111] Calculate the trend adjustment factor: 50 × 0.0023 = 0.115 (due to accelerated degradation, the life calibration factor is increased by 0.115 to reflect that "the faster the degradation, the greater the life reduction").

[0112] The sum of the trend adjustment coefficient and the life calibration coefficient is taken as the life optimization coefficient of the equipment.

[0113] Extract the historical operating power of the x cycles before the current time point, and take the average value as the average operating power;

[0114] The ratio is calculated using the average operating power as the numerator and the rated power as the denominator, and is denoted as . ;

[0115] pass Calculate the power correction factor ;in The preset power aging factor; calibrated based on accelerated aging test data, such as... =0.8, reflecting the nonlinear effect of power on aging.

[0116] like >1, then >1, further increase the lifetime optimization coefficient, if <1, then =1;

[0117] Example, =1.1, then the power correction factor is approximately 1.078.

[0118] Multiply the lifespan optimization factor by the power aging factor to obtain the estimated reduction factor;

[0119] The estimated remaining lifespan of the equipment is calculated using the formula: Estimated remaining lifespan = Theoretical lifespan of equipment - Used lifespan × Estimated reduction factor.

[0120] The theoretical lifespan of the equipment is the design lifespan at the time of manufacture.

[0121] For example, if the theoretical lifespan of the equipment is 10,000 hours and the actual lifespan is 3,000 hours, the estimated reduction factor is 1.248;

[0122] The estimated remaining lifespan of the equipment is approximately 6256 hours.

[0123] Risk boundary construction: Combining the output comprehensive health index and estimated remaining lifespan, dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer under the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window;

[0124] Specifically:

[0125] The remaining service index of the equipment is obtained by calculating the ratio between the estimated remaining life of the equipment as the numerator and the theoretical life of the equipment as the denominator.

[0126] Example: Current estimated remaining lifespan = 6256 hours, theoretical lifespan of equipment = 10000 hours, then remaining service index = 6256 / 10000 = 0.6256.

[0127] After normalizing the remaining usage index and comprehensive health index of the equipment, the equipment status assessment index is obtained by weighted calculation logic.

[0128] The supplementary explanation is that the status assessment index is obtained by multiplying the remaining usage index and the comprehensive health index by preset usage weights and health weights, respectively, and then summing them up.

[0129] Based on the rated power, a piecewise linear mapping rule is set between the state assessment index and the safety adjustment coefficient. After converting the state assessment index into the safety adjustment coefficient, it is multiplied with the benchmark to obtain the absolute safe power limit of the electrolytic cell under the current state.

[0130] Additional explanation: This involves setting up index intervals corresponding to different state assessment indices, with each interval corresponding to a safety adjustment coefficient. The safety adjustment coefficient is limited to a range of 0.5-1.0, and the lower the state assessment index, the higher the probability of matching 1.0. For example:

[0131] The state assessment index is labeled W;

[0132] when The corresponding safety adjustment factor is 1.0;

[0133] when The corresponding safety adjustment factor is 0.8;

[0134] when The corresponding safety adjustment factor is 0.6.

[0135] when The corresponding safety adjustment factor is 0.5.

[0136] Using an absolute safe power limit as a constraint, the maximum and minimum power that satisfy Q being higher than a preset reference value are identified within the constraint range, and the range between the maximum and minimum power is used as the recommended high-efficiency power range; where Q = product output / energy consumption;

[0137] Additional notes: The recommended high-efficiency power range is the power range where the energy efficiency Q is higher than the preset reference value (such as 90% of the rated energy efficiency) under the constraint of the absolute safe power upper limit, and is calibrated by combining the "power-energy efficiency" characteristic curve corresponding to the state assessment index.

[0138] Energy efficiency curves corresponding to different state assessment indices are established using offline experiments or historical operating data.

[0139] The higher the condition assessment index (the worse the condition), the leftward shift of the "optimal power point" of the energy efficiency curve (higher energy efficiency at lower power); at the same power, the higher the condition assessment index, the lower the energy efficiency.

[0140] High-efficiency power limit: Within the absolute safe power limit, find the maximum power that satisfies Q≥90% of the rated energy efficiency; Example: Absolute safe power limit = 500.5kW, when W=0.55, when P=500kW, Q=90% of the rated energy efficiency, then the high-efficiency power limit = 500kW;

[0141] High-efficiency power lower limit: Within the absolute safe power upper limit, find the minimum power that satisfies Q≥90% of the rated energy efficiency; Example: Absolute safe power upper limit = 500.5kW, when W=0.55, when P=300kW, Q=90% of the rated energy efficiency, then the high-efficiency power lower limit = 300kW; Recommended high-efficiency power range is {300kW, 500kW};

[0142] The dynamic power operation window is the intersection of the "recommended high-efficiency power range" and the "absolute safe power limit". For example: if the high-efficiency power limit = 500kW ≤ the absolute safe power limit = 500.5kW, then the window is [300kW, 500kW]; if it equals the high-efficiency power limit of 510kW, then the window is [300kW, 500.5kW].

[0143] Fusion constraint scheduling: The objective function is to maximize the total profit within the scheduling cycle. The objective function explicitly includes production revenue, energy consumption cost and aging cost. The dynamic operating window is used as the core constraint condition, and rolling optimization is performed to obtain the optimal power command.

[0144] Additional notes: auxiliary constraints also include production task constraints, which mean that the total output within the scheduling cycle must meet the order demand to avoid insufficient output.

[0145] Since the dynamic power operation window changes in real time, a "rolling optimization" strategy is required to update the data and solve the model in cycles to ensure dynamic adaptation of instructions.

[0146] Specifically:

[0147] The objective function is expressed as follows: Where T is the scheduling period and t is the time step within the period; For production revenue, For energy consumption costs, Costs related to aging;

[0148] The variables are defined as follows:

[0149] = ;in Let be the output at time t. = ×v, Let t be the power at time t, and v be the production coefficient. This refers to the unit price of the product.

[0150] ; The time step length, For electricity price;

[0151] ; The preset aging cost coefficient (calibrated based on historical maintenance data); Indicates the lifetime calibration factor;

[0152] The higher the overall health index (the worse the health), the higher the P(t) value, the faster the aging process, and the higher the cost.

[0153] in Limited to the dynamically running window;

[0154] The optimal power obtained from the solution is output as the optimal power command, and then the process enters the next optimization cycle.

[0155] The above formulas are all dimensionless calculations. Dimensionless calculations can be performed using various methods such as standardization, which will not be elaborated here. The formulas are derived from software simulations based on a large amount of collected data, and the preset parameters in the formulas can be set by those skilled in the art according to the actual situation.

[0156] Example 2

[0157] Please see Figure 2 As shown, based on the method for monitoring and optimizing the operation status of a composite membrane electrolyzer provided in Embodiment 1 of this application, Embodiment 2 of this application proposes a system for monitoring and optimizing the operation status of a composite membrane electrolyzer. Embodiment 2 is merely a preferred embodiment of Embodiment 1, and its implementation will not affect the individual implementation of Embodiment 1.

[0158] Specifically, Embodiment 2 of this application provides a composite membrane electrolyzer operation status monitoring and scheduling optimization system, comprising:

[0159] The data input module is used to input electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data, and integrate them into a multi-source feature dataset;

[0160] The health assessment module is used to fuse various features in a multi-source feature dataset using a deep learning algorithm to output a comprehensive health index. Taking the comprehensive health index and historical operating power as input, combined with the equipment's service life, it outputs the estimated remaining lifespan of the equipment.

[0161] The boundary constraint module is used to combine the output comprehensive health index and estimated remaining life to dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer in the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window.

[0162] The constraint adjustment module is used to maximize the total profit within the scheduling cycle as the objective function. The objective function explicitly includes production revenue, energy consumption cost, and aging cost. The dynamic operating window is used as the core constraint condition, and rolling optimization is performed to obtain the optimal power command.

[0163] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, ATA hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0164] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0165] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0166] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0167] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0168] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0169] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable ATA hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0170] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for monitoring and optimizing the operation status of a composite membrane electrolyzer, characterized in that, include: S1: Construct a multi-source feature dataset that includes electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data; S2: Employ deep learning algorithms to fuse multi-source feature datasets and output a comprehensive health index. Using the comprehensive health index and historical operating power as input, combined with the equipment's service life, output the equipment's estimated remaining lifespan. S3: Combining the output comprehensive health index and estimated remaining lifespan, dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer under the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window; S4: The objective function is to maximize the total profit within the scheduling cycle. The objective function explicitly includes production revenue, energy consumption cost and aging cost. The dynamic operating window is used as the core constraint. Rolling optimization is performed to obtain the optimal power command.

2. The method for monitoring and optimizing the operation status of a composite membrane electrolyzer according to claim 1, characterized in that: In step S2, a comprehensive health index is output by fusing multi-source feature datasets. Electrochemical impedance spectroscopy data were analyzed to extract charge transfer resistance and membrane resistance; The infrared thermal imaging data is analyzed to extract the highest temperature in the film region, the temperature difference between the anode and cathode, and the proportion of hot spot area. The acoustic emission data is analyzed to extract the peak amplitude of the signal and the proportion of high-frequency events; Determine the normal operating values ​​corresponding to each type of data source based on historical normal data; The feature parameters extracted from each type of data source are combined with the corresponding normal operating values ​​and weighted to obtain the electrical health index, thermal health index, and mechanical health index. After normalizing the electrical health index, thermal health index, and mechanical health index, a weighted fusion process is used to obtain the comprehensive health index of the equipment.

3. The method for monitoring and optimizing the operation status of a composite membrane electrolyzer according to claim 2, characterized in that: In step S2, the comprehensive health index and historical operating power are used as inputs; A piecewise linear mapping rule is set between the comprehensive health index and health status to convert the comprehensive health index into health status; where health status includes normal status, moderate deterioration status and severe deterioration status, and each health status corresponds to a lifespan calibration coefficient; Extract the comprehensive health index from the x periods prior to the current time point, and denot it as the time-series health index; The time-series health index and the comprehensive health index at the current time point are respectively labeled as... According to the formula Calculate the degradation rate ; Degradation rate Multiplying this by the preset rate sensitivity coefficient yields the trend adjustment coefficient; The sum of the trend adjustment coefficient and the life calibration coefficient is taken as the life optimization coefficient of the equipment. Extract the historical operating power of the x cycles before the current time point, and take the average value as the average operating power; The ratio is calculated using the average operating power as the numerator and the rated power as the denominator, and is denoted as . ; pass Calculate the power correction factor ;in This is the preset power aging factor.

4. The method for monitoring and optimizing the operation status of a composite membrane electrolyzer according to claim 3, characterized in that: In step S2, the estimated remaining lifespan of the equipment is output based on the equipment's existing lifespan. Multiply the lifespan optimization factor by the power aging factor to obtain the estimated reduction factor; The estimated remaining lifespan of the equipment is calculated using the formula: Estimated Remaining Lifespan = Theoretical Lifespan of Equipment - Actual Lifespan × Estimated Reduction Factor.

5. The method for monitoring and optimizing the operating status of a composite membrane electrolyzer according to claim 1, characterized in that: In step S3, the absolute safe power limit of the electrolyzer under the current state is dynamically calculated; The remaining service index of the equipment is obtained by calculating the ratio between the estimated remaining life of the equipment as the numerator and the theoretical life of the equipment as the denominator. After normalizing the remaining usage index and comprehensive health index of the equipment, the equipment status assessment index is obtained by weighted calculation logic. Using the rated power as a benchmark, a piecewise linear mapping rule is set between the state assessment index and the safety adjustment coefficient. After converting the state assessment index into the safety adjustment coefficient, it is multiplied by the benchmark to obtain the absolute safe power limit of the electrolytic cell under the current state.

6. The method for monitoring and optimizing the operating status of a composite membrane electrolyzer according to claim 1, characterized in that: In step S3, the recommended high-efficiency power range of the electrolyzer under the current state is dynamically calculated. With an absolute safe power limit as a constraint, identify the maximum and minimum power that satisfy Q being higher than a preset reference value within the constraint range, and use the range between the maximum and minimum power as the recommended high-efficiency power range; Where Q = product output / energy consumption.

7. The method for monitoring and optimizing the operation status of a composite membrane electrolyzer according to claim 1, characterized in that: In step S4, the objective function explicitly includes production revenue, energy consumption cost, and aging cost. Production revenue, energy consumption costs, and aging costs are respectively labeled as... , , ; The variables are defined as follows: = ;in Let be the output at time t. = ×v, Let t be the power at time t, and v be the production coefficient. This refers to the unit price of the product. ; The time step length, For electricity price; ; This is the preset aging cost coefficient; Indicates the lifetime calibration factor; in It is limited to the dynamically running window.

8. The method for monitoring and optimizing the operation status of a composite membrane electrolyzer according to claim 7, characterized in that: S4 uses the dynamic running window as the core constraint to perform rolling optimization and obtain the optimal power command; The objective function is expressed as follows: Where T is the scheduling period and t is the time step within the period; For production revenue, For energy consumption costs, Costs related to aging; The optimal power obtained from the solution is output as the optimal power command, and then the process proceeds to the next optimization cycle.

9. A composite membrane electrolyzer operation status monitoring and scheduling optimization system, applied to the composite membrane electrolyzer operation status monitoring and scheduling optimization method according to any one of claims 1-8, characterized in that, include: The data input module is used to input electrochemical impedance spectroscopy data, infrared thermal imaging data, and acoustic emission data, and integrate them into a multi-source feature dataset; The health assessment module is used to fuse various features in a multi-source feature dataset using a deep learning algorithm to output a comprehensive health index. Taking the comprehensive health index and historical operating power as input, combined with the equipment's service life, it outputs the estimated remaining lifespan of the equipment. The boundary constraint module is used to combine the output comprehensive health index and estimated remaining life to dynamically calculate the absolute safe power limit and recommended high-efficiency power range of the electrolyzer in the current state, and use the intersection of the absolute safe power limit and the recommended high-efficiency power range as the dynamic operating window. The constraint adjustment module is used to maximize the total profit within the scheduling cycle as the objective function. The objective function explicitly includes production revenue, energy consumption cost, and aging cost. The dynamic operating window is used as the core constraint condition, and rolling optimization is performed to obtain the optimal power command.

Images