A method and system for testing voltage distribution of a fuel cell membrane electrode

Abstract

The present application relates to the technical field of fuel cell membrane electrode voltage distribution test, in particular to a kind of fuel cell membrane electrode voltage distribution test method and system, the method is through the layout potential sensor array, in the loading operation process of fuel cell, through potential sensor array synchronous acquisition original voltage vector;Reference electrode is laid in non-reaction zone, stable reference signal is applied to reference electrode, and feedback signal of reference electrode is collected after delay presetting stabilization time;Feedback signal and stable reference signal are obtained system measurement drift amount by difference calculation;According to the spatial continuity constraint of system measurement drift amount and membrane electrode voltage distribution, original voltage vector is jointly calibrated, and the voltage distribution vector of fuel cell membrane electrode is obtained.By continuously compensating the drift of measurement accuracy, the authenticity and reliability of the voltage distribution data during the long-period test of the fuel cell are ensured.

Description

Technical Field

[0001] This invention relates to the field of fuel cell membrane electrode voltage distribution testing technology, specifically a fuel cell membrane electrode voltage distribution testing method and system. Background Technology

[0002] Testing the voltage distribution of the membrane electrode assembly (MEA) of a fuel cell is an important means of characterizing its internal reaction state and consistency. A common approach is to deploy multiple potential sensors in the active region of the MEA and record the voltage signals at different locations using a multi-channel acquisition circuit. For comparability and accuracy, existing systems typically perform centralized calibration of each channel using a standard voltage source before testing to establish a benchmark calibration relationship.

[0003] However, during long-term operation, the acquisition link can be affected by factors such as temperature drift and changes in contact impedance, causing the output signal to gradually deviate. Existing methods generally correct this deviation by periodically pausing the experiment and recalibrating. However, this approach interrupts the testing process and disrupts data continuity; furthermore, during the interval between calibrations, the drift continues to accumulate but cannot be detected in real time, resulting in insufficient authenticity and reliability of the test results. This deficiency is particularly prominent when conducting long-term unattended durability experiments.

[0004] Therefore, there is a need for a method and system for testing the voltage distribution of fuel cell membrane electrodes that can continuously maintain measurement accuracy without interrupting the test, thereby ensuring the authenticity and reliability of the voltage distribution data throughout the entire test cycle. Summary of the Invention

[0005] (1) Technical problems to be solved

[0006] The purpose of this invention is to provide a method and system for testing the voltage distribution of a fuel cell membrane electrode assembly, in order to solve the problem of decreased accuracy in voltage distribution measurement caused by measurement system drift during continuous long-term testing of fuel cells.

[0007] (2) Technical solution

[0008] To achieve the above objectives, in one aspect, the present invention provides a method for testing the voltage distribution of a fuel cell membrane electrode, the method comprising:

[0009] Step S1: Deploy a potential sensor array at the measurement location covering the active area of ​​the membrane electrode in the fuel cell; during the operation of the fuel cell under load, synchronously acquire the original voltage vector at the measurement location through the potential sensor array.

[0010] Step S2: Place a reference electrode in the non-reaction zone of the test fixture in the fuel cell; apply a stable reference signal to the reference electrode, and collect the feedback signal of the reference electrode after a preset stabilization time.

[0011] Step S3: Calculate the system measurement drift by subtracting the feedback signal and the stable reference signal; perform joint calibration on the original voltage vector based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode.

[0012] Preferably, the method for synchronously acquiring the original voltage vector at the measurement location through the potential sensor array includes:

[0013] The potential sensor array is divided into different zones according to the flow channel direction of the fuel cell, including an inlet zone, a middle section zone, and an outlet zone; the flow channel direction is the main flow direction of the fuel cell gas within the flow field plate.

[0014] Within a preset sampling period, voltage data from the entrance region, middle section region, and exit region are collected respectively as a first original voltage vector, a second original voltage vector, and a third original voltage vector; the first original voltage vector, the second original voltage vector, and the third original voltage vector are concatenated in the order of the regions to obtain the original voltage vector.

[0015] Preferably, the method of applying a stable reference signal to the reference electrode and acquiring the feedback signal of the reference electrode after a preset stabilization time includes:

[0016] At the start of the preset sampling period, the analog switch is controlled to apply a stable reference signal to the reference electrode; after applying the stable reference signal, a preset stabilization time is delayed, and the feedback signal of the reference electrode is acquired simultaneously with the acquisition of the original voltage vector.

[0017] Preferably, the method for calculating the system measurement drift by subtracting the feedback signal and the stable reference signal includes:

[0018] The feedback signal is preprocessed to obtain a digitized feedback signal; the difference between the digitized feedback signal and the nominal value of the stable reference signal is calculated to obtain the voltage difference value. According to the preset channel gain coefficient Regarding the voltage difference Calculations are performed to obtain the system measurement drift. .

[0019] Preferably, the method for jointly calibrating the original voltage vector based on the system's measured drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode includes:

[0020] Based on the drift measured by the system, the first original voltage vector, the second original voltage vector, and the third original voltage vector are respectively subjected to partition compensation processing to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector; the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector are then sequentially concatenated according to the flow channel direction to obtain the drift corrected voltage vector. ;

[0021] In the entrance, middle, and exit areas, a first adjacency matrix, a second adjacency matrix, and a third adjacency matrix are constructed according to the placement relationship of the potential sensors, respectively. These three matrices are then combined to obtain an overall adjacency matrix. A smoothing matrix is ​​then calculated based on this overall adjacency matrix. ;

[0022] The drift correction voltage vector is adjusted according to the smoothing matrix L. The voltage distribution vector of the fuel cell membrane electrode is obtained by performing optimization.

[0023] Preferably, the method for performing partition compensation processing on the first original voltage vector, the second original voltage vector, and the third original voltage vector based on the drift measured by the system to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector includes:

[0024] Let the first original voltage vector, the second original voltage vector, and the third original voltage vector be denoted as follows: ; Obtain the preset correction weight factor corresponding to each partition. ,in These correspond to the entrance area, middle section area, and exit area, respectively; based on the preset correction weighting factor... Calculate the partition compensation vector .

[0025] in, To and A uniform vector of all 1s; the partitioned corrected voltage vector is calculated. .

[0026] Preferably, the drift correction voltage vector is adjusted according to the smoothing matrix L. Methods for optimizing and solving to obtain the voltage distribution vector of the fuel cell membrane electrode include:

[0027] Based on the smoothing matrix L and the drift correction voltage vector Construct an optimization objective function, the expression of which is:

[0028] .

[0029] in, For regularization parameters, For data consistency items, The objective function is then solved to obtain the optimal solution, which includes a smoothing constraint term. This serves as the voltage distribution vector for the membrane electrode assembly of a fuel cell.

[0030] Based on the same inventive concept, in another aspect, the present invention also provides a fuel cell membrane electrode voltage distribution testing system, the system comprising:

[0031] The sensor acquisition module is used to deploy a potential sensor array at the measurement location covering the active area of ​​the membrane electrode in the fuel cell; during the operation of the fuel cell, the original voltage vector at the measurement location is synchronously acquired through the potential sensor array.

[0032] The reference electrode acquisition module is used to deploy a reference electrode in the non-reaction zone of the test fixture in the fuel cell; apply a stable reference signal to the reference electrode, and acquire the feedback signal of the reference electrode after a preset stabilization time.

[0033] The joint calibration module is used to calculate the system measurement drift by subtracting the feedback signal and the stable reference signal; and to perform joint calibration on the original voltage vector based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode.

[0034] Preferably, the system further includes:

[0035] The visualization module is used to generate graphical data based on the voltage distribution vector of the fuel cell membrane electrode and display it in the form of a two-dimensional curve or pseudo-color graph through a graphical interface.

[0036] (3) Beneficial effects

[0037] Compared with the prior art, the beneficial effects of the present invention are:

[0038] 1. The system measurement drift is obtained through a reference electrode, and the original voltage vector is jointly corrected by combining this with the spatial continuity constraint of the membrane electrode voltage distribution. This allows for continuous compensation of measurement accuracy drift without interrupting testing, thereby ensuring the authenticity and reliability of voltage distribution data during long-term fuel cell testing.

[0039] 2. The original voltage vector is partitioned according to the gas flow direction of the fuel cell, and drift compensation is performed in each partition. This enables differentiated calibration for regions with different characteristics, improving the accuracy and physical plausibility of the overall voltage distribution results. Attached Figure Description

[0040] Figure 1 This is a flowchart of a fuel cell membrane electrode voltage distribution testing method according to Embodiment 1 of the present invention;

[0041] Figure 2 This is a schematic diagram of the module composition of a fuel cell membrane electrode voltage distribution testing system according to Embodiment 2 of the present invention;

[0042] Figure 3 This is a schematic diagram of the planar structure of the fuel cell membrane electrode of Embodiment 1 of the present invention. Detailed Implementation

[0043] 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.

[0044] Before providing examples, it is necessary to describe the application scenarios of this invention. These scenarios primarily target long-term testing and transient process analysis of fuel cells in laboratory research and engineering applications. During fuel cell life evaluation, durability studies, and stack optimization development, continuous monitoring of the voltage distribution at different locations on the membrane electrode assembly (MEA) is required to reflect changes in local reactivity, gas distribution unevenness, and hydrothermal state. However, existing testing methods largely rely on single initial calibrations, which cannot offset systematic errors caused by temperature drift and contact impedance changes during long-term operation. This leads to data distortion over time, particularly evident in unattended durability testing and dynamic conditions with rapid load switching. Therefore, the voltage distribution testing method and system proposed in this invention can maintain accuracy continuously without interrupting testing through real-time reference electrode verification and spatial continuity constraint correction. This makes it suitable for various application scenarios such as long-term stable operation monitoring of fuel cells, dynamic transient response capture, and high-precision data acquisition.

[0045] Example 1: As Figure 1 As shown, this embodiment provides a method for testing the voltage distribution of a fuel cell membrane electrode, the method comprising:

[0046] Step S1: Deploy a potential sensor array at the measurement location covering the active area of ​​the membrane electrode in the fuel cell; during the operation of the fuel cell under load, synchronously acquire the original voltage vector at the measurement location through the potential sensor array.

[0047] Step S2: Place a reference electrode in the non-reaction zone of the test fixture in the fuel cell; apply a stable reference signal to the reference electrode, and collect the feedback signal of the reference electrode after a preset stabilization time.

[0048] Step S3: Calculate the system measurement drift by subtracting the feedback signal and the stable reference signal; perform joint calibration on the original voltage vector based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode.

[0049] For example, in this embodiment, a potential sensor array is arranged on the active region of the fuel cell membrane electrode. This array can be fabricated using printed circuit board (PCB) technology, and its arrangement can be referenced. Figure 3 The active region is divided into a 2×10 matrix with 20 measurement points to obtain the necessary spatial resolution. These sensors are preferably made of chemically stable materials, such as gold-plated materials, to reduce contact resistance and withstand the fuel cell environment.

[0050] During fuel cell operation, for example under an 80A constant current load, a multi-channel data acquisition system synchronously acquires voltage signals from all potential sensors at a preset sampling rate of no less than 10Hz to form a raw voltage vector. To ensure signal quality, appropriate signal conditioning can be performed before the signal enters the acquisition card, such as using an instrumentation amplifier for preliminary amplification. The voltage distribution vector is a column vector composed of the voltages from N measurement points. In this embodiment, the raw voltage measurement data, after multi-channel acquisition, is typically stored and processed in vector form. The voltage vector mentioned herein refers to a sequential vector formed by arranging the output voltage values ​​of potential sensors located at different measurement positions in the active region of the membrane electrode. For example, when N potential sensors are deployed in the active region, the voltage distribution vector can be represented as... in, Indicates the first The vector represents the voltage values ​​at each measurement location. This vector directly reflects the voltage distribution of the membrane electrode at each measurement point in space and serves as the basic data format for subsequent partition compensation and spatial constraint optimization.

[0051] refer to Figure 3In the non-reactive zone of the test fixture, at the edge of the flow field plate, a reference electrode 101 is installed, such as an Ag / AgCl electrode or a platinum electrode, which exhibits good stability in a fuel cell environment. This reference electrode is positioned in a location that does not directly contact the main reactants but can sense the same electrolyte environment. At the beginning of each sampling cycle, a stable reference signal is applied to this reference electrode via an analog switch; this can be a DC signal with an amplitude of 100mV or a small-amplitude AC signal. A preset settling time, typically 5ms to 500ms, is delayed to ensure interface response stability and eliminate transient interference. Subsequently, the feedback signal on the reference electrode and the voltage signal from the potential sensor array are simultaneously acquired. The difference between the feedback signal and the applied stable reference signal is... After calculation, this can be used to assess the current measurement drift of the system. This drift amount... Through formula The calculation yielded, where The known gain coefficient of the data acquisition channel is given. Subsequently, based on the system's measured drift and considering the inherent spatial continuity constraint of the membrane electrode voltage distribution (i.e., the voltage between adjacent measurement points typically does not change abruptly), the original voltage vector is jointly calibrated. This spatial continuity constraint can be quantified by constructing an appropriate mathematical model, such as an adjacency matrix in graph theory or a regularization optimization method. The calibration process aims to find an optimal voltage distribution vector that best matches the drift-corrected measurement value while satisfying the physical constraint of spatial continuity. In this way, a voltage distribution vector that more accurately reflects the actual operating state of the fuel cell membrane electrode is obtained.

[0052] The above method can compensate for the drift of the measurement system in real time and online during the test, effectively suppress the measurement error caused by factors such as temperature changes, contact resistance fluctuations and aging of electronic components, significantly improve the accuracy and reliability of voltage distribution data in long-term testing, and does not require interruption of the normal operation of the fuel cell.

[0053] The method for synchronously acquiring the original voltage vector at the measurement location using the potential sensor array includes:

[0054] The potential sensor array is divided into different zones according to the flow channel direction of the fuel cell, including an inlet zone, a middle section zone, and an outlet zone; the flow channel direction is the main flow direction of the fuel cell gas within the flow field plate.

[0055] Within a preset sampling period, voltage data from the entrance region, middle section region, and exit region are collected respectively as a first original voltage vector, a second original voltage vector, and a third original voltage vector; the first original voltage vector, the second original voltage vector, and the third original voltage vector are concatenated in the order of the regions to obtain the original voltage vector.

[0056] For example, for a flow channel with an effective length of 200 mm, an array of 20 sensors is divided into 3 functional zones along the gas flow direction, see [reference]. Figure 3 The system comprises an inlet zone 102, a middle zone 103, and an outlet zone 104. The inlet zone covers the front section of the flow channel from 0 to 60 mm and is equipped with 6 sensors; the middle zone covers the section from 60 to 140 mm and is equipped with 8 sensors; and the outlet zone covers the section from 140 to 200 mm and is equipped with 6 sensors. The data acquisition system synchronously acquires voltage data from each zone through three independent acquisition threads within each 100 ms sampling period. First, the 6 voltage values ​​from the inlet zone are obtained to form the first raw voltage vector; then, the 8 voltage values ​​from the middle zone are obtained to form the second raw voltage vector; and finally, the 6 voltage values ​​from the outlet zone are obtained to form the third raw voltage vector. Due to the synchronous sampling technology, the acquisition time deviation of the three zones is less than 1 μs, ensuring data temporal consistency. Finally, the three voltage vectors are concatenated sequentially along the flow channel direction to form a complete 20-dimensional raw voltage vector. This zoned acquisition method not only conforms to the physical characteristics of the gradual changes in parameters such as reactant gas concentration and temperature along the flow channel direction of the fuel cell, but also facilitates the implementation of differentiated drift compensation strategies for different zone characteristics, improving the physical rationality and accuracy of the voltage distribution data.

[0057] The method of applying a stable reference signal to the reference electrode and acquiring the feedback signal of the reference electrode after a preset stabilization time includes:

[0058] At the start of the preset sampling period, the analog switch is controlled to apply a stable reference signal to the reference electrode; after applying the stable reference signal, a preset stabilization time is delayed, and the feedback signal of the reference electrode is acquired simultaneously with the acquisition of the original voltage vector.

[0059] For example, at the beginning of each sampling cycle, the data acquisition system triggers and controls an analog switch array via a digital I / O port. This analog switch array switches a stable reference signal generated by a high-precision reference voltage source to the reference electrode circuit. The stable reference signal is a small-amplitude DC voltage signal of 100mV. The reason for choosing a DC signal is that it can more directly characterize the DC drift characteristics of the measurement circuit, which is the most significant source of error in fuel cell voltage measurement. Simultaneously, the small amplitude avoids significant polarization or damage to the reference electrode itself, ensuring its long-term stability. After applying the reference signal, the system starts a high-precision timer for a delay, with a preset stabilization time set to 50ms. The preset stabilization time can be experimentally determined by applying a step signal and observing the time required for the output to reach 99% of its steady-state value. This stabilization time is sufficient for the double-layer charging process at the reference electrode-electrolyte interface to reach a stable state, while eliminating transient responses in the measurement circuit. After the delay time is reached, the data acquisition system simultaneously starts two acquisition tasks: on the one hand, it acquires the voltage values ​​of each channel of the potential sensor array to form the original voltage vector through a multiplexed ADC; on the other hand, it acquires the feedback signal on the reference electrode through a dedicated ADC channel. Both acquisition tasks are triggered by the same clock source, ensuring a synchronization accuracy of better than 1 μs. This synchronous acquisition mechanism guarantees a strict time correspondence between the reference electrode feedback signal and the sensor array voltage data, providing a foundation for the accurate calculation of subsequent system drift.

[0060] The method for calculating the system measurement drift by the difference between the feedback signal and the stable reference signal includes:

[0061] The feedback signal is preprocessed to obtain a digitized feedback signal; the difference between the digitized feedback signal and the nominal value of the stable reference signal is calculated to obtain the voltage difference value. According to the preset channel gain coefficient Regarding the voltage difference Calculations are performed to obtain the system measurement drift. .

[0062] For example, the raw feedback signal acquired from the reference electrode first passes through an anti-aliasing low-pass filter with a cutoff frequency set at more than ten times the reference signal frequency to effectively suppress high-frequency noise. The filtered signal is then sampled and quantized by a high-resolution analog-to-digital converter, converting it into a digital signal, i.e., the digitized feedback signal. This digital signal is then subtracted from a known stable reference signal nominal value, i.e., one hundred millivolts, to obtain a voltage difference value. The system measures the amount of drift. By voltage difference Divide by the preset channel gain coefficient The calculated channel gain coefficient is obtained. Obtained through precise calibration before the system leaves the factory, its value is typically close to 1 but encompasses the actual gain characteristics of the entire measurement link. For example, when measured... 5mV and gain coefficient When the value is 1.02, the calculated system measurement drift is... The drift is approximately 4.9mV. This drift reflects the systematic errors caused by factors such as temperature drift and component aging throughout the entire measurement loop from signal application to acquisition completion, providing an accurate basis for subsequent precise calibration of voltage distribution data.

[0063] The method for jointly calibrating the original voltage vector based on the spatial continuity constraints of the system's measured drift and membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode includes:

[0064] Based on the drift measured by the system, the first original voltage vector, the second original voltage vector, and the third original voltage vector are respectively subjected to partition compensation processing to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector; the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector are then sequentially concatenated according to the flow channel direction to obtain the drift corrected voltage vector. ;

[0065] In the entrance, middle, and exit areas, a first adjacency matrix, a second adjacency matrix, and a third adjacency matrix are constructed according to the placement relationship of the potential sensors, respectively. These three matrices are then combined to obtain an overall adjacency matrix. A smoothing matrix is ​​then calculated based on this overall adjacency matrix. ;

[0066] The drift correction voltage vector is adjusted according to the smoothing matrix L. The voltage distribution vector of the fuel cell membrane electrode is obtained by performing optimization.

[0067] For example, obtaining the system measurement drift. Next, the original voltage vectors of each zone are first differentiated and compensated. Considering the differences in reactant gas concentration, humidity, and temperature distribution along the flow path of the fuel cell, each zone has a different sensitivity to system drift. Therefore, different correction weighting factors are assigned to the inlet, middle, and outlet zones. The inlet zone typically uses a smaller weighting factor, the outlet zone a larger value, and the middle zone a moderate value. Through this zoned compensation process, a corrected voltage vector that better reflects the characteristics of each zone is obtained.

[0068] By concatenating the corrected voltage vectors of the three partitions along the flow channel direction, a preliminary drift corrected voltage vector is obtained. Subsequently, an adjacency matrix is ​​constructed based on the spatial arrangement of the sensors. This matrix describes the spatial adjacency relationships between each measurement point. The purpose of the adjacency matrix is ​​to establish local topological constraints for each measurement point, ensuring that isolated points or unreasonable voltage jumps do not occur in subsequent calculations.

[0069] Based on the adjacency matrix, a smoothing matrix L is further calculated. Its function is to quantify the spatial constraints and penalize drastic voltage differences between adjacent measurement points, thereby ensuring that the voltage distribution meets the continuity characteristics required by the membrane electrode. In other words, the smoothing matrix introduces a physically enforced condition of "smooth voltage distribution" in the optimization solution.

[0070] Finally, an optimization algorithm is used to correct the drift voltage vector. The solution process considers both measurement data consistency and spatial smoothness constraints during optimization. The adjacency matrix ensures the authenticity of local adjacency relationships in the calculation, while the smoothness matrix ensures the rationality of the overall distribution trend. Through this trade-off, the resulting voltage distribution vector eliminates system drift errors while maintaining the spatial continuity of the membrane electrode voltage distribution, thereby significantly improving the accuracy and physical reliability of the test results.

[0071] The method for performing partition compensation processing on the first original voltage vector, the second original voltage vector, and the third original voltage vector based on the drift measured by the system to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector includes:

[0072] Let the first original voltage vector, the second original voltage vector, and the third original voltage vector be denoted as follows: ; Obtain the preset correction weight factor corresponding to each partition. ,in These correspond to the entrance area, middle section area, and exit area, respectively; based on the preset correction weighting factor... Calculate the partition compensation vector .

[0073] in, To and A uniform vector of all 1s; the partitioned corrected voltage vector is calculated. .

[0074] For example, the voltage data collected by the six sensors in the entrance area is recorded as the first original voltage vector. The data from the eight sensors in the middle section are recorded as the second original voltage vector. The data from the six sensors in the export area are denoted as the third raw voltage vector. Based on the differences in fuel cell characteristics along the flow channel, different correction weighting factors are set for each zone: inlet zone weighting factor. Take 0.8 as the weighting factor for the middle section. Set to 1.0, export area weighting factor The weighting factor is set to 0.9. These weighting factors take into account the different reaction conditions and measurement characteristics of each region. The inlet region has a high gas concentration and vigorous reaction, making it relatively insensitive to system drift; therefore, a smaller weight is assigned. The outlet region has a lower reactant concentration and increased water content, making it more sensitive to measurement drift; therefore, a larger weight is assigned. The weighting factors can be preset according to the fuel cell operating characteristics, for example, based on simulated gas concentration distribution curves, water content distribution experimental results, or error sensitivity parameters statistically obtained from long-term durability tests.

[0075] The partition compensation vector is obtained through the formula Calculation, where This is a single-dimensional vector with the same dimensions as the voltage vectors of each zone. Taking the entrance zone as an example, its dimension is six, so I1 is a six-dimensional single-dimensional vector. The calculated compensation vector... From the original voltage vector Subtracting from the middle yields the corrected partition voltage vector. This zone-specific compensation method can more accurately reflect the differences in the impact of system drift on different regions, and significantly improves the accuracy and physical rationality of voltage distribution calibration compared to a globally uniform compensation method.

[0076] The drift correction voltage vector is adjusted according to the smoothing matrix L. Methods for optimizing and solving to obtain the voltage distribution vector of the fuel cell membrane electrode include:

[0077] Based on the smoothing matrix L and the drift correction voltage vector Construct an optimization objective function, the expression of which is:

[0078] .

[0079] in, For regularization parameters, For data consistency items, The objective function is then solved to obtain the optimal solution, which includes a smoothing constraint term. This serves as the voltage distribution vector for the membrane electrode assembly of a fuel cell.

[0080] For example, in obtaining the drift correction voltage vector After obtaining the smoothing matrix L, an optimization objective function is constructed that includes data consistency constraints and smoothness constraints. This objective function consists of two parts: the data consistency term requires the obtained voltage distribution vector to be solved. As close as possible to the actual measured value The smoothing constraint term imposes a spatial continuity constraint through the smoothing matrix L, requiring that the voltage values ​​of adjacent sensors change gradually. The regularization parameter λ is used to adjust the relative weight between the data consistency term and the smoothing constraint term. Its optimal value can be determined as follows: during the initial calibration phase of the system, a known, uniformly distributed voltage field is applied, and the system automatically optimizes by minimizing the error between the reconstructed voltage field and the known voltage field. The value of is typically between 0.1 and 10, for example, for a 100cm... 2 The active region, A value of 2.5 yields good results.

[0081] The solution process employs a least-squares-based optimization algorithm, obtaining a closed-form solution for the objective function through matrix differentiation. Unlike conventional smoothing methods, this invention incorporates the voltage vector after partition compensation during optimization, ensuring that the objective function not only constrains the overall smoothness but also reflects the differences between different partitions, avoiding distortion caused by forced smoothing across partitions. Specifically, the objective function is rewritten as a quadratic form, its derivative is taken, and the derivative is set to zero, yielding a system of linear equations. The coefficient matrix of this system is a symmetric positive definite matrix, which can be efficiently solved using numerical methods such as Cholesky decomposition or the conjugate gradient method. The final optimal solution is obtained. This voltage distribution vector satisfies the dual constraints of measurement data consistency and spatial smoothness. The result is faithful to the original measurement data and conforms to the spatial continuity law of membrane electrode voltage distribution, which significantly improves the accuracy and physical rationality of the voltage distribution result.

[0082] Example 2: Based on the same inventive concept, such as Figure 2 As shown, this embodiment also provides a fuel cell membrane electrode voltage distribution testing system, the system comprising:

[0083] The sensor acquisition module is used to deploy a potential sensor array at the measurement location covering the active area of ​​the membrane electrode in the fuel cell; during the operation of the fuel cell under load, the original voltage vector at the measurement location is synchronously acquired through the potential sensor array.

[0084] The reference electrode acquisition module is used to deploy a reference electrode in the non-reaction zone of the test fixture in the fuel cell; apply a stable reference signal to the reference electrode, and acquire the feedback signal of the reference electrode after a preset stabilization time.

[0085] The joint calibration module is used to calculate the system measurement drift by subtracting the feedback signal and the stable reference signal; and to perform joint calibration on the original voltage vector based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode.

[0086] The system also includes:

[0087] The visualization module is used to generate graphical data based on the voltage distribution vector of the fuel cell membrane electrode and display it in the form of a two-dimensional curve or pseudo-color graph through a graphical interface.

[0088] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0089] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for testing the voltage distribution of a fuel cell membrane electrode, characterized in that, The method includes: A potential sensor array is deployed at the measurement location covering the active region of the membrane electrode in the fuel cell; during the operation of the fuel cell under load, the original voltage vector at the measurement location is synchronously acquired through the potential sensor array; A reference electrode is placed in the non-reaction zone of a test fixture in a fuel cell; a stable reference signal is applied to the reference electrode, and the feedback signal of the reference electrode is acquired after a preset stabilization time. The system measurement drift is calculated by subtracting the feedback signal and the stable reference signal; the original voltage vector is jointly calibrated based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode. The method for synchronously acquiring the original voltage vector at the measurement location using the potential sensor array includes: The potential sensor array is divided into different zones according to the flow channel direction of the fuel cell, including an inlet zone, a middle section zone, and an outlet zone; the flow channel direction is the main flow direction of the fuel cell gas within the flow field plate. Within a preset sampling period, voltage data from the entrance region, middle section region, and exit region are collected respectively as a first original voltage vector, a second original voltage vector, and a third original voltage vector; the first original voltage vector, the second original voltage vector, and the third original voltage vector are concatenated in the order of the regions to obtain the original voltage vector; The method for jointly calibrating the original voltage vector based on the spatial continuity constraints of the system's measured drift and membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode includes: Based on the drift measured by the system, the first original voltage vector, the second original voltage vector, and the third original voltage vector are subjected to partition compensation processing to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector; the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector are then sequentially concatenated along the flow channel direction to obtain the drift corrected voltage vector. ; In the entrance, middle, and exit areas, a first adjacency matrix, a second adjacency matrix, and a third adjacency matrix are constructed according to the placement relationship of the potential sensors, respectively. These three matrices are then combined to obtain an overall adjacency matrix. A smoothing matrix is ​​then calculated based on this overall adjacency matrix. ; The drift correction voltage vector is adjusted according to the smoothing matrix L. The voltage distribution vector of the fuel cell membrane electrode is obtained by performing optimization. The drift correction voltage vector is adjusted according to the smoothing matrix L. Methods for optimizing and solving to obtain the voltage distribution vector of the fuel cell membrane electrode include: Based on the smoothing matrix L and the drift correction voltage vector Construct an optimization objective function, the expression of which is: ； in, For regularization parameters, For data consistency items, The objective function is then solved to obtain the optimal solution, which includes a smoothing constraint term. This serves as the voltage distribution vector for the membrane electrode assembly of a fuel cell.

2. The method for testing the voltage distribution of a fuel cell membrane electrode according to claim 1, characterized in that, The method of applying a stable reference signal to the reference electrode and acquiring the feedback signal of the reference electrode after a preset stabilization time includes: At the start of the preset sampling period, the analog switch is controlled to apply a stable reference signal to the reference electrode; after applying the stable reference signal, a preset stabilization time is delayed, and the feedback signal of the reference electrode is acquired simultaneously with the acquisition of the original voltage vector.

3. The method for testing the voltage distribution of a fuel cell membrane electrode according to claim 2, characterized in that, The method for calculating the system measurement drift by the difference between the feedback signal and the stable reference signal includes: The feedback signal is preprocessed to obtain a digitized feedback signal; the difference between the digitized feedback signal and the nominal value of the stable reference signal is calculated to obtain the voltage difference value. According to the preset channel gain coefficient Regarding the voltage difference Calculations are performed to obtain the system measurement drift. .

4. The method for testing the voltage distribution of a fuel cell membrane electrode according to claim 1, characterized in that, The method for performing partition compensation processing on the first original voltage vector, the second original voltage vector, and the third original voltage vector based on the drift measured by the system to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector includes: Let the first original voltage vector, the second original voltage vector, and the third original voltage vector be denoted as follows: Obtain the preset correction weight factor corresponding to each region. ,in These correspond to the entrance area, middle section area, and exit area, respectively; based on the preset correction weighting factor... Calculate the partition compensation vector ; in, To and A uniform vector of all 1s; the partitioned corrected voltage vector is calculated. .

5. A fuel cell membrane electrode voltage distribution testing system, characterized in that, The system includes: The sensor acquisition module is used to deploy a potential sensor array at the measurement location covering the active area of ​​the membrane electrode in the fuel cell; during the operation of the fuel cell, the original voltage vector at the measurement location is synchronously acquired through the potential sensor array; A reference electrode acquisition module is used to deploy a reference electrode in the non-reaction zone of a test fixture in a fuel cell; apply a stable reference signal to the reference electrode, and acquire the feedback signal of the reference electrode after a preset stabilization time. The joint calibration module is used to calculate the system measurement drift by subtracting the feedback signal and the stable reference signal; and to perform joint calibration on the original voltage vector based on the system measurement drift and the spatial continuity constraint of the membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode. The method for synchronously acquiring the original voltage vector at the measurement location using the potential sensor array includes: The potential sensor array is divided into different zones according to the flow channel direction of the fuel cell, including an inlet zone, a middle section zone, and an outlet zone; the flow channel direction is the main flow direction of the fuel cell gas within the flow field plate. Within a preset sampling period, voltage data from the entrance region, middle section region, and exit region are collected respectively as a first original voltage vector, a second original voltage vector, and a third original voltage vector; the first original voltage vector, the second original voltage vector, and the third original voltage vector are concatenated in the order of the regions to obtain the original voltage vector; The method for jointly calibrating the original voltage vector based on the spatial continuity constraints of the system's measured drift and membrane electrode voltage distribution to obtain the voltage distribution vector of the fuel cell membrane electrode includes: Based on the drift measured by the system, the first original voltage vector, the second original voltage vector, and the third original voltage vector are subjected to partition compensation processing to obtain the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector; the first corrected voltage vector, the second corrected voltage vector, and the third corrected voltage vector are then sequentially concatenated along the flow channel direction to obtain the drift corrected voltage vector. ; In the entrance, middle, and exit areas, a first adjacency matrix, a second adjacency matrix, and a third adjacency matrix are constructed according to the placement relationship of the potential sensors, respectively. These three matrices are then combined to obtain an overall adjacency matrix. A smoothing matrix is ​​then calculated based on this overall adjacency matrix. ; The drift correction voltage vector is adjusted according to the smoothing matrix L. The voltage distribution vector of the fuel cell membrane electrode is obtained by performing optimization. The drift correction voltage vector is adjusted according to the smoothing matrix L. Methods for optimizing and solving to obtain the voltage distribution vector of the fuel cell membrane electrode include: Based on the smoothing matrix L and the drift correction voltage vector Construct an optimization objective function, the expression of which is: ； in, For regularization parameters, For data consistency items, The objective function is then solved to obtain the optimal solution, which includes a smoothing constraint term. This serves as the voltage distribution vector for the membrane electrode assembly of a fuel cell.

6. The fuel cell membrane electrode voltage distribution testing system according to claim 5, characterized in that, The system also includes: The visualization module is used to generate graphical data based on the voltage distribution vector of the fuel cell membrane electrode and display it in the form of a two-dimensional curve or pseudo-color graph through a graphical interface.

Images