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

Abstract

The present application relates to the technical field of fuel cell membrane electrode current test, in particular to a kind of fuel cell membrane electrode current distribution test method and system, the method includes installing sensor array;Current measurement data and point temperature measurement data at lattice position are obtained by sensor array;Current measurement data is cleaned according to current direction criterion, after direction abnormal measurement data is eliminated, first current cleaning data is obtained, and current two-dimensional distribution diagram is mapped to obtain;Temperature two-dimensional distribution diagram that is spatially aligned with current two-dimensional distribution diagram is mapped to obtain;Correlation analysis is carried out on current two-dimensional distribution diagram and temperature two-dimensional distribution diagram, and current measurement data value abnormal point is identified, after current measurement data value abnormal point is eliminated, fuel cell membrane electrode current distribution diagram is reconstructed to obtain.The present application identifies and eliminates current measurement data value abnormal point by correlation analysis, improves the accuracy of fuel cell membrane electrode current distribution calculation result.

Description

Technical Field

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

[0002] As a highly efficient and clean energy conversion device, the performance of the membrane electrode assembly (MEA), a core component of fuel cells, directly determines the output efficiency, lifespan, and reliability of the fuel cell stack. A key indicator of MEA performance is the uniformity of its internal current distribution. Therefore, accurate testing of the current distribution of the fuel cell MEA under operating conditions is of great significance.

[0003] Existing current distribution testing techniques typically employ embedded sensor arrays for online measurement, but these techniques suffer from several technical challenges: First, once embedded in the fuel cell stack, sensors are susceptible to changes in contact resistance and electromagnetic interference, leading to outliers in the raw current data. Current methods struggle to identify and remove these outliers. Second, existing methods often rely solely on current as a single parameter for analysis, lacking a self-verification mechanism based on multi-dimensional data, and thus failing to identify data distortions caused by sensor malfunctions or installation issues. Furthermore, the current distribution and temperature distribution in fuel cell membrane electrode assemblies are strongly coupled, allowing for cross-validation and data cleaning of current measurement results using temperature data. However, current technologies fail to fully leverage the value of both temperature and current data.

[0004] Therefore, it is necessary to propose a current distribution testing method that can integrate multi-dimensional data information such as temperature and current to identify and clean abnormal current test data, so as to improve the accuracy of current distribution test results of fuel cell membrane electrode assembly. 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 current distribution of fuel cell membrane electrode assembly (MEA), which integrates multi-dimensional data information such as temperature and current to identify abnormal current test data and improve the accuracy of the current distribution calculation results of fuel cell MEA.

[0007] (2) Technical solution

[0008] To achieve the above objectives, the present invention provides a method for testing the current distribution of a fuel cell membrane electrode assembly, the method comprising the following steps:

[0009] S1, Install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and obtain the spatial coordinates of the dot matrix position.

[0010] S2, the current measurement data is cleaned according to the pre-set current direction criterion, and the abnormal direction measurement data is removed to obtain the first current cleaned data; the two-dimensional current distribution map is obtained by interpolation algorithm based on the spatial coordinates of the dot matrix position and the first current cleaned data.

[0011] S3. Based on the spatial coordinates of the lattice positions and the temperature measurement data at the lattice positions, an interpolation algorithm is used to map and obtain a two-dimensional temperature distribution map that is spatially aligned with the two-dimensional current distribution map.

[0012] S4. Through data fusion algorithm, correlation analysis is performed on the two-dimensional current distribution map and the two-dimensional temperature distribution map to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the current distribution map of the fuel cell membrane electrode is reconstructed using an interpolation algorithm.

[0013] Furthermore, the method for cleaning the current measurement data according to a pre-set current direction criterion, and removing measurement data with abnormal direction to obtain the first current cleaned data includes:

[0014] Determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is discarded; otherwise, it is retained, and the first current cleaning data is obtained.

[0015] Furthermore, the method for identifying outliers in current measurement data by performing correlation analysis on the two-dimensional current distribution map and the two-dimensional temperature distribution map using a data fusion algorithm includes:

[0016] The two-dimensional current distribution map is normalized using the minimum-maximum normalization algorithm to obtain the normalized current distribution function; the two-dimensional temperature distribution map is normalized using the minimum-maximum normalization algorithm to obtain the normalized temperature distribution function.

[0017] The normalized current distribution function is discretized according to a pre-set sampling step size to obtain a normalized current discrete point matrix; the normalized temperature distribution function is discretized according to a pre-set sampling step size to obtain a normalized temperature discrete point matrix; the normalized current discrete points are renumbered according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix to obtain the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; the normalized temperature discrete points are renumbered according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix to obtain the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points.

[0018] The correlation index between the normalized current sequence and the normalized temperature sequence is calculated and denoted as the benchmark correlation index.

[0019] When the baseline correlation index is less than the preset baseline correlation threshold, an error is reported and the user is prompted to check the correctness of the sensor array installation; the baseline correlation threshold is a value greater than 0 and less than 1; when the baseline correlation index is greater than or equal to the baseline correlation threshold, the first current point to the Nth current point are removed from the normalized current sequence in sequence, and the first temperature point to the Nth temperature point are removed from the normalized temperature sequence in sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence; the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index; i is an integer variable ranging from 1 to N; the values ​​of i are iterated to obtain the first correlation index to the Nth correlation index; based on the first correlation index to the Nth correlation index and the baseline correlation index, abnormal discrete points are screened using a preset screening standard; the screening standard is:

[0020]

[0021] Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

[0022] Furthermore, the method for calculating the correlation index between the normalized current sequence and the normalized temperature sequence, and denoting it as the benchmark correlation index, includes:

[0023] The correlation index between the normalized current sequence and the normalized temperature sequence is calculated using the first correlation degree calculation formula, and is denoted as the benchmark correlation index; the first correlation degree calculation formula is:

[0024]

[0025] Among them, X j Indicates the j-th current point; T represents the first mean; j Represents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

[0026] Furthermore, the method for calculating the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence, denoted as the i-th correlation index, includes:

[0027] The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated using the second correlation degree calculation formula, and denoted as the i-th correlation index; the second correlation degree calculation formula is as follows:

[0028]

[0029] Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

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

[0031] The sensor array arrangement module is used to install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; to acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; to establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and to obtain the spatial coordinates of the dot matrix position.

[0032] The direction cleaning module, connected to the sensor array arrangement module, is used to clean the current measurement data according to the pre-set current direction criteria, remove abnormal direction measurement data, and obtain the first current cleaning data; the two-dimensional current distribution map is obtained by interpolation algorithm based on the spatial coordinates of the dot matrix position and the first current cleaning data.

[0033] The interpolation module, connected to the direction cleaning module, is used to map a two-dimensional temperature distribution map that is spatially aligned with the current two-dimensional distribution map by using an interpolation algorithm based on the spatial coordinates of the dot matrix position and the temperature measurement data at the dot matrix position.

[0034] The reconstruction module, connected to the interpolation module, is used to perform correlation analysis on the two-dimensional current distribution map and the two-dimensional temperature distribution map through a data fusion algorithm to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the interpolation algorithm is used to reconstruct the current distribution map of the fuel cell membrane electrode.

[0035] Furthermore, the directional cleaning module includes:

[0036] The data rejection module is used to determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is rejected; otherwise, it is retained to obtain the first current cleaning data.

[0037] Furthermore, the reconstruction module includes:

[0038] The normalization module is used to normalize the two-dimensional current distribution map using the minimum-maximum normalization algorithm to obtain the normalized current distribution function; and to normalize the two-dimensional temperature distribution map using the minimum-maximum normalization algorithm to obtain the normalized temperature distribution function.

[0039] The sampling module, connected to the normalization module, is used to discretize the normalized current distribution function according to a pre-set sampling step size, obtaining a normalized current discrete point matrix; to discretize the normalized temperature distribution function according to a pre-set sampling step size, obtaining a normalized temperature discrete point matrix; to renumber the normalized current discrete points according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix, obtaining the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; and to renumber the normalized temperature discrete points according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix, obtaining the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points.

[0040] The benchmark correlation index calculation module, connected to the sampling module, is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence, denoted as the benchmark correlation index.

[0041] A filtering module, connected to the benchmark correlation index calculation module, is used to report an error and prompt the user to check the correctness of the sensor array installation when the benchmark correlation index is less than a preset benchmark correlation threshold. The benchmark correlation threshold is a value greater than 0 and less than 1. When the benchmark correlation index is greater than or equal to the benchmark correlation threshold, the first current point to the Nth current point are sequentially removed from the normalized current sequence, and the first temperature point to the Nth temperature point are sequentially removed from the normalized temperature sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence. The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index, where i is an integer variable ranging from 1 to N. The values ​​of i are iterated to obtain the first correlation index to the Nth correlation index. Based on the first correlation index to the Nth correlation index and the benchmark correlation index, abnormal discrete points are filtered using a preset filtering standard. The filtering standard is:

[0042]

[0043] Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

[0044] Furthermore, the benchmark correlation index calculation module includes:

[0045] The first correlation calculation module is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence using the first correlation calculation formula, which is denoted as the benchmark correlation index; the first correlation calculation formula is:

[0046]

[0047] Among them, X j Indicates the j-th current point; T represents the first mean; j Represents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

[0048] Furthermore, the filtering module includes:

[0049] The second correlation calculation module is used to calculate the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence using the second correlation calculation formula, denoted as the i-th correlation index; the second correlation calculation formula is:

[0050]

[0051] Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

[0052] (3) Beneficial effects

[0053] Compared with the prior art, the beneficial effects of the present invention are:

[0054] By employing a data fusion algorithm, correlation analysis is performed on the two-dimensional current distribution map and the two-dimensional temperature distribution map to identify outliers in the current measurement data. After removing these outliers from the two-dimensional current distribution map, an interpolation algorithm is used to reconstruct the fuel cell membrane electrode current distribution map. This method can improve the accuracy of the calculated fuel cell membrane electrode current distribution. Attached Figure Description

[0055] Figure 1 This is a flowchart of a fuel cell membrane electrode current distribution testing method according to Embodiment 1 of the present invention;

[0056] Figure 2 This is a schematic diagram of the module composition of a fuel cell membrane electrode current distribution testing system according to Embodiment 2 of the present invention. Detailed Implementation

[0057] 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.

[0058] Before giving examples, it is necessary to explain the application scenarios of the present invention. The present invention is used to identify abnormal data of fuel cell membrane electrode current test, so as to obtain a more accurate fuel cell membrane electrode current distribution.

[0059] Example 1: As Figure 1 As shown, this embodiment provides a method for testing the current distribution of a fuel cell membrane electrode, the method comprising the following steps:

[0060] S1, Install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and obtain the spatial coordinates of the dot matrix position.

[0061] S2, the current measurement data is cleaned according to the pre-set current direction criterion, and the abnormal direction measurement data is removed to obtain the first current cleaned data; the two-dimensional current distribution map is obtained by interpolation algorithm based on the spatial coordinates of the dot matrix position and the first current cleaned data.

[0062] S3. Based on the spatial coordinates of the lattice positions and the temperature measurement data at the lattice positions, an interpolation algorithm is used to map and obtain a two-dimensional temperature distribution map that is spatially aligned with the two-dimensional current distribution map.

[0063] S4. Through data fusion algorithm, correlation analysis is performed on the two-dimensional current distribution map and the two-dimensional temperature distribution map to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the current distribution map of the fuel cell membrane electrode is reconstructed using an interpolation algorithm.

[0064] For example, a sensor array is installed at pre-defined dot matrix positions on the cathode side of the membrane electrode under test (MEA) of the fuel cell stack. The sensor array is a 4×4 array comprising 16 sensors. These 16 sensors are uniformly arranged in a 4×4 array on the cathode side of the MEA. The area on the cathode side of the MEA to be measured for current distribution is a square region. The distance between any two adjacent sensors, either horizontally or vertically, is equal. Current and temperature measurement data at each dot matrix position are synchronously acquired using the sensor array, ensuring temporal synchronization and spatial correspondence of the current and temperature measurement data. A two-dimensional coordinate system is established with the geometric center of the cathode side of the MEA as the origin, and each sensor is assigned a spatial coordinate point to obtain the spatial coordinates of the dot matrix position.

[0065] Since the current direction in a fuel cell reaction is always from the anode to the cathode, a current direction criterion is established. The original current measurement data is initially cleaned using this criterion, automatically identifying and removing all measurement data with abnormal directions (i.e., current values ​​displayed as flowing from the back of the sensor to the front), resulting in the first cleaned current data after direction verification. The back and front of the sensor need to be predefined to ensure that when the fuel cell current direction is from the anode to the cathode, the measurement data is correct, and the sensor installation orientation is correct, the current measured by the sensor flows from the front to the back. Based on the spatial coordinates of the dot matrix positions and the first cleaned current data, an interpolation algorithm is used to generate a continuous two-dimensional current distribution map. By employing a data cleaning method based on the current direction criterion, reversed current data caused by hardware problems such as reverse sensor installation or other reasons is prevented, providing high-quality data input and a solid foundation for subsequent analysis.

[0066] Based on the spatial coordinates of the lattice locations and the corresponding temperature measurement data, a two-dimensional temperature distribution map is generated using the same interpolation algorithm and mesh parameters as in step S2. By maintaining complete consistency in the coordinate system, interpolation method, and mesh, the generated two-dimensional temperature distribution map is ensured to be strictly aligned spatially with the two-dimensional current distribution map.

[0067] A data fusion algorithm is used to analyze the correlation between spatially aligned two-dimensional current and temperature distribution maps. Based on the strong coupling relationship between current density and local temperature during fuel cell operation, the algorithm identifies outliers that violate this physical law (e.g., extremely high current values ​​with no corresponding temperature change due to electromagnetic interference, or extremely low current values ​​with abnormally high temperature due to poor sensor contact). After removing these outliers from the current distribution data, an interpolation algorithm is used to reconstruct the regions with data gaps, ultimately obtaining a high-precision, high-reliability fuel cell membrane electrode current distribution map that has undergone dual verification of directionality and rationality.

[0068] Furthermore, the method for cleaning the current measurement data according to a pre-set current direction criterion, and removing measurement data with abnormal direction to obtain the first current cleaned data includes:

[0069] Determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is discarded; otherwise, it is retained, and the first current cleaning data is obtained.

[0070] For example, based on the electrochemical principle that the current direction is constant during fuel cell operation, the front and back sides of the sensor are predefined, and a current direction criterion is set accordingly. The sign and direction of the current measurement value at each point in the matrix are determined one by one. If the indicated current flows from the back side of the sensor to the front side, it is considered abnormal data and discarded. Only normal data where the indicated current flows from the front side to the back side are retained, thus obtaining the first current cleaning data. By using the current direction criterion for initial screening, the validity of the data can be preliminarily verified. Systematic directional errors caused by reverse sensor installation, wiring errors, or circuit faults are filtered out from a physical perspective, providing a basic data quality guarantee to prevent distortion of the subsequent current distribution map. It is worth noting that preliminary screening based on the current direction is the most intuitive, direct, and highly effective data cleaning method. This is because some steps in the sensor installation process require human operation. Human operation is prone to errors, such as reversing the sensor orientation. These errors are relatively common due to human factors. Therefore, this method can easily and directly eliminate a large portion of data errors. Traditional data anomaly detection algorithms often employ complex methods to identify raw data directly, lacking a preliminary screening process. This results in very poor robustness for these traditional algorithms, making them unsuitable for identifying fuel cell current distribution data.

[0071] Furthermore, the method for identifying outliers in current measurement data by performing correlation analysis on the two-dimensional current distribution map and the two-dimensional temperature distribution map using a data fusion algorithm includes:

[0072] The two-dimensional current distribution map is normalized using the minimum-maximum normalization algorithm to obtain the normalized current distribution function; the two-dimensional temperature distribution map is normalized using the minimum-maximum normalization algorithm to obtain the normalized temperature distribution function.

[0073] The normalized current distribution function is discretized according to a pre-set sampling step size to obtain a normalized current discrete point matrix; the normalized temperature distribution function is discretized according to a pre-set sampling step size to obtain a normalized temperature discrete point matrix; the normalized current discrete points are renumbered according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix to obtain the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; the normalized temperature discrete points are renumbered according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix to obtain the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points.

[0074] The correlation index between the normalized current sequence and the normalized temperature sequence is calculated and denoted as the benchmark correlation index.

[0075] When the baseline correlation index is less than the preset baseline correlation threshold, an error is reported and the user is prompted to check the correctness of the sensor array installation; the baseline correlation threshold is a value greater than 0 and less than 1; when the baseline correlation index is greater than or equal to the baseline correlation threshold, the first current point to the Nth current point are removed from the normalized current sequence in sequence, and the first temperature point to the Nth temperature point are removed from the normalized temperature sequence in sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence; the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index; i is an integer variable ranging from 1 to N; the values ​​of i are iterated to obtain the first correlation index to the Nth correlation index; based on the first correlation index to the Nth correlation index and the baseline correlation index, abnormal discrete points are screened using a preset screening standard; the screening standard is:

[0076]

[0077] Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

[0078] For example, a minimum-maximum normalization algorithm is used to normalize and map the two-dimensional current distribution map to a range of 0 to 1, resulting in a normalized current distribution function. Similarly, a minimum-maximum normalization algorithm is used to normalize and map the two-dimensional temperature distribution map to a range of 0 to 1, resulting in a normalized temperature distribution function. The normalization algorithm eliminates the dimensional differences between current and temperature data, thus ensuring comparability. The steps of the minimum-maximum normalization algorithm are as follows: subtract the minimum value of the normalized variable from the variable to be normalized, and then divide by the difference between the maximum and minimum values ​​of the normalized variable to obtain its normalized value.

[0079] The sensor array is a 4×4 array, comprising 16 sensors. These 16 sensors are uniformly arranged in a 4×4 array on the cathode side of the membrane electrode under test. The distance between any two adjacent sensors, either horizontally or vertically, is defined as 1 unit length. The preset sampling step size is 0.5 units. The current distribution function is discretized using this sampling step size to obtain a normalized current discrete point matrix. This normalized current discrete point matrix contains 7×7 discrete points. Similarly, the obtained normalized temperature discrete point matrix also contains 7×7 discrete points. The normalized current discrete points are renumbered according to their coordinates in the normalized current discrete point matrix, resulting in current points 1 to 49, which are then combined to obtain a normalized current sequence. Similarly, the normalized temperature discrete points are renumbered according to their coordinates in the normalized temperature discrete point matrix, resulting in temperature points 1 to 49, which are then combined to obtain a normalized temperature sequence. The renumbering order is from left to right and from top to bottom. The normalized temperature discrete points and the normalized current discrete points are numbered in the same order.

[0080] The correlation index between the normalized current series and the normalized temperature series is calculated and denoted as the baseline correlation index. This baseline correlation index reflects the degree of positive correlation between the normalized current series and the normalized temperature series. Theoretically, the normalized current series and the normalized temperature series should exhibit a certain positive correlation. However, in real-world environments, influenced by external factors, the degree of positive correlation between the normalized current series and the normalized temperature series often varies under different testing conditions. Therefore, it is necessary to first calculate the baseline correlation index to determine the overall correlation between current and temperature under the given testing environment.

[0081] When the baseline correlation index is less than the preset baseline correlation threshold, it indicates that there may be an overall problem with the data. An error is then reported, prompting the user to check the correctness of the sensor array installation. The baseline correlation threshold is a value greater than 0 and less than 1. In this embodiment, the baseline correlation threshold is set to 0.3. When the baseline correlation index is greater than or equal to the baseline correlation threshold, it indicates that the overall positive correlation between the normalized current sequence and the normalized temperature sequence is good. Therefore, points in the sequence that do not conform to the overall correlation trend need to be removed. The first current point to the 49th current point are removed sequentially from the normalized current sequence, and the first temperature point to the 49th temperature point are removed sequentially from the normalized temperature sequence, resulting in the first normalized current sequence to the 49th normalized current sequence and the first normalized temperature sequence to the 49th normalized temperature sequence. Taking the first normalized current sequence as an example, the first normalized current sequence refers to the sequence containing 48 data points obtained after removing the first current point from the normalized current sequence. A similar second normalized current sequence refers to the sequence containing 48 data points obtained by removing the second current point from the normalized current sequence. This process continues. The first to 49th normalized temperature sequences are obtained using a similar method. The correlation index between the first normalized current sequence and the first normalized temperature sequence is calculated and denoted as the first correlation index. This process continues to denominate the first to 49th correlation indices. Based on the first to 49th correlation indices and the baseline correlation index, abnormal discrete points are selected. The pre-set screening threshold is set to 0.1. Essentially, the screening criteria are used to identify data points that deviate significantly from the overall trend. The points in the two-dimensional current distribution graph corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data. It is worth noting that, based on the principles of temperature and current measurement, temperature measurement is more accurate than current measurement. Since the purpose of this invention is to obtain the current distribution, this invention uses the correlation between current and temperature to screen for abnormal points in the current measurement data. In this embodiment, a total of 5 abnormal points in the current measurement data were selected.

[0082] Furthermore, the method for calculating the correlation index between the normalized current sequence and the normalized temperature sequence, and denoting it as the benchmark correlation index, includes:

[0083] The correlation index between the normalized current sequence and the normalized temperature sequence is calculated using the first correlation degree calculation formula, and is denoted as the benchmark correlation index; the first correlation degree calculation formula is:

[0084]

[0085] Among them, X j Indicates the j-th current point; T represents the first mean; jRepresents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

[0086] For example, the benchmark correlation index reflects the correlation between the normalized current sequence and the normalized temperature sequence.

[0087] Furthermore, the method for calculating the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence, denoted as the i-th correlation index, includes:

[0088] The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated using the second correlation degree calculation formula, and denoted as the i-th correlation index; the second correlation degree calculation formula is:

[0089]

[0090] Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

[0091] For example, the i-th correlation index reflects the correlation between the i-th normalized current sequence and the i-th normalized temperature sequence.

[0092] Example 2: Based on the same inventive concept, such as Figure 2 As shown, this embodiment also provides a fuel cell membrane electrode current distribution testing system, the system comprising:

[0093] The sensor array arrangement module is used to install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; to acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; to establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and to obtain the spatial coordinates of the dot matrix position.

[0094] The direction cleaning module, connected to the sensor array arrangement module, is used to clean the current measurement data according to the pre-set current direction criteria, remove abnormal direction measurement data, and obtain the first current cleaning data; the two-dimensional current distribution map is obtained by interpolation algorithm based on the spatial coordinates of the dot matrix position and the first current cleaning data.

[0095] The interpolation module, connected to the direction cleaning module, is used to map a two-dimensional temperature distribution map that is spatially aligned with the current two-dimensional distribution map by using an interpolation algorithm based on the spatial coordinates of the dot matrix position and the temperature measurement data at the dot matrix position.

[0096] The reconstruction module, connected to the interpolation module, is used to perform correlation analysis on the two-dimensional current distribution map and the two-dimensional temperature distribution map through a data fusion algorithm to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the interpolation algorithm is used to reconstruct the current distribution map of the fuel cell membrane electrode.

[0097] Furthermore, the directional cleaning module includes:

[0098] The data rejection module is used to determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is rejected; otherwise, it is retained to obtain the first current cleaning data.

[0099] Furthermore, the reconstruction module includes:

[0100] The normalization module is used to normalize the two-dimensional current distribution map using the minimum-maximum normalization algorithm to obtain the normalized current distribution function; and to normalize the two-dimensional temperature distribution map using the minimum-maximum normalization algorithm to obtain the normalized temperature distribution function.

[0101] The sampling module, connected to the normalization module, is used to discretize the normalized current distribution function according to a pre-set sampling step size, obtaining a normalized current discrete point matrix; to discretize the normalized temperature distribution function according to a pre-set sampling step size, obtaining a normalized temperature discrete point matrix; to renumber the normalized current discrete points according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix, obtaining the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; and to renumber the normalized temperature discrete points according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix, obtaining the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points.

[0102] The benchmark correlation index calculation module, connected to the sampling module, is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence, denoted as the benchmark correlation index.

[0103] A filtering module, connected to the benchmark correlation index calculation module, is used to report an error and prompt the user to check the correctness of the sensor array installation when the benchmark correlation index is less than a preset benchmark correlation threshold. The benchmark correlation threshold is a value greater than 0 and less than 1. When the benchmark correlation index is greater than or equal to the benchmark correlation threshold, the first current point to the Nth current point are sequentially removed from the normalized current sequence, and the first temperature point to the Nth temperature point are sequentially removed from the normalized temperature sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence. The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index, where i is an integer variable ranging from 1 to N. The values ​​of i are iterated to obtain the first correlation index to the Nth correlation index. Based on the first correlation index to the Nth correlation index and the benchmark correlation index, abnormal discrete points are filtered using a preset filtering standard. The filtering standard is:

[0104]

[0105] Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

[0106] Furthermore, the benchmark correlation index calculation module includes:

[0107] The first correlation calculation module is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence using the first correlation calculation formula, which is denoted as the benchmark correlation index; the first correlation calculation formula is:

[0108]

[0109] Among them, X j Indicates the j-th current point; T represents the first mean; j Represents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

[0110] Furthermore, the filtering module includes:

[0111] The second correlation calculation module is used to calculate the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence using the second correlation calculation formula, denoted as the i-th correlation index; the second correlation calculation formula is:

[0112]

[0113] Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

[0114] 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.

[0115] 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 current distribution of a fuel cell membrane electrode assembly, characterized in that, The method includes the following steps: S1, Install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and obtain the spatial coordinates of the dot matrix position. S2, the current measurement data is cleaned according to the pre-set current direction criterion, and the measurement data with abnormal direction is removed to obtain the first current cleaned data; the two-dimensional current distribution map is obtained by interpolation algorithm based on the spatial coordinates of the dot matrix position and the first current cleaned data. S3. Based on the spatial coordinates of the lattice positions and the temperature measurement data at the lattice positions, an interpolation algorithm is used to map and obtain a two-dimensional temperature distribution map that is spatially aligned with the two-dimensional current distribution map. S4. Through data fusion algorithm, correlation analysis is performed on the two-dimensional current distribution map and the two-dimensional temperature distribution map to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the current distribution map of the fuel cell membrane electrode is reconstructed using an interpolation algorithm.

2. The method for testing the membrane electrode current distribution of a fuel cell as described in claim 1, characterized in that, The method for cleaning the current measurement data according to a pre-set current direction criterion and removing measurement data with abnormal direction to obtain the first current cleaned data includes: Determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is discarded; otherwise, it is retained, and the first current cleaning data is obtained.

3. The method for testing the membrane electrode current distribution of a fuel cell as described in claim 2, characterized in that, The method for identifying outliers in current measurement data by performing correlation analysis on two-dimensional current distribution maps and two-dimensional temperature distribution maps using a data fusion algorithm includes: The minimum-maximum normalization algorithm is used to normalize the two-dimensional current distribution map to obtain the normalized current distribution function; the minimum-maximum normalization algorithm is also used to normalize the two-dimensional temperature distribution map to obtain the normalized temperature distribution function. The normalized current distribution function is discretized according to a pre-set sampling step size to obtain a normalized current discrete point matrix; the normalized temperature distribution function is discretized according to a pre-set sampling step size to obtain a normalized temperature discrete point matrix; the normalized current discrete points are renumbered according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix to obtain the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; the normalized temperature discrete points are renumbered according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix to obtain the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points. The correlation index between the normalized current sequence and the normalized temperature sequence is calculated and denoted as the benchmark correlation index. When the baseline correlation index is less than the preset baseline correlation threshold, an error is reported and the user is prompted to check the correctness of the sensor array installation; the baseline correlation threshold is a value greater than 0 and less than 1; when the baseline correlation index is greater than or equal to the baseline correlation threshold, the first current point to the Nth current point are removed from the normalized current sequence in sequence, and the first temperature point to the Nth temperature point are removed from the normalized temperature sequence in sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence; the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index; i is an integer variable ranging from 1 to N; the values ​​of i are iterated to obtain the first correlation index to the Nth correlation index; based on the first correlation index to the Nth correlation index and the baseline correlation index, abnormal discrete points are screened using a preset screening standard; the screening standard is: Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

4. The method for testing the membrane electrode current distribution of a fuel cell as described in claim 3, characterized in that, The method for calculating the correlation index between the normalized current sequence and the normalized temperature sequence, and denoting it as the benchmark correlation index, includes: The correlation index between the normalized current sequence and the normalized temperature sequence is calculated using the first correlation degree calculation formula and is denoted as the benchmark correlation index; the first correlation degree calculation formula is: Among them, X j Indicates the j-th current point; T represents the first mean; j Represents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

5. The method for testing the membrane electrode current distribution of a fuel cell as described in claim 4, characterized in that, The method for calculating the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence, denoted as the i-th correlation index, includes: The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated using the second correlation degree calculation formula, and denoted as the i-th correlation index; the second correlation degree calculation formula is as follows: Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

6. A fuel cell membrane electrode current distribution testing system, characterized in that, The system includes: The sensor array arrangement module is used to install a sensor array at a pre-set dot matrix position on the cathode side of the membrane electrode under test in the fuel cell stack; acquire current measurement data and temperature measurement data at the dot matrix position through the sensor array; establish a two-dimensional coordinate system with the geometric center of the cathode side of the membrane electrode under test as the origin; and obtain the spatial coordinates of the dot matrix position. The orientation cleaning module, connected to the sensor array arrangement module, is used to clean the current measurement data according to a pre-set current orientation criterion, remove abnormal orientation measurement data, and obtain the first current cleaning data; the spatial coordinates of the dot matrix position and the first current cleaning data are used to map the current two-dimensional distribution map using an interpolation algorithm; The interpolation module, connected to the direction cleaning module, is used to map a two-dimensional temperature distribution map that is spatially aligned with the current two-dimensional distribution map by using an interpolation algorithm based on the spatial coordinates of the dot matrix position and the temperature measurement data at the dot matrix position. The reconstruction module, connected to the interpolation module, is used to perform correlation analysis on the two-dimensional current distribution map and the two-dimensional temperature distribution map through a data fusion algorithm to identify abnormal points in the current measurement data. After removing the abnormal points in the current measurement data from the two-dimensional current distribution map, the interpolation algorithm is used to reconstruct the current distribution map of the fuel cell membrane electrode.

7. The fuel cell membrane electrode current distribution testing system as described in claim 6, characterized in that, The directional cleaning module includes: The data rejection module is used to determine the direction of the current measurement data at the dot matrix position; if the direction of the current measurement data is from the back of the sensor to the front of the sensor, it is rejected; otherwise, it is retained to obtain the first current cleaning data.

8. The fuel cell membrane electrode current distribution testing system as described in claim 7, characterized in that, The reconstruction module includes: The normalization module is used to normalize the two-dimensional current distribution map using the minimum-maximum normalization algorithm to obtain the normalized current distribution function; and to normalize the two-dimensional temperature distribution map using the minimum-maximum normalization algorithm to obtain the normalized temperature distribution function. The sampling module, connected to the normalization module, is used to discretize the normalized current distribution function according to a pre-set sampling step size, obtaining a normalized current discrete point matrix; to discretize the normalized temperature distribution function according to a pre-set sampling step size, obtaining a normalized temperature discrete point matrix; to renumber the normalized current discrete points according to the coordinates of each normalized current discrete point in the normalized current discrete point matrix, obtaining the first current point to the Nth current point, which are then combined to obtain a normalized current sequence; and to renumber the normalized temperature discrete points according to the coordinates of each normalized temperature discrete point in the normalized temperature discrete point matrix, obtaining the first temperature point to the Nth temperature point, which are then combined to obtain a normalized temperature sequence; where N represents the number of normalized current discrete points and also the number of normalized temperature discrete points. The benchmark correlation index calculation module, connected to the sampling module, is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence, denoted as the benchmark correlation index. A filtering module, connected to the benchmark correlation index calculation module, is used to report an error and prompt the user to check the correctness of the sensor array installation when the benchmark correlation index is less than a preset benchmark correlation threshold. The benchmark correlation threshold is a value greater than 0 and less than 1. When the benchmark correlation index is greater than or equal to the benchmark correlation threshold, the first current point to the Nth current point are sequentially removed from the normalized current sequence, and the first temperature point to the Nth temperature point are sequentially removed from the normalized temperature sequence, resulting in the first normalized current sequence to the Nth normalized current sequence and the first normalized temperature sequence to the Nth normalized temperature sequence. The correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence is calculated and denoted as the i-th correlation index, where i is an integer variable ranging from 1 to N. The values ​​of i are iterated to obtain the first correlation index to the Nth correlation index. Based on the first correlation index to the Nth correlation index and the benchmark correlation index, abnormal discrete points are filtered using a preset filtering standard. The filtering standard is: Among them, R b R represents the benchmark correlation index. i Let represent the i-th correlation index, and α represent the pre-set screening threshold; the points in the two-dimensional current distribution map corresponding to the abnormal discrete points are recorded as abnormal points in the current measurement data.

9. The fuel cell membrane electrode current distribution testing system as described in claim 8, characterized in that, The benchmark correlation index calculation module includes: The first correlation calculation module is used to calculate the correlation index between the normalized current sequence and the normalized temperature sequence using the first correlation calculation formula, which is denoted as the benchmark correlation index; the first correlation calculation formula is: Among them, X j Indicates the j-th current point; T represents the first mean; j Represents the j-th temperature point; denoted as the second mean; j is an integer variable taking values ​​from 1 to N.

10. The fuel cell membrane electrode current distribution testing system as described in claim 9, characterized in that, The filtering module includes: The second correlation calculation module is used to calculate the correlation index between the i-th normalized current sequence and the i-th normalized temperature sequence using the second correlation calculation formula, denoted as the i-th correlation index; the second correlation calculation formula is: Among them, X (i,k) This represents the value of the k-th point in the i-th normalized current sequence; T represents the average value of the i-th current; (i,k) This represents the value of the k-th point in the i-th normalized temperature sequence; Let represent the average temperature of the i-th temperature; k is an integer variable taking values ​​from 1 to N-1.

Images