CN113917352B - Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic - Google Patents
Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic Download PDFInfo
- Publication number
- CN113917352B CN113917352B CN202111196890.4A CN202111196890A CN113917352B CN 113917352 B CN113917352 B CN 113917352B CN 202111196890 A CN202111196890 A CN 202111196890A CN 113917352 B CN113917352 B CN 113917352B
- Authority
- CN
- China
- Prior art keywords
- aging
- electrochemical
- impedance
- fuel cell
- preset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
Abstract
The invention discloses an online aging diagnosis method for a fuel cell catalyst layer based on impedance aging characteristics. The method comprises the following steps: measuring electrochemical impedance spectrum and electrochemical surface area parameters of the fuel cell at different aging stages; calculating two-point impedance aging characteristics corresponding to different frequency combinations in different frequency intervals at different aging stages of the current fuel cell; repeating the steps to obtain electrochemical surface area parameters and two-point impedance aging characteristics of each fuel cell at different aging stages; determining the optimal two-point impedance aging characteristics in different frequency intervals; forming a training set and training a model to obtain a trained aging diagnosis regression model; during online diagnosis, the optimal two-point impedance aging characteristic of the fuel cell to be diagnosed is measured and calculated, and the predicted electrochemical surface area parameter is obtained after diagnosis, so that the aging state of the catalyst layer is judged. The invention realizes accurate aging diagnosis of the fuel cell and is beneficial to more reliable and durable operation.
Description
Technical Field
The invention belongs to an on-line aging diagnosis method for a fuel cell catalyst layer in the field of fuel cell application, and particularly relates to an on-line aging diagnosis method for a fuel cell catalyst layer based on two-point impedance aging characteristics.
Background
The aging of the catalyst layer of the proton exchange membrane fuel cell is a complex and strong nonlinear process, and relates to complex factors such as multiple mechanisms, multiple physical domains, multiple space-time scales, multiple working conditions, multiple couplings and the like, so that the accurate and rapid estimation of the aging state of the catalyst layer becomes a great challenge. In addition, in the actual use process, the aging recovery phenomenon of the catalyst layer of the proton exchange membrane fuel cell after shutdown and restart occurs due to factors such as water management and the like, so that the aging state estimation difficulty of the catalyst layer is further increased. At present, the electrochemical surface area parameter capable of representing the aging state of a catalyst layer of a proton exchange membrane fuel cell needs to be accurately measured by adopting a cyclic voltammetry method, and the method not only needs specific operating conditions, but also needs a large amount of measuring time, thereby bringing great difficulty to the online diagnosis of the aging state of the catalyst layer. Therefore, the research on the online aging diagnosis method of the catalyst layer of the proton exchange membrane fuel cell is of great significance.
Electrochemical impedance spectroscopy is a means for effectively detecting the internal conditions of fuel cells, and is widely applied to the fields of fuel cell detection, research and development and application. However, the electrochemical impedance spectroscopy measurement time is also long, stable operation conditions are required, and the online implementation of the electrochemical impedance spectroscopy measurement method is difficult to reflect the aging state of the catalytic layer of the fuel cell, so that the electrochemical impedance spectroscopy measurement method has certain limitations. Therefore, the electrochemical impedance spectrum technology is applied to the online monitoring of the aging state of the catalyst layer of the fuel cell by adopting relevant measures, and the method has great significance for improving the reliability, the safety and the durability of the fuel cell.
Disclosure of Invention
In order to solve the defects of the prior art, the invention provides an online aging diagnosis method for a fuel cell catalyst layer based on impedance aging characteristics.
The scheme adopted by the invention is as follows:
the invention comprises the following steps:
1) measuring electrochemical impedance spectrums and electrochemical surface area parameters of the brand-new proton exchange membrane fuel cell at different aging stages;
2) respectively calculating two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of the current fuel cell in preset low-frequency intervals and preset middle-high frequency intervals of different aging stages according to the electrochemical impedance spectrums of the fuel cell in different aging stages;
3) repeating the steps 1) -2) to process each fuel cell to obtain two-point impedance aging characteristics and corresponding electrochemical surface area parameters corresponding to all electrochemical impedance spectrum frequency combinations of the preset low-frequency interval and the preset middle-high frequency interval of all the fuel cells at different aging stages;
4) selecting an optimal electrochemical impedance spectrum frequency combination in a preset low-frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset low-frequency intervals at different aging stages and corresponding electrochemical surface area parameters, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency interval as the optimal two-point impedance aging characteristics of the preset low-frequency interval;
5) selecting an optimal electrochemical impedance spectrum frequency combination in a preset middle-high frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset middle-high frequency interval at different aging stages and corresponding electrochemical surface area parameters, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency interval as the optimal two-point impedance aging characteristics of the preset middle-high frequency interval;
6) the optimal two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency interval and the preset middle-high frequency interval of all the fuel cells at different aging stages and the corresponding electrochemical surface area parameters form a training set, and the fuel cell aging diagnosis regression model is trained on the basis of the training set to obtain a trained fuel cell aging diagnosis regression model;
7) during online diagnosis, only impedance values corresponding to the optimal electrochemical impedance spectrum frequency combination of a preset low-frequency interval and a preset middle-high frequency interval measured by the fuel cell to be diagnosed in the current aging stage are collected, optimal two-point impedance aging characteristics of the fuel cell to be diagnosed in the preset low-frequency interval and the preset middle-high frequency interval are calculated and input into a trained fuel cell aging diagnosis regression model for diagnosis, a predicted electrochemical surface area parameter of the current fuel cell to be diagnosed is output, and the aging state of a catalyst layer in the current fuel cell to be diagnosed is judged according to the predicted electrochemical surface area parameter of the fuel cell.
The step 2) is specifically as follows:
in a preset low-frequency interval and a preset middle-high frequency interval, two different electrochemical impedance spectrum frequencies in the electrochemical impedance spectrum of the current fuel cell at each aging stage are used as an electrochemical impedance spectrum frequency combination, the difference value of the impedance imaginary part of the higher electrochemical impedance spectrum frequency and the impedance imaginary part of the lower electrochemical impedance spectrum frequency in the electrochemical impedance spectrum frequency combination is calculated and is used as a two-point impedance aging characteristic, all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval are traversed respectively, two-point impedance aging characteristics corresponding to all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at the current aging stage are obtained, different aging stages of the current fuel cell are traversed, and all the electrochemical impedance spectrum frequency combination pairs in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at different aging stages are obtained Two points of resistance aging characteristics.
The step 4) is specifically as follows:
calculating correlation coefficients corresponding to the two-point impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and all electrochemical impedance spectrum frequency combinations in a preset low-frequency range according to the two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations in all aging stages of all fuel cells and corresponding electrochemical surface area parameters, traversing and calculating to obtain the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations in the preset low-frequency range, forming a correlation coefficient matrix by the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations, taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the largest absolute value in the correlation coefficient matrix as the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency range, and then taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristic in the preset low-frequency range And (5) carrying out characterization.
The step 5) is specifically as follows:
calculating correlation coefficients between the two-point impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and the corresponding electrochemical surface area parameters according to the two-point impedance aging characteristics and the corresponding electrochemical surface area parameters of all the fuel cells in the preset middle-high frequency intervals of all the aging stages, traversing and calculating to obtain the correlation coefficients corresponding to all the electrochemical impedance spectrum frequency combinations in the preset middle-high frequency intervals, forming a correlation coefficient matrix by the correlation coefficients corresponding to all the electrochemical impedance spectrum frequency combinations, and taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the maximum absolute value in the correlation coefficient matrix as the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency intervals, and then, taking the two-point impedance aging characteristic corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristic of a preset middle-high frequency interval.
The correlation coefficient is a pearson correlation coefficient, and is specifically calculated by the following formula:
wherein ρ X，Y The pearson correlation coefficient between the two-point impedance aging characteristic corresponding to the same electrochemical impedance spectrum frequency combination and the corresponding electrochemical surface area parameter in the preset low frequency intervals of all the aging stages of all the fuel cells is represented by X, the set of the two-point impedance aging characteristic corresponding to the same electrochemical impedance spectrum frequency combination in the preset low frequency intervals of all the aging stages of all the fuel cells is represented by Y, the set of the electrochemical surface area parameters of the fuel cells in all the respective aging stages of all the fuel cells is represented by E (), and the desired operation is taken.
The correlation coefficient is a pearson correlation coefficient, and is specifically calculated by the following formula:
wherein ρ X，Y Representing correlation coefficients between two points of impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in preset middle-high frequency intervals of all aging stages of all fuel cells and electrochemical surface area parameters corresponding to the aging stages, X representing a set of two points of impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in the preset middle-high frequency intervals of all aging stages of all fuel cells, Y representing a set of electrochemical surface area parameters of all fuel cells in all aging stages of all fuel cells, and E () representing an expected operation.
The models of all the fuel cells in the step 6) are the same.
The preset low-frequency interval and the preset middle-high frequency interval in the step 7) are consistent with the preset low-frequency interval and the preset middle-high frequency interval in the step 1).
The beneficial effects of the invention are:
the invention solves the problem that the on-line aging diagnosis of the catalyst layer of the fuel cell is difficult in practical application. The two-point impedance aging characteristic based on the electrochemical impedance spectrum is applied to the online aging diagnosis of the fuel cell catalyst layer, the two-point impedance aging characteristic of the low-frequency and preset middle-high frequency interval can be respectively calculated only by respectively measuring impedance imaginary parts corresponding to the low-frequency and preset middle-high frequency interval within different aging stages of the fuel cell, and then the aging state of the fuel cell catalyst layer is accurately diagnosed, the data storage burden, the calculation burden and the cost burden are reduced, the online aging diagnosis of the proton exchange membrane fuel cell catalyst layer in an actual application scene is more suitable, and the more reliable and durable operation of the fuel cell is facilitated. The invention is based on the principle that the aging of the cathode catalyst layer of the fuel cell can cause the obvious increase of the oxygen transmission resistance of the cathode, thereby causing the increase of the concentration loss, and reflecting the shape change of a low-frequency interval curve on an electrochemical impedance spectrum. Meanwhile, the aging of the cathode catalyst layer of the fuel cell can also cause the obvious reduction of the electrochemical surface area parameter, thereby causing the increase of the activation loss, which is reflected in the shape change of a middle-high frequency interval curve on an electrochemical impedance spectrum.
Drawings
FIG. 1 is an overall flow diagram of the present invention.
Fig. 2 is a schematic diagram of electrochemical impedance spectra measured at different stages of aging of a fuel cell in an embodiment of the invention.
Fig. 3 is a schematic diagram of impedance points corresponding to the frequency combination of the optimal electrochemical impedance spectrum selected by the fuel cell in the low frequency and the preset middle-high frequency ranges respectively according to the embodiment of the present invention.
Fig. 4 is a graph showing the results of training a regression model for diagnosing catalyst layer degradation of a fuel cell in an embodiment of the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and the specific embodiments.
As shown in fig. 1, the present invention comprises the steps of:
1) and measuring electrochemical impedance spectrums and electrochemical surface area parameters of the brand-new proton exchange membrane fuel cell at different aging stages, as shown in figure 2. In specific implementation, the aging test cycle working condition is a voltage cycle working condition with the total duration of 40 seconds, wherein the voltage cycle working condition is specifically that 0.6V voltage lasts for 10 seconds and 0.9V voltage lasts for 30 seconds;
2) respectively calculating two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of the current fuel cell in preset low-frequency intervals and preset middle-high frequency intervals of different aging stages according to the electrochemical impedance spectrums of the fuel cell in different aging stages;
the step 2) is specifically as follows:
in a preset low-frequency interval and a preset middle-high frequency interval, two different electrochemical impedance spectrum frequencies in the electrochemical impedance spectrum of the current fuel cell at each aging stage are used as an electrochemical impedance spectrum frequency combination, the difference value of the impedance imaginary part of the higher electrochemical impedance spectrum frequency and the impedance imaginary part of the lower electrochemical impedance spectrum frequency in the electrochemical impedance spectrum frequency combination is calculated and is used as a two-point impedance aging characteristic, all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval are traversed respectively, the two-point impedance aging characteristics corresponding to all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at the current aging stage are obtained, the different aging stages of the current fuel cell are traversed, and the pairs of all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at the different aging stages are obtained Two points of impedance aging characteristics are required.
In a specific implementation, the preset electrochemical impedance spectrum frequency range is preferably 0.010Hz-3981.100Hz, and the measured electrochemical impedance spectrum frequency is 57 in total, as shown in Table 1. The preset low-frequency interval is set to be less than 100Hz, the preset middle-high frequency interval is set to be more than or equal to 100Hz, the higher the accuracy of the measured electrochemical impedance spectrum frequency in the allowable range of the measuring equipment is, the better the accuracy of the measured electrochemical impedance spectrum frequency is, the more the electrochemical impedance spectrum frequency combination is, and the more the corresponding two-point impedance aging characteristics are.
TABLE 1 Shen chemical yang antibiotic spectrum frequency (Hz) of Shen chemical yang antibiotic spectrum measurement
| 3981.1 | 3162.3 | 2511.9 | 1995.3 | 1584.9 | 1258.9 | 1000 | 794.33 | 630.96 | 501.19 |
| 398.11 | 316.23 | 251.19 | 199.53 | 158.49 | 125.89 | 100 | 79.433 | 63.096 | 50.119 |
| 39.811 | 31.623 | 25.119 | 19.953 | 15.849 | 12.589 | 10 | 7.9433 | 6.3096 | 5.0119 |
| 3.9811 | 3.1623 | 2.5119 | 1.9953 | 1.5849 | 1.2589 | 1 | 0.79433 | 0.63096 | 0.50119 |
| 0.39811 | 0.31623 | 0.25119 | 0.19953 | 0.15849 | 0.12589 | 0.1 | 0.079433 | 0.063096 | 0.050119 |
| 0.039811 | 0.031623 | 0.025119 | 0.019953 | 0.015849 | 0.012589 | 0.01 |
3) Repeating the steps 1) -2) to process each fuel cell to obtain two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset low-frequency interval and the preset middle-high frequency interval of different aging stages and electrochemical surface area parameters corresponding to different aging stages;
4) selecting an optimal electrochemical impedance spectrum frequency combination in a preset low-frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset low-frequency intervals at different aging stages and electrochemical surface area parameters corresponding to different aging stages, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency interval as the optimal two-point impedance aging characteristics of the preset low-frequency interval;
the step 4) is specifically as follows:
according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in preset low-frequency intervals of different aging stages and electrochemical surface area parameters corresponding to different aging stages, calculating correlation coefficients between two point impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and electrochemical surface area parameters corresponding to different aging stages, obtaining the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations in a preset low-frequency interval through traversal calculation, forming a correlation coefficient matrix by the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations, and taking the correlation coefficient matrix as a compact representation of the correlation between the two point impedance aging characteristics corresponding to different electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the electrochemical surface area parameters of the fuel cells, as shown in table 2.
TABLE 2 local schematic table of correlation coefficient matrix of preset low frequency interval
And taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the largest absolute value in the correlation number matrix as the optimal electrochemical impedance spectrum frequency combination of the preset low-frequency interval, and then taking the two-point impedance aging characteristic corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristic of the preset low-frequency interval, wherein the electrochemical impedance spectrum frequencies where two squares are located represent the optimal electrochemical impedance spectrum frequency combination of the preset low-frequency interval, as shown in fig. 3. The two-point impedance aging characteristic of the fuel cell in each aging stage under the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency interval is the electrochemical surface area parameter of the fuel cell in the current aging stage. As shown in table 2, the row number and the column number of the correlation coefficient in the correlation coefficient matrix respectively represent two electrochemical impedance spectrum frequencies in the electrochemical impedance spectrum frequency combination corresponding to the two points of impedance aging characteristics, and both the row and the column of the correlation coefficient matrix represent the preset electrochemical impedance spectrum frequency range.
The correlation coefficient in the step 4) is a pearson correlation coefficient, and is specifically calculated by the following formula:
where ρ is X，Y The Pearson correlation coefficient between the impedance aging characteristics of two points corresponding to the same electrochemical impedance spectrum frequency combination in the preset low frequency interval representing all the aging stages of all the fuel cells and the corresponding electrochemical surface area parameters, and X represents all the old electrochemical impedance spectrum frequency combinations of all the fuel cellsAnd E () represents the expected operation.
5) Selecting an optimal electrochemical impedance spectrum frequency combination in a preset middle-high frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset middle-high frequency interval at different aging stages and electrochemical surface area parameters corresponding to different aging stages, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency interval as the optimal two-point impedance aging characteristics of the preset middle-high frequency interval;
the step 5) is specifically as follows:
calculating correlation coefficients between the two-point impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and the electrochemical surface area parameters corresponding to different aging stages according to the two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations in the preset middle-high frequency intervals of all the fuel cells and the electrochemical surface area parameters corresponding to different aging stages, traversing and calculating to obtain the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations in the preset middle-high frequency intervals, forming a correlation coefficient matrix by the correlation coefficients corresponding to all the electrochemical impedance spectrum frequency combinations, wherein the correlation coefficient matrix is used as a compact representation of the correlation between the two-point impedance aging characteristics corresponding to different electrochemical impedance spectrum frequency combinations in the preset middle-high frequency intervals and the electrochemical surface area parameters of the fuel cells, as shown in table 3.
TABLE 3 local schematic table of correlation coefficient matrix in preset middle and high frequency interval
And taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the largest absolute value in the correlation number matrix as the optimal electrochemical impedance spectrum frequency combination of the preset middle-high frequency interval, and then taking the two-point impedance aging characteristic corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristic of the preset middle-high frequency interval, wherein the electrochemical impedance spectrum frequencies corresponding to the two circular impedance points represent the optimal electrochemical impedance spectrum frequency combination of the preset middle-high frequency interval, as shown in fig. 3. The preset middle-high frequency interval optimal two-point impedance aging characteristic is specifically the two-point impedance aging characteristic of all the fuel cells under the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency interval, and a label of the two-point impedance aging characteristic of the fuel cells under the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency interval in each aging stage is the electrochemical surface area parameter of the fuel cells in the current aging stage. As shown in table 3, the row number and the column number of the correlation coefficient in the correlation coefficient matrix respectively represent two electrochemical impedance spectrum frequencies in the electrochemical impedance spectrum frequency combination corresponding to the two points of impedance aging characteristics, and both the row and the column of the correlation coefficient matrix represent the preset electrochemical impedance spectrum frequency range.
The correlation coefficient in the step 5) is a pearson correlation coefficient, and is specifically calculated by the following formula:
where ρ is X，Y Representing correlation coefficients between two points of impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in preset middle-high frequency intervals of all aging stages of all fuel cells and electrochemical surface area parameters corresponding to the aging stages, X representing a set of two points of impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in the preset middle-high frequency intervals of all aging stages of all fuel cells, Y representing a set of electrochemical surface area parameters of all fuel cells in all aging stages of all fuel cells, and E () representing an expected operation.
6) The method comprises the steps that optimal two-point impedance aging characteristics corresponding to optimal electrochemical impedance spectrum frequency combinations in preset low-frequency intervals and preset middle-high frequency intervals of all fuel cells at different aging stages and electrochemical surface area parameters corresponding to different aging stages form a training set, and a fuel cell aging diagnosis regression model is trained on the basis of the training set to obtain a trained fuel cell aging diagnosis regression model; fig. 4 is a graph showing the result of training the regression model for the catalyst layer aging diagnosis of the fuel cell in the embodiment of the present invention. In this example, the electrochemical impedance spectroscopy measurements and the corresponding electrochemical surface area measurements of two fuel cells of the same type at different respective stages of aging were used together. In fig. 4, the triangular symbols represent the values of the electrochemical surface area actually measured and estimated for the first fuel cell at different stages of aging, and the square symbols represent the values of the electrochemical surface area actually measured and estimated for the second fuel cell at different stages of aging. The models of all the fuel cells in the step 6) are the same. In this embodiment, an adaptive fuzzy neural model is selected.
7) During on-line diagnosis, only impedance values corresponding to the optimal electrochemical impedance spectrum frequency combination of a preset low-frequency range and a preset middle-high frequency range measured by the fuel cell to be diagnosed in the current aging stage are collected, optimal two-point impedance aging characteristics of the fuel cell to be diagnosed in the preset low-frequency range and the preset middle-high frequency range are calculated and input into a trained fuel cell aging diagnosis regression model for diagnosis, a predicted electrochemical surface area parameter of the current fuel cell to be diagnosed is obtained through output, and the aging state of a catalyst layer in the current fuel cell to be diagnosed is judged according to the predicted electrochemical surface area parameter of the fuel cell.
The preset low frequency interval and the preset middle-high frequency interval in the step 7) are consistent with the preset low frequency interval and the preset middle-high frequency interval in the step 1).
Claims (7)
1. An online fuel cell catalyst layer aging diagnosis method based on impedance aging characteristics is characterized by comprising the following steps:
1) measuring electrochemical impedance spectrums and electrochemical surface area parameters of the brand-new proton exchange membrane fuel cell at different aging stages;
2) respectively calculating two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of the current fuel cell in preset low-frequency intervals and preset middle-high frequency intervals of different aging stages according to the electrochemical impedance spectrums of the fuel cell in different aging stages;
3) repeating the steps 1) -2) to process each fuel cell to obtain two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset low-frequency interval and the preset middle-high frequency interval of different aging stages and the corresponding electrochemical surface area parameters;
4) selecting an optimal electrochemical impedance spectrum frequency combination in a preset low-frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset low-frequency intervals at different aging stages and corresponding electrochemical surface area parameters, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency interval as the optimal two-point impedance aging characteristics of the preset low-frequency interval;
5) selecting an optimal electrochemical impedance spectrum frequency combination in a preset middle-high frequency interval according to two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations of all fuel cells in the preset middle-high frequency interval at different aging stages and corresponding electrochemical surface area parameters, and taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency interval as the optimal two-point impedance aging characteristics of the preset middle-high frequency interval;
6) the method comprises the steps that optimal two-point impedance aging characteristics corresponding to optimal electrochemical impedance spectrum frequency combinations in preset low-frequency intervals, preset middle-high frequency intervals and optimal electrochemical surface area parameters of all fuel cells at different aging stages form a training set, and a fuel cell aging diagnosis regression model is trained on the basis of the training set to obtain a trained fuel cell aging diagnosis regression model;
7) during online diagnosis, only acquiring impedance values corresponding to the optimal electrochemical impedance spectrum frequency combination of a preset low-frequency interval and a preset middle-high frequency interval measured by the fuel cell to be diagnosed at the current aging stage, calculating optimal two-point impedance aging characteristics of the fuel cell to be diagnosed in the preset low-frequency interval and the preset middle-high frequency interval, inputting the optimal two-point impedance aging characteristics into a trained fuel cell aging diagnosis regression model for diagnosis, outputting to obtain a predicted electrochemical surface area parameter of the current fuel cell to be diagnosed, and judging the aging state of a catalyst layer in the current fuel cell to be diagnosed according to the predicted electrochemical surface area parameter of the fuel cell;
the step 5) is specifically as follows:
calculating correlation coefficients between the two points of impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and the corresponding electrochemical surface area parameters according to the two points of impedance aging characteristics and the corresponding electrochemical surface area parameters of all the fuel cells in the preset middle-high frequency intervals of all the aging stages, traversing and calculating to obtain the correlation coefficients corresponding to all the electrochemical impedance spectrum frequency combinations in the preset middle-high frequency intervals, forming a correlation coefficient matrix by the correlation coefficients corresponding to all the electrochemical impedance spectrum frequency combinations, and taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the maximum absolute value in the correlation coefficient matrix as the optimal electrochemical impedance spectrum frequency combination in the preset middle-high frequency intervals, and then, taking the two-point impedance aging characteristic corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristic of a preset middle-high frequency interval.
2. The method for diagnosing the online aging of the catalytic layer of the fuel cell based on the impedance aging characteristic according to claim 1, wherein the step 2) is specifically as follows:
in a preset low-frequency interval and a preset middle-high frequency interval, two different electrochemical impedance spectrum frequencies in the electrochemical impedance spectrum of the current fuel cell at each aging stage are used as an electrochemical impedance spectrum frequency combination, the difference value of the impedance imaginary part of the higher electrochemical impedance spectrum frequency and the impedance imaginary part of the lower electrochemical impedance spectrum frequency in the electrochemical impedance spectrum frequency combination is calculated and is used as a two-point impedance aging characteristic, all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval are traversed respectively, the two-point impedance aging characteristics corresponding to all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at the current aging stage are obtained, the different aging stages of the current fuel cell are traversed, and the pairs of all the electrochemical impedance spectrum frequency combinations in the preset low-frequency interval and the preset middle-high frequency interval of the current fuel cell at the different aging stages are obtained Two points of impedance aging characteristics are required.
3. The method for diagnosing the online aging of the catalytic layer of the fuel cell based on the impedance aging characteristic according to claim 1, wherein the step 4) is specifically as follows:
calculating correlation coefficients corresponding to the two-point impedance aging characteristics of the same electrochemical impedance spectrum frequency combination in all aging stages of all fuel cells and all electrochemical impedance spectrum frequency combinations in a preset low-frequency range according to the two-point impedance aging characteristics corresponding to all electrochemical impedance spectrum frequency combinations in all aging stages of all fuel cells and corresponding electrochemical surface area parameters, traversing and calculating to obtain the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations in the preset low-frequency range, forming a correlation coefficient matrix by the correlation coefficients corresponding to all electrochemical impedance spectrum frequency combinations, taking the electrochemical impedance spectrum frequency combination corresponding to the correlation coefficient with the largest absolute value in the correlation coefficient matrix as the optimal electrochemical impedance spectrum frequency combination in the preset low-frequency range, and then taking the two-point impedance aging characteristics corresponding to the optimal electrochemical impedance spectrum frequency combination as the optimal two-point impedance aging characteristics in the preset low-frequency range And (5) carrying out characterization.
4. The fuel cell catalyst layer online aging diagnosis method based on the impedance aging characteristic as set forth in claim 3, wherein the correlation coefficient is a Pearson correlation coefficient, and is specifically calculated by the following formula:
where ρ is X,Y The pearson correlation coefficient between the two-point impedance aging characteristic corresponding to the same electrochemical impedance spectrum frequency combination and the corresponding electrochemical surface area parameter in the preset low frequency intervals of all the aging stages of all the fuel cells is represented by X, the set of the two-point impedance aging characteristic corresponding to the same electrochemical impedance spectrum frequency combination in the preset low frequency intervals of all the aging stages of all the fuel cells is represented by Y, the set of the electrochemical surface area parameters of the fuel cells in all the respective aging stages of all the fuel cells is represented by E (), and the desired operation is taken.
5. The fuel cell catalyst layer online aging diagnosis method based on the impedance aging characteristic as set forth in claim 1, wherein the correlation coefficient is a pearson correlation coefficient, and is specifically calculated by the following formula:
where ρ is X,Y Representing a correlation coefficient between two-point impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in preset middle-high frequency intervals of all aging stages of all the fuel cells and electrochemical surface area parameters corresponding to the aging stages, X representing a set of two-point impedance aging characteristics corresponding to the same electrochemical impedance spectrum frequency combination in the preset middle-high frequency intervals of all the aging stages of all the fuel cells, Y representing a set of the electrochemical surface area parameters of all the fuel cells in all the aging stages, and E () representing an expected operation.
6. The method for diagnosing the online aging of the catalytic layer of the fuel cell based on the impedance aging characteristic as claimed in claim 1, wherein all the fuel cells in the step 6) are of the same type.
7. The method for diagnosing online aging of the catalyst layer of the fuel cell based on impedance aging characteristics as claimed in claim 1, wherein the preset low frequency range and the preset middle-high frequency range in step 7) are consistent with the preset low frequency range and the preset middle-high frequency range in step 1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111196890.4A CN113917352B (en) | 2021-10-14 | 2021-10-14 | Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111196890.4A CN113917352B (en) | 2021-10-14 | 2021-10-14 | Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113917352A CN113917352A (en) | 2022-01-11 |
| CN113917352B true CN113917352B (en) | 2022-07-26 |
Family
ID=79240337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111196890.4A Active CN113917352B (en) | 2021-10-14 | 2021-10-14 | Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113917352B (en) |
Family Cites Families (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6519539B1 (en) * | 2000-09-29 | 2003-02-11 | Hydrogenics Corporation | Measurement of fuel cell impedance |
| CN100595964C (en) * | 2001-07-27 | 2010-03-24 | 麻省理工学院 | Battery structures, self-organizing structures and related methods |
| KR100388314B1 (en) * | 2001-09-03 | 2003-06-25 | 금호석유화학 주식회사 | method to group single cells of power sources to build optimal packs using parameters obtained by analysis of impedance spectrum |
| JP4595367B2 (en) * | 2004-03-30 | 2010-12-08 | トヨタ自動車株式会社 | Deterioration diagnosis method and apparatus for fuel cell |
| DE102009000337A1 (en) * | 2009-01-21 | 2010-07-22 | Robert Bosch Gmbh | Method for determining an aging state of a battery cell by means of impedance spectroscopy |
| AT10763U3 (en) * | 2009-05-12 | 2010-08-15 | Avl List Gmbh | METHOD AND TEST STATION FOR CHECKING HYBRID DRIVE SYSTEMS OR PARTIAL COMPONENTS THEREOF |
| FR2965360B1 (en) * | 2010-09-27 | 2013-03-29 | IFP Energies Nouvelles | METHOD FOR IN SITU DIAGNOSIS OF BATTERIES BY SPECTROSCOPY OF ELECTROCHEMICAL IMPEDANCE |
| JP2012190712A (en) * | 2011-03-11 | 2012-10-04 | Hitachi Maxell Energy Ltd | Method of inspecting membrane-electrode assembly |
| FR2994745B1 (en) * | 2012-08-21 | 2016-07-01 | Centre Nat D' Etudes Spatiales (Cnes) | METHOD FOR ESTIMATING THE AGING OF A BATTERY |
| US9236748B2 (en) * | 2012-08-30 | 2016-01-12 | Texas Instruments Incorporated | Method and apparatus of charging the battery with globally minimized integral degradation possible for predefined charging duration |
| JP6370581B2 (en) * | 2014-03-28 | 2018-08-08 | 学校法人早稲田大学 | Electrochemical analysis apparatus and electrochemical system |
| US10386422B2 (en) * | 2014-07-25 | 2019-08-20 | Lithium Balance A/S | Electrochemical impedance spectroscopy in battery management systems |
| US10955485B2 (en) * | 2015-08-25 | 2021-03-23 | Oxis Energy Limited | Battery sensor |
| JP6579068B2 (en) * | 2016-09-16 | 2019-09-25 | トヨタ自動車株式会社 | Fuel cell output performance diagnostic device, fuel cell output performance diagnostic system, fuel cell output performance diagnostic method, and fuel cell output performance diagnostic program |
| CN107607880B (en) * | 2017-09-19 | 2020-04-24 | 哈尔滨工业大学 | Lithium ion battery internal health feature extraction method based on impedance spectrum |
| CN109726452B (en) * | 2018-12-12 | 2020-10-27 | 浙江大学 | Online proton exchange membrane fuel cell fault diagnosis method based on impedance spectrum |
| CN109765496A (en) * | 2018-12-20 | 2019-05-17 | 西安交通大学 | A kind of cell health state estimation method based on online electrochemical impedance spectrometry |
| CN109633454B (en) * | 2019-01-13 | 2020-06-23 | 浙江大学 | Method for realizing on-line estimation of equivalent temperature of lithium ion battery |
| CN112415413B (en) * | 2020-11-05 | 2021-10-01 | 上海捷氢科技有限公司 | Diagnosis method and diagnosis system of fuel cell |
| CN112433095B (en) * | 2020-11-19 | 2022-02-11 | 中国科学院大连化学物理研究所 | Method for measuring water content in proton exchange membrane fuel cell membrane |
| CN113447843B (en) * | 2021-06-22 | 2022-10-28 | 中山大学 | Hydrogen fuel cell health state evaluation method based on electrochemical impedance spectrum |
| CN113466700B (en) * | 2021-06-24 | 2022-05-20 | 浙江大学 | Lithium battery online life prediction method based on two-point impedance life characteristics |
| CN113484784B (en) * | 2021-06-24 | 2022-07-08 | 浙江大学 | Lithium battery online aging diagnosis method based on two-point impedance aging characteristics |
-
2021
- 2021-10-14 CN CN202111196890.4A patent/CN113917352B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113917352A (en) | 2022-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110190306B (en) | Online fault diagnosis method for fuel cell system | |
| US10581099B2 (en) | Use of neural network and EIS signal analysis to quantify H2 crossover in-situ in operating PEM cells | |
| CN110676488B (en) | Online proton exchange membrane fuel cell fault diagnosis method based on low-frequency impedance and electrochemical impedance spectrum | |
| CN113009349A (en) | Lithium ion battery health state diagnosis method based on deep learning model | |
| CN112289385B (en) | Electrochemical impedance spectrum prediction method for high-power proton exchange membrane fuel cell stack | |
| CN107064807A (en) | Method for determining the critical operating state in fuel cell unit | |
| CN113608131B (en) | PEMFC pile recession performance detection method under dynamic locomotive working condition | |
| CN112784216B (en) | Health degree evaluation method and system of proton exchange membrane fuel cell system | |
| WO2022268144A1 (en) | Online aging diagnosis method for lithium battery based on two-point impedance aging characteristics | |
| CN106950507A (en) | A kind of intelligent clock battery high reliability lifetime estimation method | |
| CN113611900A (en) | Membrane dry fault diagnosis method of proton exchange membrane fuel cell | |
| O'Rian et al. | Application of the transmission line EIS model to fuel cell catalyst layer durability | |
| CN115248382A (en) | Control method of proton exchange membrane fuel cell | |
| Huang et al. | Correlation analysis and prediction of PEM fuel cell voltage during start-stop operation based on real-world driving data | |
| CN113917352B (en) | Online aging diagnosis method for catalyst layer of fuel cell based on impedance aging characteristic | |
| CN116203434B (en) | Method and equipment for predicting durability of fuel cell | |
| CN116053531A (en) | Evaluation system and method for fuel cell stack health state | |
| CN111628195A (en) | Fuel cell stack real-time state identification method based on logic reasoning | |
| CN114361535B (en) | Fuel cell hydrogen permeation quantity assessment method based on electrochemical impedance spectrum | |
| Peng et al. | A hybrid health prognostics method for proton Exchange membrane fuel cells with internal health recovery | |
| CN114583219A (en) | Water flooding fault diagnosis method for proton exchange membrane fuel cell | |
| CN116930804A (en) | PEMFC porous electrode performance attenuation evaluation method | |
| CN117117258B (en) | Fault monitoring method and device for hydrogen fuel cell system | |
| CN116031453A (en) | On-line estimation method for characteristic frequency impedance of proton exchange membrane fuel cell | |
| Pivac et al. | Diagnostic methods for automotive fuel cell systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Liu Hao Inventor after: Chen Jian Inventor before: Chen Jian Inventor before: Liu Hao |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |