A method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell

Abstract

This invention discloses a method for predicting the minimum auxiliary heating power required for a proton exchange membrane fuel cell (PEMFC), relating to the field of PEMFC technology. The method involves: acquiring the state data of the target PEMFC at the current moment; inputting the state data into an artificial neural network surrogate model to obtain the cold start failure time corresponding to different auxiliary heating powers at the current moment; performing thermodynamic analysis on the state data to determine the cold start success time corresponding to different auxiliary heating powers at the current moment; and determining the auxiliary heating power corresponding to the equality of the cold start success time and cold start failure time as the minimum auxiliary heating power required for the target PEMFC at the current moment. This method can accurately predict the minimum auxiliary heating power required for a successful cold start of a PEMFC.

Description

Technical Field

[0001] This invention relates to the field of proton exchange membrane fuel cell technology, and in particular to a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell. Background Technology

[0002] Currently, proton exchange membrane fuel cells (PEMFCs), as a highly efficient and clean energy conversion device, have broad application prospects in transportation and other fields. However, when the ambient temperature drops below freezing, the water produced by the PEMFC during operation freezes, clogging the pores and flow channels of the porous electrode. This prevents the electrochemical reaction from proceeding continuously, causing the PEMFC to fail to start normally—a problem known as the cold start problem. Especially in environments below -20°C, the cold start success rate is less than 60%, becoming a key factor limiting its promotion in cold regions. Furthermore, failed cold starts severely impact the lifespan of the PEMFC.

[0003] To address the cold start problem, external auxiliary heating is typically used to quickly raise the temperature of the proton exchange membrane fuel cell above its freezing point, achieving rapid heating and de-icing. However, insufficient auxiliary heating power can lead to cold start failure, while excessive auxiliary heating power results in resource waste. While some research has focused on cold start behavior, there is still insufficient research on how to predict the minimum auxiliary heating power required for a successful cold start, and accurate theoretical prediction methods are lacking. Summary of the Invention

[0004] Therefore, it is necessary to provide a method for predicting the minimum auxiliary heating power required for a proton exchange membrane fuel cell to address the aforementioned technical problems. This method can accurately predict the minimum auxiliary heating power required for a successful cold start of a proton exchange membrane fuel cell.

[0005] The present invention adopts the following technical solution:

[0006] This invention provides a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell, comprising:

[0007] Acquire the current state data of the target proton exchange membrane fuel cell;

[0008] The state data is input into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers; the training dataset of the artificial neural network surrogate model is obtained by simulating the proton exchange membrane fuel cell; the cold start failure time represents the time elapsed from when the target proton exchange membrane fuel cell starts the cold start operation to when the cold start failure is determined.

[0009] Thermodynamic analysis of the state data was performed to determine the cold start success time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers; the cold start success time is the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point.

[0010] The auxiliary heating power corresponding to the time when the cold start success time and the cold start failure time are equal is determined as the minimum auxiliary heating power of the target proton exchange membrane fuel cell at the current moment.

[0011] Optionally, the state data includes current loading density and initial temperature; the state data is input into an artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at different auxiliary heating powers at the current moment, including:

[0012] For any auxiliary heating power, the current loading density, initial temperature, and auxiliary heating power are input into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell corresponding to the auxiliary heating power at the current moment.

[0013] Optionally, the process of constructing an artificial neural network surrogate model includes:

[0014] The cold start process of a proton exchange membrane fuel cell was simulated using a one-dimensional unsteady multiphase model, and discharge curves under different operating conditions were obtained.

[0015] The discharge curves were processed to obtain a training dataset with different parameters and cold start failure time; the parameters included current loading density, initial temperature and auxiliary heating power.

[0016] The training dataset is iterated multiple times to obtain an artificial neural network proxy model.

[0017] Optionally, the state data includes current loading density and initial temperature; thermodynamic analysis is performed on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers, including:

[0018] Based on the total heat capacity, initial temperature, and temperature required for successful cold start of the target proton exchange membrane fuel cell, determine the minimum heat required for successful cold start of the target proton exchange membrane fuel cell;

[0019] The self-generated heat power of the target proton exchange membrane fuel cell is determined based on the current loading density, initial temperature, and temperature required for successful cold start.

[0020] Based on the self-generated heat power of the target proton exchange membrane fuel cell and the minimum heat required for a successful cold start, the successful cold start time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers is determined by the energy conservation equation.

[0021] Optionally, the formula for calculating the minimum heat required for a successful cold start of the target proton exchange membrane fuel cell is as follows:

[0022] ;

[0023] in, This represents the minimum amount of heat required for a successful cold start of the target proton exchange membrane fuel cell. This indicates the total heat capacity of the target proton exchange membrane fuel cell. This indicates the initial temperature of the target proton exchange membrane fuel cell. This indicates the freezing point temperature.

[0024] Optionally, the formula for calculating the self-generated heat power of the target proton exchange membrane fuel cell is:

[0025] ;

[0026] in, This indicates the self-generated heat power of the target proton exchange membrane fuel cell. This indicates the number of individual cells in the target proton exchange membrane fuel cell. This represents the self-generated heat power of a single cell in the target proton exchange membrane fuel cell;

[0027] ;

[0028] in, Indicates the heat of reaction of a single battery cell. Indicates the activation heat of a single battery cell;

[0029] ;

[0030] ;

[0031] ;

[0032] ;

[0033] ;

[0034] in, This indicates the average temperature during the cold start process. Represents the entropy change of a chemical reaction. Indicates the activation current density. This indicates the current loading density of the target proton exchange membrane fuel cell. Indicates the thickness of the cathode catalyst layer. Indicates the volume of the catalyst layer. Represents Faraday's constant. Represents the gas constant. Indicates reference oxygen concentration. Indicates the reference current density. Indicates the starting oxygen concentration. Represents the roughness factor. Represents specific surface area. This indicates the platinum loading.

[0035] Alternatively, the energy conservation equation is:

[0036] ;

[0037] in, Indicates auxiliary heating power. This indicates the time it took for a cold start to succeed.

[0038] This invention provides a device for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell, comprising:

[0039] The acquisition module is used to acquire the current state data of the target proton exchange membrane fuel cell.

[0040] The first determining module is used to input state data into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers; the training dataset of the artificial neural network surrogate model is obtained by simulating the proton exchange membrane fuel cell; the cold start failure time represents the time elapsed from when the target proton exchange membrane fuel cell starts to perform cold start operation to when the cold start failure is determined.

[0041] The second determining module is used to perform thermodynamic analysis on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers; the successful cold start time is the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point.

[0042] The third determining module is used to determine the auxiliary heating power corresponding to the time when the cold start success time and the cold start failure time are equal as the minimum auxiliary heating power of the target proton exchange membrane fuel cell at the current moment.

[0043] The present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for predicting the minimum auxiliary heating power of the proton exchange membrane fuel cell described above.

[0044] The present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for predicting the minimum auxiliary heating power of the proton exchange membrane fuel cell described above.

[0045] The above-mentioned at least one technical solution adopted in this invention can achieve the following beneficial effects:

[0046] In this invention, the cold start failure time corresponding to different auxiliary heating powers of the target proton exchange membrane fuel cell in the current state is obtained through a data-driven model, and the cold start success time corresponding to different auxiliary heating powers of the target proton exchange membrane fuel cell in the current state is obtained through thermodynamic analysis constraints. The auxiliary heating power corresponding to the intersection of the cold start failure time and the cold start success time is taken as the minimum auxiliary heating power. In this way, the method unifies the physical feasibility and engineering reliability of the cold start process under the same framework by predicting failure risk through a data-driven model and cross-validating the success conditions through thermodynamic analysis constraints. This makes the auxiliary heating power at the intersection point both the critical value that just does not trigger the failure mechanism and the minimum value that meets the thermodynamic heating requirements. Therefore, the minimum auxiliary heating power of the target proton exchange membrane fuel cell in the current state can be accurately determined. Attached Figure Description

[0047] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0048] Figure 1 A schematic flowchart of a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell provided by the present invention;

[0049] Figure 2 The curves showing the output voltage change over time under different parameter conditions provided by this invention, wherein (a) shows the output voltage change over time under different current loading densities. I The curve of the output voltage changing with time is shown in Figure (b), which shows the output voltage at different initial temperatures. T in The curve of the output voltage changing with time, (c) shows the output voltage with different auxiliary heating powers. P aux The curve of the output voltage changing over time;

[0050] Figure 3 A schematic diagram illustrating the predictive performance of an artificial neural network surrogate model provided by this invention;

[0051] Figure 4The present invention provides a schematic diagram of the heat generation of a PEMFC, wherein (a) the heat generation distribution of each component of the PEMFC and (b) the heat generation distribution in the cathode catalyst layer CLc.

[0052] Figure 5 A method provided by the present invention t 1 and t 2. With auxiliary heating power P aux A changing curve;

[0053] Figure 6 The present invention provides a battery cell with a self-generated heat power of [missing information]. The self-generated heat power in numerical simulation is A diagram illustrating the deviation;

[0054] Figure 7 A schematic diagram of the heat capacity distribution of various components in a proton exchange membrane fuel cell provided by the present invention;

[0055] Figure 8 This invention provides a method for predicting successful cold start time using a thermodynamic analysis model. Cold start success time compared to numerical simulation A diagram illustrating the deviation;

[0056] Figure 9 This invention provides a method for applying different current values. I Different cold start temperatures T in The lowest auxiliary heating power obtained Minimum auxiliary heating power compared to numerical simulation results A diagram illustrating the deviation;

[0057] Figure 10 This is a schematic diagram of a computer device for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell, as provided by the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0059] The cold start problem is influenced by numerous factors, and the interactions between these factors significantly impact the performance calculation of PEMFCs, thus hindering research on the conditions for successful cold starts. In recent years, machine learning, as an efficient tool for building data-driven models, has been widely applied in the PEMFC field, greatly reducing the computational cost of determining optimal parameters. However, building machine learning-driven models requires a large amount of basic data, and its accuracy is significantly affected by low-quality data and the degree of fit, becoming a major factor limiting its development. Furthermore, data-driven models are black-box models, with ambiguous physical meanings and a lack of explanation for optimal structures. Therefore, this invention employs a one-dimensional unsteady multiphase model to establish the output performance of a proton exchange membrane fuel cell under different cold start conditions, and utilizes an artificial neural network (ANN) to build a data-driven model to obtain the cold start failure time under different cold start conditions. Subsequently, by combining the cold start success time calculated from the thermodynamic analysis model, a general prediction method for the minimum auxiliary heating power required for successful PEMFC cold start is obtained through numerical comparison.

[0060] This invention transforms the problem of successful cold start into a numerical comparison of the successful cold start time and the failure cold start time under specific conditions. When the successful cold start time is lower than the failure cold start time under certain conditions, it indicates that the cold start can be successful under those conditions. Based on this, this invention provides a method for predicting the minimum auxiliary heating power required for a proton exchange membrane fuel cell (PEMFC). This method combines thermodynamic analysis and an artificial neural network surrogate model to determine the minimum auxiliary heating power required for a successful PEMFC cold start. This invention combines a physical model and a data-driven model, ensuring high accuracy while saving costs, and is more conducive to selecting the auxiliary heating power for a successful cold start in engineering applications.

[0061] The technical solutions provided by the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] Figure 1 This is a schematic flowchart of a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell according to the present invention, which specifically includes the following steps:

[0063] S101, Obtain the current state data of the target proton exchange membrane fuel cell.

[0064] The target proton exchange membrane fuel cell is any proton exchange membrane fuel cell with the minimum auxiliary heating power to be determined; the state data may include current loading density and initial temperature, where the initial temperature can be the start-up stack temperature of the target proton exchange membrane fuel cell at the current moment.

[0065] S102, input the state data into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers; the training dataset of the artificial neural network surrogate model is obtained by simulating the proton exchange membrane fuel cell; the cold start failure time represents the time elapsed from when the target proton exchange membrane fuel cell starts to perform cold start operation to when the cold start failure is determined.

[0066] In one embodiment, the state data is input into an artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell corresponding to different auxiliary heating powers at the current moment. This includes: for any auxiliary heating power, inputting the current loading density, initial temperature and auxiliary heating power into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell corresponding to the auxiliary heating power at the current moment.

[0067] The construction process of the artificial neural network surrogate model includes: simulating the cold start process of a proton exchange membrane fuel cell using a one-dimensional unsteady multiphase model to obtain discharge curves under different operating conditions; processing the discharge curves to obtain training datasets with different parameters and cold start failure times; the parameters include current loading density, initial temperature, and auxiliary heating power; and iteratively training the training dataset to obtain the artificial neural network surrogate model.

[0068] Specifically, a one-dimensional unsteady multiphase model was constructed using the multiphysics simulation software COMSOL. This one-dimensional unsteady multiphase model is the PEMFC simulation model. The cold start process of PEMFC was modeled and solved using the one-dimensional unsteady multiphase model to obtain the discharge curves corresponding to different current loading densities, initial temperatures, and auxiliary heating power. After outlier removal using 3σ theory and normalization using the Min-Max method, a training dataset with different parameters and corresponding cold start failure times was established.

[0069] The discharge curve is the curve of output voltage changing over time, such as... Figure 2 As shown, Figure 2 Figure 1 shows the output voltage versus time curves under different parameter conditions, illustrating the impact of different parameter conditions on the cold start performance of a proton exchange membrane fuel cell. Figure 2(a) shows the output voltage versus time curves under different current loading densities I, and Figure 3(b) shows the output voltage versus time curves under different initial temperatures. T in The curve of the output voltage changing with time, (c) shows the output voltage with different auxiliary heating powers. P aux The curve showing the change of output voltage over time.

[0070] Specifically, the output voltage drops to 0 at the moment of cold start failure. Under each set of conditions, the corresponding output parameter - cold start failure time can be read from the corresponding discharge curve, thus obtaining the training dataset.

[0071] Since cold start failure time is affected by a variety of factors, such as the initial temperature of the start-up stack, the auxiliary heating power, and the applied current density, an artificial neural network surrogate model was constructed using a training dataset to predict cold start failure time as a function of initial temperature, current density, and cold start failure. Specifically, the artificial neural network surrogate model was obtained by iteratively training the backpropagation neural network using the training dataset.

[0072] The specific settings of the artificial neural network surrogate model include: 3 hidden layers with 50 neurons, tanh activation function, root mean square error, Adam optimizer (an improved algorithm based on gradient descent) for weight correction, initialization with learning rate α = 0.0085, first-order momentum variable (exponential moving average of gradient) m0 = 0, second-order momentum variable (exponential moving average of squared gradient) v0 = 0, and bias correction parameters β1 = 0.9 and β2 = 0.999.

[0073] The predictive performance of this artificial neural network surrogate model is as follows: Figure 3 As shown, the average error is 6.94%. The horizontal axis represents the cold start failure time in the numerical simulation. The vertical axis represents the cold start failure time predicted by the neural network model. .

[0074] In this embodiment, the input parameters of the artificial neural network surrogate model are various conditional parameters (initial temperature, current loading density, and auxiliary heating power), and the output parameter is the cold start failure time. That is, the artificial neural network surrogate model establishes a simple correspondence between various conditional parameters and cold start failure time. The output parameters can be obtained through the input parameters without solving the complex PEMFC cold start model control equations, which simplifies the calculation process.

[0075] S103. Perform thermodynamic analysis on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers; the successful cold start time is the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point.

[0076] By using a thermodynamic analysis model of PEMFC, we can explore the heat capacity and heat generation of PEMFC, derive the minimum heat required for PEMFC to reach its freezing point and the heat generation power of PEMFC, and solve for the successful cold start time.

[0077] forP self This invention discovers that when the PEMFC cold start temperature is -20℃ and the current loading density is 1000 Am... -2 When the PEMFC starts up, the heat generation is as follows: Figure 4 As shown in the figure, (a) shows the heat generation distribution of each component of the PEMFC, which specifically includes the end plate (EP), flow channel (CH), bipolar plate (BP), microporous layer (MPL), gas diffusion layer (GDL), catalyst layer (CL), and proton exchange membrane (PEM). The subscript a represents the anode, and the subscript c represents the cathode. For example, EPa represents the anode end plate, EPc represents the cathode end plate, and so on for other components. It can be seen from the figure that the heat generated by the cathode catalyst layer CLc exceeds 90% of the battery's self-generated heat. (b) shows the heat generation distribution in the cathode catalyst layer CLc, specifically including ohmic heat... Heat of reaction Activation heat Latent heat By analyzing the heat generation distribution of each heat source term in the cathode catalyst layer CLc, it was found that the activation heat and reaction heat account for approximately 98% of the total heat of CLc.

[0078] Specifically, in one embodiment, the state data includes current loading density and initial temperature; performing thermodynamic analysis on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at different auxiliary heating powers at the current moment includes the following steps:

[0079] S201, based on the total heat capacity, initial temperature, and temperature required for successful cold start of the target proton exchange membrane fuel cell, determine the minimum heat required for successful cold start of the target proton exchange membrane fuel cell.

[0080] The formula for calculating the minimum heat required for a successful cold start of the target proton exchange membrane fuel cell is as follows:

[0081] (1);

[0082] in, This represents the minimum amount of heat required for a successful cold start of the target proton exchange membrane fuel cell. This indicates the total heat capacity of the target proton exchange membrane fuel cell. This indicates the initial temperature of the target proton exchange membrane fuel cell. This indicates the freezing point temperature, which is the temperature required for a successful cold start of the target proton exchange membrane fuel cell. The temperature change indicates a successful cold start.

[0083] A thermodynamic analysis model of PEMFC was established, and the analysis of the heat capacity and heat generation of PEMFC showed that more than 95% of the heat capacity of the proton exchange membrane fuel cell is concentrated in the end plates, bipolar plates, and flow channels. This allows for the determination of the total heat capacity of the target proton exchange membrane fuel cell. The heat capacity mainly comes from the end plates, bipolar plates, and flow channels of the target proton exchange membrane fuel cell. The heat capacity of these three components is added together to approximate the total heat capacity of the target proton exchange membrane fuel cell.

[0084] Substituting the heat capacity, initial temperature, and temperature required for successful cold start of the target proton exchange membrane fuel cell into formula (1), the minimum heat required for successful cold start of the target proton exchange membrane fuel cell can be obtained.

[0085] S202, determine the self-generated heat power of the target proton exchange membrane fuel cell based on the current loading density, initial temperature, and temperature required for successful cold start.

[0086] Optionally, the formula for calculating the self-generated heat power of the target proton exchange membrane fuel cell is:

[0087] (2);

[0088] in, This indicates the self-generated heat power of the target proton exchange membrane fuel cell. This indicates the number of individual cells in the target proton exchange membrane fuel cell. This represents the self-generated heat power of a single cell in the target proton exchange membrane fuel cell.

[0089] for The present invention, based on simulation data, shows that the heat generated by the cathode catalyst layer exceeds 90% of the battery's own heat generation, and can be estimated as the total heat generation. This means that... Meanwhile, the heat generated by CLc is mainly provided by the heat of activation and the heat of reaction. Therefore, the heat generated by CLc can be approximated as the sum of the heat of activation and the heat of reaction, that is:

[0090] (3);

[0091] in, Indicates the heat of reaction of a single battery cell. This indicates the activation heat of a single battery cell.

[0092] (4);

[0093] (5);

[0094] (6);

[0095] (7);

[0096] (8);

[0097] in, This indicates the average temperature during the cold start process. Represents the entropy change of a chemical reaction. Indicates the activation current density. This indicates the current loading density of the target proton exchange membrane fuel cell. Indicates the thickness of the cathode catalyst layer. Indicates the volume of the catalyst layer. Represents Faraday's constant. Represents the gas constant. Indicates reference oxygen concentration. Indicates the reference current density. Indicates the starting oxygen concentration. Represents the roughness factor. Represents specific surface area. This indicates the platinum loading.

[0098] By substituting formulas (2)-(8) into the parameters of the target proton exchange membrane fuel cell, the self-generated heat power of the target proton exchange membrane fuel cell can be obtained.

[0099] S203, based on the self-generated heat power of the target proton exchange membrane fuel cell and the minimum heat required for successful cold start, determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers through the energy conservation equation.

[0100] The minimum heat required for a successful cold start of a proton exchange membrane fuel cell represents the minimum thermal energy required to raise the average cold temperature of the proton exchange membrane fuel cell to its freezing point. Therefore, the minimum heat required for a successful cold start of the target proton exchange membrane fuel cell is... Q min The heat generated by the internal target proton exchange membrane fuel cell itself and by external auxiliary heating is provided. Therefore, the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point, i.e., the successful cold start time, can be estimated by the following energy conservation equation:

[0101] (9);

[0102] in, Indicates auxiliary heating power. This indicates the time it took for a cold start to succeed.

[0103] Under different auxiliary heating powers, the successful cold start time is obtained by dividing the minimum heat required for a successful cold start of the PEMFC by the sum of the self-heating rate and the auxiliary heating power. Therefore, based on formula (9), the successful cold start time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers can be obtained.

[0104] The meanings and specific values ​​of the above parameters can be found in Table 1.

[0105] Table 1. Parameter Meanings and Values

[0106]

[0107] In this embodiment, the minimum heat required for a successful cold start of the PEMFC (the average temperature of the PEMFC rises to the freezing point) is first calculated based on the heat capacity of the PEMFC and the temperature change for a successful cold start. Then, the self-generated heat power of the PEMFC is calculated. Under different auxiliary heating powers, the successful cold start time is obtained by dividing the minimum heat required for a successful cold start of the PEMFC by the sum of the self-generated heat power and the auxiliary heating power.

[0108] S104, the auxiliary heating power corresponding to the time when the cold start success time and the cold start failure time are equal is determined as the minimum auxiliary heating power of the target proton exchange membrane fuel cell at the current moment.

[0109] In one embodiment, nonlinear fitting can be performed on the cold start failure time and cold start success time of the target proton exchange membrane fuel cell at different auxiliary heating powers at the current time to obtain curves of cold start failure time versus auxiliary heating power and curves of cold start success time versus auxiliary heating power. The cold start success time is represented by... t 1 indicates the cold start failure time. t 2 indicates, such as Figure 5 As shown, Figure 5 for t 1 and t 2. With auxiliary heating power P aux A changing curve.

[0110] Will t 1 and t 2. Curve showing the change in auxiliary heating power t 1 and t The point of intersection of the two points determines the corresponding auxiliary heating power, which is then used as the minimum auxiliary heating power. P min .

[0111] The present invention provides a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell. This method is based on a one-dimensional unsteady multiphase model to establish parameters in the proton exchange membrane fuel cell, including the initial temperature for cold start, current loading density, auxiliary heating power, and the corresponding cold start failure time. t 2) An artificial neural network proxy model was used, and the successful cold start time of PEMFC was derived through thermodynamic analysis. t 1) Prediction formula. This is based on the cold start success time ( t 1) Compared with cold start failure time ( t 2) By comparing the numerical values, the minimum auxiliary heating power required for a successful cold start of a PEMFC is obtained. The purpose of this invention is to quickly predict the minimum auxiliary heating power required for a successful cold start of a PEMFC under different initial cold start temperatures and current loading strategies by combining data-driven and physical modeling methods. This provides a reference for the power design of auxiliary heating for cold starts of PEMFCs, preventing the adverse effects of cold start failures on PEMFCs. This method combines physical models and data-driven models, saving costs while ensuring high accuracy, and is more conducive to the design of auxiliary heating power for PEMFC cold starts in engineering applications, providing a reference for the power design of auxiliary heating for cold starts of PEMFCs and preventing the adverse effects of cold start failures on PEMFCs.

[0112] The prediction method provided by this invention is derived based on the thermodynamic analysis model of PEMFC and the assumptions of specific heat production conditions. However, its applicability is not limited to the thermodynamic analysis model of PEMFC provided by this invention. Its physical meaning can be extended to the PEMFC cold start model based on other structural parameters, and it has good versatility and reliability.

[0113] In one embodiment, when a PEMFC fails to start automatically, auxiliary heating is often required for startup. This invention establishes a rapid predictive model to estimate the minimum auxiliary heating power required for a successful cold start. P min The basic idea behind establishing this rapid prediction model is as follows: The rapid prediction model is established by comparing two typical time scales: one is the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point (successful cold start time), denoted as... t 1. The other is the cold start failure time of PEMFC under auxiliary heating conditions, denoted as... t 2. If t 1 less than t A value of 2 indicates that the porous electrodes in the PEMFC were not completely blocked by ice before the temperature reached the freezing point, signifying a successful cold start. The input parameter for this rapid prediction model is the current loading density. I Initial temperature Tin and auxiliary heating power P aux The output is the minimum auxiliary heating power required for a successful cold start. P min .

[0114] In one specific embodiment, this embodiment establishes a one-dimensional unsteady multiphase model of PEMFC, considering the heat capacity and heat generation characteristics of multiple components such as the catalyst layer, diffusion layer, proton exchange membrane, and endplates (specific parameters are shown in Table 2), and derives the successful cold start time. t 1. Train an artificial neural network agent model using multiple sets of simulation data to predict cold start failure time. t 2.

[0115] Table 2 Design and operating parameters of the one-dimensional unsteady multiphase model

[0116]

[0117] By comparing the two ( t 1 ≤ t 2) Determine the minimum auxiliary heating power P min Experiments verified the effects of different loading currents and initial temperatures. P min The average error is within 10%, proving the practicality and reliability of the method.

[0118] By investigating the heat generation of PEMFCs, it was found that the heat generated by Clc (cathode catalyst layer) exceeds 90% of the battery's self-generated heat, and can be estimated as the total heat generation. This means... Meanwhile, the heat generated by CLc is mainly provided by the heat of activation and the heat of reaction, so the heat generated by CLc can be approximated as the sum of the heat of activation and the heat of reaction.

[0119] Based on the above assumptions, the self-generated heat power of the battery cell predicted by the thermodynamic analysis model is: The self-generated heat power in numerical simulation is deviation such as Figure 6 As shown, the two are in good agreement.

[0120] Figure 7 This is a schematic diagram of the heat capacity distribution of various components in a proton exchange membrane fuel cell, revealing the total heat capacity of the proton exchange membrane fuel cell. The heat capacity of the proton exchange membrane fuel cell mainly comes from EP, BP, and CH. Therefore, the sum of the heat capacities of these three components is used to approximate the total heat capacity of the proton exchange membrane fuel cell. One can approximate this by adding the heat capacities of EP (stainless steel), BP (titanium alloy), and CH (titanium alloy), resulting in a total heat capacity of 461 J·K.-1 The successful cold start time predicted by the thermodynamic analysis model based on the above assumptions. Cold start success time compared to numerical simulation deviation such as Figure 8 As shown, the two are in good agreement.

[0121] Figure 9 The fast prediction model for assisted heating is shown under different current loading densities. I Different cold start temperatures T in The lowest auxiliary heating power obtained Minimum auxiliary heating power compared to numerical simulation results The deviations are all within 10% on average, which proves the reliability of the fast prediction model.

[0122] When applying the prediction method for the minimum auxiliary heating power of a proton exchange membrane fuel cell provided by this invention, it is not necessary to rely on... Figure 1 The steps shown are executed in sequence. The specific execution order of each step can be determined as needed, and this invention does not impose any restrictions on it.

[0123] The above describes a method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell according to one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding device for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell, the device comprising:

[0124] The acquisition module is used to acquire the current state data of the target proton exchange membrane fuel cell.

[0125] The first determining module is used to input state data into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers; the training dataset of the artificial neural network surrogate model is obtained by simulating the proton exchange membrane fuel cell; the cold start failure time represents the time elapsed from when the target proton exchange membrane fuel cell starts to perform cold start operation to when the cold start failure is determined.

[0126] The second determining module is used to perform thermodynamic analysis on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers; the successful cold start time is the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point.

[0127] The third determining module is used to determine the auxiliary heating power corresponding to the time when the cold start success time and the cold start failure time are equal as the minimum auxiliary heating power of the target proton exchange membrane fuel cell at the current moment.

[0128] Specific limitations regarding the device for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell can be found in the limitations of the prediction method for the minimum auxiliary heating power of a proton exchange membrane fuel cell described above, and will not be repeated here. Each module in the aforementioned device for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0129] The present invention also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 A method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell is provided.

[0130] The present invention also provides Figure 10 The schematic diagram of the computer device shown is as follows: Figure 10 As shown, at the hardware level, this computer device includes a processor, internal bus, network interface, memory, and non-volatile memory, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then executes it to achieve the above. Figure 1 A method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell is provided.

[0131] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0132] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this invention.

Claims

1. A method for predicting the minimum auxiliary heating power of a proton exchange membrane fuel cell, characterized in that, include: Acquire the current state data of the target proton exchange membrane fuel cell; The state data is input into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers. The training dataset for the artificial neural network surrogate model is obtained by simulating a proton exchange membrane fuel cell; the cold start failure time represents the time elapsed from when the target proton exchange membrane fuel cell begins its cold start operation to when the cold start failure is determined. When the cold start fails, the output voltage of the target proton exchange membrane fuel cell drops to 0. Thermodynamic analysis of the state data was performed to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers; Cold start success time refers to the time required for the average temperature of the target proton exchange membrane fuel cell to reach the freezing point. The auxiliary heating power corresponding to the time when the cold start success time and the cold start failure time are equal is determined as the minimum auxiliary heating power of the target proton exchange membrane fuel cell at the current moment.

2. The method according to claim 1, characterized in that, State data includes current loading density and initial temperature; the state data is input into an artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell at different auxiliary heating powers at the current moment, including: For any auxiliary heating power, the current loading density, initial temperature, and auxiliary heating power are input into the artificial neural network surrogate model to obtain the cold start failure time of the target proton exchange membrane fuel cell corresponding to the auxiliary heating power at the current moment.

3. The method according to claim 2, characterized in that, The process of constructing an artificial neural network surrogate model includes: The cold start process of a proton exchange membrane fuel cell was simulated using a one-dimensional unsteady multiphase model, and discharge curves under different operating conditions were obtained. The discharge curves were processed to obtain a training dataset with different parameters and cold start failure time; the parameters included current loading density, initial temperature and auxiliary heating power. The training dataset is iterated multiple times to obtain an artificial neural network proxy model.

4. The method according to claim 1, characterized in that, State data includes current loading density and initial temperature; thermodynamic analysis is performed on the state data to determine the successful cold start time of the target proton exchange membrane fuel cell at the current moment for different auxiliary heating powers, including: Based on the total heat capacity, initial temperature, and temperature required for successful cold start of the target proton exchange membrane fuel cell, determine the minimum heat required for successful cold start of the target proton exchange membrane fuel cell; The self-generated heat power of the target proton exchange membrane fuel cell is determined based on the current loading density, initial temperature, and temperature required for successful cold start. Based on the self-generated heat power of the target proton exchange membrane fuel cell and the minimum heat required for a successful cold start, the successful cold start time of the target proton exchange membrane fuel cell at the current moment corresponding to different auxiliary heating powers is determined by the energy conservation equation.

5. The method according to claim 4, characterized in that, The formula for calculating the minimum heat required for a successful cold start of the target proton exchange membrane fuel cell is as follows: ； in, This represents the minimum amount of heat required for a successful cold start of the target proton exchange membrane fuel cell. This indicates the total heat capacity of the target proton exchange membrane fuel cell. This indicates the initial temperature of the target proton exchange membrane fuel cell. This indicates the freezing point temperature.

6. The method according to claim 5, characterized in that, The formula for calculating the self-generated heat power of the target proton exchange membrane fuel cell is as follows: ； in, This indicates the self-generated heat power of the target proton exchange membrane fuel cell. This indicates the number of individual cells in the target proton exchange membrane fuel cell. This represents the self-generated heat power of a single cell in the target proton exchange membrane fuel cell; ； in, Indicates the heat of reaction of a single battery cell. Indicates the activation heat of a single battery cell; ； ； ； ； ； in, This indicates the average temperature during the cold start process. Represents the entropy change of a chemical reaction. Indicates the activation current density. This indicates the current loading density of the target proton exchange membrane fuel cell. Indicates the thickness of the cathode catalyst layer. Indicates the volume of the catalyst layer. Represents Faraday's constant. Represents the gas constant. Indicates reference oxygen concentration. Indicates the reference current density. Indicates the starting oxygen concentration. Represents the roughness factor. Represents specific surface area. This indicates the platinum loading.

7. The method according to claim 6, characterized in that, The energy conservation equation is: ； in, Indicates auxiliary heating power. This indicates the time it took for a cold start to succeed.

Images