Method for establishing fuel cell failure embedded model

Abstract

The application discloses a kind of establishment methods of fuel cell fault embedding model, including fault function embedding, electrochemical model calculation, membrane state water content solution, liquid water content solution gas component concentration solution and temperature equation solution sub-item calculation.The model is embedded by fault function, realizes the accurate simulation under the working condition of pressure fault, humidity fault and flow fault.Under the condition of meeting the operation requirement of fuel cell, the complex heat and mass transfer process inside fuel cell under different fault working conditions and the dynamic response of multiple transient conditions can be understood.The complex physical field change inside fuel cell under fault working condition is solved by the mutual coupling of electrochemistry, fluid and heat.The problem of unbalanced fault data type and poor data quality in current fault diagnosis research is solved, the accumulation of various fault data is realized, the model support is provided for fuel cell fault diagnosis algorithm development, and the product development cycle and bench test cost are greatly shortened.

Description

Technical Field

[0001] This invention belongs to the field of fuel cells, and specifically relates to a method for establishing fuel cell models with different embedded fault characteristics. Technical Background

[0002] Polymer electrolyte membrane fuel cells (PEMFCs) have shown great potential in rail transportation, new energy vehicles, and other fields due to their advantages of high power density, low operating temperature, and fast response. As a complex, multidimensional, nonlinear, and highly coupled system, fault prediction and health management of fuel cell systems are crucial for the system's safety and reliability. However, fault diagnosis methods for fuel cell systems are highly dependent on the quantity and quality of available samples. In real-world road conditions, fuel cell systems operate normally most of the time, making the accumulation of fault data relatively slow compared to other types of data. Besides the high cost often associated with collecting fault data, the quality of the data samples is also insufficient. Normal operating data collected through sensor monitoring is often far greater than fault data, which can lead to fault samples remaining unidentified even with highly accurate diagnostic models. While bench tests can artificially induce faults by changing experimental parameters, this method is not only extremely time-consuming but also highly likely to cause irreversible damage to the fuel cell.

[0003] This invention proposes a multiphysics-based polymer electrolyte membrane fuel cell fault embedding model applicable to various fault conditions, thereby helping to accumulate data and accelerate fault diagnosis research. Compared to experimental testing, it can reduce the R&D cycle and research costs. Simultaneously, the developed model can be used to analyze the impact of faults from simulation results, aiding in the optimization and selection of various sensors in the early stages, and reducing sensor installation costs in actual systems and computational costs during online system diagnostics. Summary of the Invention

[0004] The purpose of this invention is to propose a method for establishing a fuel cell fault embedding model. Based on fully considering the mutual coupling of the three physical fields of electrochemistry, fluidity and heat, the method accurately simulates the working process of the fuel cell under fault conditions by solving the mass conservation equation, energy conservation law and output voltage of multiple physical quantities inside the fuel cell, thereby accumulating fault data and helping to accelerate research related to fault diagnosis.

[0005] The following describes the principle of this invention and the method for establishing a fuel cell fault embedding model. The fuel cell includes: anode and cathode channels, a gas diffusion layer, a microporous layer, a catalyst layer, and a proton exchange membrane. The modeling method includes the setting and solving of six parts: fault function embedding, electrochemical model calculation, membrane water content solution, liquid water content solution, gas component concentration solution, and temperature equation solution. The model equation solution uses an explicit format update algorithm, and the specific steps are as follows:

[0006] (1) Fault function embedding

[0007] Fault functions are embedded into the inlet operating parameters to accurately describe the characteristics of faults and simulate operating conditions of pressure faults, humidity faults, and flow faults. Pressure faults are characterized by the pressure drop at the anode and cathode inlets. When a pressure fault occurs, it can be quantitatively described as follows:

[0008] P in,a =f P,a (t)P inlet,a (1)

[0009] P in,c =f P,c (t)P inlet,c (2)

[0010] Insufficient humidity is a common failure in water management within fuel cell systems. It can lead to membrane drying of the proton exchange membrane and directly affect the water vapor concentration at the fuel cell inlet. A humidity failure can be quantitatively described as follows:

[0011]

[0012]

[0013] Insufficient reactant flow can be used to characterize gas leaks in gas supply lines and insufficient gas supply flow failures under variable load conditions. When a flow failure occurs, it can be quantitatively described as follows:

[0014] q in,a =c in,a f q,a (t)v a (5)

[0015] q in,c =c in,c f q,c (t)v c (6)

[0016] (2) Electrochemical model calculation

[0017] The expression for calculating the output voltage of a fuel cell is as follows:

[0018] V = Erev -η ohm -η act,a -η act,c (7)

[0019] The formula for calculating the reversible loss is as follows:

[0020]

[0021] The formula for calculating Ohm's loss is as follows:

[0022]

[0023] Activation loss is divided into anodic activation loss and cathodic activation loss, and their calculation expressions are as follows:

[0024]

[0025]

[0026] (3) Solving for membrane water content

[0027] Membrane water exists in the catalyst layers of the membrane, anode, and cathode, and affects the calculation of ionic conductivity. The expression for calculating membrane water content is as follows:

[0028]

[0029] The diffusion flux of membrane water diffusing from adjacent layers to layer p is calculated as follows:

[0030]

[0031] In the formula φ mw,p-q This represents the diffusion flux of membrane water from the adjacent q layer to the p layer.

[0032] (4) Solving for liquid water content

[0033] When the water vapor pressure inside a fuel cell exceeds the saturation pressure, the water vapor undergoes a phase change and condenses into saturated liquid water, hereinafter referred to as liquid water. The formula for calculating the saturated water vapor pressure is as follows:

[0034]

[0035] P sat Let represent the saturated water vapor pressure. After water vapor reaches saturation, it transforms into saturated liquid water. In this model, it is assumed that the liquid water is continuous throughout the porous layer. The equation for the conservation of liquid water mass is used to solve the equation. The formula for calculating the volume fraction of liquid water within the porous layer is as follows:

[0036]

[0037] The formula for calculating the permeation flux of liquid water from an adjacent layer to layer m is as follows:

[0038]

[0039] Hydraulic p l The capillary pressure p in the porous medium was calculated using capillary pressure. c The relationship between the volume fraction of liquid water (s) and the following equation is:

[0040]

[0041] Where σ lq ε is the surface tension coefficient; θ is the contact angle; ε is the porosity; K is the liquid water permeability; p g It is the total gas pressure, from which the hydraulic pressure p at the next moment can be obtained using the liquid water volume fraction at the previous moment. l Then, the volume fraction s of liquid water at the new time point is calculated.

[0042] (5) Determining the concentration of gas components

[0043] The gas components inside a fuel cell include water vapor, hydrogen, oxygen, and nitrogen. The specific formulas for calculating the concentration of these gas components within the porous layer are as follows:

[0044]

[0045] The formula for calculating gas diffusion flux is as follows:

[0046]

[0047] (6) Temperature solution

[0048] The local temperature is solved using the energy conservation equation, and the calculation expression is as follows:

[0049]

[0050] The formula for calculating heat flux is:

[0051]

[0052] The model is solved using an explicit format update algorithm. By embedding the aforementioned fault functions and solving the equations, a fault-embedded fuel cell model can be established. Based on the settings of the fuel cell fault functions, performance parameters, and operating environment conditions, the dynamic changes of the fuel cell performance and internal parameters under fault conditions can be solved, thereby accumulating various fault data and accelerating fault diagnosis and identification.

[0053] The features and beneficial effects of this invention are as follows:

[0054] (1) Fault functions for pressure faults, humidity faults, and flow faults are proposed and embedded to simulate the inlet conditions of different fault conditions, thereby accurately describing the location of various faults and their impact on fuel cells. (2) Under different fault conditions, the complex heat and mass transfer processes inside the fuel cell and the dynamic response of the performance when a fault occurs can be simulated. (3) To address the problems of unbalanced data types and low data quality in current fault diagnosis, three physical fields—electrochemical, fluid, and thermal—are coupled to efficiently calculate gas component transport, battery heat generation, and performance changes under fault conditions, thereby accumulating fault data and accelerating related research on fault diagnosis. (4) Through the developed fault embedding model, the causes of faults can be analyzed from the simulation results, helping to optimize the selection of various sensors in the early stages and reducing the sensor installation cost in the actual system and the computational cost during online system diagnosis. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the fuel cell structure in this invention.

[0056] Figure 2 This is a polarization curve verification diagram with an ambient temperature of 343.15K and a relative humidity of 100% at both the anode and cathode.

[0057] Figure 3 This is a power density curve verification graph with an ambient temperature of 343.15K and a relative humidity of 100% at both the anode and cathode.

[0058] Figure 4 The output voltage change curve of the simulation model under pressure failure conditions.

[0059] Figure 5 The output voltage change curve of the simulation model under humidity fault conditions.

[0060] Figure 6 The simulation model outputs a voltage change curve under flow failure conditions.

[0061] Figure 7 The curve showing the change in oxygen concentration within the cathode catalyst layer under pressure failure conditions.

[0062] in Figure 2 and Figure 3 It is a comparison of simulation results and experimental data curves. Figure 4 , Figure 5 , Figure 6 as well as Figure 7 These are all effects of embodiments of the present invention. Detailed Implementation

[0063] The modeling steps and process of the present invention will be further explained below through specific embodiments and implementation effects.

[0064] Fuel cell structure as follows Figure 1 As shown, the battery includes: a cathode channel, a gas diffusion layer, a microporous layer, a catalyst layer, an anode channel, a gas diffusion layer, a microporous layer, a catalyst layer, and a proton exchange membrane. The fault embedding model proposed in this invention includes the setting and solving of six parts: fault function embedding, electrochemical model calculation, membrane water content solution, liquid water content solution, gas component concentration solution, and temperature equation solution. First, fault function embedding is performed:

[0065] The fault functions embedded in this embodiment are divided into three categories: pressure fault function, humidity fault function, and flow fault function.

[0066] (1) Pressure fault embedding can be quantitatively described as:

[0067] P in,a =f P,a (t)P inlet,a (1)

[0068] P in,c =f P,c (t)P inlet,c (2)

[0069] Insufficient humidity is a common fault in water management of fuel cell systems. Humidity fault embedding can be quantitatively described as follows:

[0070]

[0071]

[0072] Insufficient reactant flow can be used to characterize gas leaks in gas supply lines and insufficient gas supply flow failures under variable load conditions. It can be quantitatively described as follows:

[0073] q in,a =c in,a f q,a (t)v a (5)

[0074] q in,c =c in,c f q,c (t)v c (6)

[0075] In addition to embedding the ingress fault function, the proposed model requires solving for electrochemical performance, membrane water content, liquid water volume fraction, gas component concentration, and temperature. The calculations for each component are as follows:

[0076] (2) Electrochemical performance:

[0077] The expression for calculating the output voltage of a fuel cell is as follows:

[0078] V = E rev -η ohm -η act,a -η act,c (7)

[0079] The formula for calculating reversible loss is as follows:

[0080]

[0081] The formula for calculating Ohm's loss is as follows:

[0082]

[0083] Activation loss is divided into anodic activation loss and cathodic activation loss, and their calculation expressions are as follows:

[0084]

[0085]

[0086] The calculation results of each component in a fuel cell have a significant impact on its electrochemical performance. The solution process is as follows:

[0087] (3) Solving for membrane water content:

[0088] The expression for calculating membrane water content is as follows:

[0089]

[0090] The diffusion flux of membrane water from adjacent layers to layer p is calculated as follows:

[0091]

[0092] (4) Solution for liquid water content:

[0093] The formula for calculating saturated water vapor pressure is as follows:

[0094]

[0095] Calculation of liquid water volume fraction within porous layers:

[0096]

[0097] The formula for calculating the permeation flux of liquid water from an adjacent layer to layer m is as follows:

[0098]

[0099] Hydraulic p lThe capillary pressure p in the porous medium was calculated using capillary pressure. c The relationship between the volume fraction of liquid water (s) and the following equation is:

[0100]

[0101] (5) Determining the concentration of gas components:

[0102] The gas components inside a fuel cell include water vapor, hydrogen, oxygen, and nitrogen. The specific formulas for calculating the concentration of these gas components within the porous layer are as follows:

[0103]

[0104] The formula for calculating diffusion flux is as follows:

[0105]

[0106] (6) Temperature calculation:

[0107] Local temperature can be solved using the energy conservation equation:

[0108]

[0109] The formula for calculating the heat flux transferred to layer m is:

[0110]

[0111] By embedding the aforementioned fault functions and solving the equations, a fault-embedded fuel cell model can be established. Based on the settings of the fuel cell fault functions, performance parameters, and operating environment conditions, the dynamic changes of the fuel cell performance and internal parameters under fault conditions can be solved, thereby accumulating various fault data and improving fault diagnosis research. Compared with bench testing, the development of fault embedding models can reduce R&D costs and shorten the development cycle.

[0112] The fault function and some parameters involved in the calculation example are as follows:

[0113] When a pressure failure occurs, the pressure failure function is set as follows:

[0114]

[0115] When a humidity fault occurs, the humidity fault function is set as follows:

[0116]

[0117] When a traffic failure occurs, the traffic failure function is set as follows:

[0118] f q,a (t)=fq,c (t) = 0.3;

[0119] Ambient temperature: 343.15 K; Normal inlet pressure of cathode and anode: 2.0 atm; Normal inlet humidity of cathode and anode: 1.0; Activation area: 3 × 10⁻⁶ -4 m 2 The gas diffusion layer has a porosity of 0.6; the microporous layer has a porosity of 0.4; the catalyst layer has a porosity of 0.3; the electrolyte fraction in the anode catalyst layer is 0.25; the electrolyte fraction in the cathode catalyst layer is 0.25; the flow channel length is 0.1 m; the flow channel width is 0.001 m; the flow channel thickness is 0.002 m; the proton exchange membrane used is Nafion 212; the thicknesses of the diffusion layer, microporous layer, and catalyst layer are 3 × 10⁻⁶ respectively. -4 0.4×10 -4 0.1×10 -4 m; Faraday constant 96487 C mol -1 Ideal gas constant: 8.314472 Jmol -1 K -1 Membrane equivalent 1.1 kg mol⁻¹; Time step 10 mol⁻¹ -6 The specific heat capacities of hydrogen, air, liquid water, electrode plates, and proton exchange membranes are 14300 J / kg, respectively. -1 K -1 1005J kg -1 K -1 4200J kg -1 K -1 1580J kg -1 K -1 833J kg -1 K -1 A constant current model was used for startup, with a current density of 1.0 A / cm². -2 .

[0120] The following section describes a complete step to illustrate the model results when a pressure failure occurs. The time step is represented by Δt, where 10s represents the normal operating condition and 10+Δts represents the pressure failure condition. The calculation process from 10s to 10+Δts is used as an example to calculate some components:

[0121] Battery status after 10 seconds under normal conditions:

[0122] Fuel cell output voltage at 10s: V 10s =0.720623V;

[0123] The water content of the cathode catalyst layer membrane is:

[0124] The water content of the anode catalyst layer membrane is:

[0125] The water content of the proton exchange membrane is:

[0126] Volume fraction of liquid water in the cathode catalyst layer:

[0127] The hydrogen concentration in the anode catalyst layer is

[0128] The oxygen concentration in the cathode catalyst layer is:

[0129] The temperature of the anode catalyst layer is

[0130] The temperature of the cathode catalyst layer is

[0131] The temperature of the proton exchange membrane is

[0132] Based on the above calculation expression, the transient response of the membrane water, liquid water, gas components, temperature, and battery performance at 10+Δts is calculated iteratively.

[0133] The calculation shows that when a pressure failure occurs at time 10+Δt:

[0134] Fuel cell output voltage: V (10+Δt)s =0.648861V;

[0135] The water content of the cathode catalyst layer membrane is:

[0136] The water content of the anode catalyst layer membrane is:

[0137] The water content of the proton exchange membrane is:

[0138] Volume fraction of liquid water in the cathode catalyst layer:

[0139] The hydrogen concentration in the anode catalyst layer is

[0140] The oxygen concentration in the cathode catalyst layer is:

[0141] The temperature of the anode catalyst layer is

[0142] The temperature of the cathode catalyst layer is

[0143] The temperature of the proton exchange membrane is

[0144] By solving the above equations, a fault-embedded battery model with high computational efficiency can be established. Based on the fault function defined for the fuel cell, as well as performance parameters and operating environment conditions, the dynamic changes of parameters such as fuel cell output voltage, reactant concentration, temperature distribution, and liquid water volume fraction can be solved and compared with experimental data.

[0145] Figure 2 and Figure 3 This is the result of validating the model against experimental data, which verifies the model's accuracy.

[0146] Figure 4 , Figure 5 , Figure 6 as well as Figure 7 These are the curve results calculated from the model implementation example. The output voltage variation curves under different fault embedding conditions in the implementation example show a high degree of consistency between the simulation data and experimental data. Under the three fault conditions of pressure fault, humidity fault, and flow fault, the provided fault embedding model can accurately simulate the performance changes of the fuel cell.

[0147] in Figure 7 The curves showing the change in oxygen concentration in the cathode catalyst layer under the pressure fault function embedding condition are presented. This demonstrates that the proposed model can simulate the complex physical field changes inside the fuel cell under the corresponding fault conditions. This is of great value for understanding the complex heat and mass transfer processes and dynamic responses inside the fuel cell under different fault conditions.

[0148] This model achieves accurate simulations of pressure, humidity, and flow rate fault conditions through fault function embedding. While meeting the actual operating requirements of fuel cells, it provides insights into the complex heat and mass transfer processes within the fuel cell under different fault conditions and the dynamic responses to various transient conditions. Through electrochemical model calculations, solutions for membrane water content, liquid water content, gas component concentrations, and temperature equations, the complex physical field changes within the fuel cell under fault conditions are determined. This addresses the current problems of unbalanced fault data types and poor data quality in fault diagnosis research, enabling the accumulation of various fault data and providing model support for the development of fuel cell fault diagnosis algorithms, significantly shortening product development cycles and reducing bench testing costs.

Claims

1. A method for establishing a fuel cell failure embedded model, characterized by: The method for model establishment includes six parts of setting and solving: fault function embedding, electrochemical model calculation, membrane water content solving, liquid water content solving, gas component concentration solving and temperature equation solving, and the specific steps are as follows: (1) Fault function embedding The fault function is embedded into the inlet operating parameters to accurately describe the characteristics of the fault, simulate the working conditions of pressure fault, humidity fault and flow fault, and the pressure fault is characterized by the pressure drop at the anode and cathode inlet. When the pressure fault occurs, it can be quantitatively described as: P in,a = f P,a (t)P inlet,a (1) P in,c = f P,c (t)P inlet,c (2) where P in,a represents the actual inlet pressure of the fuel cell anode, P in,c represents the actual inlet pressure of the fuel cell cathode; t represents time; f P,a (t) represents a pressure fault function of the anode, f P,c (t) represents a pressure fault function of the cathode; P inlet,a represents a normal inlet pressure of the anode set operation, P inlet,c represents a normal inlet pressure of the cathode set operation, Because of the lack of humidity, water management is a common fault in fuel cell systems, which can cause membrane drying of the proton exchange membrane and directly affect the water vapor concentration at the fuel cell inlet. When the humidity fault occurs, it can be quantitatively described as: where c vap,in,a represents the inlet water vapor concentration of the fuel cell anode, c vap,in,c represents the inlet water vapor concentration of the fuel cell cathode; P sat represents the saturated water vapor pressure; RH in,a represents the actual relative humidity of the anode, RH in,c represents the actual relative humidity of the cathode; R represents the universal gas constant; T represents the ambient temperature; f RH,a (t) represents the humidity failure function of the anode, f RH,c (t) represents the humidity failure function of the cathode; RH inlet,a represents the normal relative humidity of the anode set operation, RH inlet,c represents the normal relative humidity of the cathode set operation, The lack of reactant flow can be used to characterize the gas leakage in the gas supply pipeline and the insufficient flow fault of the gas supply under variable load conditions. When the flow fault occurs, it can be quantitatively described as: q in,a = c in,a f q,a (t)v a (5) q in,c = c in,c f q,c (t)v c (6) where q in,a represents the anode inlet gas flow rate, q in,c represents the cathode inlet gas flow rate; c in,a represents the anode inlet gas concentration, c in,c represents the cathode inlet gas concentration; f q,a (t) represents the anode flow fault function, f q,c (t) represents the cathode flow fault function; v a represents the anode inlet gas flow rate, v c represents the cathode inlet gas flow rate, (2) Electrochemical model calculation The output voltage calculation expression of the fuel cell is as follows: V = E rev -η ohm -η act,a -η act,c (7) where V represents the fuel cell output voltage; E rev represents the reversible voltage loss; η ohm represents the ohmic loss; η act,a represents the activation loss of the anode; η act,c represents the activation loss of the cathode, where the calculation expression of the reversible loss is as follows: where F represents Faraday's constant; hydrogen partial pressure representing an anode flow channel, oxygen partial pressure representing a cathode flow channel, and the calculation expression of ohmic loss is as follows: where I represents fuel cell current density; d represents flow channel thickness; δ GDL represents gas diffusion layer thickness; δ MPL represents microporous layer thickness; δ CL represents catalyst layer thickness; δ MEM represents proton exchange membrane thickness; σ BP represents bipolar plate conductivity; represents gas diffusion layer electronic conductivity; represents microporous layer electronic conductivity; represents catalyst layer electronic conductivity; represents catalyst layer proton conductivity; σ MEM represents proton exchange membrane conductivity, The activation loss is divided into anode activation loss and cathode activation loss, and the calculation expression is as follows: wherein a represents a coefficient; represents the anode reference exchange current density, represents the cathode reference exchange current density; s represents the liquid water content; represents the anode flow channel hydrogen concentration, represents the cathode flow channel oxygen concentration; represents the anode reference hydrogen concentration, represents the cathode reference oxygen concentration, (3) Membrane water content solving Membrane water exists in the membrane, anode and cathode catalyst layer, and affects the calculation of ion conductivity. The calculation expression of membrane water content is as follows: where p MEM represents the density of the proton exchange membrane; EW represents the equivalent weight of the membrane; ω represents the volume fraction of the polymer; the subscript p represents the type of the water layer in the fuel cell, which includes the cathode catalyst layer, the anode catalyst layer, and the proton exchange membrane; The subscript q represents the type of the adjacent film water existing layer of the p layer, which includes the cathode catalytic layer, the anode catalytic layer, and the proton exchange membrane, λ p represents the film water content of the p layer; λ q δ represents the film water content of the q layer p δ represents the thickness of the p layer q δ represents the thickness of the q layer S represents the film water effective diffusivity between the p and q layers mw δ represents the film water source term, thereby solving for the film water content The diffusion flux of membrane water from the adjacent layer to the p layer is calculated as follows: where φ mw,p-q represents the film water diffusion flux of the adjacent q layer diffusing into the p layer, (4) Liquid water content solving When the water vapor pressure in the fuel cell is greater than the saturated gas pressure, the water vapor will condense into saturated liquid water, which is called liquid water. The calculation formula of saturated water vapor pressure is as follows: P sat The saturated water vapor pressure is represented, assuming that the liquid water is continuous throughout the porous layer, and is solved using the mass conservation equation of liquid water. The calculation formula of the liquid water volume fraction in the porous layer is as follows: where the lower index m represents the type of layer included in the fuel cell, including flow field, anode catalyst layer, cathode catalyst layer, anode gas diffusion layer, cathode gas diffusion layer, anode microporous layer, and cathode microporous layer; the lower index n represents the type of layer adjacent to the m layer, including flow field, anode catalyst layer, cathode catalyst layer, anode gas diffusion layer, cathode gas diffusion layer, anode microporous layer, and cathode microporous layer; p lq represents the density of liquid water; M w represents the molar mass of liquid water; δ m represents the thickness of the m layer; δ n represents the thickness of the n layer; ε m represents the porosity of the m layer; s m represents the volume fraction of liquid water of the m layer; φ lq,m represents the permeation flux of liquid water in the m layer; S lq is the source term for liquid water; represents the effective permeability of liquid water between the m layer and the n layer; p l,n represents the hydraulic pressure of the n layer; p l,m represents the hydraulic pressure of the m layer; μ lq represents the viscosity of liquid water, The permeation flux of liquid water from the adjacent layer to the m layer is calculated as follows: where φ lq,m-n represents the liquid water permeation flux of the adjacent n-layer diffusing to m-layer, liquid pressure p l The capillary pressure p in the porous medium is calculated by the following equation: c The relationship equation between the capillary pressure p and the liquid water volume fraction s is as follows: where σ lq is the surface tension coefficient; θ is the contact angle; ε is the porosity; K is the liquid water permeability; p g is the total gas pressure, whereby the liquid pressure p l at the next time step is obtained from the liquid water volume fraction at the previous time step, and then the liquid water volume fractions s, (5) Gas component concentration solving The gas components inside the fuel cell include water vapor, hydrogen, oxygen and nitrogen. The specific solving formula of gas component concentration in the porous layer is as follows: where the lower index i represents the type of gas component, including hydrogen, water vapor, nitrogen in the anode, and oxygen, water vapor, and nitrogen in the cathode; Dm,n,i represents the effective diffusivity of gas component i between the mth and nth layers; c i,n Cn,i represents the gas concentration of gas component i in the nth layer; c i,m Cm,i represents the gas concentration of gas component i in the mth layer, S i,m Sourceterm represents the source term of gas component i in the mth layer, thereby solving the gas component concentration, The calculation formula of gas diffusion flux is as follows: where φ i,m-n represents the diffusive flux of gas component i from layer n to layer m, (6) Temperature solving Local temperature is solved by energy conservation equation, and the calculation expression is as follows: where T m represents the temperature of the mth layer; T n represents the temperature of the mth layer; pCp m represents the effective volumetric heat capacity of the mth layer; S T represents the heat source term; represents the effective thermal conductance between the mth and nth layers, The heat flux calculation formula is: where φ T,m-n represents the heat flux transferred from the adjacent n layer to the m layer, The model solving adopts explicit format updating algorithm for calculation. Through the embedding of the above fault function and the solving of the equation, the fault embedded fuel cell model is established. According to the setting of fuel cell fault function, performance parameters and working environment conditions, the dynamic changes of fuel cell performance and internal parameters under fault working conditions can be solved, so as to realize the accumulation of various fault data and accelerate the fault diagnosis and discrimination.

Images