A fuel cell system variable scale performance prediction method combined with a neural network

Abstract

The application discloses a kind of variable scale performance prediction methods of fuel cell system combined with neural network, construct the proton exchange membrane fuel cell model combined with neural network algorithm, realize the variable scale performance prediction of proton exchange membrane fuel cell system, and the proton exchange membrane fuel cell model includes air supply module, hydrogen supply module, water-thermal management module, battery stack module and neural network fitting module;Actual output power of fuel cell system is calculated;After all module combination connection, it is constituted into a variable scale numerical simulation model of proton exchange membrane fuel cell system combined with neural network algorithm.The application combines the advantages of neural network and mathematical modeling, uses limited amount of data resources, computer resources and time cost, completes high-precision simulation modeling of proton exchange membrane fuel cell, to provide guidance for the application of proton exchange membrane fuel cell in aviation field.

Description

Technical Field

[0001] This invention belongs to the field of aviation and relates to a proton exchange membrane fuel cell, specifically a variable-scale performance prediction method for a proton exchange membrane fuel cell system that combines neural network algorithms. Background Technology

[0002] Since the beginning of this century, in order to reduce carbon emissions globally, the aviation industry has increasingly pursued clean energy, with electrification and all-electric aircraft becoming one of the mainstream trends. Within the existing technological framework, proton exchange membrane fuel cells (PEMFCs) have emerged as one of the most likely aircraft electrification technologies to be realized in the short term. However, given the stringent safety requirements of air vehicles, PEMFCs must undergo a complex and rigorous safety assessment process before they can be used as part of airborne equipment. To achieve this, establishing an accurate PEMFC model is particularly crucial.

[0003] The core objective of modeling is to accurately and rapidly simulate the performance response of fuel cells under environmental factors. While finite element modeling can provide accurate and comprehensive results, it often requires significant computing resources and time, and is generally not considered a primary approach to system modeling. Furthermore, since research on the aerospace applications of proton exchange membrane fuel cells is still in its early stages, researchers struggle to acquire sufficient data to support the construction of purely data-driven models. Simultaneously, a single mathematical model often cannot accurately describe all the external conditions that a proton exchange membrane fuel cell may encounter during flight. Therefore, it is crucial to develop efficient, accurate, and comprehensive simulation models of proton exchange membrane fuel cells using limited resources, thereby providing guidance for their application in the aerospace field. Summary of the Invention:

[0004] The purpose of this invention is to provide a variable-scale performance prediction method for fuel cell systems that combines neural networks. By combining the advantages of neural networks and mathematical modeling, and using limited data resources, computer resources and time costs, a high-precision simulation model of proton exchange membrane fuel cells can be completed.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A method for predicting the variable-scale performance of a fuel cell system using neural networks is characterized by constructing a proton exchange membrane fuel cell model incorporating neural network algorithms to achieve variable-scale performance prediction of the proton exchange membrane fuel cell system. The proton exchange membrane fuel cell model includes an air supply module, a hydrogen supply module, a hydrothermal management module, a fuel cell stack module, and a neural network fitting module. The specific steps are as follows:

[0007] Step 1: Air supply module, including air compressor, flow control system and humidifier.

[0008] The air compressor module is defined as a gas property control module used to set specific pressure and flow signals for the air in the pipeline. The flow rate at a given target pressure is determined by the flow control system.

[0009] rpm = k1·OER·IM act (1)

[0010] M set =k2·M des (2)

[0011] M des =f(rpm,π,p,T) (3)

[0012] Where rpm is the rotational speed, OER is the excess air coefficient, I is the fuel cell current, and M is the fuel cell current. act It represents the current traffic, M. set This sets the flow rate, where k1 and k2 are the speed correction coefficient and flow rate correction coefficient, respectively, and are related to the number of single batteries and the current temperature and pressure. M des Determined by the characteristics of the compressor used, it is a function of rpm, pressure ratio π, temperature T, and pressure p.

[0013] The humidifier module is used to change the proportion of water in the air duct. This is done by comparing the relative humidity of the air before and after the module, and then calculating the mass flow rate of the added water to simulate the humidification process.

[0014] M water_in =k3·(RH) out -RH in (4)

[0015] Where k3 is the humidification correction factor and RH is the relative humidity.

[0016] Step 2: Hydrogen supply module, including hydrogen tank, pressure reducing valve, and hydrogen recirculation device.

[0017] A hydrogen tank is defined as an adiabatic gas chamber that satisfies the laws of conservation of energy and mass inside the chamber and has a matter-energy exchange interface with the outside world to provide hydrogen signals to the hydrogen pipeline.

[0018] Intra-mesh mass conservation:

[0019]

[0020] Energy conservation:

[0021]

[0022] Exchange of matter with the outside world:

[0023]

[0024] Energy exchange with the outside world:

[0025] Q Ports =Φ A (8)

[0026] Where V0 is the cavity volume, ρ is the density, p is the pressure, t is the time, T is the temperature, and x i Let M be the mass fraction, h be the enthalpy, and c be the mass fraction. p Φ represents the isobaric specific heat capacity, Q represents heat, and Φ represents the specific heat capacity. A For internal heat sources, the subscript Ports indicates the interface, and Cond indicates the interior of the cavity.

[0027] A pressure reducing valve is defined as a valve that limits the outlet area, controlling the hydrogen pressure signal at the outlet by setting the throttling area.

[0028] A min =f(p act ,p set (9)

[0029] Where, p act It is actual pressure, p set It sets the pressure.

[0030] The hydrogen recirculation unit extracts a portion of the gas stream from the exhaust gas based on the fuel cell stack current, and this stream, along with the hydrogen supplied from the hydrogen tank, enters the humidifier module.

[0031] M H2_rec =f(I) (10)

[0032] Hydrogen humidifiers are consistent with oxygen humidifiers, that is...

[0033] M water_in =k3·(RH) out -RH in (11)

[0034] Step 3: The water and heat management module consists of a cooling tank, radiator, heat exchanger, and water pump.

[0035] The cooling tank is used to transmit the property signals of the coolant into the coolant piping;

[0036] The radiator is configured as a heat exchange device. Coolant and external media are located on opposite sides of the radiator, and heat flow signals are transmitted between the coolant and other media through the radiator. The heat exchange process includes convection and conduction.

[0037] Thermal convection:

[0038]

[0039] Heat conduction:

[0040]

[0041] Where Re is the Reynolds number, area is the heat transfer cross-sectional area, and D h Where is the hydraulic diameter, Pr is the Planck number, k is the thermal conductivity, and ΔT is the temperature difference between the coolant and the radiator wall.

[0042] Step 4: Battery stack module, including flow channel assembly and membrane electrode assembly.

[0043] The flow channel assembly is a cavity containing a gas mixture, within which energy and mass conservation are satisfied, as well as the condensation changes of the gas. It has four interfaces connecting to the outside world, respectively transmitting the mass fraction signal of the internal substance to the membrane electrode assembly, the mass fraction signal of the outflowing substance, the temperature signal of the outflowing substance, and the heat flow signal to the cooling system.

[0044] The energy and mass conservation equations are similar to those for hydrogen tanks, and the governing equation for condensation behavior is:

[0045]

[0046] x H2O It is the mass fraction of water, x sat tau_c is the mass fraction of water at saturation, and tau_c is the condensation time constant of water.

[0047] Membrane electrode assemblies are used to realize the electrochemical reaction process of hydrogen and oxygen, including voltage output, electrochemical heat and water transfer.

[0048] The overall electrochemical reaction can be expressed as:

[0049]

[0050] The output voltage of the fuel cell stack equals the theoretical voltage minus the actual voltage loss, which includes activation loss, ohmic loss, and concentration loss. The actual output voltage can be calculated using this formula:

[0051] V stack =V nernst -V act -V ohmic -V conc (16)

[0052] In equation (16), the theoretical voltage V nernst for:

[0053]

[0054] Where Tst For the fuel cell stack temperature, This represents the partial pressure of the corresponding substance;

[0055] In equation (16), the activation loss V act for:

[0056]

[0057] α is the charge transfer coefficient, i cell i is the stack current, and i0 is the exchange current density;

[0058] In equation (16), the ohmic loss V ohmic for:

[0059] V ohmic =i cell ·Rohm(19)

[0060] Where Rohm is the resistance, and its value is T. mem / σ, where σ is the conductivity of the membrane, which is related to the stack temperature T. st The relative humidity (RH) and the water content (λ) of the membrane are related.

[0061] In equation (16), the concentration loss V conc for:

[0062]

[0063] Where i L This represents the limiting current density.

[0064] The energy represented by electrochemical heat can be expressed as the difference between the theoretical voltage and the actual voltage, i.e.

[0065] Heat = (V theorv -V stack )·i cell ·area_cell (21)

[0066] Where area_cell is the single-cell reaction area, V stack V is calculated from equation (16). theorv The calculations take into account the higher heating value of hydrogen and the heat of vaporization of water, i.e.

[0067]

[0068] Where N cell It refers to the number of individual batteries. It is the higher heating value of hydrogen, 286 kJ / mol, h. w0 It is the heat of vaporization of water, determined by temperature and pressure. It is the molar mass of water, with a value of 18.015 kg / mol.

[0069] Water transfer n Water Including water diffusion n drag Electroosmotic drag of water drag ,Right now

[0070] n Water =n drag -n diff (twenty three)

[0071] The specific calculation formula is as follows:

[0072]

[0073] Where D H2O It is the water diffusion coefficient, C ccl C acl Let be the concentrations of water at the cathode and anode, respectively, which can be calculated using the following formulas:

[0074]

[0075] Where ρ mem It is the density of the membrane, M mem λ is the molar mass of the membrane, and λ is the water content of the membrane.

[0076] The electroosmotic drag of water can be calculated by the following formula:

[0077] n drag =0.0029·λ mem 2 +0.05·λ mem (28)

[0078] Step 5: Neural Network Fitting Module

[0079] The neural network fitting system is defined as a data-driven model. It is used to solve the problem that mathematical models cannot quickly simulate certain complex operating conditions. The system obtains current and voltage signals from the fuel cell module, corrects them using the neural network fitting results, and then transmits them back to the fuel cell module. The method is as follows: a function U is established using neural network fitting. der = f(I,Φ1,Φ2,...), where Φ represents external influencing factors, such as (tilt, vibration, impact, etc.).

[0080] Based on experimental data, a function is established to represent the influence of external influencing factors on voltage under different current densities, which is difficult to simulate, thereby enabling rapid response capabilities.

[0081] Let I, Φ1, Φ2, ... be the "response" in the neural network fitting, and U be the "predictor variable." Determine the percentage of training data Tr and the number of training layers N. After training, the fitting function U = f(I, Φ1, Φ2, ...) can be obtained. The quotient of U0 = f(I, 0, 0, ...) is U. der :

[0082]

[0083] Step 6: Calculate the actual output power of the fuel cell system

[0084] The main power-generating component in a fuel cell system is the fuel cell stack, while the power-consuming components are the compressor and water pump. Therefore, the actual output power of the system equals the stack power minus the component power, plus the effect of the tilt angle. This can be calculated using the following formula:

[0085] P act =U der ·U theo IW com -W pum (30)

[0086] Among them U theo W is the theoretical stack voltage without considering external factors. com For the compressor's power consumption, W pum This is for the power consumption of the water pump.

[0087] Step 7: After all modules are combined and connected, a variable-scale numerical simulation model of a proton exchange membrane fuel cell system incorporating neural network algorithms can be formed.

[0088] The beneficial effects of this invention are:

[0089] This invention combines the advantages of neural networks and mathematical modeling, and uses limited data resources, computer resources and time costs to complete high-precision simulation modeling of proton exchange membrane fuel cells, thereby providing guidance for the application of proton exchange membrane fuel cells in the aerospace field. Attached Figure Description

[0090] Figure 1 This is a schematic diagram illustrating the principle of variable-scale modeling for a proton exchange membrane fuel cell system that incorporates neural network algorithms. Detailed Implementation

[0091] A variable-scale performance prediction method for proton exchange membrane fuel cell systems, incorporating neural network algorithms, is proposed. A proton exchange membrane fuel cell model integrating neural network algorithms is constructed to achieve variable-scale performance prediction of the system. The proton exchange membrane fuel cell model includes an air supply module, a hydrogen supply module, a hydrothermal management module, a fuel cell stack module, and a neural network fitting module. The method considers the influence of tilt angle and ambient temperature, humidity, and pressure on system performance. Figure 1 This is a schematic diagram illustrating the principle of variable-scale modeling of a proton exchange membrane fuel cell system using a neural network algorithm. The method for constructing a fuel cell system model coupled with neural network fitting is shown below:

[0092] Step 1: Air supply module.

[0093] The air compressor module is defined as a gas property control module used to set specific pressure and flow signals for the air in the pipeline. The flow rate at a given target pressure is determined by the flow control system, which controls the flow rate by adjusting the compressor speed.

[0094] According to Faraday's law, the theoretical airflow required for the current can be calculated:

[0095]

[0096] Where OER is the excess air coefficient, N cell It refers to the number of single batteries, M. set F is the set flow rate, and F is the Faraday constant.

[0097] The difference between the theoretical and actual airflow rates, after being processed by a PID controller with a proportional gain of 5 and an integral gain of 0.5, is multiplied by the compressor's maximum speed of 3600 rpm to obtain the required airflow rate.

[0098]

[0099] Where rpm is rotational speed, M act This is the current traffic.

[0100] The compressor's output flow rate can be obtained from the compressor characteristic diagram based on the boost ratio, speed, temperature, and pressure.

[0101] M des =f(rpm,π,p,T) (33)

[0102] A humidifier module is built to change the water content in the air duct. This is done by comparing the relative humidity of the air before and after the module, and then simulating the humidification process by adding water at a specific mass flow rate.

[0103]

[0104] Where M water_in It is the mass flow rate of the added water, and RH is the relative humidity of the water.

[0105] Step 2: Hydrogen supply module.

[0106] A hydrogen tank is defined as an adiabatic gas chamber that satisfies the laws of energy and mass conservation inside the chamber and has a matter-energy exchange interface with the outside world to provide hydrogen signals to the hydrogen pipeline.

[0107] Mass conservation within the cavity:

[0108]

[0109] Energy conservation:

[0110]

[0111] Exchange of matter with the outside world:

[0112]

[0113] Energy exchange with the outside world:

[0114] Q Ports =Φ A (38)

[0115] Where V0 is the cavity volume, ρ is the density, p is the pressure, t is the time, T is the temperature, and x i Let M be the mass fraction, h be the enthalpy, and c be the mass fraction. p Φ represents the isobaric specific heat capacity, Q represents heat, and Φ represents the specific heat capacity. A For internal heat sources, the subscript Ports indicates the interface, and the subscript Cond indicates the interior of the cavity.

[0116] The pressure reducing valve is defined as a valve that limits the outlet area, controlling the hydrogen pressure signal at the outlet by setting the throttling area.

[0117]

[0118] Where, p act It is actual pressure, p set It sets the pressure, r min and r max These are the minimum and maximum throttling areas, p range It represents the percentage change in maximum pressure, and D is the average diameter of the channel.

[0119] At this point, the outlet pressure of the pressure reducing valve can be calculated as follows:

[0120]

[0121] Where M is the mass flow rate, rat area The ratio of the throttling area to the port area, rat ρ ρ is the ratio of the density at the throttling point to the density at the port. A The density at the throttling point.

[0122] A hydrogen recirculation device is established to recover hydrogen from the anode exhaust gas based on the fuel cell stack current. This recovered hydrogen, along with hydrogen supplied from the hydrogen tank, enters the humidifier module.

[0123]

[0124] Where I is the stack current and area_cell is the single cell reaction area.

[0125] Hydrogen humidifiers are consistent with oxygen humidifiers, that is...

[0126]

[0127] Step 3: Water and Heat Management Module

[0128] A cooling tank is installed to transmit the coolant's property signals into the coolant piping.

[0129] The radiator is configured as a heat exchange device. Coolant and external media are located on opposite sides of the radiator, and heat flow signals are transmitted between the coolant and other media via the radiator. The heat exchange process includes convection and conduction.

[0130] Thermal convection:

[0131]

[0132] Where Re is the Reynolds number and Pr is the Planck number, which can be calculated by the following formula:

[0133]

[0134] Pr=μ·c p ·k (45)

[0135] Where area is the flow cross-sectional area, D h Where is the hydraulic diameter, k is the thermal conductivity, μ is the dynamic viscosity, and c is the viscosity. p ΔT is the isobaric specific heat capacity, and ΔT is the temperature difference between the coolant and the radiator wall. ΔT can be calculated using the following formula:

[0136] ΔT=(T wall -T l )×(1-e -NTU (46)

[0137] Where T wall T is the wall temperature. lWhere is the fluid temperature, and NTU is the number of heat transfer units, calculated by the following formula:

[0138]

[0139] Where surface is the heat transfer area and Nu is the Nusselt number, which can be calculated by the following formula:

[0140]

[0141] Where f is the friction coefficient, which is determined by the properties of the fluid and the wall.

[0142] Heat conduction:

[0143]

[0144] Step 4: Battery stack module.

[0145] The flow channel assembly is established as a cavity containing a gas mixture, within which energy and mass conservation, as well as the condensation changes of the gas, are satisfied. It has four interfaces connecting to the outside world, respectively transmitting the mass fraction signal of the internal substance, the mass fraction signal of the outflowing substance, the temperature signal of the outflowing substance, and the heat flow signal to the cooling system.

[0146] The energy and mass conservation equations are similar to those for hydrogen tanks, and the governing equation for condensation behavior is:

[0147]

[0148] Where, x H2O It is the mass fraction of water, x sat tau_c is the mass fraction of water at saturation, determined by the temperature and pressure at that point, and is obtained in the model by referring to the saturated water vapor temperature and pressure table. tau_c is the condensation time constant of water, which is set to 1 in this model.

[0149] Membrane electrode assemblies are used to realize the electrochemical reaction process of hydrogen and oxygen, including voltage output, electrochemical heat and water transfer.

[0150] The overall electrochemical reaction can be expressed as:

[0151]

[0152] Its output voltage equals the theoretical voltage minus the actual voltage loss, which includes activation loss, ohmic loss, and concentration loss. The actual stack output voltage can be calculated using this formula:

[0153] V stack =V nernst -V act -V ohmic -V conc (52)

[0154] In equation (52), the theoretical voltage V nernst for:

[0155]

[0156] Where T st For the fuel cell stack temperature, This represents the partial pressure of the corresponding substance;

[0157] In equation (52), the activation loss V act for:

[0158]

[0159] Where α is the charge transfer coefficient, i cell i is the stack current, and i0 is the exchange current density;

[0160] In equation (52), the ohmic loss V ohmic for:,

[0161] V ohmic =i cell ·Rohm (55)

[0162] Where Rohm is the resistance, and its value is T. mem / σ, where σ is the conductivity of the membrane, which is related to the stack temperature T. st The relative humidity (RH) and the water content (λ) of the membrane are related.

[0163] In equation (52), the concentration loss V conc for:

[0164]

[0165] Where i L This represents the limiting current density.

[0166] The energy represented by electrochemical heat can be calculated from the difference between the theoretical voltage and the actual voltage, i.e.

[0167] Heat = (V theorv -V stack )·I·area_cell (57)

[0168] Where area_cell is the single-cell reaction area, V stack Calculated using the above formula, V theorv The calculations take into account the higher heating value of hydrogen and the heat of vaporization of water, i.e.

[0169]

[0170] Where N cell It refers to the number of individual batteries. This is the higher heating value of hydrogen, which is 286 kJ / mol. w0 It is the heat of vaporization of water, determined by temperature and pressure. It is the molar mass of water, with a value of 18.015 kg / mol.

[0171] Water transfer n Water Including water diffusion n drag Electroosmotic drag of water drag ,Right now

[0172] n Water =n drag -n diff (59)

[0173] The diffusion of water can be calculated using the following formula:

[0174]

[0175] Where D H2O It is the water diffusion coefficient, C ccl C acl The concentrations of water at the cathode and anode are calculated as follows:

[0176]

[0177]

[0178] Where ρ mem It is the density of the membrane, M mem λ is the molar mass of the membrane, and λ is the water content.

[0179] The electroosmotic drag of water can be calculated by the following formula:

[0180] n drag =0.0029·λ mem 2 +0.05·λ mem (64)

[0181] The method for calculating water content is as follows:

[0182]

[0183] Step 5: Neural Network Fitting Module

[0184] Define the neural network fitting system as a data-driven model. Use the neural network fitting to establish a function U. der = f(I,Til1,Til2,...), where Til represents different tilt angles.

[0185] Based on experimental data, voltage data of the fuel cell system under different tilt angles and different currents were obtained.

[0186] Set I, Til1, Til2, ... as the "response" in the neural network fitting, set U as the "predictor variable", and determine the percentage of training data Tr = 70% and the number of training layers N = 20.

[0187] After training, the neural fitted function U = f(I, Til1, Til2, ...) can be obtained. Its quotient with U0 = f(I, 0, 0, ...) is U0. der ,Right now:

[0188]

[0189] The model uses current loading, therefore, after considering the effect of the tilt angle, the actual stack output voltage can be calculated as follows:

[0190] U act =U der ·U theo (67)

[0191] Among them U act U is the actual output voltage of the fuel cell stack. thoe This represents the theoretical output voltage of the fuel cell without considering the tilt angle.

[0192] Step 6: Calculate the actual output power of the fuel cell system

[0193] The main power-generating component in a fuel cell system is the fuel cell stack, while the power-consuming components are the compressor and the water pump. The power output of the compressor can be calculated using the following formula:

[0194]

[0195] Among them, W com The power consumed by the compressor is determined by the pressures before and after the compressor (p1, p2), the initial pressure (T1), and the mass flow rate (M). pum The power consumption of the water pump is determined by the mass flow rate M and the inlet and outlet pressure p. out p in and average density ρ ave Decide.

[0196] Therefore, the actual output power of the system equals the power of the fuel cell stack minus the power of the components, plus the effect of the tilt angle. This can be calculated using the following formula:

[0197] P act =U der ·U theo IW com -W pum (70)

[0198] Step 7: Once all components are assembled, a variable-scale simulation model of a proton exchange membrane fuel cell system incorporating neural network algorithms can be created.

[0199] This invention utilizes limited resources to build an efficient, accurate, and comprehensive simulation model of proton exchange membrane fuel cells, thereby providing guidance for the application of proton exchange membrane fuel cells in the aerospace field.

Claims

1. A method for predicting the variable-scale performance of a fuel cell system using neural networks, characterized in that, A proton exchange membrane fuel cell model incorporating neural network algorithms was constructed to predict the variable-scale performance of the proton exchange membrane fuel cell system. The model includes an air supply module, a hydrogen supply module, a hydrothermal management module, a fuel cell stack module, and a neural network fitting module. The specific steps are as follows: Step 1: The air supply module includes an air compressor, a flow control system, and a humidifier; Air compressors are used to set a certain pressure and flow signal for the air in the pipeline. The flow rate at a certain target pressure is determined by the flow control system. Humidifiers are used to change the proportion of water in air ducts by comparing the relative humidity of the air before and after the humidifier and the mass flow rate of the added water to simulate the humidification process of the air. Step 2: The hydrogen supply module includes a hydrogen tank, a pressure reducing valve, and a hydrogen recirculation device; The hydrogen tank is an insulated gas chamber that satisfies the laws of conservation of energy and mass. It has a matter-energy exchange interface with the outside world to provide hydrogen signals to the hydrogen pipeline. The pressure reducing valve is a valve that can limit the outlet area, and controls the hydrogen pressure signal at the outlet by setting the throttling area; The hydrogen recirculation device extracts a portion of the gas flow from the exhaust gas based on the fuel cell current, and enters the humidifier together with the hydrogen supplied by the hydrogen tank. The hydrogen humidifier is consistent with the oxygen humidifier. Step 3: The water and heat management module includes a cooling tank, radiator, and heat exchanger; The cooling tank is used to transmit the property signals of the coolant into the coolant piping; The radiator is a heat exchange device. The coolant and the external medium are located on both sides of the radiator. The heat flow signal is transmitted through the coolant and other media via the radiator. The heat exchanger completes the heat exchange process, including heat convection and heat conduction. Step 4: The battery stack module includes flow channel components and membrane electrode components; The flow channel assembly is a cavity containing a gas mixture, within which energy and mass conservation are satisfied, as well as the condensation and change of the gas; it has four interfaces connected to the outside world, respectively transmitting the mass fraction signal of the internal substance, the mass fraction signal of the outflowing substance, the temperature signal of the outflowing substance, and the heat flow signal to the cooling system; Membrane electrode assemblies are used to realize the electrochemical reaction process of hydrogen and oxygen, including voltage output, electrochemical heat and water transfer; The output voltage of the fuel cell stack is equal to the theoretical voltage minus the actual voltage loss. The voltage loss includes activation loss, ohmic loss, and concentration loss. The energy represented by electrochemical heat is expressed as the difference between the theoretical voltage and the actual voltage. Water transfer Including water diffusion Electroosmotic dragging of water ; Step 5: The neural network fitting module is a data-driven model used to address the problem that mathematical models cannot quickly simulate certain complex operating conditions. It obtains the current and voltage signals from the fuel cell module, corrects them using the neural network fitting results, and then transmits them back to the fuel cell module. The method is as follows: Use a neural network to fit and establish a function. ,in To represent external influencing factors, based on experimental data, a function is established to represent the degree of influence of external influencing factors on voltage under different current densities, which is not easy to simulate, thereby achieving the ability to respond quickly. Will Let's define it as the "response" in neural network fitting. Set it as the "predictor variable" and determine the percentage of training data. and number of training layers Once training is complete, the fitted function can be obtained. , and The commerce is : ； Step 6: Calculate the actual output power of the fuel cell system; In a fuel cell system, the power-generating component is the fuel cell stack, while the power-consuming components are the compressor and water pump. The actual output power of the system equals the power of the fuel cell stack minus the power of each component, plus the effect of the tilt angle, calculated by the following formula: ; in This is the theoretical stack voltage without considering external factors. The compressor consumes power. The power consumed by the water pump; It uses a neural network to fit and establish a function; ; ; in, The power consumed by the compressor is determined by the pressure before and after. , Initial pressure and mass flow Decide, The power consumed by the water pump is determined by the mass flow rate. Entrance and exit pressure , and average density Decide; Step 7: After all modules are combined and connected, a variable-scale numerical simulation model of a proton exchange membrane fuel cell system incorporating neural network algorithms is formed.

2. The method for predicting the variable-scale performance of a fuel cell system using a neural network according to claim 1, characterized in that, In step 1, the flow rate under a certain target pressure is determined by the flow control system, as follows: ; ; in, It's the rotational speed. It is the excess air coefficient. It is the fuel cell current. This is the current traffic. It sets the traffic volume. , These are the rotational speed correction factor and the flow rate correction factor, respectively, which are related to the number of single cells and the current temperature and pressure. It is determined by the type of compressor used. Boost ratio ,temperature ,pressure The function; The humidifier simulates the humidification process of air as follows: ; in It is the humidification correction factor. It is the relative humidity at the outlet. It is the imported relative humidity.

3. The method for predicting the variable-scale performance of a fuel cell system using a neural network according to claim 1, characterized in that, In step 2, Conservation of mass: ; Energy conservation: ; Exchange of matter with the outside world: ; Energy exchange with the outside world: ; in The volume inside the cavity. For density, For pressure, For time, For temperature, For quality fraction, For quality fraction, For enthalpy, For isobaric specific heat capacity, For heat, Internal heat source, subscript For interface, Inside the cavity; The pressure reducing valve controls the outlet hydrogen pressure signal by setting the throttling area, that is: ; in, It's real pressure. It sets the pressure; The hydrogen recirculation unit extracts a portion of the gas stream from the exhaust gas based on the fuel cell current, and this stream, along with the hydrogen supplied from the hydrogen tank, enters the humidifier. ; Hydrogen humidifiers are consistent with oxygen humidifiers, that is... （11）。 4. The method for predicting the variable-scale performance of a fuel cell system using a neural network according to claim 1, characterized in that, In step 3, Thermal convection: ; Heat conduction: ; in, Let Reynolds number be 1. For heat exchange cross-sectional area, The hydraulic diameter, For Planck number, Thermal conductivity, This refers to the temperature difference between the coolant and the radiator wall.

5. The method for predicting the variable-scale performance of a fuel cell system using a neural network according to claim 1, characterized in that, In step 4, The energy and mass conservation equations are similar to those for hydrogen tanks, and the governing equation for condensation behavior is: ; It is the mass fraction of water. It is the mass fraction of water when saturated. It is the condensation time constant of water; The overall electrochemical reaction is expressed as follows: ; The actual output voltage of the fuel cell stack is calculated using this formula: ; Theoretical voltage for: ; in For the fuel cell stack temperature, This represents the partial pressure of the corresponding substance; Activation loss for: ; It is the charge transfer coefficient. It is the fuel cell current. It is the exchange current density; Ohm loss for: ; in , It is the conductivity of the membrane, which is related to the stack temperature. relative humidity The water content of the membrane related; Concentration loss for: ; in The limiting current density; The energy represented by electrochemical heat is expressed as the difference between the theoretical voltage and the actual voltage, i.e. ; in For the single-cell reaction area, From the formula calculate, The calculations take into account the higher heating value of hydrogen and the heat of vaporization of water, i.e. ; in It refers to the number of individual batteries. It has the higher heating value of hydrogen, 286 kJ / mol. It is the heat of vaporization of water, determined by temperature and pressure. This is the molar mass of water, with a value of 18.015 kg per mole; Water transfer Including water diffusion Electroosmotic dragging of water ,Right now ; The specific calculation formula is as follows: ; in It is the water diffusion coefficient. , The concentrations of water at the cathode and anode are respectively calculated using the following formulas: ; ; ; in It is the density of the membrane. It is the molar mass of the membrane. It is the water content of the membrane; The electroosmotic drag of water is calculated by the following formula: 。

Images