A thermoelectric co-control method and system for methanol reforming hydrogen fuel cells

Abstract

This invention discloses a thermoelectric co-control method and system for methanol reforming hydrogen fuel cells, belonging to the field of fuel cell control. The control method includes: acquiring a real-time operating parameter set of the fuel cell system; calculating the theoretical hydrogen consumption rate of the stack anode based on the instantaneous current of the fuel cell stack and a preset Faraday electrolysis mathematical equation; performing algebraic mapping calculations to generate the transient heat mismatch of the reformer; and synchronously outputting a first control command and a second control command through a hardware control interface based on the transient heat mismatch. This invention does not rely on online identification algorithms, neural networks, or frequent manual recalibration, enabling the feedforward control model to maintain good control accuracy at all stages of the equipment's life cycle.

Description

Technical Field

[0001] This invention relates to the field of fuel cell control, and in particular to a thermoelectric co-control method and system for methanol reforming hydrogen fuel cells. Background Technology

[0002] The methanol reforming hydrogen fuel cell system uses methanol-water solution as feedstock. Hydrogen-rich reformed gas is generated in situ through a methanol-water vapor reforming reaction within the reformer, and then supplied to the proton exchange membrane fuel cell stack for electrochemical power generation. Since the reforming reaction is a strongly endothermic process, the system needs to continuously supply heat to the reformer to maintain the normal operation of the reaction. In existing technologies, the heat required by the reformer mainly comes from two parts: first, the heat released by the catalytic combustion of residual hydrogen in the exhaust gas from the stack anode in the combustion chamber; and second, the heat released by the supplementary combustion of auxiliary methanol in the combustion chamber. Maintaining a dynamic balance between the heat supply and the heat endothermic by the reforming reaction is crucial to ensuring the safe and efficient operation of the system. However, existing control schemes have the following shortcomings:

[0003] On the one hand, there is a lack of real-time and accurate quantification methods for the usable calorific value carried by residual hydrogen in the anode tail gas. It is usually estimated empirically based on a preset fixed ratio. When the system operating conditions deviate from the calibration point, the error between the estimated value and the actual value will increase significantly, leading to subsequent heat allocation decisions being based on distorted data. On the other hand, the adjustment of reformer heat demand changes generally adopts a feedback control method based on temperature sensors. Due to the large thermal inertia of reformer structural components, there is a significant time delay between load changes and temperature sensor detection of temperature deviations. The control action lags severely behind actual demand changes, easily causing overheating of the reformer catalyst bed or insufficient hydrogen supply under sudden load changes. Summary of the Invention

[0004] One of the objectives of this invention is to provide a thermoelectric co-control method for hydrogen fuel cells based on methanol reforming, in order to solve the problems in the prior art where the available calorific value of the anode tail gas can only be estimated empirically on a fixed scale and cannot be accurately quantified in real time; at the same time, the adjustment of the heat demand of the reformer depends on temperature feedback control, which is limited by the thermal inertia of the reformer structural components, and the control action is seriously lagging behind the actual demand changes, resulting in the reforming-combustion thermal balance being disrupted when the load changes suddenly, causing the catalyst bed to overheat or hydrogen starvation.

[0005] This invention is achieved through the following technical solution: a thermoelectric co-control method for a methanol reforming hydrogen fuel cell, comprising the following steps: acquiring a real-time operating parameter set of the fuel cell system, wherein the real-time operating parameter set includes at least: the instantaneous current of the fuel cell stack, the current wall temperature of the reformer, the total methanol inlet flow rate, and the total flow rate of reforming gas entering the fuel cell stack; calculating the theoretical hydrogen consumption rate of the fuel cell stack anode based on the instantaneous current of the fuel cell stack and a preset Faraday electrolysis mathematical equation, and performing a difference calculation between the hydrogen partial pressure flow rate in the total flow rate of reforming gas and the theoretical hydrogen consumption rate, multiplying the difference by a fixed hydrogen lower heating constant, to generate real-time tail gas calorific value data characterizing the unreacted hydrogen energy in the anode tail gas; The current rate of change of the instantaneous current of the fuel cell stack, the current wall temperature of the reformer, and the real-time calorific value of the exhaust gas are input into a preset current-hydrogen production-waste heat demand decoupling model for algebraic mapping calculation to generate the transient heat mismatch of the reformer. Based on the transient heat mismatch, a first control command and a second control command are synchronously output through a hardware control interface. The first control command is used to drive a proportional three-way valve set at the exhaust gas emission end of the anode to adjust the exhaust gas mixing ratio entering the combustion chamber. The second control command is used to drive an auxiliary methanol afterburning pump to adjust the amount of auxiliary methanol afterburning directly entering the combustion chamber, thereby maintaining the reforming-combustion thermal balance of the reformer.

[0006] Furthermore, the real-time operating parameter set may also include: average voltage of individual cells in the fuel cell stack, coolant inlet temperature, coolant outlet temperature, ambient temperature, and system back pressure.

[0007] Furthermore, the total flow rate of the reformed gas is directly measured by a dual-channel mass flow controller before it enters the fuel cell stack. The dual-channel mass flow controller and the proportional three-way valve are connected through the same hardware control bus, and a deterministic time-division multiplexing protocol is used to synchronously transmit and receive real-time operating parameter sets and the first control command.

[0008] Further, the steps for calculating the theoretical hydrogen consumption rate of the stack anode include: obtaining the total number of individual cells in the stack; obtaining the instantaneous current of the stack at the current sampling time; and dividing the product of the total number of individual cells and the instantaneous current of the stack by the product of the Faraday constant and the number of electrons transferred per mole of hydrogen to obtain the theoretical hydrogen consumption rate.

[0009] Furthermore, the equation for calculating the theoretical hydrogen consumption rate is as follows:

[0010] In the formula, The total number of individual cells in the fuel cell stack. This refers to the instantaneous current of the fuel cell stack. is the Faraday constant, with a value of 96485 C / mol. The coefficient 2 in the denominator corresponds to the stoichiometric relationship that 2 moles of electrons are transferred per mole of hydrogen in the hydrogen oxidation half-reaction.

[0011] Further, the step of generating real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas includes: multiplying the total reforming gas flow rate by the hydrogen mole fraction to obtain the total hydrogen supply mole flow rate currently entering the stack anode; subtracting the theoretical hydrogen consumption rate from the total hydrogen supply mole flow rate to obtain the mole flow rate of unreacted hydrogen in the anode tail gas; multiplying the mole flow rate of unreacted hydrogen by the tail gas pipeline physical heat recovery efficiency constant, and then multiplying by the fixed hydrogen lower heating value constant to obtain the real-time tail gas calorific value data.

[0012] Furthermore, the real-time exhaust gas calorific value generation equation can be: where is the total flow rate of reformed gas measured in real time by the dual-channel mass flow controller, is the mole fraction of hydrogen in the reformed gas, i.e., the hydrogen partial pressure concentration coefficient difference term, is the actual residual unreacted hydrogen mole flow rate in the anode exhaust gas under the current operating conditions, is the lower heating constant of hydrogen, is the physical heat recovery efficiency constant of the exhaust gas pipeline, which characterizes the proportion of heat that can actually be recovered and utilized from the exhaust gas, and is the final calculated usable thermal power of the exhaust gas.

[0013] Furthermore, before generating real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas, a boundary anomaly handling step is also included: comparing the average voltage of the stack single cell acquired at the current moment with the theoretical voltage under the corresponding instantaneous current in the pre-stored standard polarization curve; if the average voltage of the stack single cell is lower than the theoretical voltage by more than a preset difference, and the real-time tail gas calorific value data indicates that there is sufficient unreacted hydrogen in the anode tail gas, then it is determined that a water flooding condition has occurred inside the stack.

[0014] Furthermore, after determining that a flooding condition has occurred inside the fuel cell stack, the method also includes: forcibly suspending the subsequent algebraic mapping calculation steps and the steps of synchronously outputting the first control command and the second control command; commanding the exhaust solenoid valve located at the anode exhaust gas emission end to alternately perform fully open and fully closed actions at a frequency of 10 Hz or higher, triggering the pulse purging sequence; continuously monitoring the average voltage of the fuel cell stack until the average voltage of the fuel cell stack recovers to within the preset difference range of the corresponding theoretical voltage in the standard polarization curve, releasing the suspension state, and resuming the execution of the algebraic mapping calculation steps and the steps of synchronously outputting the first control command and the second control command.

[0015] Further, the step of generating the transient heat mismatch of the reformer specifically includes: calculating the current rate of change of the instantaneous current of the fuel cell stack within three consecutive sampling periods using a first-order backward differential algorithm; mapping the current rate of change of the instantaneous current of the fuel cell stack to the target hydrogen production increment in the next time window using a preset constant ratio conversion coefficient, and converting the target hydrogen production increment into feedforward endothermic demand power based on the molar endothermic constant of the methanol steam reforming reaction; extracting the thermal inertia damping coefficient based on the specific heat capacity of the reformer structural components, subtracting the real-time tail gas calorific value data from the feedforward endothermic demand power, and multiplying by the thermal inertia damping coefficient to obtain the transient heat mismatch.

[0016] Furthermore, in the step of calculating the current rate of change of the instantaneous current of the fuel cell stack within three consecutive sampling periods using the first-order backward difference algorithm, the equation for calculating the current rate of change of the instantaneous current of the fuel cell stack is as follows:

[0017] ,in, The instantaneous current of the fuel cell stack at the current sampling moment. This represents the instantaneous current of the fuel cell stack two sampling periods prior. The sampling period.

[0018] Furthermore, in the step of converting the target hydrogen production increment into feedforward endothermic power demand, the calculation equation for the feedforward endothermic power demand is as follows: ,in, The conversion factor is a constant ratio. The current rate of change of the instantaneous current in the fuel cell stack is denoted as . is the molar endothermic enthalpy change constant of the methanol vapor reforming reaction.

[0019] Furthermore, the equation for calculating the transient heat mismatch is as follows:

[0020] ,in, The power required for feedforward heat absorption. For real-time exhaust gas calorific value data, is the thermal inertia damping coefficient; a positive value for transient heat mismatch indicates a heat deficit, while a negative value indicates a heat surplus.

[0021] Furthermore, the thermal inertia damping coefficient can be dynamically updated through the following steps:

[0022] Extract the current wall temperature of the reformer Temperature at the center of the catalyst bed located inside the reformer Within a preset sliding time window, the difference between the current wall temperature of the reformer and the center temperature of the catalyst bed is calculated using discrete-time integration to obtain the integral value of the thermal gradient. The calculation equation is as follows: ,in, This represents the number of sampling points included within the sliding time window. The sampling period is specified; the integral value of the thermal gradient is substituted into a preset polynomial fitting function to output the thermal inertia damping coefficient. The calculation equation is as follows:

[0023] ,in, , , These are the heat transfer attenuation polynomial coefficients pre-calibrated for the reformer.

[0024] Furthermore, in the step of synchronously outputting the first control command and the second control command based on the transient heat mismatch, the following mutually exclusive operating condition branches are executed according to the numerical range of the transient heat mismatch: when the transient heat mismatch is positive and greater than the first preset threshold, it is determined that the load is in a sudden increase condition; when the transient heat mismatch is negative and the absolute value is greater than the second preset threshold, it is determined that the load is in a sudden decrease condition.

[0025] Furthermore, under conditions of sudden load increase, the method further includes: setting the duty cycle of the first control command to its maximum value, driving the proportional three-way valve to guide all anode exhaust gas to the combustion chamber; and calculating the target afterburning flow rate by dividing the transient heat mismatch by the methanol combustion calorific value conversion coefficient. The calculation equation is as follows: ,in, This refers to the lower heating value of methanol. The thermal efficiency coefficient of the burner. To maintain a fixed displacement per stroke of the auxiliary methanol combustion pump, the pulse frequency of the second control command is dynamically matched to the target combustion flow rate, driving the auxiliary methanol combustion pump to inject liquid methanol into the combustion chamber.

[0026] Furthermore, under the condition of sudden load drop, the method further includes: forcibly resetting the second control command to zero and shutting off the output of the auxiliary methanol combustion pump; and calculating the exhaust gas bypass ratio based on the ratio between the absolute value of the transient heat mismatch and the real-time exhaust gas calorific value data. The calculation equation is as follows: The system generates a cutoff command as the first control command, which drives the proportional three-way valve to bypass the anode exhaust gas of the corresponding exhaust gas bypass ratio and discharge it to the condensation recovery component, preventing it from entering the combustion chamber.

[0027] Furthermore, under conditions of sudden load reduction, if the absolute value of the transient heat mismatch exceeds the upper limit threshold of the combustion chamber's heat dissipation capacity... To determine if the system is under extreme cooling conditions, the method also includes: calculating the excess residual heat power, which is the absolute value of the transient heat mismatch minus the upper limit threshold of the combustion chamber's heat dissipation capacity; and converting the residual heat power into the cold water pulse injection quantity according to a preset latent heat of vaporization formula. The calculation equation is as follows:

[0028] ,in, Let be the latent heat of vaporization of water. The density of the deionized water is determined; the bypass water valve is opened, and the deionized water is injected into the cooling jacket of the outer layer of the combustion chamber in a pulse jet manner, utilizing the water vaporization phase change process to absorb the residual heat at the bottom of the reformer.

[0029] Furthermore, in the pulse jet mode, the single opening time of the bypass water valve is on the order of microseconds to milliseconds, and the amount of water injected each time is matched with the calculated result of the cold water pulse jet amount.

[0030] Another aspect of the present invention provides a thermoelectric co-control system for a methanol reforming hydrogen fuel cell, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the thermoelectric co-control method for a methanol reforming hydrogen fuel cell as shown above.

[0031] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0032] 1. This invention is based on Faraday's electrolysis mathematical equations to calculate and generate real-time exhaust gas calorific value data. The entire conversion process does not rely on any empirical parameters, statistical fitting models, or pre-assumed fixed proportional coefficients, and has strict physical accuracy. Moreover, within each control sampling cycle, the system can obtain anode exhaust gas usable calorific value data that precisely matches the current operating conditions. This breaks through the technical blind spot of empirical estimation of exhaust gas calorific value by a fixed proportion in the prior art, eliminates the cumulative deviation between empirical estimates and actual values ​​when the system operating conditions deviate from the calibration point, and provides a reliable real-time data foundation for all subsequent heat supply and demand balance calculations.

[0033] 2. This invention uses the instantaneous current change rate of the fuel cell stack as the feedforward drive signal. Through algebraic mapping across physical domains, it achieves millisecond-level advance prediction of heat demand. Within a millisecond calculation time, the change rate signal in the electrical domain is converted into a power demand signal in the thermodynamic domain, generating the transient heat mismatch of the reformer. Through the feedforward control method, the prediction of future heat demand can be completed at the instant of load change, ensuring that the control command always leads the physical changes in the thermal process. This eliminates the control lag problem that usually requires several seconds or even tens of seconds due to the thermal inertia of the reformer structural components.

[0034] 3. The current-hydrogen production-waste heat demand decoupling model adopted in this invention processes the variables in the electrical domain (instantaneous current change rate of the fuel cell stack), the variables in the thermodynamic domain (current wall temperature of the reformer), and the variables in the chemical energy domain (real-time tail gas calorific value data) separately before coupling calculation. This layered decoupling process avoids cross-interference between multiple physical field variables during the calculation process, so that the calculation results of each physical domain have independent and clear physical meaning, ensuring the determinism and predictability of the control logic under various operating conditions. Attached Figure Description

[0035] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0036] Figure 1 This is a flowchart of the method provided in Embodiment 1 of the present invention.

[0037] Figure 2 This is a schematic diagram of the linear mapping from current to theoretical hydrogen consumption rate provided in Embodiment 1 of the present invention.

[0038] Figure 3 This is a timing response comparison diagram provided in Embodiment 1 of the present invention.

[0039] Figure 4 This is a schematic diagram comparing the image stabilization effect provided in Embodiment 1 of the present invention.

[0040] Figure 5 This is a schematic diagram of the working condition partitioning and actuator collaborative allocation logic provided in Embodiment 1 of the present invention.

[0041] Figure 6 This is a comparative schematic diagram showing the effect of catalyst bed temperature control provided in Example 1 of the present invention.

[0042] Figure 7 This is a three-dimensional surface schematic diagram of the available thermal power provided in Embodiment 1 of the present invention.

[0043] Figure 8 This is a contour map showing the available thermal power of exhaust gas provided in Embodiment 1 of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0045] Example 1

[0046] This embodiment discloses a thermoelectric co-control method for methanol-reformed hydrogen fuel cells. Figure 1 The overall method flowchart of this embodiment is shown. As can be seen from the figure, this embodiment includes the following steps:

[0047] Step 1: Obtain the real-time operating parameter set of the fuel cell system. The operating parameter set should include at least the instantaneous current of the fuel cell stack, the current wall temperature of the reformer, the total methanol inlet flow rate, and the total flow rate of reforming gas entering the fuel cell stack.

[0048] A fuel cell system refers to a complete energy conversion system that uses methanol-water solution as raw material, generates hydrogen-rich reformed gas in situ in a reformer through a methanol-water vapor reforming reaction, and supplies this reformed gas to a proton exchange membrane fuel cell (PEMFC) stack for electrochemical power generation. This system includes at least the following core functional components: a methanol-water solution storage tank, a feed pump, an evaporator, a reformer (including a catalyst bed and combustion chamber), a gas purification module, a fuel cell stack, an anode tail gas pipeline, a proportional three-way valve, an auxiliary methanol afterburner pump, and a condensation recovery assembly.

[0049] The real-time operating parameter set refers to a set of physical quantities that are synchronously collected by sensors deployed on key nodes of the fuel cell system during each control sampling cycle. This set of physical quantities can completely describe the electrochemical and thermodynamic operating states of the system at the current moment.

[0050] The instantaneous current of the fuel cell stack refers to the current current at the current sampling time. The stack output current value, measured in real time by a Hall effect current sensor or shunt installed on the main circuit of the fuel cell stack, is denoted as... The unit is ampere (A). This current value directly reflects the current electrochemical reaction intensity of the fuel cell stack and is the core input for calculating the hydrogen consumption rate and judging the load change trend.

[0051] The current wall temperature of the reformer refers to the temperature at the current sampling time. The temperature value measured by a thermocouple or platinum resistance temperature sensor attached to the outer wall of the reformer is denoted as . The unit is degrees Celsius (°C). This temperature value characterizes the outer heat source boundary conditions for heat transfer from the combustion chamber to the reforming reaction zone, and is an important characteristic quantity for evaluating the thermal state of the reformer.

[0052] Total methanol inflow rate refers to the flow rate at the current sampling time. The instantaneous feed flow rate of the methanol-water solution is measured by a mass flow meter or volumetric flow meter installed on the methanol-water solution inlet line. This flow rate value reflects the current fuel supply level of the system.

[0053] Total reformed gas flow rate refers to the flow rate at the current sampling time. The molar flow rate of the mixed gas (mainly hydrogen, carbon dioxide, and a small amount of carbon monoxide) generated after the catalytic reaction in the reformer, before entering the anode inlet of the fuel cell stack, is measured in real time by a dual-channel mass flow controller and denoted as [missing information]. The unit is moles per second (mol / s). This flow rate value is the direct basis for calculating the actual hydrogen supply to the anode of the fuel cell stack.

[0054] In this embodiment, the total flow rate of the reformed gas is directly measured by the dual-channel mass flow controller before entering the fuel cell stack. The dual-channel mass flow controller and the proportional three-way valve are connected through the same hardware control bus, and the set of operating parameters and the first control command are synchronously transmitted and received using a deterministic time-division multiplexing protocol.

[0055] Understandably, adopting a deterministic time-division multiplexing protocol means that in each communication cycle, data acquisition commands and control output commands are allocated to fixed, non-overlapping time slots for transmission, thereby avoiding communication delay jitter caused by bus contention, ensuring that the acquired operating parameters and output control commands are strictly aligned in time, and eliminating control deviations that may be introduced by asynchronous timing.

[0056] In this embodiment, the set of operating parameters may also include, but is not limited to, auxiliary parameters such as the average voltage of a single cell in the fuel cell stack, the inlet temperature of the coolant, the outlet temperature of the coolant, the ambient temperature, and the system back pressure, for subsequent calls to the anomaly diagnosis logic and boundary protection logic.

[0057] Step 2: Based on the instantaneous current and the preset Faraday electrolysis mathematical equation, calculate the theoretical hydrogen consumption rate of the stack anode, and calculate the difference between the hydrogen partial pressure flow rate in the total reforming gas flow and the theoretical hydrogen consumption rate. Multiply the difference by the fixed hydrogen lower heating constant to generate real-time tail gas calorific value data characterizing the unreacted hydrogen energy in the anode tail gas.

[0058] The Faraday electrolysis mathematical equation refers to a deterministic algebraic equation established based on Faraday's law of electrolysis, used to precisely map macroscopically measurable electrical quantities (pile current) to microscopic material consumption (hydrogen molar flow rate). The physical basis of this equation is the conservation of charge in electrochemical reactions—in the anode half-reaction of a proton exchange membrane fuel cell, for every mole of hydrogen molecule (… Oxidation reaction occurs on the surface of the catalyst layer. (This releases exactly two moles of electrons.) Therefore, when the total charge flowing through the external circuit of the fuel cell is known, the total amount of hydrogen consumed by the electrochemical reaction can be unambiguously deduced through stoichiometry, without any empirical parameters or statistical fitting.

[0059] The theoretical hydrogen consumption rate refers to the rate at which hydrogen is consumed at the current sampling time. The molar flow rate of hydrogen consumed by the electrochemical reaction at the anode side of the fuel cell stack, calculated based on the Faraday electrolysis mathematical equation, is denoted as The unit is moles per second (mol / s). The term "theoretical" indicates that this consumption rate is calculated under ideal conditions assuming a Faraday efficiency of 100% (i.e., all electrons passing through the external circuit originate from the hydrogen oxidation reaction).

[0060] Hydrogen partial pressure flow rate refers to the flow rate at the current sampling time. The molar flow rate of hydrogen component in the total reformed gas flow rate entering the anode of the fuel cell stack. Since the stoichiometric ratio of the methanol steam reforming reaction is fixed ( Under ideal complete reforming conditions, the mole fraction of hydrogen in the reformed gas. Based on the above stoichiometric relationships, this can be determined as a system constant. Therefore, the hydrogen partial pressure flow rate is equal to the total reformed gas flow rate. Multiply by the mole fraction of hydrogen .

[0061] The fixed lower heating constant of hydrogen refers to the chemical energy released per mole of hydrogen when hydrogen is completely combusted and water in the products remains in a gaseous state. It is denoted as _____. Its value is 241.8 kJ / mol. The lower heating value is a thermodynamic standard quantity relative to the higher heating value (when water in the product condenses into a liquid state). Using the lower heating value as the conversion benchmark is more in line with the actual operating conditions where water vapor does not condense under high temperature conditions in the combustion chamber.

[0062] Real-time exhaust gas calorific value data refers to the data at the current sampling time. After the above Faraday consumption calculation and difference operation, the usable chemical thermal power carried by the remaining hydrogen gas in the anode tail gas that was not consumed by the electrochemical reaction is finally calculated and denoted as . The unit is kilowatt (kW). This data is the foundational data source for all subsequent heat supply and demand balance calculations.

[0063] In this embodiment, the step of calculating the theoretical hydrogen consumption rate of the fuel cell stack anode specifically includes:

[0064] Obtain the total number of individual cells in the fuel cell stack This total number is a fixed constant of the system and is written into the read-only parameter area of ​​the controller after the fuel cell stack is assembled.

[0065] Get the current sampling time Instantaneous current of fuel cell stack ;

[0066] The product of the total number of individual cells and the instantaneous current is divided by the Faraday constant. The theoretical hydrogen consumption rate is obtained by multiplying (96485 C / mol) by the number of electrons transferred per mole of hydrogen in the hydrogen oxidation reaction (2 moles of electrons / mole of hydrogen). .

[0067] For example, in this embodiment, the theoretical hydrogen consumption rate equation can be:

[0068] ,

[0069] in, This refers to the total number of individual cells in a fuel cell stack. for The measured current of the fuel cell stack at any given time. The coefficient 2 in the denominator corresponds to the hydrogen oxidation half-reaction (96485 C / mol). The stoichiometric relationship of 2 moles of electrons transferred per mole of hydrogen gas in ().

[0070] Understandably, this equation utilizes the fundamental physical law of charge conservation in electrochemical reactions to directly convert macroscopically easily measurable electrical quantities (current) into microscopically difficult-to-measure mass flow rates (hydrogen consumption rate), thereby establishing a deterministic conversion channel from the electrical domain to the chemical domain. The entire conversion process does not rely on any empirical parameters or statistical models and has strict physical accuracy.

[0071] In this embodiment, the step of generating real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas specifically includes:

[0072] Total reformer gas flow Multiply by the mole fraction of hydrogen This yields the total molar flow rate of hydrogen currently entering the anode of the fuel cell stack.

[0073] Subtract the theoretical hydrogen consumption rate from the total hydrogen supply molar flow rate. The molar flow rate of unreacted hydrogen gas remaining in the anode tail gas was obtained.

[0074] Multiply the unreacted hydrogen molar flow rate by the physical heat recovery efficiency constant of the exhaust pipe. Then multiply by the fixed lower heating constant of hydrogen. Real-time exhaust gas calorific value data was obtained. .

[0075] For example, in this embodiment, the real-time exhaust gas calorific value generation equation can be:

[0076]

[0077] in, The total flow rate of reformed gas (mol / s) is measured in real time by a dual-channel mass flow controller. The difference term within square brackets represents the mole fraction of hydrogen in the reformed gas, i.e., the hydrogen partial pressure concentration coefficient. This refers to the actual residual molar flow rate of unreacted hydrogen in the anode tail gas under the current operating conditions. This is the lower heating constant of hydrogen (241.8 kJ / mol). The physical heat recovery efficiency constant of the exhaust gas pipeline represents the proportion of heat that can actually be recovered and utilized from the exhaust gas. The final calculated usable thermal power (kW) of the exhaust gas.

[0078] Understandably, the core logic of the above formula is a matter conservation accounting of how much enters, how much is used, and how much remains: the total amount of hydrogen entering the fuel cell stack is known (flowmeter measurement multiplied by the hydrogen mole fraction), the amount of hydrogen consumed by the electrochemical reaction is precisely calculable (Faraday's law), and the difference between the two is the actual recoverable chemical energy carrier in the anode tail gas. This precise accounting method based on matter conservation completely breaks through the technical blind spot of traditional control methods that rely on proportional empirical estimation of the tail gas calorific value. It provides a solid real-time data foundation for the subsequent realization of on-demand precise utilization of tail gas waste heat. It can complete all calculations within each control clock cycle (millisecond level), meeting the timing requirements of real-time control.

[0079] In this embodiment, before the calculation step in step 2, a boundary anomaly handling logic based on feature comparison is included. This logic is used to eliminate abnormal operating states caused by liquid water accumulation inside the fuel cell stack before the conventional thermoelectric co-control process is started. Specifically, it includes:

[0080] Compare the average voltage of a single cell in the fuel cell stack obtained at the current moment with the theoretical voltage at the corresponding instantaneous current in the standard polarization curve;

[0081] If the average voltage of a single cell is lower than the theoretical voltage by more than a preset difference, and the real-time exhaust gas calorific value data calculated in step 2 indicates that there is a sufficient amount of unreacted hydrogen in the exhaust gas, then it is determined that a severe flooding phenomenon has occurred inside the stack, resulting in gas resistance in the diffusion layer.

[0082] The standard polarization curve refers to the reference performance curve obtained by testing the steady-state current-voltage characteristics of the fuel cell stack under standard test conditions (rated temperature, rated humidity, and rated gas stoichiometry). This curve is pre-stored in the read-only memory of the controller as a reference for the voltage-current relationship under normal operating conditions of the fuel cell stack.

[0083] The preset difference refers to a pre-set voltage deviation threshold, which is determined based on the fuel cell stack type, membrane electrode assembly (MEA) characteristics, and engineering experience. When the deviation of the measured average voltage from the theoretical voltage exceeds this threshold, it indicates that the mass transfer process within the fuel cell stack has been severely hindered.

[0084] Flooding refers to an abnormal operating condition in which liquid water generated on the cathode side or liquid water that seeps back from the cathode to the anode accumulates excessively in the gas diffusion layer or flow channel during the operation of a fuel cell, blocking the channels for the transport of reactant gases to the catalyst layer, resulting in a sharp reduction in the electrochemical reaction area and a significant drop in voltage.

[0085] Understandably, the typical characteristics of a flooded operating condition are low voltage but high hydrogen content in the exhaust gas. The voltage drop is not due to insufficient hydrogen supply (there is still a large amount of unreacted hydrogen in the exhaust gas), but rather because the physical barrier of liquid water prevents the hydrogen supplied to the catalyst layer from effectively participating in the reaction. By cross-referencing this combination of characteristics, the flooded operating condition can be accurately distinguished from the hydrogen-starved operating condition (which also manifests as a voltage drop, but with extremely low hydrogen content in the exhaust gas).

[0086] Under this abnormal operating condition, the method in this step may also include:

[0087] Forcefully suspend the conventional thermoelectric co-control logic of subsequent steps 3 to 4, and immediately trigger the hardware-level pulse purge timing sequence;

[0088] The exhaust solenoid valve at the anode tail gas emission end is commanded to alternately open and close at a high frequency of 10 Hz or higher, using the generated gas pressure pulse water hammer effect to force the liquid water in the reactor to be discharged.

[0089] The average voltage of each cell in the fuel cell stack is continuously monitored until it returns to within the preset difference range of the theoretical voltage corresponding to the standard polarization curve. Only then can the suspended state be lifted and the execution of the normal control logic be resumed.

[0090] Understandably, the pulse purging sequence utilizes the water hammer effect principle in fluid mechanics. The exhaust solenoid valve alternately opens and closes at a frequency higher than 10 Hz, forming periodic pressure pulse waves in the anode channel. These pressure pulses exert an impact force on the liquid water clumps trapped in the channel that is much greater than the shear force of the steady-state airflow. This can quickly blow away the liquid water attached to the microporous structure of the diffusion layer and the corners of the channel and discharge it from the fuel cell stack, thereby restoring the normal transport channel of the reactant gas. Figure 2 This diagram illustrates the linear mapping from current to theoretical hydrogen consumption rate in this embodiment. Figure 2 The deterministic relationship between current and hydrogen consumption rate is demonstrated under different total numbers of individual cells.

[0091] Step 3: Input the current rate of change of instantaneous current, the current wall temperature of the reformer, and the real-time calorific value of exhaust gas into the preset current-hydrogen production-waste heat demand decoupling model for algebraic mapping calculation to generate the transient heat mismatch of the reformer.

[0092] Wherein, the current rate of change of instantaneous current refers to the current sampling time. Instantaneous current of fuel cell stack The rate of change relative to time is denoted as The unit is amperes per second (A / s). This rate of change is the core driving signal of the entire feedforward control logic—it appears in time before any temperature change signal, and can provide an early warning of impending changes in system heat demand the instant the load changes abruptly.

[0093] The current-hydrogen production-waste heat demand decoupling model refers to a deterministic algebraic mapping model pre-built and embedded in the controller. This model takes the instantaneous current change rate of the fuel cell stack, the reformer wall temperature, and real-time exhaust gas calorific value as inputs. Through hierarchical algebraic operations, it outputs the transient heat mismatch of the reformer at the current moment. The term "decoupling" indicates that the model processes the variables in the electrical domain (current change rate), thermodynamic domain (wall temperature), and chemical energy domain (exhaust gas calorific value) separately before coupling them, avoiding cross-interference between multiple physical field variables.

[0094] Algebraic mapping computation refers to the fact that all operations performed within the model consist of basic algebraic operations such as addition, subtraction, multiplication, division, and polynomial evaluation, without involving computationally intensive operations such as iterative solutions, matrix inversion, or numerical integration of differential equations. This ensures deterministic execution time and predictable bus latency on embedded microcontrollers.

[0095] Transient thermal mismatch refers to the amount of thermal mismatch at the current sampling time. The net difference between the heat supply side and the heat demand side of the reformer, calculated using the decoupling model, is denoted as... The unit is kilowatt (kW). When it is positive, it means that the system is in a state of heat shortage, that is, the reformer needs to add extra heat to maintain the normal progress of the reforming reaction; a negative value means that the system is in a state of heat excess, that is, the excess heat needs to be dissipated to prevent the catalyst bed from overheating and being damaged.

[0096] In this embodiment, step 3, which generates the transient thermal mismatch of the reformer, further includes:

[0097] S3.1: Calculate the instantaneous current change rate over three consecutive sampling periods using a first-order backward difference algorithm;

[0098] S3.2: By using a preset constant ratio conversion coefficient, the instantaneous current change rate is directly mapped to the target hydrogen production increment in the next time window, and the target hydrogen production increment is converted into feedforward endothermic demand power based on the molar endothermic enthalpy change constant of the methanol vapor reforming reaction.

[0099] S3.3: Extract the thermal inertia damping coefficient based on the pre-calibrated specific heat capacity of the reformer structural components, subtract the real-time exhaust gas calorific value data from the feedforward heat absorption demand power, and multiply by the thermal inertia damping coefficient to obtain the transient heat mismatch.

[0100] The following provides a detailed explanation of each of the three sub-steps.

[0101] Regarding step S3.1, the instantaneous current change rate is calculated using a first-order backward difference algorithm:

[0102] The first-order backward difference algorithm is a discrete differential approximation method based on historical sampling data, used to extract the instantaneous rate of change of the current signal from a discrete current sampling sequence. Unlike simple one-step difference (which only uses two adjacent sampling points), this embodiment adopts a wide-baseline difference scheme spanning two sampling periods, with its time baseline covering three consecutive sampling times (the current time). Previous sampling time The time corresponding to the first two sampling periods This is used to effectively suppress the interference of high-frequency electrical noise at a single sampling point on the calculation results of the rate of change without introducing a black box filter.

[0103] For example, in this embodiment, the first-order backward differential stabilization rate of change equation can be expressed as follows:

[0104]

[0105] in, The current sampling time The instantaneous current value of the fuel cell stack, This represents the instantaneous current value of the fuel cell stack two sampling periods prior. This is the fixed sampling period for the control system.

[0106] It is understandable that using a differential baseline spanning two sampling periods (i.e., three sampling points) is essentially equivalent to performing a linear smoothing process on the current signal. In engineering field environments, the output signal of a current sensor inevitably contains high-frequency electromagnetic interference noise. If a simple one-step differential (…) is used… The noise component will be significantly amplified, causing severe jitter in the calculated rate of change signal, which in turn causes meaningless high-frequency oscillations in the feedforward control command. The wide-baseline differential scheme suppresses high-frequency noise through pure deterministic algebraic operations by expanding the time span of the difference, while retaining the low-frequency trend information of load change. It does not rely on black-box filtering methods such as Kalman filtering that require parameter tuning, and has the engineering advantages of simple implementation, low computational cost and no additional parameter burden.

[0107] Regarding step S3.2, the rate of change of current is mapped to the power required for feedforward heat absorption:

[0108] The constant ratio conversion factor refers to a pre-calibrated proportionality coefficient with a clear physical meaning, denoted as . Its physical meaning is: the additional molar amount of hydrogen generated per unit current change rate (A / s) (mol / s). This coefficient maps the rate of change signal in the electrical domain to the demand increment signal in the mass flow domain across dimensions, and is the core bridge parameter for realizing feedforward control based on electricity to determine heat.

[0109] The target hydrogen production increment refers to the additional molar amount of hydrogen that the reformer needs to generate beyond the current hydrogen production rate within the next time window to meet the extra hydrogen consumption demand caused by load changes. This increment is obtained by multiplying the current change rate by a constant ratio conversion factor.

[0110] The molar endothermic enthalpy constant of the methanol steam reforming reaction refers to the main reaction of methanol steam reforming ( The heat of reaction absorbed for consuming one mole of methanol under standard conditions is denoted as . Its value is approximately 49.5 kJ / mol. Since the reforming reaction is a strongly endothermic process, for every increase in the demand for hydrogen production, a corresponding amount of additional thermal energy needs to be provided to the reformer to drive these additional endothermic chemical reactions.

[0111] Feedforward endothermic power demand refers to the additional endothermic power that the system will need to provide in the next time window due to a sudden change in load, denoted as The unit is kilowatt (kW).

[0112] For example, in this embodiment, the power equation for feedforward heat absorption demand can be expressed as follows:

[0113]

[0114] in, The conversion factor is a constant ratio. The instantaneous current change rate calculated in step S3.1 This is the standard molar endothermic enthalpy change constant for the methanol vapor reforming reaction (approximately 49.5 kJ / mol). The calculated feedforward heat absorption power requirement.

[0115] Understandably, this equation is essentially a cross-physical energy equivalence conversion chain: rate of change of current → increment of molar flow rate → heat power demand. The rate of change of current signal precedes any temperature change signal in time. When the load current begins to rise, the rate of hydrogen consumption by the fuel cell stack will increase, and the reformer needs to accelerate the reforming reaction to replenish hydrogen. Accelerating the reforming reaction requires additional heat input. Traditional feedback control methods require waiting for the reformer temperature to actually drop and for the temperature sensor to detect the temperature deviation before triggering heat replenishment. This process is limited by the physical conduction time of thermal inertia and usually takes several seconds or even tens of seconds. In this application, by directly converting the rate of change of current into heat power demand, the prediction of future heat demand is completed within millisecond-level calculation time, fundamentally eliminating the control lag problem caused by thermal inertia and realizing true feedforward control.

[0116] Regarding step S3.3, extract the thermal inertia damping coefficient and calculate the transient heat mismatch:

[0117] The thermal inertia damping coefficient is a dynamic correction factor used to compensate for deviations in the actual heat transfer efficiency of the reformer from the ideal state, denoted as . , dimensionless. Under ideal heat transfer conditions, this coefficient is close to 1, meaning that heat can be transferred from the combustion chamber to the catalyst bed with near-ideal efficiency; when heat transfer inside the reformer is hindered (e.g., due to a decrease in thermal conductivity caused by catalyst aging, sintering, or carbon buildup), this coefficient drifts away from 1, thereby applying a corresponding magnitude correction to the feedforward calculation results, so that the control command matches the actual achievable heat transfer capacity.

[0118] For example, in this embodiment, the core equation for transient heat mismatch is as follows:

[0119]

[0120] in, The feedforward heat absorption power demand calculated in step S3.2, The difference between the real-time available thermal power of the exhaust gas calculated in step 200 and the actual thermal power represents the original imbalance between heat supply and demand. The dynamic thermal inertia damping coefficient is used to correct the physical heat transfer efficiency of the original loss measure.

[0121] Positive values ​​represent a heat deficit, meaning that the heat provided by the current exhaust gas is insufficient to meet the heat endothermic demand of the reforming reaction predicted by the feedforward, and the system needs to supplement the heat by means of starting auxiliary methanol combustion; negative values ​​represent a heat surplus, meaning that the heat provided by the current exhaust gas exceeds the actual demand of the reforming reaction, and the system needs to dissipate the excess heat by means of reducing the amount of exhaust gas entering the combustion chamber.

[0122] In this embodiment, the thermal inertia damping coefficient in step S3.3 is not a fixed constant, but is dynamically updated through the following deterministic steps to compensate for the decrease in thermal conductivity caused by the long-term aging of the catalyst:

[0123] Extract the current wall temperature of the reformer Temperature at the center of the catalyst bed located inside the reformer ;

[0124] Within a sliding time window of a set length, the temperature difference between the two is calculated using discrete-time integration.

[0125] The integral result is substituted into a preset polynomial fitting function to output and update the thermal inertia damping coefficient under the current operating state in real time.

[0126] The catalyst bed center temperature refers to the temperature value measured in real time by a temperature sensor (such as a K-type thermocouple or platinum resistance thermometer) embedded in the center of the reformer catalyst packing layer, denoted as The unit is degrees Celsius (°C). This temperature reflects the actual thermal state of the region where the reforming reaction actually occurs, and the temperature gradient between this temperature and the wall temperature is a key characteristic quantity characterizing the heat transfer efficiency inside the reformer.

[0127] A sliding time window is a fixed-length time interval that traces back from the current moment to the past, during which temperature difference data is continuously collected and integrated. This window is updated over time, always maintaining a fixed coverage length.

[0128] Understandably, the length of the sliding time window should be set much longer than the time scale of load transients (usually set to cover hundreds to thousands of sampling periods, such as 100 consecutive sampling points) to ensure that what is extracted is the slow-changing low-frequency information of heat transfer decay caused by catalyst aging, rather than the high-frequency disturbance information of temperature transient fluctuations caused by load abrupt changes.

[0129] For example, in this embodiment, the discrete-time thermal gradient integral equation can be:

[0130]

[0131] in, This refers to the reformer wall temperature. This refers to the temperature at the center of the internal catalyst bed. The length of the preset sliding time window (i.e., the number of sampling points contained in the window). For the fixed sampling period of the control system, each item The time integral infinitesimal element representing the temperature difference at a certain point in history. The final calculated integral value of the thermal gradient is expressed in degrees Celsius per second (°C·s).

[0132] Understandable, This comprehensively reflects the cumulative effect of heat transfer obstruction inside and outside the reformer within the current time window. Under normal heat transfer performance, the temperature difference between the wall and the bed is maintained within a small normal range, resulting in a small integral value. However, when the catalyst undergoes sintering, carbon deposition, or poisoning due to long-term operation, leading to an increase in heat transfer resistance, the temperature difference between the wall and the bed will significantly increase under the same external heating conditions. This increase continues to accumulate over time, resulting in a significant increase in the integral value.

[0133] In obtaining the integral characteristic of the thermal gradient Then, it is mapped to the thermal inertia damping coefficient using a preset polynomial fitting function.

[0134] For example, in this embodiment, the polynomial fitting equation for the thermal damping coefficient can be:

[0135]

[0136] in, , , These are the heat transfer attenuation polynomial coefficients pre-calibrated on a test bench for a reformer of a specific material. These coefficients are obtained by measuring the integral values ​​of the thermal gradient at each stage under experimental conditions with known heat transfer states and performing regression fitting. This is the dynamically corrected thermal inertia damping coefficient.

[0137] Understandably, a second-order polynomial is used for fitting instead of a simple linear proportional relationship because there is a nonlinear relationship between heat transfer decay and the integral of the thermal gradient. In the early stages of catalyst aging, the decline in heat transfer performance is relatively gradual, and the increase in the integral value of the thermal gradient has little impact on the damping coefficient. However, in the severe aging stage of the catalyst, the accelerated deterioration of heat transfer performance leads to a significant increase in the impact of the integral value of the thermal gradient on the damping coefficient. A second-order polynomial can effectively capture this nonlinear characteristic with the fewest parameters (only three coefficients). The design value of this set of formulas lies in achieving adaptive perception of the long-term physical state of the system with extremely low computational cost through purely deterministic integration and polynomial algebraic operations. This allows the feedforward control model to maintain good control accuracy at all stages of the equipment's lifecycle without relying on additional online identification algorithms or frequent manual recalibration.

[0138] In this embodiment, the preset current-hydrogen production-waste heat demand decoupling model in step 3 is specifically implemented as a static multidimensional lookup matrix compiled in read-only memory.

[0139] The static multidimensional lookup matrix refers to the offline pre-calculation performed at all possible input variable combinations based on all algebraic equations of the aforementioned decoupling model during the system's factory calibration phase. The calculation results are stored as a three-dimensional data table (i.e., a lookup table). This matrix is ​​compiled and burned into the controller's read-only memory (ROM). During runtime, no floating-point multiplication or division operations are required; the corresponding pre-calculated results can be obtained simply through memory addressing. The specific steps of the algebraic mapping calculation include:

[0140] The rate of change of instantaneous current After quantization, it is mapped to a row index vector of the matrix;

[0141] The current wall temperature of the reformer After quantization, it is mapped to a column index vector of a matrix;

[0142] Real-time exhaust gas calorific value data After quantization, it is mapped to the depth index vector of the matrix;

[0143] Based on the calculated three-dimensional index vector coordinates, the corresponding pre-calculated mismatch value is directly extracted from the static multidimensional lookup matrix as the transient heat mismatch. .

[0144] Understandably, the advantage of using static lookup tables instead of runtime real-time computation lies in eliminating the uncertain bus latency introduced by floating-point operations. In embedded microcontrollers, the execution time of floating-point multiplication and division operations depends on the numerical characteristics of the operands, potentially resulting in inconsistent computation times between different control cycles, leading to timing jitter in control command outputs. The lookup table approach, however, moves all computationally intensive floating-point operations to the offline calibration stage, performing only memory addressing operations at runtime. Its execution time is deterministic and a constant independent of the input values, thus ensuring the timing determinism of the control loop, making it particularly suitable for hardware control scenarios with stringent real-time requirements. Figure 3 This is a comparison of the time response between the feedforward control method and the traditional temperature feedback control method under sudden load changes. Figure 3 (a) represents the change in external load current. Figure 3 (b) The response timing of the control drive signal, Figure 3 (c) shows the heat compensation response effect. These three sub-figures together demonstrate the time lead of the feedforward signal relative to the temperature feedback signal.

[0145] Step 4: Based on the transient heat mismatch, output the first control command and the second control command synchronously through the hardware control interface. The first control command is used to drive the proportional three-way valve set at the anode exhaust gas emission end to adjust the exhaust gas mixing ratio entering the combustion chamber. The second control command is used to drive the auxiliary methanol afterburning pump to adjust the amount of auxiliary methanol afterburning directly entering the combustion chamber, thereby maintaining the reforming-combustion thermal balance of the reformer.

[0146] The hardware control interface refers to the electrical ports on the controller used to output physical control signals, including but not limited to pulse width modulation (PWM) output ports and digital output ports. The first and second control commands are output synchronously through this interface within the same control cycle, ensuring strict time coordination of the actions of multiple actuators.

[0147] A proportional three-way valve is a three-way fluid control valve installed on the anode exhaust gas discharge line that allows for continuous adjustment of the flow distribution ratio. Upon receiving a first control command, the valve precisely distributes the anode exhaust gas to two downstream lines: one to the combustion chamber (for exhaust gas combustion and heating), and the other to the condensate recovery assembly (for bypass discharge). The valve opening is continuously controlled by the duty cycle or voltage amplitude of the first control command.

[0148] The auxiliary methanol combustion pump is an electromagnetic metering pump independent of the main methanol inlet pump. Its function is to inject liquid methanol from the storage tank directly into the combustion chamber for auxiliary combustion in a precisely metered manner when the heat from the exhaust gas is insufficient to meet the heat absorption requirements of the reformer. Each stroke of the pump discharges a fixed volume of methanol, and the total amount of methanol used for combustion per unit time can be precisely controlled by adjusting the pulse frequency.

[0149] Reforming-combustion thermal balance refers to the target state in which the heat absorbed by the reforming reaction inside the reformer is maintained in dynamic equilibrium with the heat released in the combustion chamber (including exhaust gas combustion and auxiliary methanol combustion). In this equilibrium state, the catalyst bed temperature is maintained within the optimal reforming reaction temperature window, so that the reforming reaction rate does not decrease due to insufficient heat (hydrogen starvation), nor does the catalyst overheat and become damaged due to excessive heat.

[0150] In this embodiment, the logic for adjusting the exhaust gas mixing ratio and the auxiliary methanol combustion amount in step 4 is based on the transient heat mismatch. The numerical range is divided into several mutually exclusive operating condition branches. By introducing a clear heat mismatch threshold, the continuous heat calculation results are discretized into operating condition ranges with clear physical meaning. Within each range, deterministic actuator allocation logic is defined to ensure that multiple actuators are in a deterministic state of complete logical decoupling at any given time, without any control blind spots that conflict with each other or have ambiguous states.

[0151] (a) Sudden load increase condition:

[0152] When the transient heat mismatch is positive and greater than the first preset threshold, it indicates that the system is in a sudden load increase condition.

[0153] Wherein, the first preset threshold is denoted as This refers to a pre-defined positive threshold for heat mismatch, measured in kilowatts (kW). The purpose of this threshold is to establish a dead zone—when the heat mismatch is positive but the value is small (i.e.,...). The system does not immediately trigger the actuator to avoid unnecessary high-frequency mechanical movements caused by minor natural fluctuations in heat during normal operation, thereby extending the mechanical life of the actuator and suppressing system oscillations.

[0154] The load surge condition refers to an operating state in which the external load current increases significantly within a short period of time. Under this condition, the rate at which the fuel cell stack consumes hydrogen accelerates, and the reformer needs to immediately provide more heat to drive a larger-scale reforming reaction, posing an urgent challenge of heat shortage to the system.

[0155] Specifically, under conditions of sudden load increase, step 4 may further include:

[0156] S4.1: Set the duty cycle of the first control command to the maximum value, and drive the proportional three-way valve to guide 100% of the anode exhaust gas to the combustion chamber.

[0157] Understandably, in heat-scarce conditions, the unreacted hydrogen remaining in the anode exhaust gas is a cost-free source of heat energy (its chemical energy comes from the methanol fuel already in the system). Therefore, when heat is scarce, the priority should be to maximize the recovery and utilization of this free waste heat by directing all exhaust gas into the combustion chamber for combustion, rather than bypassing it. This ensures that the free heat source is fully utilized before auxiliary combustion is employed.

[0158] S4.2: The difference between the transient heat mismatch and the real-time exhaust gas calorific value data is divided by the calorific value conversion coefficient of methanol combustion to calculate the target afterburning flow rate.

[0159] The calorific value conversion coefficient of methanol combustion refers to the coefficient obtained by normalizing the effective heat release of methanol combustion corresponding to the single-stroke displacement of the auxiliary methanol combustion pump. Specifically, this coefficient is equal to the lower heating value of methanol. Multiply by the thermal efficiency coefficient of the burner Multiply by the pump's fixed displacement per stroke (mol / stroke), that is Its physical meaning is the amount of heat (kJ / stroke) that the pump can effectively release during each stroke.

[0160] The target afterburning flow rate refers to the pulse frequency required for the auxiliary methanol afterburning pump to operate in order to compensate for the heat gap caused by insufficient exhaust gas heat. It is denoted as... The unit is Hertz (Hz).

[0161] For example, in this embodiment, the target afterburner flow rate equation can be:

[0162]

[0163] in, This is the transient heat mismatch calculated in step 3 (a positive value here). This is the lower heating value of methanol. This is the thermal efficiency coefficient of the burner. The pump has a fixed displacement per stroke.

[0164] Understandably, the core design idea of ​​the above formula is to losslessly equate the continuous thermal power gap to the discrete physical pulse frequency of the electromagnetic pump: the numerator is the total thermal power gap that needs to be filled, and the denominator is the heat that can be effectively released by each pumping action. Dividing the two gives the minimum pumping frequency required to fill the gap. This conversion logic of thermal gap ÷ single thermal equivalent = pulse frequency accurately maps the energy demand in the continuous domain to the actuator action frequency in the discrete domain, realizing a deterministic one-to-one correspondence between thermal demand and actuator command, and avoiding thermal overshoot or oscillation caused by excessive fuel injection based on experience in traditional control methods.

[0165] S4.3: Dynamically match the pulse frequency of the second control command to the target afterburning flow rate, and drive the auxiliary methanol afterburning pump to inject liquid methanol into the combustion chamber.

[0166] In this embodiment, the second control command is output to the electromagnetic drive circuit of the auxiliary methanol afterburning pump in the form of a PWM signal. The frequency of the PWM signal corresponds to the target afterburning flow rate calculated in step S4.2. Each rising edge of the PWM signal triggers a pump stroke, thereby precisely injecting the amount of methanol fuel that matches the heat gap per unit time.

[0167] (ii) Sudden load drop condition:

[0168] When the transient heat mismatch is negative and its absolute value is greater than the second preset threshold, it indicates that the load is in a sudden drop condition.

[0169] Wherein, the second preset threshold is denoted as This refers to a pre-set positive threshold for the absolute value of heat mismatch, measured in kilowatts (kW). Functionally symmetrical to the first preset threshold, this threshold also constitutes a dead zone on the side of excess heat, used to prevent unnecessary actuator operation when there is only a slight excess of heat.

[0170] The load descent condition refers to an operating state in which the external load current decreases significantly within a short period of time. Under this condition, the rate at which the fuel cell stack consumes hydrogen drops abruptly, but the reformer continues to operate at its previous heating level due to thermal inertia, resulting in a large increase in unreacted hydrogen in the exhaust gas. The system faces the risk of excess heat and catalyst overheating.

[0171] Specifically, under load drop conditions, step 4 may further include:

[0172] S4.4: Force the second control command to zero and immediately shut off the output of the auxiliary methanol combustion pump.

[0173] Understandably, in conditions of excess heat, the auxiliary combustion pump is the only external additional heat source introduced, and immediately forcibly shutting it off is the first priority action to eliminate excess heat—cutting off the additional heat source is more direct and effective than adjusting the distribution of the existing heat source. Forced zeroing means that the PWM signal frequency of the second control command is set to zero hertz, the electromagnetic pump completely stops working, and no more methanol is injected into the combustion chamber.

[0174] S4.5: Calculate the excess heat power to be dissipated based on the absolute value of the negative value. Based on the ratio between the real-time exhaust gas calorific value data and the excess heat power, generate a cutoff command as the first control command to drive the proportional three-way valve to bypass and discharge the corresponding proportion of anode exhaust gas to the condensation recovery component, preventing it from entering the combustion chamber.

[0175] For example, in this embodiment, the exhaust bypass ratio equation can be:

[0176]

[0177] in, It is the absolute value of the transient heat mismatch (i.e. the excess heat power that needs to be dissipated). The ratio of the total usable thermal power of the current exhaust gas to the total usable thermal power is the proportion of exhaust gas that should be bypassed. The function ensures that the bypass ratio does not exceed 100%. This is understandable. The introduction of functions reflects the constraint safety at the engineering implementation level, meaning that under any calculation conditions, the valve opening command will not exceed the physical action boundary of the actuator (i.e., the bypass ratio cannot exceed 100%). When the calculated bypass ratio is exactly equal to 100%, it means that all exhaust gas is bypassed to the condensation recovery unit, and no exhaust gas enters the combustion chamber; if there is still excess heat that cannot be absorbed through the exhaust gas bypass, then a higher level of heat dissipation needs to be activated, as follows.

[0178] (III) Extremely Sudden Drop Conditions – Reverse Cooling Water Intervention Logic:

[0179] In this embodiment, under the condition of sudden load drop, if the excess heat power represented by the absolute value of the negative value exceeds the upper limit threshold of the combustion chamber heat dissipation capacity, the method may further include: triggering the reverse cooling water intervention logic based on the execution of S4.4 and S4.5.

[0180] The upper limit threshold of the combustion chamber's heat dissipation capacity is denoted as: This refers to the maximum residual heat that continues to be released from the reformer walls and combustion chamber structural components to the catalyst bed due to thermal inertia, under the condition that all anode exhaust gases are bypassed and discharged to the condensation recovery unit (i.e., the combustion chamber no longer receives any exhaust gas heat source). When the excess heat power exceeds this upper limit, simply cutting off the gas path heat source is insufficient to prevent the catalyst bed temperature from continuing to rise, and active cooling methods are required for forced cooling.

[0181] The reverse cooling water intervention logic refers to an active heat dissipation emergency protection mechanism that is activated under the most severe heat excess conditions. Its core principle is to utilize the physical property that liquid water absorbs a large amount of latent heat of vaporization when it vaporizes into water vapor. By pulse-spraying a small amount of deionized water into the outer cooling jacket of the combustion chamber, the residual thermal inertial potential energy at the bottom of the reformer is forcibly absorbed.

[0182] Specifically, the reverse cooling water intervention logic can include:

[0183] S4.6: Calculate the excess residual heat power, which is the absolute value of the excess heat power minus the upper limit threshold of the combustion chamber's heat dissipation capacity. The remaining heat still needs to be absorbed through liquid cooling.

[0184] S4.7: Convert the remaining heat into cooling water injection volume according to the preset cold water pulse injection volume equation;

[0185] S4.8: Open the bypass water valve to inject deionized water pre-stored in the condensation recovery component into the outer cooling jacket of the combustion chamber in microsecond-level pulses, and use the water vaporization phase change process to forcibly absorb the remaining thermal inertial potential energy at the bottom of the reformer.

[0186] For example, in this embodiment, the preset cold water pulse jet quantity equation can be:

[0187]

[0188] Among them, molecules This refers to the remaining heat (kJ) that still needs to be absorbed through liquid cooling after the gas path's dissipation capacity has been fully utilized. The latent heat of vaporization of water (approximately 2260 kJ / kg) represents the amount of heat absorbed when a unit mass of liquid water completely vaporizes. This refers to the density of deionized water (kg / L), used to convert mass units to volume units for easier spray control. The final calculated volume of a single cold water pulse injection (L).

[0189] Understandably, the latent heat of vaporization of water (approximately 2260 kJ / kg) is used as the conversion benchmark instead of the sensible heat (approximately 4.2 kJ / (kg·℃)) because the heat absorbed by water during the vaporization phase change is far greater than the heat absorbed by the same mass of water during heating. The heat absorbed by 1 kg of water during vaporization is several times greater than the heat absorbed by heating the same mass of water from room temperature to its boiling point. Therefore, only a trace amount (usually on the order of tens of microliters) of deionized water is needed to achieve rapid and forced absorption of a large amount of residual heat, thereby achieving precise quantitative control of the phase change heat dissipation process. This ensures sufficient heat dissipation while avoiding the risk of thermal shock caused by a rapid drop in reformer temperature due to excessive water spraying.

[0190] In this embodiment, microsecond-level pulse jet refers to the bypass water valve opening time being on the order of microseconds to milliseconds. The amount of water jetted each time is precisely controlled to match the calculation result of the cold water pulse jet quantity equation, thereby achieving discrete and precise control of the total jet volume.

[0191] Figure 4 This diagram illustrates a comparison of the anti-jitter performance between the one-step difference method and the wide-baseline difference method of this invention in the calculation of the rate of change of current in this embodiment. Figure 4 (a) shows the current measurement signal containing high-frequency noise. Figure 4 (b) shows the rate of change signals calculated by two difference methods. Figure 5 The diagram illustrates the working condition zoning and actuator collaborative allocation logic of transient heat mismatch in this embodiment. It shows three mutually exclusive intervals divided by the first and second preset thresholds: load surge condition, dead zone, and load drop condition, as well as the normalized output curves of the proportional three-way valve and auxiliary methanol combustion pump in each interval. Figure 6 This diagram illustrates a comparison of the catalyst bed temperature control performance between the feedforward control scheme of this invention and the traditional feedback control scheme under a sudden load increase condition in this embodiment. Figure 6 (a) is a schematic diagram of the change in external load. Figure 6 (b) demonstrates the difference in transient response of catalyst bed temperature between the present invention and the conventional method under the same load surge excitation. Figure 7 and Figure 8 This embodiment illustrates the relationship between the usable thermal power of residual hydrogen in the anode tail gas and the changes in stack current and total amount of reformed gas. Figure 7 A three-dimensional surface diagram showing the usable thermal power of the exhaust gas; Figure 8 This is a contour map showing the available thermal power of the exhaust gas.

[0192] Example 2

[0193] This embodiment discloses a thermoelectric co-control system based on methanol reforming hydrogen fuel cells. Specifically, the control system can be integrated into an embedded electronic control unit, which can be a dedicated controller for fuel cell systems, a programmable logic controller (PLC), a digital signal processor (DSP), a field-programmable gate array (FPGA), or an industrial-grade microcontroller (MCU), etc. Alternatively, it can be integrated into a host computer system, which can be a single industrial control computer or a distributed control cluster composed of multiple industrial control computers.

[0194] In this embodiment, the control system can also be integrated into multiple electronic control units. For example, the data acquisition module and the calorific value calculation module are integrated into the first microcontroller, and the decoupling model analysis module and the dynamic allocation execution module are integrated into the second microcontroller. The control method in Embodiment 1 of this application is implemented collaboratively by multiple microcontrollers through a hardware control bus.

[0195] Specifically, the thermoelectric co-control system for methanol reforming hydrogen fuel cells may include the following functional modules:

[0196] The module includes a data acquisition module, a calorific value calculation module, a decoupled model analysis module, and a dynamic allocation execution module.

[0197] The data acquisition module is configured to acquire a set of real-time operating parameters of the fuel cell system, including the instantaneous current of the stack, the current wall temperature of the reformer, the total methanol inlet flow rate, and the total reforming gas flow rate.

[0198] The data acquisition module is a hardware functional unit consisting of sensors deployed at key nodes of the fuel cell system, along with their corresponding signal conditioning circuits and analog-to-digital converters (ADCs). During each control sampling cycle, this module synchronously reads the analog signals from each sensor, performs signal conditioning (including amplification, filtering, and level conversion) and analog-to-digital conversion, generates a digital set of operating parameters, and transmits these parameters to subsequent functional modules via an internal data bus.

[0199] In this embodiment, the data acquisition module also includes a hardware watchdog circuit connected to the sensor, used to detect whether the sensor signal exceeds the physical reasonable range or a communication timeout occurs, and to output an interrupt signal to trigger system-level protection logic when an abnormality is detected.

[0200] The calorific value calculation module is configured to calculate the theoretical hydrogen consumption rate based on instantaneous current, and generate real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas through difference calculation and constant multiplication.

[0201] The calorific value calculation module refers to the computational function unit that executes all the algebraic calculation logic in step 2 above. This module receives the instantaneous current of the fuel cell stack and the total flow rate of the reformed gas from the data acquisition module. First, it calculates the theoretical hydrogen consumption rate using the Faraday electrolysis mathematical equation. Then, it generates real-time tail gas calorific value data through difference operations and constant multiplication. The data is then transmitted to the decoupled model analysis module.

[0202] In this embodiment, the calorific value calculation module is implemented in a digital signal processor (DSP) or FPGA using fixed-point arithmetic to ensure the determinism and consistency of the calculation delay.

[0203] The decoupling model analysis module is configured to input instantaneous current change rate, reformer wall temperature and real-time exhaust gas calorific value data into a hard-coded algebraic decoupling model to generate the transient heat mismatch of the reformer.

[0204] The decoupling model analysis module refers to the computational function unit that executes all the algebraic mapping calculation logic in step 3 above (including first-order backward difference calculation, feedforward heat absorption demand power conversion, dynamic thermal inertia damping coefficient update, and transient heat mismatch core equation calculation). The term "hard-coded" indicates that the decoupling model parameters and calculation logic in this module are fixed in the controller's firmware at the factory and do not accept external parameter modifications during operation, thus ensuring the determinism and immutability of the control logic.

[0205] In this embodiment, the algebraic decoupling model in the decoupling model analysis module is implemented in the form of a static multidimensional lookup matrix. That is, during the factory calibration stage, all algebraic equations of the decoupling model are pre-calculated offline at all possible combinations of input variables, and the results are compiled into a three-dimensional lookup table and burned into a read-only memory. During runtime, the pre-calculated results are directly extracted through memory addressing to avoid bus latency caused by floating-point operations during runtime.

[0206] The dynamic allocation execution module includes a proportional three-way valve drive circuit and an auxiliary combustion pump control circuit. It is configured to output a first control command to adjust the exhaust gas mixing ratio entering the combustion chamber according to the transient heat mismatch, and output a second control command to adjust the amount of auxiliary methanol combustion directly entering the combustion chamber, thereby completing a thermoelectric synergistic response based on physical actions.

[0207] The dynamic allocation execution module refers to the hardware driver function unit that executes all the actuator allocation logic in step 4 above. This module receives the transient thermal mismatch output by the decoupling model analysis module. Based on its numerical range, the system determines the current operating condition branch (sudden load increase, sudden load decrease, or extreme sudden load decrease) and generates the first and second control commands accordingly.

[0208] A proportional three-way valve drive circuit is a power electronic circuit used to convert a first control command (digital signal) into an analog control voltage or PWM drive signal required to drive the electromagnetic coil of the proportional three-way valve. The output amplitude or duty cycle of this circuit corresponds linearly to the first control command, thereby achieving continuous and precise control of the valve opening.

[0209] The auxiliary combustion pump control circuit is a drive circuit used to convert the second control command (digital pulse frequency signal) into the power pulse signal required to drive the electromagnetic coil of the auxiliary methanol combustion pump. This circuit triggers the electromagnetic pump to perform a complete stroke action upon receiving each rising edge of a pulse; the number of pulses per unit time corresponds to the target combustion frequency calculated in step S4.2.

[0210] In this embodiment, the dynamic allocation execution module also includes a bypass water valve drive circuit, configured to drive the bypass water valve to open in a microsecond-level pulse manner according to the calculation result of the cold water pulse injection quantity equation under extremely sudden operating conditions, so as to inject deionized water into the outer cooling jacket of the combustion chamber.

[0211] Understandably, the data flow between the data acquisition module, calorific value calculation module, decoupled model analysis module, and dynamic allocation execution module constitutes a complete closed-loop control link, from physical quantity acquisition to energy calculation, to supply-demand mismatch judgment, and finally to precise actuator allocation. This link completes a full acquisition-calculation-decision-execution cycle within each control sampling period. Its total execution time is designed to be much smaller than the time constant of the system's thermal inertia, thereby ensuring that control commands always lead the physical changes in the thermal process.

[0212] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A thermoelectric synergistic control method for methanol reforming hydrogen fuel cells, characterized in that, The thermoelectric coordinated control method includes, Obtain the real-time operating parameter set of the fuel cell system. The real-time operating parameter set includes at least: the instantaneous current of the fuel cell stack, the current wall temperature of the reformer, the total methanol inlet flow rate, and the total flow rate of reforming gas entering the fuel cell stack; Based on the instantaneous current of the fuel cell stack and the preset Faraday electrolysis mathematical equations, the theoretical hydrogen consumption rate of the fuel cell stack anode is calculated, and The difference between the hydrogen partial pressure flow rate in the total reforming gas flow rate and the theoretical hydrogen consumption rate is calculated and multiplied by the fixed hydrogen lower heating constant to generate real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas. The current rate of change of the instantaneous current of the fuel cell stack, the current wall temperature of the reformer, and the real-time calorific value of the exhaust gas are input into a preset current-hydrogen production-waste heat demand decoupling model for algebraic mapping calculation to generate the transient heat mismatch of the reformer. Based on the transient heat mismatch, a first control command and a second control command are synchronously output through the hardware control interface. The first control command is used to drive the proportional three-way valve located at the exhaust end of the anode to adjust the exhaust gas mixing ratio entering the combustion chamber. The second control command is used to drive the auxiliary methanol afterburning pump to adjust the amount of auxiliary methanol afterburning that directly enters the combustion chamber; The steps for calculating the algebraic mapping specifically include: The current rate of change of the instantaneous current of the fuel cell stack is quantized and mapped to the row index vector of the static multidimensional lookup matrix; The current wall temperature of the reformer is quantized and mapped to the column index vector of the static multidimensional lookup matrix; The real-time exhaust gas calorific value data is quantized and mapped to the depth index vector of the static multidimensional lookup matrix; Based on the three-dimensional index coordinates formed by the row index vector, the column index vector, and the depth index vector, the corresponding pre-calculated value is extracted from the static multidimensional lookup matrix as the transient heat mismatch amount; The step of generating the transient thermal mismatch of the reformer includes: The current rate of change of the instantaneous current of the fuel cell stack over three consecutive sampling periods is calculated using a first-order backward difference algorithm. By using a preset constant ratio conversion coefficient, the current rate of change of the instantaneous current of the fuel cell stack is mapped to the target hydrogen production increment in the next time window, and the target hydrogen production increment is converted into feedforward endothermic power demand based on the molar endothermic enthalpy change constant of the methanol steam reforming reaction. Extract the thermal inertia damping coefficient based on the specific heat capacity of the reformer structural components, subtract the real-time exhaust gas calorific value data from the feedforward heat absorption demand power, and multiply by the thermal inertia damping coefficient to obtain the transient heat mismatch.

2. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 1, characterized in that, The total flow rate of the reformed gas is directly measured by a dual-channel mass flow controller before it enters the fuel cell stack, and The dual-channel mass flow controller and the proportional three-way valve are connected via the same hardware control bus, and synchronously transmit and receive the real-time operating parameter set and the first control command using a deterministic time-division multiplexing protocol.

3. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 1, characterized in that, The step of calculating the theoretical hydrogen consumption rate of the fuel cell stack anode includes: Obtain the total number of individual cells in the fuel cell stack; Obtain the instantaneous current of the fuel cell stack at the current sampling time; The theoretical hydrogen consumption rate is obtained by dividing the product of the total number of individual cells and the instantaneous current of the stack by the product of the Faraday constant and the number of electrons transferred per mole of hydrogen.

4. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 1, characterized in that, The step of generating real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas includes: Multiply the total reforming gas flow rate by the hydrogen mole fraction to obtain the total hydrogen supply mole flow rate currently entering the anode of the fuel cell stack. The molar flow rate of unreacted hydrogen in the anode tail gas is obtained by subtracting the theoretical hydrogen consumption rate from the total hydrogen supply molar flow rate. The real-time exhaust gas calorific value data is obtained by multiplying the molar flow rate of the unreacted hydrogen by the physical heat recovery efficiency constant of the exhaust gas pipeline, and then by the fixed hydrogen lower heating value constant.

5. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 4, characterized in that, Prior to the step of generating real-time tail gas calorific value data characterizing the energy of unreacted hydrogen in the anode tail gas, a boundary anomaly handling step is also included: Compare the average voltage of the single cell of the fuel cell stack obtained at the current moment with the theoretical voltage of the fuel cell stack under the instantaneous current in the pre-stored standard polarization curve; If the average voltage of a single cell in the fuel cell stack is lower than the theoretical voltage by more than a preset difference, and the real-time exhaust gas calorific value data indicates that there is a sufficient amount of unreacted hydrogen in the anode exhaust gas, then it is determined that the fuel cell stack is experiencing a water flooding condition.

6. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 5, characterized in that, After determining that the flooding condition has occurred inside the fuel cell stack, the method further includes: Forcefully suspend the subsequent algebraic mapping calculation steps and the steps of synchronously outputting the first control instruction and the second control instruction; The command is set at the exhaust solenoid valve at the anode tail gas emission end to alternately perform fully open and fully closed actions at a frequency of more than 10 Hz, triggering the pulse purging sequence; The average voltage of the single cell in the fuel cell stack is continuously monitored until it recovers to within the preset difference range corresponding to the theoretical voltage in the standard polarization curve. Then, the suspended state is released, and the execution of the algebraic mapping calculation step and the step of synchronously outputting the first control command and the second control command is resumed.

7. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 1, characterized in that, The current-hydrogen production-waste heat demand decoupling model is implemented in the form of a static multidimensional lookup matrix compiled in read-only memory.

8. The thermoelectric synergistic control method for methanol reforming hydrogen fuel cells according to claim 1, characterized in that, In the step of synchronously outputting the first control command and the second control command based on the transient heat mismatch, the following mutually exclusive operating condition branches are executed according to the numerical range of the transient heat mismatch: When the transient heat mismatch is positive and greater than the first preset threshold, it is determined that the load surge condition is underway. When the transient heat mismatch is negative and its absolute value is greater than the second preset threshold, it is determined that the load is in a sudden drop condition.

9. A thermoelectric synergistic control system for hydrogen production fuel cells based on methanol reforming, characterized in that, The thermoelectric synergistic control system includes: processor; The memory stores a computer program that, when executed by a processor, implements the thermoelectric co-control method for hydrogen fuel cells based on methanol reforming as described in any one of claims 1 to 8.

Images