A fuel cell system and method and apparatus for anode nitrogen and anode water management thereof

Abstract

The application discloses a fuel cell system and an anode nitrogen and anode water management method and device, and the method comprises the following steps: obtaining parameters of the fuel cell; determining a current anode nitrogen partial pressure change rate according to a first group of parameters in the parameters and an anode nitrogen partial pressure dynamic model, determining a current anode water content change rate according to a second group of parameters in the parameters and an anode water content dynamic model, predicting state quantities of the fuel cell in future multiple periods according to the current anode nitrogen partial pressure change rate, the current anode water content change rate and an anode prediction model, solving a target optimization problem based on a target optimization function and the anode prediction model to obtain a control sequence, and controlling a purge valve according to a first purge valve opening control quantity in the control sequence and controlling a circulating pump according to a first circulating pump control quantity. The nitrogen accumulation rate and the anode water content can be perceived and actively intervened in advance.

Description

Technical Field

[0001] This invention relates to the field of vehicles, and more particularly to a fuel cell system and a method and apparatus for managing anode nitrogen and anode water therewith. Background Technology

[0002] As a power source for vehicles, PEMFC (Proton Exchange Membrane Fuel Cell) experiences nitrogen gas on the cathode side permeating through the proton exchange membrane to the anode side during operation. This nitrogen gradually accumulates in the anode channel and hydrogen circulation loop, which not only reduces the hydrogen concentration and affects the power generation performance of the fuel cell, but also encroaches on the reaction space, further exacerbating the difficulty of draining generated water. This can easily lead to anode flooding, increase mass transfer losses, and affect the lifespan of the fuel cell stack.

[0003] Therefore, anode management is necessary. In related technologies, anode management mainly employs rule-based fixed-frequency purging or reactive control based on voltage / pressure thresholds. While these methods are simple and easy to implement, they have significant shortcomings. First, there is control lag; threshold or fixed-frequency control is a "post-hoc remedy," unable to intervene before nitrogen accumulation or flooding occurs, by which time the fuel cell stack has already suffered performance loss or irreversible damage. Second, there is poor adaptability to operating conditions; fixed parameters cannot adapt to drastic changes in load, temperature, and other operating conditions, easily leading to insufficient purging during dynamic operation, resulting in performance degradation, or excessive purging, leading to hydrogen waste and system fluctuations. Third, decoupling of multiple variables is difficult; anode purging valves, circulating pumps, humidifiers, and other actuators are coupled with each other, making it difficult for traditional control methods to coordinate their actions, easily leading to control conflicts. For PEMFC, a multivariable system with strong nonlinearity and large lag, the control accuracy and robustness of conventional closed-loop control algorithms are limited. Summary of the Invention

[0004] This invention proposes a fuel cell system and a method and apparatus for managing anode nitrogen and anode water, aiming to solve the problems described in related technologies to achieve precise and coordinated management of anode nitrogen concentration and water content.

[0005] Based on this, one embodiment of the present invention proposes a method for managing anode nitrogen and anode water in a fuel cell system, comprising: The parameters of the fuel cell are obtained based on the acquisition cycle, and the parameters include at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content; The current rate of change of anode nitrogen partial pressure is determined based on the first set of parameters and the dynamic model of anode nitrogen partial pressure; the current rate of change of anode water content is determined based on the second set of parameters and the dynamic model of anode water content. The state variables of the fuel cell in the future multiple cycles are predicted based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model, wherein the state variables include anode nitrogen partial pressure and anode water content; The control sequence is obtained by solving the objective optimization problem based on the objective optimization function and the anode prediction model. The control sequence includes control quantities for multiple future cycles. The control quantities include the purge valve opening control quantity corresponding to the anode nitrogen partial pressure and the circulation pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulation pump is controlled by the first circulation pump control quantity.

[0006] Optionally, the first set of parameters includes the stack temperature, anode flow channel volume, membrane area, purge valve opening command, anode outlet pressure, and anode inlet pressure; The dynamic model of nitrogen partial pressure at the anode is as follows: , The partial pressure of nitrogen at the anode. The nitrogen permeation flow rate is related to the nitrogen permeation coefficient, the membrane area, the partial pressure of nitrogen at the cathode, and the partial pressure of nitrogen at the anode. The purge nitrogen flow rate is related to the anode inlet pressure, the anode outlet pressure, the anode nitrogen partial pressure, and the purge valve opening command. For the anode flow channel volume, Let be the ideal gas constant. This refers to the temperature of the fuel cell stack.

[0007] Optionally, the second set of parameters includes the stack current, relative humidity in the anode channel, film thickness, and film density; The dynamic model for anode water content is as follows: ,in, For film thickness, For membrane density, The electroosmotic drag flow rate is related to the current in the fuel cell stack. The reverse diffusion flow rate is related to the water diffusion coefficient and the water concentration difference across the membrane. The water concentration difference across the membrane is related to the relative humidity within the anode channel. The water content generated by the anodic reaction. This represents the anode water content.

[0008] Optionally, the anode prediction model is an MPC model, and the state variables of the MPC model are: The control quantity is ,in, The speed of the circulating pump. The purge valve opening is; the predicted state variables for multiple future cycles are: The multiple control sequences are ; To predict the time domain, To control the time domain.

[0009] Optionally, the objective optimization function is: ; This is the fuel cell stack voltage. Let Q be the reference voltage and Q be the first weighting matrix. To control the increment, R is the second weight matrix. This is the weighting coefficient for hydrogen consumption. For the total duration of the sweeping, , All are positive integers.

[0010] Optionally, before obtaining the control sequence by solving the objective optimization problem based on the objective optimization function and the anode prediction model, the method further includes: obtaining the error and error change rate between the stack voltage and the reference voltage of the fuel cell, and adjusting the weights in the objective optimization function based on preset weight adjustment conditions.

[0011] Optionally, the preset weight adjustment conditions include: the fuzzy sets of e and Δe are {NB negative large, NM negative medium, NS negative small, ZO zero, PS positive small, PM positive medium, PB positive large}; α Q and α R The fuzzy set is {S_small, M_medium, B_large}, and the method for adjusting the weights in the objective optimization function satisfies the contents of Table 1. Table 1 Where e is the error, Δe is the rate of change of the error, and α Q α is the adjustment factor related to the first weight matrix Q. R This is the adjustment factor related to the second weight matrix R.

[0012] Based on this, a second aspect of the present invention provides an anode nitrogen and anode water management device for a fuel cell system, comprising: The acquisition module is used to acquire parameters of the fuel cell based on the acquisition cycle. The parameters include at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content. The state quantity observation module is used to determine the current rate of change of anode nitrogen partial pressure based on the first set of parameters and the dynamic model of anode nitrogen partial pressure, and to determine the current rate of change of anode water content based on the second set of parameters and the dynamic model of anode water content. The predictive control module is used to predict the state variables of the fuel cell in the future multiple cycles based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model, wherein the state variables include the anode nitrogen partial pressure and the anode water content. The target optimization module is used to solve the target optimization problem based on the target optimization function and the anode prediction model to obtain a control sequence. The control sequence includes control quantities for multiple future periods. The control quantities include a purge valve opening control quantity corresponding to the anode nitrogen partial pressure and a circulating pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulating pump is controlled by the first circulating pump control quantity.

[0013] Based on this, a third aspect of the present invention provides a fuel cell system, comprising: a parameter acquisition sensor group, a controller, a purge valve, and a circulation pump; the parameter acquisition sensor group is used to acquire parameters of the fuel cell, and the controller is connected to the parameter acquisition sensors, the purge valve, and the circulation pump respectively. The controller is used to acquire the parameters and execute the anode nitrogen and anode water management method of a fuel cell system according to any embodiment of the present invention to control the purge valve and the circulation pump.

[0014] Optionally, the controller includes a control chip, and an analog signal processing circuit, a digital signal output circuit, a speed control output circuit, and a CAN bus communication circuit connected to the control chip; The CAN bus communication circuit is used for communication between the vehicle and the controller, the analog signal processing circuit is used for processing the parameters, the digital output circuit is used for outputting digital signals, and the speed control output circuit is used for outputting speed control signals.

[0015] In summary, the fuel cell system and its anode nitrogen and anode water management method and apparatus proposed in this embodiment of the invention include: acquiring fuel cell parameters based on an acquisition cycle, the parameters including at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content; determining the current anode nitrogen partial pressure change rate based on the first set of parameters and a dynamic model of anode nitrogen partial pressure, and determining the current anode water content change rate based on the second set of parameters and a dynamic model of anode water content; predicting the state variables of the fuel cell in the future multiple cycles based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and an anode prediction model, wherein the state variables include the anode nitrogen partial pressure and the anode water content; solving the objective optimization problem based on the objective optimization function and the anode prediction model to obtain a control sequence, the control sequence including control variables in the future multiple cycles, the control variables including a purge valve opening control variable corresponding to the anode nitrogen partial pressure and a circulation pump control variable corresponding to the anode water content, and controlling the purge valve with the first purge valve opening control variable and the circulation pump with the first circulation pump control variable in the control sequence. This invention, by setting up dynamic models for anode nitrogen partial pressure, anode water content, and anode prediction models, enables the early prediction of anode nitrogen partial pressure, anode water content, purge valve opening control, and circulating pump control. This allows for proactive sensing and intervention of nitrogen accumulation rate and anode water content, preventing performance degradation. Furthermore, the entire system can adapt to changes in load and environment, achieving precise on-demand purging and water balance adjustment, avoiding control lag and poor adaptability to dynamic conditions. Simultaneously, the output purge valve opening control and circulating pump control can synergistically optimize the actions of multiple actuators such as the purge valve and circulating pump, improving control accuracy and robustness. Attached Figure Description

[0016] Figure 1 This is a flowchart of the anode nitrogen and anode water management method for a fuel cell system proposed in this embodiment of the invention; Figure 2 This is a control principle diagram of the anode nitrogen and anode water management device of the fuel cell system proposed in this embodiment of the invention; Figure 3 This is a schematic diagram of the weight adjustment mechanism of the anode nitrogen and anode water management device in the fuel cell system proposed in this embodiment of the invention. Figure 4 This is a block diagram of the anode nitrogen and anode water management device of the fuel cell system proposed in the embodiments of the present invention; Figure 5 This is a block diagram of the fuel cell system proposed in an embodiment of the present invention; Figure 6 This is a structural diagram of the controller of the fuel cell system proposed in the embodiments of the present invention; Figure 7This is a schematic diagram of the drive circuit for the anode purge valve proposed in an embodiment of the present invention; Figure 8 This is a schematic diagram of the anode circulation pump control circuit proposed in an embodiment of the present invention. Detailed Implementation

[0017] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the invention and not all structures. Various modifications and variations can be made to the present invention without departing from its spirit or scope, which will be apparent to those skilled in the art. Therefore, the present invention is intended to cover modifications and variations falling within the scope of the corresponding claims and their equivalents. It should be noted that the embodiments provided in the present invention can be combined with each other without contradiction.

[0018] Figure 1 This is a flowchart of the anode nitrogen and anode water management method for a fuel cell system proposed in an embodiment of the present invention. Figure 1 As shown, the method includes: S101, acquire parameters of the fuel cell based on the acquisition cycle, the parameters including at least a first set of parameters related to the partial pressure of nitrogen at the anode and a second set of parameters related to the water content at the anode.

[0019] The purpose of acquiring fuel cell parameters is to manage anode nitrogen and anode water, which involves controlling the purge valve and circulation pump accordingly. The acquisition period is related to the overall computation period of the anode prediction model and can be an integer multiple of the computation period; in one embodiment, the acquisition period can be 100 ms.

[0020] Furthermore, fuel cell parameters reflect the stack's operating status. Since the purpose of this invention is to manage anode nitrogen and anode water, it is necessary to obtain at least the parameters related to anode nitrogen and anode water. Anode nitrogen concentration is generally related to stack temperature, anode pressure, and flow channel volume; these parameters constitute the first set of parameters. Anode water content can generally be characterized using the relative humidity within the flow channel, and the flow rate is related to the stack current; these parameters constitute the second set of parameters.

[0021] Therefore, the parameters obtained in this embodiment include at least the fuel cell current, anode pressure (anode inlet pressure and anode outlet pressure), fuel cell temperature, and anode circulation loop humidity. In some embodiments, the parameters also include fuel cell pressure. These parameters can generally be acquired by corresponding sensors. For example, the fuel cell current can be obtained using a current sensor, the pressure can be obtained using a pressure sensor, the temperature can be obtained using a temperature sensor, the humidity can be obtained using a humidity sensor, and the voltage can be calculated from the current or directly measured using a voltmeter.

[0022] S102, determine the current anode nitrogen partial pressure change rate based on the first set of parameters and the anode nitrogen partial pressure dynamic model, and determine the current anode water content change rate based on the second set of parameters and the anode water content dynamic model.

[0023] Since the partial pressure of nitrogen cannot be directly measured, it needs to be calculated. Before calculating the nitrogen partial pressure, a dynamic model of the anode nitrogen partial pressure needs to be constructed to calculate the current rate of change of the anode nitrogen partial pressure. In one embodiment, if the anode channel is an ideal mixing container, and the dynamics of the nitrogen partial pressure are determined by the nitrogen penetration process and the purging process, then the dynamic model of the anode nitrogen partial pressure can be established by combining the ideal gas equation, the nitrogen penetration process, and the purging process. The nitrogen penetration process and the purging process are both related to the first set of parameters. Therefore, the current rate of change of the anode nitrogen partial pressure can be calculated by inputting the first set of parameters into the dynamic model of the anode nitrogen partial pressure.

[0024] Similarly, since water content cannot be directly measured and needs to be calculated, a dynamic model of anode water content needs to be constructed before calculating the current rate of change of anode water content. In one embodiment, the dynamic model of anode water content can be determined by the water migration process and thus can be established based on the water migration process. The water migration process is related to a second set of parameters; therefore, the current rate of change of anode water content can be calculated by inputting the second set of parameters into the dynamic model of anode water content.

[0025] S103, based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model, predict the state variables of the fuel cell in the future multiple cycles, where the state variables include the anode nitrogen partial pressure and the anode water content.

[0026] Among them, the anode prediction model can be a nonlinear mechanism model, with the anode nitrogen partial pressure and anode water content as state variables. The purge valve opening control quantity and the circulating pump control quantity are used as input control quantities. , ; To obtain the cycle, The rate of change. Both the initial state variable and the input control variable can be set empirically. Therefore, when the state variable is clearly defined... After obtaining the parameters of the fuel cell, the state variables for multiple future cycles can be obtained. By solving the objective optimization problem and the objective optimization function, the control sequence, i.e., the control variables for multiple future cycles, can be obtained.

[0027] S104. Based on the objective optimization function and the anode prediction model, the objective optimization problem is solved to obtain the control sequence. The control sequence includes control quantities for multiple future cycles. The control quantities include the purge valve opening control quantity corresponding to the anode nitrogen partial pressure and the circulating pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulating pump is controlled by the first circulating pump control quantity.

[0028] The dual optimization objectives are "maintaining high-performance operation of the fuel cell stack" and "maximizing hydrogen utilization efficiency." This means maintaining the anode nitrogen concentration and the anode water content within certain constraints.

[0029] It should be noted that the accumulation of nitrogen at the anode will reduce the battery's output performance, while excessive water at the anode will cause flooding of the fuel cell, affecting the mass transfer process and reducing hydrogen utilization. Therefore, the nitrogen content at the anode can be reduced by purging, and the water flow can be regulated by a circulating pump to accelerate the removal of excess water from the anode, thus preventing flooding.

[0030] Specifically, the nitrogen control quantity corresponds to the target opening degree of the purge valve. Currently, simply adjusting the purge valve to the target opening degree is sufficient to manage anode nitrogen. The larger the control quantity, the larger the purge valve opening, resulting in more nitrogen being discharged per unit time and faster reduction of anode nitrogen accumulation. The water control quantity corresponds to the target speed of the circulation pump. Currently, adjusting the circulation pump to the target speed, the larger the control quantity, the higher the circulation pump speed, and the faster the circulation flow rate of the anode water. Therefore, by predicting the anode nitrogen concentration and anode water content in advance, and using the anode prediction model and target optimization function, the optimal control sequence is obtained. The circulation pump is controlled by the first circulation pump control quantity in the optimal control sequence, and the purge valve is controlled by the first purge valve control quantity. This achieves precise control of the purge valve and circulation pump, as well as early perception and proactive intervention of the nitrogen accumulation rate and anode water content.

[0031] Therefore, the anode nitrogen and anode water management method for fuel cells provided in this embodiment of the invention can estimate the current anode nitrogen partial pressure and the current anode water volume based on the current fuel cell parameters, and predict the current anode nitrogen control quantity and anode water control quantity by combining the state prediction model, so that the current purge valve opening degree and the circulation pump speed can match the current anode nitrogen partial pressure and the current anode water volume.

[0032] Optionally, the first set of parameters includes the stack temperature, anode flow channel volume, membrane area, purge valve opening command, anode outlet pressure, and anode inlet pressure; The dynamic model of nitrogen partial pressure at the anode is as follows: , The partial pressure of nitrogen at the anode. This represents the nitrogen penetration flow rate, which is related to the nitrogen penetration coefficient, membrane area, cathode nitrogen partial pressure, and anode nitrogen partial pressure. The purge nitrogen flow rate is related to the anode inlet pressure, anode outlet pressure, anode nitrogen partial pressure, and the purge valve opening command. For the anode flow channel volume, Let be the ideal gas constant. This refers to the temperature of the fuel cell stack.

[0033] Among them, nitrogen penetration flow rate Related to the partial pressure of nitrogen on the cathode side and membrane characteristics, it can be approximated as: , Nitrogen penetration coefficient For membrane area, The partial pressure of nitrogen on the cathode side can be estimated from air pressure and composition.

[0034] Nitrogen flow rate carried away by purging Related to purge valve status and total flow channel pressure: ; in, The total pressure at the anode can be the average of the anode inlet pressure and the anode outlet pressure. This is the purge valve opening command. The flow-opening characteristic function of the purge valve can be obtained through calibration.

[0035] The membrane area and anode flow channel volume, which are related to the structure of the fuel cell stack, can be obtained in advance.

[0036] Therefore, after obtaining the parameters of the fuel cell, the rate of change of anode nitrogen can be calculated by combining the dynamic model of anode nitrogen partial pressure.

[0037] Optionally, the second set of parameters includes the stack current, relative humidity in the anode channel, film thickness, and film density; The dynamic model of anode water content is as follows ,in, For film thickness, For membrane density, This refers to the electroosmotic drag current, which is related to the fuel cell current. The reverse diffusion flow rate is related to the water diffusion coefficient and the water concentration difference across the membrane. The water concentration difference across the membrane is related to the relative humidity in the anode channel. The water content generated by the anodic reaction. This represents the anode water content.

[0038] in, , This is the electroosmotic drag coefficient. It is Faraday's constant. This represents the fuel cell current.

[0039] , Let be the water diffusion coefficient. This represents the water concentration difference across the membrane, which is related to the relative humidity on both sides.

[0040] Anode reaction produces water This is usually negligible because the main reaction, water, is generated at the cathode.

[0041] Relative humidity in the anode channel With anode water content There is a mapping relationship, which affects the condensation and evaporation of liquid water. This is influenced by the rotational speed of the circulating pump. The gas flow rate can be adjusted, thus affecting drainage capacity. High flow rates help remove liquid water and prevent flooding, but they also increase the risk of membrane drying. Therefore, it is possible to establish a relationship between anode water content and circulating pump speed; by predicting the anode water content, the corresponding circulating pump control parameters can be obtained.

[0042] Therefore, after obtaining the parameters of the fuel cell, the rate of change of the anode water content can be calculated by combining the dynamic model of anode water content.

[0043] Optionally, the anode prediction model is an MPC model, and the state variables of the MPC model are: The control quantity is ,in, The speed of the circulating pump. The purge valve opening is; the predicted state variables for multiple future cycles are: Multiple control sequences are ; To predict the time domain, To control the time domain.

[0044] In other words, by setting up the MPC model, based on the current measurement state... and future control sequence Using the above prediction model, predict the future. Step state trajectory .in, To control the input vector.

[0045] For example, the anode nitrogen partial pressure for multiple future cycles can be iterated using the following formula: .

[0046] The anode water content for future cycles can be iteratively calculated using the following formula: .

[0047] Therefore, by establishing dynamic models of anode nitrogen partial pressure, dynamic models of anode water content, and MPC models respectively, the trends of nitrogen accumulation and water content changes on the anode side during fuel cell operation can be accurately predicted, enabling proactive intervention in anode nitrogen and anode water content.

[0048] Optionally, the objective function is: ; This is the fuel cell stack voltage. Let Q be the reference voltage and Q be the first weighting matrix. To control the increment, R is the second weight matrix. This is the weighting coefficient for hydrogen consumption. For the total duration of the sweeping, , All are positive integers.

[0049] It should be noted that the first item For tracking purposes, penalize the predicted stack voltage. With reference voltage The deviation, voltage can be obtained from the state through the model. Estimate.

[0050] Second item To control for smoothness, penalize drastic changes in the control variable. R is the weight matrix, used to ensure the stability of the control.

[0051] Third item For economic purposes, the penalty is the total duration of the sweeping operation. (and (Related to points) This is a hydrogen consumption weighting coefficient used to balance performance and hydrogen efficiency.

[0052] The constraints include: State constraints: To ensure that the nitrogen concentration is within a safe range; Output constraints: Ensure that the voltage is within a safe range; Input constraints: To ensure that the implementing agency operates within an effective scope; Input rate constraints: Limit the rate of change of the implementing agency; By solving this constrained optimization problem, the control sequence that minimizes the objective function value is obtained. To determine the optimal control sequence, the first control variable is... Applied to the system.

[0053] Optionally, before obtaining the control sequence by solving the objective optimization problem based on the objective optimization function and the anode prediction model, the method further includes: obtaining the error and error change rate between the fuel cell stack voltage and the reference voltage, and adjusting the weights in the objective optimization function based on preset weight adjustment conditions.

[0054] Here, the error is the deviation between the current fuel cell voltage and the reference voltage, and the error change rate is the difference between the error of the current cycle and the error of the previous cycle. A larger error and a smaller error change rate indicate that the current fuel cell voltage is somewhat far from the reference voltage, but stable. In this case, voltage tracking should be strengthened and the restrictions on control actions should be relaxed. A smaller error and a larger error change rate indicate that the current fuel cell voltage is around the reference voltage, but oscillates significantly. In this case, control actions should be restricted to prevent oscillations. Strengthening or weakening voltage tracking and relaxing or tightening control actions can be adjusted through corresponding weights. To enhance the robustness of the model predictive control algorithm to model uncertainties and dynamic disturbances, this embodiment introduces a fuzzy adaptive tuner to adjust the weight matrices Q and R in the predictive control optimization problem online. Specifically, when voltage tracking needs to be strengthened, the weight Q is increased, and vice versa. When control actions need to be tightened, the weight R is increased, and vice versa.

[0055] Through the weight optimization adjustment in this embodiment, the control target can be flexibly adapted according to the actual operating state of the fuel cell. It can prioritize ensuring the stability of the stack output performance when the voltage deviation is large, and prioritize reducing hydrogen consumption when the operating conditions are stable, thereby improving the overall operating economy.

[0056] Optionally, the preset weight adjustment conditions include: the fuzzy sets of e and Δe are {NB negative large, NM negative medium, NS negative small, ZO zero, PS positive small, PM positive medium, PB positive large}; α Q and α R The fuzzy set is {S_small, M_medium, B_large}, and the method for adjusting the weights in the objective optimization function satisfies the contents of Table 1; Table 1 Where e is the error, Δe is the rate of change of the error, and α Q α is the adjustment factor related to the first weight matrix Q. R This is the adjustment factor related to the second weight matrix R.

[0057] It is understandable that the meanings of NB (negative and large), NM (negative and medium), NS (negative and small), ZO (zero), PS (positive and small), PM (positive and medium), and PB (positive and large) in the fuzzy sets of e and Δe can be interpreted as follows: taking the error e as an example, negative means the deviation between the current stack voltage and the reference voltage is negative, i.e., the current stack voltage is less than the reference voltage; positive means the deviation between the current stack voltage and the reference voltage is positive, i.e., the current stack voltage is greater than the reference voltage; zero means the deviation between the current stack voltage and the reference voltage is zero; large, medium, and small represent whether the error is large, medium, or small, respectively. α Q and α R The fuzzy set S is small, M is medium, and B is large can be understood as follows: when the adjustment factor is small, the influence weight becomes smaller; when the adjustment factor is large, the influence weight becomes larger.

[0058] Specific adjustment methods can be found in Table 1. For example, when the error e is positive (PB) and Δe is zero (ZO), it indicates that the voltage is far below the target and there is no improvement. Voltage tracking should be strengthened, and control action restrictions should be relaxed. α Q Choose B, α R Choose option M. When the error e is zero at ZO and Δe is zero at ZO, the system is stable near the target, maintaining the default weights. α Q In M, α R Select M. When the error e is small (negative) and Δe is large (positive) compared to PB, and the voltage slightly overshoots and shows an increasing trend, the control action should be suppressed to prevent oscillation. α Q Choose a smaller S, α R Choose option B. The remaining content in Table 1 can be found in this example. After setting the fuzzy set, the weights Q and R can be adjusted based on the input error e and the rate of change of error Δe.

[0059] The adjusted weights are Q'=α Q Q, R'=α R R, this fuzzy tuning module enables the controller parameters to adapt to the dynamic performance of the system.

[0060] In one specific embodiment, the method includes: Step 1, Data Acquisition: At time k, the sensor array acquires real-time operating data of the fuel cell stack.

[0061] Step 2, State estimation: Based on the measured values, the state that is difficult to measure directly is estimated to obtain the current state x(k).

[0062] Step 3, Fuzzy Tuning: Based on the voltage error e and the error change rate Δe, the current weight adjustment factor α is calculated through fuzzy inference and defuzzification. Q (k) and α R (k).

[0063] Step 4, Rolling Optimization: With x(k) as the initial state, using the adjusted weights Q'(k) and R'(k), solve the optimization problem through a predictive control algorithm to obtain the future N. c Optimal control sequence of steps .

[0064] Step 5, Control Implementation: The first control quantity in The command is sent to actuators such as the purge valve and the circulation pump.

[0065] Periodic update: Wait for the next sampling time k+1, repeat steps one to five to achieve closed-loop predictive control.

[0066] in, Figure 2 This is a control principle diagram of the anode nitrogen and anode water management device of the fuel cell system proposed in this embodiment of the invention. Figure 3 This is a schematic diagram of the weight adjustment mechanism of the anode nitrogen and anode water management device in the fuel cell system proposed in this embodiment of the invention. Figure 2 and Figure 3As shown, the process begins with multi-source data acquisition. The controller collects real-time measurements from sensors such as voltage, current, pressure, humidity, and temperature to construct the current operating status of the system. Next, in the state prediction phase, the controller utilizes its integrated prediction model to simulate and calculate the changing trends of nitrogen concentration and proton exchange membrane water content in the anode channel over a short period, based on the collected real-time data. This allows the controller to anticipate the system's state evolution. Following this prediction, the process enters the core optimization decision-making phase. The controller, with the dual optimization objectives of "maintaining high-performance stack operation" and "maximizing hydrogen utilization efficiency," comprehensively considers all safety constraints and quickly solves the optimal control problem, calculating the most suitable control command sequence for the anode purge valve and circulating pump. Finally, after the decision is made, the controller immediately issues commands to the actuators, driving the purge valve and circulating pump to perform precise actions, completing the physical execution of the control closed loop. To further enhance the system's adaptability and robustness in the face of model uncertainties, external disturbances, and complex dynamic conditions, the controller deeply integrates a fuzzy adaptive tuning module. This module, acting as the system's intelligent adjustment unit, continuously monitors the tracking error and its rate of change between the fuel cell output voltage and the reference voltage, and performs real-time inference and decision-making based on a pre-defined fuzzy rule base. Its output is used to automatically adjust the scaling factors of the weight matrix Q and the control quantity change weight R in the predictive control algorithm online, thereby achieving adaptive optimization of the controller's key parameters. This design allows the control algorithm to dynamically adjust according to the actual operating performance of the system, maintaining smoothness in steady state and rapid response during dynamic transient processes. This collaborative mechanism, integrating "state estimation, multi-step prediction, rolling optimization, and parameter self-calibration," constitutes a complete management closed loop capable of intelligently, accurately, and efficiently completing the process from perception and decision-making to execution, significantly improving the overall quality of control.

[0067] In summary, through its built-in anode state prediction model, this invention can predict nitrogen accumulation and water content changes several seconds to tens of seconds in advance. This allows for proactive purging or adjustment of the circulation pump before performance degradation occurs, achieving a leap from "passive response" to "proactive prevention," enabling forward-looking control and performance optimization. By integrating multi-physics information such as temperature and current density, model predictive control can coordinate the purging valve and circulation pump to achieve precise regulation of anode humidity. For example, when flooding is predicted under low current density conditions, the circulation pump speed is increased in advance; when drying is predicted under high current density conditions, the purging frequency is reduced and the pump speed is decreased. This allows the fuel cell stack to maintain optimal hydrothermal balance under various operating conditions, significantly extending the membrane electrode lifetime. This invention employs a multi-input multi-output optimization framework, fundamentally solving the coupling problem between multiple variables such as purging and circulation. Meanwhile, by explicitly incorporating the physical limits of the actuator and the safe voltage / pressure range as constraints into the optimization problem, this invention ensures that the control system operates within safe boundaries at all times, avoiding risks such as actuator saturation or system overpressure, thus improving system reliability and safety, and achieving multivariable decoupling and constraint protection. By introducing weight parameters into the fuzzy logic online tuning predictive control algorithm, this invention enables the control system to adapt to model mismatch and external disturbances, such as sudden changes in ambient temperature and drastic load fluctuations, exhibiting strong robustness and adaptability.

[0068] Figure 4 This is a block diagram of the anode nitrogen and anode water management device of the fuel cell system proposed in this embodiment of the invention, as shown below. Figure 4 As shown, it includes: The acquisition module 101 is used to acquire parameters of the fuel cell based on the acquisition cycle. The parameters include at least a first set of parameters related to the partial pressure of nitrogen at the anode and a second set of parameters related to the water content at the anode. The state quantity observation module 102 is used to determine the current anode nitrogen partial pressure change rate based on the first set of parameters and the anode nitrogen partial pressure dynamic model, and to determine the current anode water content change rate based on the second set of parameters and the anode water content dynamic model. The predictive control module 103 is used to predict the state variables of the fuel cell in the future multiple cycles based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model. The state variables include the anode nitrogen partial pressure and the anode water content. The target optimization module 104 is used to solve the target optimization problem based on the target optimization function and the anode prediction model to obtain the control sequence. The control sequence includes control quantities for multiple future cycles. The control quantities include the purge valve opening control quantity corresponding to the anode nitrogen partial pressure and the circulating pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulating pump is controlled by the first circulating pump control quantity.

[0069] Optionally, the first set of parameters includes the stack temperature, anode flow channel volume, membrane area, purge valve opening command, anode outlet pressure, and anode inlet pressure; The dynamic model of nitrogen partial pressure at the anode is as follows: , The partial pressure of nitrogen at the anode. This represents the nitrogen penetration flow rate, which is related to the nitrogen penetration coefficient, membrane area, cathode nitrogen partial pressure, and anode nitrogen partial pressure. The purge nitrogen flow rate is related to the anode inlet pressure, anode outlet pressure, anode nitrogen partial pressure, and the purge valve opening command. For the anode flow channel volume, Let be the ideal gas constant. This refers to the temperature of the fuel cell stack.

[0070] Optionally, the second set of parameters includes the stack current, relative humidity in the anode channel, film thickness, and film density; The dynamic model of anode water content is as follows ,in, For film thickness, For membrane density, This refers to the electroosmotic drag current, which is related to the fuel cell current. The reverse diffusion flow rate is related to the water diffusion coefficient and the water concentration difference across the membrane. The water concentration difference across the membrane is related to the relative humidity in the anode channel. The water content generated by the anodic reaction. This represents the anode water content.

[0071] Optionally, the anode prediction model is an MPC model, and the state variables of the MPC model are: The control quantity is ,in, The speed of the circulating pump. The purge valve opening is; the predicted state variables for multiple future cycles are: Multiple control sequences are ; To predict the time domain, To control the time domain.

[0072] Optionally, the objective function is: ; This is the fuel cell stack voltage. Let Q be the reference voltage and Q be the first weighting matrix. To control the increment, R is the second weight matrix. This is the weighting coefficient for hydrogen consumption. For the total duration of the sweeping, , All are positive integers.

[0073] Optionally, before obtaining the control sequence by solving the objective optimization problem based on the objective optimization function and the anode prediction model, the method further includes: obtaining the error and error change rate between the fuel cell stack voltage and the reference voltage, and adjusting the weights in the objective optimization function based on preset weight adjustment conditions.

[0074] Optionally, the preset weight adjustment conditions include: the fuzzy sets of e and Δe are {NB negative large, NM negative medium, NS negative small, ZO zero, PS positive small, PM positive medium, PB positive large}; α Q and α R The fuzzy set is {S_small, M_medium, B_large}, and the method for adjusting the weights in the objective optimization function satisfies the contents of Table 1; Table 1 Where e is the error, Δe is the rate of change of the error, and α Q α is the adjustment factor related to the first weight matrix Q. R This is the adjustment factor related to the second weight matrix R.

[0075] It is understood that the anode nitrogen and anode water management device for the fuel cell system proposed in this embodiment of the invention has the same technical effect as the method, and will not be described in detail here.

[0076] Figure 5 This is a block diagram of the fuel cell system proposed in an embodiment of the present invention, as shown below. Figure 5 As shown, the fuel cell system includes: a parameter acquisition sensor group 201, a controller 202, a purge valve 203, and a circulation pump 204; the parameter acquisition sensor group 201 is used to acquire parameters of the fuel cell, and the controller 202 is connected to the parameter acquisition sensor 201, the purge valve 203, and the circulation pump 204 respectively. The controller 202 is used to acquire parameters and execute the anode nitrogen and anode water management method of the fuel cell system according to any embodiment of the present invention to control the purge valve 203 and the circulation pump 204.

[0077] The parameter acquisition sensor group 201 may include temperature sensors, pressure sensors, humidity sensors, ammeters, or voltmeters. The controller 202 acquires the parameters of the fuel cell collected by the parameter acquisition sensor group 201, and then executes the anode nitrogen and anode water management method of the fuel cell system. It can predict the anode nitrogen concentration and anode water content to control the purge valve 203 and the circulation pump 204 to achieve the management of anode nitrogen and water.

[0078] Optionally, Figure 6 This is a structural diagram of the controller of the fuel cell system proposed in an embodiment of the present invention, as shown below. Figure 6As shown, the controller 202 includes a control chip, and an analog signal processing circuit, a digital signal output circuit, a speed control output circuit, and a CAN bus communication circuit connected to the control chip; wherein, the CAN bus communication circuit is used for communication between the vehicle and the controller, the analog signal processing circuit is used for processing parameters, the digital signal output circuit is used for outputting digital signals, and the speed control output circuit is used for outputting speed control signals.

[0079] Specifically, the controller hardware of this invention is based on a microcontroller (i.e., a control chip) and equipped with a full range of external interfaces. The controller features a high-speed CAN bus interface, which directly connects to the vehicle's existing in-vehicle network, enabling real-time data exchange with the upper-level vehicle controller. This allows for the acquisition of necessary operating condition commands and global status information, supporting intelligent decision-making. In terms of signal acquisition, the controller integrates a multi-channel analog signal processing circuit specifically designed to receive signals from various sensors, such as voltage or current signals output by temperature, pressure, and humidity sensors. This raw data forms the basis for state estimation and prediction. At the execution control level, the hardware provides diverse output interfaces to adapt to the needs of different actuators. The switching output circuit can drive the opening and closing of the anode purge valve; the speed control output circuit outputs a speed control signal, which can be a pulse width modulation signal, for stepless smooth adjustment of the anode circulation pump speed. Furthermore, a wide-range power supply circuit ensures stable connection to the vehicle's power system and incorporates efficient power management and protection circuitry to cope with complex fluctuations in the vehicle's electrical environment, ensuring continuous and stable operation of the controller.

[0080] Figure 7 This is a schematic diagram of the drive circuit for the anode purge valve proposed in an embodiment of the present invention. Figure 7 As shown, this circuit converts the low-voltage logic signal from the microcontroller into a high-current pulse that can directly drive the high-power solenoid valve, ensuring rapid response and long-term stability during the purging operation. The control signal from the general-purpose output pin of the control chip cannot directly drive the solenoid valve coil. The power switching amplifier circuit uses a field-effect transistor Q23 as the core switching element. Under the action of the control signal, this element turns on or off the drive current flowing to the purge valve coil. By designing a drive circuit to control its rapid switching with minimal self-loss, precise control of the coil current is achieved. Considering the characteristic that the solenoid valve coil generates an extremely high reverse induced electromotive force at the moment of power failure, a dedicated protection and energy discharge module is designed in the circuit. A freewheeling diode D62 is connected in parallel across the coil to form a freewheeling circuit. When the power switching transistor Q23 is suddenly turned off, the magnetic field energy stored in the coil can be safely released through this diode D62, effectively clamping down on harmful surges that may endanger the switching transistor and isolation components.

[0081] In this embodiment, the power switch Q23 is an NMOS transistor. The drain of the NMOS transistor is connected to the Valuecontrol node, and a Zener diode D61 is placed between Valuecontrol and POWER in. The drive circuit also includes a first resistor R161, a second resistor R162, and a third resistor R163. One end of the first resistor R161 is connected to the switch output circuit, and the other end is connected to the gate of the NMOS transistor. The source of the NMOS transistor is connected to one end of the third resistor R163, and the other end of the third resistor R163 is grounded. One end of the second resistor R162 is connected to the gate of the NMOS transistor, and the other end is grounded. The anode of the freewheeling diode D62 is grounded, and the cathode is connected to the gate of the NMOS transistor.

[0082] When the gate voltage is sufficiently high (above the turn-on threshold), the NMOS transistor turns on, pulling the ValueControl node low and reducing the opening of the purge valve. When the gate voltage is low, the NMOS transistor turns off, pulling the ValueControl node high via POWER_in and increasing the opening of the purge valve. Zener diode D61 limits the maximum voltage of the ValueControl node to prevent overvoltage. Freewheeling diode D62 clamps the gate voltage to prevent electrostatic discharge or voltage spikes from damaging the NMOS transistor gate.

[0083] Figure 8 This is a schematic diagram of the anode circulation pump control circuit proposed in an embodiment of the present invention. Figure 8 As shown, the anode circulating pump control circuit is used to achieve smooth, stepless speed regulation of the brushless DC motor. It translates the controller's intelligent decisions into precise modulation of the circulating pump's power output, thus dynamically adapting to the complex and ever-changing operating conditions of the fuel cell stack. The core drive component of the circuit is a three-phase full-bridge drive circuit, coupled with a dedicated motor control chip U3. Chip U3 receives speed commands from the speed control output circuit in the controller and implements normal commutation and precise control of the brushless DC motor. By adjusting the duty cycle of the PWM signal, the average voltage applied to the motor can be linearly controlled, thereby achieving smooth, continuous, stepless speed regulation over a wide range. Figure 8 This is the control circuit for one phase of a three-phase full-bridge drive circuit. It consists of two field-effect transistors (Q5 / Q6) and a drive circuit for the field-effect transistors. By controlling the on and off of the field-effect transistors (Q5 / Q6), the direction and timing of the current are switched, thereby generating a controllable three-phase AC current to control the operation of the circulating pump.

[0084] The second and third interfaces of chip U3 are used to receive speed commands from the speed control output circuit. The control circuit includes a first field-effect transistor Q5 (N-channel), a second field-effect transistor Q6 (N-channel), a first diode D7, a second diode D8, a third diode D9, a first capacitor C8, a second capacitor C19, a third capacitor C13, a fourth resistor R12, a fifth resistor R13, a sixth resistor R20, and a seventh resistor R21. When UL is high and UH is low, Q6 conducts, pulling U down to near the IOUT potential; when UH is high and UL is low, Q5 conducts, pulling U up to near the VCC potential. The two transistors in this circuit conduct alternately, generating a square wave / chopped signal at node U to drive the load. Connection circuit reference. Figure 8 This will not be elaborated upon here.

[0085] In this fuel control system, the control algorithm is implemented in C code and runs on a microcontroller with a sampling period of 100ms. (Prediction time domain) Choosing 20 means predicting the next 2 seconds, controlling the time domain. The choice is 5.

[0086] Specific control process: First, after the system is powered on, the controller initializes and loads the preset anode state prediction model parameters and the weights of the predictive control algorithm, Q and R.

[0087] Second, enter the main loop and acquire all sensor data at every 100ms sampling time.

[0088] Third, execute the state estimator using the measured values. , , , Wait, estimate the current ammonia partial pressure at the anode. and membrane moisture content .

[0089] Fourth, calculate the voltage error e(k) and the error change rate Δe(k), and input them into the fuzzy adaptive tuner. Based on the preset fuzzy rule base, retrieve the current weight adjustment factor α. Q (k) and α R (k).

[0090] Fifth, based on the estimated state Assuming initial conditions, construct the aforementioned optimization problem, where the weights in the performance metric J are represented by Q'(k) = α. Q (k)Q,R'(k)=α R (k)R.

[0091] Sixth, the embedded optimization solver is invoked to solve the quadratic programming problem online, obtaining the optimal control sequence for the purge valve and circulating pump for the next 5 cycles. .

[0092] Seventh, The first control variable in the process, namely and The purge valve and circulation pump are controlled by corresponding control circuits.

[0093] Eighth, wait for the next sampling period and repeat steps two through seven.

[0094] In this implementation, model predictive control is responsible for the main optimization decision-making, while fuzzy logic is responsible for parameter adaptation. The two work together to ensure that the system can maintain high accuracy and robustness in anode state management even in the complex environment of a real vehicle. Real-vehicle testing has shown that, compared with traditional timed purging strategies, this system can save hydrogen consumption and effectively smooth out fuel cell stack voltage fluctuations under dynamic operating conditions.

[0095] In summary, the fuel cell system and its anode nitrogen and anode water management method and apparatus proposed in this invention include: acquiring fuel cell parameters based on an acquisition cycle, the parameters including at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content; determining the current anode nitrogen partial pressure change rate based on the first set of parameters and a dynamic model of anode nitrogen partial pressure, determining the current anode water content change rate based on the second set of parameters and a dynamic model of anode water content; predicting the state variables of the fuel cell in the future for multiple cycles based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and an anode prediction model, wherein the state variables include the anode nitrogen partial pressure and the anode water content; solving the objective optimization problem based on the objective optimization function and the anode prediction model to obtain a control sequence, the control sequence including control quantities for the future for multiple cycles, the control quantities including a purge valve opening control quantity corresponding to the anode nitrogen partial pressure and a circulation pump control quantity corresponding to the anode water content, and controlling the purge valve with the first purge valve opening control quantity and the circulation pump with the first circulation pump control quantity in the control sequence. This invention, by setting up dynamic models for anode nitrogen partial pressure, anode water content, and anode prediction models, enables the early prediction of anode nitrogen partial pressure, anode water content, purge valve opening control, and circulating pump control. This allows for proactive sensing and intervention of nitrogen accumulation rate and anode water content, preventing performance degradation. Furthermore, the entire system can adapt to changes in load and environment, achieving precise on-demand purging and water balance adjustment, avoiding control lag and poor adaptability to dynamic conditions. Simultaneously, the output purge valve opening control and circulating pump control can synergistically optimize the actions of multiple actuators such as the purge valve and circulating pump, improving control accuracy and robustness.

[0096] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for managing anode nitrogen and anode water in a fuel cell system, characterized in that, include: The parameters of the fuel cell are obtained based on the acquisition cycle, and the parameters include at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content; The current rate of change of anode nitrogen partial pressure is determined based on the first set of parameters and the dynamic model of anode nitrogen partial pressure; the current rate of change of anode water content is determined based on the second set of parameters and the dynamic model of anode water content. The state variables of the fuel cell in the future multiple cycles are predicted based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model, wherein the state variables include anode nitrogen partial pressure and anode water content; The control sequence is obtained by solving the objective optimization problem based on the objective optimization function and the anode prediction model. The control sequence includes control quantities for multiple future cycles. The control quantities include the purge valve opening control quantity corresponding to the anode nitrogen partial pressure and the circulation pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulation pump is controlled by the first circulation pump control quantity.

2. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 1, characterized in that, The first set of parameters includes the stack temperature, anode flow channel volume, membrane area, purge valve opening command, anode outlet pressure, and anode inlet pressure; The dynamic model of nitrogen partial pressure at the anode is as follows: , The partial pressure of nitrogen at the anode. The nitrogen permeation flow rate is related to the nitrogen permeation coefficient, the membrane area, the partial pressure of nitrogen at the cathode, and the partial pressure of nitrogen at the anode. The purge nitrogen flow rate is related to the anode inlet pressure, the anode outlet pressure, the anode nitrogen partial pressure, and the purge valve opening command. For the anode flow channel volume, Let be the ideal gas constant. This refers to the temperature of the fuel cell stack.

3. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 1, characterized in that, The second set of parameters includes the stack current, relative humidity in the anode channel, film thickness, and film density; The dynamic model for anode water content is as follows: ,in, For film thickness, For membrane density, The electroosmotic drag flow rate is related to the current in the fuel cell stack. The reverse diffusion flow rate is related to the water diffusion coefficient and the water concentration difference across the membrane. The water concentration difference across the membrane is related to the relative humidity within the anode channel. The water content generated by the anodic reaction. This represents the anode water content.

4. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 1, characterized in that, The anode prediction model is an MPC model, and the state variables of the MPC model are: The control quantity is ,in, The speed of the circulating pump. The purge valve opening is; the predicted state variables for multiple future cycles are: The multiple control sequences are ; To predict the time domain, To control the time domain.

5. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 1, characterized in that, The objective optimization function is: ; This is the fuel cell stack voltage. Let Q be the reference voltage and Q be the first weighting matrix. To control the increment, R is the second weight matrix. This is the weighting coefficient for hydrogen consumption. For the total duration of the sweeping, , All are positive integers.

6. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 5, characterized in that, Before obtaining the control sequence by solving the objective optimization problem based on the objective optimization function and the anode prediction model, the method further includes: obtaining the error and error change rate between the stack voltage and the reference voltage of the fuel cell, and adjusting the weights in the objective optimization function based on preset weight adjustment conditions.

7. The method for managing anode nitrogen and anode water in a fuel cell system according to claim 6, characterized in that, The preset weight adjustment conditions include: the fuzzy sets of e and Δe are {NB negative large, NM negative medium, NS negative small, ZO zero, PS positive small, PM positive medium, PB positive large}; α Q and α R The fuzzy set is {S_small, M_medium, B_large}, and the method for adjusting the weights in the objective optimization function satisfies the contents of Table 1. Table 1 Where e is the error, Δe is the rate of change of the error, and α Q α is the adjustment factor related to the first weight matrix Q. R This is the adjustment factor related to the second weight matrix R.

8. A device for managing anode nitrogen and anode water in a fuel cell system, characterized in that, include: The acquisition module is used to acquire parameters of the fuel cell based on the acquisition cycle. The parameters include at least a first set of parameters related to the anode nitrogen partial pressure and a second set of parameters related to the anode water content. The state quantity observation module is used to determine the current rate of change of anode nitrogen partial pressure based on the first set of parameters and the dynamic model of anode nitrogen partial pressure, and to determine the current rate of change of anode water content based on the second set of parameters and the dynamic model of anode water content. The predictive control module is used to predict the state variables of the fuel cell in the future multiple cycles based on the current anode nitrogen partial pressure change rate, the current anode water content change rate, and the anode prediction model, wherein the state variables include the anode nitrogen partial pressure and the anode water content. The target optimization module is used to solve the target optimization problem based on the target optimization function and the anode prediction model to obtain a control sequence. The control sequence includes control quantities for multiple future periods. The control quantities include a purge valve opening control quantity corresponding to the anode nitrogen partial pressure and a circulating pump control quantity corresponding to the anode water content. The purge valve is controlled by the first purge valve opening control quantity in the control sequence, and the circulating pump is controlled by the first circulating pump control quantity.

9. A fuel cell system, characterized in that, include: Parameter acquisition sensor group, controller, purge valve and circulation pump; The parameter acquisition sensor group is used to acquire parameters of the fuel cell. The controller is connected to the parameter acquisition sensor, the purge valve and the circulation pump respectively. The controller is used to acquire the parameters and execute the anode nitrogen and anode water management method of the fuel cell system as described in any one of claims 1-8 to control the purge valve and the circulation pump.

10. The fuel cell system according to claim 9, characterized in that, The controller includes a control chip, and an analog signal processing circuit, a digital signal output circuit, a speed control output circuit, and a CAN bus communication circuit connected to the control chip. The CAN bus communication circuit is used for communication between the vehicle and the controller, the analog signal processing circuit is used for processing the parameters, the digital output circuit is used for outputting digital signals, and the speed control output circuit is used for outputting speed control signals.

Images