A method and system for suppressing anode pressure fluctuations in a fuel cell

Abstract

The application provides a fuel cell anode pressure fluctuation suppression method and system, and relates to the technical field of battery intelligent control, and comprises the following steps: real-time sensing of the stack operation condition of a target fuel cell to obtain a stack operation state parameter; based on the stack operation state parameter, critical condition recognition, pressure trend prediction and adaptive model prediction control are performed to determine optimal control instructions for balance pressure stability, action smoothness and system energy efficiency; the optimal control instructions are converted into the opening degree of a proportional valve at the front end of an ejector and the rotating speed of an anode circulating pump, the hydrogen flow and the recirculation gas amount entering the anode are finally adjusted by changing the opening degree and the rotating speed to suppress the anode pressure fluctuation of the target fuel cell; through multi-level information fusion and decision optimization, the application realizes precise and smooth control of the anode pressure, and significantly improves the dynamic performance and reliability of the fuel cell system.

Description

Technical Field

[0001] This invention relates to the field of battery intelligent control technology, specifically to a method and system for suppressing anode pressure fluctuations in a fuel cell. Background Technology

[0002] As a clean and efficient energy conversion device, the proton exchange membrane fuel cell (PEMFC) relies heavily on its anode hydrogen recirculation system to improve hydrogen utilization and prevent flooding. The ejector-circulation pump hybrid recirculation scheme has high operating efficiency and has become the mainstream technology.

[0003] However, this approach faces inherent technical challenges in the critical load range where the ejector begins to take effect: the system faces a switch from a pump-driven to an ejector-driven operating mode. In this critical region, the ejector flow and the pump-driven airflow are coupled, exhibiting strong nonlinearity and time-varying dynamic characteristics, leading to fluctuations in anode pressure. Existing technologies often rely on PID control algorithms, with fixed controller parameters. When the system operates under critical conditions, the control algorithm cannot adapt to the rapidly changing dynamic characteristics, resulting in anode pressure oscillations and a decline in control performance.

[0004] Furthermore, as the core power source of a vehicle, the fuel cell system's controller needs to be integrated into the vehicle's embedded control unit (ECU). This environment places extremely stringent requirements on the real-time performance, reliability, and computational resource consumption of the control algorithm. Many existing optimization algorithms that run well in simulation or laboratory environments are often difficult to deploy directly on vehicle controllers with limited resources and centered on microcontrollers (MCUs) due to their computational complexity and long iteration time, which greatly reduces their engineering application value.

[0005] Therefore, the hydrogen recirculation at the anode of proton exchange membrane fuel cells faces the challenge of controlling pressure fluctuations during critical load switching. The stringent resource constraints of automotive embedded systems make the development of intelligent collaborative management methods that balance high performance and real-time performance a key technological challenge. Summary of the Invention

[0006] To address the aforementioned problems, this invention proposes a method and system for suppressing anode pressure fluctuations in fuel cells. Through multi-level information fusion and decision optimization, it achieves precise and smooth control of anode pressure, significantly improving the dynamic performance and reliability of the fuel cell system.

[0007] According to some embodiments, the present invention adopts the following technical solution:

[0008] A method for suppressing anode pressure fluctuations in a fuel cell includes:

[0009] The stack operation status of the target fuel cell is monitored in real time to obtain stack operation status parameters;

[0010] Based on the stack operating state parameters, critical operating conditions are identified, pressure trends are predicted, and adaptive model predictive control is performed to determine the optimal control command that balances pressure stability, smoothness of action, and system energy efficiency.

[0011] The optimal control command is translated into the opening degree of the proportional valve at the front end of the ejector and the rotational speed of the anode circulation pump. By changing the opening degree and rotational speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are ultimately adjusted to suppress the anode pressure fluctuation of the target fuel cell.

[0012] According to some embodiments, the present invention adopts the following technical solution:

[0013] A fuel cell anode pressure fluctuation suppression system includes:

[0014] The status sensing module is configured to: sense the stack operation status of the target fuel cell in real time and obtain the stack operation status parameters;

[0015] The instruction decision module is configured to: identify critical operating conditions, predict pressure trends, and perform adaptive model predictive control based on the stack operating state parameters, and determine the optimal control instruction that balances pressure stability, smoothness of action, and system energy efficiency.

[0016] The instruction execution module is configured to convert the optimal control instruction into the opening degree of the proportional valve at the front end of the ejector and the rotational speed of the anode circulation pump. By changing the opening degree and rotational speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are ultimately adjusted to suppress the anode pressure fluctuation of the target fuel cell.

[0017] According to some embodiments, the present invention adopts the following technical solution:

[0018] A computer program product includes a computer program that, when executed by a processor, implements the aforementioned method for suppressing anode pressure fluctuations in a fuel cell.

[0019] According to some embodiments, the present invention adopts the following technical solution:

[0020] A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned method for suppressing anode pressure fluctuations in a fuel cell.

[0021] According to some embodiments, the present invention adopts the following technical solution:

[0022] An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the method for suppressing anode pressure fluctuations in a fuel cell.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0024] 1. By introducing a critical condition identifier based on multi-feature fuzzy fusion and a high-precision prediction model with adjustable parameters, this invention achieves accurate perception of the trend of anode pressure change. Compared with traditional fixed threshold switching methods or single-parameter model predictive control, this invention can identify the trend of the system entering the critical condition in advance and reduce pressure prediction error. This enables the model predictive controller to make optimization decisions based on more accurate prediction information, thereby reducing the anode pressure fluctuation amplitude during ejector / circulation pump switching.

[0025] 2. The dual-layer parameter adaptive mechanism adopted in this invention solves the core problem that fixed parameter controllers cannot adapt to the strong nonlinearity and time-varying characteristics of fuel cells. The first layer is parameter preset based on operating condition identification, which enables the controller to quickly switch control strategies according to the current operating mode. The second layer is feedback fine-tuning based on instantaneous performance, which can continuously compensate for model mismatch and external disturbances, and improve the operating condition adaptability and robustness of the control algorithm.

[0026] 3. This invention utilizes a multi-input multi-output model predictive control framework to simultaneously incorporate the control quantities of two actuators into the objective function and constraints for collaborative optimization. The control smoothing term in the objective explicitly penalizes drastic changes in the control quantity. Combined with an adaptive mechanism, this enables coordinated control of the slow opening of the ejector valve and the smooth reduction of the circulating pump speed. This fundamentally avoids abrupt changes in control commands, achieving truly smooth and disturbance-free switching and eliminating pressure shocks and system oscillations caused by rigid switching.

[0027] 4. This invention ensures the safe and reliable operation of the system through multiple mechanisms. First, state constraints (pressure safety boundary) and actuator constraints (stroke limits and rate of change limits) are explicitly added to the optimization problem of model predictive control, ensuring that the commands of the control system are within a safe range and effectively preventing risks such as overpressure and actuator saturation. Second, the parameter adaptive mechanism automatically adopts more conservative control parameters when it detects an increase in model mismatch or severe disturbances, prioritizing system stability and preventing control divergence. This inherent safety design greatly reduces the probability of system failure due to control failure.

[0028] 5. An economic term is introduced into the objective function to penalize the power consumption of the circulating pump, guiding the system to prioritize the use of the lower-energy-consumption ejector while meeting performance requirements. By adaptively adjusting λ, the system can automatically enter a more economical operating mode when dynamic performance requirements are not high. Compared with traditional timing or threshold control strategies, this invention can save auxiliary system energy consumption, and at the same time, due to more precise on-demand control, unnecessary hydrogen purging losses are reduced, thus improving the overall hydrogen utilization rate.

[0029] 6. The adaptive mechanism of this invention reduces the dependence on the accuracy of the controller's initial parameters. Although the underlying algorithm is complex, by employing engineering implementation methods such as lookup tables and piecewise linear approximation, and utilizing the computing power of modern embedded processors, the algorithm can run in real time on a cost-controllable hardware platform (sampling period ≤ 10ms). At the same time, the control algorithm has self-diagnosis and parameter self-learning capabilities, reducing the burden of on-site debugging and maintenance, which is conducive to the promotion and application of this control technology in large-scale commercial fuel cell vehicles. Attached Figure Description

[0030] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0031] Figure 1 This is a schematic diagram of the overall structure of the hydrogen supply and recirculation system for the fuel cell anode in Example 1.

[0032] Figure 2 This is a hierarchical technical architecture and information flow diagram of the control algorithm in Example 2.

[0033] Figure 3 This is a schematic diagram of the working principle of the critical condition identifier in Example 2.

[0034] Figure 4 This is a logical diagram of the two-layer parameter adaptive mechanism in Example 2.

[0035] Figure 5 This is a complete online rolling optimization flowchart for the integrated adaptive model predictive control in Example 2. Detailed Implementation

[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0037] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0038] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0039] Example 1

[0040] One embodiment of the present invention provides a method for suppressing anode pressure fluctuations in a fuel cell, comprising:

[0041] Step S1: Real-time sensing of the stack operation status of the target fuel cell to obtain stack operation status parameters;

[0042] Step S2: Based on the stack operating state parameters, critical operating conditions are identified, pressure trends are predicted, and adaptive model predictive control is performed to determine the optimal control command that balances pressure stability, smoothness of action, and system energy efficiency.

[0043] Step S3: The optimal control command is converted into the opening degree of the proportional valve at the front end of the ejector and the rotation speed of the anode circulation pump. By changing the opening degree and rotation speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are finally adjusted to suppress the anode pressure fluctuation of the target fuel cell.

[0044] As one embodiment, the present invention provides a method for suppressing anode pressure fluctuations in fuel cells, aiming to solve the pressure fluctuation problem caused by the anode ejector and circulation pump of a proton exchange membrane fuel cell under critical switching conditions. Its core is to construct a closed-loop control framework with real-time state perception, high-precision dynamic prediction and intelligent parameter adaptation capabilities. This framework achieves precise and smooth control of anode pressure through multi-level information fusion and decision optimization, significantly improving the dynamic performance and reliability of the fuel cell system. The specific implementation method is described in detail below.

[0045] I. Proton Exchange Membrane Fuel Cell (PEMFC) and Anode Hydrogen Recirculation System

[0046] Figure 1 This embodiment illustrates the overall hardware configuration and gas path connections of the fuel cell anode hydrogen supply and recirculation system, as shown below. Figure 1 As shown, the system consists of a hydrogen tank, solenoid valve, proportional valve, water separator, ejector, circulating pump, and the anode side of the fuel cell stack. The hydrogen tank, solenoid valve, and proportional valve belong to the anode hydrogen supply system; the water separator, ejector, and circulating pump belong to the recirculation system.

[0047] After pressure reduction, high-pressure hydrogen enters the ejector as a driving flow, using its jet effect to generate negative pressure and draw in circulating hydrogen from the anode outlet. The other flow serves as a supplementary gas flow. The ejector and circulating pump are installed in parallel, with the circulating pump ensuring hydrogen flow under low-load conditions and preventing flooding. After gas-liquid separation, most of the gas from the fuel cell reactor is recovered, with a small amount of waste gas discharged through a purge valve. Pressure and temperature sensors are positioned at key nodes to monitor the system status in real time. All sensor data is sent to the control unit, which, according to the method described in this embodiment, precisely adjusts the opening of the proportional valve at the ejector's front end and the rotational speed of the circulating pump to achieve intelligent and coordinated management of the anode pressure.

[0048] II. Overall Architecture Design and Control Process

[0049] The framework of this embodiment consists of three functionally defined layers: the perception layer, the decision-making layer, and the execution layer, ensuring efficient coordination of data acquisition, information processing, and control execution.

[0050] 1. The sensing layer is responsible for collecting key physical quantities reflecting the operating status of the fuel cell stack. The sensor array deployed in this layer includes:

[0051] A current sensor, connected in series in the main output circuit of the fuel cell stack (usually located between the positive or negative output terminal of the fuel cell stack and the load / inverter), is used to monitor the load current of the fuel cell stack in real time. This is the most direct parameter for judging the system's operating condition;

[0052] Pressure sensors, including an anode inlet pressure sensor and an anode outlet pressure sensor, are used to measure the pressure of hydrogen entering the fuel cell stack. The anode inlet pressure sensor is installed at the anode inlet of the fuel cell stack. Its dynamic change is the core objective of control; the anode outlet pressure sensor is installed at the anode outlet of the fuel cell stack to measure the pressure of hydrogen gas after the reaction. It can be used for system monitoring and diagnosis;

[0053] Temperature sensors are installed directly at the coolant outlet of the fuel cell stack or on the stack itself to monitor the stack temperature. This is because temperature changes directly affect the gas state and reaction rate.

[0054] All sensor signals are filtered, amplified, and converted from analog to digital by the signal conditioning circuit before being sent to the decision-making layer.

[0055] 2. The decision-making layer consists of a high-performance embedded microcontroller that runs the control algorithm proposed in this embodiment.

[0056] The algorithm software adopts a modular design, integrating three core functional modules: a condition identifier, a dynamic prediction model for anode pressure, and an adaptive model predictive controller. These modules work together to complete the entire control process from state judgment and trend prediction to optimal decision-making, which are described below:

[0057] 2.1 Critical Condition Identifier Based on Multi-Feature Fuzzy Fusion

[0058] Accurately identifying whether the system has entered the critical operating condition where the ejector takes effect is a prerequisite for effective control; traditional methods based on a single current threshold cannot adapt to dynamic changes. Therefore, this embodiment designs a fuzzy reasoning method based on multi-feature information fusion, which makes comprehensive judgments based on multiple state information to improve the accuracy and robustness of critical operating condition identification.

[0059] The input to this recognizer consists of several feature quantities that reflect the system approaching a critical state:

[0060] (1) Normalized current distance ( The calculation formula is as follows:

[0061]

[0062] in, It is the real-time collected load current of the fuel cell stack; The effective operating current range of the ejector was determined through preliminary experiments. It is the center value of this range; this variable quantitatively describes the position of the current load relative to the center of the critical interval; when When the absolute value is very small, it means that the system is near the center of the critical region.

[0063] (2) Rate of change of current ( The calculation formula is as follows:

[0064]

[0065] It characterizes the drasticness and direction of load changes; a rapidly increasing load, such as... Larger positive values ​​will cause the system to enter or cross the critical region more quickly, thus requiring a more forward-looking control strategy.

[0066] (3) Intensity of pressure fluctuation ( The standard deviation of the anode inlet pressure signal within a recent window is calculated, and the formula is as follows:

[0067]

[0068] in, This represents the average anode inlet pressure during the window period.

[0069] Near the critical operating condition, due to the instability of airflow switching, the pressure often exhibits specific oscillating characteristics. This feature can be effectively captured.

[0070] The next step is the fuzzification process, which transforms each precise input value into a fuzzy semantic description. For example, ... The system is divided into five fuzzy sets: "Negative Large (NB)", "Negative Small (NS)", "Zero (ZE)", "Positive Small (PS)", and "Positive Large (PB)". A corresponding membership function (such as a triangular or trapezoidal function) is defined for each set to determine the current precise membership. The value belongs to the "degree" of each fuzzy set.

[0071] The membership function here uses the normalized current distance as its independent variable. The precise value is the value of the fuzzy set (such as NB, NS, ZE, etc.), while the dependent variable is the membership degree of the value to a certain fuzzy set (such as NB, NS, ZE, etc.), which is between 0 and 1.

[0072] The core of fuzzy reasoning relies on a pre-defined rule base. These rules are based on a deep understanding of the system's physical characteristics and a summary of extensive experimental data. The rule format is typically: "IF Condition 1 AND Condition 2…THEN Conclusion"; for example, a typical rule might be: ZE (Current current is close to the center of the critical region) P (Load is increasing) M (Moderate fluctuations have occurred) THEN Critical operating condition confidence level High (high probability of being in a critical state).

[0073] The system will activate all relevant rules and ultimately merge the outputs of all rules into a precise critical condition confidence score using a "defuzzification" method (such as the centroid method). Its value is between 0 and 1. The closer the confidence level is to 1, the greater the likelihood that the system is in a critical switching state; this continuously changing confidence level provides richer information for subsequent smooth control than a simple Boolean judgment (yes / no).

[0074] The core idea of ​​the centroid method is to calculate the "centroid" or "gravity" of the synthesized fuzzy set of all activated rules, and use the x-coordinate value of this centroid as the final precise output. The standard calculation formula is as follows:

[0075]

[0076] in, The calculated confidence level for the critical operating condition is the final result.

[0077] The number of fuzzy rules that are activated;

[0078] For the first The central value of the fuzzy set corresponding to the conclusion of each rule; for example, if the conclusion is "high confidence", then the central value of this "high" set is in the universe of discourse. The set can be defined as a triangle membership function, where the x-coordinate of its vertex (e.g., 0.8) is the center value of the set. .

[0079] For the first The activation strength of a rule is obtained by performing an AND operation (usually taking the minimum value) on all the preconditions of that rule, and represents the confidence level of the rule's conclusion under the current input.

[0080] 2.2 Parameter-Adjustable Dynamic Prediction Model for Anode Pressure

[0081] The effectiveness of model predictive control is highly dependent on the accuracy of its internal predictive model. To accurately describe the dynamic characteristics of the anode pressure under the critical operating conditions of the ejector, this embodiment constructs a high-precision predictive model that combines mechanism and data-driven approaches and allows for online parameter adjustment. The core of this model is based on the law of conservation of mass and the ideal gas law, specifically:

[0082] 2.2.1 Theoretical Basis of the Model and Derivation of Core Equations

[0083] Treating the anode flow channel as a control volume, the mass of hydrogen within the volume is controlled according to the law of conservation of mass. The rate of change is equal to the inflow mass flow rate Subtract outflow mass flow rate and the mass flow rate consumed in the reaction This can be expressed as a formula:

[0084]

[0085] For the anode recirculation loop, the outflow mass flow rate is as follows: To set the value to 0, we introduce the ideal gas law. Where R is the ideal gas constant, (M is the molar mass of hydrogen gas), which can be expressed as mass. With pressure Connecting them; anode volume By taking the derivative of both sides of the equation of state and substituting it into the mass conservation equation, and then rearranging, we finally obtain the description of the anode inlet pressure. The core differential equation that changes dynamically:

[0086]

[0087] Where M is the molar mass of hydrogen gas. This is the volume of the anode.

[0088] To better integrate with traffic, quality traffic is typically... Convert to molar flow rate Substituting into the above equation, the core equation of the dynamic prediction model for anode pressure can be simplified to:

[0089]

[0090] in, This is the absolute pressure at the anode inlet, measured in Pascals (Pa). This is the core state variable of the model and the target that the control system needs to stably track.

[0091] The absolute temperature of the fuel cell stack, measured in Kelvin (K). Temperature directly affects the density and volume of a gas, and is a key parameter for accurately calculating the number of moles of a gas.

[0092] The volume of the anode flow channel is expressed in cubic meters. This parameter is a fixed parameter determined by the structure of the fuel cell.

[0093] The molar flow rate of hydrogen gas consumed in the electrochemical reaction, expressed in moles per second. The calculation formula is directly determined by the current operating state of the fuel cell stack:

[0094]

[0095] in, It refers to the number of individual cells in the fuel cell stack. It is the real-time collected load current of the fuel cell stack (unit: amperes). F is the Faraday constant ( This formula reflects the stoichiometric relationship between current and hydrogen consumption in an electrochemical reaction.

[0096] The total hydrogen molar flow rate entering the anode channel from the outside ( This parameter represents the flow rate contributed by the ejector. and the flow rate contributed by the circulating pump Sum:

[0097]

[0098] 2.2.2 Ejector Flow Model and Its Adaptive Correction

[0099] The ejector's operating characteristics are complex, and its flow rate is affected by the pressure difference across the nozzle, valve opening, and flow state. This embodiment uses an improved equivalent nozzle model to describe its flow characteristics:

[0100]

[0101] in, This is the adaptive correction coefficient for the operating condition, which is the confidence level of the critical operating condition calculated earlier. The function, i.e. When a critical operating condition with high confidence is identified (e.g.) The flow inside the ejector may become unstable, posing a risk of flow separation or slight surge. In this case, calibration using experimental data or expert knowledge... Set it to a value slightly less than 1, such as between 0.92 and 0.98, to compensate for the prediction bias of the model under this special working condition, so that the prediction results are closer to reality.

[0102] The flow coefficient is the proportional valve opening of the ejector. The function, whose relationship is obtained through bench test data, reflects the flow efficiency of the valve at different opening degrees.

[0103] The reference cross-sectional area of ​​the ejector nozzle ( ), which is a fixed geometric parameter.

[0104] The density of hydrogen gas ( The pressure and temperature of hydrogen can be calculated from the current hydrogen pressure and temperature according to the ideal gas law.

[0105] The supply pressure of high-pressure hydrogen ( ).

[0106] The critical flow function describes the choking phenomenon of a compressible fluid (hydrogen) in a constricted flow channel. Its characteristic is that when the back pressure ratio... Below the critical pressure ratio (approximately 0.53 for hydrogen), the gas velocity at the outlet reaches the local speed of sound, and the flow rate is no longer affected by the downstream back pressure. The flow rate reaches its maximum value; when the back pressure ratio is higher than the critical value, the flow is in a subcritical state, and the flow rate decreases as the back pressure increases. .

[0107] 2.2.3 Circulating Pump Flow Model

[0108] A circulating pump can be modeled as a linear or near-linear function of the pump speed:

[0109]

[0110] in, For the volumetric efficiency of the circulating pump, factors such as internal leakage of the pump are taken into account.

[0111] This refers to the pump's current speed (in revolutions per minute); it is determined by the duty cycle of the PWM signal output by the controller. Decide.

[0112] This is the maximum safe operating speed allowed for the pump.

[0113] The maximum volumetric flow rate that the pump can provide at its maximum speed ( When using it, it needs to be converted into molar flow rate according to the current temperature and pressure state of the gas. ( ).

[0114] 2.2.4 Model Discretization and Online Application

[0115] To implement the predictive function in a digital controller, the aforementioned continuous differential equation model needs to be discretized. Using the first-order Euler method or the more precise fourth-order Runge-Kutta method, the core equations of the anode pressure dynamic prediction model are transformed into a discrete state-space form suitable for computer computation.

[0116]

[0117] in, For the next discrete time point Predicted anode inlet pressure value;

[0118] For the current discrete time point The actual measured anode inlet pressure;

[0119] For the current discrete time point The opening degree of the ejector proportional valve;

[0120] For the current discrete time point The duty cycle of the PWM signal output by the controller;

[0121] For the current discrete time point The measured load current output by the fuel cell stack;

[0122] For the current discrete time point The measured temperature of the fuel cell stack;

[0123] For the current discrete time point of calculation The critical operating condition confidence level has a value between 0 and 1. The closer it is to 1, the greater the likelihood that the system is in a critical switching state.

[0124] This discrete model will be embedded into the predictive module of the controller. In each control cycle, the controller will predict the current time. Measured values Control sequence of future hypotheses and the confidence level of the critical operating condition calculated in real time. This model is called to recursively predict the anode inlet pressure trajectory at multiple future time points. This provides a basis for subsequent rolling optimization.

[0125] 2.3 Adaptive Model Predictive Controller

[0126] 2.3.1 Controller Parameter Set

[0127] The core innovation of this embodiment lies in enabling traditional model predictive control to have the ability to self-tune key parameters online, thereby actively adapting to changes in the dynamic characteristics of the system. Especially under complex critical switching conditions, a two-layer parameter adaptive mechanism is designed. This mechanism can both perform feedforward adjustment according to the operating conditions and perform feedback optimization according to the actual stack operating state of the fuel cell.

[0128] The adjustable key parameter set of the controller is defined as follows: The parameters are defined as follows:

[0129] The prediction time domain determines how many future time steps the controller algorithm can predict regarding the system behavior; a longer prediction time domain results in a more predictable system behavior. This usually means better stability and robustness, but also requires more computation.

[0130] The control time domain defines the number of future control steps that the optimizer can change.

[0131] The weighting matrix of the state (output) tracking error determines the controller's response to pressure deviation. The degree of importance attached to it; The larger the value, the more the controller is committed to quickly eliminating pressure errors.

[0132] and The weight matrix of the control quantity changes corresponds to the changes in the ejector valve opening. and changes in circulating pump speed Larger or The value will penalize drastic changes in control actions, making control smoother, but may sacrifice responsiveness.

[0133] Economic weighting coefficient, used to penalize the power consumption of the circulating pump in the objective function. The guidance system prioritizes the use of ejectors with lower energy consumption while meeting performance requirements.

[0134] 2.3.2 Implementation Process of Two-Level Parameter Adaptive Control Mechanism

[0135] The implementation process of the controller's two-level parameter adaptive control mechanism is as follows:

[0136] First layer: Pre-set feedforward parameters based on working condition identification

[0137] This layer acts as the decision-making layer of the control system, calculating the critical operating condition confidence level in real time based on the operating condition identifier. Through a pre-defined mapping function that has been optimized through simulation and experiments. Quickly generate a set of basic parameters for the MPC controller. :

[0138]

[0139] The strategy of this mapping function is as follows:

[0140] when ， At this point, the circulating pump dominates the mode: the system is relatively stable and far from the critical region; the strategy here prioritizes economy and smoothness, therefore... The function will output a set of preset parameters: a longer prediction time domain. To enhance robustness, medium-sized tracking weights Larger control variable weights and To ensure smooth movements and a greater emphasis on economic efficiency. To reduce pump consumption.

[0141] when At this point, the system enters a critical switching mode: it enters a region of critical dynamic complexity; the strategy shifts towards dynamic performance and rapid response. At this time, the mapping function is adjusted to another set of preset parameters: a moderate prediction time domain. To balance predictive power and computational speed, the tracking weights are increased. To suppress pressure fluctuations, a moderate control weight is used. and To prevent overshoot and reduce economic weights For the time being, priority will be given to ensuring control performance rather than energy efficiency.

[0142] when At this point, the system enters ejector-dominated mode: the switchover is largely complete, but stability must be ensured; the strategy focuses on a stable transition and energy efficiency optimization, with parameters set to a medium-to-long duration. ,medium smaller (The ejector is now working stably), larger (The pump will shut down smoothly), and the largest To maximize the use of the ejector.

[0143] Second layer: Closed-loop parameter fine-tuning based on instantaneous performance feedback

[0144] This layer is used to fine-tune the control parameters output by the first layer, using the "coarse-tuning" parameters provided in the first layer. Based on this, a closed-loop optimizer is introduced to make fine adjustments according to the instantaneous performance of the control system in order to compensate for the effects of model errors and random disturbances.

[0145] First, define an instantaneous performance metric. This indicator is used to quantify the control effect at the current moment, and its calculation method is as follows:

[0146]

[0147] Among them, This is the current pressure tracking error. These are all changes in the current control variables, namely changes in the ejector valve opening and changes in the circulating pump speed. and It is a weight used to balance tracking performance and the intensity of control actions.

[0148] Then, online optimization is performed using the gradient descent method; for the parameter set... One of the parameters Its adjustment amount The calculation is based on the gradient estimate of the parameter by the performance index:

[0149]

[0150] In practical applications, gradient Since direct analytical solutions are difficult to obtain, numerical approximation methods are used for calculation.

[0151]

[0152] in, It is for parameters The learning step size, and at the same time, to prevent over-adjustment or oscillation of parameters, the adjustment amount needs to be adjusted. Limitations can be imposed, and a dead zone can be set so that parameters are only updated when performance changes exceed a certain threshold.

[0153] Finally, the parameter set after two layers of adaptive adjustment is:

[0154]

[0155] Where ΔΘ(k) represents all the parameters that need to be adjusted at a specific time k. Adjustment amount A set of.

[0156] 2.3.3 Online rolling optimization process with integrated adaptive control algorithm:

[0157] Integrating the above components, each control cycle (sampling time) The complete algorithm flow within is as follows:

[0158] (1) Initialization: The system starts up and loads the default controller parameters. Initialize the state estimator.

[0159] (2) Main loop (at each sampling time) ):

[0160] 1) Data Acquisition and Preprocessing: Reading data from sensors Wait for the latest fuel cell stack operating status parameter data.

[0161] 2) Operating Condition Identification: Call the operating condition identifier to calculate the confidence level of the current critical operating condition. .

[0162] 3) Adaptive parameter adjustment:

[0163] First layer (feedforward preset): Based on Through mapping function Computational reference parameter set

[0164] Second layer (feedback fine-tuning): based on recent performance metrics Changes in calculation parameters This yields the final adaptive parameter set. .

[0165] 4) State estimation and prediction: based on the current time. Measured values Control sequence of future hypotheses and the confidence level of the critical operating condition calculated in real time. Using a parameter-adjustable prediction model, calculate the future Step Anode Pressure Trajectory .

[0166] 5) Problem Formulation and Solution: Using Adaptive Parameter Sets Based on the anode pressure trajectory, a finite-time optimization problem of the following form is constructed. The objective function consists of a pressure tracking term, a control smoothing term, and an economy term. The pressure tracking term penalizes the deviation between the predicted pressure and the reference pressure. The control smoothing term penalizes drastic changes in the ejector proportional valve opening and the circulating pump speed. The economy term penalizes the operating power consumption of the circulating pump. This can be expressed by the following formula:

[0167]

[0168] At the same time, a series of constraints must be met, including:

[0169] State constraints: This refers to the pressure safety boundary, which ensures that the pressure remains within a safe range.

[0170] Input constraints: That is, the travel limit, the physical limit of the actuator.

[0171] Input rate constraints: This refers to the rate of change limit, which restricts the speed of the actuator to ensure smooth operation.

[0172] Subsequently, an embedded optimization solver (such as the effective set method or interior point method) is invoked to solve this constrained quadratic programming (QP) problem, yielding the future... Optimal control sequence of steps .

[0173] (3) Control implementation: The rolling optimization strategy in predictive control is adopted, and only the first control variable in the optimal control sequence is used. The actual output is supplied to the actuators of the ejector proportional valve and the circulating pump.

[0174] (4) State update and loop: Update time index, Then wait for the next sampling period to arrive, and then return to step a to start a new cycle of perception, decision-making and execution.

[0175] 3. The execution layer is responsible for converting the control commands from the decision-making layer into control quantities for the ejector proportional valve and the anode circulation pump, which are the two key actuators for operation control.

[0176] These two components receive instructions from the controller and, by changing the valve opening and pump speed, ultimately regulate the flow rate of hydrogen entering the anode and the amount of recirculated gas.

[0177] 4. Overall control process

[0178] Within each fixed sampling period (10 milliseconds):

[0179] The latest data from the perception layer is first sent to the operating condition identifier, which quickly determines whether the current operating mode of the system is close to the critical point at which the ejector takes effect.

[0180] The identification results are passed to the prediction model, which predicts the trend of anode pressure changes over a future period based on the current stack operating status and different assumed control actions.

[0181] Finally, the adaptive model predictive controller integrates the current stack operating status and the predicted trend of anode pressure changes over a future period. Using optimization algorithms, it calculates the optimal control command that best balances pressure stability, smooth operation, and system energy efficiency, and issues it to the ejector valve and circulation pump at the actuator level. After the actuators operate, the system state changes, and a new cycle of perception, decision-making, and execution begins, thus forming a continuously optimized intelligent control closed loop.

[0182] 5. Real-time performance guarantee and hardware implementation

[0183] To ensure that the complex algorithm meets the stringent real-time requirements in automotive controllers, this invention has been specifically optimized at both the software and hardware levels.

[0184] 1. Computational optimization:

[0185] The fuzzy reasoning process in working condition identification is transformed into a lookup table through offline computation, allowing for direct lookup of the table during online runtime. This greatly reduces computation time.

[0186] For complex nonlinear functions in the prediction model (such as the critical stream function) This can be achieved using piecewise linear approximation or pre-computed lookup tables.

[0187] We select an efficient QP solving algorithm optimized for embedded systems and make full use of the processor's hardware acceleration capabilities.

[0188] 2. Task Scheduling:

[0189] The runtime of the adaptive model predictive controller, especially the computationally intensive second-layer fine-tuning algorithm, can be set to an integer multiple of the main loop cycle, such as performing parameter adjustments every 5 or 10 MPC cycles. This ensures that the parameters follow system changes while avoiding extensive calculations in each cycle, thus balancing computational load and adaptive speed.

[0190] 3. Hardware Platform:

[0191] The core processor is a high-performance digital signal processor (DSP) or a high-performance microcontroller, such as TI's C2000 series DSP or an ARM Cortex-M7 core MCU. These processors feature high clock speeds, hardware floating-point units, and ample on-chip memory (RAM and Flash), providing hardware support for the operation of complex algorithms. Reliable sensor signal conditioning circuits and actuator drive circuits are designed to ensure the accuracy of data acquisition and the execution of control commands.

[0192] Example 2

[0193] One embodiment of the present invention provides a method for suppressing anode pressure fluctuations in a fuel cell, which... Figure 1 The fuel cell anode hydrogen supply and recirculation system shown here suppresses anode pressure fluctuations. Figure 2 The hierarchical technical architecture and information flow of the core control algorithm in this embodiment are described. The architecture is divided into three core levels from left to right, forming a complete intelligent control closed loop.

[0194] The leftmost layer is the sensing and operating condition recognition layer, which is responsible for collecting and preprocessing sensor data in real time, and calling the multi-feature fuzzy fusion critical operating condition recognizer. By analyzing multiple feature quantities such as current and pressure change trends, it determines whether the system is currently in the circulation pump-dominated, critical switching, or ejector-dominated mode, and outputs a quantified confidence level.

[0195] The middle layer is the prediction and adaptation layer, which mainly includes the dynamic prediction model of anode pressure. It can predict the changing trend of anode pressure based on the current state and operating condition confidence level, providing a reference for the execution of the control algorithm.

[0196] The rightmost layer is the decision-making and execution layer, namely the adaptive model predictive controller. Through a two-layer parameter adaptive mechanism, it adjusts the key parameter set of the model predictive controller, such as the prediction time domain and control weights, online based on the operating condition confidence and real-time control performance. In addition, it can perform rolling optimization calculations to generate optimal control commands and issue them to the actuators.

[0197] In this embodiment, the core controller adopts a 32-bit multi-core microcontroller with a built-in hardware floating-point arithmetic unit and lockstep core; the controller communicates with the vehicle's upper-level VCU via the CAN bus to obtain driving commands; the sensor system includes: a current sensor ( ), pressure sensor ( PT100 platinum resistance temperature sensor The actuators are a proportional ejector control valve and a three-phase brushless DC circulating pump.

[0198] After the system is powered on, the controller initializes and loads preset parameters. The baseline parameter for model predictive control is set as follows: sampling period. Predicting the time domain The corresponding time is 0.3 seconds, controlling the time domain. Initial values ​​for the weight matrix: pressure-tracking weights Weighting of ejector valve opening change Weighting of circulating pump speed change Economic weight .

[0199] I. Operating Condition Identifier

[0200] Figure 3 This is the working principle of the critical condition identifier, which includes four key steps:

[0201] First, there is multi-feature extraction. The recognizer calculates key feature quantities from sensor data, including normalized current distance to reflect the relative position of the load, current change rate to characterize the load change trend, and pressure fluctuation intensity to capture specific oscillation patterns.

[0202] Next comes the fuzzification process, which transforms each precise feature value into a membership degree to different fuzzy linguistic variables.

[0203] Then comes fuzzy reasoning, which is the core step. The system performs logical reasoning based on a preset fuzzy rule base. The rule form is usually that if a certain feature satisfies a certain condition, then the critical confidence level belongs to a certain state.

[0204] Finally, there is the defuzzification step, which merges the results of fuzzy inference and transforms them into a precise critical operating condition confidence value.

[0205] Compared to a single threshold judgment, this method can identify the critical state of the system earlier and more accurately.

[0206] In this embodiment, the critical condition identifier is implemented as an independent task module in software, and its implementation steps are as follows:

[0207] 1. Feature Calculation: Read every 10ms cycle. Calculate the normalized current distance Simultaneously, the rate of change of current is calculated. The standard deviation of the anode pressure signal over the past 50 sampling points is used as the pressure fluctuation intensity. .

[0208] 2. Blurring: For Define the membership functions for each triangle. For example, The domain of discourse is [-1,1], and its fuzzy set is... The center points are set to -1, -0.5, 0, 0.5, and 1, respectively.

[0209] 3. Rule Base and Reasoning: A fuzzy rule base containing 15 rules is built based on expert experience. For example, the rule: The Mamdani min-maximum inference method and the centroid method are used to defuzzify the data, and the confidence score is output. .

[0210] II. Parameter Adaptive Model Predictive Control

[0211] Figure 4 The logical structure of a two-layer parameter adaptive mechanism is demonstrated, which aims to dynamically adjust the key parameter set of the model predictive controller to adapt it to different operating conditions.

[0212] The first layer is a feedforward parameter preset layer based on operating condition identification. This layer quickly queries or calculates a set of basic parameters based on the real-time calculated critical operating condition confidence level through a preset mapping function. For example, when the confidence level is high, the preset parameters will tend to have a shorter prediction time domain and a higher tracking weight, so that the controller responds more quickly.

[0213] The second layer is a closed-loop parameter fine-tuning layer based on performance feedback. On top of the basic parameters provided by the first layer, this layer uses optimization algorithms such as gradient descent to make small-scale fine adjustments to the parameters based on the changes in instantaneous control performance indicators, in order to compensate for model errors and disturbances.

[0214] Ultimately, the two adjustment values ​​are superimposed to form a complete adaptive parameter set for the current control cycle, thereby enabling the controller to always maintain optimal performance.

[0215] Figure 5 The complete online rolling optimization process integrating adaptive model predictive control is illustrated in flowchart form:

[0216] The process begins with system initialization and parameter loading, followed by the main loop. Within each sampling period, data acquisition and preprocessing are performed first. Next, the critical condition confidence level is calculated based on the acquired data. Then, a two-layer parameter adaptation is executed to obtain adjusted controller parameters. The next step is to use the current state as initial conditions and perform state prediction using the adjusted prediction model. Afterward, a constrained optimization problem is constructed and solved to obtain the optimal control sequence for a future period. Immediately following, the instantaneous control quantities from the optimal sequence are applied to the actuators. Finally, the process waits for the next sampling period and returns to the data acquisition step to begin a new cycle. This process repeats continuously, achieving continuous closed-loop optimal control of the anode pressure.

[0217] The online execution flow of the adaptive model predictive control algorithm is as follows:

[0218] 1. Data Acquisition and Operating Condition Identification: At each sampling time k, data is collected. and call the recognizer to get .

[0219] 2. Two-layer parameter adaptation:

[0220] First layer (feedforward preset): Based on The value of is determined by looking up a table to determine the basic parameters.

[0221] The second layer (feedback fine-tuning): calculates the integral absolute error of the pressure tracking error over the past second as a performance indicator. .like If the increase exceeds 10% compared to the previous cycle, then... Fine-tune, or decrease. Fine-tune within ±20%.

[0222] 3. Rolling optimization and control implementation:

[0223] Based on the current state estimate As initial conditions, use the adjusted parameter set. Based on the prediction model, an optimization problem is constructed. An embedded QP solver is then used to solve the problem, yielding the optimal control sequence. The first control variable in the sequence. The signal is converted into a PWM signal to control the solenoid valves of the circulating pump and ejector valve.

[0224] Example 3

[0225] One embodiment of the present invention provides a fuel cell anode pressure fluctuation suppression system, comprising:

[0226] The status sensing module is configured to: sense the stack operation status of the target fuel cell in real time and obtain the stack operation status parameters;

[0227] The instruction decision module is configured to: identify critical operating conditions, predict pressure trends, and perform adaptive model predictive control based on the stack operating state parameters, and determine the optimal control instruction that balances pressure stability, smoothness of action, and system energy efficiency.

[0228] The instruction execution module is configured to convert the optimal control instruction into the opening degree of the proportional valve at the front end of the ejector and the rotational speed of the anode circulation pump. By changing the opening degree and rotational speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are ultimately adjusted to suppress the anode pressure fluctuation of the target fuel cell.

[0229] Example 4

[0230] One embodiment of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned method for suppressing anode pressure fluctuations in a fuel cell.

[0231] Example 5

[0232] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned method for suppressing anode pressure fluctuations in a fuel cell.

[0233] Example 6

[0234] One embodiment of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the aforementioned method for suppressing anode pressure fluctuations in a fuel cell.

[0235] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0236] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0237] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method of suppressing pressure fluctuations in an anode of a fuel cell, characterized by, include: The stack operation status of the target fuel cell is monitored in real time to obtain stack operation status parameters; Based on the stack operating state parameters, critical operating conditions are identified, pressure trends are predicted, and adaptive model predictive control is performed to determine the optimal control command that balances pressure stability, smoothness of action, and system energy efficiency. The optimal control command is translated into the opening degree of the proportional valve at the front end of the ejector and the speed of the anode circulation pump. By changing the opening degree and speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are ultimately adjusted to suppress the anode pressure fluctuation of the target fuel cell. The critical operating condition identification is based on the real-time sensed stack operating status parameters, and calculates the current critical operating condition confidence of the target fuel cell through a multi-feature fuzzy fusion algorithm. The input variables of the multi-feature fuzzy fusion algorithm include normalized current distance, current change rate, and pressure fluctuation intensity. The normalized current distance is calculated based on the current stack load current and the preset effective operating current range of the ejector. The pressure fluctuation intensity is calculated by performing standard deviation calculation on recent anode pressure signals. The pressure trend prediction is achieved by using the stack operating status parameters and critical condition confidence levels, and through a parameter-adjustable dynamic prediction model of anode pressure, to predict the trajectory of anode pressure changes in the future time domain. The adjustable-parameter dynamic prediction model for anode pressure is based on the law of conservation of mass, and its differential equation is expressed as: in, It is the absolute pressure at the anode inlet. It is the ideal gas constant. It is the absolute temperature of the fuel cell stack. It is the volume of the anode flow channel. It is the molar flow rate of hydrogen gas consumed in the electrochemical reaction. This is the total molar flow rate of hydrogen entering the anode channel from the outside; this parameter represents the flow rate contributed by the ejector. and the flow rate contributed by the circulating pump sum; The adaptive model predictive control includes: A two-layer parameter adaptive mechanism is adopted to adjust the key parameter set of the model predictive control algorithm online based on the critical operating condition confidence and instantaneous control performance index. The key parameter set includes at least the prediction time domain, control time domain, state tracking weight matrix, and control quantity change weight matrix. Using the adjusted set of key parameters and the trajectory of anode pressure changes in the future time domain, a constrained finite-time optimization problem is constructed and solved to generate optimal control commands for the coordinated control of the ejector proportional valve and the anode circulation pump. The two-layer parameter adaptive mechanism includes: First-layer feedforward preset: Based on the critical operating condition confidence level, a set of basic parameters are obtained by querying or calculating through a preset mapping function; The second layer of feedback fine-tuning: using the basic parameters as initial values, and based on the changes in instantaneous performance indicators, the key parameter set is fine-tuned online using the gradient descent method.

2. The method for suppressing anode pressure fluctuations in a fuel cell as described in claim 1, characterized in that, The stack operating parameters include at least the stack load current, anode inlet pressure, anode outlet pressure, and stack temperature.

3. The method for suppressing anode pressure fluctuations in a fuel cell as described in claim 1, characterized in that, The objective function of the finite-time optimization problem consists of three terms: The pressure tracking term is used to penalize deviations between predicted pressure and reference pressure. A smoothing control term is used to penalize drastic changes in ejector proportional valve opening and circulating pump speed; The economics item is used to penalize the operating power consumption of the circulating pump.

4. A fuel cell anode pressure fluctuation suppression system, characterized in that, The method for suppressing anode pressure fluctuations in a fuel cell as described in any one of claims 1-3 includes: The status sensing module is configured to: sense the stack operation status of the target fuel cell in real time and obtain the stack operation status parameters; The instruction decision module is configured to: identify critical operating conditions, predict pressure trends, and perform adaptive model predictive control based on the stack operating state parameters, and determine the optimal control instruction that balances pressure stability, smoothness of action, and system energy efficiency. The instruction execution module is configured to convert the optimal control instruction into the opening degree of the proportional valve at the front end of the ejector and the rotational speed of the anode circulation pump. By changing the opening degree and rotational speed, the flow rate of hydrogen entering the anode and the amount of recirculated gas are ultimately adjusted to suppress the anode pressure fluctuation of the target fuel cell.

5. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement a fuel cell anode pressure fluctuation suppression method as described in any one of claims 1-3.

6. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform a method for suppressing anode pressure fluctuations in a fuel cell as described in any one of claims 1-3.

Images