Intelligent dynamic air-hydrogen pressure difference control method and control system

Abstract

The invention relates to the technical field of new energy automobile fuel cells, and provides an intelligent dynamic air-hydrogen pressure difference control method and system, and the method comprises the steps: monitoring a first pressure value of an air channel and a second pressure value of a hydrogen channel in a fuel cell stack in real time; acquiring a real-time air-hydrogen pressure difference between the air passage and the hydrogen passage according to the first pressure value and the second pressure value, and generating a control signal according to the real-time air-hydrogen pressure difference; the system state is dynamically adjusted according to the control signal, and it is ensured that the real-time air-hydrogen pressure difference is always kept within the safety interval in the operation process of the fuel cell stack. According to the invention, real-time dynamic regulation and control are carried out on the pressure difference of the air-hydrogen double-flow passage during the operation of the fuel cell system, so that the real-time air-hydrogen pressure difference is always maintained in a safe interval during the operation process of the fuel cell stack, and the membrane assembly damage caused by pressure fluctuation is prevented, thereby improving the stability and service life of the fuel cell system.

Description

Technical Field

[0001] This invention relates to the field of fuel cell technology for new energy vehicles, and more specifically, to an intelligent dynamic air-hydrogen pressure differential control method and control system. Background Technology

[0002] A hydrogen fuel cell system is a power generation device that directly converts the chemical energy of hydrogen and oxygen into electrical energy. Its basic principle is that hydrogen is decomposed into protons and electrons through the action of a catalyst. The protons react with oxygen through a proton exchange membrane to produce water, while the electrons flow from the positive electrode to the negative electrode through an external circuit to output electrical energy.

[0003] In existing fuel cell systems, pressure fluctuations in the air-hydrogen dual-flow path can easily lead to pressure differential reversal, which in turn can cause membrane module damage or performance degradation, affecting system reliability and lifespan. To prevent membrane module damage due to pressure fluctuations and ensure stable membrane module operation, it is necessary to develop an intelligent dynamic air-hydrogen pressure differential control method and control system. Summary of the Invention

[0004] Therefore, in order to prevent membrane modules from being damaged due to pressure fluctuations and to ensure the stable operation of membrane modules, this invention provides an intelligent dynamic air-hydrogen pressure differential control method and control system, the specific technical solution of which is as follows:

[0005] A smart dynamic air-hydrogen pressure differential control method includes the following steps:

[0006] Real-time monitoring of the first pressure value of the air passage and the second pressure value of the hydrogen passage within the fuel cell stack;

[0007] The real-time air-hydrogen pressure difference between the air passage and the hydrogen passage is obtained based on the first pressure value and the second pressure value, and a control signal is generated based on the real-time air-hydrogen pressure difference.

[0008] The system state is dynamically adjusted according to the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack.

[0009] Wherein, the real-time air pressure difference is equal to the second pressure value minus the first pressure value.

[0010] The intelligent dynamic air-hydrogen differential pressure control method generates a control signal by acquiring the real-time air-hydrogen differential pressure. It dynamically adjusts the pressure difference between the air and hydrogen dual-flow paths during the operation of the fuel cell system, ensuring that the real-time air-hydrogen differential pressure is always maintained within a safe range during the operation of the fuel cell stack. This prevents damage to the membrane module due to pressure fluctuations, thereby improving the stability and service life of the fuel cell system.

[0011] Preferably, the control method further includes the following steps:

[0012] Based on the pressure difference between the hydrogen passage and the air passage, different air-hydrogen pressure difference ranges are defined;

[0013] Different system control strategies are preset for different air-hydrogen pressure difference ranges.

[0014] Preferably, the specific method for dividing different air-hydrogen pressure difference ranges includes the following steps:

[0015] When the pressure difference level is less than 0 kPa, it is designated as a danger zone;

[0016] When the pressure difference level falls within the range of [0 kPa, 2 kPa), it is classified as a slightly abnormal zone.

[0017] When the pressure difference level is within [2kPa, 10kPa), it is classified as the normal working zone;

[0018] When the pressure difference level is within [10kPa, 15kPa], it is classified as a high pressure difference warning zone;

[0019] When the pressure difference level is greater than 15 kPa, it is classified as a dangerous high-pressure zone.

[0020] Preferably, the specific method for generating a control signal based on the real-time air-hydrogen pressure difference includes the following steps:

[0021] If the real-time air-hydrogen pressure difference is in a slightly abnormal zone, a control signal is generated to adjust the opening of the electronic regulating valve A in the air passage or the electronic regulating valve B in the hydrogen passage.

[0022] If the real-time air pressure difference falls within the normal operating range, a control signal is generated to maintain the current operating state.

[0023] Preferably, the control method further includes the following steps:

[0024] Based on the LSTM network, the predicted air-hydrogen pressure difference value for the next N steps is obtained, the target reference pressure difference value is obtained, and the air-hydrogen pressure difference tracking term is obtained to represent minimizing the pressure difference prediction error.

[0025] The valve opening measure at step k is obtained based on the LSTM network, and a valve opening measure penalty term for controlling the valve action amplitude is obtained based on the valve opening measure.

[0026] Based on the real-time air-hydrogen pressure difference, a risk penalty item is obtained to indicate that the penalty value increases exponentially when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone.

[0027] An optimization function is obtained based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and a predictive control model is constructed based on the optimization function.

[0028] The optimized valve opening sequence is obtained based on the predictive control model, and the valve opening is controlled based on the valve opening sequence.

[0029] Preferably, the optimization function is expressed as follows:

[0030] Wherein, ΔP pred (k), ΔP ref Let Q and R represent the predicted air-hydrogen pressure difference and the target reference pressure difference, respectively; u represents the valve opening sequence of the electronic control valve; Q and R represent the weight matrix of the output tracking error and the valve opening measurement weight matrix, respectively; Δu(k) represents the valve opening measurement at the k-th step; ρ represents the risk penalty coefficient; and Risk = e ΔP-15 Indicates risk penalty items, These represent the air-hydrogen pressure difference tracking term and the valve opening penalty term, respectively.

[0031] Preferably, the control method further includes the following steps:

[0032] Obtain the pressure difference ratio between the absolute value of the real-time air-hydrogen pressure difference and the target reference pressure difference value, and obtain the pressure difference relative deviation term based on the pressure difference ratio;

[0033] Obtain the real-time temperature value of the fuel cell stack, and obtain the temperature change rate term based on the real-time temperature value;

[0034] Obtain the initial impedance reference value of the membrane electrode and the imaginary part of the electrochemical impedance at a preset frequency, and obtain the impedance ratio term based on the impedance ratio of the imaginary part of the electrochemical impedance to the initial impedance reference value;

[0035] The air-hydrogen pressure difference health index, used to quantify the health status of the membrane electrode, is obtained by weighting the relative deviation of the pressure difference, the rate of temperature change, and the impedance ratio.

[0036] If the air-hydrogen pressure difference health index is within the preset health index warning range, then the predictive control model is invoked.

[0037] An intelligent dynamic air-hydrogen differential pressure control system, used to implement the aforementioned control method, includes:

[0038] The pressure monitoring module is used to monitor the first pressure value of the air passage and the second pressure value of the hydrogen passage in the fuel cell stack in real time.

[0039] A differential pressure controller is used to obtain the real-time air-hydrogen pressure difference between the air passage and the hydrogen passage based on the first pressure value and the second pressure value, generate a control signal based on the real-time air-hydrogen pressure difference, and dynamically adjust the system state based on the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack.

[0040] Wherein, the real-time air pressure difference is equal to the second pressure value minus the first pressure value.

[0041] Preferably, the control system further includes:

[0042] The interval division module is used to divide different air-hydrogen pressure difference intervals according to the pressure difference level between the hydrogen passage and the air passage;

[0043] The strategy preset module is used to preset different system control strategies for different air-hydrogen pressure difference ranges.

[0044] Preferably, the control system further includes:

[0045] The differential pressure tracking term acquisition module is used to acquire the predicted air-hydrogen differential pressure value for the next N steps based on the LSTM network, acquire the target reference differential pressure value, and acquire the air-hydrogen differential pressure tracking term to represent minimizing the differential pressure prediction error based on the predicted air-hydrogen differential pressure value and the target reference differential pressure value.

[0046] The valve opening measurement penalty term acquisition module is used to acquire the valve opening measurement at the k-th step based on the LSTM network, and to acquire the valve opening measurement penalty term used to control the valve action amplitude based on the valve opening measurement.

[0047] The risk penalty item acquisition module is used to acquire risk penalty items based on the real-time air-hydrogen pressure difference, which indicate that the penalty value increases exponentially when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone.

[0048] The predictive control model construction module is used to obtain an optimization function based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and to construct a predictive control model based on the optimization function.

[0049] The valve opening sequence acquisition module is used to acquire the optimized valve opening sequence according to the predictive control model, and control the valve opening based on the valve opening sequence. Attached Figure Description

[0050] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0051] Figure 1 This is a schematic diagram of the overall process of an intelligent dynamic air-hydrogen pressure difference control method in one embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of the overall process of an intelligent dynamic air-hydrogen pressure difference control method according to another embodiment of the present invention. Figure 1 ;

[0053] Figure 3 This is a flowchart illustrating a specific method for dividing different air-hydrogen pressure difference ranges in one embodiment of the present invention;

[0054] Figure 4 This is a schematic diagram of the control logic flow of an intelligent dynamic air-hydrogen pressure difference control method according to an embodiment of the present invention;

[0055] Figure 5 This is a schematic diagram of the overall process of an intelligent dynamic air-hydrogen pressure difference control method according to another embodiment of the present invention. Figure 2 ;

[0056] Figure 6 This is a schematic diagram of the overall process of an intelligent dynamic air-hydrogen pressure difference control method according to another embodiment of the present invention. Figure 3 ;

[0057] Figure 7 This is a schematic diagram of the overall structure of an intelligent dynamic air-hydrogen differential pressure control system according to an embodiment of the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to its embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of the invention.

[0059] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0060] Unless otherwise defined, 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. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0061] In this invention, "first" and "second" do not represent a specific quantity or order, but are merely used to distinguish names.

[0062] like Figure 1 As shown, an embodiment of the present invention provides an intelligent dynamic air-hydrogen pressure difference control method, comprising the following steps:

[0063] S1 monitors the first pressure value of the air passage and the second pressure value of the hydrogen passage in the fuel cell stack in real time.

[0064] S2, obtain the real-time air-hydrogen pressure difference between the air passage and the hydrogen passage based on the first pressure value and the second pressure value, and generate a control signal based on the real-time air-hydrogen pressure difference.

[0065] S3, dynamically adjust the system state according to the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack. The real-time air pressure difference is equal to the second pressure value minus the first pressure value.

[0066] Figure 4 This demonstration showcases the control logic flow of an intelligent dynamic air-hydrogen differential pressure control method, illustrating how the system dynamically responds and adjusts based on real-time differential pressure data during operation. Upon system startup, the differential pressure monitoring module first collects real-time pressure data from the air-hydrogen dual-channel system and transmits it to the differential pressure controller. The differential pressure controller analyzes the real-time air-hydrogen differential pressure information to determine if the current state is within the set safe range. If the differential pressure is within the normal range, the system maintains its current adjustment state and continues operation; if a slight anomaly occurs (the differential pressure slightly exceeds the set threshold), a small adjustment is made by fine-tuning the valve opening; and when the differential pressure enters a severely abnormal range, the controller will perform a large-scale adjustment operation to quickly restore system stability.

[0067] Once the differential pressure is adjusted to normal, the system feeds back the current status and continues to the next round of data acquisition and judgment, thus forming a continuous closed-loop control process. Through this control process, the system can achieve dynamic balance adjustment of the air-hydrogen differential pressure under various operating conditions, effectively protecting the fuel cell membrane electrode assembly from differential pressure shocks and improving system response speed and operational stability.

[0068] In summary, the intelligent dynamic air-hydrogen differential pressure control method generates a control signal by acquiring the real-time air-hydrogen differential pressure, and performs real-time dynamic regulation on the pressure difference of the air-hydrogen dual flow path during the operation of the fuel cell system. This ensures that the real-time air-hydrogen differential pressure is always maintained within a safe range during the operation of the fuel cell stack, preventing damage to the membrane module due to pressure fluctuations, thereby improving the stability and service life of the fuel cell system.

[0069] As a preferred technical solution, such as Figure 2 As shown, the control method further includes the following steps:

[0070] S4. Based on the pressure difference level between the hydrogen passage and the air passage, different air-hydrogen pressure difference ranges are defined.

[0071] For example, such as Figure 3 As shown, the specific method for dividing different air-hydrogen pressure difference ranges includes the following steps:

[0072] S41, when the differential pressure level is less than 0 kPa, it is classified as a danger zone. A danger zone means the differential pressure reverses, requiring the system to shut down immediately or undergo forced intervention to protect the membrane assembly.

[0073] S42, when the pressure difference level falls within the range of [0 kPa, 2 kPa), it is classified as a slightly abnormal zone. In this case, the system is in a critical state and the flow rate on the air or hydrogen side needs to be fine-tuned.

[0074] S43, when the differential pressure level is within the range of [2kPa, 10kPa), it is classified as the normal operating range. At this time, the differential pressure level is within the ideal differential pressure range, and the system can maintain its current operating state and maintain its current operating parameters.

[0075] S44, when the pressure difference level is within [10kPa, 15kPa], it is classified as a high pressure difference warning zone. At this time, the pressure difference level is too large, and the system needs to be adjusted in time to prevent membrane drying or stacking resistance.

[0076] S45, when the differential pressure level is greater than 15 kPa, it is classified as a dangerous high-pressure zone. In this case, the system should be quickly adjusted or the flow should be limited to prevent performance degradation or membrane damage.

[0077] For example, the specific method for generating a control signal based on the real-time air-hydrogen pressure difference includes: if the real-time air-hydrogen pressure difference is in a slightly abnormal zone, generating a control signal to adjust the opening of the electronic regulating valve A of the air passage or the electronic regulating valve B of the hydrogen passage; if the real-time air pressure difference is in a normal operating zone, generating a control signal to maintain the current operating state.

[0078] S5. For different air-hydrogen pressure difference ranges, different system control strategies are preset.

[0079] By dividing the air-hydrogen pressure difference into different ranges and determining the range to which the current real-time air-hydrogen pressure difference belongs, the electronic valves can be adjusted in real time to ensure that the air-hydrogen pressure difference of the fuel cell stack is always maintained within a safe range during operation.

[0080] As a preferred technical solution, such as Figure 5 As shown, the control method further includes the following steps:

[0081] S6. Based on the LSTM network, obtain the predicted air-hydrogen pressure difference value for the next N steps, obtain the target reference pressure difference value, and obtain the air-hydrogen pressure difference tracking term to represent minimizing the pressure difference prediction error according to the predicted air-hydrogen pressure difference value and the target reference pressure difference value.

[0082] For predicting the air-hydrogen pressure difference over the next N steps, an LSTM network can be constructed first. Historical pressure, temperature, and current data are then used as a training set and input into the LSTM network for training to obtain the predicted air-hydrogen pressure difference. The LSTM network weights can be updated online every 2 hours to adapt to membrane aging. The target reference pressure difference is generally set to a range of 2-10 kPa (ideal range), which can be understood as the safe pressure difference range with minimal mechanical stress on the membrane electrode. N represents the prediction time domain, primarily used to predict the system state over the next N control cycles.

[0083] S7. Obtain the valve opening measure at step k based on the LSTM network, and obtain the valve opening measure penalty term for controlling the valve action amplitude based on the valve opening measure.

[0084] The valve opening penalty term is used to suppress sudden changes in valve opening and smooth valve operation.

[0085] S8, based on the real-time air-hydrogen pressure difference, obtain a risk penalty item to indicate that when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone, the penalty value increases exponentially.

[0086] S9. Obtain an optimization function based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and construct a predictive control model based on the optimization function.

[0087] S10, obtain the optimized valve opening sequence according to the predictive control model, and control the valve opening based on the valve opening sequence.

[0088] For example, the optimization function is expressed as Wherein, ΔP pred (k), ΔP ref Let Q and R represent the predicted air-hydrogen pressure difference and the target reference pressure difference, respectively; u represents the valve opening sequence of the electronic control valve; Q and R represent the weight matrix of the output tracking error and the valve opening measurement weight matrix, respectively; Δu(k) represents the valve opening measurement at the k-th step; ΔP represents the real-time air-hydrogen pressure difference; ρ represents the risk penalty coefficient; and Risk = e ΔP-15 Indicates risk penalty items, These represent the air-hydrogen pressure difference tracking term and the valve opening penalty term, respectively. All are weighted L2 norms.

[0089] Both the output tracking error weight matrix and the valve opening weight matrix are dimensionless. The output tracking error weight matrix is ​​used to amplify / reduce the penalty for differential pressure tracking deviation (the larger the value, the higher the tracking accuracy requirement). The valve opening weight matrix is ​​used to smooth valve action (the larger the value, the smoother the valve action). The valve opening change at step k is used to suppress frequent valve action and ensure mechanical life. The risk penalty coefficient is greater than zero and is used to balance the weights of conventional control and risk response.

[0090] For the risk penalty term, its value approaches 1 when the real-time air-hydrogen pressure difference is close to 15 kPa. When the real-time air-hydrogen pressure difference is 20 kPa, the risk penalty term surges to e. 5 The value is approximately 148.4, indicating an exponential amplification of the optimization function's tendency to avoid this state. Therefore, this risk penalty term forces the controller to prioritize significant valve reduction when the differential pressure approaches the danger zone, sacrificing some tracking accuracy to ensure membrane safety. The risk penalty coefficient can be switched according to the risk level. For example, when in the danger zone, the risk penalty coefficient is set to 10 to achieve exponential penalty and trigger emergency pressure reduction; when in the warning zone, the risk penalty coefficient is set to 3 to increase the adjustment force; and when in the normal operating zone, the risk penalty coefficient is set to 0.1 to prioritize energy saving and valve lifespan. Thus, based on this risk penalty term, risk-sensitive control can be achieved: prioritizing economy (small ρ) in normal operating conditions and ensuring safety (large ρ) in dangerous operating conditions. Furthermore, the exponential risk penalty is naturally robust to disturbances such as sudden load changes and sensor failures, preventing over-adjustment of the controller.

[0091] The optimization function can be solved in real time using ADMM (Alternating Direction Method of Multipliers) to obtain the valve opening sequence of the electronic control valve. The electronic control valves mentioned here include, but are not limited to, electronic control valve A in the air passage and electronic control valve B in the hydrogen passage. ADMM is an optimization algorithm combining the dual ascent method and the multiplier method, mainly used to solve constrained convex optimization problems. It generally includes steps such as variable partitioning, alternating direction optimization, and residual convergence judgment. Since it is a conventional technique in this field, it will not be elaborated further here.

[0092] In summary, by obtaining the optimization function and constructing a predictive control model, the valve opening sequence of electronic regulation can be obtained, achieving a triple balance of safety, accuracy, and lifespan, and ensuring the reliability of the membrane electrode under dynamic operating conditions. Specifically, this includes: 1. minimizing the deviation between the predicted air-hydrogen pressure difference and the target value; 2. suppressing sudden changes in valve opening; 3. suppressing the probability of the pressure difference entering the dangerous high-pressure zone.

[0093] As a preferred technical solution, such as Figure 6 As shown, the control method further includes the following steps:

[0094] S11, obtain the pressure difference ratio between the absolute value of the real-time air-hydrogen pressure difference and the target reference pressure difference value, and obtain the pressure difference relative deviation term based on the pressure difference ratio.

[0095] Specifically, the relative deviation of the pressure difference is used to reflect the current pressure balance state, and is expressed as follows: When the relative deviation of the pressure difference is greater than 1.2, it can be considered that there is a risk of excessive membrane stress; when the relative deviation of the pressure difference is less than 0.8, it can be considered that there is a risk of flow channel blockage. Here, the ratio of the absolute value of the real-time air-hydrogen pressure difference to the real-time air-hydrogen pressure difference is used instead of the real-time air-hydrogen pressure difference to eliminate the influence of dimensions.

[0096] S12, obtain the real-time temperature value T of the fuel cell stack, and obtain the temperature change rate term based on the real-time temperature value.

[0097] Real-time temperature values ​​can be obtained via an embedded thermocouple array (0.1℃ resolution). The temperature change rate term is used to detect thermal management anomalies and can be represented as follows: When the temperature change rate is greater than 0.5℃ / s, it can be considered that there is a coolant pump jamming fault; when the temperature change rate is less than -1.2℃ / s, it can be considered that there is a membrane flooding leading to local low temperature fault.

[0098] S13, obtain the initial impedance reference value Z0 of the membrane electrode and the imaginary part of the electrochemical impedance at a preset frequency, and obtain the impedance ratio term based on the impedance ratio of the imaginary part of the electrochemical impedance to the initial impedance reference value.

[0099] The imaginary part of the electrochemical impedance can be obtained from an online EIS analyzer. The initial impedance reference value is the calibration value under rated operating conditions at 25℃, typically the factory calibration value stored in FLASH. The impedance ratio term is used to monitor the hydration state of the membrane electrode and can be expressed as follows: Among them, |Z”| 1kHz This represents the imaginary part of the electrochemical impedance at a frequency of 1 kHz.

[0100] For the impedance ratio, if it is less than 0.9, the membrane electrode exhibits flooding. In this case, liquid water blocks the gas diffusion layer, submerges the proton conduction channel, and causes liquid water to accumulate in the cathode channel, hindering gas diffusion and making the cathode prone to oxygen transport obstacles. A typical symptom is voltage oscillation. When the impedance ratio is between 0.9 and 1.1, the membrane electrode is in an ideal hydrated state, with minimal proton conduction resistance. When the impedance ratio is greater than 1.1, the membrane electrode is in a dried-out state. Membrane dehydration leads to a decrease in the proton dissociation ability of sulfonic acid groups, and the membrane conductivity may drop sharply.

[0101] S14. Obtain the air-hydrogen pressure difference health index for quantifying the health status of the membrane electrode based on the weighted values ​​of the pressure difference relative deviation term, temperature change rate term, and impedance ratio term.

[0102] For example, the air-hydrogen pressure differential health index Wherein, α, β, and γ are the weighting coefficients for the pressure difference relative deviation term, temperature change rate term, and impedance ratio term, respectively, with default values ​​of α = 0.6, β = 0.2, and γ = 0.2. The weighting coefficient for the impedance ratio term can be dynamically adjusted based on the impedance ratio itself. When the impedance ratio is less than 0.9, the weighting coefficient is set to 0.5; when the impedance ratio is greater than 1.1, the weighting coefficient is set to 0.7. In this way, the air-hydrogen pressure difference health index can be linked with the control strategy to provide early warnings during flooded conditions and trigger membrane protection mode during dry conditions, reducing lifespan loss.

[0103] The innovation of the air-hydrogen pressure difference health index lies in its integration of physical parameters (pressure difference), thermal parameters (temperature change), and electrochemical parameters (impedance) into a single index. This index incorporates multi-dimensional sensor data and can be used to quantify the health status of the membrane electrode. Specifically, the relative pressure difference deviation term (α-dominant) ensures the integrity of the mechanical structure, the temperature change rate term (β-dominant) can prevent thermal runaway chain reactions, and the impedance ratio term (γ-dominant) can maintain optimal membrane proton conduction.

[0104] S15, if the air-hydrogen pressure difference health index is within the preset health index warning range, then the predictive control model is invoked.

[0105] For example, the health index warning range is set to 5.1-8.0, at which point the predictive control model can be activated for feedforward compensation. When the air-hydrogen pressure difference health index is greater than 8.0, nitrogen can be injected in an emergency to address the sudden temperature rise caused by coolant pump failure.

[0106] Therefore, this air-hydrogen pressure difference health index can not only be used for real-time risk assessment, but also provide a reference for switching control strategies, providing core decision support for highly reliable fuel cell systems.

[0107] An embodiment of the present invention also provides an intelligent dynamic air-hydrogen differential pressure control system for implementing the aforementioned control method, which includes a pressure detection module and a differential pressure controller.

[0108] The pressure monitoring module is used to monitor the first pressure value of the air passage and the second pressure value of the hydrogen passage in the fuel cell stack in real time; the differential pressure controller is used to obtain the real-time air-hydrogen pressure difference between the air passage and the hydrogen passage based on the first pressure value and the second pressure value, generate a control signal based on the real-time air-hydrogen pressure difference, and dynamically adjust the system state based on the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack; wherein, the real-time air pressure difference is equal to the second pressure value minus the first pressure value.

[0109] After the system starts up, the differential pressure monitoring module continuously collects pressure data and feeds it back to the controller. The controller analyzes the data in real time and determines the current state based on the set differential pressure range: if it is within the normal range, it continues to maintain that state; if there is a slight abnormality, it quickly makes a fine adjustment; if there is a severe abnormality, it immediately makes a large adjustment. In other words, the controller generates corresponding control commands based on the real-time air-hydrogen pressure difference, driving the electronic regulating valve to perform adjustment actions to maintain a stable differential pressure environment inside the fuel cell system.

[0110] Specifically, Figure 7 This invention demonstrates the overall structure and key control paths of the intelligent dynamic air-hydrogen differential pressure control system proposed in this invention. The system comprises air flow path, hydrogen flow path, fuel cell stack, differential pressure monitoring module, electronic regulating valve, and differential pressure controller, forming a complete closed-loop differential pressure regulation structure. Air enters the system through an air filter, is pressurized by an air compressor, and then enters the air side of the fuel cell stack through electronic regulating valve A. Hydrogen is supplied from a hydrogen cylinder and flows into the hydrogen side of the fuel cell stack through a hydrogen circulation pump and electronic regulating valve B. A "differential pressure balance control area" is established between the air and hydrogen paths, which is the differential pressure sensitive area of ​​the membrane electrode assembly (MEA) and the core protected object of the system's regulation and control. The system includes a differential pressure monitoring module that collects pressure signals from both the air and hydrogen sides and sends the data to the differential pressure controller in real time. The controller determines whether the current state is within a safe range based on the differential pressure value and outputs corresponding control signals to electronic regulating valves A and B to dynamically adjust the gas flow and pressure, ensuring that the pressure on both sides of the membrane module remains within a safe balance range and preventing membrane damage caused by differential pressure reversal. The system is also equipped with a silencer to suppress gas flow noise and a cooling circuit to maintain the thermal balance of the fuel cell stack. The entire system achieves rapid response and dynamic adjustment to pressure difference through closed-loop control, effectively improving the safety and stability of the fuel cell system operation.

[0111] As a preferred technical solution, the control system further includes an interval division module and a strategy preset module.

[0112] The interval division module is used to divide different air-hydrogen pressure difference intervals according to the pressure difference level between the hydrogen passage and the air passage; the strategy preset module is used to preset different system control strategies for different air-hydrogen pressure difference intervals.

[0113] For example, when the differential pressure level is less than 0 kPa, it is classified as a danger zone. In this danger zone, the differential pressure reverses, causing the membrane electrode to deform in the opposite direction, resulting in permanent damage. In this case, the system needs to be shut down immediately or forcibly intervened to protect the membrane assembly. When the differential pressure level is between 0 kPa and 2 kPa, it is classified as a slight abnormality zone. In this case, the system is in a critical state with a certain risk of flooding, requiring fine-tuning of the air or hydrogen flow rate. When the differential pressure level is between 2 kPa and 10 kPa, it is classified as a normal operating zone. At this time, the differential pressure level is within the ideal range, and the system can maintain its current operating state and parameters. When the differential pressure level is between 10 kPa and 15 kPa, it is classified as a high differential pressure warning zone. At this time, the differential pressure level is too high, which can easily lead to membrane drying / microcracks. The system needs to be adjusted promptly to prevent membrane drying or stacking resistance. When the differential pressure level is greater than 15 kPa, it is classified as a dangerous high-pressure zone. In this case, the system should be quickly adjusted or flow-limited to prevent performance degradation or membrane damage.

[0114] As a preferred technical solution, the control system further includes a differential pressure tracking term acquisition module, an opening penalty term acquisition module, a risk penalty term acquisition module, a predictive control model construction module, and a valve opening sequence acquisition module.

[0115] The differential pressure tracking term acquisition module is used to obtain the predicted air-hydrogen differential pressure value for the next N steps based on the LSTM network, obtain the target reference differential pressure value, and obtain the air-hydrogen differential pressure tracking term to represent minimizing the differential pressure prediction error based on the predicted air-hydrogen differential pressure value and the target reference differential pressure value.

[0116] For example, the air-hydrogen pressure differential tracking term is represented as

[0117] The valve opening penalty term acquisition module is used to acquire the valve opening metric at the k-th step based on the LSTM network, and to acquire a valve opening penalty term for controlling the valve's actuation amplitude based on the valve opening metric. For example, the valve opening penalty term is represented as...

[0118] The risk penalty item acquisition module is used to acquire a risk penalty item based on the real-time air-hydrogen pressure difference, indicating that the penalty value increases exponentially when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone. For example, the risk penalty item is represented as Risk = e ΔP-15 .

[0119] The predictive control model building module is used to obtain an optimization function based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and to build a predictive control model based on the optimization function.

[0120] For example, the optimization function is represented as Wherein, ΔPpred (k), ΔP ref Let represent the predicted air-hydrogen pressure difference and the target reference pressure difference, respectively; u represent the valve opening sequence of the electronic control valve; Q and R represent the weight matrix of the output tracking error and the valve opening measurement weight matrix, respectively; Δu(k) represents the valve opening measurement at the k-th step; ΔP represents the real-time air-hydrogen pressure difference; and ρ represents the risk penalty coefficient. All are weighted L2 norms.

[0121] The valve opening sequence acquisition module is used to acquire the optimized valve opening sequence according to the predictive control model, and control the valve opening based on the valve opening sequence.

[0122] Both the output tracking error weight matrix and the valve opening weight matrix are dimensionless. The output tracking error weight matrix is ​​used to amplify / reduce the penalty for differential pressure tracking deviation (the larger the value, the higher the tracking accuracy requirement). The valve opening weight matrix is ​​used to smooth valve action (the larger the value, the smoother the valve action). The valve opening change at step k is used to suppress frequent valve action and ensure mechanical life. The risk penalty coefficient is greater than zero and is used to balance the weights of conventional control and risk response.

[0123] The optimization function can be solved in real time using the ADMM algorithm to obtain the valve opening sequence of the electronic control valves. The electronic control valves mentioned here include, but are not limited to, electronic control valve A in the air passage and electronic control valve B in the hydrogen passage. ADMM is an optimization algorithm combining the dual ascent method and the multiplier method, mainly used to solve constrained convex optimization problems. It generally includes steps such as variable partitioning, alternating direction optimization, and residual convergence judgment. Since it is a conventional technique in this field, it will not be elaborated further here.

[0124] By obtaining the optimization function and constructing a predictive control model, the valve opening sequence of electronic regulation can be obtained, achieving a triple balance of safety, accuracy, and lifespan, and ensuring the reliability of the membrane electrode under dynamic operating conditions. Specifically, this includes: 1. minimizing the deviation between the predicted air-hydrogen pressure difference and the target value; 2. suppressing sudden changes in valve opening; 3. suppressing the probability of the pressure difference entering the dangerous high-pressure zone.

[0125] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0126] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A smart dynamic air-hydrogen pressure differential control method, characterized in that, The control method includes the following steps: Real-time monitoring of the first pressure value of the air passage and the second pressure value of the hydrogen passage within the fuel cell stack; The real-time air-hydrogen pressure difference between the air passage and the hydrogen passage is obtained based on the first pressure value and the second pressure value, and a control signal is generated based on the real-time air-hydrogen pressure difference. The system state is dynamically adjusted according to the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack. Wherein, the real-time air pressure difference is equal to the second pressure value minus the first pressure value.

2. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 1, characterized in that, The control method further includes the following steps: Based on the pressure difference between the hydrogen passage and the air passage, different air-hydrogen pressure difference ranges are defined; Different system control strategies are preset for different air-hydrogen pressure difference ranges.

3. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 2, characterized in that, The specific methods for dividing different air-hydrogen pressure difference ranges include the following steps: When the pressure difference level is less than 0 kPa, it is designated as a danger zone; When the pressure difference level falls within the range of [0 kPa, 2 kPa), it is classified as a slightly abnormal zone. When the pressure difference level is within [2kPa, 10kPa), it is classified as the normal working zone; When the pressure difference level is within [10kPa, 15kPa], it is classified as a high pressure difference warning zone; When the pressure difference level is greater than 15 kPa, it is classified as a dangerous high-pressure zone.

4. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 3, characterized in that, The specific method for generating a control signal based on the real-time air-hydrogen pressure difference includes the following steps: If the real-time air-hydrogen pressure difference is in a slightly abnormal zone, a control signal is generated to adjust the opening of the electronic regulating valve A in the air passage or the electronic regulating valve B in the hydrogen passage. If the real-time air pressure difference falls within the normal operating range, a control signal is generated to maintain the current operating state.

5. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 4, characterized in that, The control method further includes the following steps: Based on the LSTM network, the predicted air-hydrogen pressure difference value for the next N steps is obtained, the target reference pressure difference value is obtained, and the air-hydrogen pressure difference tracking term is obtained to represent minimizing the pressure difference prediction error. The valve opening measure at step k is obtained based on the LSTM network, and a valve opening measure penalty term for controlling the valve action amplitude is obtained based on the valve opening measure. Based on the real-time air-hydrogen pressure difference, a risk penalty item is obtained to indicate that the penalty value increases exponentially when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone. An optimization function is obtained based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and a predictive control model is constructed based on the optimization function. The optimized valve opening sequence is obtained based on the predictive control model, and the valve opening is controlled based on the valve opening sequence.

6. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 5, characterized in that, The optimization function is expressed as follows: Wherein, ΔP pred (k), ΔP ref Let Q and R represent the predicted air-hydrogen pressure difference and the target reference pressure difference, respectively; u represents the valve opening sequence of the electronic control valve; Q and R represent the weight matrix of the output tracking error and the valve opening measurement weight matrix, respectively; Δu(k) represents the valve opening measurement at the k-th step; ρ represents the risk penalty coefficient; and Risk = e ΔP-15 Indicates risk penalty items, These represent the air-hydrogen pressure difference tracking term and the valve opening penalty term, respectively.

7. The intelligent dynamic air-hydrogen pressure difference control method as described in claim 6, characterized in that, The control method further includes the following steps: Obtain the pressure difference ratio between the absolute value of the real-time air-hydrogen pressure difference and the target reference pressure difference value, and obtain the pressure difference relative deviation term based on the pressure difference ratio; Obtain the real-time temperature value of the fuel cell stack, and obtain the temperature change rate term based on the real-time temperature value; Obtain the initial impedance reference value of the membrane electrode and the imaginary part of the electrochemical impedance at a preset frequency, and obtain the impedance ratio term based on the impedance ratio of the imaginary part of the electrochemical impedance to the initial impedance reference value; The air-hydrogen pressure difference health index, used to quantify the health status of the membrane electrode, is obtained by weighting the relative deviation of the pressure difference, the rate of temperature change, and the impedance ratio. If the air-hydrogen pressure difference health index is within the preset health index warning range, then the predictive control model is invoked.

8. An intelligent dynamic air-hydrogen differential pressure control system, used to implement the control method as described in any one of claims 1-7, characterized in that, The control system includes: The pressure monitoring module is used to monitor the first pressure value of the air passage and the second pressure value of the hydrogen passage in the fuel cell stack in real time. A differential pressure controller is used to obtain the real-time air-hydrogen pressure difference between the air passage and the hydrogen passage based on the first pressure value and the second pressure value, generate a control signal based on the real-time air-hydrogen pressure difference, and dynamically adjust the system state based on the control signal to ensure that the real-time air-hydrogen pressure difference is always maintained within a safe range during the operation of the fuel cell stack. Wherein, the real-time air pressure difference is equal to the second pressure value minus the first pressure value.

9. The intelligent dynamic air-hydrogen differential pressure control system as described in claim 8, characterized in that, The control system further includes: The interval division module is used to divide different air-hydrogen pressure difference intervals according to the pressure difference level between the hydrogen passage and the air passage; The strategy preset module is used to preset different system control strategies for different air-hydrogen pressure difference ranges.

10. The intelligent dynamic air-hydrogen differential pressure control system as described in claim 9, characterized in that, The control system further includes: The differential pressure tracking term acquisition module is used to acquire the predicted air-hydrogen differential pressure value for the next N steps based on the LSTM network, acquire the target reference differential pressure value, and acquire the air-hydrogen differential pressure tracking term to represent minimizing the differential pressure prediction error based on the predicted air-hydrogen differential pressure value and the target reference differential pressure value. The valve opening measurement penalty term acquisition module is used to acquire the valve opening measurement at the k-th step based on the LSTM network, and to acquire the valve opening measurement penalty term used to control the valve action amplitude based on the valve opening measurement. The risk penalty item acquisition module is used to acquire risk penalty items based on the real-time air-hydrogen pressure difference, which indicate that the penalty value increases exponentially when the real-time air-hydrogen pressure difference is greater than the lower limit of the high-pressure danger zone. The predictive control model construction module is used to obtain an optimization function based on the air-hydrogen pressure difference tracking term, valve opening penalty term, and risk penalty term, and to construct a predictive control model based on the optimization function. The valve opening sequence acquisition module is used to acquire the optimized valve opening sequence according to the predictive control model, and control the valve opening based on the valve opening sequence.

Images