Solid oxide stack system model prediction control method, storage medium and equipment

Abstract

The invention provides a solid oxide stack system model prediction control method, a storage medium and equipment, and belongs to the field of fuel cells, and the method comprises the following steps: constructing a flat plate type cross-flow solid oxide stack model, and modeling the stack through a double-temperature layer design; performing thermal-mass-electric coupling simulation on the solid oxide stack model, and simulating spatial distribution characteristics of a temperature field, a component concentration field and current density in the solid oxide stack; and designing a model prediction controller based on the solid oxide stack model, wherein the model prediction controller realizes accurate regulation and control on the output power of the solid oxide stack. A high-precision solid oxide galvanic pile dynamic model is established on the basis of a nodal idea, and is different from other galvanic pile models for control strategy design, the external electrochemical characteristics of the solid oxide galvanic pile are researched, the internal temperature field and the internal flow field are monitored at the same time, and the solid oxide galvanic pile is precisely regulated and controlled by designing a model prediction control strategy.

Description

Technical Field

[0001] This invention belongs to the field of fuel cells, and specifically relates to a model predictive control method, storage medium and device for a solid oxide fuel cell stack system. Background Technology

[0002] Solid oxide fuel cells (SOFCs) are considered a highly promising alternative to traditional power generation technologies due to their high efficiency, low emissions, and broad fuel adaptability. However, in practical applications, SOFCs require long-term, rapid, and frequent load changes. During operation, the internal flow field, component transport, mass transport, charge transport, electrochemistry, chemical reactions, and heat transport of the fuel cell stack are highly coupled. The thermo-mass-electrochemical multi-physics fields within the stack exhibit significant non-uniformity in both time and space, making the stack prone to harsh conditions such as localized overheating and excessive temperature gradients. This makes it difficult for traditional control strategies to achieve precise control of the fuel cell stack.

[0003] Patent application CN201611156775.3 discloses a method that assumes each individual cell has identical dynamic characteristics, models only the individual cells, and then simply combines them to form a stack model. This method employs one-dimensional node-based modeling, considering only the temperature distribution along the gas flow direction and ignoring the temperature gradient perpendicular to the airflow direction. Patent application CN202411519071.2 discloses a method that uses a distributed parameter method to divide the stack into several nodes along the direction of the reactant gas flow, establishing a one-dimensional dynamic model of a solid oxide fuel cell. Based on this, a Kalman filter-based model predictive controller KF-M is developed. PC establishes a joint simulation platform; the invention patent with application number CN202211570590.2 discloses that according to the direction of the reaction airflow, each system component is divided into a finite number of nodes in a gradual division method. The parameters inside the same node are the same and are equivalent to the node outlet parameters. That is, the characteristic value of the node outlet parameter represents the characteristic value of the parameters of the entire spatial calculation sub-model. The outlet parameter of the previous node is equal to the inlet parameter of the next node. However, since the solid oxide fuel cell model used in the design control strategy is a simple zero-dimensional or one-dimensional model, it only considers the temperature distribution in a single direction, resulting in low spatial resolution and insufficient prediction accuracy.

[0004] Existing full-scale solid oxide fuel cell (SOFC) simulation studies are mostly limited to simple thermo-mass coupling or calculations of only electrochemical fields, making it difficult to realistically reflect the complex multi-field coupling and transport processes within the fuel cell. To reduce computational load, most studies have significantly simplified SOFC models, resulting in mass, temperature, and electrochemical field distributions calculated using such methods that differ greatly from real-world conditions. Consequently, control strategies designed based on these models lack sufficient persuasiveness. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a model predictive control method, storage medium, and device for solid oxide fuel cell systems. Based on a node-based approach, a more accurate dynamic model of the solid oxide fuel cell is established. Unlike other fuel cell models used for control strategy design, this invention monitors the internal temperature and flow fields while studying its external electrochemical characteristics, and precisely regulates them through a model predictive control strategy.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] On one hand, the present invention provides a model predictive control method for a solid oxide fuel cell stack system, comprising: constructing a planar cross-flow solid oxide fuel cell stack model, dividing the stack model into n×n equal-area nodes, wherein each node represents a different physical layer of the stack, and modeling the stack through a dual-temperature-layer design; performing thermo-mass-electric coupling simulation on the solid oxide fuel cell stack model to simulate the spatial distribution characteristics of the temperature field, component concentration field, and current density inside the solid oxide fuel cell stack; and designing a model predictive controller based on the solid oxide fuel cell stack model, wherein the model predictive controller adjusts the fuel flow rate and the stack input voltage to achieve precise control of the output power of the solid oxide fuel cell stack.

[0008] Furthermore, the dual-temperature layer design includes treating the anode layer, electrolyte layer, cathode layer, and connector layer as one temperature layer, and the air layer as an independent temperature layer.

[0009] Furthermore, the physical layer includes an air layer, an anode layer, a cathode layer, an electrolyte layer, and a connector layer. The accuracy of the model is ensured by physical field coupling calculations between nodes, and the model is divided into 5×5 nodes.

[0010] Furthermore, the dual-temperature layer design includes: the anode layer, electrolyte layer, cathode layer, and connector layer are modeled as a common temperature layer, and the heat transfer of this temperature layer is coupled with the electrochemical reaction calculation; the air layer is an independent temperature layer, modeled separately from the anode layer, electrolyte layer, and connector layer.

[0011] Furthermore, the model predictive controller optimizes fuel flow and stack input voltage by predicting the output power of the solid oxide fuel cell stack in real time and adjusting the power based on the predicted output power.

[0012] Furthermore, the model predictive controller is implemented in the following ways: the dynamic behavior of the solid oxide fuel cell stack is modeled in state space, and the state vector includes fuel flow rate, pressure, temperature and current density of each gas; based on the prediction of the model predictive controller, the state at least one time step in the future is predicted, and the optimal control input is calculated to minimize the control error; when solving the optimization problem, the output power error is minimized, and the stability of the system and the minimization of control energy consumption are ensured by adjusting the weight of the control input.

[0013] Furthermore, the model predictive controller employs an adaptive algorithm to estimate key parameters of the solid oxide fuel cell stack in real time. These key parameters include the flow coefficients of hydrogen, oxygen, and water, electrochemical reactions, and the temperature of the stack.

[0014] Furthermore, the estimation of the key parameters is achieved through the following steps: the system model is decomposed using a dynamic key parameter estimation method, with some key parameters being affected by parameter changes and others being related to the control input; within each sampling period, the estimated key parameters in the solid oxide fuel cell are updated.

[0015] The present invention provides a storage medium storing a computer program, which, when executed, implements the above-described predictive control method.

[0016] The present invention provides an electronic device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the predictive control method described above is performed.

[0017] The beneficial effects of the technical solutions provided by the embodiments of the present invention include:

[0018] This invention overcomes the limitations of traditional zero-dimensional and one-dimensional models in dynamic response analysis and thermal management by establishing a node-based fuel cell stack model to simulate the spatial distribution characteristics of the internal temperature field, component concentration field, and current density. For each node, a multiphysics coupling model was developed, including mass conservation equations, energy conservation equations, and electrochemical equations. Based on this model, a model predictive control strategy was designed to achieve precise regulation of output power under safety constraints of the solid oxide fuel cell stack. Secondly, it overcomes the problem of poor prediction and tracking capabilities in existing technologies. Existing technologies can improve computational accuracy and time by increasing the computing power of the equipment, but the construction and operation costs of large servers are high. Therefore, this invention uses a specially simplified model based on existing computer equipment, which can shorten the computation time and significantly improve the accuracy and tracking performance of predictions. Under the condition that the tracking error is less than 0.8%, the tracking performance is less than 50ms. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This embodiment of the invention provides the temperature change of different nodes under varying outlet temperatures.

[0021] Figure 2 This is a schematic diagram of a 5*5 node with equal area provided in an embodiment of the present invention;

[0022] Figure 3 The power following results under different algorithms provided in Embodiment 1 of the present invention. Detailed Implementation

[0023] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] This invention discloses a model predictive control method for a solid oxide fuel cell (SOC) stack system, comprising: constructing a planar cross-flow SOC stack model, dividing the stack model into n×n equal-area nodes, where each node represents a different physical layer of the stack, and modeling the stack through a dual-temperature-layer design; performing thermo-mass-electric coupling simulation on the SOC stack model to simulate the spatial distribution characteristics of the temperature field, component concentration field, and current density inside the SOC stack; and designing a model predictive controller based on the SOC stack model, wherein the model predictive controller adjusts the fuel flow rate and the stack input voltage to achieve precise control of the SOC stack output power.

[0026] This invention overcomes the limitations of traditional zero-dimensional and one-dimensional models in dynamic response analysis and thermal management by establishing a node-based fuel cell stack model to simulate the spatial distribution characteristics of the internal temperature field, component concentration field, and current density. For each node, a multiphysics coupling model was developed, including mass conservation equations, energy conservation equations, and electrochemical equations. Based on this model, a model predictive control strategy was designed to achieve precise regulation of output power under safety constraints of the solid oxide fuel cell stack. Secondly, it overcomes the problem of poor prediction and tracking capabilities in existing technologies. Existing technologies can improve computational accuracy and time by increasing the computing power of the equipment, but the construction and operation costs of large servers are high. Therefore, this invention uses a specially simplified model based on existing computer equipment, which can shorten the computation time and significantly improve the accuracy and tracking performance of predictions. Under the condition that the tracking error is less than 0.8%, the tracking performance is less than 50ms.

[0027] Preferably, in this embodiment of the invention, a quasi-two-dimensional planar cross-flow solid oxide fuel cell model is established on the Matlab / Simulink platform.

[0028] Specifically, considering the wide application of different types of fuel cell stacks and to facilitate the verification of the rationality of subsequent models, this invention chooses to establish a flat-plate cross-flow solid oxide fuel cell stack model with a size of 15×15cm.

[0029] The dual-temperature layer design includes treating the anode layer, electrolyte layer, cathode layer, and connector layer as one temperature layer, and the air layer as an independent temperature layer.

[0030] The physical layer includes an air layer, an anode layer, a cathode layer, an electrolyte layer, and a connector layer. The accuracy of the model is ensured by physical field coupling calculations between nodes, and the model is divided into 5×5 nodes.

[0031] Based on the concept of node-based design, the planar cross-flow solid oxide fuel cell stack is divided into n×n nodes. Research has shown that, for example... Figure 1As shown, taking n to be 5 ensures both the accuracy of the established model and greatly simplifies the modeling process. A schematic diagram of a 5×5 node is shown below. Figure 2 As shown.

[0032] The dual-temperature layer design includes: the anode layer, electrolyte layer, cathode layer, and connector layer are modeled as a common temperature layer, and the heat transfer of this temperature layer is coupled with the electrochemical reaction calculation; the air layer is an independent temperature layer, which is modeled separately from the anode layer, electrolyte layer, cathode layer, and connector layer.

[0033] Based on the thermal-mass-electrical characteristics of the fuel cell stack, a mass conservation sub-model, an energy conservation sub-model, and an electrochemical sub-model are established respectively. The preceding and following nodes are connected to form a quasi-two-dimensional planar cross-flow solid oxide fuel cell stack model as shown below:

[0034] (1) Mass conservation sub-model

[0035] Based on the mass conservation equation, this invention constructs a mass conservation model for solid oxide fuel cell stacks. For each node, the main considerations are the dynamic changes in physical quantities such as volume fraction, molar flow rate, and temperature of each component in the node.

[0036] exist In the node:

[0037] ;

[0038] ;

[0039] in, For nodes The amount of substance in air, expressed in mol; For nodes The amount of substance of the fuel, in mol; For nodes Molar flow rate of air in the atmosphere, mol / s; For nodes Molar flow rate of fuel in the medium, mol / s; For nodes The mole fraction of oxygen in the solution. For nodes The mole fraction of hydrogen in the gas. For nodes The molar rate at which oxygen is consumed in an electrochemical reaction, in mol / s; For nodes The molar rate at which hydrogen is consumed in the electrochemical reaction, in mol / s; For nodes Molar flow rate of air, mol / s; For nodes The mole fraction of oxygen; For nodes Molar flow rate of fuel in the medium, mol / s; For nodes The mole fraction of hydrogen in the gas. For time, s.

[0040] in:

[0041] ;

[0042] ;

[0043] in, The Faraday constant (C / mol) For nodes The current density in the medium, A.

[0044] node The mole fractions of water vapor and nitrogen in the anode and cathode channels are given by the following formula:

[0045] ;

[0046] ;

[0047] in, For nodes Mole fraction of internal water vapor For nodes The mole fraction of nitrogen gas inside the container.

[0048] Other items can be represented individually as:

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] in, Let m be the volume of the air channel within the node. 3 ; Let m be the volume of the fuel channel within the node. 3 ; For nodes Internal air pressure, Pa; For nodes Internal fuel pressure, Pa; For nodes Inner air layer temperature, K; For nodes Inner fuel layer temperature, K; The constant is the ideal gas constant, J / (kg•K); For nodes The molar flow rate of the fuel, in mol / s.

[0054] (2) Energy conservation model

[0055] Based on the law of conservation of energy, this invention establishes a two-temperature-layer solid oxide fuel cell model, assuming that the temperatures of the PEN material, metal connector, and anode gas layer are uniformly represented as the solid layer temperature, denoted as: The temperature of the cathode gas layer is the same as the air layer temperature, and is expressed as... .

[0056] According to the law of conservation of energy, the temperature of the air layer inside the node can be calculated using the following formula:

[0057] ;

[0058] in, Let m be the gas heat transfer surface area. 2 ; The cathode heat transfer coefficient is W / (m•K); The specific heat capacity of oxygen is J / (kg•K); Oxygen enthalpy, J / (kg•K); The specific heat capacity of nitrogen is J / (kg•K); The enthalpy of nitrogen is expressed in J / (kg•K). For nodes The average temperature inside, K;

[0059] The temperature of the solid layer within the node can be calculated using the following formula:

[0060] ;

[0061] in, Enthalpy of hydrogen, J / (kg•K); Specific enthalpy of water, J / (kg•K); Density of PEN material, kg / m³ 3 ; , where is the specific heat capacity of PEN material, J / (kg•K); The thickness of the PEN material is in meters (m). Density of the connecting material, kg / m 3 ; is the specific heat capacity of the connecting material, J / (kg•K); The thickness of the connecting material is in meters (m). is the thermal conductivity of PEN material, W / (m•K); The distance between the centers of two adjacent nodes, in meters (m). m is the area of ​​the node. 2 ;

[0062] It is temperature Enthalpy of gas:

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] (3) Electrochemical Model

[0068] In the electrochemical model of solid oxide fuel cells, the main focus is on the relationship between voltage, current density, and power. In a solid oxide fuel cell, the operating voltage of a single cell is equal to the voltage of each uniformly divided node, which can be expressed as:

[0069] ;

[0070] in, The voltage of a single cell, in V; For nodes Ohmic loss voltage, V; For nodes The activation loss voltage, V; For nodes Concentration loss voltage, V; For nodes The open-circuit voltage, V;

[0071] From the Nernst equation, we can obtain:

[0072] ;

[0073] in, The Nernst reversible potential is V; The constant is the ideal gas constant, J / (kg•K); Where C is the Faraday constant, C / mol; The temperature of the solid layer within the node, in K; For nodes Partial pressure of water vapor in Pa; For nodes Partial pressure of hydrogen gas, Pa; For nodes Partial pressure of oxygen in Pa;

[0074] The relationship between the three types of energy losses in the electrochemical reaction and the Nernst voltage is an inherent electrical characteristic of solid oxide fuel cells. The relationship between the three main losses and current density in a solid oxide single cell is expressed by the equivalent resistance as follows:

[0075] ;

[0076] in, For nodes Activation impedance, ; For nodes Ohmic impedance, ; For nodes Concentration impedance, ; For nodes The equivalent resistance; For nodes Current density in A / cm 2 .

[0077] The output voltage of a single battery is:

[0078] ;

[0079] The equivalent internal resistance is in polynomial form and exhibits a non-linear relationship with battery temperature, as shown below:

[0080] ;

[0081] ;

[0082] The output power of the solid oxide fuel cell is:

[0083] ;

[0084] in, This refers to the number of batteries.

[0085] The thermoelectric output characteristics of the quasi-two-dimensional planar cross-flow solid oxide fuel cell model were analyzed to verify the rationality of the established model.

[0086] Design model predicts controller to regulate fuel flow And the input voltage of the fuel cell stack to achieve the output power of the solid oxide fuel cell system. The control specifically includes:

[0087] (1) In solid oxide fuel cell stacks, the molar constants of reactants (such as the flow coefficients of hydrogen, oxygen, and water) are key parameters affecting system performance, and these parameters change over time. Therefore, it is necessary to estimate these parameters;

[0088] To address changes in key system parameters, this invention designs a key parameter estimation algorithm. Specifically, the system model can be decomposed into two parts: one part is affected by parameter changes, and the other part is independent of the parameters, as follows:

[0089]

[0090] in, It is the system state vector. It is a control input. These are the parameters that need to be estimated. and These are matrices related to the state and control inputs, respectively.

[0091] Preferably, in this embodiment of the invention, the key parameters affected by parameter changes are the flow coefficients of hydrogen, oxygen, and water; the key parameters related to the control input are the electrochemical reaction and the temperature of the fuel cell stack.

[0092] S73: To estimate the key parameters in the solid oxide fuel cell model, the following dynamic estimation method was used:

[0093] in, It is an estimated state. It is the state estimation error. It is a positive correction factor. It is a filtering function.

[0094] S74: Parameter estimation error Defined as:

[0095]

[0096] Using the algorithm described above, the estimated parameter values ​​are updated at each sampling time.

[0097] S75: Solid oxide fuel cell systems are represented in state-space form, and the control objective is to regulate the output power. To track the reference power. The discretized state equation of the system is:

[0098]

[0099]

[0100] in, It is a state vector containing fuel flow rate and the pressure of each gas. It is the control input (fuel flow) ), It's a disturbance (load current). It is the system output (fuel cell output power) ).

[0101] S76: System Matrix It can be obtained through a linearized model:

[0102]

[0103] in, It is the fuel reaction time constant. , , These are the reaction time constants for hydrogen, water, and oxygen, respectively.

[0104] S77: Control Input Matrix Perturbation input matrix and output matrix They are respectively:

[0105]

[0106]

[0107]

[0108] S78: The controller's objective is to minimize the weighted sum of the output error and the control error. The cost function and constraints are as follows:

[0109]

[0110]

[0111] in, It is the reference output for step i. and These are the corresponding weighting coefficients. and These are the lengths of the prediction time domain and the control time domain, respectively.

[0112] S79: By optimizing the solution, we can obtain the optimal control sequence. and apply the first control action. :

[0113]

[0114]

[0115] By scrolling forward, updating parameters, and addressing optimization issues, the desired control input can be obtained.

[0116] The implementation steps of the controller are as follows: Within each sampling period, the controller updates the system's control input by solving an optimization problem. The specific steps are as follows:

[0117] (1) Linearize and discretize the solid oxide fuel cell model;

[0118] (2) Selecting the prediction time domain Control Time Domain Output tracking error weight matrix and control input weight matrix ;

[0119] (3) Initialize control input and state;

[0120] (4) Obtain real-time measurements of the solid oxide fuel cell system;

[0121] (5) Update the parameters in the solid oxide fuel cell model based on the measurement data, using the parameter update method;

[0122] (6) Solve the optimization problem to obtain the optimal control input;

[0123] (7) Apply the first control action to the solid oxide fuel cell system;

[0124] (8) Increase the time sample and repeat the optimization process.

[0125] In this embodiment of the invention, "ensuring system stability" means avoiding local overheating to 800°C inside the solid oxide fuel cell, and ensuring that the temperature gradient is no greater than 8K / cm.

[0126] This invention also provides an electronic device, including a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the predictive control method described above is performed.

[0127] This invention also provides a computer-readable storage medium storing a computer program, which is executed by a processor using the predictive control method described above.

[0128] The memory may be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory stores the program, and the processor executes the program upon receiving an execution instruction.

[0129] The processor may be an integrated circuit chip with signal processing capabilities. The aforementioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor.

[0130] Example 1

[0131] This invention provides a comparison between the predictive control method based on a dual-temperature layer and the predictive control method based on a single-temperature layer.

[0132] The input parameters for the predictive control methods of the two models are shown in Table 1.

[0133] Table 1. Initial Input

[0134] Parameter name numerical values Prediction Time Domain 20 Control Time Domain 5 Output weights 0.5 Control weight 0.1 Sampling time 1s Control input constraints [0,100]A Output constraints [830,870]KW

[0135] like Figure 3 As shown, by adopting the technical solution provided in this application, the dual-temperature-layer model disclosed in this invention has higher control accuracy and better follow-up performance compared to the single-temperature-layer model.

[0136] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A model predictive control method for a solid oxide fuel cell stack system, characterized in that, include: A planar cross-flow solid oxide fuel cell model was constructed, and the model was divided into n×n nodes of equal area, where each node represents a different physical layer of the fuel cell. The fuel cell was modeled using a dual-temperature layer design. A thermo-mass-electric coupling simulation was performed on the solid oxide fuel cell model to simulate the spatial distribution characteristics of the temperature field, component concentration field, and current density inside the solid oxide fuel cell. A model predictive controller is designed based on the solid oxide fuel cell stack model. The model predictive controller adjusts the fuel flow rate and the stack input voltage to achieve precise control of the solid oxide fuel cell stack output power.

2. The predictive control method according to claim 1, characterized in that, The dual-temperature layer design includes: The anode layer, electrolyte layer, cathode layer, and connector layer are considered as one temperature layer, while the air layer is considered as an independent temperature layer.

3. The predictive control method according to claim 1, characterized in that, The physical layer includes an air layer, an anode layer, a cathode layer, an electrolyte layer, and a connector layer. The accuracy of the model is ensured by physical field coupling calculations between nodes, and the model is divided into 5×5 nodes.

4. The predictive control method according to claim 1, characterized in that, The dual-temperature layer design includes: The anode layer, electrolyte layer, cathode layer, and connector layer are modeled as a common temperature layer, and the heat transfer of this temperature layer is coupled with the electrochemical reaction in the calculation. The air layer is an independent temperature layer, modeled separately from the anode layer, electrolyte layer, cathode layer, and connector layer.

5. The predictive control method according to claim 1, characterized in that, The model predictive controller predicts the output power of the solid oxide fuel cell stack in real time and adjusts the power accordingly to optimize fuel flow and stack input voltage.

6. The predictive control method according to claim 5, characterized in that, The model prediction controller is implemented in the following way: The dynamic behavior of solid oxide fuel cell stacks is modeled in state space, and the state vector includes fuel flow rate, pressure of each gas, temperature and current density. Based on model predictions, the state at at least one future time step is predicted, and the optimal control input is calculated to minimize the control error. When solving optimization problems, the output power error is minimized, and the stability of the system and the minimization of control energy consumption are ensured by adjusting the weights of the control inputs.

7. The predictive control method according to claim 6, characterized in that, The model predictive controller uses an adaptive algorithm to estimate key parameters of the solid oxide fuel cell stack in real time. These key parameters include the flow coefficients of hydrogen, oxygen, and water, electrochemical reactions, and the temperature of the stack.

8. The predictive control method according to claim 7, characterized in that, The estimation of the key parameters is achieved through the following steps: The system model is decomposed using the dynamic key parameter estimation method. Some of the estimated key parameters are affected by parameter changes, while the other part is related to the control input. Within each sampling period, the estimated key parameters in the solid oxide fuel cell are updated.

9. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed, implements the predictive control method according to any one of claims 1-8.

10. An electronic device, characterized in that, include: The electronic device includes a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the predictive control method as described in any one of claims 1-8 is performed.

Images