Alloy hydrogen storage hydrogen absorption and desorption control method and system based on proxy model

Abstract

The invention discloses an alloy hydrogen storage hydrogen absorption and desorption control method and system based on an agent model. The method comprises the steps that leakage detection of an alloy hydrogen storage system and activation of an adsorbent are conducted; setting a hydrogen absorption / desorption mass flow rate, taking the hydrogen absorption / desorption mass flow rate as a target value of a hydrogen absorption / desorption rate, and determining initial hydrogen pressure and temperature according to the proxy model; according to the real-time hydrogen pressure and temperature in the hydrogen storage reactor, the real-time adsorption / desorption rate is rapidly predicted; and the temperature of the hydrogen storage reactor is adjusted by adopting a PID control algorithm. The agent model based on data driving is obtained through training by adopting a small sample learning technology, and the problem of model training accuracy caused by insufficient data volume is solved. The model is small in scale and quick in response, and can be quickly deployed in a control algorithm, so that thermal environment parameter control is effectively implemented in an easy-to-implement temperature parameter control mode, and the stability of hydrogen absorption and desorption is improved.

Description

Technical Field

[0001] This application relates to a method and system for controlling hydrogen absorption and desorption in alloy hydrogen storage, belonging to the field of hydrogen energy technology, and particularly to a method and system for controlling hydrogen absorption and desorption in alloy hydrogen storage based on a proxy model. Background Technology

[0002] One of the main bottlenecks in hydrogen energy applications is the efficient storage and release of hydrogen and its isotopes. The principle of metal hydride hydrogen storage is that hydrogen atoms form hydrides through metallic valence bonds. These hydrides include rare-earth lanthanum-nickel alloys, titanium-iron alloys, magnesium alloys, vanadium, niobium, zirconium, and other multi-element alloys, such as NaH-Al-Ti, Li3N-LiNH2, MgB2-LiH, MgH2-Cr2O3, and Ni(Cu, Rh)-Cr-FeO. x Substances such as metal hydrides possess advantages such as high volumetric hydrogen storage density and high safety. However, due to the significant thermal effects of hydrogen absorption and desorption reactions in metal hydrides, and the poor heat and mass transfer characteristics of porous media packed beds, the hydrogen absorption and desorption rates in hydrogen storage tubes change significantly. Specifically, under a thermally uncontrolled system, the hydrogen absorption and desorption rates decrease drastically, thus affecting the practical application and promotion.

[0003] The hydrogen absorption and desorption rates of alloy hydrogen storage systems are mainly controlled through the following methods: 1) enhancing heat transfer characteristics by adding high thermal conductivity frameworks, cooling pipes, powder briquettes, etc.; 2) optimizing heat and mass transfer processes in the hydrogen storage reactor; and 3) enhancing mass transfer characteristics. However, these methods still cannot solve the stability and flow rate control problems of the hydrogen absorption and desorption process, further limiting the application scenarios of alloy hydrogen storage systems. Summary of the Invention

[0004] According to one aspect of this application, a method for controlling hydrogen absorption and desorption in alloy hydrogen storage based on a surrogate model is provided. This method achieves real-time control of the hydrogen absorption / desorption rate of the hydrogen storage reactor by regulating key flow field variables, thereby solving the problem of efficient storage and release of hydrogen and isotopes in hydrogen energy utilization.

[0005] The alloy hydrogen storage and hydrogen absorption / desorption control method based on the surrogate model includes: (1) Leak detection of the hydrogen storage reactor and pipelines; (2) After several adsorption-desorption cycles, the amount of adsorbed hydrogen reaches a stable state, thus activating the adsorbent. (3) Set the hydrogen absorption / desorption mass flow rate This is used as the target value for the hydrogen adsorption / desorption rate; the initial hydrogen pressure is set based on a surrogate model. Initial temperature The proxy model is a mapping relationship model: Where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate. (4) Monitor the real-time hydrogen pressure in the hydrogen storage reactor. Real-time temperature The real-time adsorption / desorption rate is calculated based on the surrogate model. ; (5) The temperature of the hydrogen storage reactor is adjusted by using a PID control algorithm to improve the real-time adsorption / desorption rate. Target value for the adsorption / desorption rate of hydrogen Approaching, achieving the target for hydrogen absorption and desorption control.

[0006] Optionally, the surrogate model is trained on CFD simulation data using a few-shot learning method. The CFD simulation data includes the hydrogen pressure P of the hydrogen storage reactor, the temperature T of the hydrogen storage reactor, and the instantaneous adsorption / desorption rate M, where P and T are the input parameters of the surrogate model, and M is the output parameter of the surrogate model. Optionally, the CFD simulation data is obtained by solving the control equations for mass and heat transfer in porous media, as shown in the following formula: Conservation of mass:

[0007] Conservation of momentum:

[0008] Energy conservation:

[0009] In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor; t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, This is an energy source term.

[0010] Optionally, the effective specific heat capacity of the hydrogen storage reactor layer It is expressed as a weighted function of the porosity of hydrogen gas and the solid metallic phase, i.e.:

[0011]

[0012] in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached.

[0013] Optionally, the quality source item Represented as:

[0014] in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature (i.e., the temperature of the hydrogen-absorbing metal in the reactor). and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to hydride temperature T and atomic ratio Closely related, as shown in the following formula: ; in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value.

[0015] Optionally, the momentum source term In porous media regions, Darcy's law can be used to determine the situation.

[0016] Optionally, the energy source item The value is positive during the hydrogen absorption reaction, and is expressed as: .

[0017] Optionally, in the balancing pressure Compared to atoms The correlation between them was obtained by fitting and training with deep learning through PCT curve data of hydrogen adsorption / desorption.

[0018] Optionally, the temperature of the hydrogen storage reactor can be adjusted online by regulating the temperature of the chiller and / or the heating rod.

[0019] According to another aspect of this application, an alloy hydrogen storage and absorption / desorption control system based on a proxy model is provided. It includes a gas source module, a hydrogen storage reactor module, a thermal environment control module, a measurement module, a proxy model module, and a control module; The gas source module is used to deliver hydrogen to the hydrogen storage reactor; The hydrogen storage reactor module is used to store hydrogen through hydrogen absorption and desorption chemical reactions; The thermal environment control module is used to monitor the temperature changes in the flow pipeline and the hydrogen storage fixed bed reactor in the hydrogen storage reactor module, and to execute the instructions of the control module. The measurement module is used to acquire data from the hydrogen storage reactor; The proxy model module is used to obtain CFD simulation data based on the data collected by the measurement module, and to train a model based on the CFD simulation data. The mapping relationship model is used as a surrogate model, where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate; the user-input hydrogen adsorption / desorption mass flow rate is obtained. This value is used as the target value for the hydrogen adsorption / desorption rate, and the initial hydrogen pressure is set according to the surrogate model. Initial temperature The output is sent to the control module; based on the real-time hydrogen pressure of the hydrogen storage reactor collected by the measurement module... Real-time temperature Predicting real-time adsorption / desorption rates ; The control module is used to set the initial hydrogen pressure of the gas source module to... The initial temperature of the hydrogen storage reactor is set by the thermal environment control module. ; Obtain the real-time adsorption / desorption rate predicted by the proxy model module. The target value for its adsorption / desorption rate with hydrogen. In comparison, a PID controller is used to generate a control signal and send it to the thermal environment control module, so that the module controls the hydrogen storage reactor module to quickly adjust the temperature of the hydrogen storage reactor to achieve the target value of hydrogen adsorption / desorption rate.

[0020] Optionally, the surrogate model module is used to train a surrogate model on CFD simulation data using a few-shot learning method. The input of the surrogate model is the hydrogen pressure P and temperature T of the hydrogen storage reactor, and the output is the instantaneous adsorption / desorption rate M.

[0021] Optionally, the surrogate model module is also used to solve the control equations for mass and heat transfer in porous media as shown in the following formula to obtain CFD simulation data: Conservation of mass:

[0022] Conservation of momentum:

[0023] Energy conservation:

[0024] In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor; t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, This is an energy source term.

[0025] Optionally, the proxy model module is used to calculate the effective specific heat capacity of the hydrogen storage reactor layer using a porosity weighting function of hydrogen and solid metal phases. :

[0026] in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached.

[0027] Optionally, the proxy model module is used to calculate the quality source item according to the following formula. :

[0028] in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature. and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to hydride temperature T and atomic ratio Closely related, as shown in the following formula: ; in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value.

[0029] Optionally, the surrogate model module is used to determine the momentum source term of the porous medium region according to Darcy's law. .

[0030] Optionally, the proxy model module is used to calculate the energy source term according to the following formula. : .

[0031] Optionally, the surrogate model module is further configured to first obtain the model through deep learning fitting training based on the PCT curve data of hydrogen adsorption / desorption. Compared to atoms The relationships between them.

[0032] Optionally, the control module is used to control the temperature of the chiller and / or heating rod in the thermal environment control module, thereby adjusting the temperature of the hydrogen storage reactor in the hydrogen storage reactor module.

[0033] The beneficial effects that this application can produce include: 1) The alloy hydrogen storage hydrogen adsorption and desorption control method and system based on the surrogate model provided in this application can quickly predict the current instantaneous adsorption / desorption rate based on the measured real-time hydrogen pressure and temperature data of the hydrogen storage projector, which provides an important guarantee for the effective implementation of thermal environment parameter control algorithm.

[0034] 2) The alloy hydrogen storage hydrogen absorption and desorption control method and system based on the surrogate model provided in this application can improve the stability of hydrogen absorption and desorption, and make it easy to achieve the hydrogen absorption and desorption control target by controlling the temperature parameters of the hydrogen storage reactor.

[0035] 3) This application provides a data-driven surrogate model. First, sampling is performed in the sample space, followed by numerical simulation of key flow field variables to obtain CFD simulation data. Then, few-shot learning techniques are used for training. The establishment of this surrogate model solves the problem of poor model training accuracy caused by insufficient data in practical applications. Furthermore, the surrogate model is small in scale and responds quickly, thus enabling rapid deployment in control algorithms. Attached Figure Description

[0036] Figure 1 This is a schematic flowchart of an alloy hydrogen storage and hydrogen absorption / desorption control method based on a proxy model in one embodiment of this application. Figure 2 This is a schematic diagram of the framework of an alloy hydrogen storage and absorption / desorption control system based on a proxy model in one embodiment of this application. Detailed Implementation

[0037] The present application is described in detail below with reference to the embodiments, but the present application is not limited to these embodiments.

[0038] This application provides a method for controlling hydrogen absorption and desorption in alloy hydrogen storage based on a surrogate model, such as... Figure 1 Shown, including: S1: Leak detection of the hydrogen storage reactor and pipelines.

[0039] S2: Activate the adsorbent.

[0040] In one embodiment, the adsorbent is an alloy adsorbent, and the activation step includes: After several adsorption-desorption cycles, the amount of adsorbed hydrogen reaches a stable state.

[0041] S3: Set the hydrogen absorption / desorption mass flow rate The initial hydrogen pressure is determined based on the surrogate model. Initial temperature .

[0042] The proxy model is a mapping relationship model, expressed as: Where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate.

[0043] During use, the hydrogen absorption / desorption mass flow rate is... Using a surrogate model, the initial hydrogen pressure of the gas source module can be determined as the target value for the hydrogen adsorption / desorption rate. Initial temperature of the hydrogen storage reactor .

[0044] In one implementation, the training method for the proxy model includes: S31: Sampling is performed in the sample space; In one embodiment, step S31 includes: using the Latin hypercube sampling method to generate approximately 500 sets of samples for the operating conditions of the hydrogen storage reactor, including pressure P and temperature T.

[0045] S32: Perform numerical simulations of key flow field variables, solve the mass and heat transfer control equations of porous media, and obtain CFD simulation data. The CFD simulation data includes hydrogen pressure P, hydrogen storage reactor temperature T, and instantaneous adsorption / desorption rate M of the hydrogen storage reactor. P and T are the input parameters of the surrogate model, and M is the output parameter of the surrogate model.

[0046] The mass and heat transfer control equations for the porous medium are shown in the following formula: Conservation of mass:

[0047] Conservation of momentum:

[0048] Energy conservation:

[0049] In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor; t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, This is an energy source term.

[0050] In one embodiment, the effective specific heat capacity of the hydrogen storage reactor layer It is expressed as a weighted function of the porosity of hydrogen gas and the solid metallic phase, i.e.:

[0051]

[0052] in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached.

[0053] In one implementation, the quality source item Represented as:

[0054] in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature. and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to hydride temperature T and atomic ratio Closely related, as shown in the following formula: ; in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value.

[0055] In a preferred embodiment, the aforementioned balancing pressure Compared to atoms The correlation between them was obtained by fitting and training with deep learning through PCT curve data of hydrogen adsorption / desorption.

[0056] In one implementation, the momentum source term In porous media, this is determined by Darcy's law.

[0057] In one implementation, the energy source item The value is positive during the hydrogen absorption reaction, and is expressed as: .

[0058] S4: Obtain the real-time hydrogen pressure inside the hydrogen storage reactor. Real-time temperature The real-time adsorption / desorption rate is calculated based on the surrogate model. ; S5: Real-time adsorption / desorption rate Target value for the adsorption / desorption rate of hydrogen By comparison, a PID control algorithm is used to regulate the temperature of the hydrogen storage reactor to achieve the desired effect. achieve .

[0059] In one embodiment, the temperature of the hydrogen storage reactor is adjusted by online regulation of the temperature of the chiller and / or heating rod.

[0060] See Figure 2 The diagram illustrates a framework of the alloy hydrogen storage and absorption control system based on a proxy model as described in this application, in one embodiment. The system includes a gas source module 1, a hydrogen storage reactor module 2, a thermal environment control module 3, a measurement module 4, a proxy model module 5, and a control module 6.

[0061] Gas source module 1 is used to deliver hydrogen to the hydrogen storage reactor.

[0062] In one embodiment, the gas source module 1 mainly includes: a commercial hydrogen cylinder, a ball valve, a mass flow meter, a vacuum pump, and a hydrogen pipeline.

[0063] Hydrogen storage reactor module 2 is used to store hydrogen through hydrogen absorption and desorption chemical reactions.

[0064] In one embodiment, the hydrogen storage reactor module 2 mainly includes: a fixed-bed reactor, a heat exchange circuit, a pressure sensor, a thermocouple, and metal hydride particles.

[0065] The thermal environment control module 3 is used to monitor the temperature changes in the flow pipeline and the hydrogen storage fixed bed reactor in the hydrogen storage reactor module, and to execute the instructions of the control module.

[0066] In one embodiment, the thermal environment control module 3 mainly includes: a heating rod, a chiller, a heat exchange pipeline, a thermocouple, a flow meter, and a control valve.

[0067] Measurement module 4 is used to collect data from the hydrogen storage reactor.

[0068] In one embodiment, the measurement module 4 mainly includes: a data acquisition system, a thermocouple, a flow meter, and a pressure gauge.

[0069] In one embodiment, the data collected by the measurement module 4 includes, but is not limited to: fixed bed temperature, reactor inlet and outlet temperatures, reactor supply pressure, reactor internal pressure, pipeline hydrogen flow rate, and heat exchanger water flow rate.

[0070] Proxy model module 5 is used to solve for CFD simulation data based on the data collected by the measurement module, and to train a model based on the CFD simulation data. The mapping relationship model is used as a surrogate model, where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate; the user-input hydrogen adsorption / desorption mass flow rate is obtained. This value is used as the target value for the hydrogen adsorption / desorption rate, and the initial hydrogen pressure is set according to the surrogate model. Initial temperature The output is sent to the control module; based on the real-time hydrogen pressure of the hydrogen storage reactor collected by the measurement module... Real-time temperature Predicting real-time adsorption / desorption rates .

[0071] In one implementation, the surrogate model module is used to train a surrogate model on CFD simulation data using a few-shot learning method. The inputs of the surrogate model are the hydrogen pressure P and temperature T of the hydrogen storage reactor, and the output is the instantaneous adsorption / desorption rate M.

[0072] In one implementation, the surrogate model module is further used to solve the porous medium mass and heat transfer control equations as shown in the following formula to obtain CFD simulation data: Conservation of mass:

[0073] Conservation of momentum:

[0074] Energy conservation:

[0075] In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor;t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, This is an energy source term.

[0076] In one embodiment, the proxy model module is used to calculate the effective specific heat capacity of the hydrogen storage reactor layer using a porosity weighted function of hydrogen and solid metal phases. :

[0077] in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached.

[0078] In one implementation, the proxy model module is used to calculate the quality source item according to the following formula. :

[0079] in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature. and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to hydride temperature T and atomic ratio Closely related, as shown in the following formula: ; in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value.

[0080] In one implementation, the surrogate model module is used to determine the momentum source term of the porous medium region according to Darcy's law. .

[0081] In one implementation, the proxy model module is used to calculate the energy source term according to the following formula. : .

[0082] In one implementation, the surrogate model module is further configured to first obtain the model through deep learning fitting training based on the PCT curve data of hydrogen adsorption / desorption. Compared to atoms The relationships between them.

[0083] Control module 6 is used to set the initial hydrogen pressure of the gas source module to... The initial temperature of the hydrogen storage reactor is set by the thermal environment control module. ; Obtain the real-time adsorption / desorption rate predicted by the proxy model module. The target value for its adsorption / desorption rate with hydrogen. In comparison, a PID controller is used to generate a control signal and send it to the thermal environment control module, so that the module controls the hydrogen storage reactor module to quickly adjust the temperature of the hydrogen storage reactor to achieve the target value of hydrogen adsorption / desorption rate.

[0084] In one embodiment, the control module 6 is used to control the temperature of the chiller and / or heating rod in the thermal environment control module, thereby adjusting the temperature of the hydrogen storage reactor in the hydrogen storage reactor module.

[0085] In this application, the temperature of the hydrogen storage tank reactor is the main variable for regulating the alloy hydrogen storage system. Traditionally, this controlled object is described through numerical simulation, followed by coupled solution of the governing equations using the finite element method or Boltzmann method to predict changes in hydrogen absorption / desorption rates, temperature, and pressure. However, numerical models based on the Navier-Stokes equations and convection-diffusion equations have long response times, requiring multiple iterations to approximate accurate solutions under boundary and initial conditions, which struggles to meet the real-time response requirements of the control algorithm.

[0086] This application establishes a data-driven surrogate model, which, after simulating and mathematically sampling the sample space, obtains an approximate mathematical model represented by a neural network through deep learning technology. The computational load for solving this model is small, and the computational accuracy is not much different from that of the high-reliability simulation model, which can realize real-time control of the hydrogen absorption and desorption rate of the hydrogen storage reactor.

[0087] The above description is merely a few embodiments of this application and is not intended to limit this application in any way. Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any changes or modifications made by those skilled in the art without departing from the scope of the technical solution of this application using the disclosed technical content are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for controlling hydrogen absorption and desorption in alloy hydrogen storage based on a surrogate model, characterized in that, The method includes: (1) Leak detection of the hydrogen storage reactor and pipelines; (2) After several adsorption-desorption cycles, the amount of adsorbed hydrogen reaches a stable state, thus activating the adsorbent. (3) Set the hydrogen absorption / desorption mass flow rate This is used as the target value for the hydrogen adsorption / desorption rate; the initial hydrogen pressure is set based on a surrogate model. Initial temperature The proxy model is a mapping relationship model: Where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate. (4) Monitor the real-time hydrogen pressure in the hydrogen storage reactor. Real-time temperature The real-time adsorption / desorption rate is calculated based on the surrogate model. ; (5) The temperature of the hydrogen storage reactor is adjusted by using a PID control algorithm to control the real-time adsorption / desorption rate. To achieve the target value of hydrogen adsorption / desorption rate .

2. The alloy hydrogen storage and hydrogen absorption / desorption control method based on a surrogate model according to claim 1, characterized in that, The surrogate model is trained on CFD simulation data using a few-shot learning method. The CFD simulation data includes hydrogen pressure P, hydrogen temperature T, and instantaneous adsorption / desorption rate M of the hydrogen storage reactor. P and T are the input parameters of the surrogate model, and M is the output parameter of the surrogate model.

3. The alloy hydrogen storage and hydrogen absorption / desorption control method based on a surrogate model according to claim 2, characterized in that, The CFD simulation data was obtained by solving the control equations for mass and heat transfer in porous media, as shown in the following formula: Conservation of mass: Conservation of momentum: Energy conservation: In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor; t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, For energy source terms; Preferably, the effective specific heat capacity of the hydrogen storage reactor layer is... It is expressed as a weighted function of the porosity of hydrogen gas and the solid metallic phase, i.e.: in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached. Preferably, the quality source item Represented as: in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature. and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to the hydride temperature T and atomic ratio inside the reactor Closely related, as shown in the following formula: ； in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value. Preferably, the momentum source term In porous media regions, this is determined by Darcy's law; Preferably, the energy source item The value is positive during the hydrogen absorption reaction, and is expressed as: 。 4. The alloy hydrogen storage and hydrogen absorption / desorption control method based on a surrogate model according to claim 3, characterized in that, The equilibrium pressure Compared to atoms The correlation between them was obtained by fitting and training with deep learning through PCT curve data of hydrogen adsorption / desorption.

5. In the alloy hydrogen storage and hydrogen absorption / desorption control method based on the proxy model according to claim 1, in step (5), the temperature of the hydrogen storage reactor is adjusted by online adjustment of the temperature of the chiller and / or the heating rod.

6. A surrogate model-based alloy hydrogen storage and absorption / desorption control system, characterized in that, The system includes a gas source module, a hydrogen storage reactor module, a thermal environment control module, a measurement module, a proxy model module, and a control module; The gas source module is used to deliver hydrogen to the hydrogen storage reactor; The hydrogen storage reactor module is used to store hydrogen through hydrogen absorption and desorption chemical reactions; The thermal environment control module is used to monitor the temperature changes in the flow pipeline and the hydrogen storage fixed bed reactor in the hydrogen storage reactor module, and to execute the instructions of the control module. The measurement module is used to acquire data from the hydrogen storage reactor; The proxy model module is used to obtain CFD simulation data based on the data collected by the measurement module, and to train a model based on the CFD simulation data. The mapping relationship model is used as a surrogate model, where P is the hydrogen pressure of the hydrogen storage reactor, T is the temperature of the hydrogen storage reactor, and M is the instantaneous adsorption / desorption rate; the user-input hydrogen adsorption / desorption mass flow rate is obtained. This value is used as the target value for the hydrogen adsorption / desorption rate, and the initial hydrogen pressure is set according to the surrogate model. Initial temperature The output is sent to the control module; based on the real-time hydrogen pressure of the hydrogen storage reactor collected by the measurement module... Real-time temperature Predicting real-time adsorption / desorption rates ; The control module is used to set the initial hydrogen pressure of the gas source module to... The initial temperature of the hydrogen storage reactor is set by the thermal environment control module. ; Obtain the real-time adsorption / desorption rate predicted by the surrogate model module. The target value for its adsorption / desorption rate with hydrogen. In comparison, a PID controller is used to generate a control signal and send it to the thermal environment control module, so that the module controls the hydrogen storage reactor module to quickly adjust the temperature of the hydrogen storage reactor to achieve the target value of hydrogen adsorption / desorption rate.

7. The alloy hydrogen storage and desorption control system based on a surrogate model according to claim 6, characterized in that, The surrogate model module is used to train a surrogate model on CFD simulation data using a few-shot learning method. The input of the surrogate model is the hydrogen pressure P and temperature T of the hydrogen storage reactor, and the output is the instantaneous adsorption / desorption rate M.

8. The alloy hydrogen storage and desorption control system based on a proxy model according to claim 7, characterized in that, The proxy model module is also used to solve the control equations for mass and heat transfer in porous media, as shown in the following formula, to obtain CFD simulation data: Conservation of mass: Conservation of momentum: Energy conservation: In the formula: Porosity of hydrogen storage materials; For Hamiltonian operators; The density of hydrogen gas inside the reactor; t For time, The velocity vector of hydrogen gas; For quality source items; This refers to the absolute pressure inside the reactor. It is the viscous stress tensor; For momentum source term; It is the acceleration due to gravity; T The temperature inside the reactor; The specific heat capacity of hydrogen; Effective thermal conductivity; The effective specific heat capacity of the hydrogen storage reactor layer, For energy source terms; Preferably, the proxy model module is used to calculate the effective specific heat capacity of the hydrogen storage reactor layer using a porosity weighting function of hydrogen and solid metal phases. : in, For the porosity of hydrogen storage materials, and These are the instantaneous densities of hydrogen gas and hydrogen-absorbing alloy, respectively. and Specific heat capacities of hydrogen and hydrogen-absorbing alloys, respectively The thermal conductivity of hydrogen gas, and These represent the thermal conductivity of the hydride and the hydrogen-absorbing alloy, respectively. Expressed as the ratio of hydrogen atoms to metal atoms, subscript sat This indicates that the hydrogen absorption state has been reached. Preferably, the proxy model module is used to calculate the quality source item according to the following formula. : in, and Represent the hydrogen absorption reaction constant and the activation energy, respectively. RT It is the product of the gas constant and the reaction temperature. and These are the hydrogen pressure and the hydrogen absorption equilibrium pressure, respectively. and These represent the instantaneous density and initial density of the hydrogen-absorbing alloy, respectively; and the equilibrium pressure serves as the driving force for hydrogen adsorption and desorption. With respect to hydride temperature T and atomic ratio Closely related, as shown in the following formula: ； in, Enthalpy change for hydrogen absorption reaction; The entropy change is due to the hydrogen absorption reaction; R g Represents the ideal gas constant; For simplified polynomial functions, their coefficients can be fitted using PCT curves; the subscript ref indicates a reference value. Preferably, the surrogate model module is used to determine the momentum source term of the porous medium region according to Darcy's law. ; Preferably, the proxy model module is used to calculate the energy source term according to the following formula. : 。 9. The alloy hydrogen storage and desorption control system based on a surrogate model according to claim 8, characterized in that, The proxy model module is further used to first obtain the model through deep learning fitting training based on the PCT curve data of hydrogen adsorption / desorption. Compared to atoms The relationships between them.

10. The alloy hydrogen storage and desorption control system based on a surrogate model according to claim 6, characterized in that, The control module is used to control the temperature of the chiller and / or heating rod in the thermal environment control module, thereby adjusting the temperature of the hydrogen storage reactor in the hydrogen storage reactor module.

Images