Model-based hydrogen fueling method for improving a given fueling protocol and controller using the same

Abstract

The hydrogen fuel filling method for a mobile body device using hydrogen as a fuel according to the present invention includes the steps of: acquiring first filling data in a first hydrogen fuel filling process that is performed based on a first filling protocol; training a control model to learn a control function of the hydrogen fuel filling process based on the first filling data; and performing a second hydrogen fuel filling process based on at least one of the first filling protocol and the control model.

Description

Technical Field

[0001] This disclosure relates to a control technology for refueling hydrogen fuel vehicles, and more specifically, to a model-based hydrogen refueling method for improving a given refueling protocol, and a control device using the model-based hydrogen refueling method, the model-based hydrogen refueling method relating to a hydrogen refueling process for improving hydrogen refueling efficiency and enhancing the convenience, speed and real-time performance of refueling, and a platform for the hydrogen refueling process. Background Technology

[0002] The content described in this section is only to provide background information about this exemplary embodiment and does not constitute related technology.

[0003] Hydrogen vehicles, or hydrogen electric vehicles, are pollution-free vehicles powered by electricity generated from the combination of high-pressure hydrogen stored in the vehicle and atmospheric air. Hydrogen electric vehicles are also known as fuel cell electric vehicles (FCEVs). Most hydrogen electric vehicles use hydrogen as an energy source, employing fuel cell systems to generate electricity. In hydrogen electric vehicles, not only is pure water (H2O) emitted during power generation, but ultrafine dust from the atmosphere is also removed during operation, thus making them a promising environmentally friendly mode of transportation. Based on the fact that hydrogen is an inexhaustible fuel on Earth and that the energy generation process is environmentally friendly, hydrogen electric vehicles are widely recognized as a technology with the potential for application across various industrial sectors.

[0004] A hydrogen fuel cell vehicle is a mobile vehicle that uses hydrogen as an energy source or as fuel to generate electricity and uses that electricity to drive an electric motor. In addition to hydrogen-electric vehicles, hydrogen fuel cell vehicles may also include airborne vehicles, industrial trucks, trains, ships, aircraft, etc., and may include devices that use hydrogen as fuel to generate electricity and use that electricity for propulsion.

[0005] Most hydrogen-electric vehicles deliver high-pressure hydrogen, safely stored in hydrogen fuel tanks, and oxygen, introduced through an air supply system, to a fuel cell stack, generating electricity through an electrochemical reaction between hydrogen and oxygen. The resulting electricity is then converted into kinetic energy to drive an electric motor, which has the advantage of emitting only pure water through its outlet.

[0006] Meanwhile, the concept of hydrogen fuel cell vehicles, in addition to hydrogen electric vehicles, also refers to vehicles that use hydrogen as fuel. These vehicles drive an electric motor by directly burning hydrogen in an engine (ICE, Internal Combustion Engine). The hydrogen refueling scheme for hydrogen fuel cell vehicles is not significantly different from that for hydrogen electric vehicles.

[0007] In control technologies used to supply hydrogen to vehicles that use hydrogen as fuel, the ultimate goal is to control the temperature and pressure of the compressed hydrogen storage system (CHSS) on the fuel cell side to operate under extreme temperature and pressure conditions required for hydrogen refueling safety.

[0008] The hydrogen refueling process, control technologies, and protocols for conventional hydrogen electric vehicles were defined when wired / wireless communication technologies or the computing technologies used for control were immature. Therefore, recent advancements in information and communication technologies (ICT) have not been adequately reflected. Consequently, conventional hydrogen refueling technologies for hydrogen electric vehicles are inefficient, slow, and unsuitable for large-capacity hydrogen refueling. Summary of the Invention

[0009] Technical issues

[0010] One object of this disclosure for solving the above problems is to propose a hydrogen refueling process that overcomes the limitations and vulnerabilities of existing one-way communication, communication protocols for hydrogen refueling processes, and refueling protocols in the hydrogen refueling process of hydrogen fuel vehicles, and improves the safety, compatibility, efficiency, and reliability of hydrogen refueling.

[0011] Another objective of this disclosure is to improve the speed and real-time performance of the hydrogen refueling process, and to enhance the accuracy of the control and prediction processes of the hydrogen refueling process, by utilizing additional means such as artificial neural network models and / or model predictive control.

[0012] Another objective of this disclosure is to provide a process for applying optimized communication and refueling protocols that can utilize conventional or advanced communication media to enable hydrogen fuel vehicles and dispensers to efficiently achieve refueling objectives.

[0013] Another objective of this disclosure is to propose a refueling process optimized for the ambient conditions and environment during hydrogen refueling.

[0014] Technical solution

[0015] According to an exemplary embodiment of this disclosure for achieving the above-described objectives, a method for refueling a hydrogen-fueled mobile vehicle, the method being executed in a dispenser for refueling the hydrogen-fueled mobile vehicle, may include: acquiring first refueling data obtained during a first hydrogen refueling process, the first hydrogen refueling process being executed based on a first refueling protocol; training a control model based on the first refueling data to learn control functions for the hydrogen refueling process; and executing a second hydrogen refueling process based on one or more of the first refueling protocol or the control model.

[0016] Performing the second hydrogen refueling process may include performing the second hydrogen refueling process by operating the first refueling protocol and control model in parallel.

[0017] Performing a second hydrogen refueling process may include performing the second hydrogen refueling process by using a control model to replace at least a portion of the functionality of the first refueling protocol.

[0018] According to an exemplary embodiment of this disclosure, the method for refueling a hydrogen-powered mobile vehicle may further include validating a control model.

[0019] At this point, performing the second hydrogen refueling process may include performing the second hydrogen refueling process by using a validated control model to replace at least a portion of the functionality of the first refueling protocol.

[0020] Performing a second hydrogen refueling process may include performing the second hydrogen refueling process by using at least a portion of the functions of the control model that support the first refueling protocol.

[0021] According to an exemplary embodiment of this disclosure, in a method of refueling a mobile vehicle with hydrogen as fuel, the control model may be a model that has learned model predictive control functions.

[0022] According to an exemplary embodiment of this disclosure, in a method for refueling a mobile vehicle with hydrogen as fuel, the control model may be an artificial neural network model that has learned the control functions of the hydrogen refueling process.

[0023] According to an exemplary embodiment of this disclosure, the method for refueling a hydrogen fuel cell vehicle may further include a process of determining whether a first refueling protocol supports the generation of control information through model predictive control.

[0024] According to an exemplary embodiment of this disclosure, the method of refueling a hydrogen fuel cell may further include determining whether a communication protocol associated with a first refueling protocol supports bidirectional communication between the cell and a dispenser for refueling the cell with hydrogen.

[0025] According to an exemplary embodiment of this disclosure, the method of refueling a hydrogen fuel cell may further include determining whether the cell and a dispenser for refueling the cell support bidirectional communication between the cell and the dispenser.

[0026] According to an exemplary embodiment of this disclosure, an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel may include: a memory storing at least one command; and a processor executing at least one command, wherein the processor can acquire first refueling data obtained in a first hydrogen refueling process through the at least one command, the first hydrogen refueling process being executed based on a first refueling protocol, a control model being trained based on the first refueling data to learn the control functions of the hydrogen refueling process, and a second hydrogen refueling process being executed based on one or more of the first refueling protocol and the control model.

[0027] The processor can execute the second hydrogen refueling process by operating the first refueling protocol and control model in parallel.

[0028] The processor can perform the second hydrogen refueling process by using a control model to replace at least a portion of the functionality of the first refueling protocol.

[0029] At this point, the processor can verify the control model and can perform the second hydrogen refueling process by replacing at least a portion of the functionality of the first refueling protocol with the verified control model.

[0030] The processor can execute the second hydrogen refueling process by using at least a portion of the functions that support the first refueling protocol in the control model.

[0031] According to an exemplary embodiment of this disclosure, in an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, the control model may be a model that has learned model predictive control functions.

[0032] According to an exemplary embodiment of this disclosure, in an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, the control model may be an artificial neural network model that has learned the control functions of the hydrogen refueling process.

[0033] According to an exemplary embodiment of this disclosure, in an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, a processor can determine whether a first refueling protocol supports a process of generating control information through model predictive control.

[0034] According to an exemplary embodiment of this disclosure, in an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, a processor may determine whether a communication protocol associated with a first refueling protocol supports bidirectional communication between the mobile vehicle and a dispenser for refueling the mobile vehicle with hydrogen.

[0035] According to an exemplary embodiment of this disclosure, in an apparatus for controlling the process of refueling a mobile body with hydrogen as fuel, a processor can determine whether the mobile body and a dispenser for refueling the mobile body with hydrogen support bidirectional communication between the mobile body and the dispenser.

[0036] Beneficial effects

[0037] According to exemplary embodiments of this disclosure, the limitations and vulnerabilities of existing one-way communication, communication protocols for hydrogen refueling, and refueling protocols in the hydrogen refueling process of hydrogen fuel vehicles can be overcome, and the safety, compatibility, efficiency, and reliability of hydrogen refueling can be improved.

[0038] According to exemplary embodiments of this disclosure, the speed and real-time performance of the hydrogen refueling process can be improved, and the accuracy of the control and prediction processes of the hydrogen refueling process can be improved by utilizing additional means, such as artificial neural network models and / or model predictive control.

[0039] According to exemplary embodiments of this disclosure, a refueling process optimized for the surrounding conditions and environment during hydrogen refueling can be implemented. Attached Figure Description

[0040] Figure 1 This is a conceptual diagram illustrating the hydrogen refueling and control process of a hydrogen fuel vehicle applied to an exemplary embodiment of this disclosure.

[0041] Figure 2 This is a conceptual diagram illustrating an example of the state changes that occur during the refueling process of a hydrogen vehicle / hydrogen fuel vehicle applied to an exemplary embodiment of this disclosure.

[0042] Figure 3 This is a conceptual diagram illustrating a platform or test platform that applies a hydrogen refueling process according to an exemplary embodiment of this disclosure.

[0043] Figure 4 This is a conceptual diagram illustrating a model predictive control process according to an exemplary embodiment of the present disclosure.

[0044] Figure 5 This is a conceptual diagram illustrating a platform or test platform used for simulation during hydrogen refueling according to an exemplary embodiment of this disclosure.

[0045] Figure 6 This is a conceptual diagram illustrating the concept of using an artificial neural network to model, predict, or control the hydrogen refueling process in an exemplary embodiment of the present disclosure.

[0046] Figure 7 This is a conceptual diagram illustrating the concept of using model predictive control to control the hydrogen refueling process to achieve a target output in an exemplary embodiment of the present disclosure.

[0047] Figure 8 This is a flowchart illustrating the training and inference processes of an artificial neural network when used for model predictive control during hydrogen refueling according to an exemplary embodiment of the present disclosure.

[0048] Figure 9 It is a framework of functional blocks for executing a series of hydrogen refueling procedures using a bidirectional hydrogen refueling communication process according to an exemplary embodiment of the present disclosure.

[0049] Figure 10 This is a flowchart illustrating a hydrogen fuel refueling method according to an exemplary embodiment of the present disclosure.

[0050] Figure 11 This shows the implementation Figure 10 A conceptual diagram illustrating the training and application processes of an artificial neural network (ANN) model according to an exemplary implementation.

[0051] Figure 12 This indicates that it can be executed. Figures 1 to 11 A conceptual diagram of an example of a generalized hydrogen refueling control device, hydrogen refueling control system, hydrogen refueling simulation device, hydrogen refueling simulation system, hydrogen refueling test platform, hydrogen refueling test system, or computing system, representing at least a part of the process. Detailed Implementation

[0052] In addition to the foregoing objectives, other objectives and features of this disclosure will become apparent from the following description of exemplary embodiments with reference to the accompanying drawings.

[0053] This disclosure can be implemented in various modifications and can have various exemplary embodiments. Therefore, specific exemplary embodiments are illustrated in the accompanying drawings and will be described in detail in the following description. However, this is not intended to limit this disclosure to the specific exemplary embodiments, and it should be understood that all modifications, equivalents, and substitutions falling within the spirit and scope of this disclosure are included. In describing the drawings, the same reference numerals are used to refer to the same elements.

[0054] Terms such as first, second, A, and B may be used to describe various elements, but these elements should not be limited by such terms. These terms are used only for the purpose of distinguishing one element from another. For example, a first element may be referred to as a second element without departing from the scope of this disclosure, and similarly, a second element may be referred to as a first element. The term "and / or" includes any one of the plurality of related listed items as well as a combination of the plurality of related listed items.

[0055] In an exemplary embodiment of this application, "at least one of A and B" may mean "at least one of A or B" or "at least one of one or more combinations of A and B". Furthermore, in an exemplary embodiment of this application, "one or more of A and B" may mean "one or more of A or B" or "one or more of one or more combinations of A and B".

[0056] When a component is referred to as being "connected to" or "coupled to" another component, it should be understood that the component can be directly connected to or coupled to that other component, or one or more intermediate components may exist between them. Conversely, when a component is referred to as being "directly connected to" or "directly coupled to" another component, it should be understood that there are no intermediate components between them.

[0057] The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to limit this disclosure. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, terms such as “comprising” and “having” are intended to specify the presence of the stated features, numbers, steps, operations, elements, components, portions or combinations thereof, and should not be construed as excluding the presence or addition of one or more other features, numbers, steps, operations, elements, components, portions or combinations thereof.

[0058] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. Terms as defined in commonly used dictionaries shall be interpreted as having the same meaning as they have in the relevant technical context and shall not be interpreted in an idealized or overly formalized sense, unless otherwise expressly defined in this application.

[0059] The following are definitions of some terms used in this specification.

[0060] Hydrogen fuel cell vehicles typically include not only hydrogen electric vehicles or hydrogen fuel cell electric vehicles (FCEVs) that use fuel cells, but also internal combustion engine-based vehicles that use hydrogen as fuel.

[0061] In addition, hydrogen fuel cell vehicles refer to vehicles that use hydrogen as fuel, and these vehicles can include conventional automobiles, motor vehicles, and hybrid transportation vehicles that can utilize both human power and fuel.

[0062] In the following exemplary embodiments, a hydrogen refueling protocol and / or a communication protocol for hydrogen refueling can be applied to hydrogen fuel vehicles.

[0063] Hydrogen fuel can be either gaseous or liquid.

[0064] "CHSS" (Compressed Hydrogen Storage System) refers to a device that stores hydrogen fuel fluid (i.e., hydrogen fuel or compressed hydrogen) as part of a vehicle / mobile body.

[0065] A “pressure relief device” (PRD) is installed in a CHSS and refers to a device that can isolate stored hydrogen from the rest of the fuel system and the environment, or conversely, vent hydrogen to the outside.

[0066] Hydrogen refueling generally refers to the process of receiving high-pressure hydrogen from a hydrogen refueling station's distributor and compressing and storing it in the vehicle's tank. In the context of supplying hydrogen fuel to fuel cell electric vehicles, hydrogen refueling can be used interchangeably with refueling. That is, in this specification, the term "fueling" can be used to mean fuel supply, hydrogen refueling, or charging, and the term "charging" can refer to the charging of hydrogen fuel. For example, a refueling agreement can be referred to as a charging agreement, a refueling session can be referred to as a charging session, and a refueling method can be referred to as a hydrogen refueling method or a fuel charging method.

[0067] The "pressure ramp rate" (PRR) refers to the rate at which the pressure of the CHSS increases, and is measured in megapascals per minute (MPa / min).

[0068] The "average pressure ramp rate" (APRR) refers to the average rate of pressure increase from the start to the end of hydrogen refueling.

[0069] "Pre-cooling" refers to the process of cooling hydrogen at a hydrogen refueling station before it is added to electric vehicles or other applications.

[0070] A "distributor" is a component that delivers pre-cooled hydrogen to the CHSS (hydrogen storage unit). Distributors can be located at hydrogen refueling stations and can perform hydrogen refueling operations between the hydrogen storage tank at the refueling station and the CHSS of the vehicle.

[0071] A "nozzle" is a device that is connected to a dispenser, coupled to a receiver in a hydrogen-electric vehicle, and allows the delivery of hydrogen fuel.

[0072] "Refueling session" can be used to refer to the communication session executed for each use case of hydrogen refueling.

[0073] "Interoperability" can refer to the state in which components of relatively different systems can work together to perform the intended operation of the entire system. Information interoperability can refer to the ability of two or more networks, systems, devices, nodes, participants, entities, applications, or components to securely and efficiently share and use information with minimal inconvenience to users.

[0074] "Related" or "associated" can include procedures used to establish a relationship between two peer communicating entities.

[0075] "Command and control communication" can refer to the communication between the hydrogen refueling device and the hydrogen fuel cell vehicle, which is used to exchange information required to start, control, and terminate the hydrogen refueling process.

[0076] Furthermore, in the following detailed description, exemplary embodiments related to hydrogen electric vehicles or fuel cell electric vehicles may be shown for ease of description; however, it will be apparent to those skilled in the art that the spirit of this disclosure can be applied to various types of hydrogen fuel cell vehicles. A hydrogen fuel cell vehicle refers to a vehicle that uses hydrogen as an energy source or uses hydrogen as fuel to generate electrical energy and uses that electrical energy to drive an electric motor. In addition to hydrogen electric vehicles, hydrogen fuel cell vehicles may also include air vehicles, industrial trucks, trains, ships, aircraft, and devices that use hydrogen as fuel to generate electrical energy and use that electrical energy for propulsion.

[0077] Furthermore, the two-way communication process for hydrogen refueling disclosed herein can be applied not only to hydrogen fuel vehicles, but also to buildings or facilities that use hydrogen as an energy source.

[0078] Furthermore, in the following description, hydrogen fuel may include one or more of gaseous hydrogen or liquid hydrogen, and may substantially refer to compressed hydrogen, but is not limited thereto.

[0079] In addition, for ease of description, vehicles using the two-way communication process for hydrogen refueling are mainly described with reference to hydrogen electric vehicles, but are not limited to this configuration, and may include hybrid EV (electric vehicle) vehicles, ICE (internal combustion engine) vehicles, etc. that use hydrogen as fuel.

[0080] In the following description, some or all of the processes of the communication method, communication protocol negotiation method, hydrogen refueling (fuel supply, fuel refueling) protocol negotiation method, and hydrogen refueling (fuel supply, fuel refueling) parameter negotiation method performed in the hydrogen fuel vehicle may be executed by the electronic control unit, communication device, or communication control device in the hydrogen fuel vehicle.

[0081] In the following description, some or all of the processes of the communication method, communication protocol negotiation method, hydrogen refueling (fuel supply, refueling) protocol negotiation method, hydrogen refueling (fuel supply, refueling) parameter negotiation method, hydrogen refueling (fuel supply, refueling) method, and hydrogen refueling (fuel supply, refueling) control method performed in the distributor can be executed by the distributor's controller, electronic control unit, communication device, or communication control device. Furthermore, some processes of the above methods can be executed by the controller, electronic control unit, communication device, or communication control device of the refueling station associated with the distributor.

[0082] Furthermore, even technologies known prior to the filing date of this application may be included as part of the configuration of this disclosure when necessary, and such technologies are described in this specification to the extent that they do not obscure the spirit of this disclosure. However, in describing the configuration of this disclosure, detailed descriptions of matters known prior to the filing date of this application and which would be clearly understood by those skilled in the art may obscure the spirit of this disclosure, therefore excessive detailed descriptions of known technologies are omitted. For example, techniques for hydrogen charging control using thermodynamic models, techniques for generalized dynamic control using model predictive control (MPC) techniques, and techniques for configuring and controlling artificial neural networks for training and inference can use technologies known prior to the filing date of this disclosure, and one or more of these known technologies can be applied as the basic technologies necessary for implementing this disclosure.

[0083] However, this disclosure is not intended to claim protection for these known technologies, and the content of the known technologies may be included as part of this disclosure without departing from its scope.

[0084] Preferred exemplary embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0085] Figure 1 This is a conceptual diagram illustrating the hydrogen refueling and control process of a hydrogen fuel vehicle applied to an exemplary embodiment of this disclosure.

[0086] refer to Figure 1 Pre-cooled hydrogen is supplied from the hydrogen refueling station (station 200) to the hydrogen fueling vehicle / mobile (hydrogen fueling vehicle / vehicle 300) via distributor 100. At this point, the hydrogen refueling process can be described by parameters including the average pressure ramp rate (APRR).

[0087] In general, hydrogen storage systems attached to vehicles can be broadly categorized into high-pressure hydrogen tanks, high-pressure pipes for pressure control, and external frames. In the case of high-pressure hydrogen tanks, commercially available capacities ranging from tens to hundreds of liters have been achieved, and in the case of vehicles, multiple small and lightweight storage tanks are connected in parallel and used to achieve high capacity.

[0088] High-pressure hydrogen tanks are well known as compressed hydrogen storage systems (CHSS) 310, and in this specification, for ease of description, the term "tank" refers to CHSS 310, and for ease of description, the terms "tank" and CHSS 310 may be used interchangeably.

[0089] In a typical hydrogen storage system, hydrogen storage can be controlled by the boss unit that allows hydrogen to enter and leave the storage tank 310. Based on the characteristic that hydrogen injection and use cannot be carried out simultaneously, hydrogen storage is controlled by attaching valves, pressure reducing mechanisms, and sensors for various types of measurements to a boss unit.

[0090] The interface between the hydrogen refueling station 200 and the hydrogen fuel vehicle 300 is handled by the distributor 100. At the distributor 100, the target pressure, injection speed, etc. are controlled by comprehensively considering the vehicle storage tank information and the refueling information of the station 200. An example of the control logic currently used can be the control logic according to the SAE J2601 (2020-05) standard.

[0091] In related technologies, schemes for transmitting information from the hydrogen fuel cell vehicle 300 to the distributor 100 include communication schemes and non-communication schemes as non-communication methods. Even when communication is used, in related technologies, the temperature and pressure values ​​of the vehicle storage tank 310 are transmitted unidirectionally to the distributor 100, and the distributor 100 does not actively utilize the corresponding information, but only uses the corresponding information as a safety standard, such as emergency stop under extreme temperature and pressure conditions.

[0092] All refueling logic for safe and rapid refueling is managed by distributor 100, and vehicle storage tank 310 has only minimal safety management equipment for automatically discharging hydrogen under conditions such as overheating via pressure relief device (PRD) 320, without any active safety management scheme.

[0093] In order to address the references later Figure 2 The hydrogen refueling station 200, which describes the phenomenon of hydrogen temperature rise during refueling, includes a high-pressure hydrogen storage unit 220 and a precooler 210. The precooler 210 supplies hydrogen to the hydrogen electric vehicle 300 via a distributor 100 after the hydrogen temperature has been reduced by precooling.

[0094] Exemplary embodiments of this disclosure utilize a basic configuration similar to that of related technologies, but the hydrogen refueling controller 110 within the dispenser 100 actively controls the hydrogen refueling process using status information such as temperature and pressure received from the hydrogen fuel carrier 300 and station 200, as well as charging status information such as state of charge (SOC) from the CHSS 310.

[0095] According to an exemplary embodiment of this disclosure, a scheme that controls the filling speed in real time based on real-time temperature data of the vehicle storage tank 310 can operate at the highest filling speed without exceeding safety limits, thus reducing filling time within the available range.

[0096] The charging protocol of the related technology has the following problems: the boundary conditions for safety are over-set, thus performing excessive pre-cooling and supply, so that the temperature of most storage tanks 310 is measured at about 40 to 50°C at the time of charging completion.

[0097] According to an exemplary embodiment of this disclosure, the cooling load of station 200 can be optimized by actively controlling the demand and supply of precooling, thereby improving the operating efficiency of hydrogen refueling station 200.

[0098] The protocol for related technologies centered on lightweight hydrogen electric vehicles has the following problem: when refueling a new vehicle, all variables need to be reset and reflected in the standard.

[0099] According to exemplary embodiments of this disclosure, a control technique that uses ANN-based learnable augmentation logic to update the logic itself through a predetermined learning and training process when applying a new device can be used to achieve a wide range of expansion and application to various mobile body regions.

[0100] In related technologies, the only measure to prevent overheating of the storage tank of hydrogen electric vehicles is to release gas through a pressure relief device / PRD 320 when overheating occurs above a predetermined temperature.

[0101] According to an exemplary embodiment of this disclosure, the cooling system 330, described later, can be installed in the storage tank 310 itself, thereby increasing the charging speed and actively responding to overheating of the storage tank 310, thus improving the safety of the hydrogen fuel vehicle 300.

[0102] According to exemplary embodiments of this disclosure, the efficiency of the hydrogen refueling / supply process can be improved while safely refueling / supplying hydrogen fuel, and the speed and real-time performance of the hydrogen refueling / supply process can be increased.

[0103] According to exemplary embodiments of this disclosure, a hydrogen refueling control technology with guaranteed real-time performance based on model predictive control (MPC) can be provided.

[0104] According to exemplary embodiments of this disclosure, a control technique based on an artificial neural network (ANN) model can be provided to improve the accuracy of hydrogen refueling result predictions. Actual measurements are reflected in the ANN model, which utilizes both actual refueling data and theoretical simulation results, thus improving the accuracy of the predictions.

[0105] According to an exemplary embodiment of this disclosure, the efficiency of hydrogen refueling control can be improved by using an intelligent meta-system (IMS) to uniformly manage actual measurement data and status information predicted from the model.

[0106] As described above, in the hydrogen refueling process of the related technology, the distributor 100 is responsible for the control between the hydrogen fuel vehicle 300 and the hydrogen refueling station 200, and the distributor 100 is equipped with a protocol that is a method for injecting hydrogen into the hydrogen fuel vehicle 300 according to predetermined rules, thereby comprehensively performing control.

[0107] Examples of protocols installed in the distributor 100 may include protocols based on the international standard SAE-J2601 (2020-05), and such protocols are equally applicable to exemplary embodiments of the present disclosure to the extent that they are consistent with the purposes of the present disclosure.

[0108] For minimum safety requirements, simulations are performed for various situations using thermodynamic modeling, and a table-based single-injection method and a partially real-time correction method based on the MC formula are used by employing parameters derived from them.

[0109] The minimum safety requirements include the upper limits of temperature and pressure conditions for CHSS 310 and the guidelines for state of charge.

[0110] Simulations can be performed using thermodynamic modeling, with boundary conditions ranging from best to worst cases.

[0111] Such a configuration can also be applied to the configuration of exemplary embodiments of this disclosure within the scope of the purpose of this disclosure.

[0112] Even if you follow Figure 1 Regarding the configuration, in related techniques where the state values ​​are not actively controlled in the distributor 100, the following problems were found. These problems were also revealed in related techniques that rely on simulations using simple thermodynamic models.

[0113] The following problems exist: the injection rate is predetermined by assuming the worst-case boundary conditions (excessive boundary conditions), thus requiring unnecessary pre-cooling and reducing the overall charging rate. In related technologies, the injection rate is simply determined by the average pressure ramp rate (APRR), which can be a factor inhibiting proactive response based on the situation. Unnecessary pre-cooling can also lead to excessive energy and operating costs.

[0114] Because it relies on simulation-based results, the following problems exist: the applicable capacity or form of storage tank 310 is limited, and in the case of new systems, separate resources are required for new development and applications, thus limiting the applicable targets.

[0115] Thermodynamic models require a significant amount of time to derive the calculation results of mathematical expressions. Therefore, in schemes that indirectly utilize variables derived from the model, their application is limited when pre-calculated variables are not available, and the following problems exist: insufficient flexibility in detailed adjustments to the scheme itself.

[0116] The table-based approach of the related technology has the following problems: it is very inefficient and it is difficult to respond flexibly to changes in the surrounding environment because the temperature of the pre-cooled hydrogen provided from station 200 or the temperature of the storage tank 310 measured in the hydrogen fuel carrier 300 are not utilized.

[0117] The related technology uses the MC formula to correct the precooling temperature in real time, but it has the following problems: the calculation and application scheme is complex, and the applicable targets are limited, so it is difficult to expand.

[0118] The protocol was developed with the primary goal of ensuring safe completion of refilling, therefore there are no alternatives that can proactively control unforeseen circumstances, such as over-precooling or overheating of storage tank 310, which would lead to problems such as increased operating costs due to overcooling and refilling delays due to overheating.

[0119] The features of this disclosure are derived to solve problems in related technologies, and are characterized by reducing reliance on simulation by reflecting real-time measurement data and actively attempting to control state variables.

[0120] Figure 2 This is a conceptual diagram illustrating an example of the state changes that occur during the refueling process of a hydrogen fuel vehicle 300 applied to an exemplary embodiment of this disclosure.

[0121] refer to Figure 2 When hydrogen is injected into the hydrogen storage tank 310, the internal temperature rises due to the heat of compression, and thus the temperature of the hydrogen inside the storage tank 310 rises.

[0122] By receiving pre-cooled hydrogen, temperature control is achieved during the hydrogen refueling process, ensuring that the internal temperature of the storage tank 310 is controlled at 85°C or lower at the final refueling completion time.

[0123] The storage tank 310 is configured such that the carbon fiber surrounding the dome and body of the tank 310 has low heat transfer efficiency in order to block heat exchange between the outside air and the hydrogen stored therein during operation.

[0124] During the refueling process, as the hydrogen temperature inside the storage tank 310 rises, due to the low heat transfer characteristics of the storage tank 310, the temperature rise on the surface of the hydrogen storage tank 310 is weaker than the temperature rise inside the storage tank 310 until the refueling is completed.

[0125] This characteristic hinders heat exchange with the outside air that can mitigate the rapid temperature rise inside the hydrogen storage tank 310 during refilling, making such heat exchange almost impossible and thus requiring separate temperature management measures.

[0126] However, the relevant technology does not include any separate cooling means other than receiving pre-cooled hydrogen from station 200.

[0127] The time for complete hydrogen refueling is managed by controlling the pre-cooling and hydrogen injection rate in the hydrogen refueling station 200 in order to keep the temperature of the vehicle's hydrogen storage tank below 85°C, which is the upper limit of temperature management. However, temperature management measures for the storage tank 310 used for the hydrogen fuel vehicle 300 are not separate.

[0128] Therefore, especially in the summer when the outside air temperature is high, problems such as refilling delays can occur because it is difficult to control the temperature of the vehicle storage tank 310 in station 200.

[0129] In an exemplary embodiment of this disclosure, Figure 2 The characteristic curves are used as the basic model, but unlike related technologies, by taking into account real-time data of variables occurring in the actual surrounding environment (such as external air temperature, atmospheric pressure, weather conditions, etc.), the operating load of Station 200 in the pre-cooling stage of Stage I and the filling speed control (pressure ramp rate, PRR) that occurs during the filling process from Stage II to Stage IV are optimized, and the optimal control conditions suitable for the actual environment can be derived.

[0130] Figure 3 This is a conceptual diagram illustrating a platform or test platform that applies a hydrogen refueling process according to an exemplary embodiment of this disclosure.

[0131] refer to Figure 3The example embodiment shows that, in addition to the hydrogen fuel refueling controller 110, an artificial neural network model 120 is also provided in the dispenser 100.

[0132] When the artificial neural network model 120 predicts the next state value by executing a model predictive control (MPC) process, hydrogen refueling control via the ANN-MPC method can be achieved. Figure 3 In this embodiment, assuming the use of a refueling protocol that allows for ANN-MPC-based hydrogen refueling control, the ANN-MPC-based hydrogen refueling protocol is typically implemented in the dispenser 100 between station 200 and mobile vehicle 300. In alternative exemplary embodiments of this disclosure, the ANN-MPC-based hydrogen refueling protocol may be implemented in the hydrogen refueling controller 110 or the artificial neural network model 120. Furthermore, in alternative exemplary embodiments of this disclosure, the ANN-MPC-based hydrogen refueling protocol, the hydrogen refueling controller 110, or the artificial neural network model 120 may communicate electronically with the dispenser 100 even if not located within the dispenser 100, and may be implemented as part of a controller capable of influencing the hydrogen refueling process within the dispenser 100.

[0133] Refer again Figure 3 The artificial neural network model 120 can receive data from the station 200 and / or the mobile body 300 using real-time communication, and can use this data to send information to the hydrogen refueling controller 110 for controlling hydrogen refueling to achieve optimal hydrogen refueling.

[0134] At this point, the real-time communication data consists of static data and dynamic data. The static data consists of the storage tank's capacity, configuration, facility type, etc., which do not change over time, while the dynamic data can indicate data such as temperature or pressure that may change during the filling process.

[0135] Static data may include, for example, the volume, pressure rating, maximum permissible pressure, tank type, number of tanks, tank dimensions, tank serial number, tank manufacturer, tank utilization data, and permissible temperature range of hydrogen fluid in the tank for the mobile body 300 or station 200.

[0136] Dynamic data may include, for example, refueling commands (e.g., start, stop, pause, abandon, increase flow rate, decrease flow rate, tank change commands, etc.), control parameters used for refueling (e.g., average pressure ramp rate (APRR), pressure ramp rate, etc.), real-time measurements of pressure and / or temperature (including temperature fluctuations and / or pressure fluctuations) in the tank that can indicate leakage and / or impending failure, the remaining state of charge (SOC) of the tank on the station side or the moving body side, the ambient temperature outside the tank, and / or real-time measurements of the flow rate of hydrogen fluid flowing into the tank, etc.

[0137] Refer again Figure 3 The hydrogen refueling control method based on the ANN-MPC refueling protocol can use static data to set the basic conditions required for hydrogen refueling before starting hydrogen refueling, and can perform real-time control based on dynamic data during the hydrogen refueling process.

[0138] refer to Figure 3 It may also include an exception and problem handling module 400. The ANN-MPC model basically assumes real-time bidirectional communication and can exhibit optimal performance under conditions where real-time bidirectional communication is possible.

[0139] However, during the pre-refueling inspection process or during the actual refueling, various situations may occur that deviate from real-time communication conditions.

[0140] (1) No communication: The mobile unit 300 or station 200 does not provide communication function, the communication function is in a failed state, or communication becomes impossible during refueling due to communication function failure.

[0141] (2) Communication error: The device is able to communicate with the mobile body 300 or the station 200, but some data is lost or erroneous due to abnormal operation of components, sensors or equipment.

[0142] (3) Uncertified equipment, components or vehicles: those that can communicate with mobile body 300 or station 200 and whose data is collected normally, but whose data transmitted from mobile body 300 or equipment has low reliability, thus requiring verification or restriction of use.

[0143] (4) Uncertified communication protocol: A protocol that can communicate with mobile body 300 or station 200, but whose communication method is not defined in the standard or certified in the relevant country, and therefore whose security, reliability and stability are not guaranteed.

[0144] While ANN-MPC's real-time communication-based control may offer the best performance in terms of functionality, in order to apply ANN-MPC to real-world applications, measures are needed to effectively, progressively, and differentially address communication problems that may occur in various forms. In the exemplary embodiments described later, apparatuses capable of addressing these problems are proposed.

[0145] Figure 4 This is a conceptual diagram illustrating a model predictive control (MPC) process according to an exemplary embodiment of the present disclosure.

[0146] In the MPC control method, the accuracy of control increases as the accuracy of the prediction model improves. Figure 4 The control method shown is ANN-MPC, and an artificial neural network model 120 is used as the prediction model, and a feedback control loop including a hydrogen refueling controller 110 is used as the control scheme.

[0147] refer to Figure 4 The temperature, pressure and external air temperature of the mobile body 300 and station 200 are used as inputs, and the temperature and pressure of the storage tank 310 of the mobile body 300 are set as outputs.

[0148] although Figure 4 An exemplary embodiment is shown that predicts the temperature and pressure of the hydrogen carrier 300 as an output, but in an alternative exemplary embodiment of this disclosure, an exemplary embodiment that predicts the temperature and pressure of the storage tank of station 200 as an output can be implemented.

[0149] When a new input is given, an artificial neural network model 120, which has learned the correlation between input and output using simulation data based on a theoretical model and data measured in the actual field, can predict the output.

[0150] When a specified SOC value (target value) is given in controller 110, the Pressure Ramp Rate (PRR) can be predicted, which is used to achieve the refueling speed of the target value. When PRR_pd is given as a control parameter, artificial neural network model 120 can predict SOC_pd. SOC_pd replaces the initial target SOC_sp and is input to controller 110. At this time, artificial neural network model 120 can be trained such that the SOC_m on the side of the mobile vehicle 300 performing hydrogen refueling through the actual application of PRR_c is transmitted to artificial neural network model 120, and the prediction accuracy of SOC_pd is improved as an objective function or loss function.

[0151] Let's refer to each other. Figure 3 and Figure 4 Real-time temperature, pressure, supply rate (which may refer to PRR), and abnormal signals of station 200 can be transmitted to artificial neural network model 120. In an alternative exemplary embodiment of this disclosure, real-time temperature, pressure, supply rate (which may refer to PRR), and abnormal signals of station 200 can also be transmitted to hydrogen refueling controller 110.

[0152] The real-time temperature, pressure, and abnormal signals of the mobile body 300 can be transmitted to the artificial neural network model 120. In an alternative exemplary embodiment of this disclosure, the real-time temperature, pressure, and abnormal signals of the mobile body 300 can also be transmitted to the hydrogen refueling controller 110.

[0153] The hydrogen refueling controller 110 generates actual control commands, and the hydrogen refueling of the dispenser 100 is performed based on these control commands. Furthermore, a portion of these control commands, such as a cooling command, can also be transmitted to the mobile body 300.

[0154] At this time, commands or information related to the control of hydrogen refueling may include a pressure ramp rate for refueling hydrogen in the tank 310 of the mobile body 300, control commands for the pressure ramp rate, and / or status information on the distributor 100 side as a result of the execution of the control commands. Furthermore, information related to the control of hydrogen refueling may include information about variables that can affect changes in the hydrogen state in the vehicle tank 310 of the mobile body 300, such as the real-time pressure ramp rate or the mass flow rate (kg / s) [m_dot] of compressed hydrogen derived during feedback control, and / or information about the results of control execution based on these variables. Information related to the control of hydrogen refueling may also affect the weights or parameters of the hidden layers of the artificial neural network model 120.

[0155] A hydrogen refueling control request based on a hydrogen refueling protocol may include a control command for the pressure ramp rate used to refuel the vehicle tank 310 with hydrogen.

[0156] Information related to hydrogen refueling control may include one or more of the following: information on hydrogen refueling control requests or status information based on the execution results of hydrogen refueling control requests; and control requests for changing the state of hydrogen may include one or more of the following: control requests for the temperature or pressure of hydrogen in vehicle tank 310.

[0157] Changes in the state of hydrogen in vehicle tank 310 may include one or more of the following: temperature, pressure, or state of charge (SOC).

[0158] As a theoretical simulation, for example, a thermodynamic model such as H2FillS (Hydrogen Filling Simulation) can be used. Thermodynamic models may include models and / or software designed to track and report one or more of the transient changes in hydrogen temperature, pressure, or mass flow rate and / or the transient changes in the state of hydrogen in the vehicle's tank during hydrogen refueling of a hydrogen fuel cell vehicle / mobile. Of course, the spirit of this disclosure is not limited to exemplary implementations of specific thermodynamic models.

[0159] Hydrogen refueling protocols may include, for example, the hydrogen refueling protocols defined in SAE J2601.

[0160] For example, when input data corresponding to the input data of an artificial neural network is input, the thermodynamic model can generate output data based on modeling and simulation. In this case, the variables of the thermodynamic model can be adjusted according to the hydrogen charging protocol, and for the same input data, different hydrogen charging protocols can yield different output data.

[0161] In an exemplary embodiment of this disclosure, the field data collected by the test platform can be provided as input and output data of an artificial neural network, replacing the input / output data of a thermodynamic model, and the artificial neural network can be trained. That is, a portion of the field data collected by the test platform can be provided as input data of the artificial neural network, and another portion can be provided as ground truth data corresponding to the output data of the artificial neural network.

[0162] The intrinsic parameters of an artificial neural network can be trained without initialization, or they can be trained based on field data after being initialized to predetermined values. For example, the input and output data (ground truth data) of an artificial neural network can be given based on a thermodynamic model, and initial learning can be performed, thus initializing the intrinsic parameters of the artificial neural network.

[0163] Artificial neural networks do not necessarily need to be deep learning; they can be shallow learning as well.

[0164] The test platform according to exemplary embodiments of this disclosure can rely on dynamic field data to optimize the hydrogen refueling process.

[0165] Exemplary embodiments of this disclosure can predict the next state using model predictive control techniques based on artificial neural networks. In this case, the artificial neural network model 120 can be trained using theoretical results, field data, or both.

[0166] The artificial neural network model 120 may be a model trained to receive information related to hydrogen refueling control and information about the state of hydrogen in the vehicle tank 310 related to the execution result of the hydrogen refueling control request, and to predict changes in the state of hydrogen in the vehicle tank 310.

[0167] The artificial neural network model 120 may be a model trained to receive information related to hydrogen refueling control, the current state of hydrogen in the vehicle tank 310, the state of hydrogen supplied from the distributor 100 to the mobile body 300, and the ambient temperature, and to predict future changes in the state of hydrogen in the vehicle tank 310.

[0168] The artificial neural network model 120 can be a model trained to predict changes in the hydrogen state in the vehicle tank 310 based on target states of hydrogen state in the vehicle tank 310 according to information related to hydrogen refueling control and according to various hydrogen refueling protocols.

[0169] The artificial neural network model 120 can update the parameters in the model by receiving information related to hydrogen refueling control and hydrogen state information in the vehicle tank 310 and using field data of hydrogen state changes in the vehicle tank as ground truth data to train the function of predicting hydrogen state changes in the vehicle tank 310.

[0170] The artificial neural network model 120 can be a model trained to use model predictive control technology to predict changes in the state of hydrogen in the vehicle tank 310.

[0171] In cases where the pre-cooling function of station 200 or the cooling system 330 of mobile body 300 is unavailable, control of hydrogen refueling can be performed by reflecting these elements.

[0172] Big data is collected based on the type and individual ID of dispenser 100, the type and individual ID of station 200, the type and individual ID of mobile body 300, the control target state (temperature, pressure, SOC), the initial state (temperature, pressure), and the type of hydrogen refueling protocol. The test platform is trained using the dynamic changes of field data corresponding to each case, so that a standardized project that can optimize and accurately describe the hydrogen refueling process can be derived.

[0173] Exemplary embodiments of this disclosure can improve the reliability of the hydrogen refueling process by collecting and processing on-site hydrogen refueling data.

[0174] Field data on changes in the state of hydrogen in vehicle tank 310 can be obtained first on the mobile body 300 side. That is, field data on changes in the state of hydrogen in vehicle tank 310 can be obtained on the mobile body 300 side, regardless of whether a control request is transmitted to the hydrogen vehicle. Alternatively, in other exemplary embodiments of this disclosure, the control request may include a field data request, and the field data may be obtained on the mobile body 300 side in response to the control request / field data request.

[0175] Field data acquisition can also be performed at the station side in addition to the mobile body 300, and similarly, field data at the station side can be collected regardless of the control request, and the control request can include a field data request. Therefore, field data on the state of hydrogen stored in the station and / or hydrogen supplied from the station to the distributor can be obtained at the station side in response to the control request / field data request.

[0176] Field data refers to data obtained at station 200 and / or mobile vehicle 300 during an actual hydrogen refueling / supply process. In this context, field data may include data obtained in test environments, in addition to actual station and vehicle conditions.

[0177] Field data may include data obtained in a test environment consisting of: a test environment or a device capable of response (feedback) corresponding to the hydrogen vehicle side, a test environment or a device capable of response (feedback) corresponding to the hydrogen storage cylinders and distributors at the station, and / or a test environment or execution device in which module A corresponding to the distribution control system side may be installed.

[0178] At this point, the field data may include data obtained in the filling site or test environment where all devices consist of the three test environments or devices described above, or may include any one of them.

[0179] For example, field data can be included in... Figure 5 The data obtained in the test environment or refueling site, which consists of the simulation model associated with modules B150 and C160 shown.

[0180] At this point, a device capable of reacting (feedback) can mean a device that includes a database built based on actual field data and is capable of responding to requests from module A.

[0181] Field data can include static data and dynamic data, and can be classified into static data (type of vehicle tank 310, volume of vehicle tank 310, number of modules / groups of vehicle tank 310, etc.) and dynamic data (hydrogen temperature, pressure, etc. of vehicle tank 310).

[0182] Multiple hydrogen storage cylinders can be installed in station 200 and operate as a bank system. Real-time field information requested by distributor 100 from station 200 and fed back to distributor 100 may include temperature and pressure information for each bank.

[0183] Multiple groups can switch and connect to distributor 100 based on requests from distributor 100 and / or selections from station 200. At this time, the temperature and pressure information of each group included in the real-time field information requested by distributor 100 from station 200 and fed back to distributor 100 can affect the selection and / or switching of groups.

[0184] At this time, based on the temperature and pressure information of each group included in the real-time field information requested and fed back to the distributor 100 from the station 200, the state of one or more groups in the group system can be adjusted.

[0185] For example, as a preparation phase, one or more groups in the station 200 can be adjusted to have the temperature and pressure required for charging, based on the current status information about the groups. Such adjustments can be performed according to a request from the distributor 100, or according to the control logic of the station 200.

[0186] Figure 5 This is a conceptual diagram illustrating a platform or test platform used for simulation during hydrogen refueling according to an exemplary embodiment of this disclosure.

[0187] refer to Figure 5 The diagram shows a mobile body section 300a for simulating an actual mobile body 300, a station section 200a for simulating an actual station 200, and a distributor section 100a for simulating an actual distributor 100.

[0188] exist Figure 5 In this process, the mobile body 300a, the station 200a, and the distributor 100a can be integrated, connected, or competed with the simulation model and the data-based model based on real-time field dynamic data.

[0189] In order to analyze the process state changes between distributor 100 and station 200, a learning model based on field data between distributor 100 and mobile body 300 can be used as a reference for the interaction process between distributor unit 100a and mobile body unit 300a.

[0190] Conversely, in order to analyze the process state changes between distributor 100 and moving body 300, a learning model based on field data between distributor 100 and station 200 can serve as a reference for the interaction process between distributor unit 100a and station unit 200a.

[0191] In alternative exemplary embodiments of this disclosure, control Figure 5 The simulated integrated management system can be implemented in the form of cloud and / or remote server.

[0192] Furthermore, although exemplary embodiments centered on the artificial neural network model 120 are shown in this application specification, other exemplary embodiments of this disclosure are not necessarily limited to the artificial neural network model 120 or model predictive control techniques. In other exemplary embodiments of this disclosure, after the transmission of information related to hydrogen refueling control and / or feedback control based on the hydrogen refueling protocol, real-time field data is fed back as a response, and the information and / or feedback control related to hydrogen refueling control can be updated based on the real-time field data, or the next information and / or feedback control related to hydrogen refueling control can be generated. According to other exemplary embodiments of this disclosure, hydrogen refueling protocols independent of model predictive control or the artificial neural network model 120, as well as test methods and test platforms for testing hydrogen refueling systems, can also be proposed.

[0193] When the station part 200a functions as a simulation model, the moving part 300a can be used as a means of modulating and optimizing the temperature of CHSS 310.

[0194] When the moving part 300a functions as a simulation model, the station part 200a can be used as a means of stabilizing and controlling the pre-cooling temperature.

[0195] control Figure 5 The simulation-integrated management system can obtain the difference between the simulation results of the changes in the hydrogen state in the vehicle tank corresponding to information related to hydrogen refueling control and the field data of the changes in the hydrogen state in the vehicle tank.

[0196] control Figure 5 The simulation-integrated management system can update the model 120 for the hydrogen refueling process of hydrogen vehicles based on the differences between simulation results and field data, corresponding to information related to hydrogen refueling control.

[0197] control Figure 5 The simulation-integrated management system can use one or more thermodynamic models that track transient changes in the temperature, pressure, or mass flow rate of hydrogen in the vehicle tank to obtain simulation results for changes in the state of hydrogen in the vehicle tank in response to a control request.

[0198] control Figure 5The simulation-integrated management system can obtain simulation results of the changes in the state of hydrogen in the vehicle tank corresponding to the hydrogen refueling control request by inputting the predicted hydrogen refueling control request into the model used for the hydrogen refueling process and using model predictive control (MPC) technology.

[0199] After updating model 120, control Figure 5 The simulated integrated management system can replace the hydrogen fuel vehicle 300 side with model 120, and can obtain one or more of the simulated data or field data on the change of state of hydrogen supplied from station 200 to distributor 100 in response to a second control request, between distributor 100 supplying hydrogen to hydrogen fuel vehicle 300 and station 200 supplying hydrogen to distributor 100.

[0200] After updating model 120, control Figure 5 The simulated integrated management system can replace station 200 with model 120, and can obtain one or more of the following data, including simulation data and field data, between the hydrogen fuel vehicle 300 receiving hydrogen from distributor 100 and distributor 100, regarding the change of hydrogen state in the vehicle tank of hydrogen fuel vehicle 300 corresponding to the third control request.

[0201] The hydrogen refueling process or hydrogen charging control technology in related technologies has the following problems: the final SOC and final temperature / pressure of the nozzle and CHSS 310 are difficult to control as expected. In addition, due to pressure variability, unstable flow rate and high environmental variability, hydrogen charging technology based on theoretical simulation has the following problems: the hydrogen charging technology based on theoretical simulation does not match the actual field data.

[0202] The mismatch between simulation results and actual field data is easily affected by equipment characteristics and environmental diversity that are difficult to fully account for in theoretical simulations. Furthermore, even when using the same hydrogen refueling protocol, the final field data may vary depending on the initial or final target values. Conversely, even assuming the same final target or initial values, the final field data may vary depending on different hydrogen refueling protocols.

[0203] To address these issues in the related technologies, exemplary embodiments of this disclosure may employ the use of real-time field data, bidirectional communication between devices, predictive control techniques, integrated control of the entire system including stations and vehicles, and the use and standardization of hydrogen refueling data based on user requests.

[0204] Furthermore, exemplary embodiments of this disclosure can enable the enhancement of existing refueling protocols, standardization of field refueling data formats, and the compilation and diagnostic of dynamic field data.

[0205] Regarding enhancements to existing refueling protocols, exemplary implementations of this disclosure may include the following.

[0206] Existing hydrogen refueling protocols can be selected as the test target.

[0207] Exemplary embodiments of this disclosure can perform artificial neural network-based model predictive control (ANN-MPC) and execute existing hydrogen refueling protocols under the same refueling conditions. Furthermore, existing hydrogen refueling protocols can be enhanced or improved by comparing the refueling control results of ANN-MPC with those of existing hydrogen refueling protocols.

[0208] The monitoring system 130 may include a protocol 131 as a test target, and may request additional control based on the included protocol, and may compare the control value with the “predicted output” input from the ANN-MPC.

[0209] The monitoring system 130 or controller can transmit information related to hydrogen refueling control to the hydrogen fuel vehicle 300 via communication interfaces 140 and 160 by executing existing refueling protocols, and can obtain on-site data on changes in the state of hydrogen in the vehicle tank via communication interfaces 140 and 160.

[0210] The monitoring system 130 or controller can obtain a hydrogen refueling control sequence based on the model 120 for the hydrogen refueling process (in the case of using ANN-MPC technology, future time-series control sequence inputs can be predicted), and can enhance existing hydrogen refueling protocols based on the hydrogen refueling control sequence based on the model 120 for the hydrogen refueling process.

[0211] The monitoring system 130 or controller can obtain the result data of the execution of the hydrogen refueling control sequence according to the model used for the hydrogen refueling process, and can use the hydrogen refueling control sequence to enhance a part of the existing hydrogen refueling protocol based on the comparison results between the result data of the execution of the hydrogen refueling control sequence and the field data according to the existing hydrogen refueling protocol.

[0212] At this point, the refueling control results using the artificial neural network model 120 can be obtained by operating the model 120 or from a previously constructed database. The basic standards used in the refueling protocol 131, which serves as the test target, remain unchanged, and the details, i.e., the refueling table or logic, can be enhanced / improved by referring to the refueling control values ​​of the artificial neural network model 120. At this point, by comparing the field data obtained using the refueling control results of the artificial neural network model 120 with the execution results of the refueling protocol 131, which serves as the test target, the details of the refueling protocol 131, which serves as the test target, can be partially improved when the performance of the refueling control results using the artificial neural network model 120 is superior. The comparison between the refueling control results and the execution results of the refueling protocol 131, which serves as the test target, can also be performed for all or part of the hydrogen refueling process.

[0213] In an exemplary embodiment of this disclosure, the lumped thermodynamic model of the artificial neural network for distributor-vehicle interaction can be used by taking into account the following characteristics.

[0214] 0-dimensional unsteady mass and energy balance

[0215] One-dimensional heat transfer in vehicle tank walls

[0216] CoolProp for hydrogen characterization

[0217] Exemplary embodiments of this disclosure can perform comparative analysis between theoretical simulation results and real-world field data.

[0218] Exemplary embodiments of this disclosure can perform predictive analysis under specific conditions, in addition to the conditions assumed in the hydrogen refueling protocol.

[0219] Figure 6 This is a conceptual diagram illustrating the concept of using an artificial neural network to model, predict, or control the hydrogen refueling process in an exemplary embodiment of the present disclosure.

[0220] refer to Figure 6 The measured value of the current state is input to the input layer.

[0221] At this time, the ambient temperature Tamb, the precooling gas temperature Tpre, and the precooling gas pressure Ppre can be measured at the nozzle of the distributor 100 or station 200.

[0222] The hydrogen temperature TCHSS and hydrogen pressure PCHSS are values ​​measured on the CHSS 310 side of the hydrogen fuel cell 300, and the actual measured values ​​can be input into the input layer.

[0223] During the training of an artificial neural network, the current measured value can be transmitted to the input layer, and the next measured value can be transmitted to the output layer and used as ground truth data during the learning process of the artificial neural network. In this case, the learning process of the artificial neural network can be a process of learning to predict the next measured value of the output layer based on a combination of input measurements. The correlation between the data input to the input layer and the data provided to the output layer is learned, and through this learned correlation, it is possible to make predictions using data from a real dynamic annotation process along with theoretical results.

[0224] During the inference or output process using an artificial neural network, the actual field measurement values ​​are transmitted to the input layer, and the predicted values ​​for the next measurement can be obtained as the output of the operation of the artificial neural network.

[0225] The learning process of the artificial neural network used in the exemplary embodiments of this disclosure can be shallow learning or deep learning, and as an artificial neural network, a neural network type that conforms to the purpose of this disclosure can be adopted from known neural networks.

[0226] The values ​​input through the input layer are transmitted to the output layer while undergoing weight-based operations in the hidden layers.

[0227] The state values ​​output by the output layer (predictions for the next state) can be used to calculate charged state variables, such as the state of charge (SOC), using at least a portion of the thermodynamic model.

[0228] In exemplary embodiments of this disclosure, hybrid control in which theoretical simulation models and artificial neural networks (ANNs) are coupled is also possible, so that predetermined performance can be achieved even by learning with a small amount of data, and performance that meets the purposes of this disclosure can be derived even with a lightweight artificial neural network.

[0229] The real-time pressure ramp rate (PRR) or the mass flow rate of compressed hydrogen (kg / s) [m_dot] derived during feedback control can affect the weights or parameters of the hidden layers in an artificial neural network.

[0230] The hydrogen refueling technology based on artificial neural networks disclosed herein can improve the accuracy of refueling result prediction through models. Actual refueling data can be used together with theoretical simulation results, thus reflecting real-time measurements and further improving the accuracy of prediction results.

[0231] The control protocols of related technologies predict results by calculating the results of simulations suitable for individual cases, while the exemplary implementation of this disclosure differs in that it uses a process of improving accuracy through repeated training for various cases.

[0232] Based on these differences, in the exemplary embodiments of this disclosure, as various theoretical values ​​and empirical results are accumulated, the accuracy is gradually improved through updates, and even if a new filling process is introduced, such as a new configuration using storage tank 310 or a change in flow rate, the functionality in the corresponding model can be updated by additional training with actual data, thus enabling wide application to various mobile body domains.

[0233] As an exemplary embodiment of the hydrogen refueling control technology for hydrogen refueling testing of a hydrogen fuel vehicle 300 according to the present disclosure, model predictive control (MPC) can be used.

[0234] Figure 7 This is a conceptual diagram illustrating the concept of using model predictive control to control the hydrogen refueling process to achieve a target output in an exemplary embodiment of the present disclosure.

[0235] refer to Figure 7 Model predictive control (MPC) is a control scheme that uses a model of the simulated process to predict the output based on future inputs and then optimizes the output to derive the control input. To comply with various boundary conditions during the control process, the process response must exist within a predetermined range, and compared to other control schemes, MPC can effectively derive the process response within this predetermined range.

[0236] MPC plans and predicts the time-domain input to obtain a process response close to the target output. It then obtains the response based on this prediction model, using only the control input as the control input. When an output is obtained from the controlled object, the same process is repeated based on that output.

[0237] In an exemplary embodiment of this disclosure, with the accuracy of the hydrogen charging model ensured to a reasonably high level, future charging results are predicted from the hydrogen charging model and current measurements, and the pressure ramp rate (PRR) can be controlled in real time based on the predicted values ​​and charging values, so that specific variables (e.g., hydrogen temperature TCHSSS or hydrogen pressure PCHSS of CHSS 310) achieve optimal charging targets without violating constraints.

[0238] refer to Figure 7 In an exemplary embodiment of this disclosure, future output values ​​can be calculated using MPC-based control based on current measurements and model predictions, and operating parameters / variables can be adjusted so that the predicted future response moves to the setpoint or target in an optimal manner.

[0239] For example, n model predictions can be derived at the current time i. The n model-based predictions can form the prediction time domain.

[0240] Each model-based prediction, i.e., the prediction time domain, corresponds to the control time domain. That is, the n control commands / actions required for n model predictions can form the control time domain.

[0241] In fact, among the n model predictions and control actions derived at the current time i, the (i+1)th control action, which is the first control action, can be transmitted to the system. As time passes, new n model predictions and control actions are derived again at the current time i+1, and these new n model predictions and control actions form new prediction time domains and new control time domains, respectively.

[0242] The technique of controlling a system in this way while extending / moving the time domain is called MPC, and in an exemplary embodiment of this disclosure, MPC-based control can be performed using measured and predicted values ​​of the state information (state values) of hydrogen temperature and pressure, including CHSS 310.

[0243] Figure 8 This is a flowchart illustrating the training and inference processes of an artificial neural network when used for model predictive control during hydrogen refueling according to an exemplary embodiment of the present disclosure.

[0244] refer to Figure 8 The diagram illustrates the training process of an artificial neural network considering an artificial neural network-model predictive control technique according to an exemplary embodiment of the present disclosure.

[0245] Figure 8 It is assumed that an artificial neural network has been learned to acquire the functions of prediction time domain and control time domain based on MPC, and in particular, an artificial neural network has been learned to acquire n future predictions and corresponding control commands so that the future response reaches the set point through MPC optimization.

[0246] refer to Figures 6 to 8 In an exemplary embodiment of this disclosure, the control system can be configured based on the ANN model 120, and a test platform system for real-time control based on model predictive control can be configured by ensuring the accuracy of the ANN model 120.

[0247] The test platform system for real-time control predicts future charging results by comparing future charging results with actual measurements and controls charging speed / pressure ramp / pressure increase rate. Limits, control time intervals, sensitivity, etc. can be set individually in the system logic, and control can be executed within the optimal value.

[0248] Mainly, optimal control is performed based on real-time data from station 200 and hydrogen fuel vehicle 300. However, when specific events occur during the operation of the respective systems, the system can provide the ability to directly control the pre-cooling temperature of pre-cooler 210 and the cooling system of hydrogen fuel vehicle 300, thereby improving the overall efficiency of the hydrogen refueling process.

[0249] refer to Figure 8 The control process begins upon receiving a specified SOCsp from the customer (t=0, step S710). For example, the current SOC could be 50%, and the SOCsp could be 85%.

[0250] SOC(t) is given as a function of TCHSS(t) and PCHSS(t), and the process can be performed based on a general dynamic model.

[0251] If the current SOC(t) is greater than or equal to SOCsp (step S720), hydrogen charging can be stopped. If the current SOC(t) is less than SOCsp (step S720), i=t is set, and the moving-time domain prediction of the accompanying artificial neural network is performed (step S730).

[0252] Step S730 can be performed by generating MPC predictions using an artificial neural network 120, etc. In step S740, it can be determined whether the obtained n predictions are the best predictions that meet the expected purpose.

[0253] Based on the case where the obtained n predictions are the best predictions, a control command PRR(t) is obtained based on the n predictions and the control command, and PRR(t) can be applied to the distributor 100-vehicle tank 310 (step S750).

[0254] Subsequently, time t increases, and new measurements TCHSS(t) and PCHSS(t) are obtained and transmitted to the input of step S720.

[0255] If the n predictions obtained in step S730 are not the best predictions, step S730 can be executed again to obtain new n predictions and control commands.

[0256] exist Figure 8 In step S730, for all arbitrary i and k, state prediction values ​​(T, P) that satisfy the temperature limit and pressure limit can be generated.

[0257] Based on the current time i (= t), n state prediction values ​​and corresponding control commands can be derived.

[0258] Figure 8Step S740 can be understood as the process of searching for a set of n predictions that minimize the cost function indicating whether the final control objective SOCsp has been achieved.

[0259] The temperature and pressure state measurements, including those of CHSS 310, can be provided as feedback inputs to the artificial neural network model 120 as the output of the hydrogen fuel cell 300.

[0260] State measurements, including the temperature and pressure of the pre-cooled hydrogen, can be provided as feedback input to the artificial neural network model 120 as the output of the hydrogen refueling station 200.

[0261] The artificial neural network model 120 transmits the predicted output to the hydrogen refueling controller 110, and the supervisory system 130 can input future inputs obtained through simulation or model-based predictions into the artificial neural network model 120 via module A 140.

[0262] ANN-MPC-based control is a control technique that utilizes both simulation and actual measurement data, and at least partially uses an artificial neural network model 120 to perform the simulation and utilize its predictive results during the control process.

[0263] The exemplary embodiments of this disclosure are intended to configure a hydrogen refueling integrated control protocol based on real-time data, and the corresponding systems are implemented using various underlying technologies.

[0264] The protocol installed in the distributor 100 can use data on precooled hydrogen provided by the slave station 200 and data on storage tank 310 provided by the hydrogen fuel carrier 300 as real-time input values, and can control the charging speed / pressure ramp rate / pressure increase rate (PRR or [m_dot]) as outputs through the installed model.

[0265] Depending on events such as changes in the external environment, overall control of charging speed / pressure ramp rate / pressure increase rate (PRR or [m_dot]), process efficiency, etc., can be performed by directly controlling the pre-cooling temperature of the control station 200 and the cooling system of the hydrogen fuel vehicle 300.

[0266] To complement the control protocol, the cooling stabilization system can be installed independently in the precooling system / precooler 210 of the hydrogen refueling station 200.

[0267] Regarding temperature stabilization, the cooling stabilization system of the precooler 210 can be controlled independently, and the target value of the control can be changed as a whole in the protocol of the distributor 100.

[0268] To improve the economic efficiency of station 200 and supplement the functionality of the integrated control protocol, additional functions related to temperature stabilization can be provided to precooler 210.

[0269] The precooling temperature varies depending on the initial temperature and flow rate of the hydrogen supplied to the precooler 210, and in order to compensate for this, as an exemplary embodiment of this disclosure, a novel precooler structure for stabilizing the temperature is proposed.

[0270] The precooler 210 according to an exemplary embodiment of this disclosure may include control logic for self-temperature control and protocol linkage.

[0271] A forced cooling system can be installed in the storage tank 310 of the hydrogen fuel vehicle 300, and can increase the charging speed by partially cooling the heat of compression generated during hydrogen charging, and the operation / control of the forced cooling system of the storage tank 310 can also be subject to an agreement.

[0272] In an exemplary embodiment of this disclosure, temperature management functionality may be provided to the storage tank 310 of the hydrogen fuel vehicle 300 to improve hydrogen refueling speed and complement the functionality of the integrated control protocol.

[0273] In an exemplary embodiment of this disclosure, the vehicle storage tank 310 is configured with a self-cooling system to improve the overall refueling speed and enhance the safety of the hydrogen fuel vehicle 300, and may include self-driving and protocol-linked control logic for the corresponding system.

[0274] In an exemplary embodiment of this disclosure, such integrated control not only improves current refill efficiency but also prepares for smooth support of the next refill.

[0275] When the precooling temperature of the precooler 210 is set to T40 at -40°C, if the precooling temperature reaches the target value but the outside air temperature is higher than the set value and the temperature rise on the storage tank 310 side is greater than expected, the control signal or information of the current state can be transmitted to the hydrogen fuel vehicle 300 / storage tank 310 side, so that the self-cooling system of the storage tank 310 can be activated.

[0276] Conversely, although the precooling temperature of the precooler 210 is set to -40°C, the target value of the precooling temperature (e.g., -35°C) can be adjusted when overcooling is determined by taking into account the external environment and actual data.

[0277] When additional control of pre-cooling target temperature and temperature on the storage tank 310 side is required, control information or control commands can be transmitted from distributor 100 to both hydrogen fuel vehicle 300 and station 200.

[0278] According to an exemplary embodiment of this disclosure, the self-cooling systems of the hydrogen fuel cell 300 and station 200 can be controlled independently or by transmitting signals from the distributor 100.

[0279] An integrated control method for hydrogen charging according to an exemplary embodiment of this disclosure may further include evaluating whether a measurement of the current state satisfies a constraint.

[0280] Constraints can indicate that the temperature and pressure of the compressed hydrogen storage system (CHSS) on the hydrogen vehicle side do not exceed the limit temperature and limit pressure, respectively.

[0281] According to exemplary embodiments of this disclosure, the efficiency of the hydrogen refueling / supply process can be improved while safely refueling / supplying hydrogen fuel, and the speed and real-time performance of the hydrogen refueling / supply process can be increased.

[0282] According to exemplary embodiments of this disclosure, a test method and test platform capable of accurately modeling the hydrogen charging / supply process based on real-time field dynamic data can be realized.

[0283] According to exemplary embodiments of this disclosure, a test method and test platform can be implemented to provide a model capable of precisely controlling the hydrogen charging / supply process, taking into account the differences between modeling and simulation results based on theoretical models and real-time field data, or taking into account both modeling and simulation results and real-time field data.

[0284] According to exemplary embodiments of this disclosure, a test technique for hydrogen charging control with guaranteed real-time performance based on model predictive control (MPC) can be implemented.

[0285] According to exemplary embodiments of this disclosure, a test technique based on an artificial neural network (ANN) model can be implemented to improve the accuracy of hydrogen charging result prediction.

[0286] By receiving a directly set charge amount from the user (e.g., SOC), it is possible to base this on the use of, for example... Figures 6 to 8 The sequence of real-time communication and computation is controlled by the control logic to charge / supply hydrogen until the target amount is reached.

[0287] Various variables can be used to set the charge amount, such as SOC, target pressure, time, temperature, etc., and the charge rate can be controlled to improve efficiency.

[0288] Figure 9 It is a framework (hereinafter referred to as the "hydrogen refueling framework") for executing a series of hydrogen refueling procedures using a bidirectional hydrogen refueling communication process according to an exemplary embodiment of the present disclosure.

[0289] refer to Figure 9 The hydrogen refueling framework may include functional blocks as various use cases: discovery and pairing functional block (hereinafter referred to as "UC1" or "UC-1"), communication security functional block (UC2 or UC-2), communication protocol negotiation functional block (UC3 or UC-3), refueling protocol negotiation functional block (UC4 or UC-4), refueling parameter negotiation functional block (UC5 or UC-5), security check-in functional block (UC6 or UC-6), monitoring and control functional block (UC7 or UC-7), security check-out functional block (UC8 or UC-8), termination functional block (UC9 or UC-9), error handling functional block (UC10 or UC-10), and emergency handling functional block (UC11 or UC-11).

[0290] Each of the UC1 to UC11 function blocks can correspond to, for example: Figure 9 The time series phases shown are (steps S401 to S411). At this time, Figure 9 It can also be understood as a flowchart that includes time series stages (steps S401 to S411).

[0291] UC10 and UC11 can each be connected to UC3 through UC8 and are configured to perform error handling and / or emergency handling in each use case.

[0292] The above use cases are functional blocks that consistently provide a complete hydrogen refueling procedure for a hydrogen refueling system, facilitating safe and secure refueling communication. Vehicles and dispensers can execute each use case sequentially in a specific order to achieve the hydrogen refueling objective.

[0293] Furthermore, after the dispenser nozzle is connected to the vehicle receiver, the corresponding vehicle and dispenser can be connected via a button. Figure 9 The sequence shown implements each use case to perform refueling communication. However, vehicles and dispensers can omit specific use cases if necessary, based on predefined requirements.

[0294] Each of the above use cases can be achieved through communication between the distributor control system of the distributor, which supplies hydrogen fuel according to the refueling protocol for hydrogen fuel vehicles, and the hydrogen fuel vehicle.

[0295] Simultaneously, the hydrogen fuel cell vehicle (hereinafter also referred to as the "vehicle") and dispenser implementing the above use cases can perform data exchange for vehicle identification within the UC-1. For this purpose, the vehicle may include sensors, an electronic control unit, a transmitter, and a receiver. In the case of bidirectional communication, the receiver can be integrally coupled to the transmitter.

[0296] Furthermore, the dispenser can be configured to receive specific data from the vehicle. The dispenser can store data recorded in a data log, or it can store data specified in the station's programmable logic controller (PLC) for the refueling protocol. Data logging can indicate a process of collecting data within a predetermined time period to analyze specific operational states of the hydrogen refueling system or to record data-based events / operations of the system or network environment, or it can indicate the data collected through such a process. In the case of bidirectional communication, the station can include sensors specified in the refueling protocol, and the station PLC or electronic control unit can obtain measurements from the sensors and transmit these measurements to the vehicle. The aforementioned vehicle or station can communicate using existing communication protocol standards, such as infrared communication, Wi-Fi, Bluetooth, etc.

[0297] Furthermore, a communication channel can be established between the vehicle and the distributor at the vehicle-distributor interface, providing a physical connection. The pairing process for establishing the communication channel can be performed using wired, optical, or wireless technologies.

[0298] The discovery and pairing procedure or pairing processor may have the prerequisite that the dispenser nozzle is inserted into the vehicle refueling receiver and securely connected to it. The vehicle refueling receiver may be simply referred to as the vehicle receiver or receiver.

[0299] Furthermore, the vehicle and dispenser are essentially aware of which communication protocol to follow. Therefore, communication after UC-1 can rely solely on the communication protocol agreed upon in the current use case as a postcondition for the discovery and pairing procedure or pairing process. If the vehicle or dispenser selects a communication protocol outside the agreed scope, the selected communication will not be executed. That is, even if the pairing process is successfully completed, approval for fuel or refueling may not be granted.

[0300] All methods used for pairing vehicles and distributors are configured not to increase the risk of ignition or explosion beyond permissible levels. For example, all wired pairing methods are configured to mitigate or block the risk of sparks due to electrostatic discharge.

[0301] Regarding the effectiveness of physical pairing, all methods for pairing vehicles and dispensers can be integrated into the vehicle-dispenser interface, or can be installed to ensure proximity between the vehicle's refueling receiver and the dispenser's nozzle and hose assembly. Here, the interface can indicate physical integration into the nozzle and receiver interface. Furthermore, proximity can be defined by the hardware associated with the pairing method. For example, the physical geometry for infrared communication can be specified, including the permissible distance between the transmitter and receiver. Additionally, the physical shape of the hydrogen refueling hardware can be pre-specified, and in this case, proximity does not include pairing methods that risk pairing physically unconnected vehicles and dispensers, such as relatively long-range wireless communication technologies like Bluetooth. Infrared communication can be referred to as Infrared Data Association (IrDA) communication and can include bidirectional IrDA communication.

[0302] In steps S401, S403, S404 and / or S405, information regarding the interoperability and compatibility between the mobile body and the distributor can be shared and mutually checked.

[0303] In steps S401, S403, S404 and / or S405, it can be confirmed whether the mobile body and the distributor each support bidirectional communication.

[0304] In steps S401, S403, S404, and / or S405, the communication protocol associated with the refueling protocol can be negotiated, and the refueling parameters according to the refueling protocol can be negotiated. At this time, in step S401, information about the interoperability and compatibility of the communication protocol and / or the refueling protocol can be shared in advance, and information about preferences or priorities on the mobile body or dispenser side can be shared.

[0305] Based on the information about interoperability, compatibility, preferences, or priorities shared in step S401, the communication protocol, refueling protocol, and refueling parameters can be negotiated in steps S403, S404, and / or S405.

[0306] At this point, if there are discrepancies between the information shared in step S401 and the information confirmed in steps S403, S404 and / or S405 (e.g., changes in the support level of bidirectional communication), the communication protocol, refueling protocol and refueling parameters can be negotiated or renegotiated based on the interoperability, compatibility, preference or priority updated in each of steps S403, S404 and / or S405.

[0307] Figures 1 to 8Whether the artificial neural network model described in the document is permitted to control or provide control support / assistance to the hydrogen refueling process can be determined through the refueling agreement, and whether the artificial neural network model is permitted to control or provide control support / assistance to the hydrogen refueling process can be determined through the refueling agreement. Figure 9 The steps are determined in at least one or more of S401, S403, S404 or S405.

[0308] Figures 1 to 8 Whether the permission for model predictive control to control or support / assist the hydrogen refueling process, as described in the document, is permitted can be determined through the refueling agreement, and whether the permission for model predictive control to control or support / assist the hydrogen refueling process is permitted can be determined through the refueling agreement. Figure 9 The steps are determined in at least one or more of S401, S403, S404 and S405.

[0309] The levels of communication supported between the mobile unit and the distributor can be represented by Tables 1 and 2 below.

[0310] Table 1

[0311] Table 1 provides examples of communication reliability levels differentiated from 0 to 3, based on various possible scenarios such as communication unavailability, failure, and lack of data security / reliability. The communication levels in Table 1 may refer to those included in the currently developing ISO 19885-1 standard.

[0312] Table 2

[0313] Table 2 shows the control levels for specifying static and dynamic variables based on the communication reliability level and permissible usage range, and for specifying alternative variables for unreliable data. For lower communication levels, some functions are limited, but the operation of the protocol annotation is possible according to each communication level.

[0314] The hydrogen refueling protocol based on ANN-MPC proposed in the exemplary embodiments of this disclosure corresponds to the situation corresponding to level 3 as defined in Tables 1 and 2, and it can be assumed that all static and dynamic data transmitted to and received from mobile body 300 and station 200 are reliable and readily available.

[0315] Figure 10 This is a flowchart illustrating a hydrogen fuel refueling method according to an exemplary embodiment of the present disclosure.

[0316] Figure 11 This shows the implementation Figure 10 A conceptual diagram illustrating the training and application processes of an artificial neural network model according to an exemplary implementation.

[0317] Let's refer to each other. Figure 10 and Figure 11 According to an exemplary embodiment of this disclosure, a method for supplying hydrogen fuel in a dispenser for supplying hydrogen fuel to a mobile vehicle using hydrogen as fuel may include: acquiring first refueling data during a first hydrogen refueling process, the first hydrogen refueling process being performed based on a first refueling protocol (step S1200); training a control model to learn control functions for the hydrogen refueling process based on the first refueling data (step S1400); and performing a second hydrogen refueling process based on one or more of the first refueling protocol or the control model (step S1600).

[0318] After training, the control model can assist in controlling the hydrogen refueling process according to the first refueling protocol. The first refueling protocol and the control model can operate in parallel, or the control model can operate by replacing some or all of the functions of the first refueling protocol.

[0319] The first betting agreement may be an existing betting agreement based on relevant technologies, but the spirit of this disclosure is not limited thereto.

[0320] The first hydrogen refueling process is performed in accordance with the existing refueling protocol, and the first refueling data can be generated as a result of the process (step S1200).

[0321] All data on the changes in temperature, pressure, and state of charge (SOC) at the distributor, mobile unit, and station side from the beginning to the end of the first hydrogen refueling process can be generated as the first refueling data.

[0322] The first refueling data is provided for the learning and training of the control model, and the hydrogen refueling process performed after the training of the control model can be referred to as the second hydrogen refueling process for ease of description.

[0323] Steps S1100 and S1600 can be executed by the controller 110 of the distributor 100.

[0324] In step S1200, the first refueling data can be generated on the controller side of the controller 110, the controller of another part of the dispenser 100, the mobile body 300, and the hydrogen refueling station 200, and can be obtained through the control model disclosed in the exemplary embodiments of this disclosure. At this time, the control model can be set in the dispenser 100 and can operate independently, but it can also be operated by connecting to and cooperating with an external server (including the cloud) via a network.

[0325] Step S1400 can be executed by the calculation system control model of the distributor 100 (see description below). Figure 12 (Configuration).

[0326] The control model can be Figure 3 and / or Figure 5 The artificial neural network model 120 shown is not limited to this.

[0327] In the step of performing the second hydrogen refueling process (step S1600), the second hydrogen refueling process can be performed by operating the first refueling protocol and control model in parallel.

[0328] In the step of performing the second hydrogen refueling process (step S1600), the second hydrogen refueling process can be performed by using a control model to replace at least a portion of the functions of the first refueling protocol.

[0329] The method for supplying hydrogen fuel to a hydrogen-fueled mobile body according to an exemplary embodiment of the present disclosure may further include verifying a control model.

[0330] At this point, the steps of performing the second hydrogen refueling process can be performed by replacing at least a portion of the functions of the first refueling protocol with a proven control model.

[0331] The steps of performing the second hydrogen refueling process can be performed by using a control model to assist at least a portion of the functions of the first refueling protocol.

[0332] In a method for supplying hydrogen fuel to a hydrogen-fueled mobile body according to an exemplary embodiment of the present disclosure, the control model may be a model that has learned the functionality of model predictive control (MPC).

[0333] In a method for supplying hydrogen fuel to a hydrogen-fueled mobile body according to an exemplary embodiment of the present disclosure, the control model may be an artificial neural network model that has learned the control functions of the hydrogen refueling process.

[0334] The method for supplying hydrogen fuel to a hydrogen fuel vehicle according to an exemplary embodiment of the present disclosure may further include a process of determining whether a first refueling protocol supports the generation of control information through model predictive control.

[0335] At this point, the process can be used as Figure 9 It is performed by a portion of one or more of S401, S403, S404 or S405 shown in the diagram.

[0336] The method for supplying hydrogen fuel to a hydrogen fuel vehicle according to an exemplary embodiment of the present disclosure may further include determining whether a communication protocol associated with a first refueling protocol supports bidirectional communication between the vehicle and a dispenser supplying hydrogen fuel to the vehicle.

[0337] At this point, the process can be used as Figure 9 It is performed by a portion of one or more of S401, S403, S404 or S405 shown in the diagram.

[0338] The method for supplying hydrogen fuel to a hydrogen fuel vehicle according to an exemplary embodiment of the present disclosure may further include determining whether the vehicle and the distributor supplying hydrogen fuel to the vehicle support bidirectional communication between the vehicle and the distributor.

[0339] At this point, the process can be used as Figure 9 The process is performed as a part of one or more of S401, S403, S404, or S405 shown in the diagram. Furthermore, as an exception, if the initially determined support for bidirectional communication changes due to a change in the communication environment, a new communication environment may be determined as part of steps S410 and / or S411, and the support for bidirectional communication may be re-determined.

[0340] refer to Figure 11 It can provide a dedicated artificial neural network model 120 for each detailed specification related to hydrogen refueling stations and / or mobile vehicles.

[0341] Furthermore, based on the detailed specifications related to hydrogen refueling stations and mobile vehicles, the first refueling protocol supported between the dispenser and the mobile vehicle can be determined.

[0342] The first refueling protocol is transmitted to the controller 110, and the controller 110 can perform the first hydrogen fuel refueling process (step S1100) based on the first refueling protocol.

[0343] The first hydrogen refueling data obtained during the first hydrogen refueling process (step S1200) can be provided to a dedicated artificial neural network model 120 for learning / training.

[0344] A dedicated artificial neural network model 120 can be trained based on the first hydrogen refueling data until predetermined conditions are met in order to optimize the hydrogen refueling process (step S1400).

[0345] Several existing refueling protocols, including the commonly used SAE J-2601, predetermine the range of mobile vehicles and hydrogen refueling stations to be operated, and implement models for control by predicting the number of possible scenarios within the corresponding range through simulation and empirical verification.

[0346] A universal protocol that comprehensively covers various conditions and vehicles can ensure safety and versatility. However, the boundary conditions set to meet the versatility requirements may be excessive, and there are attempts to optimize the universal protocol by applying various additional exception conditions to mitigate excessive boundary conditions. These attempts to add exception conditions have led to the problem of the control model becoming more complex.

[0347] In the refueling protocols of related technologies, the control model is based on simulation results calculated by assuming various conditions within a predetermined range, and the actual measurement data as a result of supplying fuel using the corresponding protocol is not reflected in the protocol's control model.

[0348] The difference between the simulated results and the actual results of the refueling protocol is determined by the level of the assumed boundary conditions, and the boundary conditions may be an important factor affecting the operating efficiency and refueling speed of the hydrogen refueling station.

[0349] To ensure generality, it is often difficult to consider changes in boundary conditions for special cases, and there are the following limitations: in terms of security, inefficient partial cases or special cases are difficult to consider.

[0350] Because existing refueling protocols have a structure that makes problem analysis and self-improvement difficult based on refueling result data, there is a need for methods that optimize individual speeds for individual vehicles or stations.

[0351] In the ANN model, the overall architecture and operation scheme of the program are standardized, while the dedicated detailed model can be flexibly generated according to the learned data and conditions. Furthermore, such a detailed model can be configured as an initial model for applying this disclosure based on theoretical simulation results, similar to existing protocols.

[0352] The betting process can be executed using an initial model, and reliability and efficiency can be gradually improved by retraining the initial model using generated actual betting data. Due to this characteristic, betting speed optimization can be achieved relatively easily through the accumulation of betting data.

[0353] ANN models can achieve functions such as problem analysis of the betting process, self-improvement through analysis, and speed optimization, which are difficult to achieve in existing protocols. In addition, because ANN models are based on the correlation between inputs and outputs and do not utilize the mathematical or theoretical relationships between them, betting result data generated according to another existing protocol can be used as is without modification or improvement within the model.

[0354] In the case of stations operating using existing protocols such as SAE J2601, inefficiencies in the refueling process can be identified and analyzed, but improvements to the protocol itself may be structurally difficult. According to an exemplary embodiment of this disclosure, even in stations operating using existing protocols in this manner, dedicated refueling speed optimizations for each mobile body type can be performed using an ANN model, and the optimized ANN model may have a clear advantage over existing protocols in terms of refueling speed.

[0355] By utilizing refueling data generated at stations operating under hydrogen refueling protocols including SAE J2601 that employ related technologies, a refueling profile for each mobile body type at the corresponding station can be determined. The method for obtaining a specialized ANN basic model based on the determined refueling profile can then be categorized into the following two exemplary implementations.

[0356] ① External Acquisition of the ANN Basic Model: If the ANN basic model is stored within the mobile device, it can be directly received and obtained through communication, and is generally available from the mobile device manufacturer. If the manager of the basic model is a distributor, station operator, or other third-party organization, the basic model can be obtained from the respective organization.

[0357] ② Self-construction of ANN models: Distributor manufacturers can independently build basic models using ANN model development tools. First, based on the refueling configuration file, station specifications, environmental conditions, etc. for each mobile body type, theoretical data is exported through the simulation module in the development tool, and the basic model can be built using the ANN learning module.

[0358] The above process can be performed using an ANN model development tool framework.

[0359] ANN model development tool framework is a tool that can generate specialized ANN base models based on the injection profile and station specifications for each mobile body type and perform retraining based on actual measurement data generated through the injection process.

[0360] The framework can be composed of a simulation module that generates theoretical data and an ANN learning module that is responsible for learning the ANN model.

[0361] The framework can instruct software packages that can automatically / semi-automatically generate specialized ANN base models based on basic inputs, and the amount and reliability level of data in the initial model configuration can be selected by the user, as can the level of the retraining process.

[0362] According to an exemplary embodiment of this disclosure, the process of replacing existing protocols with artificial neural network models can be implemented in stations operating based on existing hydrogen refueling protocols.

[0363] The injection data of mobile entities injected based on existing injection protocols can be applied to specialized ANN base models according to the corresponding mobile entity types, and can be retrained and optimized.

[0364] Specialized ANN models, retrained using data accumulated at the station, can be experimentally applied to betting for the corresponding movement type, and safety and speed optimizations can be evaluated. Reliability and safety are improved by utilizing betting data obtained during betting operations performed using the ANN-based model.

[0365] In the initial stage of ANN model application, station operating efficiency can be improved by simultaneously and complementaryly operating existing general but speed-deficient protocols and ANN models that enable speed optimization but lack versatility. Furthermore, the optimized ANN model can be used independently to completely replace existing protocols.

[0366] In the long run, the ANN model may be advantageous for the expansion and operational efficiency of all mobile body types at stations, but its application may need to be reviewed if there is no refueling requirement for the corresponding vehicle at the station or no customer demand for improvements in refueling speed.

[0367] Therefore, the replacement of existing protocols by the ANN model can be carried out under the following conditions.

[0368] ① By statistically analyzing the refueling data of hydrogen refueling stations currently operating under SAE J2601, we can first identify the main mobile vehicle types and main mobile vehicle types that are refueled at the corresponding stations, and then review and select the target mobile vehicle type based on user complaints or needs related to refueling speed optimization.

[0369] ② Priority review of ANN model application: Since ANN models can maximize functionality by utilizing real-time data from mobile bodies and stations during the control process, ANN models should be applied first to mobile body types that support real-time communication. The target mobile body type can be reviewed and selected based on whether a profile for each mobile body type is provided as static data for simulation and whether the requirements are sufficient.

[0370] ③ Specialized ANN models can be implemented using ANN model development tools. In the initial operational phase where the learning using real data is insufficient, retraining can be performed by operating existing protocols and using the corresponding results. Afterward, stability can be verified while operating in parallel. And based on the fact that sufficient reliability has been ensured through retraining, the model itself can be replaced.

[0371] The artificial neural network model 120 according to the above exemplary embodiments of this disclosure can use simulated data and real-world data to learn functions for controlling the hydrogen refueling process, or can learn parts of the control functions.

[0372] The actual field data used for learning can vary depending on the specifications of the storage tank or device on the station 200 side, and the reliability of the artificial neural network model 120 may be reduced when the learned data differs from the hardware specifications at the actual application time point, based on the actual refueling situation performed for a specific mobile body 300.

[0373] To improve and optimize the reliability and efficiency of the artificial neural network model 120, specialized individual models can be learned for each type and detailed specification of the station 200 and the mobile body 300.

[0374] In the case of station 200, the conditions used to classify lower-level individual models may include maximum supply pressure, pre-cooling temperature, maximum flow rate, etc.

[0375] In the case of mobile body 300, the conditions may include vehicle model name, total CHSS volume, maximum container volume, maximum allowable flow rate, etc., and the vehicle model name may include a serial number classified by manufacturer, etc.

[0376] At this point, the CHSS of the mobile body 300 can be implemented for safety by including multiple storage tanks of different types and sizes, and in the mobile body 300, not only the total volume of the CHSS can be considered, but also the detailed specifications of the storage tanks can be considered.

[0377] Individually specialized and trained artificial neural network (ANN) model 120 can be set up and operated in distributor 100, and according to an exemplary embodiment, artificial neural network model 120 can also be stored in mobile body 300 or a third-party location and operated via network communication.

[0378] A basic artificial neural network model can be provided in the allocator 100, and lower-level artificial neural network models trained and differentiated for individual body-specific conditions can be set up and managed in the allocator 100. The basic artificial neural network model can be initialized by the manufacturer of the mobile body 300.

[0379] According to an exemplary embodiment, even when the basic artificial neural network model is stored in the mobile body 300 or a third-party location, the individual model for hydrogen refueling can be stored and operated on the dispenser 100 side during actual field operations.

[0380] In other exemplary embodiments of this disclosure, the manufacturer of the mobile unit 300 may generate and distribute a basic artificial neural network model initialized using data based on the characteristics of the storage tank. The basic artificial neural network model may be stored in the storage device of the mobile unit 300 and transmitted to the distributor 100 after the mobile unit 300 enters the station 200.

[0381] The differentiation of specialized artificial neural network models and the learning of lower-level individual models can be primarily performed by the allocator 100. Lower-level individual models can learn or relearn based on the actual injection data performed in the allocator 100.

[0382] To ensure the reliability of the lower-level individual models, the relearning of the artificial neural network model can be performed after sufficient real-world data has been accumulated. This process can be performed periodically or irregularly in the allocator 100.

[0383] To ensure the reliability of the lower-level individual models, the manufacturer of mobile unit 300 can collect the refueling data received from station 200 and relearn the basic artificial neural network model. The updated basic artificial neural network model (which can be interpreted as a model specific to each mobile unit 300) can then be transmitted back to dispenser 100 for further updates. The basic artificial neural network model can be transmitted to dispenser 100 when mobile unit 300 enters station 200.

[0384] It can operate a third party that can ensure the reliability of lower-level individual models and provide integrated management.

[0385] Third parties can be government agencies, associations, professional management companies, etc.

[0386] The association can be a technology-related association or a standards-related association.

[0387] The refueling data generated in the actual field can be collected through cloud servers and relearned using artificial neural network models. Third parties can generate and distribute basic artificial neural network models on behalf of Mobile300.

[0388] Furthermore, lower-level individual models can be collected and upgraded periodically or irregularly, and the upgraded individual models can be provided back to each station. This process can be implemented as a form of federated learning.

[0389] Figure 12 This indicates that it can be executed. Figures 1 to 11 A conceptual diagram of an example of a generalized hydrogen refueling control device, hydrogen refueling control system, hydrogen refueling simulation device, hydrogen refueling simulation system, hydrogen refueling test platform, hydrogen refueling test system, or computing system, representing at least a part of the process.

[0390] refer to Figure 12 The computing system 3000 according to an exemplary embodiment of the present disclosure can be configured to include a processor 3100, a memory 3200, a communication interface 3300, a storage device 3400, an input interface 3500, an output interface 3600, and a bus 3700.

[0391] The artificial neural network model 120 of the above exemplary embodiments can be as follows: Figure 12 It is configured like the artificial neural network model 3800 and can communicate with other components in the computing system via bus 3700.

[0392] A computing system 3000 according to an exemplary embodiment of the present disclosure may include at least one processor 3100 and a memory 3200 storing instructions that instruct the at least one processor 3100 to execute at least one program. At least a portion of a program of a method according to an exemplary embodiment of the present disclosure may be loaded from the memory 3200 by the at least one processor 3100 and executed by executing the instructions.

[0393] Processor 3100 may refer to a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor that performs methods according to exemplary embodiments of the present disclosure.

[0394] Each of the memory 3200 and the storage device 3400 may be configured with one or more of volatile or non-volatile storage media. For example, the memory 3200 may be configured with one or more of read-only memory (ROM) or random access memory (RAM).

[0395] In addition, the computing system 3000 may include a communication interface 3300 for performing communication via a wireless network.

[0396] In addition, the computing system 3000 may also include a storage device 3400, an input interface 3500, an output interface 3600, etc.

[0397] Furthermore, the various components included in the computing system 3000 can be connected via bus 3700 and can communicate with each other.

[0398] Examples of the computing system 3000 disclosed herein may be communicable desktop computers, laptop computers, notebooks, smartphones, tablet PCs, mobile phones, smartwatches, smart glasses, e-book readers, portable multimedia players (PMPs), portable game consoles, navigation devices, digital cameras, digital multimedia broadcast players (DMBs), digital audio recorders, digital audio players, digital video recorders, digital video players, personal digital assistants (PDAs), etc.

[0399] An apparatus installed in a hydrogen-fueled mobile body 300 and controlling the hydrogen refueling process according to an exemplary embodiment of the present disclosure may include a memory 3200 storing at least one command; and a processor 3100 executing at least one command.

[0400] The processor 3100 can acquire first refueling data obtained during the first hydrogen refueling process through at least one command. The first hydrogen refueling process is executed based on the first refueling protocol (step S1200). It can train a control model to learn the control functions of the hydrogen refueling process based on the first refueling data (step S1400). It can also execute a second hydrogen refueling process based on one or more of the first refueling protocol or the control model (step S1600).

[0401] The processor 3100 can execute the second hydrogen refueling process by operating the first refueling protocol and control model in parallel.

[0402] The processor 3100 can perform the second hydrogen refueling process by using a control model to replace at least a portion of the functionality of the first refueling protocol.

[0403] At this point, the processor 3100 can verify the control model and can perform the second hydrogen refueling process by replacing at least a portion of the functionality of the first refueling protocol with the verified control model.

[0404] The processor 3100 can perform the second hydrogen refueling process by using at least a portion of the functions of the control model-assisted first refueling protocol.

[0405] In an apparatus for controlling the process of refueling a hydrogen-fueled mobile body according to an exemplary embodiment of the present disclosure, the control model may be a model that has learned the functionality of Model Prediction Control (MPC).

[0406] In an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, according to an exemplary embodiment of the present disclosure, the control model may be an artificial neural network (ANN) model that has learned the control functions of the hydrogen refueling process.

[0407] In an apparatus for controlling the process of refueling a hydrogen-fueled mobile body according to an exemplary embodiment of the present disclosure, processor 3100 may determine whether a first refueling protocol supports a process for generating control information through model predictive control.

[0408] In an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, according to an exemplary embodiment of the present disclosure, processor 3100 may determine whether a communication protocol associated with a first refueling protocol supports bidirectional communication between the mobile vehicle and a dispenser that refuels the mobile vehicle with hydrogen.

[0409] In an apparatus for controlling the process of refueling a mobile vehicle with hydrogen as fuel, according to an exemplary embodiment of the present disclosure, processor 3100 may determine whether the mobile vehicle and the dispenser for refueling the mobile vehicle support bidirectional communication between the mobile vehicle and the dispenser.

[0410] The control device for the hydrogen refueling process according to an exemplary embodiment of this disclosure can be located in a dispenser that dispenses hydrogen to a mobile body, or it can be implemented as a controller that is not located in the dispenser but communicates electronically with the dispenser and can affect the dispensing of the dispenser.

[0411] The operation of the methods according to exemplary embodiments of this disclosure can be implemented as computer-readable programs or code in a computer-readable recording medium. A computer-readable recording medium includes all types of recording devices in which information readable by a computer system is stored. Furthermore, the computer-readable recording medium can be distributed across computer systems connected via a network, and the computer-readable programs or code can be stored and executed in a distributed manner.

[0412] Furthermore, computer-readable recording media can include hardware devices specifically configured to store and execute program instructions, such as read-only memory (ROM), random access memory (RAM), flash memory, etc. Program instructions can include not only machine language code, such as code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or similar tool.

[0413] Some aspects of this disclosure have been described in the context of a device, but this description may also represent a description based on the corresponding method, and here, a block or device corresponds to a program of the method or a feature of the program of the method. Similarly, aspects described in the context of a method may also represent a corresponding block or item or a feature of the corresponding device. Some or all of the program of the method may be executed by (or using) a hardware device, such as, for example, a microprocessor, a programmable computer, or an electronic circuit. In some exemplary embodiments, one or more of the most important programs of the method may be executed by such a device.

[0414] In exemplary embodiments, a programmable logic device (e.g., a field-programmable gate array) may be used to perform some or all of the functions of the methods described herein. In exemplary embodiments, the field-programmable gate array may operate in conjunction with a microprocessor to perform one of the methods described herein. Generally, the methods are preferably performed by a hardware device.

[0415] Although the present disclosure has been described above with reference to preferred exemplary embodiments thereof, those skilled in the art will understand that various modifications and changes may be made to the present disclosure without departing from the spirit and scope of the present disclosure as set forth in the appended claims.

Claims

1. A method of refueling (filling) hydrogen, said method being performed in a dispenser for refueling a mobile vehicle using hydrogen as fuel, said method comprising: Acquire first refueling data during the first hydrogen refueling process, wherein the first hydrogen refueling process is executed based on a first refueling protocol; The control model is trained based on the first refueling data to learn the control functions of the hydrogen refueling process; as well as The second hydrogen refueling process is executed based on one or more of the first refueling protocol and the control model.

2. The method according to claim 1, wherein, Performing the second hydrogen refueling process includes: performing the second hydrogen refueling process by operating the first refueling protocol and the control model in parallel.

3. The method according to claim 1, wherein, Performing the second hydrogen refueling process includes: performing the second hydrogen refueling process by replacing at least a portion of the functions of the first refueling protocol with the control model.

4. The method according to claim 3, further comprising: Verify the control model. The second hydrogen refueling process includes performing the second hydrogen refueling process by replacing at least a portion of the functionality of the first refueling protocol with the validated control model.

5. The method according to claim 1, wherein, Performing the second hydrogen refueling process includes: using the control model to assist at least a portion of the functions of the first refueling protocol to perform the second hydrogen refueling process.

6. The method according to claim 1, wherein, The control model is a model that has learned the functions of model predictive control (MPC).

7. The method according to claim 1, wherein, The control model is an artificial neural network (ANN) model that has learned the control functions of the hydrogen refueling process.

8. The method according to claim 1, further comprising: The process of determining whether the first refueling protocol supports generating control information through the Model Predictive Control (MPC) function.

9. The method according to claim 1, further comprising: Determine whether the communication protocol associated with the first refueling protocol supports bidirectional communication between the mobile body and the dispenser that refuels the mobile body with hydrogen.

10. The method according to claim 1, further comprising: Determine whether the mobile body and the dispenser that supplies hydrogen to the mobile body support bidirectional communication between the mobile body and the dispenser.

11. An apparatus for controlling the process of refueling (adding) hydrogen to a mobile body using hydrogen as fuel, the apparatus comprising: A memory that stores at least one command; as well as The processor that executes at least one of the commands, Specifically, the processor is configured to: Acquire first refueling data during the first hydrogen refueling process, which is executed based on a first refueling protocol. The control model is trained based on the first refueling data to learn the control functions of the hydrogen refueling process, and The second hydrogen refueling process is executed based on one or more of the first refueling protocol and the control model.

12. The device according to claim 11, wherein, The processor is configured to execute the second hydrogen refueling process by operating the first refueling protocol and the control model in parallel.

13. The device according to claim 11, wherein, The processor is configured to perform the second hydrogen refueling process by replacing at least a portion of the functionality of the first refueling protocol with the control model.

14. The device according to claim 13, wherein, The processor is configured to: verify the control model and perform the second hydrogen refueling process by replacing at least a portion of the functionality of the first refueling protocol with the verified control model.

15. The device according to claim 11, wherein, The processor is configured to perform the second hydrogen refueling process by using the control model to assist at least a portion of the functions of the first refueling protocol.

16. The device according to claim 11, wherein, The control model is a model that has learned the functions of model predictive control (MPC).

17. The device according to claim 11, wherein, The control model is an artificial neural network (ANN) model that has learned the control functions of the hydrogen refueling process.

18. The device according to claim 11, wherein, The processor is configured to: determine whether the first refueling protocol supports the generation of control information via model predictive control (MPC).

19. The device according to claim 11, wherein, The processor is configured to determine whether a communication protocol associated with the first refueling protocol supports bidirectional communication between the mobile body and a dispenser that refuels the mobile body with hydrogen.

20. The device according to claim 11, wherein, The processor is configured to determine whether the mobile body and the hydrogen dispenser that dispenses hydrogen to the mobile body support bidirectional communication between the mobile body and the dispenser.

Images