Process control method of electrolytic hydrogen production system and electronic equipment

Abstract

The invention discloses a process control method of an electrolytic hydrogen production system and electronic equipment. The method comprises the steps that working condition information is obtained, the working condition information is used for determining the current working state of the electrolytic hydrogen production system, and the working condition information comprises state parameters of target components in the electrolytic hydrogen production system; the working condition information serves as model input, control information is generated through a process adjustment model, the process adjustment model is obtained through training based on a selected reinforcement learning algorithm, and the control information is used for representing an optimal control strategy for adjusting operation parameters of the target component; based on the control information, the operation parameters of the target component are adjusted, and the operation parameters comprise at least one of the hydrogen storage tank pressure, the electrolytic bath working electrical parameters and the compressor rotating speed. According to the invention, the technical problem of low control precision caused by slow response speed of an existing control method in response to renewable energy fluctuation is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrogen production control, and in particular to a process control method and electronic equipment for an electrolytic hydrogen production system. Background Art

[0002] With the rapid development of renewable energy, especially the widespread application of wind and solar energy, the use of these renewable energy sources for electrolytic hydrogen production has become an important way of energy conversion and storage. However, the intermittent and volatile nature of wind and solar energy has brought many challenges to the electrolytic hydrogen production system. The traditional PID control method has the problems of slow response speed and low control accuracy when dealing with these fluctuations, resulting in a narrow load range and slow power regulation rate of the hydrogen production system, which cannot keep up with the changes in wind and solar power in a timely manner, thus affecting the safety and efficiency of the system.

[0003] In recent years, model predictive control (MPC) has been introduced into hydrogen electrolysis systems to improve the speed and accuracy of control. However, the performance of MPC is highly dependent on the establishment of physical models and requires a large amount of experimental data for manual fitting, which is not only time-consuming and labor-intensive, but also difficult to adapt to complex and changing working conditions in practical applications.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a process control method and electronic equipment for an electrolytic hydrogen production system, so as to at least solve the technical problem of low control accuracy caused by the slow response speed of the existing control method when dealing with fluctuations in renewable energy.

[0006] According to one aspect of an embodiment of the present invention, a process control method for an electrolytic hydrogen production system is provided, comprising: obtaining operating condition information, the operating condition information being used to determine the current operating state of the electrolytic hydrogen production system, the operating condition information including state parameters of target components in the electrolytic hydrogen production system; using a process adjustment model to generate control information based on the operating condition information as a model input, wherein the process adjustment model is trained based on a selected reinforcement learning algorithm, and the control information is used to characterize an optimal control strategy for adjusting operating parameters of the target component; and adjusting operating parameters of the target component based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

[0007] Furthermore, obtaining the operating condition information includes: obtaining at least one of temperature information, pressure information, current information, and voltage information of the target component.

[0008] Furthermore, the training of the process regulation model includes: establishing an initial model based on a reinforcement learning algorithm; obtaining a target training set, the target training set including at least one of the following: historical operating data of the electrolysis hydrogen production system, preset simulation training data; training the initial model based on the target training set to obtain the process regulation model.

[0009] Further, after adjusting the operating parameters of the target component based on the control information, the method includes: determining a reward function of the process adjustment model; and adjusting the parameters of the process adjustment model based on the reward function.

[0010] Furthermore, determining the reward function of the process regulation model includes: determining an operating objective of the electrolytic hydrogen production system, wherein the operating objective includes at least one of the following: maximum hydrogen production and minimum system energy consumption; and determining a positive reward sub-function in the reward function based on the operating objective.

[0011] Furthermore, determining the reward function of the process control model includes: determining the adverse state conditions of the hydrogen production system by electrolysis, wherein the adverse state conditions include at least one of the following: the temperature of the target component exceeds the safety value, the electrical parameter of the target component exceeds the limit value, and the hydrogen production is less than the set value; and determining the negative reward sub-function in the reward function based on the adverse state conditions.

[0012] Furthermore, establishing an initial model based on a reinforcement learning algorithm includes: establishing an initial model based on a deep Q algorithm or a policy gradient algorithm.

[0013] According to another aspect of an embodiment of the present invention, there is further provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein the method in each embodiment of the present invention is executed when the program is running.

[0014] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present invention.

[0015] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0016] In an embodiment of the present invention, a method of dynamically adjusting the control strategy based on reinforcement learning is adopted. The process adjustment model obtained by training based on the reinforcement learning algorithm can generate control information according to real-time operating condition information, thereby achieving the purpose of accurately and dynamically adjusting the operating parameters of the target components of the electrolysis hydrogen production system, thereby achieving the technical effect of improving the system response speed, control accuracy and adapting to complex and changeable operating conditions, and further solving the technical problem of low control accuracy caused by the slow response speed of the existing control method when dealing with fluctuations in renewable energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flow chart of an optional process control method for an electrolytic hydrogen production system according to an embodiment of the present invention;

[0019] Figure 2 is a system architecture diagram of an optional process control method for an electrolysis hydrogen production system according to an embodiment of the present invention;

[0020] Figure 3 is a reinforcement learning framework diagram of an optional process control method for an electrolytic hydrogen production system according to an embodiment of the present invention;

[0021] Figure 4 The present invention is a flowchart of a control process of an optional process control method for an electrolytic hydrogen production system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] PID control method, namely proportional-integral-derivative control (Proportional-Integral-Derivative Control), is a closed-loop control algorithm widely used in industrial automation. The PID controller calculates the deviation between the set value and the process variable, and then adjusts the controller output according to the three control actions of proportional (P), integral (I), and differential (D) to achieve the control target.

[0025] Proportional control (P): This is the most basic control mode, where the controller output is proportional to the deviation. The proportional term responds directly to the error, but it cannot completely eliminate the error, especially for persistent deviations.

[0026] Integral control (I): The role of the integral term is to eliminate the deviation and make the control amount change with time accumulation until the deviation is zero. Integral control can eliminate steady-state errors, but too large an integral effect may cause the system to respond slowly or even become unstable.

[0027] Differential control (D): The role of the differential term is to predict the trend of the deviation, so as to adjust the control amount in advance to prevent the deviation from being too large. Differential control can speed up the system's response, reduce overshoot, and improve the system's dynamic performance, but it is sensitive to noise.

[0028] The application of PID control methods in electrolytic hydrogen production systems is usually to stabilize the operating parameters of the electrolyzer, such as current, voltage, temperature and pressure, to ensure that the system operates near the set values, thereby improving the output and quality of hydrogen while ensuring system safety. However, in the face of the volatility of renewable energy such as wind and solar energy, the response speed and adaptability of PID control may be insufficient.

[0029] According to an embodiment of the present invention, a method embodiment of a process control method for an electrolytic hydrogen production system is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] Figure 1 The process control method of the electrolytic hydrogen production system according to an embodiment of the present invention is as follows: Figure 1 As shown, the method comprises the following steps:

[0031] Step S102, obtaining operating condition information, where the operating condition information is used to determine the current operating state of the hydrogen production system by electrolysis, and the operating condition information includes state parameters of target components in the hydrogen production system by electrolysis;

[0032] Step S104, based on the operating condition information as a model input, using the process adjustment model to generate control information, wherein the process adjustment model is trained based on the selected reinforcement learning algorithm, and the control information is used to represent the optimal control strategy for adjusting the operating parameters of the target component;

[0033] Step S106, adjusting the operating parameters of the target component based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

[0034] Specifically, the hydrogen storage tank pressure refers to the pressure level of the hydrogen stored in the tank. Pressure control directly affects the storage density and safety of hydrogen. Controlling the pressure of the hydrogen storage tank can ensure the effective storage of hydrogen, avoiding the danger of tank rupture caused by excessive pressure, or the decrease in hydrogen storage efficiency caused by too low pressure. In dynamic process control, the regulation of hydrogen storage tank pressure helps to balance system output and demand, ensuring a continuous and stable supply of hydrogen.

[0035] The working electrical parameters of the electrolyzer mainly include current and voltage. The current of the electrolyzer is used to control the rate of hydrogen generation, while the voltage affects the energy efficiency of the electrolysis process. Adjusting the current of the electrolyzer can dynamically adjust the hydrogen production to match the changes in renewable energy input. Voltage control can optimize electrolysis efficiency, reduce energy consumption, and improve the economic efficiency of the hydrogen production process. In the face of fluctuations in wind energy input, precise control of the current and voltage of the electrolyzer is the key to achieving efficient and safe hydrogen production.

[0036] The compressor speed refers to the rotation speed of the compressor per unit time. The speed adjustment affects the compression efficiency and energy consumption of hydrogen. Controlling the compressor speed can adjust the compression rate of hydrogen to adapt to changes in hydrogen production. Increasing the compressor speed when hydrogen production is high can compress hydrogen into the hydrogen storage tank in time to avoid hydrogen backlog; while reducing the speed when production is low can reduce unnecessary energy consumption and improve the operating efficiency and economic benefits of the entire system.

[0037] Through the above steps S102 to S106, in an embodiment of the present invention, a method of dynamically adjusting the control strategy based on reinforcement learning is adopted. The process adjustment model obtained by training based on the reinforcement learning algorithm can generate control information according to real-time operating condition information, thereby achieving the purpose of accurately and automatically dynamically adjusting the operating parameters of the target components of the electrolysis hydrogen production system, thereby achieving the technical effect of improving the system response speed, control accuracy and adapting to complex and changeable operating conditions, and further solving the technical problem of low control accuracy caused by the slow response speed of the existing control method when coping with fluctuations in renewable energy.

[0038] Optionally, acquiring the operating condition information includes: acquiring at least one of temperature information, pressure information, current information, and voltage information of the target component.

[0039] Specifically, temperature sensors such as thermocouples, RTDs, infrared temperature sensors, etc. are installed in key locations (such as inside the electrolyzer, hydrogen storage tanks, compressors, etc.) to monitor real-time temperature and collect temperature information. The temperature sensor converts temperature changes into electrical signals, which are read and recorded by the data acquisition system.

[0040] Use pressure sensors, such as piezoresistive sensors and capacitive pressure sensors, installed in electrolyzers, hydrogen storage tanks, compressor inlets and outlets, etc. to monitor real-time pressure and collect pressure information. The pressure sensor converts pressure changes into electrical signals, which are read and recorded by the data acquisition system.

[0041] Use current sensors, such as Hall effect sensors, current transformers, etc., installed in the power line of the electrolyzer to monitor real-time current and collect current information. The current sensor converts current changes into electrical signals, which are read and recorded by the data acquisition system.

[0042] Use a voltage sensor or multimeter installed in the electrolytic cell power line to monitor the real-time voltage and collect voltage information. The voltage sensor converts voltage changes into electrical signals, which are read and recorded by the data acquisition system.

[0043] In this embodiment, the collection of these operating information is usually achieved through sensors and data acquisition systems. By collecting these operating information, the system can more comprehensively understand the current operating conditions, help the system adjust control parameters according to real-time operating conditions, and ensure the efficiency and stability of the hydrogen production process. Abnormal temperature, pressure, current and voltage can also indicate problems in the system, such as equipment failure, system blockage, etc. Timely fault detection is helpful for maintenance and troubleshooting.

[0044] Optionally, the training of the process regulation model includes: establishing an initial model based on a reinforcement learning algorithm; obtaining a target training set, the target training set including at least one of the following: historical operating data of the electrolysis hydrogen production system, preset simulation training data; training the initial model based on the target training set to obtain the process regulation model.

[0045] Reinforcement learning algorithm is a machine learning method, a type of unsupervised learning, whose main goal is to enable the agent to learn how to make decisions in interaction with the environment to maximize a certain cumulative reward. In the process control method of electrolysis hydrogen production system, reinforcement learning algorithm can automatically adjust the control strategy through real-time feedback and self-learning to cope with the fluctuation of renewable energy input and achieve more efficient and safer hydrogen production.

[0046] Specifically, the target training set includes historical operating data of the electrolytic hydrogen production system or preset simulation training data. Historical operating data can be collected from the system operation records, including normal operation, abnormal conditions and fault data; simulation training data is generated by establishing a mathematical model of the system or using software tools for simulation. The target data set should include various states and action instances in the operation of the system, as well as the corresponding result feedback. The initial model is pre-trained through historical operating data and preset simulation training data to obtain a process adjustment model, which can preliminarily grasp the operating laws of the target components. The model can make optimal decisions under a variety of working conditions, is suitable for electrolytic hydrogen production systems under different environments and conditions, and improves the generalization ability of the model. This training method reduces the dependence on physical models, reduces the complexity and cost of manual fitting, and improves the adaptability and robustness of the model.

[0047] Optionally, after adjusting the operating parameters of the target component based on the control information, the method includes: determining a reward function of the process adjustment model; and adjusting the parameters of the process adjustment model based on the reward function. By dynamically adjusting the reward function, the system can continuously optimize the control strategy to adapt to various changes in the electrolytic hydrogen production process.

[0048] The reward function is a criterion used in reinforcement learning algorithms to evaluate the quality of actions. It defines the immediate rewards or penalties for system states and actions. The process control model adjusts the operating parameters of the target component through the rewards or penalties fed back by the reward function to learn the optimal control strategy.

[0049] Specifically, the process regulation model predicts the next action (such as adjusting the electrolyzer current and adjusting the hydrogen storage tank pressure) based on the current state. According to the action predicted by the model, the operating parameters of the target components of the electrolytic hydrogen production system are adjusted. After executing the action, the system will update the state parameters of the target components, including hydrogen production, energy consumption, equipment status, etc. The reward or penalty for this regulation action is calculated according to the reward function. If the result meets the operating target (such as increased hydrogen production and reduced energy consumption), a positive reward is given; otherwise, a negative reward is given. The model adjusts the parameters of the reward feedback process regulation model through the back propagation algorithm to optimize the accuracy of the predicted action. The model continues the process of prediction-execution-observation-calculation of rewards-update parameters, and gradually masters the optimal control strategy through multiple iterative learning. By optimizing the control strategy, the safety and efficiency of the system are improved, energy consumption is reduced, and hydrogen production is increased. It not only improves the response speed and control accuracy of the system, but also significantly improves the safety and efficiency of the system, reduces operating costs, and has important economic and social benefits.

[0050] Optionally, determining the reward function of the process regulation model includes: determining an operating objective of the electrolytic hydrogen production system, wherein the operating objective includes at least one of the following: maximum hydrogen production and minimum system energy consumption; and determining a positive reward sub-function in the reward function based on the operating objective.

[0051] Specifically, the design of the positive reward subfunction should target one or more operational goals to ensure that the model learning behavior is consistent with the goal. When the model's control behavior approaches or achieves the operational goal, a significant positive reward is given.

[0052] According to a specific embodiment, the operating goal is set to maximize hydrogen production, while considering system energy consumption and stability, and it is not desired to increase production at the expense of system stability and a significant increase in energy consumption. To achieve this operating goal, the steps are as follows: an initial model is established based on a reinforcement learning algorithm, and historical operating data of the electrolytic hydrogen production system is collected, including hydrogen production, system energy consumption, electrolyzer temperature and pressure, etc., and the model is trained to learn how to dynamically adjust the electrolyzer current, hydrogen storage tank pressure, compressor speed, etc. to maximize hydrogen production, while controlling energy consumption and maintaining system stability. The trained model is applied to the actual electrolytic hydrogen production system. The model generates control information based on real-time operating condition information, thereby adjusting the operating parameters of the target component, and determining the reward function parameters based on the actual results after adjustment. The positive reward sub-function is determined by calculating the deviation between the current system operating condition information and the operating target.

[0053] Optionally, determining the reward function of the process control model includes: determining an adverse state condition of the hydrogen production system by electrolysis, wherein the adverse state condition includes at least one of the following: the temperature of the target component exceeds a safe value, the electrical parameter of the target component exceeds a limit value, and the hydrogen production is less than a set value; and determining a negative reward sub-function in the reward function based on the adverse state condition.

[0054] If the temperature of the target component exceeds the safety threshold, it may cause material damage, reduced chemical reaction efficiency, or increased safety risks. Electrical parameter out-of-limit: Including current and voltage exceeding the design limit, which may cause reduced electrolysis efficiency, equipment overload or damage. Hydrogen production is less than the set value, which may affect the economic benefits and operating efficiency of the system.

[0055] Specifically, the negative reward sub-function is determined by calculating the severity of the deviation of the current operating condition information of the system from the operating target.

[0056] In this embodiment, by introducing the negative reward sub-function, it can be ensured that the process regulation model not only pursues maximization of output and minimization of energy consumption when controlling the electrolysis hydrogen production system, but also can actively avoid or reduce the occurrence of adverse conditions such as system overheating, electrical parameter exceeding the limit and low output, thereby improving the overall operation safety and efficiency of the system.

[0057] Optionally, establishing an initial model based on a reinforcement learning algorithm includes: establishing an initial model based on a deep Q algorithm or a policy gradient algorithm. The present invention applies reinforcement learning to the process control method of the electrolysis hydrogen production system for the first time. Through the reinforcement learning algorithm, the system can adaptively adjust the control strategy to cope with the intermittency and volatility of renewable energy sources such as wind and solar energy. Compared with traditional PID control and model predictive control (MPC), reinforcement learning can achieve faster response speed and higher control accuracy, effectively improving the safety and efficiency of the system.

[0058] Specifically, Deep Q-Network (DQN), the core idea of ​​DQN is to approximate the Q function through a deep neural network (usually a convolutional neural network), that is, (Q(s,a;\theta)), where (s) is the state, (a) is the action, and (\theta) is the parameter of the network. It tries to learn a strategy that enables the selection of action (a) to maximize the future cumulative reward when given a state (s).

[0059] The Policy Gradient algorithm is a method for directly optimizing policy parameters in reinforcement learning. Representative algorithms include REINFORCE, Actor-Critic, Proximal Policy Optimization (PPO), etc. The Policy Gradient algorithm uses a policy network (usually a deep neural network) to parameterize the policy, selects actions and executes them according to the current policy, and the environment returns a reward signal, which is used to update the policy parameters to increase the probability of actions that will obtain higher cumulative rewards in the future.

[0060] like Figure 3 The figure shows a reinforcement learning framework diagram of an optional process control method for an electrolytic hydrogen production system according to an embodiment of the present invention. In this embodiment, a deep Q algorithm (DQN) or a policy gradient algorithm (Policy Gradient) is selected as a reinforcement learning algorithm to design a reinforcement learning framework, thereby establishing a model. The reinforcement learning framework includes a state space, an action space, a reward function, and a penalty mechanism, as follows:

[0061] The state space includes state variables such as temperature, pressure, current, and voltage.

[0062] The action space includes operations such as adjusting the current of the electrolyzer, adjusting the pressure of the hydrogen storage tank, and controlling the speed of the compressor.

[0063] The reward function is designed according to the operating objectives of the system, such as maximizing hydrogen production, minimizing energy consumption, etc.

[0064] At the same time, a penalty mechanism is set up, such as giving negative rewards when the system temperature exceeds the safety threshold or the hydrogen production is lower than the target value, to prevent the system from entering a bad state. The penalty mechanism can effectively constrain the operating range of the system, ensure that the system operates within a safe range, and improve the reliability and stability of the system.

[0065] Figure 2It is a system architecture diagram of a process control method for an optional electrolytic hydrogen production system according to an embodiment of the present invention. Specifically, the electrolytic hydrogen production system includes target components such as an electrolyzer, a hydrogen storage tank, and a compressor, and various sensors are used to collect system operation data, including parameters such as temperature, pressure, current, voltage, and speed. The arrows in the figure represent information flow and control flow, and sensor data such as temperature, pressure, current, voltage, and speed are continuously collected and transmitted to the control system. The control system combines the process regulation model to calculate the optimal control parameters, such as electrolyzer current, hydrogen storage tank pressure, and compressor speed, based on these data and preset operation goals. The received control parameters are applied to the electrolyzer, hydrogen storage tank, and compressor in real time to adjust their operating status. The system operation results and state changes are recorded as operation data as input for learning and optimization of the process regulation model. The process regulation model conducts online learning through operation data and continuously adjusts its strategy in order to achieve better control in future operations.

[0066] Figure 4 1 is a flow chart of a control process of an optional process control method for an electrolytic hydrogen production system according to an embodiment of the present invention, and the specific steps are as follows:

[0067] 1. Start.

[0068] 2. Initialize the system: prepare the sensors, control system and process adjustment model to enter the working state.

[0069] 3. Collect initial state: Obtain the current state data of the target component from the sensor and input it into the model.

[0070] 4. Loading process adjustment model: The model is loaded from the pre-trained state or the learning state of the previous cycle, ready for real-time control.

[0071] 5. Real-time monitoring of system status: Continuously collect system status and provide real-time data for the model.

[0072] 6. Determine whether adjustment is needed: Determine whether operating parameters need to be adjusted based on system status changes and preset control objectives.

[0073] If the current state meets the operating target (such as hydrogen production reaching the target), it is considered that no adjustment is required and the process ends.

[0074] If the system status deviates from the operating target or shows an unstable trend, proceed to the next step to adjust the control strategy.

[0075] 7. Calculate the optimal action: The model calculates and outputs the optimal control action based on the current state and the learned strategy.

[0076] 8. Execute control actions: The control system converts the calculated optimal actions into actual equipment operation instructions, such as adjusting the electrolyzer current, adjusting the hydrogen storage tank pressure, etc.

[0077] 9. Update system status: After executing the action, the system status changes and is updated to the current latest status.

[0078] 10. Record operation data: Store the system status and results after executing the action as a basis for model training and optimization.

[0079] 11. Feedback to process adjustment model: Feedback the operating target data to the model for online learning, and continuously adjust and optimize the control strategy.

[0080] By applying the trained process regulation model to the electrolysis hydrogen production system, the state variables of the system are monitored in real time. According to the current state, the process regulation model outputs the optimal control action and dynamically adjusts the current of the electrolyzer, the pressure of the hydrogen storage tank, the speed of the compressor and other parameters. The operating status of the system is monitored in real time, and the control strategy is adjusted according to the feedback information to ensure the stable operation of the system.

[0081] The embodiment of the present application also provides an electronic device, comprising: a memory storing an executable program; a processor for running the program, wherein when the program is running, the process control method of the electrolysis hydrogen production system in each embodiment of the present invention is executed. The specific steps are as follows:

[0082] Step S102, obtaining operating condition information, where the operating condition information is used to determine the current operating state of the hydrogen production system by electrolysis, and the operating condition information includes state parameters of target components in the hydrogen production system by electrolysis;

[0083] Step S104, based on the operating condition information as a model input, using the process adjustment model to generate control information, wherein the process adjustment model is trained based on the selected reinforcement learning algorithm, and the control information is used to represent the optimal control strategy for adjusting the operating parameters of the target component;

[0084] Step S106, adjusting the operating parameters of the target component based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

[0085] The embodiment of the present application also provides a computer-readable storage medium, which includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the process control method of the electrolytic hydrogen production system in each embodiment of the present invention. The specific steps are as follows:

[0086] Step S102, obtaining operating condition information, where the operating condition information is used to determine the current operating state of the hydrogen production system by electrolysis, and the operating condition information includes state parameters of target components in the hydrogen production system by electrolysis;

[0087] Step S104, based on the operating condition information as a model input, using the process adjustment model to generate control information, wherein the process adjustment model is trained based on the selected reinforcement learning algorithm, and the control information is used to represent the optimal control strategy for adjusting the operating parameters of the target component;

[0088] Step S106, adjusting the operating parameters of the target component based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

[0089] The embodiments of the present application also provide a computer program product, including a computer program, which, when executed by a processor, implements the process control method of the electrolysis hydrogen production system in each embodiment of the present invention. The specific steps are as follows:

[0090] Step S102, obtaining operating condition information, where the operating condition information is used to determine the current operating state of the hydrogen production system by electrolysis, and the operating condition information includes state parameters of target components in the hydrogen production system by electrolysis;

[0091] Step S104, based on the operating condition information as a model input, using the process adjustment model to generate control information, wherein the process adjustment model is trained based on the selected reinforcement learning algorithm, and the control information is used to represent the optimal control strategy for adjusting the operating parameters of the target component;

[0092] Step S106, adjusting the operating parameters of the target component based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

[0093] An embodiment of the present application further provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium is used to store a computer program, and when the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0094] The embodiments of the present application further provide a computer program, which implements the methods in the above-mentioned embodiments of the present invention when executed by a processor.

[0095] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0097] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0098] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0099] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0100] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A process control method for an electrolytic hydrogen production system, characterized in that: include: Acquiring operating condition information, wherein the operating condition information is used to determine the current operating state of the hydrogen production system by electrolysis, and the operating condition information includes state parameters of target components in the hydrogen production system by electrolysis; Based on the operating condition information as a model input, a process adjustment model is used to generate control information, wherein the process adjustment model is trained based on a selected reinforcement learning algorithm, and the control information is used to represent an optimal control strategy for adjusting operating parameters of a target component; The operating parameters of the target component are adjusted based on the control information, wherein the operating parameters include at least one of the following: hydrogen storage tank pressure, electrolyzer operating electrical parameters, and compressor speed.

2. The process control method according to claim 1, characterized in that: The working condition information obtained includes: At least one of temperature information, pressure information, current information, and voltage information of the target component is obtained.

3. The process control method according to claim 1, characterized in that: The training of the process adjustment model includes: Establishing an initial model based on the reinforcement learning algorithm; Acquire a target training set, wherein the target training set includes at least one of the following: historical operation data of the hydrogen production by electrolysis system, and preset simulation training data; The initial model is trained based on the target training set to obtain the process adjustment model.

4. The process control method according to claim 1, characterized in that: After adjusting the operating parameter of the target component based on the control information, the method includes: determining a reward function of the process regulation model; Parameters of the process regulation model are adjusted based on the reward function.

5. The process control method according to claim 4, characterized in that: Determining a reward function of the process regulation model includes: Determining an operation target of the electrolytic hydrogen production system, wherein the operation target includes at least one of the following: maximum hydrogen production and minimum system energy consumption; A positive reward sub-function in the reward function is determined based on the operational goal.

6. The process control method according to claim 4, characterized in that: Determining a reward function of the process regulation model includes: Determining an unfavorable state condition of the electrolytic hydrogen production system, wherein the unfavorable state condition includes at least one of the following: the temperature of the target component exceeds a safe value, the electrical parameter of the target component exceeds a limit value, and the hydrogen production is less than a set value; A negative reward sub-function in the reward function is determined based on the bad state condition.

7. The process control method according to claim 3, characterized in that: An initial model is established based on the reinforcement learning algorithm, including: The initial model is established based on the deep Q algorithm or the policy gradient algorithm.

8. An electronic device, characterized in that: include: A memory storing an executable program; A processor, configured to run the program, wherein the program executes the method according to any one of claims 1 to 7 when running.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The invention comprises a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images