Electrolytic cell liquid level control system, method, apparatus, and storage medium

Abstract

The electrolytic tank liquid level control system, method, device and storage medium provided by the embodiments of the present disclosure comprise: a data acquisition module, a reinforcement learning controller and an execution module; the data acquisition module is configured to acquire operation data of a hydrogen separator and an oxygen separator, and transmit the operation data to the reinforcement learning controller; the reinforcement learning controller is configured to receive the operation data, determine a working condition according to the operation data, generate an adjustment instruction based on the determined working condition, and send the adjustment instruction to the execution module; and the execution module is configured to receive the adjustment instruction, and perform a valve opening degree adjustment operation according to the adjustment instruction. The automatic and intelligent control of the electrolytic tank liquid level is realized, the production efficiency is improved through precise liquid level control, the production interruption is reduced, the product quality stability is improved, the safety risk caused by abnormal liquid level is reduced, and strong support is provided for the efficient, stable and safe operation of industrial production.

Description

Technical Field

[0001] This disclosure relates to the field of automation control technology, and in particular to an electrolytic cell liquid level control system, method, device and storage medium. Background Technology

[0002] Alkaline water electrolyzers, as an important method for large-scale hydrogen production, are widely used due to their low cost and mature technology. In this system, hydrogen separators and oxygen separators are responsible for separating the generated hydrogen and oxygen. Maintaining the liquid level balance between the two is crucial for ensuring the safe operation of the system. If the liquid level difference is too large, the risk of hydrogen and oxygen cross-contamination increases significantly, which can easily lead to equipment damage or even serious explosions.

[0003] Traditional alkaline electrolyzer systems regulate pressure and achieve liquid level control by controlling the opening of the separator outlet valve, with valve adjustment largely relying on PID controllers. However, this control method has several significant limitations. First, the parameters of the PID controller are fixed, making it difficult to adjust in real time according to changes in actual operating conditions. It fails to provide optimal control when facing highly volatile conditions such as intermittent renewable energy supply and frequent load fluctuations. Second, its response speed and accuracy are poor. In actual operation, especially when operating conditions change frequently, the valve opening, response speed, and control accuracy are difficult to match, leading to liquid level control failure, increasing the risk of downtime, raising the possibility of hydrogen-oxygen cross-contamination, and seriously threatening system safety. Furthermore, under complex operating conditions, such as fluctuations in energy input and changes in load demand, traditional PID control struggles to maintain stable liquid level control. For example, when solar or wind power supply is unstable, the difficulty of electrolyzer liquid level control increases significantly, and uncontrolled situations are highly likely to occur.

[0004] In summary, to meet the need for precise and stable control of the liquid level in alkaline water electrolyzers, there is an urgent need for an intelligent control system that can adjust control parameters in real time according to operating conditions, so as to cope with complex operating conditions and improve system safety and reliability. Summary of the Invention

[0005] This disclosure provides an electrolytic cell level control system, method, device, and storage medium to address the problems existing in the prior art.

[0006] To address the aforementioned problems, in a first aspect, an electrolytic cell level control system is provided, comprising:

[0007] It includes: a data acquisition module, a reinforcement learning controller, and an execution module;

[0008] The data acquisition module is used to collect the operating data of the hydrogen separator and the oxygen separator, and transmit the operating data to the reinforcement learning controller.

[0009] The reinforcement learning controller is used to receive the running data, determine the operating conditions based on the running data, generate adjustment instructions based on the determined operating conditions, and send the adjustment instructions to the execution module.

[0010] The execution module is used to receive the adjustment command and perform valve opening adjustment operation according to the adjustment command.

[0011] In conjunction with the first aspect, in one possible implementation, the reinforcement learning controller is configured to acquire historical operating data and use the historical operating data for offline training to generate a valve opening control strategy. In the case of online operation, it selects an action based on the valve opening control strategy and the received operating data, obtains rewards or penalties through interaction with the actual environment to evaluate the action, generates a new valve opening control strategy, and generates adjustment instructions based on the new valve opening control strategy.

[0012] In conjunction with the first aspect, in one possible implementation, the system further includes: a data processing unit; the data processing unit is connected to the data acquisition module and the reinforcement learning controller;

[0013] The data processing unit is used to receive the running data transmitted by the data acquisition module, perform data cleaning, normalization and feature extraction on the running data, and transmit the processed running data to the reinforcement learning controller.

[0014] In conjunction with the first aspect, in one possible implementation, the execution module includes: a valve actuator, a hydrogen separator outlet valve, and an oxygen separator outlet valve;

[0015] The valve actuator is used to receive the adjustment command output by the reinforcement learning controller and drive the hydrogen separator outlet valve and the oxygen separator outlet valve to adjust their opening degree so that the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve conforms to the target opening degree represented by the adjustment command.

[0016] In conjunction with the first aspect, in one possible implementation, the system further includes: a feedback module; the feedback module is connected to the data acquisition module.

[0017] The feedback module is used to acquire liquid level data from the data acquisition module according to a preset sampling period, compare the actual liquid level with the target liquid level, calculate the liquid level deviation, and feed back the liquid level deviation information to the reinforcement learning controller.

[0018] The reinforcement learning controller is used to adjust the adjustment command according to the liquid level deviation.

[0019] In conjunction with the first aspect, in one possible implementation, the system further includes: a security monitoring module; the security monitoring module is connected to the data acquisition module;

[0020] The safety monitoring module is used to monitor the operating data, and when a preset potential risk is detected, it issues an early warning and executes a preset emergency response plan.

[0021] The preset potential risks include: excessive liquid level difference and / or abnormal pressure; the preset emergency handling plan includes: closing the valve or stopping the system operation.

[0022] In conjunction with the first aspect, in one possible implementation, the system further includes: a fault diagnosis module; the fault diagnosis module is connected to the data acquisition module;

[0023] The fault diagnosis module is used to acquire the operating data from the data acquisition module; identify equipment faults based on the operating data and match corresponding solutions.

[0024] Secondly, a method for controlling the liquid level in an electrolyzer is provided, including:

[0025] Collect operating data from the hydrogen separator and oxygen separator;

[0026] Based on the operating data, the operating conditions are determined, and adjustment instructions are generated based on the determined operating conditions;

[0027] Perform the valve opening adjustment operation according to the adjustment command.

[0028] In conjunction with the second aspect, in one possible implementation, determining the operating condition based on the operating data and generating adjustment instructions based on the determined operating condition includes:

[0029] Historical operating data is acquired and used for offline training to generate a valve opening control strategy. In online operation, an action is selected based on the valve opening control strategy and the received operating data. The action is evaluated by obtaining rewards or penalties through interaction with the actual environment, a new valve opening control strategy is generated, and an adjustment command is generated based on the new valve opening control strategy.

[0030] In conjunction with the second aspect, in one possible implementation, the method further includes:

[0031] The operational data is then cleaned, normalized, and its features are extracted.

[0032] In conjunction with the second aspect, in one possible implementation, the step of performing the valve opening adjustment operation according to the adjustment command includes:

[0033] The opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve is adjusted so that the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve conforms to the target opening degree represented by the adjustment command.

[0034] In conjunction with the second aspect, in one possible implementation, the method further includes: acquiring liquid level data from a data acquisition module according to a preset sampling period, comparing the actual liquid level with the target liquid level represented by the adjustment command, calculating the liquid level deviation, and adjusting the adjustment command based on the liquid level deviation.

[0035] In conjunction with the second aspect, in one possible implementation, the method further includes: monitoring the operational data, and issuing an early warning and executing a preset emergency response plan when a preset potential risk is detected;

[0036] The preset potential risks include: excessive liquid level difference and / or abnormal pressure; the preset emergency handling plan includes: closing the valve or stopping the system operation.

[0037] In conjunction with the second aspect, in one possible implementation, the method further includes: acquiring the operational data from the data acquisition module; identifying equipment faults based on the operational data and matching corresponding solutions.

[0038] Thirdly, a computer device is provided, comprising: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, perform the steps of an electrolytic cell level control method as described in the second aspect, or in combination with any possible embodiment of the second aspect.

[0039] Fourthly, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, performs the steps of an electrolytic cell level control method as described in the second aspect, or in conjunction with any possible embodiment of the second aspect.

[0040] The beneficial effects of the embodiments disclosed herein include:

[0041] This disclosure provides an electrolyzer liquid level control system, method, equipment, and storage medium, applicable to industrial production scenarios involving hydrogen and oxygen separation, such as hydrogen production via water electrolysis. In such production, liquid level control of the hydrogen and oxygen separators is crucial for production efficiency, product quality, and safety. The system's data acquisition module collects operating data from the hydrogen and oxygen separators, providing accurate data for subsequent decisions, ensuring real-time system awareness of operating status, and avoiding control errors. The reinforcement learning controller receives data, determines operating conditions, and generates adjustment commands. Compared to traditional fixed-rule control methods, it can make intelligent decisions based on complex operating conditions, better cope with changing operating situations, improve control accuracy and flexibility, and optimize system performance. The execution module receives commands and executes valve opening adjustment operations, ensuring the effective implementation of the control strategy and achieving precise control of the hydrogen and oxygen separator liquid levels, maintaining stable production operation. Overall, this system achieves automated and intelligent control of the electrolyzer liquid level, improving production efficiency, reducing production interruptions, enhancing product quality stability, and reducing safety risks caused by abnormal liquid levels, providing strong support for the efficient, stable, and safe operation of industrial production. Attached Figure Description

[0042] Figure 1 This is one of the schematic diagrams of an electrolytic cell level control system provided in an embodiment of this disclosure;

[0043] Figure 2 A second schematic diagram of an electrolytic cell level control system provided in an embodiment of this disclosure;

[0044] Figure 3 A third schematic diagram of an electrolytic cell level control system provided in an embodiment of this disclosure;

[0045] Figure 4 A fourth schematic diagram of an electrolytic cell level control system provided in this embodiment of the present disclosure;

[0046] Figure 5 Fifth schematic diagram of an electrolytic cell level control system provided in this embodiment of the present disclosure;

[0047] Figure 6 This is a schematic diagram illustrating the application of an electrolyzer level control system in an industrial hydrogen production scenario, as provided in an embodiment of this disclosure.

[0048] Figure 7 An internal operation flowchart of an electrolytic cell level control system provided in this embodiment of the present disclosure;

[0049] Figure 8 A flowchart of an electrolytic cell level control method provided in an embodiment of this disclosure. Detailed Implementation

[0050] This disclosure provides an electrolytic cell level control system, method, apparatus, and storage medium. Preferred embodiments of this disclosure are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the scope of this disclosure. Furthermore, the embodiments and features described herein can be combined with each other unless otherwise specified.

[0051] This disclosure provides an electrolytic cell level control system, such as... Figure 1 As shown, it includes: a data acquisition module 101, a reinforcement learning controller 102, and an execution module 103;

[0052] The data acquisition module 101 is used to collect the operating data of the hydrogen separator and the oxygen separator, and transmit the operating data to the reinforcement learning controller 102;

[0053] The reinforcement learning controller 102 is used to receive operating data, determine the operating conditions based on the operating data, generate adjustment instructions based on the determined operating conditions, and send the adjustment instructions to the execution module 103.

[0054] The execution module 103 is used to receive adjustment instructions and perform valve opening adjustment operations according to the adjustment instructions.

[0055] Traditional control methods often employ PID controllers with fixed parameters, which cannot adjust in real time according to complex and changing operating conditions, resulting in poor liquid level control. Therefore, the system first incorporates a data acquisition module 101. Its function is to collect operating data from the hydrogen separator and oxygen separator, which forms the basis for subsequent control decisions. By acquiring operating data in real time and accurately, it provides a basis for subsequent system adjustments, overcoming the lack of real-time data support in traditional control methods. The reinforcement learning controller 102 is the core of the system, receiving operating data from the data acquisition module 101 and determining the operating conditions based on this data. Traditional control methods, unable to dynamically determine operating conditions, have fixed parameters that are difficult to adapt to changes. The reinforcement learning controller 102, trained based on operating data, can generate adjustment commands based on real-time operating conditions. This embodies the concept of adaptive liquid level control, dynamically adjusting the valve opening based on real-time operating conditions to achieve optimal liquid level control, solving the problem of traditional control methods being unable to adapt to changing operating conditions. The execution module 103 is responsible for receiving adjustment commands and executing valve opening adjustment operations, effectively implementing the commands generated by the reinforcement learning controller 102, ensuring that the system can accurately control the liquid level according to the control strategy.

[0056] In this embodiment, the system employs reinforcement learning technology to achieve adaptive liquid level control, overcoming the limitations of traditional fixed-parameter control and allowing for flexible adjustment of valve opening according to different operating conditions. Real-time data-driven optimization automatically adjusts the control strategy based on data collected by sensors in real time, improving response speed and control accuracy, and eliminating the inaccuracies caused by complex manual intervention and fixed PID parameters. This comprehensively enhances the level of liquid level control in the electrolytic cell.

[0057] In addition, the electrolytic cell level control system provided in this disclosure can be applied to a variety of scenarios and has high adaptability.

[0058] For example, large-scale industrial hydrogen production plants typically deploy numerous alkaline water electrolyzers. For instance, a chemical company uses alkaline water electrolyzers to produce hydrogen for chemical raw material synthesis. The data acquisition module 101 collects real-time operating data such as liquid level, pressure, and temperature from numerous hydrogen and oxygen separators, rapidly transmitting this data to the reinforcement learning controller 102. The reinforcement learning controller 102 uses this data to determine the current operating conditions. For example, during peak daytime electricity consumption, power supply fluctuations can cause unstable gas production from the electrolyzers. In this case, the controller 102 generates adjustment instructions based on a reinforcement learning algorithm. After receiving the instructions, the execution module 103 precisely adjusts the valve openings to ensure stable liquid levels in the hydrogen and oxygen separators, guaranteeing the safe and stable operation of the chemical production process and ensuring a continuous supply of hydrogen required for chemical raw material synthesis.

[0059] For example, in a distributed energy demonstration park, solar power is used for water electrolysis to produce hydrogen for energy storage. During the day, when there is ample sunlight, the electrolyzer operates actively, and the data acquisition module 101 continuously collects operating data from the hydrogen and oxygen separators. When clouds block sunlight, causing fluctuations in solar power generation, the reinforcement learning controller 102 judges the changes in operating conditions based on the operating data and generates adjustment commands. The execution module 103 responds quickly, adjusting the valve opening to maintain the liquid level balance in the hydrogen and oxygen separators, ensuring the smooth operation of the hydrogen production process. Excess electrical energy is efficiently converted into hydrogen energy for storage, which can be used at night or during peak energy demand periods, improving the stability and energy utilization efficiency of the distributed energy system.

[0060] For example, in hydrogen refueling stations for hydrogen-powered vehicles, hydrogen production through water electrolysis is one source of hydrogen. The data acquisition module 101 collects operational data from the hydrogen and oxygen separators in the electrolyzer within the refueling station. For instance, during peak refueling periods, increased hydrogen demand leads to changes in the electrolyzer load. The reinforcement learning controller 102 determines the operating conditions based on the data and promptly generates adjustment commands. The execution module 103 adjusts the valve opening to ensure stable liquid levels in the hydrogen and oxygen separators, guaranteeing stable hydrogen production and supply. This meets the rapid refueling needs of hydrogen-powered vehicles, improves the operational efficiency and safety of the refueling station, and promotes the widespread application of hydrogen-powered vehicles.

[0061] In another embodiment provided in this disclosure, the reinforcement learning controller 102 is used to acquire historical operating data and use the historical operating data for offline training to generate a valve opening control strategy. When operating online, it selects an action based on the valve opening control strategy and the received operating data, obtains rewards or penalties through interaction with the actual environment to evaluate the action, generates a new valve opening control strategy, and generates adjustment instructions based on the new valve opening control strategy.

[0062] In this embodiment, the reinforcement learning controller 102 first acquires historical operating data, which includes information on the past operating status of the electrolyzer under different conditions, such as the valve opening status of the hydrogen separator and oxygen separator under various pressures, temperatures, and gas production rates. This historical operating data is used for offline training. Through a specific reinforcement learning algorithm, the relationship between different operating conditions and valve opening is deeply mined and learned, thereby generating a valve opening control strategy. For example, during offline training, a series of operating condition features in the historical data are correlated with corresponding successful valve opening control operations to identify more effective valve opening adjustment methods under similar operating conditions, forming a preliminary control strategy.

[0063] During the online operation phase, the reinforcement learning controller 102 selects actions based on the generated valve opening control strategy and the real-time received current operating data. For example, if the current real-time operating data shows an upward trend in the hydrogen separator level, and according to the control strategy generated during offline training, appropriately increasing the valve opening helps maintain a stable level under similar conditions, the controller will choose to execute the action of increasing the valve opening. Then, by interacting with the actual environment—that is, observing the actual level change after executing the action—the controller obtains rewards or penalties to evaluate the action. If the level successfully stabilizes, a positive reward is given; otherwise, a negative penalty is given. Based on the reward or penalty feedback, the controller further optimizes and generates a new valve opening control strategy. For instance, if increasing the valve opening restores the level to stability, the weight of increasing the valve opening under similar conditions may be increased in the newly generated control strategy. Finally, an adjustment command is generated based on the new valve opening control strategy to control the execution module 103 to adjust the valve opening.

[0064] The reinforcement learning controller 102 continuously adapts to the complex and ever-changing operating conditions of the electrolyzer. Through learning from historical data and interacting with the actual environment during online operation, it continuously optimizes the valve opening control strategy, thereby improving the accuracy and stability of level control. Its significance lies in significantly enhancing the system's ability to cope with various operating conditions compared to traditional fixed-parameter control methods, reducing safety risks caused by improper level control, ensuring the stable operation of the electrolyzer, and improving production efficiency and safety.

[0065] In yet another embodiment provided in this disclosure, such as Figure 2 As shown, the system also includes: a data processing unit 104; the data processing unit 104 is connected to the data acquisition module 101 and the reinforcement learning controller 102;

[0066] The data processing unit 104 is used to receive the running data transmitted by the data acquisition module 101, perform data cleaning, normalization and feature extraction on the running data, and transmit the processed running data to the reinforcement learning controller 102.

[0067] In this embodiment of the disclosure, in the electrolyzer level control system, the data processing unit 104 is located between the data acquisition module 101 and the reinforcement learning controller 102. After the data acquisition module 101 acquires the operating data of the hydrogen separator and the oxygen separator, it transmits it to the data processing unit 104. The operating data includes various types of information, such as liquid level, pressure, temperature, and flow rate, but these raw data often suffer from problems such as noise, inconsistent formats, and lack of prominent features.

[0068] The data processing unit 104 first performs data cleaning on the received operating data. For example, the raw data may contain outliers caused by sensor malfunctions, such as a sudden spike in the hydrogen separator level data to a value far exceeding the normal range. The data cleaning process identifies and corrects these outliers to ensure data accuracy. Next, normalization is performed. Since different types of data have significantly different numerical ranges—for example, pressure data may range from tens to hundreds, while temperature data may range from tens to thousands—normalization maps these different ranges to a specific interval, such as [0,1]. This makes the reinforcement learning controller 102 more efficient in processing data and avoids learning bias caused by different data units. Finally, feature extraction is performed to extract key features that significantly impact level control from the large amount of raw data. For example, analysis reveals that features such as the hydrogen separator level change rate and the correlation between oxygen separator pressure and level are important for level control decisions; the data processing unit 104 extracts these features.

[0069] The operational data, after being cleaned, normalized, and processed by feature extraction, is transmitted to the reinforcement learning controller 102. This processed, high-quality data helps the reinforcement learning controller 102 to more accurately determine the operating conditions and generate more effective valve opening control strategies.

[0070] The data processing unit 104 improves the quality of the raw operating data, making it more suitable for analysis and decision-making by the reinforcement learning controller 102. Its significance lies in providing a reliable data foundation for the entire liquid level control system, thereby improving the accuracy and stability of the system's control over the electrolytic cell's liquid level and ensuring the system's safe and efficient operation. Through a series of processing operations on the raw data, data defects are eliminated, key information is highlighted, and subsequent control decisions are made more scientific and reasonable, avoiding control errors caused by data problems, which is crucial for improving the overall performance of the system.

[0071] In yet another embodiment provided in this disclosure, the execution module 103 includes: a valve actuator, a hydrogen separator outlet valve, and an oxygen separator outlet valve;

[0072] The valve actuator receives adjustment commands output by the reinforcement learning controller 102 and drives the hydrogen separator outlet valve and the oxygen separator outlet valve to adjust their openings so that the openings of the hydrogen separator outlet valve and the oxygen separator outlet valve conform to the target openings characterized by the adjustment commands.

[0073] In this embodiment of the disclosure, in the electrolyzer level control system, the execution module 103, as the final implementation part of the system control commands, consists of a valve actuator, a hydrogen separator outlet valve, and an oxygen separator outlet valve. After the reinforcement learning controller 102 determines the operating conditions based on the operating data and generates adjustment commands, the valve actuator undertakes the task of receiving these commands.

[0074] After receiving the adjustment command, the valve actuator converts the command into actual mechanical action to drive the hydrogen separator outlet valve and the oxygen separator outlet valve to adjust their opening. For example, in actual operation, if the reinforcement learning controller 102 determines, based on operating conditions such as liquid level changes, that to maintain the liquid level balance between the hydrogen separator and the oxygen separator, the opening of the hydrogen separator outlet valve needs to be increased by 10%, and the opening of the oxygen separator outlet valve needs to be decreased by 5%, the valve actuator will, according to the command, precisely control the movement of the valve core of the hydrogen separator outlet valve through motor drive or hydraulic drive, increasing its opening to the target 10%, while simultaneously controlling the movement of the valve core of the oxygen separator outlet valve, decreasing its opening to the target 5%. In terms of implementation, if the valve actuator uses motor drive, the valve opening can be precisely controlled by controlling the rotation angle and speed of the motor; if it uses hydraulic drive, the valve opening can be precisely adjusted by regulating the flow rate and pressure of the hydraulic oil.

[0075] The execution module 103 is a crucial component for achieving precise liquid level control. Its significance lies in ensuring the effective execution of the entire liquid level control system's strategy, directly impacting whether the liquid levels in the hydrogen and oxygen separators can be adjusted as expected, thereby guaranteeing the safe and stable operation of the electrolyzer. By accurately driving the outlet valves of the hydrogen and oxygen separators to the target opening degree through valve actuators, it maintains stable liquid levels within the separators, preventing dangerous situations such as hydrogen and oxygen cross-contamination due to abnormal liquid levels. This ensures efficient production in the electrolyzer and plays an indispensable role in the normal operation of the entire system.

[0076] In yet another embodiment provided in this disclosure, such as Figure 3 As shown, the system also includes: a feedback module 105; the feedback module 105 is connected to the data acquisition module 101.

[0077] Feedback module 105 is used to acquire liquid level data from data acquisition module 101 according to a preset sampling period, compare the actual liquid level with the target liquid level, calculate the liquid level deviation, and feed back the liquid level deviation information to reinforcement learning controller 102.

[0078] The reinforcement learning controller 102 is used to adjust the adjustment command according to the liquid level deviation.

[0079] In this embodiment of the disclosure, in the electrolyzer level control system, the feedback module 105 establishes a connection with the data acquisition module 101 and plays a crucial information feedback role in system operation. According to a preset sampling period, the feedback module 105 acquires the level data of the hydrogen separator and oxygen separator from the data acquisition module 101. This level data represents the actual level status in the electrolyzer at the current moment. Subsequently, the feedback module 105 compares the acquired actual level with a preset target level. For example, assuming the target level of the hydrogen separator is set to 50 cm, and the actual level acquired from the data acquisition module 101 at a certain sampling moment is 48 cm, the feedback module 105 will calculate the level deviation, i.e., the difference between 48 cm and 50 cm, which is -2 cm.

[0080] The feedback module 105 feeds back the calculated liquid level deviation information to the reinforcement learning controller 102. Upon receiving the liquid level deviation information, the reinforcement learning controller 102 adjusts the previously generated adjustment instructions based on its own algorithm and logic. For example, if the original adjustment instruction was to increase the opening of the hydrogen separator outlet valve by 5% to raise the liquid level, but the feedback module 105 indicates that the increase in liquid level is insufficient, the reinforcement learning controller 102 may adjust the instruction to increase the valve opening to 8% to make the actual liquid level approach the target liquid level more quickly.

[0081] In terms of implementation, the preset sampling period can be determined according to actual needs and system operational stability. If the system is sensitive to changes in liquid level, a shorter sampling period can be set, such as sampling once per second; if changes in liquid level are relatively slow and system stability requirements are not so stringent, the sampling period can be appropriately extended, such as sampling once every 5 seconds.

[0082] The feedback module 105 provides the reinforcement learning controller 102 with deviation information between the actual liquid level and the target liquid level, enabling the controller to dynamically adjust the control strategy based on the real-time liquid level situation, thereby optimizing the adjustment commands. Its significance lies in improving the accuracy and stability of liquid level control by forming a closed-loop control system. Continuously adjusting the control commands based on the liquid level deviation effectively avoids liquid level runaway caused by inaccurate initial control commands or changes in operating conditions, ensuring that the liquid levels of the hydrogen separator and oxygen separator are always maintained within the target range, guaranteeing the safe and stable operation of the electrolyzer, and playing a crucial role in improving the performance and reliability of the entire liquid level control system.

[0083] In yet another embodiment provided in this disclosure, such as Figure 4 As shown, the system also includes: a security monitoring module 106; the security monitoring module 106 is connected to the data acquisition module 101;

[0084] The safety monitoring module 106 is used to monitor the operation data, and when a preset potential risk is detected, it issues an early warning and executes a preset emergency response plan.

[0085] The preset potential risks include: excessive liquid level difference and / or abnormal pressure; the preset emergency response plan includes: shutting down the valve or stopping the system.

[0086] In this embodiment of the disclosure, the safety monitoring module 106 and the data acquisition module 101 are closely connected in the electrolyzer level control system. The data acquisition module 101 is responsible for collecting the operating data of the hydrogen separator and the oxygen separator. This data includes key information such as liquid level and pressure, and is then transmitted to the safety monitoring module 106. The core task of the safety monitoring module 106 is to monitor this operating data. It constantly monitors changes in the data and takes action when the data meets preset potential risk conditions. The preset potential risks cover two situations: excessive liquid level difference and / or abnormal pressure. For example, under normal circumstances, the liquid level difference between the hydrogen separator and the oxygen separator should be maintained within 5 cm. If the safety monitoring module 106 detects that the liquid level difference reaches 8 cm, exceeding the preset range, it determines that there is a potential risk of excessive liquid level difference. As another example, the pressure should be stable within a certain range during normal system operation. If the pressure suddenly rises or falls, exceeding the preset pressure range, it is considered an abnormal pressure.

[0087] Once a pre-set potential risk is detected, the safety monitoring module 106 will immediately issue a warning. The warning can be issued via an audible and visual alarm, such as a sharp alarm sounding in the control room accompanied by flashing bright lights, to attract the attention of operators. Simultaneously, a pre-set emergency response plan will be executed. This plan includes closing valves or stopping system operation. If the risk is determined to be caused by an excessive liquid level difference, the safety monitoring module 106 may control the closure of relevant valves to prevent further expansion of the liquid level difference and to prevent hydrogen and oxygen from mixing and causing a hazard. If the pressure anomaly is severe, it may directly trigger a command to stop system operation, avoiding serious consequences such as equipment damage or even explosion due to pressure issues.

[0088] In terms of implementation, the safety monitoring module 106 can use real-time data analysis algorithms to monitor operating data and continuously compare the collected data with preset risk thresholds. The preset potential risk thresholds need to be determined based on the design parameters of the electrolyzer, operating experience, and safety standards. For example, regarding the setting of the liquid level difference threshold, after a large number of experiments and theoretical analyses, it was determined that in a certain type of electrolyzer, a liquid level difference exceeding 6 cm may pose a significant safety risk. Therefore, 6 cm is set as the warning threshold for excessive liquid level difference.

[0089] The function of the safety monitoring module 106 is to promptly detect potential risks during the operation of the electrolytic cell and take effective early warning and emergency response measures to prevent danger from occurring. Its significance lies in providing a solid guarantee for the safe operation of the entire system, greatly reducing the possibility of safety accidents caused by excessive liquid level differences or abnormal pressure, protecting equipment and personnel safety, ensuring the continuity and stability of the production process, and being crucial for maintaining the reliable operation of the electrolytic cell liquid level control system.

[0090] In yet another embodiment provided in this disclosure, such as Figure 5 As shown, the system also includes: a fault diagnosis module 107; the fault diagnosis module 107 is connected to the data acquisition module 101;

[0091] The fault diagnosis module 107 is used to acquire operating data from the data acquisition module 101; identify equipment faults based on the operating data and match corresponding solutions.

[0092] In this embodiment of the disclosure, in the electrolyzer level control system, the fault diagnosis module 107 is closely connected to the data acquisition module 101. The data acquisition module 101 collects operating data from the hydrogen separator and oxygen separator, including various parameters such as liquid level, pressure, and temperature, and transmits this data to the fault diagnosis module 107. After acquiring this operating data, the fault diagnosis module 107 performs fault identification based on the data.

[0093] The fault diagnosis module 107 uses specific algorithms and pre-established fault models to conduct in-depth analysis of operating data. For example, during normal operation, the liquid level of the hydrogen separator fluctuates steadily within a certain range. If the fault diagnosis module 107 detects abnormally large fluctuations in the liquid level data from the operating data, and the duration exceeds the normal fluctuation duration, while the pressure data also shows corresponding abnormal changes, it can identify that the liquid level sensor of the hydrogen separator may be faulty by comparing it with the pre-set fault model.

[0094] Once a device malfunction is identified, the fault diagnosis module 107 immediately matches the corresponding solution. For the aforementioned liquid level sensor malfunction, the solution might be to prompt maintenance personnel to check if the sensor connection wiring is loose; if the wiring is normal, the liquid level sensor needs to be replaced. In terms of implementation, the fault diagnosis module 107 can employ machine learning-based algorithms, training the model with a large amount of historical fault data and normal operating data to improve the accuracy and efficiency of fault identification. For example, the operating data features corresponding to many different types of faults can be input into the machine learning model for training, enabling the model to learn typical data feature patterns of various faults. When new operating data is acquired, the model can quickly determine whether a fault exists and its type.

[0095] The fault diagnosis module 107 is designed to promptly detect equipment faults in the electrolytic cell level control system and provide targeted solutions to prevent the faults from escalating and causing serious impacts on system operation. Its significance lies in improving system reliability and maintainability, reducing downtime caused by equipment failures, minimizing production losses, and ensuring the smooth operation of the production process. Through real-time analysis of operating data and fault matching, it provides strong support for stable system operation, ensuring the continuous and efficient operation of the entire electrolytic cell level control system.

[0096] The electrolyzer level control system disclosed herein interacts with the hydrogen production system in a large-scale hydrogen production plant as follows: Figure 6 As shown, sensors collect real-time operational data such as liquid level and pressure from the separator in the hydrogen production system and transmit it to the reinforcement learning-based intelligent control system. The data processing unit in this system cleans, normalizes, and extracts features from the received operational data, providing high-quality data for the reinforcement learning controller.

[0097] The reinforcement learning controller determines the current operating condition based on the processed data, and then generates corresponding adjustment commands, which are sent to the execution module. Upon receiving the commands, the valve actuators in the execution module drive the hydrogen separator outlet valves and oxygen separator outlet valves to adjust their openings, thereby maintaining liquid level balance and controlling pressure to ensure stable operation of the hydrogen production system.

[0098] The feedback module acquires liquid level data from the data acquisition module according to a preset sampling period, compares the actual liquid level with the target liquid level, calculates the liquid level deviation, and feeds it back to the reinforcement learning controller so that it can optimize the adjustment instructions.

[0099] The safety monitoring module continuously monitors operational data. Once it detects potential risks such as excessive liquid level difference or abnormal pressure, it immediately issues an early warning and executes pre-set emergency response plans such as closing valves or stopping system operation to ensure the safety of the hydrogen production system.

[0100] The fault diagnosis module acquires operational data from the data acquisition module, identifies equipment faults in the hydrogen production system based on this data, and matches corresponding solutions to promptly repair faults, reduce system downtime, and improve the reliability and operational efficiency of the hydrogen production system.

[0101] For ease of understanding, such as Figure 7 As shown, the internal operation flow of the electrolyzer level control system is provided:

[0102] The level, pressure, temperature, and flow sensors in the sensor module (data acquisition module) collect real-time operating data of the electrolyzer and transmit it to the data processing unit. The data processing unit cleans and normalizes the raw data, removes noise, and standardizes the data format before sending the processed data to the reinforcement learning controller.

[0103] The reinforcement learning controller first generates an initial control strategy through offline training, and then optimizes it online based on real-time state evaluation. According to the evaluation results, the optimal action is selected, that is, the appropriate opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve is determined, and adjustment commands are sent to the actuator module.

[0104] After receiving the adjustment command, the actuator module drives the outlet valve of the hydrogen-oxygen separator to actuate and adjust the opening to achieve liquid level balance. Simultaneously, the data feedback module feeds back the real-time operating data after the valve actuation to the reinforcement learning controller, enabling it to continuously adjust the control strategy.

[0105] The safety monitoring module continuously monitors the data from the data acquisition module. If it detects potential risks such as excessive level difference or abnormal pressure, it immediately triggers actions such as closing valves or issuing alarms to ensure safe system operation. The fault diagnosis module, based on the collected data, identifies equipment faults and matches solutions for timely equipment maintenance, ensuring the stable and efficient operation of the entire electrolytic cell level control system.

[0106] Based on the same disclosed concept, this disclosure also provides an electrolytic cell level control method, device, and storage medium. Since the principle of solving the problem by these methods, devices, and storage media is similar to that of the aforementioned electrolytic cell level control system, the implementation of these methods, devices, and storage media can refer to the implementation of the aforementioned system, and repeated details will not be repeated.

[0107] With the above Figure 1 Correspondingly, this disclosure also provides a method for controlling the liquid level in an electrolyzer, such as... Figure 8 As shown, it includes:

[0108] S801, Collect operating data from the hydrogen separator and oxygen separator;

[0109] S802. Determine the operating conditions based on the operating data, and generate adjustment instructions based on the determined operating conditions;

[0110] S803. Perform valve opening adjustment operation according to the adjustment command.

[0111] In another embodiment provided in this disclosure, step S802, determining the operating condition based on the operating data and generating an adjustment instruction based on the determined operating condition, includes:

[0112] Historical operating data is acquired and used for offline training to generate valve opening control strategies. In online operation, actions are selected based on the valve opening control strategies and the received operating data. Rewards or penalties are obtained through interaction with the actual environment to evaluate the actions, generate new valve opening control strategies, and generate adjustment instructions based on the new valve opening control strategies.

[0113] In another embodiment provided in this disclosure, the method further includes:

[0114] The operational data is cleaned, normalized, and its features are extracted.

[0115] In another embodiment provided in this disclosure, step S803, performing valve opening adjustment operation according to adjustment instructions, includes:

[0116] The opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve is adjusted so that the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve conforms to the target opening degree represented by the adjustment command.

[0117] In another embodiment provided in this disclosure, the method further includes: acquiring liquid level data from a data acquisition module according to a preset sampling period, comparing the actual liquid level with the target liquid level represented by the adjustment command, calculating the liquid level deviation, and adjusting the adjustment command based on the liquid level deviation.

[0118] In another embodiment provided in this disclosure, the method further includes: monitoring the operating data, issuing an early warning and executing a preset emergency handling plan when a preset potential risk is detected;

[0119] The preset potential risks include: excessive liquid level difference and / or abnormal pressure; the preset emergency response plan includes: shutting down the valve or stopping the system.

[0120] In another embodiment provided in this disclosure, the method further includes: acquiring operational data from a data acquisition module; identifying equipment faults based on the operational data and matching corresponding solutions.

[0121] This disclosure provides a computer device, including a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of an electrolytic cell level control method provided in any embodiment of this disclosure.

[0122] The computer device provided in this disclosure includes a processor, a memory, and a bus. The memory, also known as internal memory, stores execution instructions and includes main memory and external memory. The main memory temporarily stores data processed by the processor, as well as data exchanged with external storage devices such as hard disks. The processor exchanges data with external storage devices through main memory. When the electronic device is running, the processor and memory communicate via the bus, causing the processor to execute the following instructions:

[0123] Collect operating data from the hydrogen separator and oxygen separator;

[0124] The operating conditions are determined based on the operational data, and adjustment instructions are generated based on the determined operating conditions.

[0125] Perform valve opening adjustment operations according to the adjustment instructions.

[0126] This disclosure provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program performs the steps of an electrolytic cell level control method provided in any embodiment of this disclosure. The storage medium can be volatile or non-volatile computer-readable storage.

[0127] Through the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of this disclosure can be implemented in hardware or by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.

[0128] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes in the drawings are not necessarily essential for implementing this disclosure.

[0129] Those skilled in the art will understand that the modules in the system of the embodiments can be distributed in the system of the embodiments as described in the embodiments, or they can be located in one or more systems different from this embodiment with corresponding changes. The modules of the above embodiments can be combined into one module, or they can be further divided into multiple sub-modules.

[0130] The sequence numbers of the embodiments disclosed above are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0131] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.

Claims

1. An electrolytic cell level control system, characterized in that, include: Data acquisition module, reinforcement learning controller, and execution module; The data acquisition module is used to collect the operating data of the hydrogen separator and the oxygen separator, and transmit the operating data to the reinforcement learning controller. The reinforcement learning controller is used to receive the running data, determine the operating conditions based on the running data, generate adjustment instructions based on the determined operating conditions, and send the adjustment instructions to the execution module. The execution module is used to receive the adjustment command and perform valve opening adjustment operation according to the adjustment command.

2. The system as described in claim 1, characterized in that, The reinforcement learning controller is used to acquire historical operating data and use the historical operating data for offline training to generate a valve opening control strategy. When operating online, it selects an action based on the valve opening control strategy and the received operating data, obtains rewards or penalties through interaction with the actual environment to evaluate the action, generates a new valve opening control strategy, and generates adjustment instructions based on the new valve opening control strategy.

3. The system as described in claim 1, characterized in that, The system further includes: a data processing unit; the data processing unit is connected to the data acquisition module and the reinforcement learning controller; The data processing unit is used to receive the running data transmitted by the data acquisition module, perform data cleaning, normalization and feature extraction on the running data, and transmit the processed running data to the reinforcement learning controller.

4. The system as described in claim 1, characterized in that, The execution module includes: a valve actuator, a hydrogen separator outlet valve, and an oxygen separator outlet valve; The valve actuator is used to receive the adjustment command output by the reinforcement learning controller and drive the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve to adjust the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve so that the opening degree of the hydrogen separator outlet valve and the oxygen separator outlet valve conforms to the target opening degree represented by the adjustment command.

5. The system as described in claim 1, characterized in that, The system further includes a feedback module; the feedback module is connected to the data acquisition module. The feedback module is used to acquire liquid level data from the data acquisition module according to a preset sampling period, compare the actual liquid level with the target liquid level, calculate the liquid level deviation, and feed back the liquid level deviation information to the reinforcement learning controller. The reinforcement learning controller is used to adjust the adjustment command according to the liquid level deviation.

6. The system as described in claim 1, characterized in that, The system further includes: a security monitoring module; the security monitoring module is connected to the data acquisition module; The safety monitoring module is used to monitor the operating data, and when a preset potential risk is detected, it issues an early warning and executes a preset emergency response plan. The preset potential risks include: excessive liquid level difference and / or abnormal pressure; the preset emergency handling plan includes: closing the valve or stopping the system operation.

7. The system as described in claim 1, characterized in that, The system further includes: a fault diagnosis module; the fault diagnosis module is connected to the data acquisition module; The fault diagnosis module is used to acquire the operating data from the data acquisition module; identify equipment faults based on the operating data and match corresponding solutions.

8. A method for controlling the liquid level in an electrolytic cell, characterized in that, include Collect operating data from the hydrogen separator and oxygen separator; Based on the operating data, the operating conditions are determined, and adjustment instructions are generated based on the determined operating conditions; Perform the valve opening adjustment operation according to the adjustment command.

9. A computer device, characterized in that, include: The system includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of the electrolytic cell level control method as described in claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the electrolytic cell level control method as described in claim 8.

Images