Big data analysis method and system for ion exchange membrane electrolysis water reactor

Abstract

The invention relates to the field of information management automation, and particularly discloses a real-time big data analysis method and system for an ion exchange membrane electrolysis water reactor. According to the method, sensors are deployed at key parts of the electrolytic water reactor to collect multi-source data, after cleaning and normalization preprocessing, operation parameters are optimized by using a deep learning model, fault prediction and diagnosis are performed by using a deep belief network or a support vector machine, and an operation mode is adjusted in combination with an adaptive control strategy. And data-driven decision making and visual management are realized by means of big data storage analysis and a visual monitoring platform. The system comprises a data acquisition module, a preprocessing module, an optimization module, a diagnosis module, a control module, a storage management module, a visual monitoring module and the like. The ion exchange membrane electrolysis water reactor can improve the electrolysis efficiency, reduce the energy consumption, enhance the system adaptability, reduce faults and promote the development of the ion exchange membrane electrolysis water reactor technology.

Description

Technical Field

[0001] The present invention relates to the fields of water electrolysis technology and big data analysis, and in particular to a real-time big data analysis method and system for ion exchange membrane water electrolysis reactors. Background Art

[0002] Ion exchange membrane water electrolysis reactors are an important hydrogen production device that utilizes the properties of ion exchange membranes to electrolyze water to produce hydrogen and oxygen. However, current IEM water electrolysis reactors face numerous operational challenges, severely hindering their performance improvement and large-scale application.

[0003] Difficulty in Precisely Optimizing Operating Parameters: The operating efficiency of a water electrolysis reactor is influenced by numerous parameters, such as electrolysis voltage, current density, electrolyte flow rate, temperature, and ion exchange membrane performance. These parameters are interdependent and mutually constrained, making precise optimization difficult using traditional methods based on experience or simple control strategies. In actual operation, it is often impossible to find the optimal parameter combination based on real-time operating conditions, resulting in low electrolysis efficiency and high energy consumption.

[0004] Difficulties in Fault Prediction and Diagnosis: During long-term operation, ion exchange membrane electrolysis reactors may experience faults such as ion exchange membrane aging, electrode corrosion, and seal leakage. Due to a lack of comprehensive monitoring and in-depth analysis of operating data, it is difficult to predict these faults in advance. When a fault does occur, it is also difficult to quickly and accurately diagnose its cause and location, resulting in high repair costs and prolonged downtime.

[0005] Poor system adaptability: The actual operating environment of water electrolysis reactors is complex and volatile, with factors such as power supply fluctuations and changes in feedwater quality. Existing systems lack a fast and effective adaptive adjustment mechanism to these changes, making it impossible to adjust operating parameters in a timely manner to maintain stable and efficient operation, impacting the continuity and stability of hydrogen production. Summary of the Invention

[0006] The present invention provides a real-time big data analysis method for an ion exchange membrane water electrolysis reactor. The method aims to optimize the operating parameters of the water electrolysis reactor, improve electrolysis efficiency, reduce energy consumption, and enhance system stability and reliability by collecting, analyzing, and processing the operating data of the water electrolysis reactor in real time. The method includes the following steps:

[0007] Install a variety of sensors at key locations of the ion exchange membrane water electrolysis reactor to collect real-time operating data such as electrolysis voltage V, current density J, electrolyte temperature T, pressure P, flow rate Q, and ion concentration C;

[0008] The collected data is cleaned, denoised, and normalized. The normalization formula is: Where x is the original data, xmin is the minimum value of the data type, xmax is the maximum value of the data type, and xnorm is the normalized data;

[0009] A neural network model based on deep learning is constructed. The preprocessed data is used as input. The model is trained through a large amount of historical operation data to learn the relationship between operating parameters and electrolysis efficiency eta and energy consumption E. The model training adopts the mean square error loss function, which is: Where n is the number of samples, is the model prediction value, and yi is the actual value. After the training is completed, the optimized operating parameters are output and the actual operation of the electrolytic water reactor is adjusted through the controller.

[0010] Build a fault prediction and diagnosis model using deep belief networks or support vector machine algorithms, collect data from normal operation and fault conditions to train the model, enabling it to learn the characteristics of different fault modes, monitor operating data in real time, predict the possibility of faults, and diagnose the fault type and location;

[0011] Based on the real-time collected data and the above model results, an adaptive control strategy is designed to automatically adjust the operation mode of the water electrolysis reactor;

[0012] Establish a big data storage and management system, use distributed file systems and databases to store historical data, and use data mining and analysis techniques to discover potential patterns and trends;

[0013] Build a visual monitoring platform to intuitively display the real-time operating status, historical data, fault information, optimization suggestions, etc. of the water electrolysis reactor.

[0014] Furthermore, the multiple sensors include voltage sensors and current sensors installed on the electrode surface, temperature sensors, pressure sensors, flow sensors on the electrolyte pipeline, and ion concentration sensors near the ion exchange membrane.

[0015] Furthermore, the neural network model based on deep learning is a long short-term memory network or a convolutional neural network.

[0016] Furthermore, during the training process of the fault prediction and diagnosis model, the cross-validation method is used to improve the generalization ability of the model.

[0017] Furthermore, in the adaptive control strategy, when the power supply fluctuates, the electrolysis voltage is adjusted according to the fluctuation amplitude ΔP according to the formula ΔV=k×ΔP, where k is a preset voltage adjustment coefficient.

[0018] Furthermore, in the big data storage and management system, the distributed file system adopts the Hadoop distributed file system HDFS, and the database adopts the NoSQL database.

[0019] Furthermore, the visual monitoring platform is developed through Web technology, using HTML, CSS, JavaScript languages, and combining Vue.js or React framework to realize data display function.

[0020] On the other hand, a real-time big data analysis system for an ion exchange membrane water electrolysis reactor is also provided, including:

[0021] The multi-source data real-time acquisition module is composed of multiple sensors installed in key parts of the ion exchange membrane water electrolysis reactor to collect operating data in real time;

[0022] Data preprocessing module, which cleans, denoises and normalizes the collected data;

[0023] The operating parameter optimization module builds a neural network model based on deep learning and outputs optimized operating parameters based on the training results;

[0024] Fault prediction and diagnosis module, which uses deep belief network or support vector machine algorithm to build models to predict faults and diagnose fault types and locations;

[0025] Adaptive control module, automatically adjusts the water electrolysis reactor operation mode based on real-time data and model results;

[0026] Big data storage and management module, which uses distributed file systems and databases to store historical data and conduct data mining and analysis;

[0027] Visual monitoring module, builds a visual monitoring platform to display information related to the water electrolysis reactor.

[0028] Furthermore, the sensor in the multi-source data real-time acquisition module is a high-precision sensor with a measurement error of no more than pm1%.

[0029] Furthermore, in the operating parameter optimization module, the training of the neural network model adopts GPU accelerated computing to increase the training speed.

[0030] Beneficial effects

[0031] Deeply integrated multi-source data collection and preprocessing: Breaking through the traditional monitoring model of single or a few parameters, comprehensive multi-source data from the water electrolysis reactor is collected. Advanced data preprocessing techniques provide a high-quality data foundation for subsequent analysis, enabling comprehensive understanding of the water electrolysis reactor's operating status. Deep learning models, such as LSTM or CNN, are used to explore the complex relationships between operating parameters and electrolysis efficiency and energy consumption. This allows for real-time optimization of operating parameters, improving electrolysis efficiency and reducing energy consumption, surpassing traditional empirical or simple control strategies. Fault prediction and diagnosis models, using algorithms such as deep belief networks (DBNs) or support vector machines (SVMs), are constructed to predict faults in advance, accurately diagnose fault types and locations, reduce failure losses, and improve equipment reliability. Adaptive control strategies are designed to automatically adjust operating modes based on real-time data and model analysis results, enabling the water electrolysis reactor to quickly adapt to complex and changing operating environments and enhance system stability and continuity. Big data storage and management: Data value mining: Efficient big data storage and management systems are established, combined with data mining and analysis techniques, to deeply mine the value of historical data, providing strong support for system optimization and fault prediction, and enabling data-driven decision-making. Visual monitoring platform facilitates efficient management: Building a visual monitoring platform provides intuitive display of system operation information, facilitating real-time monitoring and decision-making by operators, and improving management efficiency and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure (1) Flowchart of system operation principle. DETAILED DESCRIPTION

[0033] Example 1

[0034] A real-time big data analysis method and system for an ion exchange membrane water electrolysis reactor are provided. Through innovative big data analysis technology and intelligent control strategies, real-time optimization of the operating parameters of the water electrolysis reactor is achieved, faults are predicted and diagnosed in advance, and the system's adaptability to complex environments is enhanced, thereby improving the electrolysis efficiency of the water electrolysis reactor, reducing energy consumption, reducing the occurrence of faults, extending the service life of the equipment, and promoting the development and large-scale application of ion exchange membrane water electrolysis reactor technology.

[0035] A real-time big data analysis method for an ion exchange membrane water electrolysis reactor comprises the following steps:

[0036] Install a variety of sensors at key locations of the ion exchange membrane water electrolysis reactor to collect real-time operating data such as electrolysis voltage V, current density J, electrolyte temperature T, pressure P, flow rate Q, and ion concentration C;

[0037] The collected data is cleaned, denoised, and normalized. The normalization formula is: Where x is the original data, xmin is the minimum value of the data type, xmax is the maximum value of the data type, and xnorm is the normalized data;

[0038] A neural network model based on deep learning is constructed. The preprocessed data is used as input. The model is trained through a large amount of historical operation data to learn the relationship between operating parameters and electrolysis efficiency eta and energy consumption E. The model training adopts the mean square error loss function, which is: Where n is the number of samples, is the model prediction value, and yi is the actual value. After the training is completed, the optimized operating parameters are output and the actual operation of the electrolytic water reactor is adjusted through the controller.

[0039] Build a fault prediction and diagnosis model using deep belief networks or support vector machine algorithms, collect data from normal operation and fault conditions to train the model, enabling it to learn the characteristics of different fault modes, monitor operating data in real time, predict the possibility of faults, and diagnose the fault type and location;

[0040] Based on the real-time collected data and the above model results, an adaptive control strategy is designed to automatically adjust the operation mode of the water electrolysis reactor;

[0041] Establish a big data storage and management system, use distributed file systems and databases to store historical data, and use data mining and analysis techniques to discover potential patterns and trends;

[0042] Build a visual monitoring platform to intuitively display the real-time operating status, historical data, fault information, optimization suggestions, etc. of the water electrolysis reactor.

[0043] The various sensors include voltage sensors and current sensors installed on the electrode surface, temperature sensors, pressure sensors, flow sensors on the electrolyte pipeline, and ion concentration sensors near the ion exchange membrane.

[0044] The neural network model based on deep learning is a long short-term memory network or a convolutional neural network.

[0045] During the training process of the fault prediction and diagnosis model, the cross-validation method is used to improve the generalization ability of the model.

[0046] In the adaptive control strategy, when the power supply fluctuates, the electrolysis voltage is adjusted according to the fluctuation amplitude ΔP according to the formula ΔV=k×ΔP, where k is a preset voltage adjustment coefficient.

[0047] In the big data storage and management system, the distributed file system adopts the Hadoop distributed file system HDFS, and the database adopts the NoSQL database.

[0048] The visual monitoring platform is developed through Web technology, using HTML, CSS, and JavaScript languages, and combined with Vue.js or React framework to realize data display function.

[0049] On the other hand, a real-time big data analysis system for an ion exchange membrane water electrolysis reactor is also provided, including:

[0050] The multi-source data real-time acquisition module is composed of multiple sensors installed in key parts of the ion exchange membrane water electrolysis reactor to collect operating data in real time;

[0051] Data preprocessing module, which cleans, denoises and normalizes the collected data;

[0052] The operating parameter optimization module builds a neural network model based on deep learning and outputs optimized operating parameters based on the training results;

[0053] Fault prediction and diagnosis module, which uses deep belief network or support vector machine algorithm to build models to predict faults and diagnose fault types and locations;

[0054] Adaptive control module, automatically adjusts the water electrolysis reactor operation mode based on real-time data and model results;

[0055] Big data storage and management module, which uses distributed file systems and databases to store historical data and conduct data mining and analysis;

[0056] Visual monitoring module, builds a visual monitoring platform to display information related to the water electrolysis reactor.

[0057] Multi-source data real-time acquisition and preprocessing module: A variety of sensors, including voltage, current, temperature, pressure, flow, and ion concentration sensors, are installed at key locations within the ion exchange membrane water electrolysis reactor, such as electrodes, electrolyte pipes, and near the ion exchange membranes, to collect real-time operational data from the reactor. The collected data undergoes preprocessing operations such as cleaning, denoising, and normalization to remove outliers and noise interference, unifying data from different dimensions to the same scale for subsequent analysis.

[0058] Deep Learning-Based Operating Parameter Optimization Model: A deep learning-based neural network model, such as a long short-term memory network (LSTM) or a convolutional neural network (CNN), is constructed, using preprocessed data as input. The model is trained using a large amount of historical operating data, allowing it to learn the complex relationships between operating parameters, electrolysis efficiency, and energy consumption. The model outputs optimized operating parameters, such as electrolysis voltage, current density, and electrolyte flow rate. The controller then adjusts the actual operating parameters of the electrolysis stack, achieving real-time optimization of the stack's operation.

[0059] Fault prediction and diagnosis model: This model is constructed using algorithms such as deep belief networks (DBNs) or support vector machines (SVMs). Data from the water electrolysis reactor during normal operation and various fault conditions is collected and trained to learn the characteristics of different fault modes. By monitoring operational data in real time, the model predicts the likelihood of a fault occurring and accurately diagnoses the fault type and location when one occurs, issuing timely warnings to guide maintenance personnel.

[0060] Adaptive Control Strategy Module: This module designs an adaptive control strategy that automatically adjusts the electrolysis reactor's operating mode based on real-time data collected and the results of optimization models and fault prediction and diagnosis models. When power supply fluctuates, the electrolysis voltage and current density are adjusted to ensure stable operation. When ion exchange membrane performance degradation is detected, the electrolyte flow rate and temperature are adjusted to slow membrane aging. This adaptive adjustment improves the electrolysis reactor's adaptability to complex environments.

[0061] Big Data Storage and Management System: Establish a big data storage and management system, using a combination of distributed file systems (such as the Hadoop Distributed File System (HDFS)) and databases (such as NoSQL databases) to store large amounts of historical water electrolysis reactor operation data. Using data mining and analysis techniques, conduct in-depth analysis of historical data to uncover potential patterns and trends, providing data support for operating parameter optimization, fault prediction, and diagnosis.

[0062] Visual monitoring platform: A visual monitoring platform is built to display the real-time operating status, historical data, fault information, optimization suggestions, etc. of the electrolysis reactor through intuitive charts and graphs. Operators can use this platform to monitor the operation of the electrolysis reactor in real time, understand the system status in a timely manner, and make appropriate decisions.

[0063] Sensors in the multi-source data real-time acquisition and preprocessing module collect real-time operating data from the water electrolysis reactor. After preprocessing, data is transmitted to the deep learning-based operating parameter optimization model and the fault prediction and diagnosis model. The operating parameter optimization model outputs optimized operating parameters based on patterns learned from historical data, and the controller adjusts the actual operation of the water electrolysis reactor. The fault prediction and diagnosis model monitors data in real time, predicts faults, and diagnoses their type and location. The adaptive control strategy module automatically adjusts the operating mode of the water electrolysis reactor based on real-time data and model results. The big data storage and management system stores and analyzes historical data to support model training and decision-making. The visual monitoring platform presents system operation information to operators in an intuitive manner, facilitating monitoring and decision-making.

[0064] Example 2

[0065] (1) Hardware deployment

[0066] Sensor Installation: High-precision voltage and current sensors are installed on the electrode surfaces of the ion exchange membrane water electrolysis reactor to accurately measure the electrolysis voltage and current density. Temperature sensors, pressure sensors, and flow sensors are installed on the electrolyte inlet, outlet, and circulation pipes to monitor the electrolyte temperature, pressure, and flow rate. Ion concentration sensors are installed near the ion exchange membrane to monitor ion concentration changes in real time. Ensure that the sensors are installed in the appropriate locations, do not affect the normal operation of the water electrolysis reactor, and can accurately collect data.

[0067] Connecting the controller to the actuator: Select a reliable controller, such as a programmable logic controller (PLC) or industrial computer. Connect the controller to the sensor to obtain real-time sensor data. Also, connect the controller to the actuator, such as an electric control valve to adjust the electrolyte flow, a heating or cooling device controller to adjust the electrolyte temperature, and a power supply controller to adjust the electrolysis voltage and current density. Ensure that the controller can accurately receive sensor data and precisely control the actuator's movements based on the control instructions output by the algorithm.

[0068] (2) Software Programming

[0069] Data acquisition and preprocessing program development: Use Python or C++ to write data acquisition programs to implement real-time sensor data acquisition, calibration, and filtering. Use data normalization algorithms, such as min-max normalization or Z-score normalization, to map data from different ranges to a unified interval for subsequent analysis. Develop data cleaning and denoising programs to remove outliers and noise.

[0070] Deep Learning Model Training and Deployment: Build and train operating parameter optimization models and fault prediction and diagnosis models based on deep learning frameworks such as TensorFlow or PyTorch. Collect extensive historical operating data from the water electrolysis reactor, including both normal operation and fault conditions, to train and optimize the models. Once trained, deploy the models to servers or edge computing devices to rapidly process and analyze real-time data.

[0071] Adaptive control strategy implementation: Write an adaptive control strategy program to automatically generate control instructions based on real-time collected data and model analysis results. Implement communication with the controller to ensure that control instructions can be accurately sent to the controller to adjust the operating parameters of the water electrolysis reactor.

[0072] Big Data Storage and Management System Development: Build a big data storage and management system, using the Hadoop Distributed File System (HDFS) to store large amounts of historical data and NoSQL databases (such as MongoDB) for data management. Develop data mining and analysis programs, applying techniques such as cluster analysis and association rule mining to uncover underlying patterns and trends in the data.

[0073] Visual monitoring platform development: Using web development technologies such as HTML, CSS, and JavaScript, combined with relevant frameworks (such as Vue.js or React), a visual monitoring platform was developed. This platform provides real-time data display, historical data query, fault warning prompts, and optimization suggestion display, ensuring that operators can easily monitor and manage the water electrolysis reactor.

[0074] (3) System testing and optimization

[0075] Simulated Environment Testing: A test platform simulating an ion exchange membrane water electrolysis reactor was constructed and tested under various operating conditions, such as power supply fluctuations, electrolyte concentration variations, and temperature changes. This verified whether the multi-source data real-time acquisition and preprocessing module could accurately collect and process data, whether the deep learning model could correctly learn the relationship between operating parameters and electrolysis efficiency and energy consumption, whether the fault prediction and diagnosis model could accurately predict and diagnose faults, and whether the adaptive control strategy could effectively adjust the operating mode. Through simulation testing, data was collected and analyzed for model performance indicators, such as electrolysis efficiency improvement, energy consumption reduction, and fault prediction accuracy, to optimize the algorithm's parameters and structure.

[0076] Actual system testing: The real-time big data analysis system for ion exchange membrane water electrolysis reactors was deployed and tested on actual water electrolysis equipment. During actual operation, the system's operating status and performance indicators were monitored and compared with traditional control methods. The system's responsiveness and stability in the face of external disturbances such as power fluctuations and changes in feedwater quality were observed. Based on the actual test results, the system's hardware configuration and software algorithms were further optimized to ensure stable and efficient operation, achieving the goals of improving electrolysis efficiency, reducing energy consumption, and enhancing system adaptability.

[0077] Beneficial effects

[0078] Improve electrolysis efficiency: By optimizing operating parameters in real time, the ion exchange membrane water electrolysis reactor can maintain high electrolysis efficiency under different operating conditions. Compared with traditional control methods, it can significantly increase hydrogen production and meet the growing demand for clean energy.

[0079] Reduced energy consumption: Precise parameter control effectively reduces energy consumption in the water electrolysis process, reduces energy waste, reduces hydrogen production costs, improves the economic feasibility of water electrolysis reactors, and helps promote their large-scale commercial application.

[0080] Reduce the occurrence of failures: Fault prediction and diagnosis models can detect potential faults in advance and take timely measures to repair them, thereby reducing the occurrence of failures, reducing maintenance costs and downtime, and improving equipment reliability and service life.

[0081] Enhanced system adaptability: The adaptive control strategy enables the water electrolysis reactor to quickly adapt to complex and changing operating environments, ensuring the continuity and stability of the hydrogen production process and improving the overall performance of the system.

[0082] Improve management efficiency: The visual monitoring platform and big data storage and management system provide operators with intuitive operation information and data support, facilitating monitoring and decision-making, and improving management efficiency and scientificity.

Claims

1. A real-time big data analysis method for an ion exchange membrane water electrolysis reactor, characterized in that: The following steps are involved: Install a variety of sensors at key locations of the ion exchange membrane water electrolysis reactor to collect real-time operating data such as electrolysis voltage V, current density J, electrolyte temperature T, pressure P, flow rate Q, and ion concentration C; The collected data is cleaned, denoised, and normalized. The normalization formula is: Where x is the original data, xmin is the minimum value of the data type, xmax is the maximum value of the data type, and xnorm is the normalized data; A neural network model based on deep learning is constructed. The preprocessed data is used as input. The model is trained through a large amount of historical operation data to learn the relationship between operating parameters and electrolysis efficiency eta and energy consumption E. The model training adopts the mean square error loss function, which is: Where n is the number of samples, is the model predicted value, and yi is the actual value; After the training is completed, the optimized operating parameters are output and the actual operation of the water electrolysis stack is adjusted through the controller; Build a fault prediction and diagnosis model using deep belief networks or support vector machine algorithms, collect data from normal operation and fault conditions to train the model, enabling it to learn the characteristics of different fault modes, monitor operating data in real time, predict the possibility of faults, and diagnose the fault type and location; Based on the real-time collected data and the above model results, an adaptive control strategy is designed to automatically adjust the operation mode of the water electrolysis reactor; Establish a big data storage and management system, use distributed file systems and databases to store historical data, and use data mining and analysis techniques to discover potential patterns and trends; Build a visual monitoring platform to intuitively display the real-time operating status, historical data, fault information, optimization suggestions, etc. of the water electrolysis reactor.

2. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: The various sensors include voltage sensors and current sensors installed on the electrode surface, temperature sensors, pressure sensors, flow sensors on the electrolyte pipeline, and ion concentration sensors near the ion exchange membrane.

3. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: The neural network model based on deep learning is a long short-term memory network or a convolutional neural network.

4. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: During the training process of the fault prediction and diagnosis model, the cross-validation method is used to improve the generalization ability of the model.

5. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: In the adaptive control strategy, when the power supply fluctuates, the electrolysis voltage is adjusted according to the fluctuation amplitude ΔP according to the formula ΔV=k×ΔP, where k is a preset voltage adjustment coefficient.

6. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: In the big data storage and management system, the distributed file system adopts the Hadoop distributed file system HDFS, and the database adopts the NoSQL database.

7. The real-time big data analysis method for an ion exchange membrane water electrolysis reactor according to claim 1, characterized in that: The visual monitoring platform is developed through Web technology, using HTML, CSS, and JavaScript languages, and combined with Vue.js or React framework to realize data display function.

8. A real-time big data analysis system for an ion exchange membrane water electrolysis reactor, characterized in that: include: The multi-source data real-time acquisition module is composed of multiple sensors installed in key parts of the ion exchange membrane water electrolysis reactor to collect operating data in real time; Data preprocessing module, which cleans, denoises and normalizes the collected data; The operating parameter optimization module builds a neural network model based on deep learning and outputs optimized operating parameters based on the training results; Fault prediction and diagnosis module, which uses deep belief network or support vector machine algorithm to build models to predict faults and diagnose fault types and locations; Adaptive control module, automatically adjusts the water electrolysis reactor operation mode based on real-time data and model results; Big data storage and management module, which uses distributed file systems and databases to store historical data and conduct data mining and analysis; Visual monitoring module, builds a visual monitoring platform to display information related to the water electrolysis reactor.

9. The real-time big data analysis system for an ion exchange membrane water electrolysis reactor according to claim 8, characterized in that: The sensor in the multi-source data real-time acquisition module is a high-precision sensor with a measurement error of no more than pm1%.

10. The real-time big data analysis system for ion exchange membrane water electrolysis reactor according to claim 8, characterized in that: In the operating parameter optimization module, the training of the neural network model uses GPU accelerated computing to increase the training speed.

Images