Hydrogen leakage monitoring and system arrangement method for hydrogen-electric coupling system

Abstract

The application discloses a hydrogen leakage monitoring and system arrangement method for a hydrogen-electricity coupling system in oxygen, and relates to the technical field of hydrogen-electricity coupling systems. The method comprises the following steps: building an HTO test platform, monitoring the hydrogen concentration in oxygen in real time under different operation conditions according to the operation characteristic data set of the hydrogen-electricity coupling system under wind and light output fluctuation conditions, and establishing a basic data set of hydrogen leakage in oxygen; multi-scale modeling and simulation analysis are conducted to establish a hydrogen leakage and diffusion data set in oxygen; a data set of consequences of hydrogen explosion accidents in oxygen is established; a multi-objective optimization algorithm is used to optimize the electrolyzer arrangement scheme; a real-time monitoring model of hydrogen leakage in oxygen is constructed; and a hydrogen leakage early warning mechanism and emergency response strategy based on risk classification are established. The application combines the wind and light output fluctuation characteristics, HTO test, multi-scale numerical simulation and intelligent optimization algorithm, realizes accurate evaluation of the hydrogen leakage risk of the hydrogen-electricity coupling system in oxygen and optimization of the electrolyzer arrangement, and improves the intrinsic safety of the hydrogen-electricity coupling system.

Description

Technical Field

[0001] This invention belongs to the field of hydrogen energy safety technology, specifically a method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system. Background Technology

[0002] Hydrogen-electric coupling systems based on renewable energy have become an important technological path for constructing new power systems. The intermittent and fluctuating nature of wind and solar power generation leads to frequent load changes in electrolyzers, exacerbating diaphragm aging and damage, and increasing the risk of hydrogen-oxygen cross-contamination. Excessive hydrogen concentration in oxygen not only reduces oxygen purity but may also form explosive gas mixtures, causing serious safety accidents. Existing safety research on hydrogen-electric coupling systems focuses primarily on steady-state operating conditions, lacking a deep understanding of the hydrogen leakage mechanism under extreme conditions of fluctuating wind and solar power output. This has resulted in a failure to achieve real-time monitoring and accurate early warning of hydrogen leakage, leading to poor safety and reliability of hydrogen-electric coupling systems. Traditional electrolyzer layouts rely on empirical design, failing to fully consider leakage diffusion characteristics and monitoring optimization requirements.

[0003] Therefore, there is an urgent need for a method to monitor hydrogen leakage in oxygen in hydrogen-electric coupling systems that are subject to fluctuations in wind and solar power output, so as to achieve intrinsic safety optimization of electrolyzer layout. Summary of the Invention

[0004] To address the technical problems existing in the monitoring of hydrogen leakage in oxygen in existing hydrogen-electric coupling systems, this invention provides a method for monitoring and arranging hydrogen leakage in oxygen in hydrogen-electric coupling systems based on the fluctuation characteristics of wind and solar power output. This method combines the fluctuation characteristics of wind and solar power output, oxygen hydrogen concentration detection experiments, multi-scale numerical simulation, and intelligent optimization algorithms to achieve accurate assessment of the risk of hydrogen leakage in oxygen in hydrogen-electric coupling systems and optimization of the electrolyzer layout, thereby improving the inherent safety of hydrogen-electric coupling systems.

[0005] Therefore, the present invention adopts the following technical solution: a method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system, comprising: An HTO test platform was built, and the hydrogen-electric coupling system operating characteristics dataset under different operating conditions was used to monitor the hydrogen concentration in oxygen in real time, based on the data set of hydrogen-electric coupling system operating characteristics under fluctuating wind and solar power output, and to establish a basic dataset of hydrogen leakage in oxygen. Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer and a macroscale diffusion model of the system are constructed. Fluid dynamic simulation analysis of hydrogen leakage and diffusion behavior in oxygen is carried out to obtain the spatiotemporal evolution characteristics of hydrogen concentration under different leakage scenarios and establish a dataset of hydrogen leakage and diffusion in oxygen. Based on the oxygen-hydrogen leakage and diffusion dataset, combined with the characteristics of the ignition source and the geometric constraints of the confined space, a fluid dynamics simulation analysis of oxygen-hydrogen combustion and explosion accidents is carried out to obtain the temperature field, overpressure and flame propagation characteristics under different combustion and explosion scenarios, determine the safety boundary of oxygen-hydrogen combustion and explosion, and establish a dataset of consequences of oxygen-hydrogen combustion and explosion accidents. Based on the datasets of oxygen-hydrogen leakage and diffusion and oxygen-hydrogen combustion and explosion accident consequences, the electrolyzer spacing, ventilation layout and monitoring point arrangement are used as optimization variables, and the optimization objectives are minimizing the volume of combustible cloud, minimizing leakage detection time and minimizing equipment investment cost. A multi-objective optimization algorithm is used to optimize the electrolyzer layout scheme. Based on HTO test data and oxygen-hydrogen leakage diffusion dataset, a real-time monitoring model for oxygen-hydrogen leakage was trained using machine learning algorithms to achieve rapid prediction of oxygen-hydrogen concentration and leakage source location under fluctuating wind and solar power output conditions. Based on the safety boundary of hydrogen combustion and explosion in oxygen and the real-time monitoring model of hydrogen leakage in oxygen, a risk classification and early warning mechanism and emergency response strategy are established to achieve dynamic control of the risk of hydrogen leakage in oxygen in hydrogen-electric coupling system.

[0006] Furthermore, the process of establishing the data set of operating characteristics of the hydrogen-electric coupling system under fluctuating wind and solar power output specifically includes: Collect historical wind and solar power output data for the target area, analyze the daily characteristics, seasonal characteristics and extreme fluctuation events of wind and solar power output, and extract characteristic parameters of typical operating conditions, including power fluctuation rate, fluctuation amplitude and duration. Based on the topology and control strategy of the hydrogen-electric coupling system, a dynamic simulation model of the system is established to simulate the variation of the operating parameters of the electrolyzer under different wind and solar power output fluctuations. Design an HTO test matrix covering typical and extreme operating conditions. The test conditions include normal fluctuation conditions, rapid load reduction conditions, rapid loading conditions, and intermittent operation conditions. Establish a dataset of the operating characteristics of the hydrogen-electric coupling system under fluctuating wind and solar power output conditions.

[0007] Furthermore, the process of establishing the basic dataset for hydrogen leakage in oxygen specifically includes: Based on the membrane aging mechanism and mechanical damage mode of alkaline electrolyzers, the main causes of hydrogen-oxygen crosstalk were identified, including membrane thinning, pinhole formation and sealing failure. Based on the aforementioned factors, an HTO test platform was constructed, comprising an alkaline electrolyzer, an anode gas circulation system, a cathode gas circulation system, a cooling system, and a control system. Hydrogen and oxygen generated on the cathode and anode sides of the alkaline electrolyzer, respectively, enter the cathode and anode gas circulation systems. The control system outputs control parameters related to current, voltage, pressure, and flow rate to the alkaline electrolyzer, directly managing its operating conditions. The control system also outputs temperature control commands to the cooling system, ensuring that the electrolyzer's operating temperature meets the system's temperature control requirements. An oxygen-hydrogen concentration analyzer is installed on the anode side of the electrolyzer, and a hydrogen-oxygen concentration analyzer is installed on the cathode side. Based on the operating characteristic dataset of the hydrogen-electric coupling system, HTO tests were conducted under different current densities, operating temperatures, and membrane aging degrees to obtain the evolution law of hydrogen concentration in oxygen with operating parameters and establish a basic dataset of hydrogen leakage in oxygen.

[0008] Furthermore, the process of establishing the dataset on hydrogen leakage and diffusion in oxygen specifically includes: Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer was established to obtain the transmembrane hydrogen flux. Based on the three-dimensional geometric model of the hydrogen-electric coupling system, a macroscopic computational fluid dynamics diffusion model was established. The transmembrane hydrogen flux was used as the input of the boundary condition, and the inside of the anode cavity of the electrolyzer was used as the leakage source. Considering the coupling effect of mechanical ventilation and natural ventilation, fluid dynamics simulation analysis was carried out under different leakage intensities and different ventilation conditions to obtain the three-dimensional distribution characteristics of hydrogen concentration in the system. By coupling the transmembrane hydrogen flux and the three-dimensional distribution characteristics of hydrogen concentration in the system, a dataset of hydrogen leakage diffusion in oxygen was established.

[0009] Furthermore, the process of establishing the dataset of consequences of hydrogen combustion and explosion accidents in oxygen specifically includes: Based on the oxygen-hydrogen leakage and diffusion dataset, the range and concentration distribution of flammable hydrogen clouds under different leakage scenarios were determined. Using the large eddy simulation model and the flame surface tracking model, the hydrodynamic simulation of combustion and explosion accidents under different ignition locations, ignition times, and oxygen-hydrogen concentrations was carried out to obtain the deflagration-to-detonation characteristics, flame propagation velocity, overpressure peak value, and temperature field distribution. The critical concentration and critical ignition energy of oxygen-hydrogen combustion and explosion were determined, and a dataset of consequences of oxygen-hydrogen combustion and explosion accidents was established.

[0010] Furthermore, the specific process of optimizing the electrolytic cell layout includes: Using the spacing between electrolyzers, ventilation layout, and monitoring point arrangement as optimization variables, and minimizing the volume of combustible clouds, the shortest leakage detection time, and the minimum equipment investment cost as optimization objectives, a multi-objective optimization function is constructed. The NSGA-II or MOEA / D multi-objective optimization algorithm is used, combined with the oxygen-hydrogen leakage diffusion dataset and the combustion and explosion accident consequence dataset, to search for the Pareto optimal solution and obtain the optimal electrolyzer layout scheme that balances safety and economy.

[0011] Furthermore, the process of establishing a real-time monitoring model for hydrogen leakage in oxygen specifically includes: By integrating HTO test data and oxygen-hydrogen leakage diffusion datasets, a comprehensive dataset containing operating parameters, environmental parameters, and historical leakage characteristics is constructed. Random forest or gradient boosting tree algorithms are used to perform feature importance analysis on all feature variables in the comprehensive dataset, screening for sensitive features related to oxygen-hydrogen leakage. Using the screened sensitive features and their time-series data as input, a mapping relationship between wind and solar power output fluctuations, operating parameters, and oxygen-hydrogen concentration is established using a long short-term memory neural network or Transformer model. This yields a real-time monitoring model for oxygen-hydrogen leakage, enabling rapid prediction of oxygen-hydrogen concentration and leak source location under fluctuating wind and solar power output conditions.

[0012] Furthermore, the operating parameters include current, voltage, temperature, and pressure; the environmental parameters include ambient temperature, humidity, and ventilation rate; and the oxygen-hydrogen leakage sensitivity characteristics include current density fluctuation rate, membrane impedance change rate, and anode pressure.

[0013] Furthermore, the process of establishing an early warning mechanism and emergency response strategy for hydrogen leakage in oxygen based on risk classification specifically includes: Based on the critical concentration of hydrogen combustion and explosion in oxygen and the real-time monitoring model of hydrogen leakage in oxygen, a three-level early warning threshold is set; for different early warning levels, differentiated emergency response strategies are formulated: Level 1 early warning initiates automatic load reduction and enhanced ventilation, Level 2 early warning executes forced ventilation and leak investigation, and Level 3 early warning triggers emergency shutdown and personnel evacuation; an early warning-response linkage mechanism is established to achieve closed-loop management of monitoring-early warning-response.

[0014] Furthermore, the three-level warning thresholds are as follows: Level 1 warning, which is the attention level, is when the hydrogen concentration in oxygen is ≥1%vol, i.e., 25%LFL; Level 2 warning, which is the warning level, is when the hydrogen concentration in oxygen is ≥2%vol, i.e., 50%LFL; and Level 3 warning, which is the danger level, is when the hydrogen concentration in oxygen is ≥3%vol, i.e., 75%LFL.

[0015] The beneficial effects of this invention are as follows: By analyzing the coupling relationship between the fluctuation characteristics of wind and solar power output and the operating characteristics of the hydrogen-electric coupling system, this invention designs an HTO test scheme covering typical and extreme operating conditions, revealing the mechanism of hydrogen leakage in oxygen under fluctuating wind and solar power output; by constructing a microscale mass transfer model inside the electrolyzer and a macroscale diffusion model of the system, the invention accurately characterizes the hydrogen leakage and diffusion behavior in oxygen; by using a multi-objective optimization algorithm, the invention achieves an inherently safe design for the electrolyzer layout; and by using a data-driven model, the invention achieves real-time monitoring and accurate early warning of hydrogen leakage in oxygen, significantly improving the safety and reliability of the hydrogen-electric coupling system. Attached Figure Description

[0016] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to the present invention; Figure 2 This is a schematic diagram illustrating the analysis of wind and solar power output fluctuation characteristics according to the present invention. Detailed Implementation

[0018] The following examples illustrate specific implementations of the present invention. These examples are carried out based on the solution described in the present invention, and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following examples.

[0019] Example 1 This embodiment describes a method for monitoring and arranging a hydrogen-electric coupling system for hydrogen leakage in oxygen. Figure 1 As shown, the steps are as follows: Step 1) Based on the analysis of wind and solar power output fluctuation characteristics, identify the typical operating conditions of the hydrogen-electric coupling system, design an HTO (Hydrogen to Oxygen) test scheme covering typical operating conditions and extreme fluctuation scenarios, and establish a dataset of the operating characteristics of the hydrogen-electric coupling system under wind and solar power output fluctuation conditions.

[0020] The specific process includes: A. Collect historical wind and solar power output data for the target area, analyze the daily characteristics, seasonal characteristics, and extreme fluctuation events of wind and solar power output, and extract typical operating condition characteristic parameters, including power fluctuation rate, fluctuation amplitude, and duration.

[0021] B. Based on the topology and control strategy of the hydrogen-electric coupling system, a dynamic simulation model of the system is established to simulate the variation of operating parameters such as current density, temperature, and pressure of the electrolyzer under different wind and solar power output fluctuations.

[0022] C. Design an HTO test matrix covering typical and extreme operating conditions. The test conditions include normal fluctuation conditions, rapid load reduction conditions, rapid loading conditions, and intermittent operation conditions. Establish a dataset of the operating characteristics of the hydrogen-electric coupling system under fluctuating wind and solar power output conditions.

[0023] Step 2) Based on the membrane structure characteristics and gas diffusion mechanism of the electrolyzer in the hydrogen-electric coupling system, analyze the physicochemical process of hydrogen-oxygen cross-contamination, construct an HTO test platform, conduct real-time monitoring of hydrogen concentration in oxygen under different operating conditions based on the operating characteristic dataset of the hydrogen-electric coupling system, and establish a basic dataset of hydrogen leakage in oxygen.

[0024] The basic dataset for hydrogen leakage in oxygen is a structured dataset formed by organizing, calibrating, and classifying operating conditions based on HTO test data. The HTO test data is the raw data obtained from actual measurements on the HTO test platform, such as the concentration of hydrogen in oxygen and temperature under various operating conditions.

[0025] Specifically: Based on the membrane aging mechanism and mechanical damage mode of alkaline electrolyzers, the main causes of hydrogen-oxygen crosstalk were identified, including membrane thinning, pinhole formation and sealing failure. Based on the aforementioned factors, an HTO test platform was constructed, comprising an alkaline electrolyzer, an anode gas circulation system, a cathode gas circulation system, a cooling system, and a control system. Hydrogen and oxygen generated on the cathode and anode sides of the alkaline electrolyzer, respectively, enter the cathode and anode gas circulation systems. The control system outputs control parameters related to current, voltage, pressure, and flow rate to the alkaline electrolyzer, directly managing its operating conditions. The control system also outputs temperature control commands to the cooling system, ensuring that the electrolyzer's operating temperature meets the system's temperature control requirements. An oxygen-hydrogen concentration analyzer is installed on the anode side of the electrolyzer, and a hydrogen-oxygen concentration analyzer is installed on the cathode side, with a sampling frequency of no less than 1Hz. Based on the operating characteristic dataset of the hydrogen-electric coupling system, HTO tests were conducted under different current densities, operating temperatures, and membrane aging degrees to obtain the evolution law of hydrogen concentration in oxygen with operating parameters and establish a basic dataset of hydrogen leakage in oxygen.

[0026] Step 3) Based on the basic dataset of hydrogen leakage in oxygen, establish a microscale mass transfer model inside the electrolyzer and a macroscale diffusion model of the system, conduct hydrodynamic simulation analysis of hydrogen leakage and diffusion behavior in oxygen, obtain the spatiotemporal evolution characteristics of hydrogen concentration under different leakage scenarios, and establish a dataset of hydrogen leakage and diffusion in oxygen.

[0027] Specifically: Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer is established. The permeation process of hydrogen molecules in the membrane is simulated based on the dissolution-diffusion theory, and the quantitative relationship between transmembrane hydrogen flux and operating parameters (current density, temperature, membrane thickness) is obtained.

[0028] This embodiment simulates the dissolution-diffusion process of hydrogen molecules within the membrane based on Fick's diffusion law and Henry's law, calculating the transmembrane hydrogen flux. The formula for calculating the transmembrane hydrogen flux using Henry-Fick coupling is:

[0029] In the formula, This represents the transmembrane hydrogen flux (mol / (m²·s)). This represents the diffusion coefficient of H2 in the membrane (m² / s); Henry's solubility coefficient (mol / (m³·Pa)) is expressed as a constant. Indicates film thickness (m); This represents the partial pressure of hydrogen gas on the cathode side surface of the diaphragm, in Pa. This represents the partial pressure of hydrogen on the anode side surface of the diaphragm, expressed in Pa.

[0030] Based on the three-dimensional geometric model of the hydrogen-electric coupling system, a macroscopic computational fluid dynamics diffusion model is established, using standard... A turbulence model and component transport equations were used, with transmembrane hydrogen flux as the boundary condition input. The leakage source was the interior of the electrolyzer anode cavity, and the coupling effect of mechanical and natural ventilation was considered. Fluid dynamics simulations were conducted under different leakage intensities (small leakage <1 g / s, medium leakage 1-10 g / s, large leakage >10 g / s) and ventilation conditions to obtain three-dimensional hydrogen concentration distribution data within the system. The leakage intensity was determined by the transmembrane hydrogen flux output from the microscale mass transfer model. Obtained through conversion.

[0031] Adopting standards The formula for calculating the hydrogen concentration field using the turbulence model and component transport equations is as follows:

[0032]

[0033] In the formula, This indicates the density of the gas mixture, expressed in kg / m³. It represents turbulent kinetic energy, with units of m² / s². This represents the average velocity component of the fluid, with units of m / s (i is the directional index, corresponding to the three directions of the Cartesian coordinate system). This represents the turbulent kinetic energy dissipation rate, with units of m² / s³. , All represent spatial coordinate components, in meters, with i and j being summation indices; This represents the molecular dynamic viscosity, with units of Pa·s; This represents turbulent viscosity, expressed in Pa·s. G k This represents the turbulent kinetic energy generation term caused by the average velocity gradient, with units of kg / (m·s³), at the ventilation inlet. k Larger; turbulent kinetic energy =1.0, the turbulent Prandtl number of the turbulent kinetic energy dissipation rate. =1.3, =1.44, =1.92, =0.09.

[0034]

[0035] In the formula, Indicates the mass flux of the gas mixture; This represents the average velocity vector of the gas mixture, measured in m / s. In a hydrogen leak scenario, due to the significant buoyancy of hydrogen, ... The vertical component plays a key role in concentration stratification; Indicates the mass fraction of H2; This represents the effective diffusion coefficient (m² / s). ,in The diffusion coefficient of hydrogen molecules, ≈0.7 is the turbulent Schmidt number. From the above The model was obtained; This indicates the hydrogen source quantity (kg / (m³·s)), set at the location of the leakage source (such as the anode outlet, flange sealing surface, etc.); This represents the Hamiltonian operator.

[0036] By coupling transmembrane hydrogen flux and the three-dimensional distribution characteristics of hydrogen concentration within the system, a dataset of hydrogen leakage and diffusion in oxygen was established.

[0037] Step 4) Based on the oxygen-hydrogen leakage and diffusion dataset, combined with the ignition source characteristics and confined space geometric constraints, conduct fluid dynamics simulation analysis of oxygen-hydrogen combustion and explosion accidents, obtain temperature field, overpressure and flame propagation characteristics under different combustion and explosion scenarios, determine the safety boundary of oxygen-hydrogen combustion and explosion, and establish an oxygen-hydrogen combustion and explosion accident consequence dataset.

[0038] Specifically: Based on the oxygen-hydrogen leakage and diffusion dataset, the range and concentration distribution of flammable hydrogen clouds under different leakage scenarios were determined. Using the Large Eddy Simulation (LES) model and the flame front tracking model, hydrodynamic simulations of combustion and explosion accidents under different ignition locations, ignition times, and oxygen-hydrogen concentrations were conducted to obtain deflagration-to-detonation characteristics, flame propagation velocity, overpressure peak value, and temperature field distribution. The critical concentration and critical ignition energy of oxygen-hydrogen combustion and explosion were determined, and a dataset of consequences of oxygen-hydrogen combustion and explosion accidents was established.

[0039] Large eddy simulation (LES) models can more accurately simulate the evolution of transient flow fields, temperature fields, and concentration fields during combustion and explosion processes by directly solving large-scale turbulent structures and performing subgrid modeling only on small-scale turbulence.

[0040] ,

[0041] In the formula, the incompressible approximation of the isotropic part is ignored. This represents subgrid stress (Pa), which characterizes the feedback effect of filtered-out small-scale turbulence on large-scale motion; C represents the subgrid eddy viscosity coefficient, Pa·s; s ≈0.1: Smagorinsky constant; Δ: mesh filtering scale (m). In hydrogen combustion and explosion simulation, the typical mesh scale is 10-50 mm. It should be located in the turbulent inertial sub-region and needs to be able to analyze large-scale turbulent structures and key geometric feature dimensions. Since Δ is much larger than the thickness of the laminar flame (about 0.3~0.5 mm), the internal structure of the flame cannot be directly analyzed. Therefore, a flame surface tracking model is used to treat the flame as an infinitely thin interface. The modulus of the solvable scale strain rate tensor is defined as: | |=(2 ij ij ) 0.5 , This represents the solvable scale strain rate tensor after filtering.

[0042] The flame surface tracking model is as follows:

[0043] In the formula, G: Level Set Function, scalar field, the flame surface is defined as the isosurface of G=G0 (usually G0=0), G<0 is the burned region, and G>0 is the unburned region; Laminar flame velocity (m / s), hydrogen The hydrogen gas velocity is the highest among all combustible gases (approximately 2.9 m / s at stoichiometric ratio), which is one of the fundamental reasons for the explosion hazard of hydrogen. In practical applications... Often using effective flame velocity that includes sublattice wrinkling effect replace; : Favre filter (density-weighted filter) velocity vector (m / s).

[0044] Step 5) Based on the oxygen-hydrogen leakage and diffusion dataset and the oxygen-hydrogen combustion and explosion accident consequence dataset, the electrolyzer spacing, ventilation layout, and monitoring point arrangement are used as optimization variables, and the optimization objectives are minimizing the combustible cloud volume, minimizing the leakage detection time, and minimizing the equipment investment cost. A multi-objective optimization algorithm is used to optimize the electrolyzer layout scheme.

[0045] Specifically: Using electrolytic cell spacing, ventilation layout, and monitoring point arrangement as optimization variables, and minimizing combustible cloud volume, leakage detection time, and equipment investment cost as optimization objectives, a multi-objective optimization function is constructed. The NSGA-II or MOEA / D multi-objective optimization algorithm is employed, combined with oxygen-hydrogen leakage diffusion datasets and explosion accident consequence datasets, to perform Pareto optimal solution search, obtaining the optimal electrolytic cell layout scheme that balances safety and economy. The constraints of the multi-objective optimization algorithm are: combustible cloud volume < 100 m³; meeting explosion safety boundaries; leakage detection time < 30 s; constrained spatial geometric constraints and electrolytic cell spacing engineering requirements; and minimizing equipment investment cost.

[0046] Step 6) Based on HTO test data and oxygen-hydrogen leakage diffusion dataset, a real-time monitoring model for oxygen-hydrogen leakage is trained using machine learning algorithms to achieve rapid prediction of oxygen-hydrogen concentration and leakage source location under fluctuating wind and solar power output conditions.

[0047] Specifically: By integrating HTO test data and oxygen-hydrogen leakage diffusion datasets, a comprehensive dataset was constructed, including operating parameters (current, voltage, temperature, pressure), environmental parameters (ambient temperature, humidity, ventilation rate), and historical leakage characteristics. Random forest or gradient boosting tree algorithms were used to perform feature importance analysis on all feature variables in the comprehensive dataset, screening for oxygen-hydrogen leakage sensitive features (current density fluctuation rate, membrane impedance change rate, anode pressure, etc.). These screened oxygen-hydrogen leakage sensitive features were used as model input, employing a Long Short-Term Memory Neural Network (LSTM) or Transformer model to establish a nonlinear mapping relationship between wind and solar power output fluctuations, electrolyzer operating parameters, and oxygen-hydrogen concentration. After network training converged, all weight parameters were fixed, and the system was deployed as a low-latency online inference engine. During runtime, a fixed-length sliding time-series window constructed from the sensitive features collected in real-time by each sensor was used as input, and a predicted oxygen-hydrogen concentration value could be output within seconds through a single forward computation. When the predicted concentration exceeds the alarm threshold or the deviation from the predicted baseline without leakage exceeds the limit, a leakage alarm is triggered. Based on the source location probability distribution in the multi-task output, the leakage source is located, thereby obtaining a real-time monitoring model for hydrogen leakage in oxygen, enabling rapid prediction of hydrogen concentration in oxygen and location of leakage sources under fluctuating wind and solar power output.

[0048] Step 7) Based on the oxygen-hydrogen combustion and explosion safety boundary and the real-time monitoring model of oxygen-hydrogen leakage, establish an oxygen-hydrogen leakage early warning mechanism and emergency response strategy based on risk classification.

[0049] Specifically: Based on the critical concentration of hydrogen combustion in oxygen (4% vol), a three-level warning threshold is set: Level 1 warning (attention level, concentration ≥ 1% vol, i.e., 25% LFL), Level 2 warning (warning level, concentration ≥ 2% vol, i.e., 50% LFL), and Level 3 warning (hazard level, concentration ≥ 3% vol, i.e., 75% LFL). Differentiated emergency response strategies are developed for different warning levels: Level 1 warning initiates automatic load reduction and enhanced ventilation; Level 2 warning executes forced ventilation and leak detection; and Level 3 warning triggers emergency shutdown and personnel evacuation. A warning-response linkage mechanism is established to achieve closed-loop management of monitoring, warning, and response.

[0050] Example 2 Taking a wind-solar-hydrogen integrated demonstration project as the implementation object, the method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system as described in Example 1 includes the following steps: 1) such as Figure 2 As shown, one year's worth of wind and solar power data was collected at the project site, with a time resolution of 15 minutes. The average daily fluctuation rate of wind power was 25%, and the average daily fluctuation rate of photovoltaic power was 35%. In extreme scenarios, wind and solar power output could drop by 80% within 10 minutes. Based on the dynamic response characteristics of an alkaline electrolyzer (rated power 1MW, rated current density 1.0A / cm²), a dynamic simulation model of the hydrogen-electric coupling system was established. An HTO test condition matrix was designed: normal fluctuation condition (power change rate 3% / min), rapid load reduction condition (power change rate 20% / min), rapid loading condition (power change rate 15% / min), and intermittent operation condition (30min operation - 15min shutdown cycle), establishing a dataset of the operating characteristics of the hydrogen-electric coupling system under wind and solar power output fluctuation conditions.

[0051] 2) Construct an HTO test platform, including: a 5kW alkaline electrolyzer (single cell active area 200cm², membrane thickness 150μm), an anode gas circulation system, a cathode gas circulation system, a cooling system, and a control system. An oxygen-to-hydrogen concentration analyzer is installed on the anode side of the electrolyzer, and a hydrogen-to-oxygen concentration analyzer is installed on the cathode side (range 0-4%vol, accuracy 0.05%vol), with a sampling frequency of 10Hz. Conduct HTO tests under different current densities (0.4, 0.8, 1.2, 1.6 A / cm²), different temperatures (65, 70, 75℃), and different membrane aging degrees (new membrane, accelerated aging thin film, and artificial pinhole membrane). Simultaneously collect membrane impedance data, record the evolution of oxygen-to-hydrogen concentration with operating parameters, and establish a basic dataset for oxygen-to-hydrogen leakage.

[0052] 3) Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer was established. Based on Fick's diffusion law and Henry's law, the dissolution-diffusion process of hydrogen molecules within the membrane was simulated, and the transmembrane hydrogen flux was calculated. Based on the geometric model of the electrolyzer, a CFD model of the electrolysis system was established, using standard... A turbulence model and component transport equations were used to establish a dataset of hydrogen leakage diffusion in oxygen, with the leakage source being the interior of the anode cavity of the electrolyzer and the leakage intensity set as small leakage (0.5 g / s) and medium leakage (5 g / s).

[0053] 4) Based on the leakage and diffusion results, the region with a concentration >4%vol was selected as the initial condition for the combustion and explosion simulation. The LES model and the G equation flame surface model were adopted, and the ignition source was set as an electrical spark (energy 50mJ). The safety boundary of hydrogen combustion and explosion in oxygen was determined, and a dataset of consequences of hydrogen combustion and explosion accidents in oxygen was established.

[0054] 5) Electrolytic cell spacing, ventilation layout, and monitoring point arrangement were used as optimization variables. The optimization objectives were: leak detection time <30s, combustible cloud volume <100m³, and minimum equipment investment cost. The NSGA-II algorithm was used with a population size of 100 and 200 iterations to obtain the Pareto front solution set.

[0055] 6) Integrate HTO test data with oxygen-hydrogen leakage diffusion data to construct a comprehensive dataset. Use the XGBoost algorithm for feature selection, identifying eight key features including current density fluctuation rate, membrane impedance, anode pressure, and ventilation rate. Construct an LSTM network (8 neurons in the input layer, 2 hidden layers with 64 neurons each, and 1 neuron in the output layer), with a time window set to 5 minutes.

[0056] 7) Set three-level early warning thresholds: Level 1 (1.0% vol), Level 2 (2.0% vol), and Level 3 (3.0% vol). The oxygen-hydrogen concentration analyzer on the anode side collects real-time data on the oxygen-hydrogen leakage real-time monitoring model; simultaneously, a distributed fiber optic hydrogen sensor network (5m spacing) is deployed in the electrolysis workshop to collect concentration data in real time. When the oxygen-hydrogen concentration reaches the Level 1 warning level, the system automatically reduces the electrolyzer load by 20% and activates backup ventilation; when the Level 2 warning level is reached, forced ventilation is implemented and the leak source is investigated; when the Level 3 warning level is reached, the power supply is immediately cut off and inert gas purging is initiated. A central monitoring platform is established to achieve fully automated control of the monitoring-early warning-response process.

[0057] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. It will be apparent to those skilled in the art that various modifications can be made to the above embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A method for monitoring and arranging a hydrogen-electric coupling system for hydrogen leakage in oxygen, characterized in that, include: An HTO test platform was built, and real-time monitoring of hydrogen concentration in oxygen under different operating conditions was carried out based on the data set of operating characteristics of hydrogen-electric coupling system under fluctuating wind and solar power output. A basic dataset of hydrogen leakage in oxygen was established. Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer and a macroscale diffusion model of the system are constructed to conduct hydrodynamic simulation analysis of hydrogen leakage and diffusion behavior in oxygen and establish a dataset of hydrogen leakage and diffusion in oxygen. Based on the oxygen-hydrogen leakage and diffusion dataset, combined with the characteristics of the ignition source and the geometric constraints of the confined space, a fluid dynamics simulation analysis of an oxygen-hydrogen combustion and explosion accident is conducted to determine the safety boundary of an oxygen-hydrogen combustion and explosion accident and to establish a dataset of consequences of an oxygen-hydrogen combustion and explosion accident. Based on the datasets of hydrogen leakage and diffusion in oxygen and hydrogen combustion and explosion accident consequences in oxygen, the electrolytic cell layout scheme is optimized using the spacing between electrolytic cells, ventilation layout and monitoring point arrangement as optimization variables, and the optimization objectives are minimizing the volume of combustible cloud, minimizing the leakage detection time and minimizing the equipment investment cost. Based on HTO test data and oxygen-hydrogen leakage diffusion dataset, a real-time monitoring model for oxygen-hydrogen leakage was trained using machine learning algorithms to achieve rapid prediction of oxygen-hydrogen concentration and leakage source location under fluctuating wind and solar power output conditions. Based on the safety boundary of hydrogen combustion and explosion in oxygen and the real-time monitoring model of hydrogen leakage in oxygen, an early warning mechanism and emergency response strategy for hydrogen leakage in oxygen based on risk classification are established.

2. The method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to claim 1, characterized in that, The process of establishing the operating characteristic dataset of the hydrogen-electric coupling system specifically includes: Collect historical wind and solar power output data for the target area, analyze the daily characteristics, seasonal characteristics and extreme fluctuation events of wind and solar power output, and extract characteristic parameters of typical operating conditions, including power fluctuation rate, fluctuation amplitude and duration. Based on the topology and control strategy of the hydrogen-electric coupling system, a dynamic simulation model of the system is established to simulate the variation of the operating parameters of the electrolyzer under different wind and solar power output fluctuations. Design an HTO test matrix covering typical and extreme operating conditions. The test conditions include normal fluctuation conditions, rapid load reduction conditions, rapid loading conditions, and intermittent operation conditions. Establish a dataset of the operating characteristics of the hydrogen-electric coupling system under fluctuating wind and solar power output conditions.

3. The method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to claim 1, characterized in that, The process of establishing the basic dataset for hydrogen leakage in oxygen specifically includes: Based on the membrane aging mechanism and mechanical damage mode of alkaline electrolyzers, the main causes of hydrogen-oxygen crosstalk were identified, including membrane thinning, pinhole formation and sealing failure. Based on the aforementioned factors, an HTO test platform was constructed, comprising an alkaline electrolyzer, an anode gas circulation system, a cathode gas circulation system, a cooling system, and a control system. Hydrogen and oxygen generated on the cathode and anode sides of the alkaline electrolyzer, respectively, enter the cathode and anode gas circulation systems. The control system outputs control parameters related to current, voltage, pressure, and flow rate to the alkaline electrolyzer, directly managing its operating conditions. The control system also outputs temperature control commands to the cooling system, ensuring that the electrolyzer's operating temperature meets the system's temperature control requirements. An oxygen-hydrogen concentration analyzer is installed on the anode side of the electrolyzer, and a hydrogen-oxygen concentration analyzer is installed on the cathode side. Based on the operating characteristic dataset of the hydrogen-electric coupling system, HTO tests were conducted under different current densities, operating temperatures, and membrane aging degrees to obtain the evolution law of hydrogen concentration in oxygen with operating parameters and establish a basic dataset of hydrogen leakage in oxygen.

4. The method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to claim 1, characterized in that, The process of establishing the oxygen-hydrogen leakage diffusion dataset specifically includes: Based on the basic dataset of hydrogen leakage in oxygen, a microscale mass transfer model inside the electrolyzer was established to obtain the transmembrane hydrogen flux. Based on the three-dimensional geometric model of the hydrogen-electric coupling system, a macroscopic computational fluid dynamics diffusion model was established. The transmembrane hydrogen flux was used as the input of the boundary condition, and the inside of the anode cavity of the electrolyzer was used as the leakage source. Considering the coupling effect of mechanical ventilation and natural ventilation, fluid dynamics simulation analysis was carried out under different leakage intensities and different ventilation conditions to obtain the three-dimensional distribution characteristics of hydrogen concentration in the system. By coupling the transmembrane hydrogen flux and the three-dimensional distribution characteristics of hydrogen concentration in the system, a dataset of hydrogen leakage diffusion in oxygen was established.

5. The method for monitoring and arranging a hydrogen-electric coupling system for monitoring hydrogen leakage in oxygen according to claim 1, characterized in that, The process of establishing the dataset of consequences of hydrogen combustion and explosion accidents in oxygen specifically includes: Based on the oxygen-hydrogen leakage and diffusion dataset, the range and concentration distribution of flammable hydrogen clouds under different leakage scenarios were determined. Using the large eddy simulation model and the flame surface tracking model, the hydrodynamic simulation of combustion and explosion accidents under different ignition locations, ignition times, and oxygen-hydrogen concentrations was carried out to obtain the deflagration-to-detonation characteristics, flame propagation velocity, overpressure peak value, and temperature field distribution. The critical concentration and critical ignition energy of oxygen-hydrogen combustion and explosion were determined, and a dataset of consequences of oxygen-hydrogen combustion and explosion accidents was established.

6. The method for monitoring and arranging a hydrogen-electric coupling system for monitoring hydrogen leakage in oxygen according to claim 1, characterized in that, The specific process of optimizing the layout of electrolytic cells includes: Using the spacing between electrolyzers, ventilation layout, and monitoring point arrangement as optimization variables, and minimizing the volume of combustible clouds, the shortest leakage detection time, and the minimum equipment investment cost as optimization objectives, a multi-objective optimization function is constructed. The NSGA-II or MOEA / D multi-objective optimization algorithm is used, combined with the oxygen-hydrogen leakage diffusion dataset and the combustion and explosion accident consequence dataset, to search for the Pareto optimal solution and obtain the optimal electrolyzer layout scheme that balances safety and economy.

7. The method for monitoring and arranging a hydrogen-electric coupling system for monitoring hydrogen leakage in oxygen according to claim 1, characterized in that, The construction process of the real-time monitoring model for hydrogen leakage in oxygen specifically includes: By integrating HTO test data and oxygen-hydrogen leakage diffusion datasets, a comprehensive dataset containing operating parameters, environmental parameters, and historical leakage characteristics is constructed. Random forest or gradient boosting tree algorithms are used to perform feature importance analysis on all feature variables in the comprehensive dataset, screening for sensitive features related to oxygen-hydrogen leakage. Using the screened sensitive features and their time-series data as input, a mapping relationship between wind and solar power output fluctuations, operating parameters, and oxygen-hydrogen concentration is established using a long short-term memory neural network or Transformer model. This yields a real-time monitoring model for oxygen-hydrogen leakage, enabling rapid prediction of oxygen-hydrogen concentration and leak source location under fluctuating wind and solar power output conditions.

8. The method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to claim 7, characterized in that, The operating parameters include current, voltage, temperature, and pressure; the environmental parameters include ambient temperature, humidity, and ventilation rate; and the oxygen-hydrogen leakage sensitivity features include current density fluctuation rate, membrane impedance change rate, and anode pressure.

9. The method for monitoring and arranging hydrogen leakage in oxygen in a hydrogen-electric coupling system according to claim 1, characterized in that, The process of establishing an early warning mechanism and emergency response strategy for hydrogen leakage in oxygen based on risk classification specifically includes: Based on the critical concentration of hydrogen combustion and explosion in oxygen and the real-time monitoring model of hydrogen leakage in oxygen, a three-level early warning threshold is set; for different early warning levels, differentiated emergency response strategies are formulated: Level 1 early warning initiates automatic load reduction and enhanced ventilation, Level 2 early warning executes forced ventilation and leak investigation, and Level 3 early warning triggers emergency shutdown and personnel evacuation; an early warning-response linkage mechanism is established to achieve closed-loop management of monitoring-early warning-response.

10. The method for monitoring and arranging a hydrogen-electric coupling system for monitoring hydrogen leakage in oxygen according to claim 9, characterized in that, The three-level warning thresholds are as follows: Level 1 warning, which is the attention level, is when the hydrogen concentration in oxygen is ≥1%vol, i.e., 25%LFL; Level 2 warning, which is the warning level, is when the hydrogen concentration in oxygen is ≥2%vol, i.e., 50%LFL; and Level 3 warning, which is the danger level, is when the hydrogen concentration in oxygen is ≥3%vol, i.e., 75%LFL.

Images