Method for predicting hydrogen leakage range of hydrogen production and refueling station based on meshless particle method simulation

Abstract

The invention relates to a hydrogen production and refueling station hydrogen leakage range prediction method based on meshless particle method simulation, and belongs to the technical field of hydrogen energy safety and numerical simulation. Aiming at the problems that a traditional CFD grid method is poor in complex geometric boundary adaptability, low in calculation efficiency and insufficient in dynamic early warning capacity, a three-dimensional scene of a hydrogen production and refueling station is constructed by combining a geographic information system (GIS) and a building information model (BIM), a mathematical model is discretized by adopting a smoothed particle hydrodynamics (SPH) method, and the three-dimensional scene of the hydrogen production and refueling station is obtained. Real-time sensor data are integrated, and simulation is accelerated in parallel through a GPU. And carrying out numerical simulation of hydrogen leakage of the hydrogen production / hydrogen refueling station under multiple factors in combination with large eddy simulation, and generating high-precision simulation of evolution and distribution of a hydrogen leakage concentration field of a hydrogen production / hydrogen refueling station scene. The method has the advantages of being high in modeling speed, high in simulation precision, wide in applicability and the like, the problem of precise simulation of hydrogen leakage and diffusion under the complex geometric structure of the hydrogen production and refueling station is effectively solved, and reliable technical support is provided for safe operation of the hydrogen production and refueling station.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrogen energy safety and numerical simulation, and relates to a method for predicting the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle method simulation. Background Art

[0002] As a major secondary clean energy with abundant sources, green and low carbon, and wide applications, hydrogen energy is of great significance in building a clean, low-carbon, safe and efficient energy system. It has been widely used in industrial production, transportation, energy storage and other fields. However, hydrogen has the characteristics of low density, fast diffusion speed, flammability and explosiveness. During the preparation, storage, transportation and use of hydrogen, the complex geographical environment, building structure and fluid dynamics behavior are intertwined, and the high-pressure hydrogenation system has a great risk of leakage. Once a leak occurs, it is very likely to cause serious fires and explosions, posing a serious threat to industrial production safety. Therefore, it is particularly important to monitor the hydrogen concentration, leakage range and dangerous areas of hydrogen production and hydrogenation stations. How to quickly predict and efficiently control the leakage range and diffusion of hydrogen in hydrogen production and hydrogenation stations to reduce the possibility of explosions and their losses has become a key issue that needs to be solved urgently. Currently, traditional hydrogen leak warning systems primarily rely on real-time monitoring using physical sensing devices such as distributed sensor networks and fiber-optic monitoring technology. These systems suffer from technical pain points such as high monitoring and warning costs, weak adaptability to external interference, high false alarm rates, a lack of forward-looking risk prediction capabilities, and the difficulty of fixed monitoring points in capturing transient diffusion processes. These technical shortcomings have led to the prominent problem of "high investment and low efficiency" in traditional warning systems, making them unable to meet the safety and control requirements of modern hydrogen production and refueling station infrastructure. The development of a new intelligent warning technology system is urgently needed.

[0003] Numerical simulation technology provides a new path for quickly and efficiently predicting hydrogen leakage and diffusion behavior. Currently, numerical simulations of hydrogen leakage at hydrogen production and refueling stations mostly use grid methods (such as the finite volume method (FVM) and the finite element method (FEM)) or simplified empirical models, which have been widely used in areas such as combustion characteristics and monitoring and early warning. The invention patent (CN 115206443 A) proposes a method and system for predicting the spontaneous combustion of high-pressure hydrogen leakage using CFD simulations, which helps improve the accuracy of simulations of high-pressure hydrogen leakage and spontaneous combustion. The invention patent (CN 118153948 B) proposes a method and device for hydrogen energy storage device leakage fault warning based on deep learning, which can effectively solve key problems in hydrogen energy storage device leakage fault warning and improve the accuracy and reliability of hydrogen energy storage device leakage warning. The invention patent (CN 117313191 A) proposes a large-scale underground garage hydrogen leakage prediction and risk assessment method based on the informer model, which can perform high-precision and high-efficiency predictions on long time series and simulate the spatiotemporal evolution of hydrogen leakage. An invention patent (CN 117391126 A) discloses an intelligent and efficient prediction method for hydrogen leakage and diffusion based on deep learning of graph physics information. By mapping the physical information of hydrogen jets with hydrogen diffusion characteristics, the method can quickly and accurately predict the concentration and velocity fields of hydrogen leakage and diffusion. An invention patent (CN 118936757 A) discloses a detection and early warning system for hazardous gas leaks in hydrogen refueling stations. This system solves the problem of leak point detection by using tetrahedral hydrogen detection sensors.

[0004] In summary, it is found that the existing technology has the following deficiencies:

[0005] The CFD simulation method and system for predicting the spontaneous combustion of high-pressure hydrogen leakage proposed in the invention patent (CN 115206443 A) has difficulty in meshing the complex geometric structures (pipelines, valves, and storage tanks) of hydrogen production stations, and it is difficult to dynamically capture the transient diffusion process of hydrogen leakage; the deep learning-based hydrogen energy storage equipment leakage fault warning method and device proposed in the invention patent (CN 118153948 B) lacks a high-precision dynamic description of the coupling effect of hydrogen diffusion and turbulent mixing, and cannot accurately predict the critical conditions for explosion. The Informer model-based hydrogen leakage prediction and risk assessment method for large underground garages proposed in the invention patent (CN 117313191 A) has low computational efficiency and cannot be combined with real-time monitoring data for dynamic risk warning. The intelligent and efficient prediction method for hydrogen leakage and diffusion based on deep learning of graph physics information disclosed in the invention patent (CN 117391126 A) has poor adaptability to complex geometric boundaries (such as hydrogen storage tanks and pipeline valves) and is highly dependent on the grid, making it difficult to dynamically capture the transient process of leakage and diffusion. The detection and early warning system for hazardous gas leaks in hydrogen refueling stations disclosed in the invention patent (CN 118936757 A) relies too much on sensors for physical real-time monitoring, and has problems such as high technical cost, great difficulty, and many interference factors.

[0006] In existing technologies, simulations of hydrogen leakage in hydrogen production stations mostly use grid methods (such as FVM and FEM) or simplified empirical models, which have the following limitations:

[0007] The hydrogen leakage and diffusion process is subject to the complex influence of environmental factors, building structure and airflow distribution. Existing leakage assessment methods are mostly based on experimental tests or traditional CFD (computational fluid dynamics) simulation calculations. Traditional hydrogen leakage assessment methods such as experimental simulation and simplified models have the disadvantages of high cost, long cycle and low accuracy. Traditional CFD numerical simulation methods have poor adaptability to complex geometric boundaries (such as hydrogen storage tanks and pipeline valves) and are highly dependent on the grid, making it difficult to dynamically capture the transient process of leakage diffusion. Traditional CFD methods (such as the finite volume method) have difficulty in meshing and processing complex boundaries in complex scenarios of hydrogen production stations / refueling stations, resulting in complex modeling and long simulation time. Traditional CFD numerical simulation methods have poor environmental adaptability. Complex terrain and meteorological conditions (such as wind speed and temperature gradient) have a significant impact on hydrogen diffusion, but existing CFD models are difficult to quickly integrate multi-dimensional environmental parameters, resulting in limited prediction accuracy. The leakage range estimation based on empirical formulas has large errors, especially in open or semi-open scenarios. The reliability is insufficient. Traditional CFD numerical simulation methods are highly grid-dependent, making them extremely difficult to handle complex boundary conditions and multi-physics coupling. Furthermore, their reliance on periodic data collection results in inefficient and inefficient data collection, making it difficult to comprehensively and rapidly capture the dynamic processes and diffusion behaviors of hydrogen leaks. Furthermore, traditional CFD methods suffer from high costs, long cycles, low computational efficiency, and low accuracy, making them inadequate for hydrogen leak concentration prediction. Therefore, an efficient and accurate assessment method and system are urgently needed. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide a method for predicting the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation. This method comprehensively considers factors such as the surrounding environment, building structure, and equipment layout of the hydrogen production and refueling station, and quickly and accurately constructs a three-dimensional model of the hydrogen production and refueling station scene based on the geographic information system (GIS) and building information model (BIM) method. It combines SPH simulation to achieve accurate simulation and risk assessment of the hydrogen leakage diffusion process, and uses image processing technology to identify dangerous areas. This method can quickly and efficiently predict the hydrogen diffusion range of hydrogen production and refueling station scenes in different geographical environments and meteorological conditions, and provide technical support for the safety evaluation and emergency plan formulation of hydrogen production and refueling stations.

[0009] In order to achieve the above object, the present invention provides the following technical solutions:

[0010] A method for predicting hydrogen leakage range of a hydrogen production and refueling station based on a gridless particle simulation method includes the following steps:

[0011] S1: Establish a 3D environmental model around the hydrogen production and refueling station, including: Establish a 3D environmental model around the hydrogen production and refueling station. Acquire accurate geographic and environmental data and process the data to generate a detailed 3D environmental model. Process and reconstruct the collected data to generate an accurate 3D environmental model. The model includes geospatial data (topography, geographic coordinates, buildings and structures, hydrological data, roads and traffic), infrastructure data (underground pipelines, surrounding buildings and facilities), and environmental and risk data (meteorological data: air quality data, climate change data).

[0012] S2: Build a 3D physical model of the hydrogen production and refining plant. This involves creating a 3D physical model of the plant's interior based on the plant's architectural design drawings, field data, and other information. This model includes not only the spatial layout of facilities such as gas storage vessels, compressors, pipelines, and valves, as well as the operating status of these equipment, but also the plant's structural layout, walls, doors and windows, vents, and equipment distribution.

[0013] S3: Build a 3D physical model of the hydrogen production and refueling station scenario. This involves integrating the 3D environmental model and plant model established in Steps 1 and 2 to construct a complete 3D scene model encompassing the station's surroundings, interior, and equipment. This model's coordinate systems are unified, and the environmental model from S1 is spatially aligned with the plant model from S2 to ensure consistent coordinate systems, resulting in a complete 3D physical scene for the hydrogen production and refueling station. Furthermore, the model undergoes a comprehensive inspection to ensure its completeness, accuracy, and error-free nature.

[0014] S4: 3D data collection and integration of the hydrogen production and refueling station environment. This involves utilizing a sensor network to collect real-time environmental parameters within the station, including temperature, humidity, wind speed, wind direction, and hydrogen concentration. This collected environmental data is integrated with the 3D scene model established in S3 to correlate environmental parameters with spatial information. Using data fusion and interpolation techniques, the environmental data is processed and analyzed to provide accurate boundary and initial conditions for subsequent numerical simulations. This enables real-time monitoring and updating of environmental data, enabling timely response to the impact of environmental changes on hydrogen diffusion.

[0015] S5: Establish a complete mathematical model of hydrogen leakage in complex scenarios at hydrogen production and refueling stations, including control equations, turbulence models, boundary conditions, and numerical model verification. The control equations include continuity equations, momentum equations, energy equations, diffusion models, etc. Introduce the large eddy simulation (LES) method to describe the hydrogen diffusion process, set appropriate boundary conditions according to actual conditions, such as the leakage rate of the leakage source, the shape of the leakage port, ambient temperature, wind speed, etc., and consider the hydrogen buoyancy effect and air resistance. Use existing experimental data or standard examples to verify the established mathematical model to ensure the accuracy and reliability of the model.

[0016] S6: Construction and parameter setting of the SPH numerical discretization format for the mathematical model, including the use of the smoothed particle hydrodynamics (SPH) method to discretize the mathematical model established in S5. Reasonable settings of SPH method parameters such as particle spacing, smoothing length, and time step are performed based on the actual scenario and simulation requirements. Adaptive particle refinement technology is used to increase particle density in areas with large hydrogen concentration gradients to improve simulation accuracy. Appropriate boundary condition processing methods are used to simulate fluid flow behavior in complex geometric environments. The numerical dissipation and stability of the SPH method are analyzed and optimized to improve the accuracy and reliability of the simulation results.

[0017] S7: Validation of the SPH-based numerical simulation method for hydrogen leakage at a hydrogen production and refueling station. This involves validating the SPH numerical simulation method developed in S6 using existing experimental data or standard examples. Simulation results are compared with experimental data to evaluate the accuracy and reliability of the simulation method. Sensitivity analysis is conducted to investigate the impact of simulation parameters on the results and determine the optimal parameter combination. The computational efficiency of the simulation method is evaluated, and algorithms and procedures are optimized to increase computational speed.

[0018] S8: Conduct numerical simulations of hydrogen leakage at hydrogen production and refueling stations under multiple factors. This includes using the numerical simulation method validated in S7 to simulate hydrogen leakage and diffusion under various leakage parameters (leak source location, leak port size, leak pressure, and leak flow), environmental parameters (temperature, humidity, wind speed, and direction), equipment layout parameters (leak source location, pipeline layout, and ventilation system), and other factors (surrounding buildings, obstacles, and topography). Using the SPH simulation method, combined with GPU acceleration technology (such as CUDA) to parallelize multiple simulation tasks, sensitivity analysis will be conducted on the diffusion range, concentration distribution, and risk level of hydrogen leakage under different leakage scenarios, environmental conditions, and equipment layouts to determine the impact of each factor on hydrogen leakage and diffusion.

[0019] S9: Outputs a dynamic distribution cloud map of the hydrogen leakage concentration field in the hydrogen production and refueling station scenario. This involves visualizing the hydrogen concentration field data obtained in S8 to generate a dynamic distribution cloud map. Appropriate color mapping and contour line display methods are used to highlight hazardous areas with high hydrogen concentrations. Animation is used to display the dynamic process of hydrogen diffusion, enhancing the visualization effect.

[0020] S10: Post-processing software is used to process data and images, determine hazardous areas based on thresholds, and generate warning signs. This includes: using post-processing software to analyze the cloud map data generated in S9 and determine the extent of the hydrogen leakage hazard area based on the set hydrogen concentration threshold. For areas exceeding the safety threshold, warning signs are generated on the cloud map, clearly indicating the location and size of the hazardous area.

[0021] Furthermore, S1 specifically includes the following steps:

[0022] S11: Use laser radar scanners (LiDAR), drone remote sensing, photogrammetry and other technologies to perform high-precision scanning of environmental elements such as terrain, buildings, vegetation and meteorological observation stations around the hydrogen production and refueling station to obtain point cloud data (accuracy ≤ 5cm) and geographic coordinate information. Obtain high-precision three-dimensional point cloud data of the terrain and landforms (such as slope and elevation), building distribution (such as surrounding residential buildings and commercial facilities), road traffic (such as road width and vehicle flow), vegetation cover, obstacles (such as walls and utility poles) around the hydrogen production and refueling station. For specific areas, combine geographic information system (GIS) data to supplement geographic location, terrain elevation and other information. Use panoramic cameras to collect high-definition images for texture mapping and model authenticity verification.

[0023] S12: Perform pre-processing such as denoising, filtering, and registration on the collected point cloud data, and perform image matching, sparse reconstruction, and dense reconstruction on the photogrammetric data to generate a high-quality point cloud model. Use point cloud processing software (such as CloudCompare and Cyclone) to perform point cloud segmentation and feature extraction to identify buildings, roads, vegetation, etc. Use 3D modeling software (such as AutoCAD, Blender, and 3ds Max) to reconstruct the surface of the pre-processed point cloud or mesh model to generate a 3D digital terrain model (DTM) containing elevation, slope, and obstacle distribution. Use GIS technology to establish a 3D environmental model around the hydrogen production and refueling station, including geospatial data, infrastructure data, and environmental and risk data. Perform detailed modeling of buildings, roads, terrain, etc. to ensure the geometric accuracy and topological relationship of the model. Perform texture mapping on the model to increase the realism and visualization effect of the model.

[0024] S13: Compare the generated 3D model with field photos, maps, and other data to verify the model's accuracy and completeness. Use GIS data and field measurements to verify key dimensions, locations, and other information in the model to ensure a high degree of consistency with actual conditions. Simplify and optimize the model to reduce its complexity and improve the computational efficiency of subsequent numerical simulations.

[0025] Furthermore, S2 specifically includes the following steps:

[0026] S21: Obtain detailed information such as architectural design drawings, construction drawings, equipment layouts, and piping layouts for the hydrogen production and refueling plant. Use a laser scanner or 3D measuring instrument to capture plant layout information, including key information such as equipment location, piping routing, hydrogen storage tanks, compressors, valves, and electrical equipment. This captured equipment information includes the size, function, and location of each piece of equipment within the plant.

[0027] S22: Based on the obtained drawings and measured data, BIM technology is used and Revit is used to build a three-dimensional model of the hydrogen production and refueling station plant, including detailed data such as the building structure, equipment information, and pipeline layout in the station, to accurately reflect the structural layout, walls, doors and windows, vents, equipment distribution and other information of the plant. BIM technology is used to perform refined three-dimensional modeling of key equipment in the plant (such as hydrogen storage tanks, hydrogenation machines, pipelines, valves, etc.), including their shape, size, position, etc., and integrate physical properties (material, pressure, volume), and use parametric design tools (such as SolidWorks) to define equipment size, material and connection relationships. Physical properties are given to the equipment in the model, including material thermal conductivity, compressive strength, sealing performance, etc., and potential leakage risk points (such as welds and flange interfaces) are marked.

[0028] S23: Define the material properties of the model, setting parameters such as thermal conductivity and specific heat capacity for different materials. Simplify and optimize the model to reduce its complexity and improve the computational efficiency of subsequent numerical simulations. This physical model should undergo multiple verifications and optimizations to ensure it accurately reflects the actual conditions within the hydrogen production and refueling station.

[0029] Furthermore, S3 specifically includes the following steps:

[0030] S31: Data Format and Standardization. GIS data is typically stored in geographic data formats (such as Shapefile, GeoJSON, KML, etc.) and can be spatially analyzed and visualized using GIS software. BIM data uses standard BIM file formats, such as Revit .RVT files, IFC (Industry Foundation Classes) files, and 3D DWG files. During the integration process, the coordinate systems of GIS and BIM data must be unified (such as WGS84, UTM, etc.) to ensure that the two can be correctly connected and merged.

[0031] S32: 3D Modeling and Scene Integration. Using ArcGIS Pro, a software platform supporting GIS and BIM data integration, the data from the GIS model in Step 1 and the BIM model in Step 2 were integrated to construct a complete 3D scene model encompassing the surrounding environment, plant interior, and station equipment of the hydrogen production and refueling station. This model included geographic spatial information, building facility information, hydrogen production and refueling equipment information, and pipeline layout information to ensure the accuracy and completeness of the simulated environment. The spatial relationships between the various parts of the model were ensured to be correct, avoiding model overlap or omissions.

[0032] S33: Unify the coordinate systems of all parts of the model, spatially align the S1 environmental model with the S2 plant model to ensure coordinate consistency, and form a complete 3D physical scene of the hydrogen production and refueling station. Align and lightweight the GIS and BIM data to generate a high-precision 3D visualization scene that supports dynamic rendering and interactive operations. Define the scene boundaries of the hydrogen production and refueling station and determine the simulation calculation area. Define the location, shape, size, and other information of the leak source.

[0033] S34: Conduct a comprehensive inspection of the integrated 3D scene model to ensure that it is complete, accurate, and error-free. Use visualization software to browse and inspect the model to ensure that it conforms to the actual situation.

[0034] Furthermore, S4 specifically includes the following steps:

[0035] S41: Hydrogen concentration sensors (accuracy ±1%LEL), temperature sensors (±0.5℃) and anemometers (±0.1m / s) are installed in key areas of hydrogen production and refueling stations, such as hydrogen storage areas, refueling areas, ventilation openings, and inside factory buildings to form a real-time monitoring network.

[0036] S42: Real-time collection of environmental parameters, including temperature, humidity, wind speed, wind direction, hydrogen concentration, and pressure. Sensors should be deployed across hydrogen production and refueling stations. A data acquisition system collects sensor data in real time and stores it in a database. The collected data is timestamped to ensure data consistency and relevance. Through the sensor network, the data is transmitted to an integrated GIS platform for real-time updates of environmental parameters.

[0037] S43: Aggregates data from various sensors and uses data processing software to clean, analyze, and process the real-time data. A Kalman filter algorithm is used to clean noisy data to ensure input data reliability. Data fusion and interpolation techniques are used to process and analyze environmental data, providing accurate boundary and initial conditions for subsequent numerical simulations. Meteorological data (wind speed, direction, temperature, humidity) and geographic information (such as slope and obstacle distribution) are integrated with the 3D scene model established in S3 through an API interface, linking environmental parameters with spatial information.

[0038] Furthermore, S5 specifically includes the following steps:

[0039] S51: The control equations include the continuity equation, the momentum equation, and the energy equation. The continuity equation describes the equation for conservation of hydrogen mass and is used to calculate the distribution changes of hydrogen in space and time. The momentum equation (Navier-Stokes equation) describes the equation for hydrogen flow and is used to calculate the velocity and pressure distribution of hydrogen in the environment. The energy equation is used to calculate the temperature distribution of hydrogen during the diffusion process.

[0040] Continuity equation:

[0041]

[0042] Where ρ is the hydrogen density, t is the time, and u is the velocity vector. represents the divergence operator;

[0043] Momentum equation:

[0044]

[0045] Where P is the pressure, τ is the viscous stress tensor, g is the gravitational acceleration vector, represents the tensor product;

[0046] Energy equation:

[0047]

[0048] Where T is temperature, C P is the specific heat capacity at constant pressure, k is the thermal conductivity, is the temperature gradient, represents the viscous dissipation term.

[0049] S52: Turbulence model: The large eddy simulation (LES) model is selected to describe the concentration change process of hydrogen during the diffusion process.

[0050] S53: Diffusion model: A calculation model based on the real gas state equation developed by NIST (National Institute of Standards and Technology) is selected to describe hydrogen. It is used to calculate the relationship between density, temperature and pressure during the diffusion of hydrogen in air.

[0051] S54: Equation of state: describes the relationship between hydrogen density, temperature and pressure.

[0052] S55: Leakage model. The leakage concentration is described using a non-point source Gaussian plume mixture model. The leakage intensity requires the establishment of a corresponding leakage model based on the specific leakage scenario, such as an orifice leakage model or a pipeline leakage model.

[0053] S56: Define initial and boundary conditions. Leaks are described using a pressure outlet, with calculations performed based on the shape of the leak. Other conditions are free boundaries.

[0054] Furthermore, S6 specifically includes the following steps:

[0055] S61: SPH discretization of the equation model. Using a quintic spline smoothing function with a variable smoothing length, the mass conservation equation, the momentum conservation equation, the energy conservation equation, and the diffusion equation are discretized.

[0056] S62: SPH model parameters: Based on the actual scenario and simulation requirements, the SPH method parameters such as particle spacing, smoothing length, time step, and artificial viscosity coefficient are optimized through orthogonal experimental method.

[0057] S63: Boundary Condition and Source Term Processing. At each time step, boundary conditions and source terms must be processed. For a leak source, a source term is typically set to simulate the amount of hydrogen released, and the leak rate is updated along with the concentration of surrounding particles. At the same time, the particle states at the boundary need to be updated based on reflections or open boundary conditions.

[0058] S64: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the particle injection position is set according to the leakage source position, and the particle injection amount is calculated according to the leakage port size, leakage pressure and leakage rate.

[0059] S65: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the positions of the particles are distributed according to the requirements of the physical model, and the initial parameters of the air particles are defined, including density, viscosity, diffusion coefficient, temperature, rate, direction, etc.; the parameters of the hydrogen leakage source are defined, including thermal conductivity, density, viscosity, diffusion coefficient, temperature, rate, and direction, and the particle motion equations of air and hydrogen are dynamically corrected in combination with real-time meteorological data (wind speed, temperature).

[0060] S66: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the initial conditions are set to include wind speed, wind direction, temperature and humidity, etc., and the boundary conditions are set to a no-slip boundary.

[0061] S67: The method for efficiently predicting hydrogen leakage range at a hydrogen production and refueling station based on meshless particle simulation according to claim 1, characterized in that, in S5, the time step is updated by iterative processing using a fourth-order Runge-Kutta method, and the particle state is updated with the time step. Each time, information such as the particle velocity, concentration, pressure, and temperature is calculated, and the results are updated to the next time step.

[0062] S68: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the SPH simulation software is DualSPHysics, which performs simulation based on a GPU parallel computing architecture.

[0063] Furthermore, S7 specifically includes the following steps:

[0064] S71: Verification Objective: Verify the accuracy of the SPH numerical simulation method in simulating the hydrogen leakage and diffusion behavior of hydrogen production and refueling stations, evaluate the applicability of the SPH model to different leakage scenarios, environmental conditions, and equipment layouts, and determine the parameter settings and calculation accuracy requirements of the SPH model.

[0065] S72: Verification Method. Leverage existing hydrogen leak experimental data to compare SPH simulation results with the experimental data. Comparison parameters include hydrogen concentration distribution, diffusion range, diffusion velocity, and temperature distribution. For some simple leak scenarios, analytical solutions (theoretical solutions) can be compared with SPH simulation results. Leverage a hydrogen leak accident at a hydrogen production and refueling station to compare SPH simulation results with the accident investigation report.

[0066] S73: Verification parameters. These mainly include leakage parameters (leak size, leakage pressure, leakage rate), environmental parameters (wind speed, wind direction, temperature, humidity, atmospheric stability), equipment layout parameters (leak source location, pipeline layout, ventilation system), and SPH model parameters (kernel function type, particle spacing, time step, turbulence model parameters).

[0067] S74: Validation indicators. These mainly include concentration error, diffusion range error, velocity error, and temperature error. Quantitative evaluation of these errors is achieved by calculating the root mean square error (RMSE) and correlation coefficient. If the error is ≤5%, the model is considered valid.

[0068] S75: Numerical Stability and Convergence Analysis. This section verifies numerical stability by examining whether the changes in physical quantities (such as hydrogen concentration and velocity) during the simulation are reasonable and whether they tend to stabilize over time. The convergence of the model is determined by evaluating the SPH simulation results as the interparticle spacing and time step change.

[0069] S76: Sensitivity Analysis. Perform a sensitivity analysis to examine the model's sensitivity to various input parameters (such as wind speed, leak source size, pipe material, etc.), leak source location and size, and ambient temperature. By analyzing the impact of different parameters on the simulation results, the stability and reliability of the SPH model are evaluated.

[0070] Furthermore, S9 specifically includes the following steps:

[0071] S91, save the hydrogen concentration field data obtained in S8 in VTK or CSV format, including timestamps (interval 0.1s), spatial coordinates, and concentration values.

[0072] S92 uses appropriate color mapping and contour display to highlight hazardous areas with high hydrogen concentrations. ParaView is used to generate 3D data, dynamic hydrogen concentration cloud maps, and videos. Using the Matplotlib library, Python scripts are used to create dynamic 3D concentration cloud maps, annotating concentration gradients and diffusion paths. The leak diffusion process is monitored at a time step (e.g., 0.1 second / frame), and obstacle information from the environmental model is overlaid.

[0073] Furthermore, S10 specifically includes the following steps:

[0074] S101: The Min-Max method is used to perform normalization processing on the hydrogen concentration field dynamic distribution data and the hydrogen concentration field distribution cloud map.

[0075] S102: Setting the threshold for determining hazardous areas. Referencing GB / T 34532 "General Technical Guidelines for Hydrogen Safety," GB 50177 "Design Specifications for Hydrogen Stations," and NFPA 2 "Technical Specifications for Hydrogen," set the threshold for determining hazardous areas based on the safety distances and concentration limits specified in these standards. A 4% volume concentration is used as the hazardous threshold; areas exceeding this value are marked as high-risk areas. Safety levels: Low-risk area: <2% (warning zone); Medium-risk area: 2% to 4% (intervention zone); High-risk area: ≥4% (evacuation zone).

[0076] S103: Region segmentation. Create a concentration threshold filter in post-processing software (such as the Threshold function in ParaView); segment regions based on the threshold and assign different color labels (red / yellow / green).

[0077] S104: Determine the hazardous area based on the threshold. Post-processing software analyzes the cloud image data generated in S9 to identify the spread of the hydrogen leak and determine the size of the leak area. A hydrogen concentration threshold (such as the lower explosion limit (LEL)) is set and the normalized hydrogen concentration is compared with the set threshold. If the hydrogen concentration exceeds the threshold, the area is determined to be hazardous.

[0078] S105: Generate warning signs. Alert levels are assigned based on factors such as hydrogen concentration and diffusion range. First, color coding is used to visually display the levels of hazardous areas. Red indicates high-risk areas, yellow indicates medium-risk areas, and blue indicates low-risk areas. Second, a timestamp, concentration scale, and risk level description are overlaid on the cloud map. Finally, the warning signs are integrated into the GIS system to enable real-time monitoring and visualization of hazardous areas.

[0079] S106: Visualization and Simulation Demonstration. Based on set concentration thresholds, safe, warning, and hazardous areas are demarcated, and corresponding warning signs are generated. Based on hydrogen leak simulation and hazardous area determination, visualization software can be used to present information such as zone demarcation and warning signs in graphical or animated form, helping personnel intuitively understand the spread and risk of the leak.

[0080] The beneficial effects of the present invention are:

[0081] The beneficial effects of the present invention are:

[0082] (1) Compared to traditional experimental testing methods, the simulation method based on the gridless particle method in this invention completely avoids the explosion risk of high-pressure hydrogen leakage experiments, eliminating the need for expensive experimental platforms and saving equipment investment and labor costs. By replacing physical experiments with numerical simulations, the operational complexity is reduced by approximately 60%, making it particularly suitable for rapid risk assessment in complex scenarios.

[0083] (2) The smoothed particle hydrodynamics (SPH) method, combined with a large eddy simulation (LES) turbulence model, solves the mesh distortion problem of traditional meshing methods in complex geometric boundaries such as hydrogen storage tanks and pipeline valves. Verified by experimental data, the root mean square error between the simulation results and the measured values is controlled within 5%, and the accuracy of leakage diffusion range prediction is improved by more than 20%, making it particularly suitable for dynamic capture of transient diffusion processes.

[0084] (3) Through GPU parallel acceleration technologies such as the CUDA architecture, the SPH particle calculation task is decomposed into thread blocks to handle local particle interactions. Global memory and shared memory are optimized for data access, reducing the single simulation time by 30% to 50% compared to traditional CFD methods. It supports rapid simulation analysis using multi-factor orthogonal experimental methods and can simultaneously process multiple parameter combinations such as leak diameters ranging from 2 mm to 20 mm, pressures ranging from 0.1 MPa to 70 MPa, and wind speeds ranging from 1 m / s to 15 m / s, meeting real-time dynamic warning requirements.

[0085] (4) A 3D scene model is constructed by integrating Geographic Information System (GIS), Building Information Model (BIM), and 5-centimeter-precision LiDAR point cloud data. Environmental parameters, including temperature, wind speed, and concentration, are dynamically corrected through a real-time sensor network. The model verification error is less than 3%, and the match between prediction results and actual scenes is improved by 25%, significantly enhancing reliability in complex terrain and weather conditions.

[0086] (5) Risk levels are divided according to the national standard GB / T 34532, and high-risk areas with hydrogen concentrations of 4% and above, medium-risk areas with concentrations of 2% to 4%, and low-risk areas with concentrations below 2% are dynamically marked in the three-dimensional scene. The LSTM neural network is used to predict the evolution trend of dangerous areas in the next 30 seconds, shortening the warning response time to seconds, providing an intuitive visualization basis for emergency response plans.

[0087] (6) Through orthogonal test method, the sensitivity weights of multiple factors such as leak diameter, pressure, wind speed, etc. to the area of explosion hazard zone are quantified, and a sensitivity analysis report is output. The scope of application covers open, semi-open and closed scenarios, providing quantitative data support for equipment layout optimization and ventilation system design of hydrogen production and refueling stations, and improving engineering adaptability by 40%.

[0088] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0090] Figure 1 This is a flow chart of a method for efficiently predicting the scope of hydrogen leakage in a hydrogen production and refueling station based on gridless particle simulation according to an embodiment of the present invention;

[0091] Figure 2 This is the hydrogen leakage and diffusion solution process based on the SPH method used in the embodiment of the present invention;

[0092] Figure 3 This is the main program framework flow of the SPH for hydrogen leakage and diffusion in the hydrogen production and refueling station used in the embodiment of the present invention;

[0093] Figure 4 The CUDA parallel accelerated computing method used in the embodiment of the present invention;

[0094] Figure 5This is a hydrogen leakage distribution cloud map of a hydrogen production and refueling station simulated based on the gridless particle method in an embodiment of the present invention. DETAILED DESCRIPTION

[0095] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0096] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0097] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0098] like Figure 1 As shown, the present invention provides an efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on a gridless particle method simulation, comprising the following steps:

[0099] Step 1: Establish a three-dimensional environmental model around the hydrogen production and refueling station. The specific process is as follows:

[0100] S101: Use laser radar scanners (LiDAR), drone remote sensing, photogrammetry and other technologies to perform high-precision scanning of environmental elements such as terrain, buildings, vegetation and meteorological observation stations around hydrogen production and refueling stations to obtain point cloud data (accuracy ≤ 5cm) and geographic coordinate information. Obtain high-precision three-dimensional point cloud data of the terrain and landforms (such as slope and elevation), building distribution (such as surrounding residential buildings and commercial facilities), road traffic (such as road width and vehicle flow), vegetation cover, obstacles (such as walls and utility poles) around the hydrogen production and refueling stations. For specific areas, combine geographic information system (GIS) data to supplement geographic location, terrain elevation and other information. Use panoramic cameras to collect high-definition images for texture mapping and model authenticity verification.

[0101] S102: Perform pre-processing such as denoising, filtering, and registration on the collected point cloud data, and perform image matching, sparse reconstruction, and dense reconstruction on the photogrammetry data to generate a high-quality point cloud model. Use point cloud processing software (such as CloudCompare, Cyclone) to perform point cloud segmentation and feature extraction to identify buildings, roads, vegetation, etc. Use 3D modeling software (such as AutoCAD, Blender, 3ds Max) to reconstruct the surface of the pre-processed point cloud or mesh model to generate a 3D digital terrain model (DTM) containing elevation, slope, and obstacle distribution. Use GIS technology to establish a 3D environmental model around the hydrogen production and refueling station, including geospatial data, infrastructure data, and environmental and risk data. Perform detailed modeling of buildings, roads, terrain, etc. to ensure the geometric accuracy and topological relationship of the model. Perform texture mapping on the model to increase the realism and visualization effect of the model.

[0102] S103: Compare the generated 3D model with field photos, maps, and other data to verify the model's accuracy and completeness. Use GIS data and field measurements to verify key dimensions, locations, and other information in the model to ensure a high degree of consistency with actual conditions. Simplify and optimize the model to reduce its complexity and improve the computational efficiency of subsequent numerical simulations.

[0103] Step 2: Establish a three-dimensional physical model of the hydrogen production and hydrogenation plant. The specific process is as follows:

[0104] S201: Obtain detailed information such as architectural design drawings, construction drawings, equipment layouts, and piping layouts for the hydrogen production and refueling plant. Use a laser scanner or 3D measuring instrument to collect plant layout information, including key information such as equipment location, piping routing, hydrogen storage tanks, compressors, valves, and electrical equipment. This collected equipment information includes the size, function, and location of each piece of equipment within the plant.

[0105] S202: Based on the obtained drawings and measured data, BIM technology is used and Revit is used to build a three-dimensional model of the hydrogen production and refueling station plant, including detailed data such as the building structure, equipment information, and pipeline layout in the station, to accurately reflect the structural layout, walls, doors and windows, vents, equipment distribution and other information of the plant. BIM technology is used to perform refined three-dimensional modeling of key equipment in the plant (such as hydrogen storage tanks, hydrogenation machines, pipelines, valves, etc.), including their shape, size, position, etc., and integrate physical properties (material, pressure, volume), and use parametric design tools (such as SolidWorks) to define equipment size, material and connection relationships. Physical properties are given to the equipment in the model, including material thermal conductivity, compressive strength, sealing performance, etc., and potential leakage risk points (such as welds and flange interfaces) are marked.

[0106] S203: Define the material properties of the model, setting parameters such as thermal conductivity and specific heat capacity for different materials. Simplify and optimize the model to reduce complexity and improve the computational efficiency of subsequent numerical simulations. This physical model should undergo multiple verifications and optimizations to ensure it accurately reflects the actual conditions within the hydrogen production and refueling station.

[0107] Step 3: Establish a 3D physical model of the hydrogen production and refueling station scenario. The specific process is as follows:

[0108] S301: Data Format and Standardization. GIS data is typically stored in geographic data formats (such as Shapefile, GeoJSON, KML, etc.) and can be spatially analyzed and visualized using GIS software. BIM data uses standard BIM file formats, such as Revit .RVT files, IFC (Industry Foundation Classes) files, and 3D DWG files. During the integration process, the coordinate systems of GIS and BIM data must be unified (such as WGS84, UTM, etc.) to ensure that the two can be correctly connected and merged.

[0109] S302: 3D Modeling and Scene Integration. Using ArcGIS Pro, a software platform supporting GIS and BIM data integration, the data from the GIS model in Step 1 and the BIM model in Step 2 are integrated to construct a complete 3D scene model of the hydrogen production and refueling station's surroundings, plant interior, and equipment. This model includes geographic spatial information, building and facility information, hydrogen production and refueling equipment information, and pipeline layout information to ensure the accuracy and completeness of the simulated environment. The spatial relationships between the various parts of the model are ensured to avoid overlap or omissions.

[0110] S303: Unify the coordinate systems of all parts of the model and spatially register the S1 environmental model with the S2 plant model to ensure coordinate consistency, thus forming a complete 3D physical scene of the hydrogen production and refueling station. GIS and BIM data are aligned and lightweighted to generate a high-precision 3D visualization scene that supports dynamic rendering and interactive operations. Define the scene boundaries of the hydrogen production and refueling station and determine the simulation calculation area. Define the location, shape, size, and other information of the leak source.

[0111] S304: Conduct a comprehensive inspection of the integrated 3D scene model to ensure that the model is complete, accurate, and error-free. Use visualization software to browse and inspect the model to ensure that the model conforms to the actual situation.

[0112] Step 4: 3D data collection and integration of hydrogen production and refueling station environment. The specific process is as follows:

[0113] S401: Hydrogen concentration sensors (accuracy ±1%LEL), temperature sensors (±0.5°C) and anemometers (±0.1m / s) are installed in key areas of the hydrogen production and refueling station, such as the hydrogen storage area, refueling area, ventilation holes, and inside the plant, to form a real-time monitoring network.

[0114] S402: Real-time collection of environmental parameters, including temperature, humidity, wind speed, wind direction, hydrogen concentration, and pressure. Sensors should be deployed across hydrogen production and refueling stations. A data acquisition system collects sensor data in real time and stores it in a database. The collected data is timestamped to ensure data consistency and relevance. Through the sensor network, the data is transmitted to an integrated GIS platform for real-time updates of environmental parameters.

[0115] S403: Aggregate data from various sensors and use data processing software to clean, analyze, and process the real-time data. A Kalman filter algorithm is used to clean noisy data to ensure the reliability of the input data. Data fusion and interpolation techniques are used to process and analyze environmental data, providing accurate boundary and initial conditions for subsequent numerical simulations. Meteorological data (wind speed, direction, temperature, humidity) and geographic information (such as slope and obstacle distribution) are integrated with the 3D scene model established in S3 through an API interface to link environmental parameters with spatial information.

[0116] Step 5: Establish a complete mathematical model of hydrogen leakage in complex scenarios of hydrogen production and refueling stations. The specific process is as follows:

[0117] Specifically, the governing equations include the continuity equation, momentum equation, energy equation, and diffusion model. Large eddy simulation (LES) is used to describe the hydrogen diffusion process. Appropriate boundary conditions are set based on the actual situation, such as the leak rate of the source, leak port shape, ambient temperature, and wind speed. The effects of hydrogen buoyancy and air resistance are also considered. The established mathematical model is validated using existing experimental data or standard examples to ensure its accuracy and reliability.

[0118] S51: The control equations include the continuity equation, the momentum equation, and the energy equation. The continuity equation describes the equation for conservation of hydrogen mass and is used to calculate the distribution changes of hydrogen in space and time. The momentum equation (Navier-Stokes equation) describes the equation for hydrogen flow and is used to calculate the velocity and pressure distribution of hydrogen in the environment. The energy equation is used to calculate the temperature distribution of hydrogen during the diffusion process.

[0119] S52: Turbulence model: The large eddy simulation (LES) or improved subparticle scale model is selected in the model to describe the equation of hydrogen concentration change, which is used to calculate the diffusion process of hydrogen in the air.

[0120] S53: Diffusion model: An equation describing the change in hydrogen concentration, used to calculate the diffusion process of hydrogen in the air. The diffusion equation is used to describe the mass content of substances in the multi-substance transport flow process after a hydrogen leak. The substance diffusion equation is expressed as follows:

[0121]

[0122] Where Y i is the mass content of species i, is the mass diffusion rate of species i, μ t is the turbulent viscosity, is the turbulent Schmidt number, D i,m is the thermal diffusivity of the species.

[0123] S54: Equation of State: Describes the relationship between hydrogen density, temperature, and pressure. A calculation model based on the real gas equation of state developed by NIST (National Institute of Standards and Technology) is selected to describe the relationship between hydrogen density, temperature, and pressure. The specific expression is as follows:

[0124]

[0125] Where α is the real gas state constant.

[0126] S55: Leakage model. Leakage parameter selection. Leakage concentration is described using a non-point source Gaussian plume mixture model. Leakage intensity requires establishing a corresponding leakage model based on the specific leakage scenario, such as an orifice leakage model or a pipeline leakage model.

[0127] The concentration of the gas that causes leakage in the three-dimensional space (x, y, z) is calculated as follows:

[0128]

[0129] Where x, y, and z are the coordinates of any point in space, t is the leakage time, C is the gas concentration, m is the total mass of the leaked gas, x, y, and z are the atmospheric diffusion coefficients corresponding to the x, y, and z directions respectively, u is the atmospheric velocity, and H is the height of the leakage source.

[0130] The leakage intensity is estimated based on the type of leakage source. The leakage intensity of the small hole leakage source model needs to be calculated based on the flow velocity at the leakage port.

[0131] when When , the gas flow at the leakage port is subsonic, and the gas leakage intensity is calculated as follows:

[0132]

[0133] when When , the gas flow at the leak port is sonic flow, and the gas leakage intensity is calculated as follows:

[0134]

[0135] Where q m is the leakage intensity, C g is the gas leakage coefficient, A is the leakage port area, p1 is the absolute pressure of the gas at the leakage port, M is the molar mass of the gas, Z is the compression factor, R is the molar constant, T i is the gas temperature at the leak point.

[0136] The leakage intensity of the pipeline leakage source model is calculated as follows:

[0137]

[0138] Where C g is the pipeline leakage coefficient; T2 and T3 are the fluid temperatures at the leakage source and pipeline L; P2 and P3 are the fluid absolute pressures at the leakage source and pipeline L.

[0139] S56: Define initial and boundary conditions. Leaks are described using a pressure outlet, with calculations performed based on the shape of the leak. Other conditions are free boundaries.

[0140] Step 6: Construction of SPH numerical discretization format and parameter setting of mathematical model. The specific process is as follows:

[0141] Specifically, the smoothed particle Galerkin method (SPG) is selected to discretize the mathematical model. According to the three-dimensional scene model, the particles are initialized in the fluid area and the particles are given initial velocity, temperature, concentration and other properties. The Gaussian kernel function is selected as the kernel function to control the interaction range of the particles and the density calculation rule. According to the actual scene and simulation requirements, the orthogonal test method is used to optimize the parameters such as the particle spacing, smoothing length, time step, artificial viscosity coefficient of the SPH method. The adaptive particle refinement technology is used to increase the particle density in areas with large hydrogen concentration gradients to improve the simulation accuracy. The boundary particle method is used to deal with the fluid flow behavior in complex geometric environments, and the numerical dissipation and stability of the SPH method are analyzed and optimized to improve the accuracy and reliability of the simulation results. Real-time environmental data (such as wind speed vector and temperature field) are introduced to dynamically correct the particle motion equation to improve the simulation accuracy under complex meteorological conditions.

[0142] In order to discretize the NS equations using SPH, the smoothing function selected is the quintic spline smoothing function, and the smoothing length is a variable smoothing length, which is described as follows:

[0143] Quintic spline smoothing function:

[0144]

[0145] Where, α d The distribution of the values of in one-dimensional, two-dimensional and three-dimensional space are and

[0146] Variable smooth length:

[0147]

[0148] Where h0 and are the initial smoothing length and density, respectively, and d is the dimension.

[0149] S62: The method for efficiently predicting the scope of hydrogen leakage in a hydrogen production and refueling station based on meshless particle simulation according to claim 1 is characterized in that the mass conservation equation, momentum conservation equation, and energy equation are converted into SPH discrete format using a particle approximation method and a quintic spline function that is closer to a Gaussian kernel function, and are expressed as follows:

[0150]

[0151] Where m is mass, ρ is density, e is internal energy, v is velocity vector, σ αβ is the total stress tensor, ε αβis the strain rate tensor, α and β represent the component indices of the vector and tensor. The above solution process is performed separately at each coordinate and time-integrated on the Lagrangian time frame.

[0152] S63: The method for efficiently predicting the scope of hydrogen leakage in a hydrogen production and refueling station based on meshless particle simulation according to claim 1 is characterized in that, using a particle approximation method, the density of a fluid particle can be obtained by taking a weighted average of the densities of all particles within its scope, as expressed as follows:

[0153]

[0154] Where, ρ i is the density of particle i, m j is the mass of particle j, r ij is the distance between particles i and j.

[0155] S64: The method for efficiently predicting the scope of hydrogen leakage in a hydrogen production and refueling station based on meshless particle physics simulation according to claim 1 is characterized in that the diffusion equation is numerically discretized using SPH as follows:

[0156]

[0157] Where C is the concentration of the mixture and D is the diffusion coefficient. The SPH discretization of the isotropic diffusion equation is expressed as follows:

[0158]

[0159] Among them, r ij =r i -r j

[0160] S65: The method for efficiently predicting the hydrogen leakage range of a hydrogen production and refueling station based on meshless particle simulation according to claim 1 is characterized in that, in order to solve the flow field pressure, the SPH method is used to describe the state equation of the weakly compressible fluid as follows:

[0161]

[0162] Where ρ0 is the reference density, constant γ = 7, and coefficient B = ρ0c 2 γ / , the speed of sound c is usually taken as 10 times the maximum speed of the flow field.

[0163] S66: The efficient prediction method for hydrogen leakage range of hydrogen production and refueling stations based on gridless particle simulation according to claim 1 is characterized in that the basic Smagorinsky model is selected as the LES model, and the viscosity of the sub-particles caused by turbulence The calculation is as follows:

[0164]

[0165] S67: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the particle injection position is set according to the leakage source position, and the particle injection amount is calculated according to the leakage port size, leakage pressure and leakage rate.

[0166] S68: The efficient prediction method for the hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that, in S5, the initial positions of the particles are distributed according to the requirements of the physical model, and the initial parameters of the air particles are defined, including density, viscosity, diffusion coefficient, temperature, rate, direction, etc.; the parameters of the hydrogen leakage source are defined, including thermal conductivity, density, viscosity, diffusion coefficient, temperature, rate, and direction, and the particle motion equations of air and hydrogen are dynamically corrected in combination with real-time meteorological data (wind speed, temperature).

[0167] S69: The method for efficiently predicting hydrogen leakage range in hydrogen production and refueling stations based on meshless particle physics simulation according to claim 1 is characterized in that in S5, initial conditions are set including wind speed, wind direction, etc., and the Lennard-Jones method is used to process the no-slip motion boundary of particles in the SPH simulation. The specific description is as follows:

[0168]

[0169] Where n1=12, n2=4; D depends on the specific problem and is generally taken to be equal to the square of the maximum velocity; r0 is the cutoff radius.

[0170] S610: The method for efficiently predicting hydrogen leakage range in a hydrogen production and refueling station based on meshless particle simulation according to claim 1, wherein in S5, the time step is estimated using the time step model proposed by Monaghan:

[0171]

[0172] The time step Δt satisfies the Courant-Friedrichs-Lewy condition based on the artificial sound speed c and the smoothing length h.

[0173] Δt≤0.4h / c (19)

[0174] S611: The method for efficiently predicting the scope of hydrogen leakage in hydrogen production and refueling stations based on gridless particle simulation according to claim 1 is characterized in that in S5, the SPH simulation software is DualSPHysics, and the solution process is shown in Figure 2; Using CUDA-GPU based high performance parallel method to accelerate computing ( Figure 3 ).

[0175] S62: SPH model parameters: Based on the actual scenario and simulation requirements, the SPH method parameters such as particle velocity, particle position, particle acceleration, viscosity, and artificial viscosity coefficient are optimized through orthogonal experimental method.

[0176] The Leap-Frog display algorithm with second-order accuracy is used to describe the particle velocity and position updates. The specific format is as follows: at the end of the initial step of the calculation, the density and velocity advance by half a step, and the position advances by one step:

[0177]

[0178] In order to maintain the continuity of subsequent time step calculations, the density and velocity are predicted half a time step ahead before the start of each time step:

[0179]

[0180] After each time step, the density, velocity, and position advance by one standard time step:

[0181]

[0182] When determining the position of each particle, the velocity also needs to be corrected, and the expression is:

[0183]

[0184] Where: ε is a correction coefficient ranging from 0 to 1. The purpose of the correction is to appropriately increase the numerical viscosity, so that the arrangement of the entire particle system is more orderly and the mutual penetration between particles can be avoided. In general numerical simulations, ε can be taken as 0.02.

[0185] The viscous force on fluid particles is calculated using the following formula:

[0186]

[0187] Wherein, subscript i represents the fluid particle at the center of the support domain that is affected by the viscous force of the surrounding fluid particles, subscript j represents the fluid particles other than the central fluid particle in the support domain, N represents the number of particles other than the central particle in the support domain, P represents the pressure of the particle, ρ represents the density of the particle, and v represents the velocity of the particle.

[0188] The Monaghan type artificial viscosity П is used to describe the particle kinematic viscosity. The Monaghan type artificial viscosity formula is:

[0189]

[0190] Where, α Π and β Π It is usually a constant and can be set to 1.0. c represents the speed of sound and v represents the velocity of the particle.

[0191] Step 7: Verification of the numerical simulation method of hydrogen leakage in hydrogen production and refueling stations based on SPH. The specific process is as follows:

[0192] Specifically, a small-scale hydrogen leak was simulated in a laboratory setting. The leak diffusion process was recorded using a high-speed camera and a gas concentration meter. The simulation results were compared with actual experimental data to verify the effectiveness of SPH simulation of hydrogen leaks. Through multiple verifications, the simulation model parameters were adjusted to ensure that the simulation results were consistent with the experimental data. Simulations were also conducted under different scenarios (different environments, different facility layouts, different meteorological conditions, etc.) to ensure that the model maintained its accuracy under various conditions.

[0193] Furthermore, the simulation results were compared with experimental data to evaluate the accuracy and reliability of the simulation method. The root mean square error (RMSE) between the simulated concentration field and the actual measured values was calculated. If the error was ≤ 5%, the model was considered valid. The RMSEP was calculated as follows.

[0194]

[0195] Where N is the number of measurement points at a given time t, V x (y n ,t) and Analytical solutions and numerical simulation results are shown in Figure 2.

[0196] Furthermore, sensitivity analysis was used to study the impact of simulation parameters on the simulation results and determine the optimal parameter combination. Model parameters (such as the number of particles and time step) were adjusted to verify computational stability and convergence. Sensitivity analysis was used to study the impact of simulation parameters on the simulation results and determine the optimal parameter combination.

[0197] Furthermore, the computational efficiency of the simulation method is evaluated, and the algorithms and procedures are optimized to increase the computational speed. The computational efficiency of the simulation method is evaluated, and the algorithms and procedures are optimized to increase the computational speed.

[0198] S701: Verify Objectives. Verify the accuracy of the SPH numerical simulation method in simulating hydrogen leakage and diffusion behavior in hydrogen production and refueling stations, evaluate the applicability of the SPH model to different leakage scenarios, environmental conditions, and equipment layouts, and determine the parameter settings and computational accuracy requirements for the SPH model.

[0199] S702: Verification Method. Leveraging existing hydrogen leakage experimental data, compare the SPH simulation results with the experimental data. Comparison parameters include hydrogen concentration distribution, diffusion range, diffusion velocity, and temperature distribution. For some simple leakage scenarios, analytical solutions (theoretical solutions) can be compared with SPH simulation results. Leveraging a hydrogen leakage accident at a hydrogen production and refueling station, compare the SPH simulation results with the accident investigation report.

[0200] S703: Verify parameters. These mainly include leakage parameters (leak size, leakage pressure, leakage rate), environmental parameters (wind speed, wind direction, temperature, humidity, atmospheric stability), equipment layout parameters (leak source location, pipeline layout, ventilation system), and SPH model parameters (kernel function type, particle spacing, time step, turbulence model parameters).

[0201] S704: Verification indicators. These mainly include concentration error, diffusion range error, velocity error, and temperature error. Quantitative evaluation of these errors is achieved by calculating the root mean square error (RMSE) and correlation coefficient. If the error is ≤5%, the model is considered valid.

[0202] S705: Numerical Stability and Convergence Analysis. This section verifies numerical stability by checking whether the changes in physical quantities (such as hydrogen concentration and velocity) during the simulation are reasonable and whether they tend to stabilize over time. The convergence of the model is determined by evaluating the SPH simulation results as the interparticle spacing and time step change.

[0203] S706: Sensitivity Analysis. This sensitivity analysis examines the model's sensitivity to various input parameters (such as wind speed, leak source size, pipe material, etc.), leak source location and size, and ambient temperature. By analyzing the impact of various parameters on the simulation results, the stability and reliability of the SPH model are evaluated.

[0204] Step 8: Conduct numerical simulations of hydrogen leakage at hydrogen production and refueling stations under multiple factors. Based on the numerical simulation method verified in Step 7, design a simulation scheme for hydrogen leakage and diffusion under multiple factors, including different leakage source locations, leakage rates, environmental conditions (wind speed, wind direction, temperature, humidity), leakage time, leakage hole size, etc. Through simulation analysis, the changes in hydrogen leakage range, concentration field, and diffusion rate under different meteorological conditions are analyzed. GPU acceleration technology (such as CUDA) is used to parallelize multiple simulation tasks, shortening calculation time and improving the efficiency of multi-factor simulation calculations. Analyze the impact of different factors on the hydrogen diffusion range and concentration distribution to provide a basis for safety risk assessment of hydrogen production and refueling stations.

[0205] Step 9: Output the dynamic distribution cloud map of hydrogen leakage concentration field in the hydrogen production and refueling station scenario. The specific process is as follows:

[0206] S901: Save the hydrogen concentration field data obtained in S8 in VTK or CSV format, including timestamps (intervals of 0.1s), spatial coordinates, and concentration values. In particular, the particle concentration needs to be converted into hydrogen concentration as follows:

[0207]

[0208] In the table below, j is the adjacent particle; W ij =W(r ij ,h),W ij ′=W′(r ij ,h). Here, r ij , h and W are the particle distance || X i -X j ||, cutoff radius, and fifth-order Wendland function.

[0209] S902 uses appropriate color mapping and contour display to highlight hazardous areas with high hydrogen concentrations. ParaView is used to generate 3D data, dynamic hydrogen concentration cloud maps, and videos. Using the Matplotlib library, a Python script is used to create dynamic 3D concentration cloud maps, annotating concentration gradients and diffusion paths. The leak diffusion process is monitored at a time step (e.g., 0.1 second / frame), and obstacle information from the environmental model is overlaid.

[0210] Step 10: Use post-processing software to process data and images, determine the dangerous area based on the threshold value, and generate warning signs. The specific process is as follows:

[0211] S1001: The Min-Max method is used to perform normalization processing on the dynamic distribution data of the hydrogen concentration field and the hydrogen concentration field distribution cloud map.

[0212] S1002: Setting the threshold for determining hazardous areas. Referencing GB / T 34532 "General Technical Guidelines for Hydrogen Safety," GB 50177 "Design Specifications for Hydrogen Stations," and NFPA 2 "Technical Specifications for Hydrogen," set the threshold for determining hazardous areas based on the safety distances and concentration limits specified in these standards. A 4% volume concentration is used as the hazardous threshold; areas exceeding this value are marked as high-risk areas. Safety levels: Low-risk area: <2% (warning zone); Medium-risk area: 2% to 4% (intervention zone); High-risk area: ≥4% (evacuation zone).

[0213] S1003: Region segmentation. Create a concentration threshold filter in post-processing software (such as the Threshold function in ParaView). Segment regions based on the threshold and assign different color identifiers (red / yellow / green).

[0214] S1004: Determine the hazardous area based on the threshold. Post-processing software analyzes the cloud image data generated in S9 to identify the spread of the hydrogen leak and determine the size of the leak area. A hydrogen concentration threshold (such as the lower explosion limit (LEL)) is set and the normalized hydrogen concentration is compared with the set threshold. If the hydrogen concentration exceeds the threshold, the area is determined to be hazardous.

[0215] S1005: Generate warning signs. Alert levels are assigned based on factors such as hydrogen concentration and diffusion range. First, color coding is used to visually display the levels of hazardous areas. Red indicates high-risk areas, yellow indicates medium-risk areas, and blue indicates low-risk areas. Second, a timestamp, concentration scale, and risk level description are overlaid on the cloud map. Finally, the warning signs are integrated into the GIS system to enable real-time monitoring and visualization of hazardous areas.

[0216] S1006: Visualization and Simulation Demonstration. Based on set concentration thresholds, safe, warning, and hazardous areas are demarcated, and corresponding warning signs are generated. Based on hydrogen leak simulation and hazardous area determination, visualization software can be used to present information such as zone demarcation and warning signs in graphical or animated form, helping personnel intuitively understand the spread and risk of the leak.

[0217] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for predicting hydrogen leakage range in hydrogen production and refueling stations based on meshless particle simulation is characterized by: The following steps are involved: S1: Use geographic information system (GIS) and laser scanning technology to build a three-dimensional environmental model around the hydrogen production and refueling station, including topography, building distribution, and meteorological observation data; S2: Build a 3D physical model of the hydrogen production and refueling plant based on Building Information Modeling (BIM) technology, including the 3D spatial layout of hydrogen storage tanks, pipelines, valves, and vents; S3: spatially registering the three-dimensional environment model with the plant model to construct a complete three-dimensional scene model including environmental boundary conditions; S4: Real-time data collection of station temperature, wind speed, and hydrogen concentration is collected through a sensor network and integrated with the 3D scene model; S5: Establish a hydrogen leakage mathematical model including the continuity equation, momentum equation, energy equation and large eddy simulation (LES) turbulence model; S6: Use the smoothed particle hydrodynamics (SPH) method to discretize the mathematical model and set the particle spacing, smoothing length, and time step parameters; S7: Use experimental data to verify the accuracy of the SPH simulation method and conduct multi-factor numerical simulation to calculate the diffusion range under different leakage scenarios; S8: Generate dynamic distribution cloud map of hydrogen concentration field based on GPU parallel acceleration technology; S9: Post-processing software is used to divide hazardous areas according to preset concentration thresholds and generate visual warning signs.

2. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: Said S1 specifically includes: S11: Use LiDAR to obtain 5cm accurate 3D point cloud data around the hydrogen production and refueling station; S12: Identify building outlines through point cloud segmentation and build a digital terrain model (DTM) based on GIS elevation data; S13: Perform spatial coordinate system registration and verification on the three-dimensional model and the satellite map.

3. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: In S2, the three-dimensional physical model of the plant includes: The pressure resistance parameters of the hydrogen storage tank, the sealing performance parameters of the pipeline welds, and the marking of potential leakage points of the valve flange interface.

4. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: In S5, the control equations of the mathematical model include: Continuity equation: Where ρ is the hydrogen density, t is the time, and u is the velocity vector. represents the divergence operator; Momentum equation: Where P is the pressure, τ is the viscous stress tensor, g is the gravitational acceleration vector, represents the tensor product; Energy equation: Where T is temperature, C P is the specific heat capacity at constant pressure, k is the thermal conductivity, is the temperature gradient, represents the viscous dissipation term.

5. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 4 is characterized in that: The LES model adopts the Smagorinsky subgrid model, and the turbulent viscosity is calculated as: where c s is the Smagorinsky constant, Δ is the filter width, ||S i || is the strain rate tensor modulus.

6. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: The SPH discretization process in S6 specifically includes: The quintic spline kernel function is used for particle approximation calculation, and the kernel function expression is: W(r,h)=α d (1-q) 4 (1+4q) Where q = r / h, α d is the dimension normalization coefficient, h is the smoothing length; The time step Δt satisfies the CFL condition: Δt≤0.25h / c.

7. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: The multi-factor numerical simulation described in S7 includes: Set simulation conditions with different leakage diameters d∈[2mm,20mm], leakage pressures p∈[0.1MPa,70MPa], and wind speeds v∈[1m / s,15m / s]; The sensitivity weights of various factors to the area of explosion hazardous area were analyzed by orthogonal test method.

8. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: The GPU parallel acceleration described in S8 uses the CUDA architecture to divide the SPH particle computing task into: Thread blocks handle local particle interaction calculations; The global memory stores the particle position, velocity, and concentration parameters; The shared memory caches the neighboring particle search data.

9. The method for predicting hydrogen leakage range of a hydrogen production and refueling station based on gridless particle simulation according to claim 1 is characterized in that: The criteria for determining hazardous areas in S9 are: High-risk area: hydrogen volume concentration ≥ 4%; Medium risk area: 2% ≤ concentration < 4%; Low risk area: concentration <2%; The warning signs dynamically mark the boundaries of each risk level area in the three-dimensional scene model.

Images