Atmospheric pollutant diffusion multi-working condition CFD simulation device and method

Abstract

The application discloses a kind of atmospheric pollutant diffusion multi-working condition CFD simulation device and method, using wind field library pre-construction-diffusion superposition solving decoupling type computing framework, based on wind speed-wind direction combination set, by CFD solving unstructured grid wind field containing terrain / building, adaptive encryption is realized to wind field library by combining global sensitivity analysis, reusable wind field library with structured index is constructed;For different working conditions matching reusable pre-computed wind field, by unstructured grid topology automatic pre-processing, dynamic diffusion coefficient adaptive model based on turbulence statistics, finite volume method is used to solve convection-diffusion equation, multi-working condition rapid iteration is realized.Multiple working condition simulation power consumption is reduced by more than 80%, the calculation period is shortened to 12% of the traditional method or less, while the diffusion calculation error under complex underlying surface is reduced by 35%, both calculation efficiency and accuracy are considered, suitable for multi-working condition simulation scene such as construction project environmental impact assessment, emergency response of sudden environmental event.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of atmospheric environmental engineering and computational fluid dynamics (CFD), specifically involving a CFD simulation device and method for atmospheric pollutant diffusion under multiple operating conditions that is computationally efficient, highly accurate, adaptable to complex underlying surfaces, and supports rapid iteration under multiple operating conditions. Background Technology

[0002] With the continuous improvement of my country's atmospheric environmental control requirements, scenarios such as environmental impact assessments of construction projects, pollution source control in industrial parks, and emergency response to sudden environmental incidents all require high-precision, multi-condition numerical simulations of the diffusion process of air pollutants. In complex underlying surface scenarios including topographic relief and urban building clusters, unstructured mesh CFD simulation is currently the mainstream technique. Its core is to obtain the wind field distribution by solving the Navier-Stokes equations, and then solve the pollutant diffusion process based on the passive scalar transport equations.

[0003] In existing technologies, multi-condition simulations of pollutant diffusion generally adopt a fully coupled approach integrating "wind field solution - diffusion calculation." This means that for each set of conditions (including combinations of variables such as wind speed, wind direction, pollution source location, release intensity, and pollutant type), the entire process from solving the Navier-Stokes equations to diffusion calculations is fully executed. This approach has the following core technical shortcomings:

[0004] 1. Severe Redundancy in Computing Resources: Wind speed and direction are the core control parameters determining wind field distribution. Under the same wind speed-direction combination, the wind field distribution characteristics remain constant when only pollution source-related parameters are adjusted. However, the fully coupled mode still requires repeated solving of the Navier-Stokes equations. In complex underlying surface scenarios, the time consumed by solving a single wind field typically accounts for more than 90% of the total computation cycle. The simulation cycle for 100 working conditions can reach more than 72 hours, resulting in a large-scale waste of computing power.

[0005] 2. Insufficient timeliness of multi-condition iteration: In emergency response scenarios of sudden environmental events, it is necessary to complete the diffusion simulation of multiple leakage conditions within 1 hour. The traditional fully coupled mode cannot meet this timeliness requirement. At the same time, due to the limitation of computational efficiency, it is difficult to carry out systematic sensitivity analysis of pollution source parameters and pollutant characteristics.

[0006] 3. Insufficient calculation accuracy under complex underlying surfaces: Existing technologies generally use constant diffusion coefficients, which cannot adapt to the spatial heterogeneity of turbulence intensity under complex underlying surfaces, resulting in diffusion calculation errors of more than 40% in building wake zones and terrain undulation zones; at the same time, the automatic processing capability of the lower boundary of unstructured grids is insufficient, which easily leads to numerical oscillations and non-physical results.

[0007] 4. Poor reusability of wind field data: Existing technologies only pre-calculate wind fields for specific operating conditions of a single project, and have not formed a standardized wind field library that can be reused across operating conditions and projects. Wind field data between different projects and different operating conditions cannot be effectively reused.

[0008] To address the aforementioned shortcomings, existing technologies have proposed a simplified approach called the "frozen wind field method," which involves first solving the wind field once and then solving the diffusion equation based on the fixed wind field. However, this approach still has the following limitations: it lacks a standardized system for constructing and reusing wind field libraries, making it unsuitable for simulations under multiple operating conditions across the entire parameter domain; it fails to optimize the discretization format and boundary treatment methods for complex underlying unstructured meshes; and it does not solve the accuracy problem caused by the constant diffusion coefficient. The overall solution remains at the level of a general process, lacking substantial technological innovation, which is obvious to those skilled in the art and cannot meet the actual needs of engineering applications. Summary of the Invention

[0009] The purpose of this invention is to address the aforementioned deficiencies of existing technologies by providing a CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions that is computationally efficient, highly accurate, adaptable to complex underlying surfaces, and supports rapid iteration under multiple operating conditions. This method solves the technical problems of redundant computing power, poor timeliness, and insufficient accuracy under complex underlying surfaces in existing multi-operating condition simulations.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: a CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions, characterized by comprising the following steps:

[0011] S1: Wind field library pre-construction: Determine the wind speed-direction combination set of the target area according to the atmospheric environmental assessment specifications. For each combination in the wind speed-direction combination set, use unstructured grid CFD steady / unsteady wind field solution that fits the terrain and buildings to solve the wind field. Output standardized VTK format wind field files and construct a reusable wind field library with a structured index table.

[0012] S2: Unstructured mesh topology preprocessing: Read the target wind field file in the reusable wind field library, extract the mesh topology data and wind field data, obtain the cell center coordinates, cell volume, and topological relationship of adjacent cells, automatically identify the terrain and building solid wall boundary cells, and complete the mesh quality screening and wind field data completion.

[0013] S3: Operating Condition Matching and Wind Field Reuse: Read the parameter set of the target operating condition, and match the corresponding pre-calculated wind field data from the reusable wind field library through a structured index table based on the wind speed-direction parameters in the operating condition; when the wind speed-direction parameters of the new operating condition are consistent with those of the calculated operating condition, the loaded wind field data is directly reused.

[0014] S4: Numerical solution for pollutant diffusion: Initialize the global concentration field, locate the corresponding grid cells according to the coordinates of the pollution source in the working condition, discretize the convection-diffusion equation using the finite volume method, combine the upwind scheme to handle the convection term and the second-order central difference scheme to handle the diffusion term, and solve the time step using the explicit Euler method to obtain the time evolution law and spatial distribution results of pollutant concentration.

[0015] S5: Multi-condition iteration: Repeat steps S3-S4 to complete the simulation calculation of all target conditions and output the multi-condition comparison analysis results.

[0016] Preferably, in step S1, the basic parameter range of the wind speed-wind direction combination set is: wind speed 1~10m / s, step size 1m / s; wind direction 0~360°, step size 10°, for a total of 360 basic combinations.

[0017] Preferably, the reusable wind farm library is an adaptively encrypted reusable wind farm library, and the adaptive encryption process is as follows:

[0018] S1-1. Based on the global sensitivity analysis method, calculate the sensitivity coefficient of different wind speed-direction combinations on the pollutant diffusion results;

[0019] S1-2. For the prevailing wind direction and critical wind speed range where the sensitivity coefficient is higher than the preset threshold, the wind speed step size is increased to 0.5 m / s and the wind direction step size is increased to 5°; for the non-sensitive range where the sensitivity coefficient is lower than the preset threshold, the wind speed step size is increased to 2 m / s and the wind direction step size is increased to 20°.

[0020] S1-3. Supplement the CFD wind field solution for the newly added combinations after encryption / relaxation, and update the wind field library and structured index table.

[0021] Preferably, in step S1, the fields of the structured index table include combination ID, wind speed, wind direction, absolute path of the wind field file, wind field type, solution method, total number of grid cells, and calculation completion time; the wind field file adopts a Binary format VTK file, which includes node three-dimensional coordinates, cell topology information, three-dimensional wind speed components at the cell center, terrain elevation label, and building solid wall cell label.

[0022] Preferably, in step S2, the specific criteria for mesh quality screening are: removing cells with an aspect ratio > 10 and a volume < 1. e-6 For the m³ malformed cells, the inverse distance weighted interpolation method is used to complete the wind field data for the missing regions after removal;

[0023] The specific method for boundary element identification is as follows: elements with less than 3 adjacent elements in a two-dimensional triangular mesh and elements with less than 4 adjacent elements in a three-dimensional tetrahedral mesh are identified as boundary elements.

[0024] Preferably, in step S4, the diffusion term is calculated using a dynamic diffusion coefficient adaptive model based on CFD turbulence statistics, with the following formula:

[0025]

[0026] In the formula, , which is a constant for the turbulence model, is taken as 0.09; The turbulent kinetic energy at the center of the unit output by the CFD wind field; The turbulent kinetic energy dissipation rate at the center of the unit output of the CFD wind field.

[0027] Preferably, in step S4, the time step of the explicit Euler method is... Satisfying the CFL stability condition:

[0028]

[0029] In the formula, the CFL number is taken as 0.5~0.8; Unit volume; The unit center wind speed modulus; This represents the maximum surface area of ​​the unit.

[0030] Preferably, step S4 further includes boundary condition processing after the numerical solution of pollutant diffusion. Specifically, the boundary condition processing is as follows: the boundary of the terrain and the solid wall of the building adopts a zero-flux boundary condition, and the flux of the adjacent surface of the boundary cell is set to 0; the boundary of the computational domain outlet adopts a zero-gradient open boundary condition; and the boundary of the computational domain inlet and far-field adopts a zero-concentration boundary condition.

[0031] Preferably, in step S1, the method for processing unsteady wind fields is as follows: the unsteady wind field time series is processed by intrinsic orthogonal decomposition (POD) to reduce the order, and the first 20 dominant modes are extracted to construct a wind field reduction model, which is stored in the wind field library; in step S3, when matching unsteady wind fields, the wind field data at any time is quickly reconstructed through the reduction model and matched with the time step of diffusion calculation.

[0032] A multi-condition CFD simulation device for atmospheric pollutant diffusion is provided to implement a multi-condition CFD simulation method for atmospheric pollutant diffusion. The multi-condition CFD simulation device for atmospheric pollutant diffusion includes a wind field database construction module, a mesh preprocessing module, a condition matching module, a diffusion solution module, and a multi-condition iteration module. The wind field database construction module performs wind field database pre-construction operations. The mesh preprocessing module performs unstructured mesh topology preprocessing operations. The condition matching module performs condition matching and wind field reuse operations. The diffusion solution module performs pollutant diffusion numerical solution operations. The multi-condition iteration module performs multi-condition iteration and result output operations.

[0033] The advantages of this invention are as follows: By introducing a decoupled computational framework of "preconstruction of wind field database - diffusion superposition solution", combined with an adaptive encryption strategy for wind field database, a dynamic diffusion coefficient model based on turbulence statistics, and a method for order reduction and reconstruction of unsteady wind field POD, this invention systematically solves the three core defects of existing technologies: redundant computing power, insufficient accuracy, and poor reusability.

[0034] Compared with the prior art, the present invention has the following significant advantages:

[0035] 1. Significant improvement in computing power efficiency: This invention decouples wind field solution and diffusion calculation by pre-constructing and reusing a standardized wind field library. Multi-condition simulations under the same wind speed-direction combination do not require repeated solving of the Navier-Stokes equations, reducing overall computing power consumption by more than 80%. Combined with the adaptive encryption method of the wind field library, the workload of wind field pre-calculation is further reduced by 42%, and the simulation cycle of 100 sets of conditions is shortened from 72 hours in the traditional method to less than 8 hours, meeting the timeliness requirements of emergency response to sudden environmental events.

[0036] 2. Significantly improved calculation accuracy under complex underlying surfaces: This invention adopts a dynamic diffusion coefficient adaptive model based on turbulence statistics to replace the constant diffusion coefficient of the existing technology, realizing spatial adaptive matching between the diffusion coefficient and the local turbulence intensity, reducing the diffusion calculation error in the building wake zone and topographic undulation zone by more than 35%; at the same time, through automated preprocessing of unstructured grids and precise implementation of boundary conditions, numerical oscillations are avoided, ensuring the calculation stability and accuracy under complex underlying surfaces.

[0037] 3. Significantly enhanced wind field data reusability: This invention constructs a standardized wind field library that conforms to atmospheric environmental assessment standards. Through a structured index table, it enables rapid matching and cross-condition and cross-project reuse of wind field data, avoiding repetitive wind field solution work and significantly reducing the computational cost of engineering applications.

[0038] 4. Strong adaptability to multiple operating conditions: This invention supports steady / unsteady wind fields and two-dimensional / three-dimensional unstructured grids, and can flexibly adjust parameters such as pollution source location, release intensity, and pollutant type to meet the multi-condition simulation needs of various scenarios such as environmental impact assessment of construction projects, pollution source control in industrial parks, and emergency response to sudden environmental incidents.

[0039] 5. Strong engineering applicability: This invention provides a complete standardized calculation process, parameter value specifications, and boundary condition implementation methods. Those skilled in the art can fully reproduce the solution by following the instructions. No deep numerical calculation background is required, and it can be directly applied to engineering practice. Attached Figure Description

[0040] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0041] Figure 1 This is an overall flowchart of the CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions described in this invention;

[0042] Figure 2 This is a flowchart of the adaptive encryption sub-step of the wind field database in this invention;

[0043] Figure 3 This is a diagram showing the operating parameters for the five operating conditions in Embodiment 1 of the present invention.

[0044] Figure 4 This is a schematic diagram of the structure of the device of the present invention. Detailed Implementation

[0045] This invention discloses a multi-condition CFD simulation device for atmospheric pollutant diffusion, used to implement a multi-condition CFD simulation method for atmospheric pollutant diffusion. The device includes a wind field database construction module 1, a mesh preprocessing module 2, a condition matching module 3, a diffusion solution module 4, and a multi-condition iteration module 5. The wind field database construction module performs wind field database pre-construction operations; the mesh preprocessing module performs unstructured mesh topology preprocessing operations; the condition matching module performs condition matching and wind field reuse operations; the diffusion solution module performs pollutant diffusion numerical solution operations; and the multi-condition iteration module performs multi-condition iteration and result output operations. This invention systematically solves the three core defects of existing technologies—redundant computing power, insufficient accuracy, and poor reusability—by introducing a decoupled computational framework of "wind field database pre-construction-diffusion superposition solution," combined with an adaptive wind field database densification strategy, a dynamic diffusion coefficient model based on turbulence statistics, and a non-steady wind field POD order reduction reconstruction method.

[0046] The multi-condition CFD simulation method for atmospheric pollutant diffusion of the present invention includes the following steps:

[0047] S1: Wind field library pre-construction: Determine the wind speed-direction combination set of the target area according to the atmospheric environmental assessment specifications. For each combination in the wind speed-direction combination set, use unstructured grid CFD steady / unsteady wind field solution that fits the terrain and buildings to solve the wind field. Output standardized VTK format wind field files and construct a reusable wind field library with a structured index table.

[0048] S2: Unstructured mesh topology preprocessing: Read the target wind field file in the reusable wind field library, extract the mesh topology data and wind field data, obtain the cell center coordinates, cell volume, and topological relationship of adjacent cells, automatically identify the terrain and building solid wall boundary cells, and complete the mesh quality screening and wind field data completion.

[0049] S3: Operating Condition Matching and Wind Field Reuse: Read the parameter set of the target operating condition, and match the corresponding pre-calculated wind field data from the reusable wind field library through a structured index table based on the wind speed-direction parameters in the operating condition; when the wind speed-direction parameters of the new operating condition are consistent with those of the calculated operating condition, the loaded wind field data is directly reused.

[0050] S4: Numerical solution for pollutant diffusion: Initialize the global concentration field, locate the corresponding grid cells according to the coordinates of the pollution source in the working condition, discretize the convection-diffusion equation using the finite volume method, combine the upwind scheme to handle the convection term and the second-order central difference scheme to handle the diffusion term, and solve the time step using the explicit Euler method to obtain the time evolution law and spatial distribution results of pollutant concentration.

[0051] S5: Multi-condition iteration: Repeat steps S3-S4 to complete the simulation calculation of all target conditions and output the multi-condition comparison analysis results.

[0052] This invention achieves efficient coverage of the entire wind speed-direction parameter domain and cross-condition reuse of wind field data through a standardized wind field database pre-construction and structured indexing mechanism, combined with an adaptive encryption strategy guided by global sensitivity analysis. This reduces the computational power consumption of multi-condition simulation by more than 80%. Furthermore, it utilizes a dynamic diffusion coefficient adaptive model based on turbulence statistics and CFD to solve for the output turbulent kinetic energy. and dissipation rate This method achieves spatial adaptive matching between the diffusion coefficient and the local turbulence intensity on complex underlying surfaces, significantly improving the calculation accuracy of building wake regions and topographic undulation regions. Through unsteady wind field POD order reduction and rapid reconstruction methods, and by employing intrinsic orthogonal decomposition to extract dominant modes and construct a reduced-order model, it achieves efficient storage and rapid reconstruction of wind fields at any time in unsteady scenarios, significantly shortening the response time to sudden environmental events.

[0053] In step S1, the basic parameter range of the wind speed-wind direction combination set is: wind speed 1~10m / s, step size 1m / s; wind direction 0~360°, step size 10°, for a total of 360 basic combinations.

[0054] The reusable wind farm library is an adaptively encrypted reusable wind farm library. The adaptive encryption process is as follows:

[0055] S1-1. Based on the global sensitivity analysis method, calculate the sensitivity coefficient of different wind speed-direction combinations on the pollutant diffusion results;

[0056] S1-2. For the prevailing wind direction and critical wind speed range where the sensitivity coefficient is higher than the preset threshold, the wind speed step size is increased to 0.5 m / s and the wind direction step size is increased to 5°; for the non-sensitive range where the sensitivity coefficient is lower than the preset threshold, the wind speed step size is increased to 2 m / s and the wind direction step size is increased to 20°.

[0057] S1-3. Supplement the CFD wind field solution for the newly added combinations after encryption / relaxation, and update the wind field library and structured index table.

[0058] In step S1, the fields of the structured index table include combination ID, wind speed, wind direction, absolute path of wind field file, wind field type, solution method, total number of grid cells, and calculation completion time; the wind field file adopts a Binary format VTK file, which includes node three-dimensional coordinates, cell topology information, three-dimensional wind speed components at cell center, terrain elevation label, and building solid wall cell label.

[0059] In step S2, the specific criteria for mesh quality screening are: removing cells with an aspect ratio > 10 and a volume < 1. e-6 For malformed cells of m³, inverse distance weighted interpolation is used to complete the wind field data for the missing regions after removal. The specific method for boundary cell identification is as follows: cells with less than 3 adjacent cells in a two-dimensional triangular mesh and cells with less than 4 adjacent cells in a three-dimensional tetrahedral mesh are identified as boundary cells.

[0060] In step S4, the diffusion term is calculated using a dynamic diffusion coefficient adaptive model based on CFD turbulence statistics, with the following formula:

[0061]

[0062] In the formula, , which is a constant for the turbulence model, is taken as 0.09; The turbulent kinetic energy at the center of the unit output by the CFD wind field; The turbulent kinetic energy dissipation rate at the center of the unit output of the CFD wind field.

[0063] In step S4, the time step of the explicit Euler method Satisfying the CFL stability condition:

[0064]

[0065] In the formula, the CFL number is taken as 0.5~0.8; Unit volume; The unit center wind speed modulus; This represents the maximum surface area of ​​the unit.

[0066] Step S4, after the numerical solution of pollutant diffusion, also includes boundary condition processing. Specifically, the boundary condition processing is as follows: the topographic and building solid wall boundaries adopt zero flux boundary conditions, and the flux of adjacent surfaces of the boundary cells is set to 0; the computational domain outlet boundary adopts zero gradient open boundary conditions; and the computational domain inlet and far-field boundaries adopt zero concentration boundary conditions.

[0067] In step S1, the method for processing unsteady wind fields is as follows: the time series of unsteady wind fields is processed by intrinsic orthogonal decomposition (POD) to reduce the order, and the first 20 dominant modes are extracted to construct a wind field reduction model, which is stored in the wind field library; in step S3, when matching unsteady wind fields, the wind field data at any time is quickly reconstructed through the reduction model and matched with the time step of diffusion calculation.

[0068] The present invention provides a multi-condition CFD simulation device for atmospheric pollutant diffusion, used to implement a multi-condition CFD simulation method for atmospheric pollutant diffusion. It includes a wind field database construction module, a mesh preprocessing module, a condition matching module, a diffusion solution module, and a multi-condition iteration module. The wind field database construction module performs wind field database pre-construction operations; the mesh preprocessing module performs unstructured mesh topology preprocessing operations; the condition matching module performs condition matching and wind field reuse operations; the diffusion solution module performs pollutant diffusion numerical solution operations; and the multi-condition iteration module performs multi-condition iteration and result output operations.

[0069] Example 1

[0070] Taking a complex underlying surface scenario with buildings in an industrial park as an example, the method of the present invention will be described in detail. The computational domain of this embodiment is 5km×5km, including 12 industrial buildings, and adopts a three-dimensional tetrahedral unstructured mesh with a total of 256,000 mesh cells.

[0071] The specific steps of this embodiment are as follows:

[0072] S1. Wind farm storage pre-construction steps:

[0073] S1-1. According to the "Technical Guidelines for Environmental Impact Assessment - Atmospheric Environment" (HJ 2.2-2018), the basic wind speed-wind direction combination set for the target area is determined: wind speed 1~10m / s, step size 1m / s; wind direction 0~360°, step size 10°, totaling 360 basic combinations.

[0074] S1-2. For each combination, the steady wind field is solved using the Realizable k-ε turbulence model. The computational domain inlet uses a velocity inlet boundary, the outlet uses a free flow outlet boundary, and the ground and building surfaces use no-slip wall boundaries. After the solution is completed, a Binary format VTK wind field file is output, and the file naming rule is wind_U{wind speed}_theta{wind direction}.vtk.

[0075] S1-3. Construct a structured index table with fields including combination ID, wind speed, wind direction, absolute path of wind field file, wind field type (steady), solution method (Realizable k-ε), total number of grid cells, and calculation completion time.

[0076] S1-4. Based on the Sobol global sensitivity analysis method, the sensitivity coefficients of different wind speed-direction combinations on the pollutant diffusion results are calculated. For the dominant wind direction (NE, ENE) and critical wind speed (3~5m / s) range with sensitivity coefficients higher than 0.1, the wind speed step size is increased to 0.5m / s and the wind direction step size is increased to 5°. For the non-sensitive range with sensitivity coefficients lower than 0.02, the wind speed step size is increased to 2m / s and the wind direction step size is increased to 20°. The wind field solution is supplemented for the 28 newly added combinations, and the wind field library and index table are updated. Finally, the wind field library contains a total of 388 wind field files.

[0077] S2. Unstructured mesh topology preprocessing steps:

[0078] S2-1. Read the target wind field file (wind speed 5m / s, wind direction 30°) from the wind field library, and extract the node three-dimensional coordinates, unit topology data, unit center wind speed components (u,v,w), turbulent kinetic energy k, and turbulent kinetic energy dissipation rate ε.

[0079] S2-2. Calculate the coordinates of the cell center, the cell volume, and the topological relationship between adjacent cells. The volume of the three-dimensional tetrahedral cell is directly calculated using the VTK built-in function cell.GetVolume().

[0080] S2-3. Grid quality screening: Remove malformed cells with an aspect ratio >10 and a volume <1e-6m³. In this embodiment, a total of 124 malformed cells were removed. The inverse distance weighted interpolation method was used to complete the wind field data for the missing areas.

[0081] S2-4 Boundary Element Identification: Tetrahedral elements with fewer than 4 adjacent elements are identified as boundary elements. Terrain and building solid boundary elements are automatically identified, totaling 21,000 boundary elements.

[0082] S3. Operating condition matching and wind farm reuse steps:

[0083] This embodiment sets up a total of 5 working conditions, and the working condition parameters are as follows: Figure 3 As shown:

[0084] For operating condition 1, the wind field file wind_U5_theta30.vtk is matched through the index table and the wind field data is loaded; the wind speed-direction parameters of operating conditions 2-4 are the same as those of operating condition 1, and the already loaded wind field data is directly reused without repeated reading and solving; for operating condition 5, the wind field file wind_U8_theta60.vtk is matched and the corresponding wind field data is loaded.

[0085] S4. Steps for numerical solution of pollutant diffusion:

[0086] S4-1. Initialize the global concentration field to 0. Based on the pollution source coordinates (2500, 2500) of working condition 1, calculate the Euclidean distance between the unit center and the target coordinates. Take the unit with the smallest distance as the pollution source unit, with ID 125864.

[0087] S4-2. The diffusion coefficient is calculated using a dynamic diffusion coefficient model based on turbulence statistics, and the formula is as follows: The diffusion coefficient correction factor for SO2 is taken as 1.2, and the diffusion coefficient correction factor for PM2.5 is taken as 0.3.

[0088] S4-3. Determine the time step: Set the CFL number to 0.5, and calculate the time step based on the CFL stability condition. The total simulation time is 3600s, and the output interval is 60s.

[0089] S4-4. The convection-diffusion equations are discretized using the finite volume method. The convection term is discretized using a first-order upwind scheme, and the diffusion term is discretized using a second-order central difference scheme. The solution is obtained by time-stepping using the explicit Euler method.

[0090] S4-5 Boundary Condition Treatment: The terrain and building solid wall boundaries adopt the flux-free boundary condition, and the flux of adjacent surfaces of the boundary cells is set to 0; the computational domain outlet adopts the zero gradient open boundary condition; the inlet and far field adopt the zero concentration boundary condition.

[0091] S4-6. Store the concentration field data every 60 seconds, complete the simulation calculation for 3600 seconds, and obtain the temporal evolution and spatial distribution results of pollutant concentration.

[0092] S5 Multi-condition Iteration Steps:

[0093] Repeat steps S3-S4 to complete the simulation calculations for conditions 2-5, and finally output the concentration field VTK files for 5 conditions, GIF animations of the diffusion process, and a comparison report of the maximum concentration and diffusion range for multiple conditions.

[0094] Comparative Example 1

[0095] This comparative example uses the existing fully coupled simulation method. For the five working conditions in Example 1, the entire process of wind field solution and diffusion calculation is fully executed for each working condition, and the remaining parameters are completely consistent with Example 1.

[0096] Implementation effect verification:

[0097] 1. Verification of calculation accuracy: The calculation results of Example 1 and Comparative Example 1 were compared with the measured values ​​from the wind tunnel experiment. The results are as follows: Figure 3As shown: the average relative error between the calculated and measured values ​​in Example 1 is 8.2%, and the average relative error in Comparative Example 1 is 12.6%. The present invention reduces the calculation error by 34.9%, especially significantly improving the calculation accuracy of the building wake zone;

[0098] 2. Verification of computational efficiency: Comparison of the computation cycles of Example 1 and Comparative Example 1. Figure 4 As shown: the total calculation cycle of the 5 working conditions in Comparative Example 1 is 78 hours, and the total calculation cycle of Example 1 is 8.5 hours. The calculation cycle is shortened to 10.9% of the traditional method, and the computing power consumption is reduced by 89.1%, which verifies the efficiency of the present invention.

[0099] 3. Wind field data reuse effect: In Example 1, the wind field data of Case 1 is directly reused in Case 2-4. The calculation time for a single case is only 12 minutes, while the wind field solution time for each case in Comparative Example 1 is 15 hours, which fully verifies the technical effect of wind field data reuse in this invention.

[0100] Example 2

[0101] This embodiment focuses on emergency response scenarios for sudden environmental events, using unsteady wind fields for simulation. The specific steps are as follows:

[0102] S1. Wind field database pre-construction steps: For the unsteady wind field time series of the target area (time step 1s, total duration 3600s), the intrinsic orthogonal decomposition (POD) is used for order reduction processing to extract the top 20 dominant modes (cumulative energy accounting for 98.2%), construct the wind field order reduction model, and store it in the wind field database;

[0103] S2. Unstructured mesh topology preprocessing step: Same as step S2 in Example 1;

[0104] S3. Operating condition matching and wind field reuse steps: For 10 emergency operating conditions with different leakage intensities, the same unsteady wind field reduction model is matched. The wind field data at any time is quickly reconstructed through the reduction model and matched with the time step of diffusion calculation.

[0105] S4. Numerical solution steps for pollutant diffusion: Same as step S4 in Example 1;

[0106] S5. Multi-condition iterative steps: Complete the simulation calculation of 10 emergency conditions, with a total calculation cycle of 28 minutes, meeting the timeliness requirement of producing results within 1 hour in emergency response scenarios.

[0107] Compared with the prior art, the present invention:

[0108] 1. Pioneering a decoupled computational framework of "pre-construction of wind farm storage - diffusion superposition solution".

[0109] Breaking away from the traditional fully coupled model of "wind field solution-diffusion calculation", this method achieves efficient reuse of wind field data across operating conditions and projects by pre-calculating standardized wind speed-direction combination sets and constructing structured index tables, fundamentally eliminating repetitive wind field solutions in multi-operating condition simulations.

[0110] 2. An adaptive encryption method for wind field databases based on global sensitivity analysis is proposed.

[0111] The Sobol global sensitivity analysis method is introduced to quantify the impact of different wind speed-direction combinations on pollutant diffusion results. The sampling is intensified in the prevailing wind direction and critical wind speed range (wind speed step 0.5 m / s, wind direction step 5°), while the sampling is relaxed in the non-sensitive range (wind speed step 2 m / s, wind direction step 20°). The workload of wind field pre-calculation is reduced by 42% while ensuring accuracy.

[0112] 3. Design an adaptive model for the dynamic diffusion coefficient based on turbulence statistics.

[0113] Using CFD wind field to solve for the turbulent kinetic energy at the center of the unit cell and dissipation rate Constructing the dynamic diffusion coefficient ( This allows for spatial adaptive matching of the diffusion coefficient with the local turbulence intensity on complex underlying surfaces, replacing the traditional constant diffusion coefficient.

[0114] 4. Innovative methods for order reduction and rapid reconfiguration of POD in unsteady wind fields

[0115] For unsteady wind field time series, the first 20 dominant modes (energy percentage ≥ 98%) are extracted using intrinsic orthogonal decomposition (POD) to construct a reduced-order model, which is stored in the wind field library. During diffusion calculation, the wind field at any time is quickly reconstructed using the reduced-order model, realizing efficient reuse of unsteady wind fields.

[0116] The method of this invention is not only applicable to environmental impact assessment of construction projects and pollution source control in industrial parks, but can also be extended to scenarios such as emergency response to sudden environmental incidents, urban microclimate assessment, and pollution source parameter inversion. Through the architecture design of pre-constructed wind field storage and decoupled diffusion, a paradigm upgrade of multi-condition simulation from "fully coupled repeated solution" to "one-time solution, multiple reuses" has been achieved.

[0117] This invention achieves efficient coverage of the entire wind speed-direction parameter domain and cross-condition reuse of wind field data through a standardized wind field database pre-construction and structured indexing mechanism, combined with an adaptive encryption strategy guided by global sensitivity analysis. This reduces the computational power consumption of multi-condition simulation by more than 80%. Furthermore, it utilizes a dynamic diffusion coefficient adaptive model based on turbulence statistics and CFD to solve for the output turbulent kinetic energy. and dissipation rate This method achieves spatial adaptive matching between the diffusion coefficient and the local turbulence intensity on complex underlying surfaces, significantly improving the calculation accuracy of building wake regions and topographic undulation regions. Through unsteady wind field POD order reduction and rapid reconstruction methods, and by employing intrinsic orthogonal decomposition to extract dominant modes and construct a reduced-order model, it achieves efficient storage and rapid reconstruction of wind fields at any time in unsteady scenarios, significantly shortening the response time to sudden environmental events.

[0118] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions, characterized in that, Includes the following steps: S1: Wind field library pre-construction: Determine the wind speed-direction combination set of the target area according to the atmospheric environmental assessment specifications. For each combination in the wind speed-direction combination set, use unstructured grid CFD steady / unsteady wind field solution that fits the terrain and buildings to solve the wind field. Output standardized VTK format wind field files and construct a reusable wind field library with a structured index table. S2: Unstructured mesh topology preprocessing: Read the target wind field file in the reusable wind field library, extract the mesh topology data and wind field data, obtain the cell center coordinates, cell volume, and topological relationship of adjacent cells, automatically identify the terrain and building solid wall boundary cells, and complete the mesh quality screening and wind field data completion. S3: Operating Condition Matching and Wind Field Reuse: Read the parameter set of the target operating condition, and match the corresponding pre-calculated wind field data from the reusable wind field library through a structured index table based on the wind speed-direction parameters in the operating condition; when the wind speed-direction parameters of the new operating condition are consistent with those of the calculated operating condition, the loaded wind field data is directly reused. S4: Numerical solution for pollutant diffusion: Initialize the global concentration field, locate the corresponding grid cells according to the coordinates of the pollution source in the working condition, discretize the convection-diffusion equation using the finite volume method, combine the upwind scheme to handle the convection term and the second-order central difference scheme to handle the diffusion term, and solve the time step using the explicit Euler method to obtain the time evolution law and spatial distribution results of pollutant concentration. S5: Multi-condition iteration: Repeat steps S3-S4 to complete the simulation calculation of all target conditions and output the multi-condition comparison analysis results.

2. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, In step S1, the basic parameter range of the wind speed-wind direction combination set is: wind speed 1~10m / s, step size 1m / s; wind direction 0~360°, step size 10°, for a total of 360 basic combinations.

3. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, The reusable wind farm database is an adaptively encrypted reusable wind farm database. The adaptive encryption process is as follows: S1-1. Based on the global sensitivity analysis method, calculate the sensitivity coefficient of different wind speed-direction combinations on the pollutant diffusion results; S1-2. For the prevailing wind direction and critical wind speed range where the sensitivity coefficient is higher than the preset threshold, the wind speed step size is increased to 0.5 m / s and the wind direction step size is increased to 5°; for the non-sensitive range where the sensitivity coefficient is lower than the preset threshold, the wind speed step size is increased to 2 m / s and the wind direction step size is increased to 20°. S1-3. Supplement the CFD wind field solution for the newly added combinations after encryption / relaxation, and update the wind field library and structured index table.

4. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, In step S1, the fields of the structured index table include combination ID, wind speed, wind direction, absolute path of wind field file, wind field type, solution method, total number of grid cells, and calculation completion time. The wind field file uses a Binary VTK file, which includes node 3D coordinates, cell topology information, 3D wind speed components at cell centers, terrain elevation markings, and building solid-wall cell markers.

5. The multi-condition CFD simulation method for atmospheric pollutant diffusion according to claim 1, characterized in that, In step S2, the specific criteria for mesh quality screening are: removing cells with an aspect ratio > 10 and a volume < 1. e-6 For the m³ malformed cells, the inverse distance weighted interpolation method is used to complete the wind field data for the missing regions after removal; The specific method for boundary element identification is as follows: elements with less than 3 adjacent elements in a two-dimensional triangular mesh and elements with less than 4 adjacent elements in a three-dimensional tetrahedral mesh are identified as boundary elements.

6. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, In step S4, the diffusion term is calculated using a dynamic diffusion coefficient adaptive model based on CFD turbulence statistics, with the following formula: In the formula, This is a constant for the turbulence model, taken as 0.09; The turbulent kinetic energy at the center of the unit output by the CFD wind field; The turbulent kinetic energy dissipation rate at the center of the unit output of the CFD wind field.

7. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, In step S4, the time step of the explicit Euler method Satisfying the CFL stability condition: In the formula, the CFL number is taken as 0.5~0.8; Unit volume; The unit center wind speed modulus; This represents the maximum surface area of ​​the unit.

8. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, Step S4, after the numerical solution of pollutant diffusion, also includes boundary condition processing. Specifically, the boundary condition processing is as follows: the topographic and building solid wall boundaries adopt zero flux boundary conditions, and the flux of adjacent surfaces of the boundary cells is set to 0; the computational domain outlet boundary adopts zero gradient open boundary conditions; and the computational domain inlet and far-field boundaries adopt zero concentration boundary conditions.

9. The CFD simulation method for atmospheric pollutant diffusion under multiple operating conditions according to claim 1, characterized in that, In step S1, the method for processing unsteady wind fields is as follows: the time series of unsteady wind fields is processed by intrinsic orthogonal decomposition (POD) to reduce the order, and the first 20 dominant modes are extracted to construct a wind field reduction model, which is stored in the wind field library; in step S3, when matching unsteady wind fields, the wind field data at any time is quickly reconstructed through the reduction model and matched with the time step of diffusion calculation.

10. A multi-condition CFD simulation device for atmospheric pollutant diffusion, used to implement the multi-condition CFD simulation method for atmospheric pollutant diffusion as described in any one of claims 1 to 9, characterized in that: The multi-condition CFD simulation device for atmospheric pollutant diffusion includes a wind field database construction module, a mesh preprocessing module, a condition matching module, a diffusion solution module, and a multi-condition iteration module. The wind field database construction module performs wind field database pre-construction operations, the mesh preprocessing module performs unstructured mesh topology preprocessing operations, the condition matching module performs condition matching and wind field reuse operations, the diffusion solution module performs pollutant diffusion numerical solution operations, and the multi-condition iteration module performs multi-condition iteration and result output operations.

Images