Heat Source Location System Based on CFD Model

Abstract

The present invention discloses a heat source position positioning system based on a CFD model. The present invention relates to the technical field of building energy conservation and includes a simulation analysis platform. The simulation analysis platform is communicatively connected to a data input and processing module, a geometric model construction module, a parameter setting module, a CFD simulation execution module, a polymorphism analysis module, a flow state prediction module, and a result display module. Among them, the modules are electrically connected to each other. The data input and processing module is used to collect and process input data. The heat source position positioning system based on the CFD model evaluates the influence of different heat source positions on the flow state by simulating and predicting the flow state under different heat source positions, and predicts the possible thermo-pressure ventilation effect, so as to guide designers to optimize the heat source layout and improve the ventilation effect. This not only helps to improve the comfort of buildings, but also can effectively reduce energy consumption and achieve the goal of energy conservation and emission reduction.

Description

Technical Field

[0001] The present invention relates to the technical field of building energy conservation, and specifically to a heat source position positioning system based on a CFD model. Background Art

[0002] The CFD model is the computational fluid dynamics model. In underground buildings, the position of the heat source directly affects the temperature distribution and the flow field of thermosiphon ventilation. The thermosiphon ventilation in underground buildings often exhibits polymorphism, that is, under the same boundary conditions and geometric structures, due to the influence of factors such as initial conditions, heat source position and intensity, there will be multiple stable ventilation states. The existence of this phenomenon increases the difficulty of ventilation system design, and at the same time provides the possibility of optimizing the design. By using the simulation ability of the CFD model, the influence of different heat source positions inside the underground building on thermosiphon ventilation can be analyzed by simulating the ventilation states under different heat source positions, and further optimize the design and performance of the ventilation system.

[0003] In the prior art, there are polymorphism or multiple solution phenomena in the thermosiphon ventilation of underground buildings, that is, there are multiple stable flow states under the same conditions. It is difficult for the CFD model to accurately predict the final flow state during simulation. Therefore, how to analyze the stability and conversion conditions of different flow states and predict the final flow state is the problem we need to solve. For this reason, a heat source position positioning system based on a CFD model is proposed. Summary of the Invention

[0004] To achieve the above object, the present invention is realized through the following technical solutions: A heat source position positioning system based on a CFD model, including a simulation and analysis platform, which is communicatively connected to a data input and processing module, a geometric model construction module, a parameter setting module, a CFD simulation execution module, a polymorphism analysis module, a flow state prediction module, and a result display module. Among them, the modules are electrically connected to each other;

[0005] The data input and processing module is used to collect and process input data, including building geometric structure, heat source characteristics, and boundary conditions, to ensure the accuracy and effectiveness of the simulation, and provide necessary basic data for subsequent modules;

[0006] The geometric model construction module constructs a geometric model of the underground building according to the input data, conducts mesh division, and conducts mesh convergence analysis and time step independence verification to generate a computational domain for CFD simulation, ensuring the accuracy and computational efficiency of the CFD simulation, reducing truncation errors and computational convergence problems;

[0007] The parameter setting module is used to set the position, intensity, and distribution parameters of the heat source, simulate the thermo - pressure ventilation effect under different heat source configurations, analyze its influence on the flow state, and set the initial temperature, pressure conditions, and wall, inlet, and outlet boundary conditions for the simulation to ensure that the simulation process conforms to the actual situation and improve the reliability of the simulation results;

[0008] The CFD simulation execution module runs the CFD model to simulate the thermo - pressure ventilation process in the underground building, generating detailed temperature and velocity distribution data to provide data support for analyzing the flow state;

[0009] The polymorphism analysis module is used to analyze the CFD simulation results, identify different flow states and their stability, determine the conditions for the existence of polymorphism, and evaluate the stability and transition conditions of each flow state;

[0010] The flow state prediction module combines machine learning algorithms, trains the model using historical data and CFD simulation results, predicts the final flow state under different heat source positions, improves the prediction accuracy and efficiency, and solves the problem that the CFD model is difficult to directly predict the final flow state;

[0011] The result display module provides a user interface to assist the user in inputting the heat source position, viewing the simulation results and prediction results, enhancing the usability of the system, and enabling the user to intuitively understand the complex process and polymorphism phenomenon of thermo - pressure ventilation in the underground building.

[0012] Preferably, in the data input and processing module, the process of collecting and processing input data includes:

[0013] Collect input data on the building geometric structure, heat source characteristics, and boundary conditions based on building design drawings, field measurement data, and historical record data. Among them, the building geometric structure includes the size, shape, internal layout, and construction information of the building, the heat source characteristics include the position, intensity, type (such as equipment heat dissipation, human heat dissipation, etc.) and its working time information of the heat source, and the boundary conditions include outdoor temperature, humidity, wind speed, and wind direction information;

[0014] Conduct a preliminary verification of the collected data, check whether the data is complete, accurate, and meets the simulation requirements, check whether there are logical errors in the building geometric structure data, whether the heat source characteristics data is within a reasonable range, clean the data, process the problem data found during the verification process, and at the same time, standardize the data to ensure the consistency of data format and units;

[0015] Convert the data format of the collected input data and integrate data from different sources into a unified data set to ensure the logical relationship and time consistency between the data;

[0016] Convert the integrated data into a format acceptable to the CFD model, including generating geometric files for mesh generation and case files for simulation, store the processed data in the database of the simulation analysis platform for subsequent modules to call and analyze, establish a data management system, and classify, index, and back up the stored data to ensure data security and accessibility.

[0017] Preferably, in the geometric model construction module, the process of generating the CFD simulation computational domain includes:

[0018] According to the collected input data of the building geometric structure (including the size, shape, internal layout, and construction information of the building), use the Geometry module (CFD pre-processing software) provided by ANSYS Fluent to create a preliminary geometric model of the underground building;

[0019] Refine the preliminary geometric model of the underground building to depict the complex structure and boundary conditions inside the building, including the main structure of the building, the location of local heat sources, and the inlet and outlet channels, ensure that all building features and internal structures are accurately reflected, and verify whether the geometric model is consistent with the actual building and design drawings, and check whether there are defects in the model, such as overlapping surfaces, missing walls, or other structural problems;

[0020] After the geometric model is constructed, perform mesh generation on the geometric model to generate a mesh system for CFD simulation;

[0021] By gradually refining the mesh and comparing the simulation results at different mesh densities, determine whether the simulation results tend to be stable as the mesh density increases;

[0022] Set different time steps for simulation and observe the variation of the simulation results with the time step;

[0023] After completing mesh generation and verification, generate a computational domain for CFD simulation according to the mesh generation and boundary condition settings, ensure that the computational domain completely covers the entire geometric model, and in the computational domain, include all geometric structures and mesh information of the underground building, providing a basis for subsequent boundary condition settings, initial condition settings, and simulation solution steps;

[0024] Perform a pre-simulation before the formal simulation, check the physical reasonableness of the model and calculate the stability, adjust the simulation parameters to solve any convergence problems, and optimize the geometric model according to the results of the pre-simulation. The optimization includes adjusting the mesh density and modifying the boundary conditions, and output the final geometric model and mesh system to the CFD simulation execution module.

[0025] Preferably, in the parameter setting module, the process of setting the heat source location, intensity, and distribution parameters includes:

[0026] According to the actual building situation, determine the exact location of the heat source, and in the Geometry module provided by ANSYS Fluent, mark the location of the heat source, mark the selected location as the heat source area. Based on the type of heat source, define the heat flux of the heat source in the material properties, and set the distribution methods of the heat source, namely point heat source, line heat source, surface heat source, and the corresponding distribution densities, to simulate the thermal pressure ventilation effect under different heat source configurations, and analyze the influence of the heat source configuration on the flow velocity, flow direction, and temperature distribution;

[0027] Based on the temperature of the surrounding environment of the underground building, set the initial temperature of the entire simulation area, and set the initial pressures of the internal and external environments of the building at the start of the simulation. For natural ventilation simulation, set the static pressure. For an open system (such as an underground building with inlet and outlet channels), set the static pressure at the inlet and outlet to the ambient pressure, and perform iterative calculations in the internal area;

[0028] Set the walls of the underground building as non-slip and adiabatic surfaces, so that the walls do not participate in heat exchange and the fluid does not slide on the walls. For the inlet and outlet channels, based on the simulation requirements, set them as pressure inlets. If the simulation is of natural ventilation, set the static pressure at the inlet to the ambient pressure, set the boundary conditions when the air leaves the building, set the outlet channel as a pressure outlet, and set the static pressure to the ambient pressure.

[0029] Preferably, in the CFD simulation execution module, the generation process of the temperature and velocity distribution data includes:

[0030] Load the verified geometric model and grid system in the CFD software, import the heat source location, intensity, distribution parameters, as well as the boundary conditions and initial conditions defined in the parameter setting module;

[0031] Select the pressure-based solver and set the fluid properties, define the physical properties of air such as density, specific heat capacity, and thermal conductivity, and select the fluid dynamics model to simulate the fluid behavior. Synchronously perform grid independence verification to ensure that the generated grid meets the grid independence requirements to guarantee the accuracy of the simulation results;

[0032] Configure the time step, relaxation factor, and convergence criterion parameters of the pressure-based solver, start the pressure-based solver in the CFD software, perform iterative calculations, monitor the residual curve during the solution process, monitor the progress of the simulation and the usage of computing resources, calculate the metric for convergence determination, and detect whether there are problems of numerical instability and non-convergence until the convergence criterion is reached;

[0033] Collect the distribution data of the temperature and velocity physical quantities generated during the simulation process, and use the post-processing tool of the CFD software to view the temperature distribution map inside the underground building, view the velocity vector map, and analyze the flow state and direction of the fluid inside the underground building;

[0034] Conduct a preliminary analysis of the simulation results, check for abnormal or physically unrealistic data, and post-process the simulation results to extract key data on the flow state and visualize it.

[0035] Preferably, the calculation expression for the convergence determination metric value is:

[0036]

[0037] where C v is the metric value for convergence determination, i.e., the square root of the sum of the squares of the residual changes, R i is the current value of the residual at the i-th iteration step, R b，i is the reference value of the residual at the i-th iteration step, which is the residual value at the initial iteration step, ∈ is a pre-set convergence criterion, a small positive number, usually less than 1, n is the number of residuals involved in the calculation, and C v has a value range of [0, +∞). When the value of C v is smaller, it indicates that the residual change is smaller and the simulation is closer to convergence. The value ranges of R i and R b，i are [0, +∞), where R i decreases as the number of iterations increases.

[0038] Preferably, in the polymorphism analysis module, the process of determining the conditions for the existence of polymorphism includes:

[0039] Extract the distribution data of physical quantities such as temperature, velocity, and pressure from the CFD simulation results, and import the extracted data into the polymorphism analysis module to ensure the integrity and accuracy of the data;

[0040] Analyze the simulation data, automatically identify the flow states in the simulation results, including laminar flow, turbulent flow, and transitional flow. For each flow state, extract the characteristic parameters of its velocity distribution, vortex structure, and turbulence intensity;

[0041] For each identified flow state, evaluate its stability by analyzing the change trend of the flow state over time, determine whether it is a stable state, analyze the conversion conditions between different flow states, and determine the critical parameters of the heat source intensity and inlet velocity for state conversion;

[0042] Search for regions in the simulation results that satisfy the polymorphism conditions, identify the critical values of the key parameters of velocity, pressure, and temperature that affect polymorphism, conduct a sensitivity analysis of the key parameters, evaluate the impact of parameter changes on the flow state, and verify whether the determined polymorphism conditions are accurate by comparing the simulation results under different conditions.

[0043] Preferably, in the flow state prediction module, the process of predicting the final flow state at different heat source positions includes:

[0044] Collect historical parameter data and generated CFD simulation data, and clean and standardize the collected historical parameter data and CFD simulation data. Data cleaning removes noise and outliers, and data standardization enables data with different dimensions to be compared and calculated on the same scale. Among them, the historical parameter data is the flow state data at different heat source positions, including key parameters of temperature distribution, velocity distribution, and pressure distribution, which are derived from actual experimental measurements and existing CFD simulation results. The generated CFD simulation data is obtained by using CFD software to simulate different heat source positions and acquiring the corresponding flow state data;

[0045] Divide the collected data into a training set and a test set, select a physics-informed neural network model for model training, use the training set data to train the physics-informed neural network model, construct a flow state prediction model, and use the test set data to verify the model and evaluate the prediction accuracy and generalization ability of the model;

[0046] Based on the flow state prediction model, historical parameter data, and generated CFD simulation data, comprehensively analyze to obtain a flow state evaluation coefficient, and predict the flow state and trend at different heat source positions;

[0047] Deploy the trained flow state prediction model to the CFD simulation platform, input new heat source position parameters into the flow state prediction model, and the flow state prediction model predicts according to the rules obtained from historical data and CFD simulation results, and outputs prediction results, including key parameters of temperature distribution, velocity distribution, and pressure distribution;

[0048] Analyze and verify the prediction results, evaluate their accuracy and practicality, and feedback the prediction results to relevant researchers or engineers to provide strong support for solving practical problems.

[0049] Preferably, the calculation expression of the flow state evaluation coefficient is:

[0050]

[0051] where FE is the flow state evaluation coefficient, used to measure the overall change of the flow state, T j is the temperature at the j-th measurement point under a specific heat source position, T b is the reference temperature, which is the initial temperature under the heat source position, ΔP j is the pressure change at the j-th measurement point, P b is the reference pressure, which is the initial pressure of the system, V jis the velocity of the jth measurement point at a specific heat source position, V b is the reference velocity, is the initial velocity at the heat source position, m is the total number of measurement points, and the value range of FE is [0, +∞), where FE=0 means no flow state change, and the larger the FE value, the more drastic the flow state change.

[0052] Preferably, in the result display module, the process of assisting the user to input the heat source location, view the simulation results and the prediction results includes:

[0053] Provide a user interface that allows users to easily input parameters, such as the location of the heat source, provide clear instructions and prompts, and guide users to operate, integrating the fluid dynamics data obtained from CFD simulation and the prediction results of the machine learning model into the result display module;

[0054] Visualize temperature distribution, velocity vectors, and pressure using graphics and animations, and provide views at different perspectives and zoom levels to fully understand the data;

[0055] Present simulation and prediction results, including stability of flow regimes and flow trends, and interpret simulation results, including description of flow regimes and explanation of polymorphic phenomena;

[0056] Output simulation reports and display prediction results, provide comparative views, and show the differences between different prediction results. During the result display process, provide interactive annotation functions, allowing users to add comments, arrows, text boxes and other elements to the graph to record and share their understanding and analysis of the results.

[0057] The present invention provides a heat source location positioning system based on a CFD model. It has the following beneficial effects:

[0058] 1. The heat source location positioning system based on the CFD model simulates and predicts the flow state under different heat source positions, evaluates the impact of different heat source positions on the flow state, and predicts the possible thermal pressure ventilation effect, thereby guiding designers to optimize the heat source layout and improve the ventilation effect. It not only helps to improve the comfort of the building, but also effectively reduces energy consumption and achieves the goal of energy conservation and emission reduction. By integrating the physical information neural network machine learning algorithm and making full use of historical data and CFD simulation results, it can realize accurate prediction of the flow state under different heat source positions, significantly improve the accuracy of the prediction, and at the same time, make the prediction process more efficient, and be able to complete the processing and analysis of a large amount of data in a short time, providing timely decision support for engineering design.

[0059] II. The heat source location positioning system based on the CFD model conducts a detailed CFD simulation of the underground building, simulates the thermal pressure ventilation effect under different heat source positions, analyzes the influence of heat source configuration on the flow velocity, flow direction, and temperature distribution, uses the polymorphism analysis module to identify and evaluate the stability of different flow states, determines the critical parameters for state conversion, and can effectively control the heat distribution inside the building, reduce energy waste, and improve comfort by adjusting the heat source position. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a block diagram of the module composition of the heat source location positioning system based on the CFD model of the present invention;

[0061] Figure 2 It is a flowchart for generating temperature and velocity distribution data of the present invention;

[0062] Figure 3 It is a flowchart for predicting the final flow state under different heat source positions of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] The present invention will be further described in detail below with reference to the drawings and specific embodiments. The embodiments of the present invention are given for purposes of illustration and description, and are not exhaustive or limit the present invention to the disclosed form. Many modifications and variations are obvious to those of ordinary skill in the art. The embodiments are selected and described to better illustrate the principles and practical applications of the present invention, and enable those of ordinary skill in the art to understand the present invention and design various embodiments with various modifications suitable for specific purposes.

[0064] The first embodiment, as Figure 1 , Figure 2 shown, the present invention provides a technical solution: a heat source location positioning system based on the CFD model, including a simulation analysis platform, which is communicatively connected to a data input and processing module, a geometric model construction module, a parameter setting module, a CFD simulation execution module, a polymorphism analysis module, a flow state prediction module, and a result display module. Among them, the modules are electrically connected to each other;

[0065] Data input and processing module, which is used to collect and process input data, including building geometric structure, heat source characteristics, and boundary conditions, to ensure the accuracy and effectiveness of the simulation and provide necessary basic data for subsequent modules. Based on building design drawings, on-site measurement data, and historical record data, collect input data of building geometric structure, heat source characteristics, and boundary conditions. Among them, the building geometric structure includes the size, shape, internal layout, and construction information of the building; the heat source characteristics include the location, intensity, type (such as equipment heat dissipation, human heat dissipation, etc.), and working time information of the heat source; the boundary conditions include outdoor temperature, humidity, wind speed, and wind direction information. Conduct preliminary verification on the collected data, check whether the data is complete, accurate, and meets the simulation requirements, check whether there are logical errors in the building geometric structure data, whether the heat source characteristics data is within a reasonable range, and clean the data to process the problematic data found during the verification process. At the same time, standardize the data to ensure the consistency of data format and unit. Convert the format of the collected input data, and integrate data from different sources into a unified dataset to ensure the logical relationship and time consistency between the data. Convert the integrated data into a format acceptable to the CFD model, including generating geometric files for mesh generation and case files for simulation. Store the processed data in the database of the simulation analysis platform for subsequent modules to call and analyze. Establish a data management system to classify, index, and back up the stored data to ensure data security and accessibility;

[0066] The geometric model construction module constructs the geometric model of the underground building according to the input data, performs meshing, and performs mesh convergence analysis and time step independence verification, generates the calculation domain for CFD simulation, ensures the accuracy and calculation efficiency of CFD simulation, reduces truncation errors and calculation convergence problems, and uses the Geometry module (CFD pre-processing software) provided by ANSYS Fluent to create a preliminary geometric model of the underground building based on the collected building geometry input data (including the building's size, shape, internal layout and construction information). The preliminary geometric model of the underground building is refined to depict the complex structure and boundary conditions inside the building, including the main structure of the building, the location of local heat sources, and import and export channels, to ensure that all building features and internal structures are accurately reflected, and to verify whether the geometric model is consistent with the actual building and design drawings. Check whether the model has defects, such as overlapping surfaces, missing walls or other structural problems. After the geometric model is built, the geometric model is meshed to generate a grid system for CFD simulation. The quality of meshing directly affects the accuracy and computational efficiency of CFD simulation. Complex geometric structures in underground buildings and areas with drastic temperature changes require finer meshes to ensure that the detailed features of the flow field can be accurately captured. For key areas such as around heat sources, vents, and connecting channels, local encryption processing should be performed. In the process of meshing, pay attention to the shape, size, and direction of the mesh. In order to avoid large truncation errors in the calculation, use high-quality meshes (such as structured meshes, hexahedral meshes), and control the distortion and aspect ratio of the mesh. By gradually refining the mesh and comparing the simulation results under different mesh densities, the accuracy of the mesh can be judged. Whether the simulation results tend to be stable with the increase of grid density. If the changes of simulation results are within an acceptable range after the grid is encrypted to a certain extent, it is considered that the current grid density has reached convergence. Determine a grid density that can ensure simulation accuracy and has a reasonable calculation cost, set different time steps for simulation, and observe the changes of simulation results with time steps. If the changes of simulation results are very small after the time step is reduced to a certain extent, it is considered that the current time step setting is reasonable. Ensure that the selected time step can accurately capture the dynamic changes of physical phenomena and avoid numerical instability. After completing mesh division and verification , based on the grid division and boundary condition settings, generate the computational domain for CFD simulation, ensure that the computational domain completely covers the entire geometric model, including all the geometric structures and grid information of the underground building in the computational domain, and provide a basis for the subsequent boundary condition settings, initial condition settings, and simulation solution steps. Perform a pre-simulation before the formal simulation to check the physical rationality of the model and calculate the stability, adjust the simulation parameters to solve any convergence problems, and optimize the geometric model based on the results of the pre-simulation. The optimization includes adjusting the grid density and modifying the boundary conditions. The final geometric model and grid system are output to the CFD simulation execution module;

[0067] The parameter setting module is used to set the location, intensity, and distribution parameters of the heat source, simulate the heat pressure ventilation effect under different heat source configurations, analyze its impact on the flow state, and set the initial temperature, pressure conditions, and wall, inlet, and outlet boundary conditions of the simulation to ensure that the simulation process conforms to the actual situation and improve the reliability of the simulation results. According to the actual building situation, the exact location of the heat source is set and the simulation results are analyzed in ANSYS. In the Geometry module that comes with Fluent, mark the location of the heat source and mark the selected location as the heat source area. Based on the type of heat source, define the heat flux of the heat source in the material properties and set the distribution mode of the heat source, which are point heat source, line heat source, surface heat source, and the corresponding distribution density. Simulate the heat pressure ventilation effect under different heat source configurations, analyze the influence of heat source configuration on flow rate, flow direction and temperature distribution, set the initial temperature of the entire simulation area based on the temperature of the surrounding environment of the underground building, and set the initial pressure of the internal and external environments of the building at the beginning of the simulation. For natural ventilation simulation, set the static pressure. For open systems (such as underground buildings with inlet and outlet channels), set the static pressure of the inlet and outlet to the ambient pressure, and perform iterative calculations in the internal area. Set the wall of the underground building as a non-slip and adiabatic surface so that the wall does not participate in heat exchange and the fluid does not slide on the wall. For the inlet and outlet channels, set them as pressure inlets based on the simulation requirements. If natural ventilation is simulated, set the static pressure of the inlet to the ambient pressure. Set the boundary conditions when the air leaves the building, set the outlet channel to the pressure outlet, and set the static pressure to the ambient pressure.

[0068] CFD simulation execution module, which runs the CFD model to simulate the thermosiphon ventilation process in an underground building, generates detailed temperature and velocity distribution data to provide data support for analyzing the flow state. In the CFD software, load the verified geometric model and grid system, import the heat source position, intensity, distribution parameters, as well as boundary conditions and initial conditions defined in the parameter setting module, select the pressure-based solver and set the fluid properties, define the physical properties of air such as density, specific heat capacity, and thermal conductivity, and select the fluid dynamics model to simulate the fluid behavior. Synchronously perform grid independence verification to ensure that the generated grid meets the grid independence requirements to guarantee the accuracy of the simulation results. Configure the time step, relaxation factor, and convergence criterion parameters of the pressure-based solver, start the pressure-based solver in the CFD software, perform iterative calculations, monitor the residual curve during the solution process, monitor the progress of the simulation and the usage of computing resources, calculate the metric value for convergence determination, detect whether there are problems of numerical instability and non-convergence until the convergence criterion is reached, collect the distribution data of the temperature and velocity physical quantities generated during the simulation process, and use the post-processing tool of the CFD software to view the temperature distribution map in the underground building, view the velocity vector map, analyze the flow state and direction of the fluid in the underground building, conduct a preliminary analysis of the simulation results, check whether there are abnormal or physically unrealistic data, and post-process the simulation results to extract the key data of the flow state and visualize the display;

[0069] Furthermore, the calculation expression of the convergence determination metric value is:

[0070]

[0071] where C v is the metric value for convergence determination, that is, the square root of the sum of the squares of the residual changes, R i is the current value of the residual at the i-th iteration step, R b，i is the reference value of the residual at the i-th iteration step, which is the residual value at the initial iteration step, ∈ is a pre-set convergence criterion, a small positive number, usually less than 1, n is the number of residuals involved in the calculation, and the value range of C v is [0, +∞). When the value of C v is smaller, it indicates that the residual change is smaller and the simulation is closer to convergence. The value ranges of R i and R b，i are [0, +∞), where R i decreases as the number of iterations increases. The value range of ∈ is (0, 1], which is a pre-set threshold used to control the strictness of convergence. C v calculates the square root of the sum of the squares of the deviations of all monitored residuals from their respective reference values, giving a comprehensive convergence determination index. When C vWhen the value is lower than the set threshold ∈, it is considered that the solver has reached the convergence criterion. If C v If the value does not decrease or even increases after several consecutive iterations, it indicates numerical instability during the solution process. During the iteration process, it is necessary to monitor the residual curve in real time to see if it decreases as the number of iterations increases. If the residual curve shows an oscillating or increasing trend, it may be necessary to adjust the time step, relaxation factor, or recheck the physical model and boundary conditions;

[0072] The polymorphism analysis module is used to analyze the CFD simulation results, identify different flow states and their stabilities, determine the conditions for the existence of polymorphism, evaluate the stabilities and transition conditions of each flow state, extract the distribution data of physical quantities such as temperature, velocity, and pressure from the CFD simulation results, and import the extracted data into the polymorphism analysis module to ensure the integrity and accuracy of the data, analyze the simulation data, automatically identify the flow states in the simulation results, including laminar flow, turbulent flow, and transitional flow, extract the characteristic parameters of velocity distribution, vortex structure, and turbulent intensity for each flow state, for each identified flow state, evaluate its stability by analyzing the change trend of the flow state over time, determine whether it is a stable state, analyze the transition conditions between different flow states, determine the critical parameters of heat source intensity and inlet flow velocity for state transition, search for regions that meet the polymorphism conditions in the simulation results, identify the critical values of key parameters of velocity, pressure, and temperature that affect polymorphism, conduct a sensitivity analysis of the key parameters, evaluate the impact of parameter changes on the flow state, and verify whether the determined polymorphism conditions are accurate by comparing the simulation results under different conditions;

[0073] The flow state prediction module combines machine learning algorithms, uses historical data and CFD simulation results to train the model, predicts the final flow state under different heat source positions, improves the prediction accuracy and efficiency, and solves the problem that it is difficult for the CFD model to directly predict the final flow state;

[0074] The result display module provides a user interface to assist the user in inputting the heat source position, viewing the simulation results and prediction results, enhances the usability of the system, and enables the user to intuitively understand the complex process and polymorphism phenomenon of natural ventilation by thermal pressure in underground buildings.

[0075] The second embodiment, on the basis of the first embodiment, please refer to Figure 3 As shown, in the flow state prediction module, the process of predicting the final flow state under different heat source positions includes:

[0076] Collect historical parameter data and generated CFD simulation data, clean and standardize the collected historical parameter data and CFD simulation data, remove noise and outliers, and standardize data so that data of different dimensions can be compared and calculated on the same scale. The historical parameter data are flow state data at different heat source positions, including key parameters of temperature distribution, velocity distribution, and pressure distribution, which are derived from actual experimental measurements and existing CFD simulation results. The generated CFD simulation data are simulated at different heat source positions using CFD software to obtain the corresponding flow state data. The collected data are divided into training sets and test sets, and a physical information neural network model is selected for model training. The physical information neural network model is trained using the training set data to construct a flow Dynamic state prediction model, and use test set data to verify the model, evaluate the prediction accuracy and generalization ability of the model, based on the flow state prediction model, historical parameter data and generated CFD simulation data, comprehensive analysis to obtain the flow state evaluation coefficient, predict the flow state and trend under different heat source positions, deploy the trained flow state prediction model to the CFD simulation platform, input new heat source position parameters into the flow state prediction model, the flow state prediction model predicts according to the rules obtained by training with historical data and CFD simulation results, outputs the prediction results, including key parameters of temperature distribution, velocity distribution, and pressure distribution, analyzes and verifies the prediction results, evaluates their accuracy and practicality, and feeds back the prediction results to relevant researchers or engineers to provide strong support for solving practical problems;

[0077] Furthermore, the calculation expression of the flow state evaluation coefficient is:

[0078]

[0079] Where FE is the flow state evaluation coefficient, which is used to measure the overall change of flow state, T j is the temperature of the jth measurement point at a specific heat source location, T b is the reference temperature, is the initial temperature at the heat source position, ΔP j is the pressure change at the jth measurement point, P b is the reference pressure, the initial pressure of the system, V j is the velocity of the jth measurement point at a specific heat source position, V b is the reference velocity, is the initial velocity at the heat source position, m is the total number of measurement points, and the value range of FE is [0, +∞), where FE = 0 means no flow state change, and the larger the FE value, the more drastic the flow state change, T j , ΔP j , and V jThe value range depends on the actual measurement data. As the position of the heat source changes, the value of FE will vary according to the change in the flow state. The exponential function is used to measure the temperature change. The exponential function is very effective in capturing the non-linear characteristics of temperature change. When the temperature change is small, the value of the exponential function is close to 1 and the change is not obvious; when the temperature change is large, the value of the exponential function increases significantly, indicating a significant change in the flow state. By measuring the impact of pressure change on the flow state, the square is used to emphasize the impact of large pressure changes. By measuring the impact of velocity change on the flow state, the square is used to emphasize the impact of large velocity changes. The changes in temperature, pressure, and velocity are integrated through a summation function to form a flow state evaluation coefficient. Taking the square root converts the flow state evaluation coefficient into a unitless quantity for easy comparison and analysis;

[0080] In the result display module, the process of assisting the user to input the heat source position, view the simulation results and prediction results includes:

[0081] Providing a user interface that allows the user to easily input parameters such as the heat source position, providing clear instructions and prompts, and guiding the user to operate. Integrating the hydrodynamic data obtained from the CFD simulation and the prediction results of the machine learning model into the result display module, visualizing physical quantities such as temperature distribution, velocity vector, and pressure using graphics and animations, and providing views with different perspectives and zoom levels so that the user can comprehensively understand the data. Displaying the simulation and prediction results, including the stability of the flow state and the flow trend, and interpreting the simulation results, including the description of the flow state and the explanation of the polymorphism phenomenon. Outputting a simulation report and displaying the prediction results, providing a comparison view to show the differences between different prediction results. During the result display process, providing an interactive annotation function that allows the user to add elements such as annotations, arrows, and text boxes to the graphics for recording and sharing the understanding and analysis of the results.

[0082] Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. All other embodiments obtained by those of ordinary skill in the art and related fields based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention, unless otherwise specified and limited, are implemented according to the conventional means in the art.

Claims

1. A heat source location positioning system based on a CFD model, including a simulation analysis platform, characterized in that: The simulation analysis platform is connected with a data input and processing module, a geometric model building module, a parameter setting module, a CFD model simulation execution module, a polymorphism analysis module, a flow state prediction module and a result display module, wherein the electrical signals between the modules are connected; Data input and processing module, used to collect and process input data, including building geometry, heat source characteristics, and boundary conditions; The geometric model building module is used to build the geometric model of the underground building according to the input data, perform meshing, mesh convergence analysis and time step independence verification, and generate the calculation domain for CFD model simulation; The parameter setting module is used to set the heat source position, intensity, and distribution parameters, simulate the heat pressure ventilation effect under different heat source configurations, analyze the influence of heat source configuration on the flow state, and set the initial temperature, pressure conditions, and wall, inlet, and outlet boundary conditions of the simulation; The CFD model simulation execution module is used to run the CFD model, simulate the heat pressure ventilation process in the underground building, generate the distribution data of temperature, velocity and pressure, and import the generated data into the polymorphism analysis module; Polymorphism analysis module, used to analyze CFD model simulation results, identify different flow states and their stability, determine the conditions for the existence of polymorphism, and evaluate the stability and conversion conditions of each flow state; Conditions that determine the presence of polymorphism include: Analyze the CFD model simulation results and automatically identify the flow states in the CFD model simulation results, including laminar flow, turbulent flow, and transitional flow. For each flow state, extract the characteristic parameters of velocity distribution, vortex structure, and turbulence intensity; For each identified flow state, determine whether it is a stable state by analyzing the change trend of the flow state over time, analyze the conversion conditions between different flow states, and determine the critical parameters of the heat source intensity and inlet flow velocity of the state conversion; Search for areas that meet the polymorphism conditions in the CFD model simulation results, identify the key parameters of temperature distribution, velocity distribution, and pressure distribution that affect the polymorphism, perform sensitivity analysis on the key parameters, and verify whether the conditions for the polymorphism determined are correct by comparing the CFD model simulation results under different conditions; The flow state prediction module is used to combine the machine learning algorithm, use the historical data and CFD model simulation results to train the flow state prediction model, and predict the final flow state under different heat source positions; based on the flow state prediction model, historical data and the generated CFD model simulation results, the flow state evaluation coefficient is obtained through comprehensive analysis; The calculation expression of flow state evaluation coefficient is: ； in, is the flow state assessment coefficient, It is The temperature at the heat source location, is the reference temperature, It is The pressure change at the heat source position is is the base pressure, It is The speed at the heat source position, is the base speed, is the total number of heat source locations; The result display module is used to provide a user interface to assist users in inputting heat source locations, viewing simulation results and prediction results.

2. The heat source location positioning system based on the CFD model according to claim 1, characterized in that: In the data input and processing module, the process of collecting and processing input data includes: Collect input data of building geometry, heat source characteristics, and boundary conditions based on building design drawings, field measurement data, and historical data, where the building geometry includes the size, shape, internal layout, and construction information of the building; the heat source characteristics include the location, intensity, type, and working time of the heat source; and the boundary conditions include outdoor temperature, humidity, wind speed, and wind direction information; Conduct preliminary verification of the input data of the building geometry, heat source characteristics, and boundary conditions, clean the input data, process the problematic data found during the verification process, and standardize the input data; Convert the collected input data into different formats and integrate the input data from different sources into a unified data set; Convert the integrated input data into a format acceptable to the CFD model, including generating geometry files for meshing and case files for simulation, storing the processed input data in the database of the simulation analysis platform, and establishing a data management system to classify, index and back up the stored input data.

3. The heat source location positioning system based on the CFD model according to claim 2, characterized in that: In the geometric model building module, the process of generating a computational domain for CFD model simulation includes: Based on the collected building geometry input data, the Geometry module provided by ANSYS Fluent was used to create a preliminary geometric model of the underground building; Refine the preliminary geometric model of the underground building to depict the complex structure and boundary conditions inside the building, including the main structure of the building, the location of local heat sources, and the entrance and exit passages. Verify whether the refined geometric model is consistent with the actual building and design drawings, and check whether the refined geometric model has any defects. After the refined geometric model is constructed, the refined geometric model is meshed to generate a mesh system for CFD model simulation; By gradually refining the grid and comparing the CFD model simulation results under different grid densities, it is determined whether the CFD model simulation results tend to be stable with the increase of grid density; Set different time steps for simulation and observe how the CFD model simulation results change with the time step; After the meshing and verification are completed, the computational domain for CFD model simulation is generated according to the meshing and boundary condition settings. The computational domain includes all the geometric structures and mesh information of the underground building. A pre-simulation is performed before the formal simulation to check the physical rationality of the refined geometric model and calculate the stability. According to the results of the pre-simulation, the refined geometric model is optimized. The optimization includes adjusting the grid density, modifying the boundary conditions, and obtaining the final geometric model. The final geometric model and grid system are output to the CFD model simulation execution module.

4. The heat source location positioning system based on the CFD model according to claim 3, characterized in that: In the parameter setting module, the process of setting the heat source position, intensity, and distribution parameters includes: According to the actual building conditions, the exact location of the heat source is set, and the location of the heat source is marked in the Geometry module of ANSYS Fluent. The selected location is marked as the heat source area. Based on the type of heat source, the heat flux of the heat source is defined in the material properties, and the distribution mode of the heat source is set, which are point heat source, line heat source, surface heat source, and the corresponding distribution density. The heat pressure ventilation effect under different heat source configurations is simulated, and the influence of heat source configuration on flow rate, flow direction and temperature distribution is analyzed; Based on the temperature of the surrounding environment of the underground building, set the initial temperature of the entire simulation area, and set the initial pressure of the building's internal and external environment at the beginning of the simulation. For natural ventilation simulation, set the static pressure. For open systems, set the static pressure of the inlet and outlet to the ambient pressure, and perform iterative calculations in the internal area. Set the wall surface of the underground building as a non-slip and insulating surface. For the inlet and outlet channels, set them as pressure inlets based on the simulation requirements. If natural ventilation is simulated, set the static pressure of the inlet to ambient pressure. Set the boundary conditions when the air leaves the building, set the outlet channel to a pressure outlet, and set the static pressure to ambient pressure.

5. The heat source location positioning system based on the CFD model according to claim 4, characterized in that: In the CFD model simulation execution module, the generation process of the distribution data of temperature, velocity and pressure includes: Load the final geometric model and grid system into the CFD model software, and import the heat source position, intensity, distribution parameters, initial temperature, pressure conditions, and wall, inlet, and outlet boundary conditions defined in the parameter setting module; Select the pressure-based solver and set the fluid properties, define the physical properties of air density, specific heat capacity, and thermal conductivity, select the CFD model to simulate the fluid behavior, and perform grid independence verification simultaneously; Configure the time step, relaxation factor and convergence criteria parameters of the pressure-based solver, start the pressure-based solver in the CFD model software, perform iterative calculations, monitor the residual curve during the solution process, monitor the progress of the simulation and the use of computing resources, and calculate the metrics for convergence determination; Collect the distribution data of temperature, velocity, and pressure generated during the simulation process, and use the post-processing tools of the CFD model software to view the temperature distribution map in the underground building, view the velocity vector map, and analyze the flow state and direction of the fluid in the underground building; Conduct preliminary analysis on the CFD model simulation results to check for abnormal data, and post-process the CFD model simulation results to extract key data of the flow state and visualize them; The calculation expression of the metric value of the convergence judgment is: ； in, is the measure of convergence, It is The current value of the residual for the iteration step, It is The baseline value of the residual of the iteration step, is a pre-set convergence criterion, is the number of residuals.

6. The heat source location positioning system based on the CFD model according to claim 5, characterized in that: In the flow state prediction module, the process of predicting the final flow state at different heat source positions includes: Collect historical data and generated CFD model simulation results, and clean and standardize the collected historical data and CFD model simulation results. The historical data is the flow state data at different heat source positions, including key parameters of temperature distribution, velocity distribution, and pressure distribution. The generated CFD model simulation results are obtained by simulating different heat source positions using CFD model software to obtain corresponding flow state data. The collected historical data and the generated CFD model simulation results are divided into a training set and a test set, a physical information neural network model is selected for training, the training set is used to train the physical information neural network model, a flow state prediction model is constructed, and the test set is used to verify the flow state prediction model; Deploy the trained flow state prediction model to the CFD model software, input the new heat source position parameters into the flow state prediction model, and the flow state prediction model makes predictions based on the rules obtained by training the historical data and CFD model simulation results, and outputs the prediction results, including the key parameters of temperature distribution, velocity distribution, and pressure distribution; Analyze and verify the forecast results, and provide feedback to relevant personnel.

7. The heat source location positioning system based on CFD model according to claim 6, characterized in that: In the result display module, the process of assisting the user to input the heat source location, view the simulation results and the prediction results includes: Provide a user interface to allow the user to input parameters and guide the user to perform operations, and integrate the obtained fluid dynamics data and the prediction results of the flow state prediction model into the result display module; Visualize temperature, velocity, and pressure physical quantities using graphics and animations, and provide views at different perspectives and zoom levels; Present simulation and prediction results, including flow stability and flow trends, and interpret CFD model simulation results; Output simulation reports and display prediction results, provide comparison views, show the differences between different prediction results, and provide interactive annotation functions during the result display process.

Images