Multi-source simulation method of heat island effect in underground space of cold cities

Abstract

A multi-source simulation method for the urban heat island effect in underground spaces in frigid cities. This invention relates to the interdisciplinary fields of building technology science and smart city, specifically a multi-source simulation method for the urban heat island effect in underground spaces in frigid cities. This invention integrates field measurements, CFD numerical simulation, and climate experiments to simulate the urban heat island effect in underground spaces in frigid cities and obtain parameters to guide building design. The method includes the following steps: selecting measurement point A and control point B, and collecting measured meteorological data; constructing a three-dimensional geometric model; calculating the CFD discrete domain of the geometric model; setting up a CFD model based on the measured meteorological data and the CFD discrete domain, and iteratively solving the simulation results of the CFD model using the set CFD model; constructing an artificial climate chamber for comparative experiments and obtaining comparative experimental results; correcting the CFD model based on the comparative experimental results; and simulating the urban heat island effect in underground spaces in frigid cities using the CFD model.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of building technology science and smart city technology, specifically to a multi-source simulation method for the heat island effect in underground spaces of frigid cities. Background Technology

[0002] As urban construction moves towards intensive and green development, the scale of underground space development and utilization is continuously expanding, and the resulting changes in the underground thermal environment are gradually attracting attention. In frigid regions, due to unique climatic conditions such as significant freeze-thaw cycles, large annual temperature differences, and long periods of snow cover, the heat and moisture migration process of the soil surrounding underground spaces is more complex, and its thermal environment evolution patterns have a significant impact on building energy conservation, the durability of underground structures, and urban renewal projects. Therefore, conducting research on changes in the underground thermal environment in frigid cities is of practical significance.

[0003] Currently, research on changes in the underground thermal environment mainly relies on methods such as remote sensing inversion, historical observation data analysis, and numerical simulation. These methods can reflect the temperature characteristics of the surface and shallow soil to a certain extent, but due to the limited data sources, spatial resolution, and lack of effective verification between different types of data, the simulation results still have shortcomings in terms of refinement and reliability. In addition, existing methods are mostly based on thermal environment models of general climate regions, and do not adequately consider the perturbation of the soil temperature field by freeze-thaw cycles, snow insulation effects, and the operating conditions of underground structures in frigid regions, making it difficult to accurately reflect the dynamic evolution of the underground temperature field under extreme low-temperature conditions.

[0004] In engineering applications, existing research findings are difficult to directly translate into quantitative parameters required for fields such as architectural design and underground engineering planning. For example, the impact of soil cover thickness on the underground thermal environment lacks quantifiable design guidance, making it difficult for related research to support parameter selection and performance optimization in engineering practice.

[0005] Furthermore, existing studies largely rely on field monitoring data for verification, but observation conditions are easily affected by environmental interference and it is difficult to simulate extreme working conditions, resulting in an incomplete verification system and limiting the wider application of related simulation technologies. At the same time, the inconsistent data formats and standards generated by different research methods make it difficult to integrate with commonly used urban information systems or digital platforms, hindering the formation of a shareable and reusable technology system.

[0006] In summary, existing underground thermal environment simulation technologies have significant shortcomings in terms of data fusion capabilities, adaptability to severe cold climates, engineering usability, verification systems, and digital compatibility. They cannot fully meet the technical requirements for thermal environment simulation in underground space development and urban construction in severe cold regions. There is an urgent need to develop a new technology solution that can simultaneously address both simulation and application. Summary of the Invention

[0007] To address the aforementioned problems, the purpose of this invention is to propose a multi-source simulation method for the urban heat island effect in underground spaces in frigid cities. This method integrates field measurements, CFD numerical simulations, and climate experiments to simulate the urban heat island effect in underground spaces in frigid cities and obtain parameters to guide building design.

[0008] The method includes the following steps:

[0009] S1. Select measuring point A and control point B, and collect the measured meteorological data set;

[0010] S2. Construct three-dimensional geometric models of measuring point A and control point B respectively;

[0011] S3. Calculate the CFD discrete domain of the geometric model constructed in step S2 by meshing;

[0012] S4. Based on the measured meteorological dataset and the CFD discrete domain, set up a CFD model, and iteratively solve the simulation results of the CFD model using the set up CFD model.

[0013] S5. Based on the set CFD model, construct an artificial climate chamber for comparative experiments and obtain the comparative experimental results.

[0014] S6. Correct the soil thermal property coefficients of the CFD model based on the results of comparative experiments;

[0015] S7. The corrected CFD model and the artificial climate chamber were compared again in an experiment.

[0016] When the root mean square error between the simulation results of the corrected CFD model and the experimental results of the artificial climate chamber stabilizes at... Within, and the coefficient of determination When the value is greater than 0.9, a CFD model for simulation is obtained;

[0017] Otherwise, proceed to step S6;

[0018] S8. Using the CFD model obtained in step S7, simulate the heat island effect in the underground space of a frigid city to obtain parameters for guiding architectural design.

[0019] Furthermore, the underground space at measuring point A has been developed; the underground space at control point B has not been developed.

[0020] The measured meteorological dataset includes: air temperature, surface temperature, air humidity, and soil moisture at measuring point A and control point B.

[0021] Furthermore, the three-dimensional geometric model of measuring point A includes: the main underground structure of measuring point A, the overburden layer, the surface paving, the vegetation, and the surrounding building forms;

[0022] The three-dimensional geometric model of control point B includes: the soil cover, surface paving, vegetation, and surrounding building forms of control point B.

[0023] Furthermore, in step S3, the geometric model constructed in step S2 is imported into the mesh generation tool ICEM CFD, and meshing is performed on the CFD computational domain. The mesh after meshing is denoted as the CFD discrete domain.

[0024] The CFD computational domain represents the cubic region obtained by extending the geometric model constructed in step S2.

[0025] Furthermore, in step S4, the operation of setting the CFD model is as follows: import the CFD discrete domain into the ANSYS Fluent solver, set the solver turbulence model to RNG k-ε model, set the solver radiation model to P1 model, and activate the solver energy equation.

[0026] The measured meteorological dataset is input into the ANSYS Fluent solver as a dynamic condition that changes over time.

[0027] Furthermore, the comparative experiment specifically involves simulating temperature changes under the same conditions using a pre-configured CFD model and an artificial climate chamber.

[0028] Furthermore, in step S6, the soil thermal properties include: thermal conductivity. and specific heat capacity .

[0029] Furthermore, in step S8, the parameters used to guide architectural design are sequentially processed through structuring, standardization, and encapsulation, and then integrated into the City Information Model (CIM) or Building Information Model (BIM) to guide architectural design.

[0030] The beneficial effects of the method described in this invention are as follows:

[0031] (1) To address the problems of limited technical means and weak data support in existing technologies, this invention establishes a high-confidence simulation model by integrating winter field measurements and climate chamber experimental data to dynamically correct the CFD model (CFD governing equations). Verification shows that the model's prediction error for surface and shallow soil temperatures in the study area can be controlled within 0.5℃, significantly outperforming traditional simulation methods that rely on constant parameters. This achieves a high-precision quantitative simulation of the underground heat island effect in frigid cities during winter.

[0032] (2) To address the shortcomings of existing research in insufficient consideration of the characteristics of severe cold, this invention, by coupling actual freeze-thaw cycles with extreme conditions simulation in artificial climate chambers, systematically quantifies for the first time the warming and pressure-stabilizing effects of underground spaces as heat sources on the surface environment during winter. The soil cover thickness selected using the method described in this invention (e.g., 1 meter) can reduce the surface temperature around underground spaces by approximately 40% compared to areas without underground spaces, effectively enhancing the buffering and adaptability of urban spaces to extreme climates.

[0033] (3) To address the disconnect between architectural science and engineering design, this invention ultimately produces a quantified library of design parameters and construction guidelines. For example, it clearly provides the optimal range of soil cover thickness (0.8-1.2 meters) for different climate zones, balancing thermal buffering effects and engineering costs. This allows designers to directly utilize the invention in software such as Revit and ArchiCAD, transforming architectural science into actionable and acceptable design practices, achieving a seamless transition from macro-level strategy to micro-level design. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of measuring point A and control point B as described in this invention;

[0035] Figure 2 This is the three-dimensional geometric model of measuring point A described in this invention;

[0036] Figure 3 This is the three-dimensional geometric model of measuring point B as described in this invention. Detailed Implementation

[0037] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] Example 1

[0039] This embodiment provides a multi-source simulation method for the heat island effect in underground spaces of cold cities.

[0040] Traditional simulation methods struggle to accurately quantify the "peak-shaving and valley-filling" effect of underground space on surface temperature fluctuations through the overburden layer, i.e., the intensity and extent of the thermal buffering effect. Due to the lack of in-depth analysis of energy exchange within the underground structure-soil-atmosphere continuum, reliable simulation methods cannot be established to guide design.

[0041] This embodiment employs a three-pronged approach: "long-term on-site monitoring to obtain empirical data + CFD simulation to analyze energy pathways + climate chamber experiments to verify key parameters," to construct a physical model that accurately reflects the thermal buffering effect of underground spaces. Quantitative evidence of the thermal buffering effect is directly obtained through comparative monitoring at on-site measurement points. The underlying energy redistribution mechanism is revealed through CFD model simulation. Finally, the process is reproduced in a climate chamber to verify the model parameters, forming a complete technical chain from phenomenon observation to mechanism analysis and simulation prediction.

[0042] The multi-source simulation method for the heat island effect in underground spaces of frigid cities includes the following steps:

[0043] S1. Select measuring point A and control point B, and collect the measured meteorological data set;

[0044] like Figure 1 As shown, in this embodiment, in a typical area of ​​a frigid city, a precisely selected developed underground space was chosen. For example, any building above an underground parking garage was designated as the experimental monitoring point A, and a dormitory building adjacent to an undeveloped green space (site) was designated as the control point B. A monitoring network composed of high-precision sensors was deployed: small microclimate stations were used to simultaneously collect four elements: air temperature and humidity, wind speed, wind direction, and solar radiation; soil temperature and humidity sensors were buried vertically downwards at a depth of -0.5m. All sensors were networked through an Internet of Things (IoT) module and continuously and automatically monitored for four days at 30-minute intervals to obtain measured meteorological data, which was used as the boundary conditions for the CFD control equations.

[0045] This embodiment extends the research perspective from above-ground space to underground for the first time, providing direct evidence of the existence of the "underground heat island." Measured temperature data (unit: °C) were collected during typical winter periods, as shown in Table 1. In this embodiment, a JT2020 building thermal environment tester was used to monitor the air temperature at pedestrian height (1.5m), the surface temperature at ground level (0m), and the shallow soil temperature at shallow underground depth (-0.5m) at points A and B from December 6th to 9th, 2025.

[0046] Table 1

[0047]

[0048] Based on the field monitoring data analysis in Table 1, the temperature variation at experimental point A (above the underground parking garage) was an average daily fluctuation of -4.2℃, significantly higher than the -7.0℃ at control point B, with a temperature difference of 2.8℃. This directly proves that the surface received heat replenishment from the underground. Meanwhile, the variance of the surface temperature at point A was... The temperature at point A was approximately 39.5% lower than the 5.06℃ at point B. At a depth of -0.5 meters, the average temperature at point A remained at 1.5℃ (unfrozen), while at point B it had dropped to -0.5℃. This 2.0℃ temperature difference indicates that the underground space effectively prevented the freezing of the soil above it. This demonstrates that the underground space significantly raised the surface temperature baseline in winter and greatly suppressed temperature fluctuations, exhibiting a "thermal buffering" effect. This confirms that the underground structure has a significant thermal buffering effect on the surface thermal environment.

[0049] S2. Construct three-dimensional geometric models of measuring point A and control point B respectively;

[0050] Based on on-site surveying and architectural design drawings, a precise 3D geometric model containing measurement point A and control point B was constructed using the Rhinoceros platform and its parametric plugin Grasshopper. The 3D geometric models of measurement point A and control point B are shown below. Figure 2 and 3 As shown, the 3D geometric model of measuring point A includes: the main underground structure, overburden layer, surface paving, vegetation, and surrounding building morphology of measuring point A; the 3D geometric model of control point B includes: the overburden layer, surface paving, vegetation, and surrounding building morphology of control point B. Utilizing Grasshopper's logic definition capabilities, key parameters in the 3D geometric model (such as overburden thickness, ventilation opening size, and building density) are set as adjustable variables, laying the foundation for subsequent parametric simulation and scheme comparison.

[0051] S3. Calculate the CFD discrete domain of the geometric model constructed in step S2 by meshing;

[0052] Based on the 3D geometric model constructed in step S2, a sufficiently large cubic region is established as the CFD computation domain. This computation domain should extend horizontally to at least five times the height of the target building and vertically from at least 20 meters below the underground space floor to a sufficient height above the ground surface to ensure that boundary effects do not affect the flow and temperature field calculations in the core area. Subsequently, the core areas requiring detailed simulation are clearly defined within the computation domain, primarily including: the underground space structure itself, the overburden layer on the underground space structure, surface vegetation, adjacent building surfaces, and the near-surface air region. These areas are key channels for heat transfer and exchange.

[0053] The geometric model is imported into professional mesh generation tools such as ANSYS Fluent Meshing or ICEM CFD. A hybrid meshing strategy combining unstructured meshes and boundary layer meshes is adopted to mesh the CFD computational domain. Mesh refinement is performed on key areas around the underground space to ensure that the y+ value near the wall meets the requirements of the selected turbulence model, thus balancing the adaptability of complex geometry with computational accuracy.

[0054] After mesh generation, a rigorous mesh quality check is required to avoid divergence or accuracy loss during calculation. Finally, mesh independence verification is performed, which involves generating three sets of meshes with different densities (coarse, medium, and fine) sequentially, simulating the same typical working condition, and comparing the predicted temperature values ​​at the core observation points (those skilled in the art can set the criteria for passing mesh independence verification according to actual needs). The mesh that passes the independence verification is the CFD discrete domain.

[0055] The core purpose of mesh independence verification is to ensure that simulation results are independent of the number of meshes, but only controlled by the physical model and boundary conditions. In other words, if further refining the mesh after it is sufficiently fine does not significantly change the results, then mesh independence has been achieved.

[0056] Through the above systematic discretization method, the method described in this invention can accurately capture the details of heat exchange between underground space and the surface, while maintaining reasonable computational efficiency, thus laying a reliable numerical foundation for subsequent CFD simulation setup and high-precision solution.

[0057] S4. Based on the measured meteorological dataset and the CFD discrete domain, set up a CFD model, and iteratively solve the simulation results of the CFD model using the set up CFD model.

[0058] Import the pre-defined mesh model (CFD discrete domain) into the ANSYS Fluent solver. Select the Realizable k-ε turbulence model suitable for outdoor microclimate simulation, activate the energy equation and radiation model (those skilled in the art can select DO or P1 according to actual needs), and input the measured data collected in step S1 as dynamic conditions that change over time into the ANSYS Fluent solver. The specific settings of the CFD model (governing equations) in this embodiment are shown in Table 2:

[0059] Table 2:

[0060]

[0061] Turbulence models describe the turbulent characteristics of airflow, energy equations control the heat transfer process, and radiation models calculate the propagation and absorption of solar radiation and longwave radiation.

[0062] The CFD model (governing equations) is not directly generated from the grid (CFD discrete domain). Instead, it is obtained by integrating within a discrete control volume using the finite volume method, based on fundamental physical laws such as mass conservation, momentum conservation, and energy conservation. Subsequently, the convection, diffusion, and source terms are discretized using a suitable numerical scheme, ultimately forming a solvable set of algebraic equations. By setting different physical models, momentum equations, energy equations, and turbulent transport equations can be enabled, thereby constructing a governing equation system for simulating underground thermal environments.

[0063] In this embodiment, the "three-dimensional unsteady thermal-fluid coupling process of the underground space-soil-surface air system" was simulated by iterating 300 times using the established CFD model. The CFD simulation results were obtained within the CFD computational domain and are shown in Table 3 (unit: °C).

[0064] Table 3

[0065]

[0066] The CFD model simulation results include: air temperature at pedestrian height (1.5m), surface temperature at ground height (0m), and shallow soil temperature at shallow subsurface (-0.5m) at points A and B, and attempts to predict the soil temperature field distribution characteristics at deeper layers (-2m).

[0067] The CFD simulation results are similar to the measured data (Table 1). The simulation shows that the average surface temperature at point A (-4.0℃) is significantly higher than that at point B (-7.2℃), with a temperature difference of 3.2℃; at the same time, the variance of the surface temperature at point A ( The temperature at point A is approximately 40.8% lower than at point B (5.25℃). This numerically replicates and verifies the dual thermal buffering effect of underground spaces in winter, which "raises the temperature baseline and suppresses temperature fluctuations." Simulations show that the average temperature at -0.5 meters below ground at point A is positive (…). ), and the variance is extremely low ( This indicates that the soil layer is in a stable, unfrozen thermal state; while the average temperature at the same depth at point B is already negative. ), with larger variance ( This precisely reveals the extent and intensity of the "insulating" effect of the underground space on the soil directly above it.

[0068] The core advantage of CFD simulation lies in revealing deep information that cannot be obtained through field measurements. Data shows that at a depth of -2.0 meters, a stable heat source with an average temperature as high as 5.2°C forms in region A, which is 2.2°C higher than that at the same depth (3.0°C) at point B, and both have extremely small variances. This directly confirms that the "underground heat island" phenomenon not only exists in winter, but its core heat source region, characterized by high temperature and high stability, is located several meters underground, providing a precise energy source explanation for the thermal effects observed on the surface. Comparing Tables 1 and 3, it can be seen that the CFD simulation values ​​are highly consistent with the measured values ​​in terms of trend and magnitude, but there are slight deviations in specific values ​​(e.g., the simulated surface temperature at point A is slightly higher than the measured value). This deviation mainly stems from the model's physical simplification of water migration and latent heat exchange during the complex freeze-thaw phase change of soil. However, the model successfully predicted all key phenomena, with a correlation coefficient exceeding 0.95, fully demonstrating the high reliability and practical value of the CFD model established by this method for analyzing the mechanism of the winter underground heat island.

[0069] After the simulation results are obtained, post-processing tools such as CFD-Post are used to extract and visualize the three-dimensional temperature field distribution characteristics within the CFD computational domain.

[0070] Based on the simulation results, the temperature difference between the experimental point and the control point is quantitatively calculated, i.e., the heat island intensity (not a direct physical field, but rather calculated by extracting the simulated (or measured) surface temperature T values ​​of the experimental point (point A) and the control point (point B). A and T B Then, a scalar or spatiotemporal distribution field is obtained by calculating the difference between the two at the same moment. The formula can be simplified to: Heat island intensity (t) = T A - T B It directly quantifies the temperature rise effect caused by underground space.

[0071] Analysis from the perspectives of energy balance and heat island intensity reveals that the average temperature in region A at the -2 meter depth experimental point reaches [value missing]. A is a stable heat source. Because the thermal conductivity of the overlying soil increases dramatically after freezing in winter, heat is efficiently conducted upwards. This results in a significantly higher upward soil heat flux at point A compared to point B, effectively compensating for the heat lost from the surface to the cold air, thus creating a "warm island" effect that raises the temperature baseline (insulation) and suppresses temperature fluctuations (pressure stabilization). This mechanism is fundamentally different from the "cold island" effect where heat is conducted downwards in summer.

[0072] S5. Based on the set CFD model, construct an artificial climate chamber for comparative experiments and obtain the comparative experimental results.

[0073] Seasonal freeze-thaw cycles in frigid cities significantly alter soil thermal properties and the heat exchange process between underground structures and surrounding soil. Traditional steady-state or simple dynamic models cannot accurately capture this nonlinear and transient heat and moisture migration process, resulting in predictions of winter heat loss and spring snowmelt heat imbalance that deviate significantly from reality.

[0074] To compensate for the limitations of pure numerical simulation and to verify key parameters, this embodiment conducts an empirical study within an artificial climate chamber. The experiment aims to reproduce typical freeze-thaw cycle conditions in frigid regions, with the following specific setup: Standard-sized (1m × 1m × 0.5m) soil-insulation composite samples were prepared within the artificial climate chamber to simulate different soil cover and structural conditions. A high-precision temperature control system drove the chamber environment through a typical freeze-thaw cycle from +5℃ to -18℃ and back to +5℃, with the entire simulation lasting 48 hours to fully capture the phase transition process. Temperature sensors were placed at the interfaces of different soil layers and at key locations near the surface within the sample. A data acquisition system continuously recorded the temperature response curves at each measuring point throughout the entire freeze-thaw cycle.

[0075] S6. Correct the soil thermal property coefficients of the CFD model based on the results of comparative experiments;

[0076] After collecting experimental data from the climate chamber, the experimental data were compared with the simulation results under the corresponding conditions in the CFD governing equations. The core object of the correction was the key parameters describing soil thermal properties in the CFD model, especially the thermal conductivity. ) and specific heat capacity ( These components exhibit strong nonlinear changes during freeze-thaw cycles. Traditional constant or simple linear assumptions introduce significant errors.

[0077] When the matching degree between the simulation results and the climate chamber experimental data reaches a preset accuracy threshold, it is confirmed as a high-confidence model. Specifically, after parameter inversion optimization, the root mean square error (RMSE) between the simulated temperature curves of the CFD model for all verification conditions and the corresponding experimental data remains consistently within 0.5°C, and the coefficient of determination (... When the value is greater than 0.9, the correction is considered complete.

[0078] S7. The corrected CFD model and the artificial climate chamber were compared again in an experiment.

[0079] When the root mean square error between the simulation results of the corrected CFD model and the experimental results of the artificial climate chamber stabilizes within 0.5°C, and the coefficient of determination... When the value is greater than 0.9, a CFD model for simulation is obtained;

[0080] Otherwise, proceed to step S6;

[0081] The high-confidence CFD model (the CFD model used for simulation), corrected for climate chamber data, was applied to simulate winter conditions across the entire study area (including experimental point A and control point B). The model was calculated using field-measured meteorological data from winter as the driving boundary conditions. The results are shown in Table 4 (unit: °C).

[0082] Table 4

[0083]

[0084] It can be observed that the results of this calculation are highly consistent with the measured data in winter (Table 1). For example, the simulated average surface temperature at points A and B ( ) and measured

[0085] ( The trends are completely identical, and the temperature difference (simulated 3.2°C, measured 2.8°C) is very close, verifying the predictive reliability of the corrected model at the actual site scale. Furthermore, the calculation results clearly reveal the temperature field from deep underground to the surface. At point A, the temperature reaches as high as 8.1°C at -5.0 meters, dropping to -3.9°C at the surface, forming a strong bottom-up temperature gradient, direct evidence of continuous upward heat transport. At the same depth, the temperature at point A is significantly higher than at point B. Particularly at a depth of -2.0 meters, point A (5.3°C) is 2.2°C higher than point B (3.1°C), quantitatively defining the intensity and location of the core area of ​​the "underground heat island" in winter. Across all depths, the variance of the temperature at point A is much smaller than that at point B. For example, the variance of the surface temperature at point A is... Point B is The thermal fluctuation at point A was approximately 41.3% lower than that at point B. This statistically confirms the remarkable "thermal stability" effect brought about by the underground space. These results demonstrate that the corrected CFD governing equations successfully predicted the complete three-dimensional thermal environment driven by the underground space as a heat source at the site scale during winter. The output data not only reproduced the observed surface phenomena but, more importantly, for the first time, numerically and completely characterized the three-dimensional structure of the underground heat island, from the heat source core (-5m), the main influence zone (-2m), the buffer layer (-0.5m), to the surface manifestation (0m), providing an irreplaceable data foundation for the accurate assessment and utilization of this effect.

[0086] S8. Using the CFD model obtained in step S7, simulate the heat island effect in the underground space of a frigid city to obtain parameters for guiding architectural design.

[0087] This embodiment compiles the optimal parameter set and quantified relationships obtained in step S7 (parameters used to guide architectural design) into a design guideline for optimizing the thermal environment of underground spaces in frigid cities, including recommended structural details, material selection tables, and key design parameter tables. Finally, these structured and standardized design rules and their underlying analytical models are encapsulated into plugins or data interfaces and connected to mainstream City Information Modeling (CIM) or Building Information Modeling (BIM) platforms. This allows designers to directly utilize the results of this invention during the design phase, achieving data-driven scientific design and ensuring the direct transformation and application of research findings.

[0088] In this embodiment, the parameters used to guide building design are obtained as follows: a high-confidence CFD control model, validated through climate chamber experiments, is used for large-scale parametric simulation. Hundreds of simulation conditions are automatically run to quantitatively analyze the impact of each parameter on the winter "heat island" (warm island) effect, and the optimal solution (parameters used to guide building design) that balances thermal performance and engineering economy is selected. The temperature field distribution under different soil cover thicknesses (0.5m to 5.0m) is systematically simulated (see Table 4). By analyzing the simulation results of all conditions, a quantitative relationship curve between soil cover thickness and thermal performance is established. The simulation results of key indicators of winter thermal performance under different soil cover thicknesses are shown in Table 5.

[0089] Table 5

[0090]

[0091] It can be observed that as the cover thickness increases, both the average surface temperature and the temperature at -0.5 meters below ground level continuously rise. However, the temperature increase decreases sharply after the cover thickness exceeds 1.2 meters (indicating diminishing marginal returns). Furthermore, the surface temperature variance continuously decreases, indicating enhanced thermal stability, but this improvement also becomes less significant after approximately 1.2 meters. Therefore, considering the three main objectives of balancing insulation, frost protection, and stability (low variance) with engineering costs, a cover thickness between 0.8 and 1.2 meters exhibits the best cost-effectiveness. The average surface temperature in this range is lower than that in areas without underground space (approximately...). It rose to .underground The soil temperature in rice is stable at to To ensure no risk of freezing. Surface temperature variance decreased to The following results demonstrate significant thermal stability. When the thickness is less than 0.8 meters, the insulation and frost protection effects are insufficient; when it is greater than 1.2 meters, the improvement in thermal performance is slight, but the earthwork and structural costs increase significantly. Therefore, a soil cover thickness of 1.0 meter is a core recommended value under the simulation conditions described in this invention. At this thickness, the average surface temperature is approximately... The variance is The temperature of shallow soil reaches It achieves an excellent balance between effectiveness and cost.

Claims

1. A multi-source simulation method for the urban heat island effect in underground spaces in frigid cities, characterized in that, The method includes the following steps: S1. Select measuring point A and control point B, and collect the measured meteorological data set; S2. Construct three-dimensional geometric models of measuring point A and control point B respectively; S3. Calculate the CFD discrete domain of the geometric model constructed in step S2 by meshing; S4. Based on the measured meteorological dataset and the CFD discrete domain, set up a CFD model, and iteratively solve the simulation results of the CFD model using the set up CFD model. S5. Based on the set CFD model, construct an artificial climate chamber for comparative experiments and obtain the comparative experimental results. S6. Correct the soil thermal property coefficients of the CFD model based on the results of comparative experiments; S7. The corrected CFD model and the artificial climate chamber were compared again in an experiment. When the root mean square error between the simulation results of the corrected CFD model and the experimental results of the artificial climate chamber stabilizes at... Within, and the coefficient of determination When the value is greater than 0.9, a CFD model for simulation is obtained; Otherwise, proceed to step S6; S8. Using the CFD model obtained in step S7, simulate the heat island effect in the underground space of a frigid city to obtain parameters for guiding architectural design.

2. The multi-source simulation method for the urban heat island effect in underground spaces in frigid cities according to claim 1, characterized in that, The underground space at measuring point A has been developed; The underground space at control point B is undeveloped; The measured meteorological dataset includes: air temperature, surface temperature, air humidity, and soil moisture at measuring point A and control point B.

3. The multi-source simulation method for the urban heat island effect in underground spaces in frigid cities according to claim 2, characterized in that, The three-dimensional geometric model of measuring point A includes: the main underground structure of measuring point A, the overburden layer, the surface paving, the vegetation and the surrounding building forms; The three-dimensional geometric model of control point B includes: the soil cover, surface paving, vegetation, and surrounding building forms of control point B.

4. The multi-source simulation method for the heat island effect in underground space of frigid cities according to claim 3, characterized in that, In step S3, the geometric model constructed in step S2 is imported into the mesh generation tool ICEM CFD, and the mesh is generated on the CFD computational domain. The resulting mesh is denoted as the CFD discrete domain. The CFD computational domain represents the cubic region obtained by extending the geometric model constructed in step S2.

5. The multi-source simulation method for the heat island effect in underground space of frigid cities according to claim 4, characterized in that, In step S4, the operation of setting the CFD model is as follows: import the CFD discrete domain into the ANSYS Fluent solver, set the solver turbulence model to RNG k-ε model, set the solver radiation model to P1 model, and activate the solver energy equation. The measured meteorological dataset is input into the ANSYS Fluent solver as a dynamic condition that changes over time.

6. The multi-source simulation method for the heat island effect in underground space of severely cold cities according to claim 5, characterized in that, The comparative experiment specifically involved simulating temperature changes under the same conditions using a pre-configured CFD model and an artificial climate chamber.

7. The multi-source simulation method for the urban heat island effect in underground spaces in frigid cities according to claim 6, characterized in that, In step S6, the soil thermal properties include: thermal conductivity. and specific heat capacity .

8. The multi-source simulation method for the heat island effect in underground space of frigid cities according to claim 7, characterized in that, In step S8, the parameters used to guide architectural design are sequentially processed through structuring, standardization, and encapsulation, and then integrated into the City Information Model (CIM) or Building Information Model (BIM) to guide architectural design.

Images