Method, device and equipment for predicting ozone permeation flux of high-rise building gap structure and storage medium

Abstract

The application provides a high-rise building gap structure ozone permeation flux prediction method, device, equipment and storage medium, relating to the environmental pollution control technical field.The method comprises the following steps: obtaining multi-source parameters from BIM, material library and meteorological data; using CFD simulation to obtain the dynamic wind pressure field of the building surface, and extracting the wind pressure difference, wind speed, temperature and concentration data and other key driving data at the permeation node; deploying a lightweight PINN proxy model at the permeation node, receiving the driving data and gap geometric parameters, training through embedding the Darcy law physical constraint, and directly predicting the microscale ozone permeation flux; through online dynamic correction of the mass transfer coefficient and comprehensive equivalent pressure difference, coupling the wind pressure enhancement effect, material aging mechanism and thermal pressure effect, realizing real-time coupling simulation of wind, heat and material multi-physical fields. The application realizes high-precision and high-efficiency ozone permeation dynamic prediction.

Description

Technical Field

[0001] This application relates to the field of environmental pollution control technology, and in particular to a method, apparatus, equipment and storage medium for predicting ozone permeation flux in the gap structure of high-rise buildings. Background Technology

[0002] Ozone, as a pollutant with low exposure concentrations but high health risks, has become a key indicator for air quality assessment. Outdoor ozone mainly infiltrates indoors through three major channels in the building envelope, becoming the primary source of indoor ozone. The current "Healthy Building Evaluation Standard" T / ASC 02-2021 simplifies the calculation of ozone infiltration in high-rise buildings. It calculates ozone infiltration based on the mass transfer coefficient range of different channels (such as window seams and curtain wall joints), combined with the product of the indoor and outdoor concentration difference and the effective infiltration area, and adds correction factors such as wind pressure and chimney effect coefficients.

[0003] Currently, publicly available technical solutions for achieving dynamic prediction mainly fall into two categories and have inherent drawbacks: Purely data-driven methods (such as LSTM and random forest) rely on training with large amounts of historical monitoring data. While they offer fast inference speeds, these models lack physical interpretability and suffer from large prediction errors (typically ±15-20%) when data is insufficient or when encountering unseen operating conditions. Furthermore, they fail to reflect long-term evolution processes such as material aging. High-fidelity numerical simulation methods (such as CFD simulations based on RANS models) possess clear physical mechanisms but face significant bottlenecks. First, simulating flow in micrometer-level gaps in buildings requires extremely fine meshes due to the vast differences between macro and micro scales, leading to high computational costs and low efficiency. Second, parameters such as the mass transfer coefficient in these models are typically fixed, making it impossible to respond in real-time to material performance degradation or dynamic changes in environmental conditions; long-term prediction errors can reach ±25-35%.

[0004] Furthermore, existing technical solutions are mostly limited to single prediction models or algorithms, and a complete prediction system or dedicated tool integrating efficient macroscopic simulation, physically constrained microscopic prediction, multi-field dynamic coupling, and long-term performance evolution functions has not yet been formed. Currently, there is no publicly available solution for dynamic prediction of ozone infiltration flux in high-rise building gaps that is based on the deep integration of Physical Information Neural Network (PINN) and Computational Fluid Dynamics (CFD) and encompasses methods, devices, and equipment. Summary of the Invention

[0005] This application provides a method, apparatus, device, and storage medium for predicting ozone infiltration flux in high-rise building gap structures. It integrates computational fluid dynamics (CFD) with physical information neural networks (PINN), achieving accurate simulation of the ozone infiltration mass transfer process through real-time coupling of multi-channel mass transfer coefficients with the CFD-PINN wind pressure field. The technical solution establishes a dynamic interface between CFD macroscopic simulation and PINN microscopic prediction, directly predicting the ozone flux inside building gaps, effectively solving the computational and physical consistency problems of traditional methods in microscale flow simulation.

[0006] In a first aspect, this application provides a method for predicting the ozone permeation flux of gap structures in high-rise buildings, including: Obtain the geometric parameters of the gaps in the building, material properties and aging kinetic parameters, as well as the environmental field driving parameters; Computational fluid dynamics (CFD) was used to simulate the external wind field of a building, obtain the dynamic wind pressure distribution on the surface, and extract macroscopic driving data vectors at multiple preset infiltration nodes. Where Δ P i For wind pressure difference, U i For local wind speeds, T out Outdoor temperature C out This refers to the outdoor ozone concentration. The macroscopic driving data at each of the aforementioned permeation nodes, along with their corresponding gap geometry and material parameters, are input into a pre-trained PINN surrogate model, which incorporates Darcy's law as a physical constraint. The microscale ozone permeation flux Φ at each node is then output. PINN ; Based on the real-time local wind speed in the macro driving data vector U i Dynamically correct the mass transfer coefficient; update the baseline mass transfer coefficient online based on the cumulative simulation time to reflect material aging, and calculate the comprehensive equivalent pressure difference Δ by coupling the hot-pressing effect. Peff ; The updated mass transfer coefficient is compared with the comprehensive equivalent pressure difference Δ. Peff The data is fed back to the PINN agent model in real time, driving it to make predictions for the next time step.

[0007] In one possible design, the gap geometry parameters include the gap length. L Equivalent hydraulic diameter D h Surface roughness height k s and tortuosity factor τ Wherein, the equivalent hydraulic diameter Dh The calculation formula is: In the formula, A For cross-sectional area, P For wet perimeter; The material properties include the initial molecular diffusion coefficient, the initial material surface reaction rate constant with ozone, and the initial distribution coefficient of ozone in the material; the aging kinetic parameters include the pre-exponential factor, activation energy, and empirical coefficient of performance degradation.

[0008] In one possible design, the total loss function during training of the PINN proxy model is... L total Data loss items L data and physical loss items L physics Composition, that is L total = L data +λ L physics Where λ is a hyperparameter; the physical loss term L physics This is achieved by constructing the Darcy's law residual, which is expressed as: In the formula, K eff To achieve an effective penetration rate of implicit learning in networks, A d The characteristic area of ​​permeation used in Darcy's law calculations. L d This corresponds to the permeation feature length. μ This represents the fluid viscosity.

[0009] In one possible design, the convection enhancement component in the dynamically corrected mass transfer coefficient K p The following formula is used: K p =K p0 +aU b In the formula, K p0 As the reference mass transfer coefficient, a and b The parameters are fitted based on experimental data. U This refers to local wind speed.

[0010] In one possible design, the material aging kinetic model adopts the Arrhenius form for online updating of the baseline mass transfer coefficient, and its formula is: In the formula, K p0 ( t Let be the reference mass transfer coefficient at time t. K p0 (0) is the initial reference mass transfer coefficient, R is the ideal gas constant, and T is the temperature. t For time, α、β is the empirical coefficient for performance degradation, and exp is an exponential function with the natural constant as its base.

[0011] In one possible design, the equivalent pressure difference Δ generated by the thermo-pressure effect P stack The calculation formula is as follows: In the formula, ξ As a correction factor, ρ out and ρ in These are the outdoor and indoor air densities calculated from indoor and outdoor temperatures, respectively. g Let H be the acceleration due to gravity, and H be the effective height.

[0012] In one possible design, the combined equivalent pressure difference is obtained by vector superposition of wind pressure difference and thermal pressure equivalent pressure difference.

[0013] Secondly, this application provides a device for predicting the ozone permeation flux of a high-rise building's gap structure, the device comprising: The data acquisition module is configured to acquire the building's gap geometry parameters, material properties and aging kinetic parameters, as well as environmental field driving parameters; and initialize the CFD solver and PINN surrogate model respectively. The CFD calculation module is configured to construct a macroscopic computational domain containing the target building and perform mesh generation. It uses computational fluid dynamics (CFD) simulation to obtain the dynamic wind pressure distribution on the building surface and extracts macroscopic driving data vectors at preset infiltration nodes. Where Δ P i For wind pressure difference, U i For local wind speeds, T out Outdoor temperature C out This refers to the outdoor ozone concentration. The PINN proxy module is configured to deploy PINN proxy models at each of the penetration nodes, wherein the PINN proxy models receive the corresponding macro-driving data vectors. X CFD In addition to the gap geometry and material parameters; after training, the PINN surrogate model outputs high-precision ozone permeation flux Φ at each node. PINN ; The correction module is configured to adjust based on the real-time local wind speed in the macroscopic driving data vector. U i The convection enhancement component in the mass transfer coefficient is dynamically corrected; the baseline mass transfer coefficient is updated online by calling the material aging kinetic model based on the simulation cumulative time; the equivalent pressure difference generated by the thermo-pressure effect is coupled and vector-superimposed with the wind pressure difference to obtain the comprehensive equivalent pressure difference Δ. Peff The updated mass transfer coefficient is compared with the combined equivalent pressure difference Δ. Peff The data is fed back to the PINN agent model in real time, driving it to make predictions for the next time step.

[0014] Thirdly, embodiments of this application provide an electronic device, including: at least one processor and a memory; the memory stores computer execution instructions; the at least one processor executes the computer execution instructions stored in the memory, causing the at least one processor to perform the method for predicting ozone permeation flux in high-rise building gap structures as described in the first aspect and various possible designs of the first aspect.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions. When a processor executes the computer-executable instructions, it implements the method for predicting ozone permeation flux in high-rise building gap structures as described in the first aspect and various possible designs of the first aspect.

[0016] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the method for predicting ozone permeation flux in high-rise building gap structures as described in the first aspect and various possible designs of the first aspect.

[0017] The method, apparatus, equipment, and storage medium for predicting ozone permeation flux in high-rise building gap structures provided in this application have at least the following beneficial effects: This application aims to solve the problem of accurately calculating the dynamic infiltration of ozone through building envelopes. It employs a technical approach that couples multiple physical fields, including wind pressure, thermal pressure, and material aging, and uses a multi-channel mass transfer coefficient coupled with a CFD-PINN (Physical Information Neural Network) wind pressure field in real time. This approach also addresses the problem of increased computational complexity and difficulty in iterative convergence caused by the significant difference in scale between the macroscopic mesh (≥1m) and the microscopic mesh (≤0.1mm) of the gaps when simulating building gaps (characteristic size 0.1-10mm) using traditional CFD methods. This innovatively sets up a PINN surrogate model at the gap boundary of the CFD mesh. This model receives driving parameters such as macroscopic wind pressure ΔP calculated by CFD, and uses Darcy's law as a physical constraint to ensure the physical rationality of the microscopic infiltration flux prediction, directly outputting the microscopic flux and thus accurately predicting the ozone flux within the gap. By constructing a hybrid architecture that combines PINN for solving microflow at the infiltration interface and CFD for simulating macroscopic wind pressure fields, this method achieves a 15-20 times faster computation speed than pure CFD-based precision simulations, while maintaining accuracy and eliminating the need for supercomputing resources. The prediction error remains stable within ±7%. This application is applicable to the airtightness design of new buildings and the prediction and assessment of ozone infiltration in the renovation of existing buildings. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] Figure 1 A flowchart illustrating a method for predicting ozone permeation flux through a gap structure in a high-rise building, provided as an embodiment of this application; Figure 2 A flowchart of multiphysics parameter input and system initialization provided for embodiments of this application; Figure 3 A CFD calculation flowchart provided for an embodiment of this application; Figure 4 A flowchart for microscale permeation flux prediction provided in the embodiments of this application; Figure 5 A flowchart illustrating the multi-effect dynamic coupling and online correction of material aging coefficient provided in this application embodiment; Figure 6 Another flowchart of a method for predicting ozone permeation flux in a high-rise building with gap structure provided in an embodiment of this application; Figure 7 A structural diagram of the ozone permeation flux prediction device for high-rise building gap structures provided in this application embodiment.

[0020] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concepts of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0021] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0022] The collection, storage, use, processing, transmission, provision, and disclosure of financial data or user data involved in the technical solution of this application all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0023] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0024] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0025] This application provides a method for predicting ozone permeation flux through gaps in high-rise buildings, particularly suitable for building energy-saving design and indoor health environment control. A dedicated training dataset containing wind tunnel test data for 20 types of curtain wall materials has been constructed. This method can be used for optimizing the airtightness design of new buildings and for evaluating the renovation of existing buildings, providing an innovative technical means for the prevention and control of building environmental health risks. Specifically, as... Figure 1 As shown, the method for predicting ozone infiltration flux through the gap structure of this high-rise building includes the following steps S10-S40.

[0026] S10: Obtain the geometric parameters of the building's gaps, material properties, aging kinetic parameters, and environmental field driving parameters; and initialize the CFD solver and PINN surrogate model respectively.

[0027] The purpose of step S10 is to implement multiphysics parameter input and system initialization. In some embodiments, such as Figure 2 As shown, step S10 can be implemented through the following steps S101-S104.

[0028] S101: Extract building gap information using BIM / CAD.

[0029] In this embodiment, the gap information of the building is divided into three categories: curtain wall joints, window gaps, and structural joints. Among them, the key geometric parameters include gap length. L (m), equivalent hydraulic diameter , A For cross-sectional area, P For wetted perimeter, surface roughness height k s (m), tortuosity factor τ(-).

[0030] S102: Obtain material and aging kinetic parameters.

[0031] In this embodiment, materials and aging kinetic parameters are divided into two categories: initial material properties and aging kinetic parameters. Initial material properties include the molecular diffusion coefficient. D m ,0 (m 2 / s), the rate constant of the reaction between the material surface and ozone k r ,0 (m / s), the partition coefficient of ozone in the material K aw (-); Aging kinetic parameters include pre-exponential factors. A aging (s -1 ),activation energy E a,aging (J / mol), empirical coefficient of performance degradation α , β (-) is used to quantify the evolution of material properties over time.

[0032] S103: Obtain environmental field driving parameters.

[0033] In this embodiment, the environmental field driving parameters are divided into two categories: wind field parameters and concentration and thermal environment parameters. The wind field parameters are input using real-time or typical meteorological data of the target building's location, including wind speed time series. v ( t (m / s), wind direction time series θ ( t (rad), turbulence parameters; concentration and thermal environment parameters input from the time series of background ozone concentration outside the building. c out ( t(kg / m 3 Indoor and outdoor air temperatures T in , T out ( t (K) and solar irradiance G ( t (W / m) 2 ).

[0034] S104: Model initialization.

[0035] In this embodiment, model initialization is divided into two parts: CFD solver initialization and PINN surrogate model initialization. CFD solver initialization employs a transient model, initializes the time step Δt, and configures the turbulence model with the pressure-velocity coupled SIMPLE algorithm to ensure stable flow field calculations. For the PINN surrogate model, a 6-8 layer fully connected network is constructed, using Tanh or Swish activation functions, initializing weights, and embedding Darcy's law through an optimizer and physical constraint loss function.

[0036] S20: Construct a macroscopic computational domain containing the target building and mesh it. Use computational fluid dynamics (CFD) simulation to obtain the dynamic wind pressure distribution on the building surface and extract the macroscopic driving data vectors at the preset infiltration nodes. Where Δ P i For wind pressure difference, U i For local wind speeds, T out Outdoor temperature C out This refers to the outdoor ozone concentration.

[0037] The purpose of step S20 is to simulate and output macroscopic wind pressure fields through CFD calculations. In some embodiments, such as Figure 3 As shown, step S20 can be achieved through the following steps S201-S203.

[0038] S201: Computational domain construction and mesh generation.

[0039] Centered on the target building, a macroscopic computational domain conforming to fluid computation standards was established (inlet ≥ 5H, outlet ≥ 15H, top and side ≥ 5H, where H is the building height). A CFD-RANS (Realizable k-ε) model was adopted, using a coarse mesh of 0.5-2m for the building's external field and applying boundary layer refinement to the building surface to ensure that the dimensionless distance y+ value on the wall is within a reasonable range of 30-300, accurately capturing and preventing flow.

[0040] S202: CFD solution and wind pressure field calculation.

[0041] Boundary conditions such as velocity inlet, pressure outlet, and wall function are set, and transient CFD calculations are performed using the SIMPLE algorithm to accurately obtain the dynamic wind pressure distribution P(x,y,z,t) on the building surface.

[0042] S203: Extract driver data.

[0043] After the CFD calculation is completed, for all predefined permeability nodes on the building's exterior surface, i.e., the outdoor locations of the gaps, the output includes the wind pressure difference Δ. P i Local wind speed U i ,temperature T out and ozone concentration C out Key physical quantities are used to form a data vector. As input to the PINN model.

[0044] S30: Deploy PINN proxy models at each penetration node. The PINN proxy model receives the corresponding macro-level driving data vector. X CFD In addition to the gap geometry and material parameters; after training, the PINN surrogate model outputs high-precision ozone permeation flux Φ at each node. PINN .

[0045] Step S30 is based on the PINN surrogate model to predict microscale penetration flux. In some embodiments, such as... Figure 4 As shown, step S30 can be achieved through the following steps S301-S303.

[0046] S301: PINN Deployment and Interfaces.

[0047] At each penetration node, a lightweight PINN proxy model is deployed, which receives macroscopic driving data vectors from the CFD. X CFD It also receives the gap geometry and material parameters corresponding to the node.

[0048] S302: Physical constraint embedding and model training.

[0049] In PINN training, the loss function design includes data loss and physical loss, with the total loss function being... L total =L data +λL physics Among them, data loss items L dataThe mean square error is calculated based on high-fidelity CFD microgrid simulation or wind tunnel experimental data to ensure the consistency between model predictions and real data; physical loss term. L physics By constructing Darcy's law residuals To achieve this, in which K eff To achieve an effective penetration rate of implicit learning in networks, A d The characteristic area of ​​permeation used in Darcy's law calculations. L d For the corresponding permeation characteristic length, this constraint forces the network prediction to conform to the basic physical quantities of porous media flow. This is achieved by adjusting the hyperparameters. λ By balancing data fitting and physical consistency, the model's predictions ultimately achieve a consistency of over 92% with Darcy's law, ensuring that the predictions are both data-driven and physically reliable.

[0050] S303: Flux Prediction.

[0051] The trained PINN model, by integrating all input parameters, directly outputs the high-precision ozone permeation flux Φ at each node. PINN .

[0052] S40: Real-time local wind speed based on macroscopic driving data vectors U i The convection enhancement component in the mass transfer coefficient is dynamically corrected; the baseline mass transfer coefficient is updated online by calling the material aging kinetic model based on the simulation cumulative time; the equivalent pressure difference generated by the thermo-pressure effect is coupled and vector-superimposed with the wind pressure difference to obtain the comprehensive equivalent pressure difference Δ. Peff The updated mass transfer coefficient is compared with the overall equivalent pressure difference Δ. Peff The data is fed back to the PINN agent model in real time, driving it to make predictions for the next time step.

[0053] Step S40 is the step of multi-effect dynamic coupling and online correction of material aging coefficient. In some embodiments, such as Figure 5 As shown, step S40 can be achieved through the following steps S401-S404.

[0054] S401: Dynamic wind pressure enhancement correction.

[0055] Based on the real-time local wind speed (U) provided by CFD, the convective enhancement component in the transmission coefficient is dynamically corrected: K p =K p0 + aU b ,in K p0As the reference mass transfer coefficient, a and b These are parameters fitted based on experimental data. This allows for the quantification of the instantaneous amplification effect of wind pressure on the mass transfer process, enabling the model to respond to transient wind field changes.

[0056] S402: Material aging coefficient updated online.

[0057] Based on the cumulative simulation time, the material aging database is accessed. An aging kinetic model of the Arrhenius form is then used. Online update of reference mass transfer coefficient K p0 This enables quantitative simulation of material performance degradation, giving the model long-term predictive capabilities and allowing it to reflect changes in material permeability after many years of use.

[0058] S403: Thermo-pressure coupling based on the chimney effect.

[0059] Introducing chimney effect correction, the equivalent pressure difference generated by hot pressing is calculated. ,in ξ As a correction factor, The density of air is calculated from the temperature. g It is the acceleration due to gravity. H The effective height is determined by vector superposition of wind pressure and thermal pressure to obtain the comprehensive equivalent pressure difference acting on the gap. Δ Peff =ΔP wind +ΔP stack .

[0060] S404: Multi-effect integrated coupling simulation.

[0061] Integrate all the above corrections and use the updated baseline mass transfer coefficient. K p0 (t) and the comprehensive equivalent pressure difference Δ P eff (t) is used as a key input parameter and is transmitted in real time to the PINN proxy model in step three to drive it to make accurate predictions for the next time step.

[0062] Through the steps S10-S40 described above, a comprehensive coupled simulation of multiple physical effects, including wind pressure, thermal pressure, and material aging, was finally achieved. The PINN model receives dynamically corrected input parameters that better reflect real-world physical scenarios, thereby significantly improving the accuracy and reliability of ozone permeation flux prediction.

[0063] In some embodiments, such as Figure 6 As shown, the method for predicting ozone permeation flux through the gap structure of a high-rise building, after step S40, further includes the following steps: S50: System verification, fine-tuning, and performance evaluation.

[0064] The model predicts the dynamic ozone permeation flux. J ( t As a source term, it is input into building energy simulation software such as EnergyPlus, and the indoor concentration is received in feedback from it. C in (t) is fed back to the model to form an accurate closed-loop verification system, and the verification results are shown in Table 1 below.

[0065] Table 1 Performance Verification Results

[0066] By using 1-2 sets of field test data, the pre-trained PINN surrogate model can be transferred or rapidly fine-tuned, significantly reducing the dependence on massive labeled data and keeping the final prediction error stably within ±7%.

[0067] By comparing with full CFD encrypted simulations and wind tunnel experimental data, the superior performance of this method is evaluated in terms of computational efficiency (more than 20 times faster), resource consumption (95% reduction in memory usage), and prediction accuracy (error ~6.3%). This application's method, through a CFD-PINN multi-scale coupling architecture, improves computational efficiency by more than 20 times and reduces memory usage by 95% while maintaining high accuracy (error 6.3%). It innovatively integrates wind pressure dynamic correction, material aging updates, and thermo-pressure effect coupling mechanisms, improving the accuracy of reproducing sudden pollution events from 52% to 89%. By integrating a material aging database and embedding physical constraints, it achieves for the first time accurate prediction and physically reliable simulation of the entire life cycle of building ozone infiltration.

[0068] This application also provides a device for predicting the ozone permeation flux of gap structures in high-rise buildings, such as... Figure 7 As shown, the ozone permeation flux prediction device for the gap structure of this high-rise building includes: The data acquisition module 701 is configured to acquire the building's gap geometry parameters, material properties and aging kinetic parameters, and environmental field driving parameters; and initialize the CFD solver and PINN surrogate model respectively. The CFD calculation module 702 is configured to construct a macroscopic computational domain containing the target building and perform mesh generation, use computational fluid dynamics (CFD) simulation to obtain the dynamic wind pressure distribution on the building surface, and extract macroscopic driving data vectors at preset infiltration nodes. Where Δ P i For wind pressure difference, U i For local wind speeds, T outOutdoor temperature C out This refers to the outdoor ozone concentration. PINN proxy module 703 is configured to deploy PINN proxy models at each of the penetration nodes, wherein the PINN proxy model receives the corresponding macro-driving data vector. X CFD In addition to the gap geometry and material parameters; after training, the PINN surrogate model outputs high-precision ozone permeation flux Φ at each node. PINN ; The correction module 704 is configured to adjust the real-time local wind speed based on the macroscopic driving data vector. U i The convection enhancement component in the mass transfer coefficient is dynamically corrected; the baseline mass transfer coefficient is updated online by calling the material aging kinetic model based on the simulation cumulative time; the equivalent pressure difference generated by the thermo-pressure effect is coupled and vector-superimposed with the wind pressure difference to obtain the comprehensive equivalent pressure difference Δ. Peff The updated mass transfer coefficient is compared with the combined equivalent pressure difference Δ. Peff The data is fed back to the PINN agent model in real time, driving it to make predictions for the next time step.

[0069] This application provides an electronic device. The electronic device may include a processor and a memory, wherein the processor and the memory can communicate; exemplarily, the processor and the memory communicate via a communication bus.

[0070] The processor executes computer execution instructions stored in memory, causing the processor to perform the scheme in the above embodiments. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0071] The communication bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The system bus can be divided into address bus, data bus, control bus, etc. Transceivers are used to enable communication between database access devices and other computers (e.g., clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) and may also include non-volatile memory.

[0072] The electronic device provided in this application embodiment can be the terminal device described in the above embodiments.

[0073] This application also provides a computer-readable storage medium storing computer instructions. When the computer instructions are executed on a computer, the computer performs the technical solution of the method for predicting ozone infiltration flux in high-rise building gap structures described in the above embodiments.

[0074] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When the at least one processor executes the computer program, it can implement the technical solution of the method for predicting ozone infiltration flux in high-rise building gap structures described in the above embodiments.

[0075] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0076] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to implement the solution of this embodiment according to actual needs.

[0077] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The unit composed of the above modules can be implemented in hardware or in the form of hardware plus software functional units.

[0078] The integrated modules described above, implemented as software functional modules, can be stored in a computer-readable storage medium. These software functional modules, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods of the various embodiments of this application.

[0079] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly manifested as execution by a hardware processor, or execution by a combination of hardware and software modules within the processor.

[0080] The memory may include high-speed RAM, and may also include non-volatile storage (NVM), such as at least one disk storage device, and may also be a USB flash drive, external hard drive, read-only memory, disk or optical disc, etc.

[0081] Buses can be Industry Standard Architecture (ISA) buses, Peripheral Component Interconnect (PCI) buses, or Extended Industry Standard Architecture (EISA) buses, etc. Buses can be categorized into address buses, data buses, control buses, etc.

[0082] The aforementioned storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0083] An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Alternatively, the storage medium can be an integral part of the processor. The processor and storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and storage medium can exist as discrete components in an electronic control unit or main control device.

[0084] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for predicting ozone permeation flux through gap structures in high-rise buildings, characterized in that, The method includes: Obtain the geometric parameters of the gaps in the building, material properties and aging kinetic parameters, as well as the environmental field driving parameters; Computational fluid dynamics (CFD) was used to simulate the external wind field of a building, obtain the dynamic wind pressure distribution on the surface, and extract macroscopic driving data vectors at multiple preset infiltration nodes. Where Δ P i For wind pressure difference, U i For local wind speeds, T out Outdoor temperature C out This refers to the outdoor ozone concentration. The macroscopic driving data at each of the aforementioned permeation nodes, along with their corresponding gap geometry and material parameters, are input into a pre-trained PINN surrogate model, which incorporates Darcy's law as a physical constraint. The microscale ozone permeation flux Φ at each node is then output. PINN ; Based on the real-time local wind speed in the macro driving data vector U i Dynamically correct the mass transfer coefficient; update the baseline mass transfer coefficient online based on the cumulative simulation time to reflect material aging, and calculate the comprehensive equivalent pressure difference Δ by coupling the hot-pressing effect. Peff ; The updated mass transfer coefficient is compared with the comprehensive equivalent pressure difference Δ. Peff The data is fed back to the PINN agent model in real time, driving it to make predictions for the next time step.

2. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 1, characterized in that, The gap geometry parameters include the gap length. L Equivalent hydraulic diameter D h Surface roughness height k s and tortuosity factor τ Wherein, the equivalent hydraulic diameter D h The calculation formula is: In the formula, A For cross-sectional area, P It is a wetted period.

3. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 1, characterized in that, When training the PINN proxy model, its total loss function L total Data loss items L data and physical loss items L physics Composition, that is L total = L data +λ L physics Where λ is a hyperparameter; the physical loss term L physics This is achieved by constructing the Darcy's law residual, which is expressed as: In the formula, K eff To achieve an effective penetration rate of implicit learning in networks, A d The characteristic area of ​​permeation used in Darcy's law calculations. L d This corresponds to the permeation feature length. μ This represents the fluid viscosity.

4. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 1, characterized in that, Mass transfer coefficient dynamically corrected based on real-time local wind speed U K p The following formula is used: K p =K p0 +aU b In the formula, K p0 As the reference mass transfer coefficient, a and b The parameters are fitted based on experimental data. U This refers to local wind speed.

5. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 4, characterized in that, The baseline mass transfer coefficient is updated online based on the simulated cumulative time t. K p0 ( t The material aging kinetics model is as follows: In the formula, K p0 (0) is the initial reference mass transfer coefficient, A aging E is a pre-exponential factor. a,aging Let R be the activation energy, R be the ideal gas constant, and T be the temperature. α、β This is an empirical coefficient for performance degradation.

6. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 1, characterized in that, The equivalent pressure difference Δ generated by the thermo-pressure effect P stack The calculation formula is as follows: In the formula, ξ As a correction factor, ρ out and ρ in These are the outdoor and indoor air densities calculated from indoor and outdoor temperatures, respectively. g Let H be the acceleration due to gravity, and H be the effective height.

7. The method for predicting ozone permeation flux through gap structures in high-rise buildings according to claim 1, characterized in that, Before using the CFD method for simulation, the steps include: constructing a macroscopic computational domain centered on the target building and performing mesh generation, and initializing the CFD solver and PINN proxy model.

8. A device for predicting ozone permeation flux through gap structures in high-rise buildings, characterized in that, The apparatus includes functional modules for performing each step of the method according to any one of claims 1-7.

9. An electronic device, characterized in that, include: A processor and a memory, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the method as described in any one of claims 1-7.

10. A computer-readable storage medium, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1-7.

Images