A design method and system for underground public buildings that optimizes ventilation and lighting.

Abstract

This invention proposes a design method and system for the coordinated optimization of ventilation and lighting in underground public buildings, comprising: simulating natural ventilation and lighting under different control design factors and their combinations, and screening out key control design factors; establishing a mathematical model for the coordinated optimization of natural ventilation and natural lighting using Bayesian inference; determining quantitative evaluation indicators for underground public buildings through factor analysis; approximating the mathematical model for the coordinated optimization of natural ventilation and natural lighting using Bayesian inference with key control design factors as input variables and the quantitative evaluation indicators of underground public buildings as the objective function, thereby obtaining a control model for natural ventilation and natural lighting; and obtaining the optimal threshold range of key control design factors that make the objective function optimal by using the key control design factors of the underground public building to be optimized as input variables and the control model for natural ventilation and natural lighting as the objective function.

Description

Technical Field

[0001] This invention belongs to the field of building physics and building energy conservation technology, specifically a design method and system for underground public buildings that optimizes ventilation and lighting. Background Technology

[0002] Due to their enclosed physical characteristics, underground public buildings generally suffer from insufficient natural ventilation and lighting. This limitation leads to a series of problems, including increased environmental pollution, higher energy consumption, and harm to human health. Specifically: First, due to obstructed air circulation, pollutants easily accumulate in underground public buildings; insufficient natural light impairs physical and mental health, resulting in a significantly worse spatial experience compared to surface environments. Second, to compensate for insufficient natural ventilation and lighting, underground public buildings typically rely heavily on mechanical ventilation and artificial lighting, causing significant energy consumption and hindering the achievement of energy conservation and carbon reduction goals. Furthermore, poor ventilation easily leads to the accumulation of pathogens, bacteria, and viruses in enclosed environments, especially during peak hours, exacerbating the risk of disease transmission.

[0003] Currently, common design methods for natural ventilation in underground public buildings include: creating sunken plazas and courtyards to form airflow channels; arranging ventilation shafts to promote air exchange; using facilities such as wind towers and air deflectors to enhance airflow organization; and utilizing natural dynamic mechanisms such as wind pressure difference and thermal pressure difference to promote airflow. Natural lighting measures for underground public buildings mainly include: introducing natural light through open spaces such as sunken plazas and courtyards; installing vertical lighting facilities such as light wells and skylights; and using technologies such as light pipes and fiber optic lighting.

[0004] It is worth noting that natural ventilation and natural lighting do not exist in isolation in underground public buildings; they overlap to some extent in form and effect. For example, sunken plazas, courtyards, and light wells can both introduce natural light and organize airflow, achieving a synergistic effect of ventilation and lighting. Some components (such as window louvers and light guides) can also guide airflow and enhance ventilation. This coupling phenomenon of natural ventilation and natural lighting has a significant impact on core environmental indicators such as air quality, thermal comfort, and visual comfort in underground public buildings.

[0005] However, existing studies mostly focus on single factors, optimizing natural ventilation or natural lighting separately, with little consideration for the coupling relationship and synergistic mechanisms between the two, resulting in the overall environmental benefits being difficult to fully realize. Furthermore, the comfort needs of users in underground public buildings exhibit dynamic changes, influenced by real-time variations in environmental parameters such as temperature, humidity, wind speed, and light intensity. Different groups also demonstrate diverse comfort needs at different times, in different spatial areas, and due to individual physiological differences. This complexity places higher demands on the dynamic responsiveness and personalized adaptability of environmental control systems. Summary of the Invention

[0006] Purpose of the invention: To address the current situation where existing technologies generally neglect the coupling relationship between natural ventilation and natural lighting, and fail to fully consider human comfort, this invention proposes a design method and system for underground public buildings that optimizes ventilation and lighting in a coordinated manner. Based on multiphysics simulation, environmental and human perception data collection, and intelligent optimization algorithms, it achieves coordinated regulation of natural ventilation and natural lighting to improve the environmental quality of underground public buildings and reduce their energy consumption.

[0007] Technical solution: In a first aspect, the present invention provides a design method for underground public buildings that optimizes ventilation and lighting in a coordinated manner, comprising the following steps:

[0008] Step 1: Determine the control design factors for underground public buildings from the selected design cases;

[0009] Step 2: Simulate natural ventilation and natural lighting under different control design factors and their combinations to obtain the performance of natural ventilation and natural lighting under different control design factors and their combinations; based on the performance of natural ventilation and natural lighting under different control design factors and their combinations, screen out the key control design factors that affect the performance of natural lighting and natural ventilation; the combination of control design factors shall consist of at least two control design factors.

[0010] Step 3: Clarify the mathematical expressions for the natural ventilation performance evaluation index and the natural lighting performance evaluation index. Using the Bayesian inference method, construct a probabilistic mathematical relationship model between key control design factors and performance evaluation indexes to obtain a mathematical model for the synergistic optimization of natural ventilation and natural lighting.

[0011] Step 4: Collect human perception data of users under different environmental conditions; the environmental data refers to the environmental data under the selected typical underground public building design case.

[0012] Step 5: Perform spatiotemporal alignment and normalization on the environmental data and human perception data in sequence to obtain human perception parameters and environmental variables; explore the correlation between human perception parameters and environmental variables through factor analysis to determine quantitative evaluation indicators for underground public buildings.

[0013] Step 6: Using a Bayesian neural network, with the key control design factors selected in Step 2 as input variables and the quantitative evaluation index of underground public buildings determined in Step 5 as the objective function, the mathematical model for the coordinated optimization of natural ventilation and natural lighting constructed in Step 3 is approximated by Bayesian inference to obtain the natural ventilation and natural lighting control model.

[0014] Step 7: Using the key control design factors of the underground public building to be optimized as input variables, and the natural ventilation and natural lighting control model obtained in Step 6 as the objective function, the optimal threshold range of the key control design factors that make the objective function optimal is obtained, thereby obtaining the control design scheme of the underground public building to be optimized.

[0015] Furthermore, it also includes:

[0016] Step 8: Construct a virtual reality environment based on the usage scenario of the proposed optimized control design scheme for underground public buildings; collect human perception data of subjects in real time in this virtual reality environment, and adjust the proposed optimized control design scheme for underground public buildings based on the collected human perception data.

[0017] Furthermore, the control design factors include: the location coordinates, opening size, shape outline, and orientation angle of the spatial interface; the spatial interface includes: sunken plaza, sunken courtyard, underground atrium, solar chimney, underground ventilation, light well, ventilation tower, side window, skylight, and light well.

[0018] Furthermore, the environmental data includes: lighting data and ventilation data. The lighting data includes illuminance, daylight factor, lighting uniformity, color temperature, autonomous daylighting rate, spatial autonomous daylighting rate, daylight availability, effective daylighting, annual sunshine hours, uncomfortable glare index, and daytime glare probability. The ventilation data includes: air age, wind speed, number of air changes per hour, and minimum fresh air volume.

[0019] Furthermore, the user's human perception data includes the user's physiological data and the user's behavioral data; the user's physiological data includes: pupillary instability index, saccade rate, total fixation time, first fixation time, skin conductance, heart rate, and heart rate variability; the user's behavioral data includes movement path and subjective comfort evaluation data.

[0020] Furthermore, the environmental data and human perception data are sequentially subjected to spatiotemporal alignment and normalization to obtain human perception parameters and environmental variables. Specific operations include:

[0021] A dynamic time warping method is used to keep the temporal sequence of environmental data and human perception data consistent.

[0022] By using spatial coordinate mapping, a spatial mapping is established between human perception data and environmental data, resulting in environmental data and human perception data that are consistent in time and spatially aligned.

[0023] The Z-score normalization method is used to normalize environmental data and human perception data that are consistent in time and spatially aligned to obtain human perception parameters and environmental variables.

[0024] Furthermore, in step 6, the natural ventilation and natural lighting control model is expressed as follows:

[0025] ;

[0026] ;

[0027] in, As an indicator for evaluating ventilation efficiency, The light environment quality evaluation index is a quantitative evaluation index composed of ventilation efficiency evaluation index and light environment quality evaluation index. , These are the weighting coefficients for ventilation efficiency evaluation indicators and light environment quality evaluation indicators, respectively. Indicates the control design factor, This represents the optimized control design factor; Represents the set of control design factors. Let be the posterior probability, and represent the optimized distribution of the control design factors.

[0028] Secondly, this invention discloses a design system for underground public buildings that optimizes ventilation and lighting in a coordinated manner, comprising:

[0029] The key control design factor screening module is used to simulate natural ventilation and natural lighting under different control design factors and their combinations, and obtain the natural ventilation performance and natural lighting performance under different control design factors and their combinations. Based on the natural ventilation performance and natural lighting performance under different control design factors and their combinations, the key control design factors that affect the natural lighting performance and natural ventilation performance are screened. The combination of control design factors consists of at least two control design factors.

[0030] The module for establishing a mathematical model for the synergistic optimization of natural ventilation and natural lighting is used to clarify the mathematical expressions for the performance evaluation indicators of natural ventilation and natural lighting. Using the Bayesian inference method, a probabilistic mathematical relationship model between key control design factors and performance evaluation indicators is constructed to obtain the mathematical model for the synergistic optimization of natural ventilation and natural lighting.

[0031] The module for determining quantitative evaluation indicators for underground public buildings is used to collect human perception data of users under different environmental data. The environmental data is the environmental data under the selected typical design case of underground public buildings. The environmental data and human perception data are sequentially processed by spatiotemporal alignment and normalization to obtain human perception parameters and environmental variables. The correlation between human perception parameters and environmental variables is explored by factor analysis to determine the quantitative evaluation indicators for underground public buildings.

[0032] The module for establishing a natural ventilation and natural lighting control model is used to approximate the mathematical model of natural ventilation and natural lighting synergistic optimization by using a Bayesian neural network, with key control design factors as input variables and quantitative evaluation indicators of underground public buildings as objective functions, through Bayesian inference, and to obtain the natural ventilation and natural lighting control model.

[0033] The control design module for underground public buildings is used to obtain the optimal threshold range of the key control design factors that make the target function optimal, with the key control design factors of the underground public building to be optimized as input variables and the natural ventilation and natural lighting control model as the objective function. This results in the control design scheme of the underground public building to be optimized.

[0034] Furthermore, it also includes:

[0035] The dynamic optimization module is used to construct a virtual reality environment for the application scenario of the proposed optimization design scheme for underground public buildings. In this virtual reality environment, human perception data of the subjects is collected in real time, and the control design scheme of the proposed optimization underground public buildings is adjusted based on the collected human perception data.

[0036] Furthermore, the control design factors include: the location coordinates, opening size, shape profile, and orientation angle of the spatial interface; the spatial interface includes: sunken plaza, sunken courtyard, underground atrium, solar chimney, underground ventilation, light well, ventilation tower, side window, skylight, and light well.

[0037] Beneficial effects: Compared with existing technologies, the method of this invention achieves efficient coordinated control of natural ventilation and natural lighting by constructing a multiphysics simulation model, coupling analysis of environmental and human perception data, and using an intelligent optimization algorithm based on Bayesian neural networks. It has the following advantages:

[0038] (1) By establishing a mathematical model for the synergistic optimization of natural ventilation and natural lighting, this invention clarifies the interaction mechanism of different control design factors (such as the opening ratio of the sunken plaza, the window-to-wall ratio of the side windows, and the type of light guiding equipment) on indicators such as wind speed, illuminance, and effective daylighting, thus making up for the shortcomings of the existing technology in independently studying the two and lacking analysis of the coupling mechanism.

[0039] (2) The method of the present invention collects environmental data and human perception data, and uses factor analysis to establish a quantitative evaluation index for underground public buildings driven by human comfort perception, so that the environmental optimization target is more in line with the physiological and psychological needs of actual users, and improves the accuracy and scientific nature of environmental regulation.

[0040] (3) The method of the present invention adopts a Bayesian neural network, which can adjust the natural ventilation and lighting strategies in real time according to different seasons, time periods and user groups, realize the dynamic adaptive optimization of the underground public building environment, and overcome the problems of static design and response lag in the existing methods.

[0041] (4) The method of the present invention can effectively improve the air circulation, natural lighting level and user environmental comfort of underground public buildings, reduce the dependence on mechanical ventilation and artificial lighting systems, and achieve the dual goals of energy conservation and emission reduction and improving space quality.

[0042] (5) The method of the present invention can provide technical support for the design and operation of various underground public buildings such as underground commercial space, underground administrative office space, underground cultural and leisure space, underground education and scientific research space, underground medical space, underground sports space, underground pedestrian passage, underground bus station, underground passenger station, and underground parking lot. It has good feasibility, adaptability and promotion and application value, and meets the requirements of green and low-carbon development and healthy city construction. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of an underground public building design method for synergistic optimization of ventilation and lighting, provided as an embodiment of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the following description, in conjunction with the accompanying drawings and embodiments, will further illustrate the design method for underground public buildings that optimizes ventilation and lighting in synergy.

[0045] This invention proposes a design method for underground public buildings that optimizes ventilation and lighting in a synergistic manner. This method improves air quality, thermal comfort, and visual comfort, while reducing reliance on mechanical ventilation and artificial lighting, thereby effectively enhancing the environmental quality of underground public buildings and reducing energy consumption. Figure 1As shown, the method mainly includes the following steps:

[0046] Step 1: Extract relevant forms and parameters of natural ventilation and natural lighting from typical underground public building cases, and determine the control design factors. These control design factors include the geometric parameters of the spatial interface, which includes, but is not limited to, sunken plazas, sunken courtyards, underground atriums, solar chimneys, underground ventilation, light wells, ventilation towers, side windows, skylights, and light wells. The geometric parameters include: the location coordinates, opening size, shape outline, and orientation angle of the spatial interface.

[0047] Step 2: Perform natural ventilation simulation using the CFD simulation software Fluent. Set boundary conditions such as inlet wind speed range (e.g., 1~5 m / s), ambient temperature (e.g., 15~35℃), and air pressure (e.g., standard atmospheric pressure). Simulate the natural ventilation performance index values ​​under different control design factors and their combinations. These natural ventilation performance index values ​​include, but are not limited to, indoor wind speed distribution and air replacement rate. In this embodiment, the CFD simulation software Fluent is set to: standard k-ε turbulence model, unsteady-state simulation, and time step of 1 s. The control design factor combination refers to a combination of at least two control design factors.

[0048] Natural daylighting was simulated using the Radiance lighting environment simulation software. The solar altitude angle and climate data files (e.g., typical meteorological year TYC files) were set, and the natural daylighting performance index values ​​under different operating conditions were analyzed. These operating conditions consisted of control design factors and their combinations, solar altitude angle, and climate data. The natural daylighting performance index values ​​included, but were not limited to, indoor illuminance (unit: lx) and effective daylight utilization (DA value). In this embodiment, the Radiance lighting environment simulation software was set to: rendering parameters - quality Q1, accuracy A0.15, and light source tracking steps 5000.

[0049] Step 3: Record the natural ventilation performance index value and natural lighting performance index value corresponding to each control design factor and its combination, as the natural ventilation performance and natural lighting performance.

[0050] Step 4: Based on the natural ventilation performance and natural lighting performance corresponding to different control design factors and their combinations, screen out the key control design factors that are more sensitive to and have a greater impact on natural lighting and natural ventilation performance.

[0051] Step 5: Based on the content recorded in Step 3, clarify the mathematical expressions of the natural ventilation performance evaluation index and the natural lighting performance evaluation index. Then, using the Bayesian inference method, construct a probabilistic mathematical relationship model between the key control design factors selected in Step 4 and the performance evaluation index, thereby establishing a mathematical model for the synergistic optimization of natural ventilation and natural lighting, also known as the natural lighting-ventilation coupling model.

[0052] Step 6: Conduct on-site environmental testing on selected typical underground public building cases by deploying monitoring equipment at multiple points to collect environmental data. Simultaneously, conduct human factors experiments using wearable devices (such as skin conductance sensors and heart rate monitors) to collect users' human sensory data. Details are shown in Table 1.

[0053] Table 1 Environmental Data and Human Perception Data

[0054]

[0055] Step 7: Due to significant differences in individual user characteristics (such as age, gender, and behavioral preferences), both human perception data and environmental data exhibit multimodal and heterogeneous characteristics. Traditional single-data analysis methods are insufficient to accurately depict the interaction between humans and the environment, failing to meet the needs of dynamic environmental adjustment in underground public buildings. Therefore, this embodiment of the invention performs spatiotemporal alignment and normalization on multimodal data, transforming it into unified human perception parameters and environmental variables. Based on factor analysis, it delves into the correlation between the two, thereby determining quantitative evaluation indicators for underground public buildings. In this embodiment, these quantitative evaluation indicators include: ventilation efficiency evaluation indicators and light environment quality evaluation indicators. Specific operations include:

[0056] First, the Dynamic Time Warping (DTW) method is used to ensure the temporal consistency of human perception data and environmental monitoring data, as shown below:

[0057] ;

[0058] in, This represents the matching cost between the i-th and j-th data points. For Euclidean distance , These represent the values ​​at the i-th time point in the human perception data and environmental data sequences, respectively.

[0059] Then, using spatial coordinate mapping technology, the user location data is... Environmental data (e.g., temperature) Establish spatial mapping to achieve spatial alignment and effective integration of multi-source data, represented as:

[0060] ;

[0061] Furthermore, the Z-score normalization method is used to ensure the scale consistency of different physical quantities.

[0062] After the above data processing, human perception data is processed into human perception parameters, and environmental data is processed into environmental variables.

[0063] Finally, focusing on feature extraction and cross-domain fusion of multimodal human perception data, the correlation between human perception parameters (S) and environmental variables (E) is explored through factor analysis, thereby determining quantitative evaluation indicators for underground public buildings, namely ventilation efficiency evaluation indicators and light environment quality evaluation indicators.

[0064] ;

[0065] in, This is the factor loading matrix, where each element represents the loading of the i-th original variable on the j-th factor variable. This is the noise term.

[0066] For example, wind speed and air exchange rate are significantly correlated with heart rate and skin conductance, indicating that the better the airflow, the more relaxed a person is; illuminance and uniformity are significantly correlated with glare score and pupil instability, indicating that the softer the light, the more comfortable a person is.

[0067] In this embodiment, factor analysis methods may be employed, but are not limited to, principal component analysis (PCA).

[0068] Step 8: Natural ventilation and lighting in underground public buildings have complex interactions. Controlling design factors not only affects light levels but also significantly impacts the thermal environment and airflow. Traditional optimization methods typically focus on single variables, making it difficult to achieve multi-objective synergistic optimization. Therefore, this embodiment uses a Bayesian neural network (BNN), with the key control design factors selected in Step 4 as input variables and the quantitative evaluation indicators of underground public buildings determined in Step 7 as the objective function. Bayesian inference is used to approximate the mathematical model for the synergistic optimization of natural ventilation and natural lighting constructed in Step 5, resulting in a natural ventilation and natural lighting control model, expressed as:

[0069] ;

[0070] ;

[0071] in, As an indicator for evaluating ventilation efficiency, The light environment quality evaluation index is a quantitative evaluation index composed of ventilation efficiency evaluation index and light environment quality evaluation index. , These are the weighting coefficients for ventilation efficiency evaluation indicators and light environment quality evaluation indicators, respectively. Indicates the control design factor, This represents the optimized control design factor; Represents the set of control design factors. Let be the posterior probability, and represent the optimized distribution of the control design factors.

[0072] The Bayesian neural network (BNN) used in this embodiment has 50 hidden layer nodes, a learning rate of 0.001, and 1000 iterations.

[0073] Step 9: Using the key control design factors of the underground public building to be optimized as input variables, and the natural ventilation and natural lighting control model obtained in Step 8 as the objective function, obtain the optimal threshold range of the key control design factors that make the objective function optimal.

[0074] Step 10: Using the key control design factors of the proposed underground public building and the optimal threshold range of the obtained key control design factors, obtain the control design scheme of the proposed underground public building.

[0075] Based on the above design method, a control design platform integrating simulation calculation, dynamic regulation, and human factor experimental feedback can be further developed. This control design platform includes: a simulation module (Fluent + Radiance linkage), an optimization module (Bayesian optimizer), and a human factor feedback module (VR experimental system). Since the underground public building to be modified using the method of this embodiment has not yet been constructed, the usage scenarios of the underground public building under different control design schemes of the proposed optimization can be simulated in a virtual reality (VR) environment. The different control design schemes of the proposed optimization underground public building are design schemes that conform to the optimal threshold range of key control design factors. The comfort level of the subjects is collected in real time in the virtual reality (VR) environment, and the control design scheme of the proposed optimization underground public building is dynamically adjusted based on the comfort feedback.

[0076] The virtual reality (VR) environment of this invention is developed using the Unity engine and supported by the Oculus Quest 2 device. This invention uses VR human factors experiments to verify the optimization effect, ultimately forming a proposed optimized control design scheme for underground public buildings. This scheme guides underground public buildings to achieve adaptive optimization and adjustment of the environment under different seasons, time periods, and user needs.

Claims

1. A design method for underground public buildings that optimizes ventilation and lighting in a coordinated manner, characterized in that: Includes the following steps: Step 1: Determine the control design factors for underground public buildings from the selected design cases; Step 2: Under different control design factors and their combinations, natural ventilation and natural lighting simulations were conducted to obtain the performance of natural ventilation and natural lighting under different control design factors and their combinations. Based on the performance of natural ventilation and natural lighting under different control design factors and their combinations, key control design factors that affect natural lighting and natural ventilation were screened. The combination of control design factors must consist of at least two control design factors; Step 3: Clarify the mathematical expressions for the natural ventilation performance evaluation index and the natural lighting performance evaluation index. Using the Bayesian inference method, construct a probabilistic mathematical relationship model between key control design factors and performance evaluation indexes to obtain a mathematical model for the synergistic optimization of natural ventilation and natural lighting. Step 4: Collect human perception data of users under different environmental conditions; the environmental data refers to the environmental data under the selected typical underground public building design case. Step 5: Perform spatiotemporal alignment and normalization on the environmental data and human perception data in sequence to obtain human perception parameters and environmental variables; explore the correlation between human perception parameters and environmental variables through factor analysis to determine quantitative evaluation indicators for underground public buildings. Step 6: Using a Bayesian neural network, with the key control design factors selected in Step 2 as input variables and the quantitative evaluation index of underground public buildings determined in Step 5 as the objective function, the mathematical model for the coordinated optimization of natural ventilation and natural lighting constructed in Step 3 is approximated by Bayesian inference to obtain the natural ventilation and natural lighting control model. Step 7: Using the key control design factors of the underground public building to be optimized as input variables, and the natural ventilation and natural lighting control model obtained in Step 6 as the objective function, the optimal threshold range of the key control design factors that make the objective function optimal is obtained, thereby obtaining the control design scheme of the underground public building to be optimized.

2. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: Also includes: Step 8: Construct a virtual reality environment based on the usage scenarios of the proposed optimized control design scheme for underground public buildings; The human perception data of the subjects is collected in real time in this virtual reality environment, and the control design scheme of the proposed optimized underground public building is adjusted based on the collected human perception data.

3. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: The control design factors include: the location coordinates, opening size, shape outline and orientation angle of the spatial interface; the spatial interface includes: sunken plaza, sunken courtyard, underground atrium, solar chimney, underground ventilation, light well, ventilation tower, side window, skylight and light well.

4. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: The environmental data includes lighting data and ventilation data. The lighting data includes illuminance, daylight factor, daylight uniformity, color temperature, autonomous daylight rate, spatial autonomous daylight rate, daylight availability, effective daylight intensity, annual sunshine hours, uncomfortable glare index, and daytime glare probability. The ventilation data includes air age, wind speed, number of air changes per hour, and minimum fresh air volume.

5. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: The user's human sensory data includes the user's physiological data and the user's behavioral data; The user's physiological data includes: pupillary instability index, saccade rate, total fixation time, first fixation time, skin conductance, heart rate, and heart rate variability. The user's behavioral data includes movement paths and subjective comfort evaluation data.

6. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: The process of sequentially performing spatiotemporal alignment and normalization on environmental data and human perception data to obtain human perception parameters and environmental variables includes the following specific operations: A dynamic time warping method is used to keep the temporal sequence of environmental data and human perception data consistent. By using spatial coordinate mapping, a spatial mapping is established between human perception data and environmental data, resulting in environmental data and human perception data that are consistent in time and spatially aligned. The Z-score normalization method is used to normalize environmental data and human perception data that are consistent in time and spatially aligned to obtain human perception parameters and environmental variables.

7. The design method for underground public buildings that optimizes ventilation and lighting according to claim 1, characterized in that: In step 6, the natural ventilation and natural lighting control model is expressed as follows: ； ； in, As an indicator for evaluating ventilation efficiency, The light environment quality evaluation index is a quantitative evaluation index composed of ventilation efficiency evaluation index and light environment quality evaluation index. , These are the weighting coefficients for ventilation efficiency evaluation indicators and light environment quality evaluation indicators, respectively. Indicates the control design factor, This represents the optimized control design factor; Represents the set of control design factors. Let be the posterior probability, and represent the optimized distribution of the control design factors.

8. A design system for underground public buildings that optimizes ventilation and lighting in a coordinated manner, characterized in that: include: The key control design factor screening module is used to simulate natural ventilation and natural lighting under different control design factors and their combinations, and obtain the natural ventilation performance and natural lighting performance under different control design factors and their combinations. Based on the performance of natural ventilation and natural lighting under different control design factors and their combinations, key control design factors that affect natural lighting and natural ventilation were screened. The combination of control design factors must consist of at least two control design factors; The module for establishing a mathematical model for the synergistic optimization of natural ventilation and natural lighting is used to clarify the mathematical expressions for the performance evaluation indicators of natural ventilation and natural lighting. Using the Bayesian inference method, a probabilistic mathematical relationship model between key control design factors and performance evaluation indicators is constructed to obtain the mathematical model for the synergistic optimization of natural ventilation and natural lighting. The module for determining quantitative evaluation indicators for underground public buildings is used to collect human perception data of users under different environmental data. The environmental data is the environmental data under the selected typical design case of underground public buildings. The environmental data and human perception data are sequentially processed by spatiotemporal alignment and normalization to obtain human perception parameters and environmental variables. Factor analysis was used to explore the correlation between human perception parameters and environmental variables, and quantitative evaluation indicators for underground public buildings were determined. The module for establishing a natural ventilation and natural lighting control model is used to approximate the mathematical model of natural ventilation and natural lighting synergistic optimization by using a Bayesian neural network, with key control design factors as input variables and quantitative evaluation indicators of underground public buildings as objective functions, through Bayesian inference, and to obtain the natural ventilation and natural lighting control model. The control design module for underground public buildings is used to obtain the optimal threshold range of the key control design factors that make the target function optimal, with the key control design factors of the underground public building to be optimized as input variables and the natural ventilation and natural lighting control model as the objective function. This results in the control design scheme of the underground public building to be optimized.

9. The underground public building design system for synergistic optimization of ventilation and lighting according to claim 8, characterized in that: Also includes: The dynamic optimization module is used to construct a virtual reality environment for the application scenarios of the control design scheme of the underground public building to be optimized; The human perception data of the subjects is collected in real time in this virtual reality environment, and the control design scheme of the proposed optimized underground public building is adjusted based on the collected human perception data.

10. The underground public building design system for synergistic optimization of ventilation and lighting according to claim 8, characterized in that: The control design factors include: the location coordinates, opening size, shape outline and orientation angle of the spatial interface; the spatial interface includes: sunken plaza, sunken courtyard, underground atrium, solar chimney, underground ventilation, light well, ventilation tower, side window, skylight and light well.

Images