Health space design optimization method and system based on multi-modal human factor data

Abstract

The invention discloses a health space design optimization method and system based on multi-modal human factor data, and relates to the technical field of health buildings. The method comprises the following steps: synchronously acquiring physiological, behavior, subjective and other multi-modal human factor data and environmental data of a person in a target space; performing data fusion and feature extraction; constructing an environmental parameter-health efficiency prediction model; carrying out design parameter sensitivity analysis and setting an optimization target; the parameterization design tool and the prediction model are linked, and an optimization design scheme set is generated through automatic iteration; and finally outputting an evidence-based design report. According to the method, the real response of the human body is deeply fused into the design process through a data driving mode, quantitative, refined and personalized optimization of health space design is achieved, the method is particularly suitable for solving health living challenges in special environments such as plateau, and design scientificity and health efficiency guarantee are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of healthy building and smart city technology, and to a method and system for optimizing healthy space design based on multimodal human factors data. Background Technology

[0002] Designing architectural spaces that promote the physical and mental well-being of users has become a crucial issue. Currently, spatial design relies heavily on experience, standards, and static environmental simulations, lacking quantitative evidence-based analysis of the dynamic and comprehensive responses of the human body in real or simulated environments. This is especially true in challenging environments such as high-altitude and mountainous regions, where stressors like low oxygen and intense radiation have significant impacts on human health, making it difficult for traditional design methods to accurately address these issues.

[0003] Existing technologies include those that monitor environmental parameters using sensors or collect subjective feedback through questionnaires, but these generally suffer from problems such as limited data dimensions, separation of subjective and objective data, and an inability to reveal the deep mechanisms of human-environment interaction. There is a lack of a systematic approach and technological system capable of collecting and integrating multimodal data, including physiological, behavioral, subjective, and physical environmental data, and using this data to drive spatial design optimization in a closed loop. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for optimizing health space design based on multimodal human factors data. This method systematically collects, integrates, and analyzes multi-dimensional data on human physiology, behavior, and subjective feelings in a specific space, and combines this with high-precision environmental data to construct a quantitative relationship model of "environmental stimulus-human response". This provides objective and quantitative evidence-based basis for the health effectiveness of space design, and intelligently generates or optimizes design schemes, realizing the transformation from "experience-based design" to "evidence-based design".

[0005] According to the purpose of this invention, a method for optimizing health space design based on multimodal human factors data is provided, comprising the following steps: S1: In the target space or simulated space, simultaneously collect the subject's multimodal human factors data and corresponding physical environment data, wherein the multimodal human factors data includes physiological data, behavioral data and subjective evaluation data; S2: Clean, align and fuse the spatiotemporally labeled synchronous data collected in step S1 to form a structured dataset, and extract key indicators reflecting human condition and environmental characteristics from it. S3: Based on machine learning or statistical methods, analyze the correlation between the key indicators and construct a prediction model or influencing mechanism model with spatial environment parameters as input and human health efficacy indicators as output. S4: Based on the model built in step S3, conduct sensitivity analysis of design parameters, identify key influencing factors, and set optimization goals in conjunction with health standards or project requirements; S5: Use parametric design tools in conjunction with the prediction model built in step S3 to automatically or semi-automatically adjust the design parameters, simulate and predict health performance indicators under different schemes, and select the set of optimized schemes that meet the optimization objectives. S6: Based on the set of optimization schemes, generate an evidence-based design report that includes specific design strategies, parameter recommendations, and expected health benefits.

[0006] Further, in step S1, the physiological data includes at least one of heart rate variability, electroencephalogram, skin conductance, and blood oxygen saturation collected by a portable physiological monitoring device; The behavioral data includes at least one of visual attention distribution, spatial activity trajectory, dwell time and distribution density collected by eye trackers, indoor positioning systems, accelerometers or computer vision analysis devices. The physical environment data includes at least one of the following collected by environmental sensors: air temperature, humidity, illuminance, color temperature, wind speed, carbon dioxide concentration, atmospheric pressure, and solar radiation intensity.

[0007] Further, step S1 is performed in an artificial climate chamber capable of reproducing the climatic conditions of plateaus or special regions, including low oxygen, low air pressure, large temperature differences, and strong solar radiation; and / or, Step S1 is performed in a real built environment, and large-scale, long-term data collection is carried out using drones, mobile monitoring equipment and fixed sensor networks.

[0008] Furthermore, in step S3, the human health performance indicators include at least one of the following: cognitive task performance score, physiological stress / recovery index, emotional state index, and subjective comfort attainment rate. The algorithms used to construct the prediction model include at least one of random forest, support vector machine, neural network, and geographically weighted regression.

[0009] Furthermore, in step S4, the design parameters include at least one of the following: building or space lighting environment parameters, thermal environment parameters, air quality parameters, spatial geometric shape parameters, material property parameters, and facility layout parameters; In step S5, the linkage between the parametric design tool and the prediction model is achieved through an application programming interface (API), and the optimization process uses a multi-objective optimization algorithm to search for the Pareto optimal solution set.

[0010] According to another objective of the present invention, the present invention provides a healthy space design optimization system for implementing the above-described method, comprising: The data acquisition module is configured to simultaneously collect physiological data, behavioral data, subjective evaluation data, and physical environment data. The data fusion and processing module is communicatively connected to the data acquisition module and is used to perform time synchronization, spatial registration, cleaning and feature extraction on multi-source data, and output a structured fusion dataset. The data analysis and modeling module is communicatively connected to the data fusion and processing module. It is used to train a health efficacy prediction model or perform an impact mechanism analysis based on the fused dataset, and to perform a design parameter sensitivity analysis. The parametric design and optimization engine module is communicatively connected to the data analysis and modeling module. It is used to integrate the parametric design platform and the health performance prediction model, automatically adjust the design parameters according to the optimization objectives, perform health performance simulation, and output an optimization scheme set. The report generation and visualization module communicates with the parametric design and optimization engine module and is used to output the analysis results, optimization schemes and predictive performance in the form of visual charts and structured reports.

[0011] Furthermore, the data acquisition module includes at least two of the following: a wearable physiological signal acquisition device, an eye tracker, an indoor positioning base station and tag, an environmental sensor network, a digital questionnaire terminal, and a remote sensing sensor carried by a drone.

[0012] Furthermore, the parametric design and optimization engine module has built-in or can call various parametric modeling software and multi-objective optimization algorithm libraries; The report generation and visualization module can generate specific healthy space design guidelines for different regional climate conditions.

[0013] The beneficial effects of this invention are: This invention constructs a quantitative prediction model of "environmental parameters-health efficacy" by simultaneously collecting and integrating multimodal human factors data (physiological, behavioral, and subjective) with high-precision environmental data, thus realizing a paradigm shift in healthy space design from experience-driven to data-driven approaches. This method can accurately reveal the impact mechanisms of specific environmental factors on human condition and, relying on parametric tools and model linkage, achieve automatic iteration and evidence-based optimization of design schemes. The resulting design strategy report possesses objective quantitative evidence, significantly improving the scientific rigor, personalization, and health efficacy assurance of design interventions. It is particularly suitable for the challenges of healthy living in special environments such as high-altitude areas, effectively promoting the transformation of research and development achievements in the field of healthy building. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall process of the method of the present invention; Figure 2 This is a schematic diagram of the module composition of the system of the present invention. Detailed Implementation

[0015] The specific embodiments of the present invention will be further described below. It should be noted that these descriptions are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0016] Example 1 like Figure 1 As shown, a method for optimizing health space design based on multimodal human factors data includes the following steps: S1: In the target space or simulated space, simultaneously collect the subject's multimodal human factors data and corresponding physical environment data. The multimodal human factors data includes physiological data, behavioral data and subjective evaluation data. S2: Clean, align and fuse the spatiotemporally labeled synchronous data collected in step S1 to form a structured dataset, and extract key indicators reflecting human condition and environmental characteristics from it. S3: Based on machine learning or statistical methods, analyze the correlation between the key indicators and construct a prediction model or influencing mechanism model with spatial environment parameters as input and human health efficacy indicators as output. S4: Based on the model built in step S3, conduct sensitivity analysis of design parameters, identify key influencing factors, and set optimization goals in conjunction with health standards or project requirements; S5: Use parametric design tools in conjunction with the prediction model built in step S3 to automatically or semi-automatically adjust the design parameters, simulate and predict health performance indicators under different schemes, and select the set of optimized schemes that meet the optimization objectives. S6: Based on the optimized solution set, generate an evidence-based design report that includes specific design strategies, parameter recommendations, and expected health benefits.

[0017] Specifically, in step S1, the physiological data includes at least one of heart rate variability (HRV), electroencephalogram (EEG), skin conductance (GSR), blood oxygen saturation (SpO2), and electrocardiogram (ECG) collected by a portable physiological monitoring device.

[0018] In step S1, the behavioral data includes at least one of the following: visual attention distribution, spatial activity trajectory, action posture, dwell time, and distribution density, collected by an eye tracker, indoor positioning system, accelerometer, or computer vision analysis device.

[0019] In step S1, the subjective evaluation data includes at least one of the following: emotional state score based on a standardized scale, environmental restorative perception score, thermal comfort vote, visual comfort score, and satisfaction score, collected through a digital interactive terminal.

[0020] In step S1, the physical environment data includes at least one of the following collected by environmental sensors: air temperature, humidity, black sphere temperature, wind speed, wind direction, illuminance, color temperature, spectral power distribution, noise level, atmospheric pressure, carbon dioxide concentration, fine particulate matter concentration, and solar radiation intensity.

[0021] Step S1 is performed in an artificial climate chamber, which can reproduce the climate conditions of plateaus or special regions, including low oxygen, low air pressure, large temperature difference, and strong solar radiation.

[0022] Step S1 is performed in a real built environment, and large-scale, long-term data collection is carried out using drones, mobile monitoring equipment and fixed sensor networks.

[0023] Specifically, in step S2, data fusion uses a unified time server to timestamp multi-source data streams and aligns them with the high-frequency physiological signal or the clock of the environmental master control system.

[0024] Specifically, in step S3, the human health performance indicators include at least one of the following: cognitive task performance score, physiological stress / recovery index, emotional state index, subjective comfort achievement rate, and space utilization efficiency index.

[0025] In step S3, the algorithm used to construct the prediction model includes at least one of random forest, support vector machine, neural network, and geographically weighted regression.

[0026] Specifically, in step S4, the design parameters include at least one of the following: building or space lighting environment parameters, thermal environment parameters, acoustic environment parameters, air quality parameters, spatial geometric shape parameters, material property parameters, and facility layout parameters.

[0027] Specifically, in step S5, the linkage between the parametric design tool and the prediction model is achieved through an application programming interface, and the optimization process uses a multi-objective optimization algorithm to search for the Pareto optimal solution set.

[0028] Step S5 also includes: importing the optimized solution into the digital twin model for collaborative simulation and comprehensive evaluation of health performance, energy consumption, and cost.

[0029] like Figure 2 As shown, this embodiment also provides a system for implementing the above method, comprising: The data acquisition module is used to simultaneously collect physiological data, behavioral data, subjective evaluation data, and physical environment data; The data fusion and processing module is used to perform time synchronization, spatial registration, cleaning and feature extraction on multi-source data, and output a structured fusion dataset. The data analysis and modeling module is used to train health efficacy prediction models or perform impact mechanism analysis based on fused datasets, and to perform design parameter sensitivity analysis. The parametric design and optimization engine module integrates the parametric design platform and the health performance prediction model. It automatically adjusts the design parameters and performs health performance simulations based on the optimization objectives, and outputs a set of optimized solutions. The report generation and visualization module is used to output analysis results, optimization schemes, and predictive performance in the form of visual charts and structured reports.

[0030] Specifically, the data acquisition module includes at least two of the following: wearable physiological signal acquisition devices, eye trackers, indoor positioning base stations and tags, environmental sensor networks, digital questionnaire terminals, and remote sensing sensors carried by drones.

[0031] The data fusion and processing module also includes a functional sub-module for performing computer vision analysis on video streams to extract crowd behavior features.

[0032] The parametric design and optimization engine module has built-in or can call various parametric modeling software and multi-objective optimization algorithm libraries.

[0033] The report generation and visualization module can generate specialized health space design guidelines for different regional climate conditions such as plateau, hot and humid, and frigid areas.

[0034] Example 2 This embodiment presents a health space design optimization method based on multimodal human factors data, and optimizes the light-oxygen environment of a plateau healing space based on an artificial climate chamber. This embodiment is applied to the design optimization of healing environments in hotel rooms or health care wards in plateau areas.

[0035] like Figure 1 As shown, this embodiment includes the following steps: Step S1: Data Acquisition.

[0036] The artificial climate chamber recreates the low air pressure, low temperature, and strong ultraviolet radiation environment of high-altitude regions (such as 3500 meters above sea level). A simulated guest room space is set up inside.

[0037] Physiological data collection: Healthy subjects were recruited and wore portable physiological monitoring devices to continuously collect their blood oxygen saturation (SpO2), heart rate variability (HRV), electroencephalogram (EEG), and skin conductance (GSR).

[0038] Behavioral data collection: Subjects wore eye trackers to record their visual fixations and saccades during standardized tasks such as reading and resting. Simultaneously, an indoor UWB positioning system was used to record their movement trajectories.

[0039] Subjective data collection: After each environmental condition test, the subjects completed the digital Universal Emotional Scale (POMS) and Restorative Environment Scale (PRS) on a tablet computer, as well as a targeted questionnaire on light comfort and air freshness.

[0040] Environmental data acquisition: The artificial climate chamber control system records and outputs environmental parameters in real time, including illuminance, color temperature, spectrum, temperature, humidity, air pressure, CO2 concentration, and PM2.5 concentration. All data is timestamped to the millisecond level via a unified time server.

[0041] Step S2: Data fusion and feature extraction.

[0042] The data fusion platform receives all timestamped data streams. Using the EEG data timeline as a baseline, it aligns HRV time-domain / frequency-domain indicators (such as LF / HF), eye-tracking heatmaps, the ratio of sedentary to active time during positioning, and subjective scale scores with environmental parameters. Features are extracted, such as theta brainwave power, average fixation duration, subjective attention scores, and average illuminance and oxygen concentration during the cognitive task performed by subjects under specific color temperature (4000K) illumination.

[0043] Step S3: Model building.

[0044] Using light parameters (illuminance, color temperature) and air parameters (oxygen concentration, CO2 concentration) as core independent variables, and "cognitive task accuracy," "HRV recovery index," and "subjective recovery perception score" as dependent variables, a health performance prediction model was constructed using a random forest regression algorithm. The model results show that, in a simulated altitude environment of 3500 meters, moderately increasing local illuminance (to 500 lx) combined with a 24% oxygen-enriched environment has a significant synergistic effect on improving cognitive performance and promoting physiological recovery.

[0045] Steps S4 & S5: Optimization and Simulation.

[0046] The above model was integrated into parametric design software (such as Rhino + Grasshopper). The optimization objective was set as maximizing the "cognitive-recovery integrated score" while meeting basic energy consumption constraints. The optimization engine automatically adjusted the size and location of guest room windows (affecting natural light), the layout, illuminance, and color temperature of the artificial lighting system, and the location and air supply parameters of the fresh air system (affecting oxygen distribution). After multiple iterations, a set of guest room design schemes that optimally improve guests' cognitive and health recovery efficiency in a high-altitude, low-oxygen environment was generated.

[0047] Step S6: Report generation.

[0048] The system outputs the "Guidelines for Optimizing the Light-Oxygen Environment in High-Altitude Healing Rooms," which clearly states: It is recommended to set up an independently adjustable bedside reading oxygen-enriched lighting environment (500lx, 4000K, oxygen concentration 24%±2%); and to adopt low-illuminance, warm-color-temperature indirect lighting in the rest area to promote melatonin secretion, among other specific strategies, and includes model validation data.

[0049] Example 3 This embodiment presents a healthy space design optimization method based on multimodal human factors data. It optimizes the microclimate and behavior of public spaces in historical districts in humid and hot regions based on on-site monitoring. This embodiment is applied to the renewal projects of historical districts in humid and hot regions such as Xishuangbanna, Yunnan, aiming to improve the health, comfort and vitality of public spaces.

[0050] like Figure 1 As shown, this embodiment includes the following steps: Step S1: Data Acquisition.

[0051] The squares and alleyway nodes in the neighborhood were selected as the research area.

[0052] Environmental data collection: Deploy mobile monitoring vehicles and fixed weather stations to collect air temperature and humidity, black sphere temperature, wind speed and direction, and solar radiation intensity.

[0053] Behavioral data collection: Computer vision analysis is performed using installed security cameras (with privacy settings) to extract the distribution density, dwell time, and activity types (walking, standing, socializing) of people at different times and in different weather conditions.

[0054] Subjective data collection: On the survey day, intercept interviews were conducted with tourists and residents in the neighborhood, and thermal perception voting (TSV) and space satisfaction surveys were completed using tablet computers.

[0055] Spatial data acquisition: Using drone oblique photogrammetry, a 3D real-world model of the street is obtained, from which geometric and material parameters such as building height, street aspect ratio, underlying surface material, and green coverage are extracted.

[0056] Steps S2 & S3: Data fusion and model building.

[0057] Behavioral density data and thermal perception voting data were spatially gridded and matched with corresponding time- and location-specific microclimate data (such as PET physiological equivalent temperature). A geographically weighted regression (GWR) model was used to analyze the spatial heterogeneity of spatial morphological parameters (such as H / W ratio and sky visibility factor SVF) on local PET and crowd gathering heat. It was found that certain narrow and unshaded alleyways exhibited extremely high PET levels in the afternoon, leading to crowd avoidance and creating "thermal health depressions."

[0058] Steps S4 & S5: Optimization and Simulation.

[0059] The aforementioned GWR relationship model was incorporated into the digital twin model of the neighborhood. The optimization objective was set as follows: while preserving the historical character of the neighborhood, to minimize the average PET (Peak-to-Active) level in the summer afternoons and enhance behavioral activity along main paths. The optimization engine simulated various intervention measures: A. Installing traditional-style retractable awnings; B. Adding misting systems at specific nodes; C. Replacing some hard paving with permeable grass pavers; D. Optimizing street furniture layout to guide airflow. The simulation predicted the improvement effects of each option on PET and behavioral distribution.

[0060] Step S6: Report generation.

[0061] The system generated an "Optimization Plan for Enhancing the Thermal Health and Vitality of Public Spaces in Historic Districts," recommending a combined strategy of "C (paving renovation) + A (node ​​shading)," and providing precise suggestions on the location and size of shading installations. The plan predicts that this solution can reduce the average afternoon PET temperature in the target area by 3.5°C and increase the willingness of people to stay by 40%. This report provides quantitative decision support for protective renewal design.

[0062] This invention, by integrating multimodal objective human factors data with subjective feedback, provides robust empirical evidence for design that transcends traditional experience. The constructed "environment-human" response model reveals the impact paths and extent of key environmental factors on specific human health indicators, making design optimization more targeted. This invention achieves a rapid iterative closed loop from design scheme to performance prediction and parameter adjustment, significantly improving the efficiency and scientific rigor of design optimization. The methodological framework of this invention is universally applicable; by adjusting environmental parameters and human factors testing tasks, it can be applied to special environments such as high altitudes, extreme cold, and high temperatures, as well as various building types including offices, medical facilities, educational institutions, and residences. This invention directly produces design guidelines, optimization schemes, and predictive models, easily integrating with industrial processes such as design, healthcare, and product manufacturing, achieving efficient transformation of research results.

[0063] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing health space design based on multimodal human factors data, characterized in that, Includes the following steps: S1: In the target space or simulated space, simultaneously collect the subject's multimodal human factors data and corresponding physical environment data, wherein the multimodal human factors data includes physiological data, behavioral data and subjective evaluation data; S2: Clean, align and fuse the spatiotemporally labeled synchronous data collected in step S1 to form a structured dataset, and extract key indicators reflecting human condition and environmental characteristics from it. S3: Based on machine learning or statistical methods, analyze the correlation between the key indicators and construct a prediction model or influencing mechanism model with spatial environment parameters as input and human health efficacy indicators as output. S4: Based on the model built in step S3, conduct sensitivity analysis of design parameters, identify key influencing factors, and set optimization goals in conjunction with health standards or project requirements; S5: Use parametric design tools in conjunction with the prediction model built in step S3 to automatically or semi-automatically adjust the design parameters, simulate and predict health performance indicators under different schemes, and select the set of optimized schemes that meet the optimization objectives. S6: Based on the set of optimization schemes, generate an evidence-based design report that includes specific design strategies, parameter recommendations, and expected health benefits.

2. The method according to claim 1, characterized in that, In step S1, the physiological data includes at least one of heart rate variability, electroencephalogram, skin conductance, and blood oxygen saturation collected by a portable physiological monitoring device. The behavioral data includes at least one of visual attention distribution, spatial activity trajectory, dwell time and distribution density collected by eye trackers, indoor positioning systems, accelerometers or computer vision analysis devices. The physical environment data includes at least one of the following collected by environmental sensors: air temperature, humidity, illuminance, color temperature, wind speed, carbon dioxide concentration, atmospheric pressure, and solar radiation intensity.

3. The method according to claim 1 or 2, characterized in that, Step S1 is performed in an artificial climate chamber capable of reproducing the climatic conditions of plateaus or special regions, including low oxygen, low air pressure, large temperature differences, and strong solar radiation; and / or, Step S1 is performed in a real built environment, and large-scale, long-term data collection is carried out using drones, mobile monitoring equipment and fixed sensor networks.

4. The method according to claim 1, characterized in that, In step S3, the human health performance indicators include at least one of the following: cognitive task performance score, physiological stress / recovery index, emotional state index, and subjective comfort achievement rate. The algorithms used to construct the prediction model include at least one of random forest, support vector machine, neural network, and geographically weighted regression.

5. The method according to claim 1, characterized in that, In step S4, the design parameters include at least one of the following: building or space lighting environment parameters, thermal environment parameters, air quality parameters, spatial geometric shape parameters, material property parameters, and facility layout parameters; In step S5, the linkage between the parametric design tool and the prediction model is achieved through an application programming interface, and the optimization process uses a multi-objective optimization algorithm to search for the Pareto optimal solution set.

6. A healthy space design optimization system for implementing the method of any one of claims 1-5, characterized in that, include: The data acquisition module is configured to simultaneously collect physiological data, behavioral data, subjective evaluation data, and physical environment data. The data fusion and processing module is communicatively connected to the data acquisition module and is used to perform time synchronization, spatial registration, cleaning and feature extraction on multi-source data, and output a structured fusion dataset. The data analysis and modeling module is communicatively connected to the data fusion and processing module. It is used to train a health efficacy prediction model or perform an impact mechanism analysis based on the fused dataset, and to perform a design parameter sensitivity analysis. The parametric design and optimization engine module is communicatively connected to the data analysis and modeling module. It is used to integrate the parametric design platform and the health performance prediction model, automatically adjust the design parameters according to the optimization objectives, perform health performance simulation, and output an optimization scheme set. The report generation and visualization module communicates with the parametric design and optimization engine module and is used to output the analysis results, optimization schemes and predictive performance in the form of visual charts and structured reports.

7. The system according to claim 6, characterized in that, The data acquisition module includes at least two of the following: a wearable physiological signal acquisition device, an eye tracker, an indoor positioning base station and tag, an environmental sensor network, a digital questionnaire terminal, and a remote sensing sensor carried by a drone.

8. The system according to claim 6, characterized in that, The parametric design and optimization engine module has built-in or can call various parametric modeling software and multi-objective optimization algorithm libraries; The report generation and visualization module can generate specific healthy space design guidelines for different regional climate conditions.

Images