A Method and System for Human Comfort Modeling Based on Meteorological Data

Abstract

This invention discloses a method and system for human comfort modeling based on meteorological data, belonging to the field of human comfort data modeling technology. The method includes: integrated sensing of meteorological parameters, calculation of the dust stress index, stratification of meteorological health risks, and human comfort data modeling. By fusing and collecting multi-source meteorological data, optimizing quality, and co-enhancing spectral exposure, a comprehensive meteorological feature set for dust processes is constructed. A physical data fusion model is built based on an individual comfort baseline index, a dust breathing impedance penalty term, and a radiation anomaly correction term. Then, a health risk level classification is achieved through multi-level threshold rules. A multi-objective optimized comfort decision-making model is constructed, considering energy consumption, risk, and control stability. This invention can accurately express the degree of human discomfort in dusty environments and provide executable output of health risks, applicable to scenarios such as smart meteorological services, environmental health monitoring, and comfort-assisted decision-making.

Description

Technical Field

[0001] This invention relates to the field of human comfort data modeling technology, specifically to a method and system for human comfort modeling based on meteorological data. Background Technology

[0002] Human comfort modeling based on meteorological data refers to the process of collecting and analyzing meteorological factors such as temperature, humidity, wind speed, particulate matter concentration, solar radiation, and ultraviolet radiation intensity to establish a quantitative model reflecting the perceived load and degree of discomfort of the human body under different environmental conditions. This model is used to assess the human body's comfort state at a specific moment or in a specific environment. Its core function is to identify environmental characteristics that may cause respiratory discomfort, thermal and humidity stress, and sudden changes in light intensity by fusing multidimensional meteorological factors, providing a scientific basis for health risk warnings, outdoor activity recommendations, environmental regulation strategy formulation, and intelligent device control.

[0003] This invention further incorporates the characteristics of sandstorm scenarios by constructing a sandstorm stress index and risk level, enabling the model to provide more accurate and actionable comfort assessment results in complex environments such as sandstorms and dust storms.

[0004] However, existing human comfort modeling methods based on meteorological data generally only build models around thermal and humid comfort (temperature, humidity, wind speed), making it difficult to simultaneously express typical non-thermal and humid factors such as respiratory load, spectral attenuation, and particulate matter disturbance under sandstorm scenarios. This results in a technical problem where the model output deviates significantly from the subjective discomfort experience of the human body when a sandstorm occurs.

[0005] In the existing meteorological parameter integrated sensing process, there is a common problem that only parameter acquisition and conventional interpolation are performed, but the coupling relationship between particulate matter, solar radiation and ultraviolet radiation caused by dust in the spectral band attenuation is not modeled. This results in insufficient feature expression and model input distortion when there are sudden changes in concentration or rapid changes in spectral occlusion.

[0006] In the existing calculation of meteorological stress index, there are common problems such as simply adding up different load items (thermal and humid load caused by warm and humid wind, particulate matter breathing resistance, and radiation anomaly) or using empirical threshold methods, which lack differentiability, trainability, and physical consistency. This makes it impossible to effectively characterize the composite load contribution in dust storms and also makes it difficult to achieve regional adaptation or individual differentiated calibration.

[0007] In the existing human comfort data modeling process, there is a common problem that only a single comfort evaluation value is output, which cannot integrate multiple objectives such as dust stress risk, indoor regulation energy consumption and control command stability into the decision-making process. This leads to problems such as high energy consumption or frequent oscillations in regulation during actual control, making the project unusable. Summary of the Invention

[0008] To address the above issues and overcome the shortcomings of existing technologies, this invention provides a method and system for human comfort modeling based on meteorological data. The technical solution adopted by this invention is as follows: The human comfort modeling method based on meteorological data provided by this invention includes the following steps:

[0009] Step S1: Integrated sensing of meteorological parameters;

[0010] Step S2: Calculation of the dust stress index;

[0011] Step S3: Meteorological health risk stratification;

[0012] Step S4: Human comfort data modeling.

[0013] Further, in step S1, the meteorological parameter integrated sensing is used to collect multi-dimensional meteorological parameters synchronously and accurately. Specifically, it involves sensing and collecting data on particulate matter concentration, temperature, humidity, wind speed, solar radiation, and ultraviolet intensity in the environment, fusing the data, and constructing meteorological features based on spectral exposure synergistic enhancement to obtain a multi-dimensional meteorological feature dataset, including the following steps:

[0014] Step S11: Meteorological sensor acquisition, specifically, constructing a sensor array, collecting node data, performing linear interpolation and time synchronization on data such as particulate matter concentration, temperature, humidity, wind speed, solar radiation and ultraviolet intensity in the environment, to obtain raw meteorological sensor data;

[0015] Step S12: Data quality optimization, specifically, based on the original meteorological sensor data, by sequentially performing numerical threshold range checks, outlier removal, and outlier imputation, to obtain optimized meteorological data;

[0016] Step S13: Spatiotemporal data fusion, specifically, based on the quality-optimized meteorological data, the target spatial location coordinates for human comfort modeling are determined, and the spatial distance from each sensor node to the target spatial location coordinates is calculated. A spatiotemporal data fusion weight based on the spatial distance is constructed, and spatiotemporal data fusion is performed based on the spatiotemporal data fusion weight to obtain spatiotemporal fused meteorological data.

[0017] Step S14: Spectral exposure synergistic feature enhancement, specifically, based on the spatiotemporal fusion meteorological data, constructing meteorological features based on spectral exposure synergistic enhancement, performing meteorological feature enhancement, and obtaining a comprehensive meteorological feature set for dust data, including the following steps:

[0018] Step S141: Historical exposure feature construction, used to jointly model the instantaneous and accumulated values ​​of various meteorological features, specifically by calculating historical exposure features through exponential decay memory and sliding window calculation;

[0019] Step S142: Construction of spectral abrupt change features, used to jointly model the spectral attenuation and abrupt change effects of dust on solar radiation and ultraviolet radiation. Specifically, a clean reference spectral band is introduced, radiation attenuation ratio and ultraviolet attenuation ratio features are constructed respectively, and short-time abrupt change rate features are calculated to obtain spectral abrupt change features.

[0020] The clean reference spectrum is specifically based on historical meteorological sensor data. The time period in the annual data where the particulate matter concentration is below a set threshold is selected as the clean sample day, and feature modeling is performed based on the statistics of each time period of the clean sample day.

[0021] Step S143: Construction of dust disturbance features, used to jointly model the fluctuation characteristics of particulate matter concentration in a short period of time. Specifically, it involves calculating the set of statistical features characterizing the rapid dust disturbance behavior, including mean, standard deviation, skewness, kurtosis, and maximum rate of change, to obtain the dust disturbance features.

[0022] Step S144: Feature integration, specifically, based on the historical exposure features, the spectral abrupt change features, and the dust disturbance features, feature splicing is performed to obtain a comprehensive meteorological feature set;

[0023] Step S15: Construction of multidimensional meteorological feature vectors, specifically, based on the comprehensive meteorological feature set, feature normalization and feature concatenation operations based on Z-score are performed sequentially to obtain a multidimensional meteorological feature dataset.

[0024] Further, in step S2, the dust stress index calculation is used to quantify the comprehensive stress degree of dust and its accompanying meteorological conditions on the human body. Specifically, based on the multidimensional meteorological feature dataset, a weighted coupling model integrating the individual comfort baseline index, dust breathing impedance penalty, and radiation anomaly correction is adopted, and a dust stress index physical data fusion model is established to calculate the stress index and obtain dust stress index reference data, including the following steps:

[0025] Step S21: Construction of the basic index of individual comfort, which is used to quantitatively model the thermal and humidity comfort load caused by temperature, humidity and wind speed without considering sandstorms and radiation anomalies. Specifically, based on the temperature, relative humidity and wind speed characteristics in the multidimensional meteorological feature dataset, a quadratic deviation model centered on the comfort reference point is constructed to obtain the basic index of individual comfort.

[0026] Step S22: Modeling the dust breathing resistance penalty term, which is used to model the resistance and irritation load caused by dust and suspended particulate matter to the respiratory system. Specifically, based on the particulate matter concentration, historical exposure characteristics and dynamic statistical characteristics in the multidimensional meteorological feature dataset, a penalty model combining linear and nonlinear methods is constructed to calculate the dust breathing resistance penalty term.

[0027] Step S23: Radiation anomaly correction term modeling, used to model the abnormal changes in solar radiation and ultraviolet spectrum caused by dust, specifically, to construct the radiation anomaly correction term based on the abrupt change characteristics of the spectrum;

[0028] Step S24: Physical data fusion model construction, specifically, the individual comfort baseline index, dust breathing resistance penalty term and radiation anomaly correction term are physically constrained and weighted to construct a dust stress index physical data fusion model and obtain the initial dust stress index;

[0029] Step S25: Calculation of normalization of dust stress index, specifically, linear normalization of the initial dust stress index to obtain reference data of dust stress index; the reference data of dust stress index includes the normalized dust stress index and the unnormalized initial dust stress index.

[0030] Furthermore, in step S3, the meteorological health risk stratification is used to transform the abstract stress index into a concrete and actionable risk level and protection strategy. Specifically, based on the dust stress index reference data, the risk level is divided into a ladder-like level of concern, warning, severe and high risk by applying a preset multi-level risk judgment rule, and specific health warnings and protection recommendations are mapped to each level to obtain risk stratification reference data.

[0031] Further, in step S4, the human comfort data modeling is used to generate a comprehensive comfort decision model that integrates sandstorm stress. Specifically, based on the sandstorm stress index reference data and risk stratification reference data, a structured comfort assessment conclusion is obtained through integrated modeling, and a comfort decision model based on multi-objective optimization is driven to generate control commands and obtain the human comfort decision model.

[0032] The comfort decision modeling based on multi-objective optimization is specifically calculated based on minimizing the excessive load of sandstorm stress, minimizing energy consumption operating costs, and minimizing frequent fluctuations in control commands.

[0033] The human comfort modeling system based on meteorological data provided by this invention includes an integrated sensing module, an index calculation module, a risk stratification module, and a comfort modeling module.

[0034] The integrated sensing module is used for integrated sensing of meteorological parameters. Through integrated sensing of meteorological parameters, a multidimensional meteorological feature dataset is obtained, and the multidimensional meteorological feature dataset is sent to the index calculation module.

[0035] The index calculation module is used to calculate the dust stress index. Through the calculation of the dust stress index, reference data of the dust stress index is obtained, and the reference data of the dust stress index is sent to the risk stratification module and the comfort modeling module.

[0036] The risk stratification module is used for meteorological health risk stratification. Through meteorological health risk stratification, risk stratification reference data is obtained, and the risk stratification reference data is sent to the comfort modeling module.

[0037] The comfort modeling module is used for human comfort data modeling, and through human comfort data modeling, a human comfort decision model is obtained.

[0038] The beneficial effects achieved by the present invention using the above solution are as follows:

[0039] (1) In view of the technical problem that existing human comfort modeling methods based on meteorological data generally only build models around thermal and humid comfort (temperature, humidity, wind speed), making it difficult to simultaneously express typical non-thermal and humid factors such as respiratory load, spectral attenuation, and particulate matter disturbance under sandstorm scenarios, resulting in a serious deviation between the model output and the subjective discomfort experience of the human body when a sandstorm occurs, this solution creatively adopts a comprehensive comfort modeling method that combines special meteorological characteristics for sandstorm scenarios, sandstorm stress index calculation, and hierarchical risk modeling. It can unify factors such as particulate matter disturbance behavior, solar radiation / ultraviolet spectrum anomalies, and historical exposure cumulative effects into the comfort evaluation framework, realizing the leap from general comfort to sandstorm scenario-specific comfort; through the two-layer mapping of sandstorm stress index and risk level, the model's response during the sandstorm process is more in line with the real subjective experience of the human body, and generates actionable health protection recommendations;

[0040] (2) In the existing meteorological parameter integration sensing process, there is a common problem that only parameter acquisition and conventional interpolation are performed, but the coupling relationship of spectral attenuation caused by dust between particulate matter, solar radiation and ultraviolet radiation is not modeled. This results in insufficient feature expression and model input distortion when there are sudden changes in concentration or rapid changes in spectral occlusion. This solution creatively adopts a meteorological feature construction method based on spectral exposure synergistic enhancement. It jointly models instantaneous exposure, historical exposure, spectral attenuation ratio, short-term mutation rate and particulate matter disturbance features to realize the expression of the dual effects of continuous exposure and instantaneous mutation in the dust process. The multidimensional features obtained in this way not only enhance the model's ability to distinguish typical processes such as dust intrusion and dust tail attenuation, but also significantly improve the stability and trend tracking ability of stress index calculation.

[0041] (3) In the existing meteorological stress index calculation process, there is a common problem that different load items (heat and humidity load caused by warm and humid wind, particulate breathing resistance, and radiation anomaly) are simply added together or empirical threshold methods are used. This lacks differentiability, trainability, and physical consistency, which makes it impossible to effectively characterize the composite load contribution in dust storms and also makes it difficult to achieve regional adaptation or individual differentiated calibration. This solution creatively adopts a weighted coupling model that integrates the individual comfort baseline index, dust breathing resistance penalty, and radiation anomaly correction. It also establishes a physical data fusion model for dust stress index to calculate the stress index. By introducing trainable weights and monotonic physical constraints into the model at the same time, the stress index maintains a stable physical directionality under different load conditions (the higher the load, the stronger the discomfort). At the same time, it has the ability to automatically adapt to different regions / climate backgrounds, which significantly improves the generalization performance and interpretability of the stress index.

[0042] (4) In the existing human comfort data modeling process, there is a common problem that only a single comfort evaluation value is output, which cannot integrate multiple objectives such as dust stress risk, indoor regulation energy consumption and control command stability into the decision-making process. This leads to problems such as high energy consumption or frequent oscillations in regulation during actual control, making the project unusable. This solution creatively adopts comfort decision modeling based on multi-objective optimization. It takes "reducing the probability of exceeding the stress index, reducing energy consumption cost, and reducing regulation command fluctuations" as joint optimization objectives, constructs a mathematically reproducible multi-objective optimization framework, and provides two implementation methods based on traditional optimization algorithms and reinforcement learning to realize automatic generation of control strategies. Thus, in practical applications, it simultaneously meets the engineering requirements of comfort steady-state, energy consumption economy and system stability, and greatly improves the feasibility of the overall decision model. Attached Figure Description

[0043] Figure 1 A flowchart illustrating the human comfort modeling method based on meteorological data provided by this invention;

[0044] Figure 2 A schematic diagram of the human comfort modeling system based on meteorological data provided by the present invention;

[0045] Figure 3 This is a flowchart illustrating the process of integrating meteorological parameters in step S1.

[0046] Figure 4 This is a flowchart illustrating the process of enhancing the spectral band exposure cooperative features in step S14.

[0047] Figure 5 This is a flowchart illustrating the process of calculating the dust stress index in step S2.

[0048] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0050] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0051] Example 1, see Figure 1 The present invention provides a method for human comfort modeling based on meteorological data, which includes the following steps:

[0052] Step S1: Integrated sensing of meteorological parameters;

[0053] Step S2: Calculation of the dust stress index;

[0054] Step S3: Meteorological health risk stratification;

[0055] Step S4: Human comfort data modeling.

[0056] By performing the above operations, this solution addresses the technical problem in existing meteorological data-based human comfort modeling methods, which generally only focus on thermal and humid comfort (temperature, humidity, wind speed) and fail to simultaneously represent typical non-thermal and humid factors such as respiratory load, spectral attenuation, and particulate matter disturbance under dust storm scenarios. This leads to a significant deviation between the model output and the subjective discomfort experienced by humans during dust storms. For example, in scenarios with low wind speeds but high PM10 fluctuations, traditional models would still classify it as "mild discomfort," while in reality, humans experience significant respiratory irritation and blurred vision. This solution creatively adopts a comprehensive comfort modeling method that combines specific meteorological characteristics for dust storm scenarios, dust stress index calculation, and hierarchical risk modeling. It can uniformly incorporate factors such as particulate matter disturbance behavior, solar radiation / UV spectrum anomalies, and the cumulative effect of historical exposure into the comfort evaluation framework, achieving a leap from general comfort to dust storm-specific comfort. Through a two-layer mapping between the dust stress index and risk level, the model's response during dust storms is more consistent with the actual subjective experience of humans, and it generates actionable health protection recommendations.

[0057] Example 2, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S1, the meteorological parameter integrated sensing is used to collect multi-dimensional meteorological parameters synchronously and accurately. Specifically, it involves sensing data collection, data fusion, and meteorological feature construction based on spectral exposure synergistic enhancement of particulate matter concentration, temperature, humidity, wind speed, solar radiation, and ultraviolet intensity data in the environment to obtain a multi-dimensional meteorological feature dataset, including the following steps:

[0058] Step S11: Meteorological sensor acquisition, specifically, constructing a sensor array, collecting node data, performing linear interpolation and time synchronization on data such as particulate matter concentration, temperature, humidity, wind speed, solar radiation and ultraviolet intensity in the environment, to obtain raw meteorological sensor data;

[0059] The calculation formula for the raw meteorological sensor data is as follows:

[0060] ;

[0061] In the formula, z k (s j ,t n ) is the corresponding sampling node s j and sampling time t n The original meteorological sensor data is given by: j is the sensor node index, n is the acquisition time index, interp is the interpolation function used to perform linear interpolation and imputation on the original meteorological sensor data, and z... k (s j,t) is the raw, uninterpolated sequence of node data acquisition, t is the total time of the sequence, and k is the index of the raw data type of meteorological sensor, used to represent the specific meteorological parameter type, including PM2.5, PM10, temperature T, relative humidity RH, wind speed WS, solar radiation intensity SR, and ultraviolet radiation intensity UV.

[0062] Step S12: Data quality optimization, specifically, based on the original meteorological sensor data, by sequentially performing numerical threshold range checks, outlier removal, and outlier imputation, to obtain optimized meteorological data;

[0063] Preferably, the threshold range check is used to impose physical reasonable range constraints on the data variables in the raw meteorological sensor data, and the calculation formula is as follows:

[0064] ;

[0065] In the formula, This is a threshold range check to optimize data, z k (s j ,t n ) is the corresponding sampling node s j and sampling time t n Raw meteorological sensor data, It is the physically reasonable minimum value of meteorological parameters. It is the physically reasonable maximum value of the meteorological parameters;

[0066] The outlier removal specifically employs outlier imputation based on Z-scores; the outlier imputation specifically employs the nearest neighbor interpolation method for outlier imputation.

[0067] Step S13: Spatiotemporal data fusion, specifically, based on the quality-optimized meteorological data, the target spatial location coordinates for human comfort modeling are determined, and the spatial distance from each sensor node to the target spatial location coordinates is calculated. A spatiotemporal data fusion weight based on the spatial distance is constructed, and spatiotemporal data fusion is performed based on the spatiotemporal data fusion weight to obtain spatiotemporal fused meteorological data.

[0068] The calculation formula for the spatiotemporal fusion meteorological data is as follows:

[0069] ;

[0070] In the formula, x k (t n () represents spatiotemporal fusion meteorological data, used to represent the fusion of similar meteorological quantities from multiple sensor nodes into a single representative value. J is the total number of sensor nodes, and j is the sensor node index. It is a weighting system for spatiotemporal data fusion based on spatial distance. It is to optimize meteorological data in terms of quality;

[0071] Step S14: Enhancement of spectral exposure synergy features, specifically, based on the spatiotemporal fusion meteorological data, construct meteorological features based on spectral exposure synergy enhancement, enhance meteorological features, and obtain a comprehensive meteorological feature set for dust data;

[0072] Step S15: Construction of multidimensional meteorological feature vectors, specifically, based on the comprehensive meteorological feature set, feature normalization and feature concatenation operations based on Z-score are performed sequentially to obtain a multidimensional meteorological feature dataset;

[0073] Preferably, the multidimensional meteorological feature dataset includes at least: instantaneous sensor observation features of particulate matter concentration, temperature, humidity, wind speed, solar radiation and ultraviolet intensity; historical exposure features calculated based on exponential decay memory and sliding window; radiation attenuation ratio and ultraviolet attenuation ratio features constructed based on clean reference spectrum; and dynamic statistical features of PM10 and PM2.5 that characterize the rapid disturbance behavior of dust storms, including fluctuation, skewness, kurtosis and maximum rate of change.

[0074] By performing the above operations, this solution addresses the technical problem in existing meteorological parameter integrated sensing processes that generally only perform parameter acquisition and conventional interpolation, failing to model the spectral attenuation coupling relationship between particulate matter, solar radiation, and ultraviolet radiation caused by dust storms. This leads to insufficient feature representation and distorted model input when there are sudden changes in concentration or rapid changes in spectral obstruction. For example, in a dust storm scenario, when PM10 suddenly increases from 80 μg / m³ to 600 μg / m³, the radiation attenuation is not linear but shows a significant "instantaneous drop." Traditional methods cannot reflect this spectral abrupt change. This solution creatively adopts a meteorological feature construction method based on synergistic enhancement of spectral exposure, jointly modeling instantaneous exposure, historical exposure, spectral attenuation ratio, short-term abrupt change rate, and particulate matter disturbance characteristics to express the dual effects of continuous exposure and instantaneous abrupt changes in the dust storm process. The resulting multidimensional features not only enhance the model's ability to distinguish typical processes such as dust storm intrusion and dust tail attenuation but also significantly improve the stability and trend tracking ability of stress index calculation.

[0075] Example 3, see Figure 1 , Figure 2 and Figure 4 This embodiment is based on the above embodiment. Step S14 specifically includes the following steps:

[0076] Step S141: Historical exposure feature construction, used to jointly model the instantaneous and accumulated values ​​of various meteorological features, specifically by calculating historical exposure features through exponential decay memory and sliding window calculation;

[0077] The formula for calculating the historical exposure characteristics is as follows:

[0078] ;

[0079] In the formula, HIS(t) n This is a characteristic of historical exposure. It is a cumulative exposure characteristic calculated based on exponential decay memory. It is based on the characteristics of short-term continuous exposure effects calculated using a sliding window.

[0080] The specific calculation formula for the cumulative exposure characteristic based on exponential decay memory calculation is as follows:

[0081] ;

[0082] In the formula, It is the attenuation coefficient, used to control the effective memory length of exposure, with a value range of (0,1], x k (t n This is spatiotemporal fusion meteorological data. It is a feature of the cumulative exposure amount at the previous moment;

[0083] The specific calculation formula for the short-term continuous exposure effect characteristics based on the sliding window calculation is as follows:

[0084] ;

[0085] In the formula, L w Where is the total length of the time window, and 'i' is the time index within the time window. yes Spatiotemporal fusion meteorological data at any given moment;

[0086] In this embodiment, exposure features refer to cumulative indicators used to characterize the potential impact of meteorological factors on the human body over a time scale. The core meaning is to simultaneously reflect the instantaneous state of meteorological variables and the continuous exposure intensity over past periods. This can more realistically reflect the continuous exposure state of the human body under complex dust and meteorological conditions, providing more stable and time-trend expressive feature inputs for subsequent spectral mutation modeling and dust stress index calculation.

[0087] Step S142: Construction of spectral abrupt change features, used to jointly model the spectral attenuation and abrupt change effects of dust on solar radiation and ultraviolet radiation. Specifically, a clean reference spectral band is introduced, radiation attenuation ratio and ultraviolet attenuation ratio features are constructed respectively, and short-time abrupt change rate features are calculated to obtain spectral abrupt change features.

[0088] The clean reference spectrum is specifically based on historical meteorological sensor data. The time period in the annual data where the particulate matter concentration is below a set threshold is selected as the clean sample day, and feature modeling is performed based on the statistics of each time period of the clean sample day.

[0089] Preferably, the set threshold is specifically PM2.5 < 35 μg / m³ and PM10 < 50 μg / m³;

[0090] The calculation formula for feature modeling based on daily time-period statistics of clean samples is as follows:

[0091] ;

[0092] In the formula, SR clear (t n ) is a characteristic of the clean reference spectrum of solar radiation, N c SR is the total number of clean sample days, d is the clean sample day index, and SR is the total number of clean sample days. d (t n ) is the clean sample day d at t n Radiation value at any time, UV clear (t n () is a characteristic of the ultraviolet cleaning reference spectrum, UV d (t n ) is the clean sample day d at t n Ultraviolet intensity value at any given time;

[0093] The formula for calculating the abrupt change characteristics of the spectral band is:

[0094] ;

[0095] In the formula, R(t) n ) is a spectral mutation feature, R SR (t n R is the characteristic of solar radiation attenuation ratio. UV (t n This is the characteristic of ultraviolet attenuation ratio. It is a characteristic of the short-term abrupt change rate of solar radiation. It is a characteristic of short-term mutation rate under ultraviolet light;

[0096] Preferably, the formula for calculating the solar radiation attenuation ratio characteristic is:

[0097] ;

[0098] In the formula, x SR (t n () represents the corresponding solar radiation intensity in the spatiotemporal fusion meteorological data;

[0099] The formula for calculating the ultraviolet attenuation ratio characteristic is as follows:

[0100] ;

[0101] In the formula, x UV (t n () represents the corresponding ultraviolet radiation intensity in the spatiotemporal fusion meteorological data;

[0102] The formula for calculating the short-term abrupt change rate characteristic of solar radiation is as follows:

[0103] ;

[0104] In the formula, x SR (t n-1 () represents the solar radiation intensity corresponding to the spatiotemporal fusion meteorological data of the previous moment. It is a time interval;

[0105] The formula for calculating the ultraviolet short-term mutation rate characteristic is as follows:

[0106] ;

[0107] In the formula, x UV (t n-1 () represents the ultraviolet intensity in the spatiotemporal fusion meteorological data of the previous moment;

[0108] Step S143: Construction of dust disturbance features, used to jointly model the fluctuation characteristics of particulate matter concentration in a short period of time. Specifically, it involves calculating the set of statistical features characterizing the rapid dust disturbance behavior, including mean, standard deviation, skewness, kurtosis, and maximum rate of change, to obtain the dust disturbance features.

[0109] The formula for calculating the characteristics of the sand and dust disturbance is as follows:

[0110] ;

[0111] In the formula, DisT(t) n This is a characteristic of sand and dust disturbance. This is a characteristic of the PM10 mean. It is a characteristic of PM10 standard deviation. It is a characteristic of PM10 skewness. It is a characteristic of PM10 peak intensity. This is the characteristic of the maximum rate of change of PM10;

[0112] In the specific implementation step S143, the mean characteristic, standard deviation characteristic, skewness characteristic, kurtosis characteristic, and maximum rate of change characteristic can be calculated based on conventional meteorological data using existing statistical methods. The calculation method is consistent with the time series fluctuation analysis method commonly used in the field of mathematical statistics. In this embodiment, the above statistical characteristics are applied to characterize the disturbance behavior of PM10 in a short time scale to reflect the amplitude, directionality, and degree of anomaly of rapid changes in dust. The basic calculation method of such statistical characteristics is not modified or redefined.

[0113] Preferably, to further enhance the ability to characterize the disturbance behavior of fine particulate matter, this embodiment also performs the same statistical feature calculation on the short-term changes of PM2.5 to obtain the mean, standard deviation, skewness, kurtosis and maximum rate of change characteristics of PM2.5. By jointly modeling the disturbance characteristics of PM10 and PM2.5, the synergistic disturbance effect of coarse and fine particulate matter can be expressed simultaneously in dust events, thereby enhancing the ability to distinguish the comprehensive disturbance characteristics.

[0114] Step S144: Feature integration, specifically, based on the historical exposure features, the spectral abrupt change features, and the dust disturbance features, feature splicing is performed to obtain a comprehensive meteorological feature set;

[0115] The formula for calculating the comprehensive meteorological feature set is as follows:

[0116] F total ={HIS(t n ),R(t n ),DisT(t n )};

[0117] In the formula, F total It is a comprehensive set of meteorological characteristics, HIS(t) n ) is a historical exposure characteristic, R(t) n DisT(t) is a spectral mutation feature. n This is a characteristic of sand and dust disturbance.

[0118] Example 4, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S2, the dust stress index is calculated to quantify the comprehensive stress degree of dust and its accompanying meteorological conditions on the human body. Specifically, based on the multidimensional meteorological feature dataset, a weighted coupling model integrating the individual comfort baseline index, dust breathing impedance penalty, and radiation anomaly correction is adopted, and a dust stress index physical data fusion model is established to calculate the stress index and obtain dust stress index reference data. The steps include:

[0119] Step S21: Construction of the basic index of individual comfort, which is used to quantitatively model the thermal and humidity comfort load caused by temperature, humidity and wind speed without considering sandstorms and radiation anomalies. Specifically, based on the temperature, relative humidity and wind speed characteristics in the multidimensional meteorological feature dataset, a quadratic deviation model centered on the comfort reference point is constructed to obtain the basic index of individual comfort.

[0120] The formula for calculating the individual comfort baseline index is as follows:

[0121] ;

[0122] In the formula, CI base (t n ) is the basic index of individual comfort, p1 is the temperature weight, and T(t) is the temperature weight. n ) represents the temperature characteristic, T0 is the comfort reference temperature, with a preferred value of 22℃, p2 is the humidity weight, and RH(t) is the humidity weight. n ) represents the humidity characteristic, RH0 is the comfort reference humidity, with a preferred value range of [45%, 55%], p3 is the wind speed weight, and WS(t) is the humidity characteristic. n ) represents the wind speed characteristic, and WS0 is the minimum comfortable wind speed threshold, with a preferred value of 1m / s;

[0123] As a further optimization of this embodiment, the quadratic deviation model centered on the comfort reference point is not a simple threshold judgment, but rather a nonlinear deviation function that constructs the comfort reference point of temperature, humidity and wind speed as the global optimal point, so that the comfort index reaches a minimum value at the reference point and increases quadratically with the degree of deviation, thereby realizing a continuous, differentiable and trainable quantitative modeling of the intensity of human body heat and humidity load.

[0124] Step S22: Modeling the dust breathing resistance penalty term, which is used to model the resistance and irritation load caused by dust and suspended particulate matter to the respiratory system. Specifically, based on the particulate matter concentration, historical exposure characteristics and dynamic statistical characteristics in the multidimensional meteorological feature dataset, a penalty model combining linear and nonlinear methods is constructed to calculate the dust breathing resistance penalty term.

[0125] The formula for calculating the dust breathing resistance penalty term is as follows:

[0126] ;

[0127] In the formula, P dust (t n ) is the dust respiration resistance penalty term, q1 is the PM10 concentration weight, and PM10(t) is the PM10 concentration weight. n ) represents the instantaneous concentration characteristic of PM10, q2 represents the concentration weight of PM2.5, and PM2.5(t) represents the instantaneous concentration characteristic of PM10. n ) represents the instantaneous concentration characteristics of PM2.5, q3 represents the weights of historical exposure characteristics of PM10, and HIS PM10 (t n ) represents historical exposure characteristics of PM10, q4 represents the weights of historical exposure characteristics of PM2.5, and HIS PM2.5 (t n ) represents historical exposure characteristics of PM2.5, and q5 represents the weight of the maximum rate of change of PM10. This is the characteristic of the maximum rate of change of PM10;

[0128] Step S23: Radiation anomaly correction term modeling, used to model the abnormal changes in solar radiation and ultraviolet spectrum caused by dust, specifically, to construct the radiation anomaly correction term based on the abrupt change characteristics of the spectrum;

[0129] The formula for calculating the radiation anomaly correction term is as follows:

[0130] ;

[0131] In the formula, C rad (t n ) is the radiation anomaly correction term, r1 is the weight of the solar radiation attenuation ratio, (·) + It is a non-negative function, and its calculation formula is (x). + =max(0,x), r2 is the weight of ultraviolet attenuation ratio, r3 is the weight of short-term solar radiation mutation rate, |·| is the modulo operator, and r4 is the weight of ultraviolet short-term mutation rate.

[0132] Step S24: Physical data fusion model construction, specifically, the individual comfort baseline index, dust breathing resistance penalty term and radiation anomaly correction term are physically constrained and weighted to construct a dust stress index physical data fusion model and obtain the initial dust stress index;

[0133] The formula for calculating the initial dust stress index is as follows:

[0134] ;

[0135] In the formula, DSI raw (t n w1 is the initial dust stress index, w2 is the physical constraint weight of the individual comfort baseline index, w3 is the physical constraint weight of the dust breathing resistance penalty term, and w4 is the physical constraint weight of the radiation anomaly correction term.

[0136] Preferably, to ensure that the dust stress index maintains a consistent trend with the actual degree of human discomfort, physical constraints are introduced during the calculation of the weighting parameters, making the dust stress index monotonically constant with respect to various load terms. The calculation formula is as follows:

[0137] ;

[0138] As a further optimization of this embodiment, all weight parameters in the individual comfort baseline index, dust breathing impedance penalty term, and radiation anomaly correction term mentioned in steps S21 to S24 are trainable. By pre-setting a historical sample dataset and combining least squares regression, gradient descent optimization, or other differentiable optimization methods, it is possible to jointly calibrate all the above weight parameters.

[0139] By introducing trainable weight parameters, the dust stress index model described in this embodiment can be adaptively updated based on actual meteorological exposure data of different regions, seasons and populations, thereby significantly improving the fitting accuracy and generalization ability of the index calculation results.

[0140] Step S25: Calculation of normalization of dust stress index, specifically, linear normalization of the initial dust stress index to obtain reference data of dust stress index; the reference data of dust stress index includes the normalized dust stress index and the unnormalized initial dust stress index.

[0141] The formula for calculating the linear normalization is:

[0142] ;

[0143] In the formula, DSI(t) n The normalized dust stress index (DSI) is the normalized dust stress index. min It is the minimum value of the stress index, DSI max It is the maximum value of the stress index.

[0144] By performing the above operations, this approach addresses the technical problems in existing meteorological stress index calculations, which commonly involve simply adding different load items (thermal and humid load caused by warm and humid winds, particulate matter breathing resistance, and radiation anomalies) or using empirical threshold methods. These methods lack differentiability, trainability, and physical consistency, resulting in an inability to effectively characterize the composite load contribution in dust storms and making it difficult to achieve regional adaptation or individual-differentiated calibration. For example, under the same PM10 concentration, the superposition effect of radiation anomalies and respiratory stimulation is often ignored in traditional models. This solution creatively adopts a weighted coupling model that integrates the individual comfort baseline index, dust breathing resistance penalty, and radiation anomaly correction. It also establishes a physical data fusion model for dust stress index calculation. By simultaneously introducing trainable weights and monotonic physical constraints into the model, the stress index maintains a stable physical directionality under different load conditions (the higher the load, the stronger the discomfort). It also has the ability to automatically adapt to different regional / climate backgrounds, significantly improving the generalization performance and interpretability of the stress index.

[0145] Example 5, see Figure 1 , Figure 2 This embodiment is based on the above embodiment. In step S3, the meteorological health risk stratification is used to transform the abstract stress index into a concrete and actionable risk level and protection strategy. Specifically, based on the dust stress index reference data, the risk level is divided into a ladder-like level of concern, warning, severe and high risk by applying a preset multi-level risk judgment rule, and specific health warnings and protection suggestions are mapped to each level to obtain risk stratification reference data.

[0146] Preferably, this embodiment adopts a linear threshold division strategy, which does not rely on individual privacy data of the population, nor does it require the input of sensitive individual information, and can achieve a highly interpretable risk classification mechanism while maintaining the transparency of the model; preferably, the linear threshold can be adaptively adjusted according to the historical meteorological statistical distribution of each region to improve the regional adaptability of risk level division.

[0147] To achieve the above objectives, Table 1 is a reference classification table for the multi-level risk assessment rules. As shown in the table, this embodiment presets four assessment intervals based on the normalized distribution characteristics of the stress index in the [0,100] interval. When the dust stress index is low, it corresponds to a lighter health risk level, and when the stress index increases, it corresponds to a higher level of health protection requirements.

[0148] Table 1. Reference classification table for multi-level risk assessment rules;

[0149]

[0150] This embodiment completes the risk level mapping by referring to a table. Each level corresponds to a clear description of the health impact and corresponding protective measures. This classification method is flexible and can be adjusted according to the historical meteorological distribution and health statistics of each region, thereby enhancing the adaptability of the risk level classification to the actual characteristics of dust storm events.

[0151] In this embodiment, the risk levels can be stratified through a mapping function. The mapping method ensures the certainty and interpretability of the risk level determination, while maintaining a correspondence with the normalized dust stress index calculated in step S2.

[0152] By performing the above operations, this approach addresses the technical problems in existing meteorological stress index calculations, which commonly involve simply adding different load items (thermal and humid load caused by warm and humid winds, particulate matter breathing resistance, and radiation anomalies) or using empirical threshold methods. These methods lack differentiability, trainability, and physical consistency, resulting in an inability to effectively characterize the composite load contribution in dust storms and making it difficult to achieve regional adaptation or individual-differentiated calibration. For example, under the same PM10 concentration, the superposition effect of radiation anomalies and respiratory stimulation is often ignored in traditional models. This solution creatively adopts a weighted coupling model that integrates the individual comfort baseline index, dust breathing resistance penalty, and radiation anomaly correction. It also establishes a physical data fusion model for dust stress index calculation. By simultaneously introducing trainable weights and monotonic physical constraints into the model, the stress index maintains a stable physical directionality under different load conditions (the higher the load, the stronger the discomfort). It also has the ability to automatically adapt to different regional / climate backgrounds, significantly improving the generalization performance and interpretability of the stress index.

[0153] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S4, the human comfort data modeling is used to generate a comprehensive comfort decision model that integrates sandstorm stress. Specifically, based on the sandstorm stress index reference data and risk stratification reference data, a structured comfort assessment conclusion is obtained by integrating and modeling, and a comfort decision model based on multi-objective optimization is driven to generate control commands and obtain the human comfort decision model.

[0154] The comfort decision modeling based on multi-objective optimization is specifically calculated based on minimizing the excessive load of sandstorm stress, minimizing energy consumption operating costs, and minimizing the frequent fluctuation of control commands.

[0155] In this embodiment, the indoor environment regulation system is used as the control object of the human comfort decision model, and the control variables include at least the following parameters: the operating power or air volume control quantity of the air purification equipment, the air exchange intensity control quantity of the fresh air system, and the indoor circulation and external circulation adjustment coefficient.

[0156] Preferably, the calculation formula for the multi-objective optimization function specifically constructed in the comfort decision modeling based on multi-objective optimization is as follows:

[0157] ;

[0158] In the formula, U(t) n ) is the control variable, J(t) n ) is a multi-objective optimization function. It is the risk weight of dust stress, J risk (t n This is a penalty item for the risk of dust storms. It is the energy consumption weight, J energy (t n ) is an energy consumption penalty item. It is the weight for controlling stationarity, J smooth (t n ) is a penalty term for controlling stability;

[0159] The excessive dust stress load is used as a penalty item for dust stress risk, and the calculation formula is as follows:

[0160] ;

[0161] In the formula, DSI(t) n The normalized dust stress index (DSI) is the normalized dust stress index. th This is the risk trigger threshold, with a preferred value range of [40, 60].

[0162] The energy consumption operating cost is used as an energy consumption penalty item, and the calculation formula is as follows:

[0163] ;

[0164] In the formula, It is the energy consumption weight of air purification equipment, u1(t) n This refers to the operating intensity control quantity of air purification equipment (such as air purifiers, filter fans, or integrated purification modules). It is the energy consumption weight of the fresh air system equipment, u2(t) n This refers to the air exchange intensity control value of the fresh air system. It is the weight of energy consumption adjustment between internal and external circulation, u3(t) n ) is the indoor circulation and outdoor circulation adjustment coefficient; among which, each energy consumption weight can be calibrated based on the proportion of rated power or based on the marginal energy consumption per unit control quantity;

[0165] The frequent fluctuation term of the control command is used as a control stability penalty term, and its calculation formula is as follows:

[0166] ;

[0167] In the formula, U(t) n-1 ) represents the control variable from the previous moment, and ||·||2 is the L2 norm operator;

[0168] As a further optimization of this embodiment, the weights of each objective in the multi-objective optimization function satisfy the following constraints:

[0169] ;

[0170] In the formula, The preferred value range is [0.4, 0.6], which is used to indicate prioritizing the protection of human health and dust storms. The preferred value range is [0.2, 0.4], which is used to balance the energy consumption of system operation. The preferred value range is [0.1, 0.2], used to control the stability of the regulation process; when the dust storm risk level increases, it can be dynamically increased. The proportion of values ​​is used to enhance protection priority;

[0171] In a preferred embodiment, the multi-objective optimization problem is transformed into a single-objective optimization problem through a weighted objective function, and solved using gradient descent, sequential quadratic programming, or other constrained optimization algorithms to obtain the optimal control vector;

[0172] In another embodiment, the human comfort decision modeling problem can be transformed into a Markov decision process, and the strategy can be trained by reinforcement learning algorithm to achieve adaptive optimization of the control strategy. The state space includes a normalized dust stress index, risk level and current control state, the action space is the control variables, and the reward function is constructed based on the multi-objective optimization function to guide the strategy to tend to achieve the minimum solution of the multi-objective optimization function.

[0173] By performing the above operations, this solution addresses the common problem in existing human comfort data modeling processes that only output a single comfort evaluation value, failing to integrate multiple objectives such as dust stress risk, indoor regulation energy consumption, and control command stability into the decision-making process. This leads to engineering unavailability issues in actual regulation, such as high energy consumption or frequent regulation fluctuations. For example, frequent start-stop cycles of air purification systems can shorten equipment lifespan and degrade user experience. This solution creatively adopts comfort decision modeling based on multi-objective optimization, taking "reducing the probability of exceeding the stress index, reducing energy consumption costs, and reducing regulation command fluctuations" as joint optimization objectives. It constructs a mathematically reproducible multi-objective optimization framework and provides two implementation methods based on traditional optimization algorithms and reinforcement learning to automatically generate control strategies. Thus, in practical applications, it simultaneously meets the engineering requirements of comfort steady-state performance, energy economy, and system stability, significantly improving the feasibility of the overall decision model.

[0174] Example 7, see Figure 1 and Figure 2 This embodiment is based on the above embodiments. The human comfort modeling system based on meteorological data provided by the present invention includes an integrated sensing module, an index calculation module, a risk stratification module and a comfort modeling module.

[0175] The integrated sensing module is used for integrated sensing of meteorological parameters. Through integrated sensing of meteorological parameters, a multidimensional meteorological feature dataset is obtained, and the multidimensional meteorological feature dataset is sent to the index calculation module.

[0176] The index calculation module is used to calculate the dust stress index. Through the calculation of the dust stress index, reference data of the dust stress index is obtained, and the reference data of the dust stress index is sent to the risk stratification module and the comfort modeling module.

[0177] The risk stratification module is used for meteorological health risk stratification. Through meteorological health risk stratification, risk stratification reference data is obtained, and the risk stratification reference data is sent to the comfort modeling module.

[0178] The comfort modeling module is used for human comfort data modeling, and through human comfort data modeling, a human comfort decision model is obtained.

[0179] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0180] Although embodiments of the 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 invention.

[0181] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for modeling human comfort based on meteorological data, characterized in that: The method includes the following steps: Step S1: Meteorological parameter integration sensing performs sensor data acquisition, data fusion, and meteorological feature construction based on spectral exposure synergistic enhancement to obtain a multidimensional meteorological feature dataset; the meteorological feature construction based on spectral exposure synergistic enhancement includes the following steps: historical exposure feature construction, spectral abrupt change feature construction, dust disturbance feature construction, and feature integration; The construction of historical exposure features involves joint feature modeling of the instantaneous and accumulated values ​​of various meteorological features. Specifically, historical exposure features are obtained through exponential decay memory and sliding window calculation. The construction of the spectral abrupt change features involves joint feature modeling of the spectral attenuation and abrupt change effects of dust on solar radiation and ultraviolet radiation. Specifically, a clean reference spectral band is introduced, and radiation attenuation ratio and ultraviolet attenuation ratio features are constructed respectively. Short-time abrupt change rate features are also calculated to obtain the spectral abrupt change features. The construction of the dust disturbance characteristics involves jointly modeling the fluctuation characteristics of particulate matter concentration over a short period of time. Specifically, it involves calculating a set of statistical features characterizing the rapid dust disturbance behavior, including mean, standard deviation, skewness, kurtosis, and maximum rate of change, to obtain the dust disturbance characteristics. The feature integration is based on the historical exposure features, the spectral abrupt change features, and the dust disturbance features, and the feature splicing is performed to obtain a comprehensive meteorological feature set; Step S2: Calculation of dust stress index. Based on the multidimensional meteorological feature dataset, a weighted coupling model integrating the basic individual comfort index, dust breathing resistance penalty, and radiation anomaly correction is adopted, and a physical data fusion model for dust stress index is established to calculate the stress index and obtain reference data for dust stress index. Step S3: Meteorological health risk stratification. Based on the dust stress index reference data, the risk level is divided into a ladder-like level of concern, warning, severe and high risk by applying the preset multi-level risk judgment rules. Specific health warnings and protection recommendations are mapped to each level to obtain risk stratification reference data. Step S4: Human comfort data modeling. Based on the dust stress index reference data and risk stratification reference data, a structured comfort assessment conclusion is obtained through integrated modeling. This drives comfort decision modeling based on multi-objective optimization, generates control commands, and obtains a human comfort decision model.

2. The method for human comfort modeling based on meteorological data according to claim 1, characterized in that: In step S1, the clean reference spectrum is specifically based on historical meteorological sensor data. The time period in the annual data where the concentration of particulate matter in the environment is lower than a set threshold is selected as the clean sample day, and feature modeling is performed based on the statistical data of each time period of the clean sample day.

3. The method for human comfort modeling based on meteorological data according to claim 2, characterized in that: In step S2, the calculation of the dust stress index includes the following steps: constructing the basic index of individual comfort, modeling the dust breathing impedance penalty term, modeling the radiation anomaly correction term, constructing the physical data fusion model, and normalizing the calculation of the dust stress index.

4. The method for human comfort modeling based on meteorological data according to claim 3, characterized in that: In step S2, the individual comfort baseline index is constructed by quantitatively modeling the thermal and humidity comfort load caused by temperature, humidity and wind speed without considering sandstorms and radiation anomalies. Specifically, based on the temperature, relative humidity and wind speed characteristics in the multidimensional meteorological feature dataset, a quadratic deviation model centered on the comfort reference point is constructed to obtain the individual comfort baseline index. The modeling of the dust breathing resistance penalty term involves modeling the resistance and irritation load caused by dust and suspended particulate matter to the respiratory system. Specifically, based on the particulate matter in the environment and its historical exposure characteristics and dynamic statistical characteristics in the multidimensional meteorological feature dataset, a penalty model combining linear and nonlinear elements is constructed to calculate the dust breathing resistance penalty term. The radiation anomaly correction term modeling is used to model the abnormal changes in solar radiation and ultraviolet spectrum caused by dust storms. Specifically, the radiation anomaly correction term is constructed based on the abrupt change characteristics of the spectrum.

5. The method for human comfort modeling based on meteorological data according to claim 4, characterized in that: In step S2, the physical data fusion model is constructed by physically constraining and weighting the individual comfort baseline index, the dust breathing resistance penalty term, and the radiation anomaly correction term to construct a dust stress index physical data fusion model and obtain the initial dust stress index. The normalization calculation of the dust stress index specifically involves linearly normalizing the initial dust stress index to obtain reference data for the dust stress index. The dust stress index reference data includes the normalized dust stress index and the unnormalized initial dust stress index.

6. The method for human comfort modeling based on meteorological data according to claim 5, characterized in that: In step S4, the comfort decision modeling based on multi-objective optimization is specifically calculated based on minimizing the excessive load of sandstorm stress, minimizing energy consumption operating costs, and minimizing the frequent fluctuation of control commands.

7. A human comfort modeling system based on meteorological data, used to implement the human comfort modeling method based on meteorological data as described in any one of claims 1-6, characterized in that: It includes an integrated sensing module, an index calculation module, a risk stratification module, and a comfort modeling module.

8. The human comfort modeling system based on meteorological data according to claim 7, characterized in that: The integrated sensing module is used for integrated sensing of meteorological parameters. Through integrated sensing of meteorological parameters, a multidimensional meteorological feature dataset is obtained, and the multidimensional meteorological feature dataset is sent to the index calculation module. The index calculation module is used to calculate the dust stress index. Through the calculation of the dust stress index, reference data of the dust stress index is obtained, and the reference data of the dust stress index is sent to the risk stratification module and the comfort modeling module. The risk stratification module is used for meteorological health risk stratification. Through meteorological health risk stratification, risk stratification reference data is obtained, and the risk stratification reference data is sent to the comfort modeling module. The comfort modeling module is used for human comfort data modeling, and through human comfort data modeling, a human comfort decision model is obtained.

Images