Pedestrian high-temperature health risk early warning method based on thermal health risk index

Abstract

The invention discloses a pedestrian high-temperature health risk early warning method based on a thermal health risk index, and relates to the technical field of high-temperature risk early warning, and the method comprises the steps: collecting environment parameters, individual data and perception feedback data; obtaining spatial features and environmental features based on the environmental parameters, obtaining physiological features based on the individual data, and obtaining target variables based on the perception feedback data; inputting the spatial features, the environmental features, the physiological features and the target variables into a pre-training model to obtain a prediction probability; and obtaining a thermal health risk index based on the prediction probability and the target variable, and obtaining risk early warning information based on the thermal health risk index, thereby solving the problems of low early warning precision, high false alarm rate and low individuation degree caused by the adoption of a simple meteorological parameter or a fixed threshold value for early warning in a current urban high temperature early warning method.

Description

Technical Field

[0001] The present invention relates to the technical field of high temperature risk warning technology, and in particular to a pedestrian high temperature health risk warning method based on a heat health risk index. Background Art

[0002] At present, the issuance of urban high temperature warnings usually relies on the judgment method based on macro-meteorological parameters and fixed thresholds adopted by the meteorological department. This method monitors or forecasts one or a few key meteorological parameters such as air temperature and humidity at representative meteorological stations in the city (usually limited in number and with a wide coverage); the obtained meteorological parameter value (for example, the maximum daily temperature) is compared with a pre-set fixed threshold applicable to the entire warning area (for example, a specific temperature limit of 35°C, 37°C or 40°C stipulated by the national or local meteorological department). Sometimes it may also include time conditions, for example, to determine whether a certain threshold has been exceeded for multiple consecutive days; if the monitored or forecasted meteorological parameter value reaches or exceeds this fixed threshold, a hierarchical high temperature warning signal (such as yellow, orange or red warning) is triggered and issued to all members of the public in the area.

[0003] However, the core of the above method is to judge whether a few regional meteorological parameters have exceeded static limits. It has obvious limitations and fails to take the following key factors into consideration:

[0004] 1) Individual physiological differences: This approach fails to take into account the significant differences in physiological tolerance among different pedestrians (e.g., the elderly, children, and patients with underlying diseases);

[0005] 2) Activity and path differences: This does not take into account the specific activity intensity (e.g., walking, brisk walking), exposure duration, and the microclimate environment along the actual travel path (e.g., thermal environment differences caused by different underlying surfaces and shading conditions such as street canyons, open squares, and boulevards);

[0006] 3) Differences in the severity of heat symptoms: Warning information is usually general and cannot distinguish between health risks of varying severity, such as heat stress, heat exhaustion, and heat stroke, that may result from different exposure levels.

[0007] Therefore, the above-mentioned method based on simple meteorological parameters and fixed thresholds often leads to low accuracy of warning information (which cannot accurately reflect the actual risks faced by individuals), high false alarm rate (excessive warnings for low-risk groups) and low degree of personalization. It is difficult to guide pedestrians to take effective protective measures in a targeted manner, and it is impossible to effectively prevent the health risks caused by high temperatures at the individual level. Summary of the Invention

[0008] In order to solve the problems that the current urban high temperature warning method uses simple meteorological parameters or fixed thresholds for warning, resulting in low warning accuracy, high false alarm rate and low personalization, the present invention provides a pedestrian high temperature health risk warning method based on the heat health risk index, the method comprising: collecting environmental parameters, individual data and perception feedback data; obtaining spatial characteristics and environmental characteristics based on the environmental parameters, obtaining physiological characteristics based on the individual data, and obtaining target variables based on the perception feedback data; inputting the spatial characteristics, the environmental characteristics, the physiological characteristics and the target variables into a pre-trained model to obtain a predicted probability; obtaining a heat health risk index based on the predicted probability and the target variable, and obtaining risk warning information based on the heat health risk index.

[0009] The present invention collects real-time environmental parameters of the target path, takes into account the differences in the path environment, and is more in line with the actual environment; introduces individual characteristics and dynamic perception feedback, takes into account individual differences, is more targeted and effective, improves the degree of personalization, and can reduce the false alarm rate; obtains a dynamic adaptive health risk index based on real-time environmental parameters, individual characteristics, real-time thermal sensation and symptom feedback, integrates spatial characteristics and individual characteristics, dynamically responds to path differences, environmental differences and individual differences, and is more in line with actual application scenarios; introduces dynamic weight adaptation, effectively improves the accuracy and personalization of index calculation, solves the problem of insufficient accuracy of traditional fixed threshold methods, and provides users with personalized high-temperature health risk warnings through health risk indexes, accurately reflects the current heat risk status of pedestrians, reduces false alarm rates, improves the accuracy of personalized warnings and practical application effects, and achieves improved accuracy in pedestrian high-temperature health risk prediction and warnings.

[0010] Sunlight-exposed points and sun-shaded points refer to the areas of the path that are unshaded and shaded, respectively, under sunlight.

[0011] Furthermore, the environmental parameters are obtained as follows:

[0012] Collection equipment is set at the sunlight exposure point and the sunlight shadow point of the target path, and the environmental parameters are collected based on the collection equipment. The environmental parameters include air temperature, relative humidity, wind speed and solar radiation.

[0013] On the target path, selecting a location that can typically reflect the thermal environment characteristics of the unshaded sunny area of ​​the path as a sunlight exposure point, and selecting a location that can typically reflect the thermal environment characteristics of the shaded shadow area of ​​the path as a sunlight shadow point can improve the accuracy of environmental perception: it can simultaneously capture the key parameters of the two main and significantly different thermal environments on the path, obtain more comprehensive and detailed environmental data input, and lay the foundation for the subsequent accurate calculation of the thermal health risk index; support path differentiated assessment: after obtaining differentiated environmental data of the sunlight area and the shaded area, it can be combined with the subsequently calculated path sunlight / shadow ratio to dynamically and differentially evaluate the contribution of different road sections to pedestrian thermal risks; enhance the reality fit of risk assessment: enable the calculation of the risk index to more realistically simulate the cumulative thermal effects when pedestrians alternately pass through the sunlight and shaded areas on the actual path, significantly improving the fit and accuracy of the risk assessment results with the individual's actual feelings.

[0014] Furthermore, the spatial features are obtained as follows:

[0015] The sunlit area and the shadow area of ​​the target path are obtained, the sunlit ratio and the shadow ratio of the sunlit area and the shadow area are calculated respectively, and the spatial feature is obtained based on the sunlit ratio and the shadow ratio.

[0016] Real-time analysis of the ratio of sunlight and shade on the path ahead of the user, dynamic fusion of sunlight and shaded road section environmental data, and dynamic determination of the proportion of environmental data usage. This ratio directly and quantitatively describes the degree and distribution characteristics of solar radiation received in space by the path segment the user is about to travel, reflects the structural thermal environment characteristics of the path, and provides a weight basis for effectively combining the two different environmental data collected from sunlight exposure points and sunlight shadow points.

[0017] Furthermore, the environmental characteristics include air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics. The calculation formula of the air temperature characteristics is:

[0018] Ta=Ta sun ×Sum percentage +Ta shade ×Shade percentage ;

[0019] The calculation formula of relative humidity characteristic is:

[0020] RH=RH sun ×Sun percentage +RH shade ×Shade percentage ;

[0021] The calculation formula of wind speed characteristics is:

[0022] Va=Va sun×Sun percentage +Va shade ×Shade percentage ;

[0023] The calculation formula for solar radiation characteristics is:

[0024] G=G sun ×sun percentage +G shade ×Shade percentage ;

[0025] Among them, Ta, RH, Va and G represent the air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics respectively. sun RH sun 、Va sun and G sun Represent the air temperature, relative humidity, wind speed and solar radiation in the sunshine area respectively, Ta shade RH shade 、Va shade and G shade Represent the air temperature, relative humidity, wind speed and solar radiation in the shadow area, Sun percentage Indicates the proportion of sunlight, Shade percentage Indicates the shadow scale.

[0026] Furthermore, the target variables include a thermal sensation index, a thermal acceptance index, a number of thermal symptoms, and a thermal symptom presence index. The specific steps for obtaining the predicted probability include:

[0027] Model parameters are obtained based on the spatial characteristics, the environmental characteristics, the physiological characteristics and the target variables, and the model parameters are input into a pre-trained thermal sensation model, a pre-trained thermal acceptance model and a pre-trained thermal symptom model to obtain thermal sensation prediction probability, thermal acceptance prediction probability and thermal symptom prediction probability respectively, and the prediction probability is obtained based on the thermal sensation prediction probability, the thermal acceptance prediction probability and the thermal symptom prediction probability.

[0028] Use multi-model integrated prediction to improve the robustness and reliability of model prediction.

[0029] Furthermore, the calculation formula for the heat health risk index is:

[0030]

[0031] TSV norm =(TSV+3) / 6;

[0032] Among them, DSHRI represents the heat health risk index, W_TSV, W_TAV and W_Symptom represent the thermal sensation weight factor, thermal acceptance weight factor and thermal symptom weight factor respectively, TSV_norm represents the relevant parameters of the thermal sensation index, TSV represents the thermal sensation index, TAV represents the thermal acceptance index, and Symptom_score represents the weighted score of the heat symptom classification.

[0033] Furthermore, the thermal sensation weight factor, thermal acceptance weight factor, and thermal symptom weight factor are obtained as follows:

[0034] The thermal sensation weighting factor is obtained based on the thermal sensation prediction probability, the thermal acceptance weighting factor is obtained based on the thermal acceptance prediction probability, and the thermal symptom weighting factor is obtained based on the thermal symptom prediction probability.

[0035] Prediction probability reflects credibility. The predicted probabilities output by the pre-trained model (such as the predicted probability of heat sensation and heat symptoms) not only predict the likelihood of a certain condition (such as feeling hot or experiencing heatstroke symptoms) but also, to a certain extent, reflect the model's confidence in the prediction. A high probability generally means that the model is highly confident that the condition will occur based on the current input data. Conversely, a probability close to random guessing (e.g., around 50%) indicates a high degree of uncertainty in the model's judgment of the condition. Therefore, using the predicted probability as a weighting factor dynamically adjusts the influence of various indicators in the DSHRI calculation, more accurately reflecting the current situation.

[0036] When a pre-trained model (e.g., a heat symptom model) outputs a potential risk signal with only a low predicted probability (e.g., a 30% probability of a headache), the present invention also sets the corresponding weight factor to a correspondingly low value. This means that even if this low-probability signal itself may be a false alarm, its contribution to the final Heat Health Risk Index (DSHRI) is significantly limited. The system will not easily raise the overall risk rating due to an uncertain and low-credibility risk prediction by a model, thereby avoiding false alarms based on uncertain information and achieving the purpose of reducing the risk of false alarms. On the contrary, when a model outputs a clear risk signal with a very high prediction probability (for example, the thermal sensation model predicts a 95% probability of feeling hot, or the heat symptom model predicts a 90% probability of sweating symptoms), its corresponding weight factor will become very high, which makes this high-risk, high-credibility indicator dominate in the DSHRI calculation. Even if the predicted risk of other indicators is not high or the credibility is low, this high-weighted risk signal can effectively improve the final DSHRI value, ensuring that this real risk with high confidence from the model will not be underestimated or ignored, thereby reducing the possibility of missed reports and achieving the purpose of reducing the risk of missed reports.

[0037] Therefore, using prediction probability as a weighting factor can effectively reduce the risk of false positives or negative negatives from the model, improving the accuracy and stability of the overall warning. Furthermore, when the model's prediction credibility is high, the weight of this indicator automatically increases, better highlighting the impact of important indicators.

[0038] Through this probability-based dynamic weighting, the calculation results of DSHRI are less sensitive to the forecast fluctuations of each model (especially those with large uncertainties), and are more dependent on the stable and confident forecast output of the model, thereby improving the robustness and stability of the overall early warning system.

[0039] Furthermore, the heat symptom grading weighted score is obtained as follows:

[0040] A first weight factor for each heat stress symptom is obtained based on the pre-trained heat symptom model, and the heat symptom graded weighted score is obtained based on the first weight factor and a human body heat tolerance benchmark reference table.

[0041] Taking into account individual differences and differences in the severity of fever symptoms and improving the degree of personalization can improve the accuracy of early warnings and reduce the false alarm rate.

[0042] Simple judgments on the presence or absence of heat symptoms or quantitative statistics cannot distinguish the actual threat levels of different heat-related symptoms to human health. For example, excessive sweating and gastrointestinal discomfort are both heat symptoms, but their risk levels are completely different. In order to more accurately quantify the complex health risks posed by multiple heat symptoms of varying severity, the present invention introduces a heat symptom hierarchical weighted score. This scoring mechanism combines the occurrence of heat symptoms with the medical assessment of the inherent severity of the symptoms (reflected in the weight factor of each heat stress symptom), thereby calculating an indicator that can comprehensively reflect the overall severity of the current predicted symptom combination.

[0043] Furthermore, the calculation formula for the heat symptom grading weighted score is:

[0044]

[0045] Among them, Symptom_score represents the weighted score of fever symptoms, symptom_weight i represents the weight factor of the i-th heat stress symptom, Symptom_baseline represents the human body's heat tolerance baseline, n represents the number of heat stress symptoms, and i represents an integer greater than or equal to 1.

[0046] Furthermore, the method further comprises:

[0047] If the environmental parameter is greater than or equal to a preset threshold, a first reminder message is pushed;

[0048] If the heat stress symptom is a preset symptom, a second reminder message is pushed.

[0049] Provide users with redundant protection and reduce safety issues under high temperature risks.

[0050] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0051] 1. The present invention collects real-time environmental parameters of the target path, takes into account the differences in the path environment, and is more in line with the actual environment; introduces individual characteristics and dynamic perception feedback, takes into account individual differences, is more targeted and effective, improves the degree of personalization, and can reduce the false alarm rate; obtains a dynamic adaptive health risk index based on real-time environmental parameters, individual characteristics, real-time thermal sensation and symptom feedback, integrates spatial characteristics and individual characteristics, dynamically responds to path differences, environmental differences and individual differences, and is more in line with actual application scenarios. It introduces dynamic weight adaptive technology to effectively improve the accuracy and personalization of index calculation, solves the problem of insufficient accuracy of traditional fixed threshold methods, and provides users with personalized high-temperature health risk warnings through health risk indexes, accurately reflects the current heat risk status of pedestrians, reduces false alarm rates, improves the accuracy of personalized warnings and practical application effects, and achieves improved accuracy in pedestrian high-temperature health risk prediction and warnings.

[0052] 2. The thermal sensation weight factor is obtained based on the predicted probability of thermal sensation, the thermal acceptance weight factor is obtained based on the predicted probability of thermal acceptance, and the thermal symptom weight factor is obtained based on the predicted probability of thermal symptoms. Using the predicted probability as a weight factor can dynamically adjust the influence of each indicator during the DSHRI calculation, and more accurately reflect the actual situation at the current moment. When the model prediction credibility is high, the weight of the indicator is automatically increased to better highlight the impact of important indicators. This effectively reduces the risk of false alarms or omissions caused by the model, and improves the accuracy and stability of the overall warning.

[0053] 3. Based on the first weight factor and the human body's heat tolerance benchmark reference table, a weighted score for heat symptom grading is obtained; considering individual differences and differences in the severity of heat symptoms, improving the degree of personalization can improve warning accuracy and reduce false alarm rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation of the embodiments of the present invention;

[0055] Figure 1 It is a flow chart of a pedestrian high temperature health risk early warning method based on a heat health risk index in the present invention. DETAILED DESCRIPTION

[0056] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features therein can be combined with each other without conflict.

[0057] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0058] Example 1

[0059] refer to Figure 1 This embodiment provides a pedestrian high temperature health risk warning method based on a heat health risk index, the method comprising:

[0060] Collect environmental parameters, individual data and sensory feedback data;

[0061] The environmental parameters are obtained in the following manner: a collection device is set at the sunlight exposure point and the sunlight shadow point of the target path, and the environmental parameters are collected based on the collection device. The environmental parameters include air temperature, relative humidity, wind speed and solar radiation. For example, a sunlight exposure point and a shadow point are selected at the central position of the unshaded area and the shaded area of ​​the target path, and a micro-meteorological station is set up at each. The micro-meteorological station model can be TRM-ZS2-BX, and the micro-meteorological station obtains air temperature (Ta), relative humidity (RH), wind speed (Va), and solar radiation (G) in real time. In this embodiment, the center of the open and unobstructed area is usually selected as the sunlight exposure point, and the center of the stable and large shadow area is selected as the sunlight shadow point, avoiding the selection of those instantaneous points that are not representative due to temporary, small-scale obstruction or light spots. Among them, representativeness means that these points should be able to reflect the general or average conditions of such areas (sunlight area or shadow area) in a section of the path or a region.

[0062] Individual data can be obtained through mobile terminals, including user's gender, height, weight, clothing, travel route planning and other information.

[0063] Perceptual feedback data can be recorded in real time via mobile devices or electronic questionnaires: thermal sensation (TSV), thermal acceptance (TAV), and self-reported heat symptoms. The thermal sensation level can be set on a seven-point scale: -3 is cold, -2 is cool, -1 is slightly cool, 0 is neutral, +1 is slightly warm, +2 is warm, and +3 is hot; the thermal acceptance level can be set on a two-point scale: 0 is unacceptable and 1 is acceptable. Self-reported heat symptoms can be selected in multiple options, including: none, weakness, dizziness, nausea, sweating, chest tightness, headache, rapid heartbeat, irritability, pale complexion, gastrointestinal discomfort, and others (users can describe additional symptoms themselves).

[0064] Obtaining spatial features and environmental features based on the environmental parameters, obtaining physiological features based on the individual data, and obtaining target variables based on the sensory feedback data;

[0065] Among them, the spatial features are obtained by using a GIS model to obtain the sunlight area and shadow area of ​​the target path, respectively calculating the sunlight ratio and shadow ratio of the sunlight area and the shadow area, and obtaining the spatial features based on the sunlight ratio and the shadow ratio.

[0066] The environmental characteristics include air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics. The calculation formula of air temperature characteristics is:

[0067] Ta=Ta sum ×Sun percentage +Ta shade ×SHade percentage ; (1)

[0068] The calculation formula of relative humidity characteristic is:

[0069] RH=RH sun ×Sun percentage +RH shade ×Shade percentage ; (2)

[0070] The calculation formula of wind speed characteristics is:

[0071] Va=Va sun ×Shade percentage +Va shade ×Shade percentage ; (3)

[0072] The calculation formula for solar radiation characteristics is:

[0073] G=G sun ×Sun percentage +G shade ×Shade percentage ; (4)

[0074] Among them, Ta, RH, Va and G represent the air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics respectively. sun RH sun 、Va sun and G sun Represent the air temperature, relative humidity, wind speed and solar radiation in the sunshine area respectively, Ta shade RH shade 、Va shade and G shadeRepresent the air temperature, relative humidity, wind speed and solar radiation in the shadow area, Sun percentage Indicates the proportion of sunlight, Shade percentage Indicates the shadow scale.

[0075] Dynamically integrate the environmental data of sunlight and shaded road sections to obtain weighted environmental characteristics.

[0076] Physiological characteristics may include: BMI value calculated based on height and weight, gender, and thermal resistance calculated based on clothing. In this embodiment, international standards (such as ISO 9920: Ergonomics of the thermal environment—Estimation of thermal insulation and water vapor resistance of a clothing ensemble or ASHRAE Standard 55: Thermal Environmental Conditions for Human Occupancy) can be used to assign a standard inherent thermal resistance value to each clothing item or each clothing combination category in the database. This thermal resistance value is usually measured in clo (1 clo ≈ 0.155 m2·K / W) and represents the thermal insulation capacity of the clothing. The clo values ​​of each individual piece of clothing are simply added together to obtain the total thermal resistance of the entire clothing combination.

[0077] Inputting the spatial features, the environmental features, the physiological features, and the target variable into a pre-trained model to obtain a predicted probability;

[0078] The target variables include the thermal sensation index, the thermal acceptance index, the number of heat symptoms, and the presence index of heat symptoms (whether heat symptoms exist). The specific steps for obtaining the predicted probability include:

[0079] Model parameters are obtained based on the spatial features, the environmental features, the physiological features, and the target variables. The model parameters are input into a pre-trained thermal sensation model, a pre-trained thermal acceptance model, and a pre-trained thermal symptom model to obtain thermal sensation prediction probability, thermal acceptance prediction probability, and thermal symptom prediction probability, respectively. The prediction probability is obtained based on the thermal sensation prediction probability, the thermal acceptance prediction probability, and the thermal symptom prediction probability. In this embodiment, gender is a binary variable and can be encoded using one-hot encoding. The target variables are all categorical variables. The model parameters can also be integer-encoded using LabelEncoder to ensure compatibility with the machine learning model.

[0080] In this embodiment, the pre-trained thermal sensation model, the pre-trained thermal acceptance model and the pre-trained thermal symptom model can be a multi-classification logistic regression model, an XGBoost model and a multi-classification logistic regression model, respectively, which are obtained by training with historical data. The model training and hyperparameter optimization adopt a grid search method with five-fold cross-validation.

[0081] The thermal sensation index reflects the subjective intensity of heat stress (-3 to +3), the thermal acceptance index represents the user's subjective willingness, and the thermal symptom graded weighted score (symptom_score) in the thermal symptom assessment jointly determines the thermal safety assessment.

[0082] Based on the predicted probability and the target variable, a heat health risk index is obtained, and risk warning information is obtained based on the heat health risk index.

[0083] The calculation formula for the heat health risk index is:

[0084]

[0085] TSV norm =(TSV+3) / 6; (6)

[0086] Among them, DSHRI represents the heat health risk index, W_TSV, W_TAV and W_Symptom represent the thermal sensation weight factor, thermal acceptance weight factor and thermal symptom weight factor respectively, TSV_norm represents the relevant parameters of the thermal sensation index, TSV represents the thermal sensation index, TAV represents the thermal acceptance index, and Symptom_score represents the weighted score of the heat symptom classification.

[0087] Formula 6 maps the thermal sensation index from -3 to +3 to the range of 0-1. The original thermal sensation index scale (for example, the ASHRAE seven-point thermal sensation scale: +3 hot, +2 warm, +1 slightly warm, 0 neutral, -1 slightly cool, -2 cool, -3 cold) is an ordered categorical scale. Its numerical range and meaning are inconsistent with other indicators used in the calculation of the Heat Health Risk Index (DSHRI) and the predicted probabilities used as weighting factors. The DSHRI calculation formula requires a weighted summation of each indicator and normalization to the range [0, 1] to ensure that the TSV has the same mathematical meaning as other indicators when weighted summed, ensuring the effectiveness of weighted fusion.

[0088] Among them, the thermal sensation weight factor, thermal acceptance weight factor and thermal symptom weight factor are obtained as follows: the thermal sensation weight factor is obtained based on the thermal sensation prediction probability, the thermal acceptance weight factor is obtained based on the thermal acceptance prediction probability, and the thermal symptom weight factor is obtained based on the thermal symptom prediction probability.

[0089] The predicted probability output by each model represents the model's degree of certainty about a particular prediction category (between 0 and 1). The closer the probability is to 1, the more reliable the model's prediction. Using this probability as a weighting factor dynamically adjusts the influence of each indicator in the DSHRI calculation, more accurately reflecting the current situation. When the model's prediction is more reliable, the weight of that indicator automatically increases, further highlighting the impact of important indicators. This effectively reduces the risk of false positives or negatives, improving the accuracy and stability of the overall warning.

[0090] The heat symptom grading weighted score is obtained by obtaining a first weighting factor for each heat stress symptom based on the pre-trained heat symptom model, and then obtaining the heat symptom grading weighted score based on the first weighting factor and a reference table of human heat tolerance. Referring to the clinical heat stress level assignments, as shown in Table 1, the pre-trained heat symptom model predicts all possible symptoms in real time, calculates the sum of these weights, and then divides them by the human heat tolerance baseline to calculate the score. The human heat tolerance baseline (Symptom_baseline) is derived from pre-experimental results and varies depending on individual health conditions, as shown in Table 2.

[0091] Table 1 Reference table of heat stress symptom weights

[0092]

[0093] Table 2 Human body heat tolerance reference table

[0094]

[0095] The calculation formula for the weighted score of fever symptom grading is:

[0096]

[0097] Among them, Symptom_score represents the weighted score of fever symptoms, symptom_weight i represents the weight factor of the i-th heat stress symptom, Symptom_baseline represents the human body's heat tolerance baseline, n represents the number of heat stress symptoms, and i represents an integer greater than or equal to 1.

[0098] Warning based on DSHRI calculation results:

[0099] Low risk (DSHRI < 0.4): Continue monitoring;

[0100] Medium risk (0.4≤DSHRI<0.6): push water replenishment reminder;

[0101] High risk (DSHRI ≥ 0.6): A high temperature danger alert has been triggered. Please immediately seek a nearby summer retreat to rest and continue traveling only after the danger alert disappears.

[0102] In this embodiment, the method further includes:

[0103] If the environmental parameter is greater than or equal to the preset threshold, the first reminder message is pushed; for example, if the solar radiation is ≥500W / m 2 When the sun is on, an additional sunshade reminder will be pushed.

[0104] If the heat stress symptom is a preset symptom, a second reminder message is pushed. If the heat symptom prediction detects excessive sweating, an additional hydration reminder is pushed.

[0105] The dual insurance design of dynamic risk grading and secondary redundant channels significantly improves the reliability and safety of risk warning.

[0106] This application can also be applied to high temperature health warnings in smart cities. Through the dynamic and real-time calculation of the DSHRI index, personalized high temperature health risk warnings and suggestions can be provided to urban residents. It can be integrated with the city's intelligent management system to achieve accurate and personalized push of high temperature risk avoidance information. It can be applied to urban walking path planning and optimization. In urban renewal or new block planning, real-time dynamic thermal comfort risk data can be used to optimize pedestrian route layout, recommend safer and more comfortable walking paths for residents, reduce health risks brought by heat exposure, and improve the comfort and safety of slow-moving spaces in blocks. It can also optimize the layout of urban greening shade and guide the planning and design of landscape and pedestrian facilities. It can be used in public health monitoring and risk management to provide key indicators of the Walk Score system, helping planners optimize the walking environment; it can be used in public health monitoring and risk management to provide health departments with accurate heat stress data, supporting real-time monitoring and emergency response to health risks caused by high temperatures in summer; it can provide specialized risk assessment and prevention strategies for sensitive groups (such as the elderly, children, and people working outdoors); it can be used in mobile applications and services to integrate the DSHRI method into mobile applications and wearable devices to provide outdoor users with real-time health warnings and recommendations (such as sun protection reminders and hydration reminders); it can also improve the practicality and accuracy of meteorological service applications and reduce the health risks caused by heat waves.

[0107] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0108] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A pedestrian high temperature health risk early warning method based on a heat health risk index, characterized in that: The method comprises: Collect environmental parameters, individual data and sensory feedback data; Obtaining spatial features and environmental features based on the environmental parameters, obtaining physiological features based on the individual data, and obtaining target variables based on the sensory feedback data; Inputting the spatial features, the environmental features, the physiological features, and the target variable into a pre-trained model to obtain a predicted probability; Based on the predicted probability and the target variable, a heat health risk index is obtained, and risk warning information is obtained based on the heat health risk index.

2. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 1 is characterized in that: The environmental parameters are obtained as follows: Collection equipment is set at the sunlight exposure point and the sunlight shadow point of the target path, and the environmental parameters are collected based on the collection equipment. The environmental parameters include air temperature, relative humidity, wind speed and solar radiation.

3. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 2 is characterized in that: The spatial features are obtained as follows: The sunlit area and the shadow area of ​​the target path are obtained, the sunlit ratio and the shadow ratio of the sunlit area and the shadow area are calculated respectively, and the spatial feature is obtained based on the sunlit ratio and the shadow ratio.

4. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 3 is characterized in that: The environmental characteristics include air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics. The calculation formula of air temperature characteristics is: Ta=Ta sun ×Sun percentage +Ta shade ×Shade percentage ; The relative humidity characteristic is calculated as: RH=RH sun ×Sun percentage +RH shade ×Shade percentage ; The calculation formula of wind speed characteristics is: Go=Go sun ×Sun percentage +Will shade ×Shade percentage ; The calculation formula for solar radiation characteristics is: G=G sun ×Sun percentage +G shade ×Shade percentage ; Among them, Ta, RH, Va and G represent the air temperature characteristics, relative humidity characteristics, wind speed characteristics and solar radiation characteristics respectively. sun RH sun 、Va sun and G sun Represent the air temperature, relative humidity, wind speed and solar radiation in the sunshine area respectively, Ta shade RH shade 、Va shade and G shade Represent the air temperature, relative humidity, wind speed and solar radiation in the shadow area, Sun percentage Indicates the proportion of sunlight, Shade percentage Indicates the shadow scale.

5. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 1 is characterized in that: The target variables include thermal sensation index, thermal acceptance index, number of heat symptoms, and presence index of heat symptoms. The specific steps of obtaining the predicted probability include: Model parameters are obtained based on the spatial characteristics, the environmental characteristics, the physiological characteristics and the target variables, and the model parameters are input into a pre-trained thermal sensation model, a pre-trained thermal acceptance model and a pre-trained thermal acceptance model to obtain thermal sensation prediction probability, thermal acceptance prediction probability and thermal symptom prediction probability respectively, and the prediction probability is obtained based on the thermal sensation prediction probability, the thermal acceptance prediction probability and the thermal symptom prediction probability.

6. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 5 is characterized in that: The calculation formula for the heat health risk index is: TSV norm =(TSV+3) / 6; Among them, DSHRI represents the heat health risk index, W_TSV, W_TAV and W_Symptom represent the thermal sensation weight factor, thermal acceptance weight factor and thermal symptom weight factor respectively, TSV_norm represents the relevant parameters of the thermal sensation index, TSV represents the thermal sensation index, TAV represents the thermal acceptance index, and Symptom_score represents the weighted score of the heat symptom classification.

7. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 6 is characterized in that: The thermal sensation weight factor, thermal acceptance weight factor, and thermal symptom weight factor are obtained as follows: The thermal sensation weighting factor is obtained based on the thermal sensation prediction probability, the thermal acceptance weighting factor is obtained based on the thermal acceptance prediction probability, and the thermal symptom weighting factor is obtained based on the thermal symptom prediction probability.

8. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 6 is characterized in that: The weighted score for fever symptom grading is obtained as follows: A first weight factor for each heat stress symptom is obtained based on the pre-trained heat symptom model, and the heat symptom graded weighted score is obtained based on the first weight factor and a human body heat tolerance benchmark reference table.

9. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 8 is characterized in that: The calculation formula for the weighted score of fever symptom grading is: Among them, Symptom_score represents the weighted score of fever symptoms, symptom_weight i represents the weight factor of the i-th heat stress symptom, Sympton_baseline represents the human body's heat tolerance baseline, n represents the number of heat stress symptoms, and i represents an integer greater than or equal to 1.

10. The pedestrian high temperature health risk early warning method based on the heat health risk index according to claim 8, characterized in that: The method further comprises: If the environmental parameter is greater than or equal to a preset threshold, a first reminder message is pushed; If the heat stress symptom is a preset symptom, a second reminder message is pushed.

Images