A hierarchical personalized early warning system and method based on air quality health index

Abstract

This invention relates to the field of air quality monitoring technology, specifically disclosing a hierarchical personalized early warning system and method based on the Air Quality Health Index (AHI). The system dynamically acquires basic air quality data and user-specific characteristic data through a data acquisition module. After preprocessing, the health index calculation module innovatively integrates real-time pollutant concentrations, short-term peak values, and dynamic trends to calculate a general air quality health index. The hierarchical early warning module then determines the initial public warning level based on this index. The personalized adaptation module calculates a personalized correction factor based on the user's age, underlying diseases, and historical exposure accumulation characteristics, amplifying the general index in real time to generate a health risk score reflecting the individual. Finally, it outputs the final personalized warning level and customized protection recommendations based on the highest possible score. This invention represents a leap from static environmental monitoring to dynamic personal health risk management, significantly improving the protection effectiveness for sensitive populations.

Description

Technical Field

[0001] This invention relates to the field of air quality monitoring technology, specifically to a hierarchical personalized early warning system and method based on the air quality health index. Background Technology

[0002] With the acceleration of industrialization and urbanization, air quality problems are becoming increasingly prominent. Pollutants such as particulate matter and harmful gases in the air can have serious effects on human health, especially on the elderly, children, and special groups with underlying diseases.

[0003] Currently, most existing air quality early warning systems suffer from the following problems: 1. They are mainly based on instantaneous pollutant concentrations, failing to reflect dynamic trends in pollution (such as rapid deterioration or improvement) and the historical residual effects of short-term high exposure peaks; 2. They ignore the drastically different vulnerabilities and health risks exhibited by different population groups due to individual differences in age, health status, exposure history, etc. For children, the elderly, and sensitive groups with respiratory or cardiovascular diseases, uniform early warning standards may severely underestimate the actual health threats they face, leading to inadequate protection. Therefore, this invention proposes a tiered personalized early warning system and method based on the Air Quality Health Index. Summary of the Invention

[0004] The purpose of this invention is to provide a tiered personalized early warning system and method based on the air quality health index, thereby solving the above-mentioned technical problems: The objective of this invention can be achieved through the following technical solutions: A hierarchical personalized early warning system based on the air quality health index, the system comprising a data acquisition module, a health index calculation module, a hierarchical early warning module, and a personalized adaptation module; The data acquisition module is used to collect basic air quality data and personalized user characteristic data; The data preprocessing module, connected to the data acquisition module, is used to clean, denoise, standardize, and fuse the collected basic air quality data and user personalized feature data. The health index calculation module is connected to the data preprocessing module and is used to calculate the air quality health index based on the preprocessed air quality baseline data. The tiered early warning module is connected to the health index calculation module. It is used to preset multi-level early warning thresholds and compare the personalized air quality health index with the preset multi-level early warning thresholds to determine the initial early warning level. The personalized adaptation module is connected to the health index calculation module and the hierarchical early warning module respectively. It performs real-time correction of the air quality health index based on personalized feature data to obtain a personalized air quality health index. The initial early warning level is adjusted according to the correction result to obtain the final early warning level.

[0005] As a further description of the technical solution of the present invention, the working process of the data acquisition module includes: Basic air quality data is collected in real time through a sensor network deployed in the region, and personalized user data is collected through user terminals. The collection frequency is dynamically adjusted according to changes in air quality. The collection frequency is once per hour during periods of severe pollution and once every 6 hours during periods of good air quality.

[0006] As a further description of the technical solution of the present invention, the working process of the data preprocessing module includes: Missing values ​​in the basic air quality data are supplemented using linear interpolation or mean imputation, and outliers are identified and removed using the 3σ criterion. User-personalized feature data are classified and coded, and the classified data is converted into calculable numerical data. All preprocessed data sources are normalized to ensure that the data are on the same order of magnitude.

[0007] As a further description of the technical solution of the present invention, the working process of the health index calculation module includes: Collect n important pollutants, for the i-th pollutant Indicates the current time concentration, This indicates that the i-th pollutant is present in the current set time period. The maximum value, Indicates the current set time period The rate of change of , where, ; The Air Quality Health Index is calculated by combining the current pollutant concentration, the peak pollutant concentration over the current set time period, and the rate of change of pollutant concentration over the current set time period. ; In the formula, These are the reference concentrations for the i-th pollutant set by the system. The i-th pollutant set for the system within a set time period The reference rate of change within, The indicator is the trend of change of the i-th pollutant. The weighting coefficient is the one corresponding to the i-th pollutant. , and These are the weighting coefficients corresponding to real-time values, peak values, and trends of change, respectively. It is the peak attenuation coefficient. This represents the air quality health index at the current moment.

[0008] As a further description of the technical solution of the present invention, the peak attenuation coefficient The value is assigned based on the time when the peak occurs, if the time of the peak occurrence is different from the current time. The interval is less than ,but The value is 1 if the peak occurs at a time that is 1 from the current time. The interval is greater than or equal to Less than ,but The value is 0.5. If the peak occurs at a time that is 0.5 away from the current time... The interval is greater than or equal to ,but The value is 0.2, and the peak value is 0.2 from the current time. The farther away, the smaller the impact.

[0009] As a further description of the technical solution of the present invention, the calculation process of the indicator of the influence of the change trend of the i-th pollutant includes: Obtain the current concentration of the i-th pollutant. Initial concentration for the current set time period Concentration at the midpoint of the current set time period The acceleration of change of the i-th pollutant in the current time period can be calculated using the following formula. : ; Let be the rate of change of the i-th pollutant in the current time period. Let be the acceleration of the change of the i-th pollutant in the current time period, so ,in, and These are the weighting coefficients corresponding to the rate of change and the acceleration, respectively. and The system is set to operate within a specified time period. The reference rate of change and acceleration within.

[0010] As a further description of the technical solution of the present invention, the working process of the hierarchical early warning module includes: The air quality health index (AQI) levels are categorized as: Good, Lightly Polluted, Moderately Polluted, and Heavily Polluted or Worst. Each pollution level has a corresponding AQI threshold range. The current AQI is then used to determine the air quality health index. Map the values ​​to the corresponding threshold ranges to determine the initial warning level.

[0011] As a further description of the technical solution of the present invention, the working process of the personalized adaptation module includes: Personalized user data includes: age characteristics, basic respiratory or cardiovascular disease characteristics, and historical exposure accumulation characteristics; The age characteristic sensitivity index ,in, Based on user age, the criteria are: under 12 years old or over 65 years old. =1, between 12 and 65 years old =0; Sensitivity index of basic respiratory or cardiovascular disease characteristics ,in, Using categorical variables, if there is no underlying respiratory or cardiovascular disease =0, if either a basic respiratory or cardiovascular disease is present. =1, if both underlying respiratory and cardiovascular diseases exist. =2; The historical exposure cumulative characteristic sensitivity index = ,in, The cumulative effect coefficient set for the system. The cumulative exposure time within the currently set time period. The safety threshold for exposure time within the current set time period is set for the writing system; Sensitivity indicators for age characteristics and sensitivity indicators for baseline respiratory or cardiovascular disease characteristics, respectively. The correction factor is obtained by weighting and summing the sensitivity index of historical exposure cumulative characteristics. Therefore, personalized air quality health index for users ,Will Map the data to the corresponding threshold range to determine the final warning level.

[0012] A tiered, personalized early warning method based on the air quality health index, the method comprising the following steps: Step S1: Collect basic air quality data and personalized user characteristic data of various pollutants in real time at a dynamic frequency through a sensor network deployed in the target area. Step S2: Preprocess the raw data collected in step S1; Step S3: Calculate the general air quality health index based on the preprocessed air quality baseline data; Step S4: Initial warning level determination; Step S5: Based on various user-personalized data, correct the general air quality health index to a personalized health index; Step S6: Determine the personalized early warning level.

[0013] The beneficial effects of this invention are: This invention dynamically acquires basic air quality data and personalized user characteristic data through a data acquisition module. After preprocessing, a health index calculation module innovatively integrates real-time pollutant concentrations, short-term peak values, and dynamic trends to calculate a general air quality health index. A tiered early warning module then determines the initial public warning level based on this index. A personalized adaptation module calculates personalized correction factors based on user age, underlying diseases, and historical exposure accumulation characteristics, amplifying the general index in real time to generate an individual's health risk score. Finally, based on the highest possible score, it outputs the final personalized warning level and customized protection recommendations. This invention represents a leap from static environmental monitoring to dynamic personal health risk management, significantly improving the protective effectiveness for sensitive populations. Attached Figure Description

[0014] The invention will now be further described with reference to the accompanying drawings.

[0015] Figure 1 This is a partial structural diagram of the hierarchical personalized early warning system based on the air quality health index of this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1 As shown, the present invention provides a hierarchical personalized early warning system based on the air quality health index. The system includes a data acquisition module, a health index calculation module, a hierarchical early warning module, and a personalized adaptation module. The data acquisition module is used to collect basic air quality data and personalized user characteristic data; The data preprocessing module, connected to the data acquisition module, is used to clean, denoise, standardize, and fuse the collected basic air quality data and user personalized feature data. The health index calculation module is connected to the data preprocessing module and is used to calculate the air quality health index based on the preprocessed air quality baseline data. The tiered early warning module is connected to the health index calculation module. It is used to preset multi-level early warning thresholds and compare the personalized air quality health index with the preset multi-level early warning thresholds to determine the initial early warning level. The personalized adaptation module is connected to the health index calculation module and the hierarchical early warning module respectively. It performs real-time correction of the air quality health index based on personalized feature data to obtain a personalized air quality health index. The initial early warning level is adjusted according to the correction result to obtain the final early warning level.

[0018] Through the above technical solution, this invention achieves an intelligent transformation from general air quality assessment to precise personal health risk early warning by dynamically integrating multi-dimensional environmental data with user personal characteristics. The system first collects data on various pollutants (such as PM2.5, ... The system collects basic air quality data (including age, history of basic respiratory / cardiovascular diseases, and duration of historical exposure) through user terminals, and dynamically adjusts the data collection frequency based on the severity of pollution to ensure the timeliness and representativeness of the data. Subsequently, the data preprocessing module cleans, denoises, standardizes, and integrates the raw data to provide a high-quality data foundation for subsequent calculations. The core innovation of the system lies in the health index calculation module. This module does not only evaluate based on real-time pollutant concentrations but also innovatively constructs a composite calculation model that comprehensively considers real-time pollutant concentrations, peak concentrations within a set time period, and dynamic trends. The trend indicator quantifies the deterioration or improvement of pollution by calculating the rate and acceleration of pollutant changes, while the peak concentration incorporates a coefficient that decays over time to scientifically reflect the residual health impacts of recent high-exposure events. Through weighted summation, a general air quality health index is generated that sensitively reflects the dynamic intensity and potential risks of pollution. This index is input into the tiered early warning module and compared with preset health risk threshold ranges for good, light pollution, moderate pollution, and heavy pollution and above to preliminarily determine the warning level for the public. To achieve truly personalized service, the personalized adaptation module calculates correction factors based on user characteristic data: sensitivity indicators are generated for age (e.g., a risk coefficient of 1 for those under 12 or over 65), disease status (0 for no disease, 1 for one disease, and 2 for two diseases), and historical cumulative exposure duration (calculated according to exposure ratio). These are then weighted and summed to obtain the personalized correction factor. Finally, the general index is multiplied by the correction factor to obtain a personalized air quality health index for the specific user. The system then maps the personalized air quality health index to a warning threshold range, generating a final personalized warning level and customized protection recommendations for that user (e.g., issuing a higher level of risk warning to COPD patients than to the general public). This entire process, by deeply integrating macro-environmental monitoring with micro-level personal health information, solves the problems of insufficient risk warnings for sensitive groups and lack of dynamic foresight in traditional air quality indices. It provides the public, especially vulnerable groups, with a more timely, targeted, and health-guiding air pollution risk management system.

[0019] As a further description of the technical solution of the present invention, the working process of the data acquisition module includes: Basic air quality data is collected in real time through a sensor network deployed in the region, and personalized user data is collected through user terminals. The collection frequency is dynamically adjusted according to changes in air quality. The collection frequency is once per hour during periods of severe pollution and once every 6 hours during periods of good air quality.

[0020] As a further description of the technical solution of the present invention, the working process of the data preprocessing module includes: Missing values ​​in the basic air quality data are supplemented using linear interpolation or mean imputation, and outliers are identified and removed using the 3σ criterion. User-personalized feature data are classified and coded, and the classified data is converted into calculable numerical data. All preprocessed data sources are normalized to ensure that the data are on the same order of magnitude.

[0021] As a further description of the technical solution of the present invention, the working process of the health index calculation module includes: Collect n important pollutants, for the i-th pollutant Indicates the current time concentration, This indicates that the i-th pollutant is present in the current set time period. The maximum value, Indicates the current set time period The rate of change of , where, ; The Air Quality Health Index is calculated by combining the current pollutant concentration, the peak pollutant concentration over the current set time period, and the rate of change of pollutant concentration over the current set time period. ; In the formula, These are the reference concentrations for the i-th pollutant set by the system. The i-th pollutant set for the system within a set time period The reference rate of change within, The indicator is the trend of change of the i-th pollutant. The weighting coefficient is the one corresponding to the i-th pollutant. , and These are the weighting coefficients corresponding to real-time values, peak values, and trends of change, respectively. It is the peak attenuation coefficient. This represents the air quality health index at the current moment.

[0022] Through the above technical solution, this embodiment constructs a dynamic health risk quantification model that integrates current intensity, short-term historical peak values, and dynamic change trends. This model first targets the n important pollutants monitored (such as PM2.5, ...). , Parallel processing is performed on the i-th pollutant: its real-time concentration at the current time t is captured simultaneously. During the current preset observation period The highest concentration (e.g., the highest concentration in the past 6 hours) and the average rate of change during that time period. This allows for a complete characterization of three key features of pollutants over time: their state points, extreme peaks, and directions of change. Based on this, the module uses formulas... By scientifically integrating and normalizing these three characteristics, a comprehensive air quality health index is ultimately generated. In the formula, the first term This represents the immediate health risk of current exposure, compared with the system's preset pollutant health reference concentration. (Typically based on environmental quality standards or epidemiological thresholds) to assess the current level of hazard; the second item This introduces the concept of peak health impact residue, which not only considers the highest concentration reached over a period of time, but also simulates the lag and decay effects of high-concentration exposure on health through a peak decay coefficient λ (assigned based on the time elapsed since the peak, such as 1, 0.5, 0.2), allowing the index to remember recent pollution shock events; the third item The influence of the trend indicator G (based on the rate of change) and changing acceleration (Calculated) quantifies the dynamic evolution trend of pollutants—when pollutant concentration rises rapidly (i.e. and When all three characteristics are positive and relatively large, the G value will increase significantly even if the current absolute concentration has not yet reached an extremely high value, thus providing an early warning of potential and impending escalation of risks. Conversely, under a trend of rapid improvement, the G value can moderately reduce the risk score, reflecting a positive signal of environmental improvement. Ultimately, the risk components of the three characteristics are weighted by pollutants. and feature weights , and After being adjusted and aggregated, a dynamic and forward-looking health risk index is formed that can reflect the current state of pollution and capture its historical peaks and future evolution trends, providing a more sensitive and accurate quantitative basis for subsequent graded early warning.

[0023] As a further description of the technical solution of the present invention, the peak attenuation coefficient The value is assigned based on the time when the peak occurs, if the time of the peak occurrence is different from the current time. The interval is less than ,but The value is 1 if the peak occurs at a time that is 1 from the current time. The interval is greater than or equal to Less than ,but The value is 0.5. If the peak occurs at a time that is 0.5 away from the current time... The interval is greater than or equal to ,but The value is 0.2, and the peak value is 0.2 from the current time. The farther away, the smaller the impact.

[0024] Through the above technical solution, this embodiment determines the time interval between the peak occurrence time and the current time t, and compares it with a preset time period. and segmentation parameters (e.g., K=3) Compare and implement a step-wise assignment: if the peak value has just occurred (interval less than...) If λ reaches its maximum value of 1, it indicates that the health effects are completely preserved; if the peak has occurred for a period of time (interval between 1 and 2), then λ takes the maximum value of 1, indicating that the health effects are completely preserved; and If the peak value is between [a certain value] and [a certain value], then λ decays to 0.5, indicating that the impact is halved; if the peak value is far from the present (interval greater than or equal to [a certain value]), ... If λ further decreases to 0.2, it indicates that the impact has been significantly reduced. Through this piecewise decay model, the system ensures that the health index not only focuses on the current instantaneous pollution level, but also scientifically reflects and remembers short-term high-exposure events that occurred over a period of time in the past, making risk assessment more continuous and in line with the basic law of exposure dose-time effect in toxicology.

[0025] As a further description of the technical solution of the present invention, the calculation process of the indicator of the influence of the change trend of the i-th pollutant includes: Obtain the current concentration of the i-th pollutant. Initial concentration for the current set time period Concentration at the midpoint of the current set time period The acceleration of change of the i-th pollutant in the current time period can be calculated using the following formula. : ; Let be the rate of change of the i-th pollutant in the current time period. Let be the acceleration of the change of the i-th pollutant in the current time period, so ,in, and These are the weighting coefficients corresponding to the rate of change and the acceleration, respectively. and The system is set to operate within a specified time period. The reference rate of change and acceleration within.

[0026] Using the above technical solution, this embodiment first obtains three key concentration values: the current time, the start point of the time period, and the midpoint. Then, it uses a formula... The system accurately calculates the acceleration of pollutant changes within the current time period, revealing whether the concentration change is accelerating or decelerating. Subsequently, the calculated rate of change and acceleration are compared with preset reference values, and finally, a weighted formula is used for the result. Synthetic single trend-influence indicator. This mechanism allows the health index to not only reflect the current level of pollutants, but also to keenly capture their changing trends: when the concentration rises rapidly (to (positive and large) and deteriorating rapidly ( When the concentration is positive, the G value increases significantly, thus providing an early warning of the risk escalation trend before the absolute concentration value reaches the high-risk threshold; conversely, under an improving trend, the risk score can be appropriately reduced.

[0027] As a further description of the technical solution of the present invention, the working process of the hierarchical early warning module includes: The health risk levels are categorized as follows: Good (low health risk, positive trend), Lightly Polluted (increased risk for sensitive groups, warning issued if trend worsens), Moderately Polluted (increased health risk for all, significant peak impact), and Heavy or worse (serious health risk, requiring emergency protection). Each pollution level has a corresponding air quality health index threshold range. The current air quality health index is then used to determine the appropriate level. Map the values ​​to the corresponding threshold ranges to determine the initial warning level.

[0028] As a further description of the technical solution of the present invention, the working process of the personalized adaptation module includes: Personalized user data includes: age characteristics, basic respiratory or cardiovascular disease characteristics, and historical exposure accumulation characteristics; The age characteristic sensitivity index ,in, Based on user age, the criteria are: under 12 years old or over 65 years old. =1, between 12 and 65 years old =0; Sensitivity index of basic respiratory or cardiovascular disease characteristics ,in, Using categorical variables, if there is no underlying respiratory or cardiovascular disease =0, if either a basic respiratory or cardiovascular disease is present. =1, if both underlying respiratory and cardiovascular diseases exist. =2; The historical exposure cumulative characteristic sensitivity index = ,in, The cumulative effect coefficient set for the system. The cumulative exposure time within the currently set time period. The safety threshold for exposure time within the current set time period is set for the writing system; Sensitivity indicators for age characteristics and sensitivity indicators for baseline respiratory or cardiovascular disease characteristics, respectively. The correction factor is obtained by weighting and summing the sensitivity index of historical exposure cumulative characteristics. Therefore, personalized air quality health index for users ,Will Map the data to the corresponding threshold range to determine the final warning level.

[0029] Using the above technical solution, this embodiment first acquires and processes three key user characteristic data: age characteristics, basic respiratory or cardiovascular disease characteristics, and historical exposure accumulation characteristics. For the age characteristic, the system uses binarization processing to assign risk coefficients to children under 12 years old and the elderly over 65 years old. =1, while healthy young adults aged 12 to 65 are assigned a value of 0, directly identifying age-related vulnerable groups. For disease characteristics, a graded quantification strategy is used; if no relevant disease is present... =0, when suffering from an underlying respiratory or cardiovascular disease =1, if both are present, then 2. This accurately reflects the multiplicative effect of different disease burdens on pollution sensitivity. Historical exposure cumulative characteristics are expressed through a formula. = Perform calculations, where This refers to the user's actual cumulative exposure time within the currently set time period. The safety duration threshold set for the system, The cumulative effect coefficient was used to simulate the cumulative health risk from continuous exposure. Finally, the module weighted and summed the three sensitivity indices to obtain a comprehensive correction factor. This factor was then compared with the general health index. Multiply by each product to obtain a personalized air quality health index. This ensures that the final output accurately reflects the actual health threat level to this specific individual under this polluted environment, thus laying a scientific foundation for generating differentiated warning levels and customized protection recommendations.

[0030] A tiered, personalized early warning method based on the air quality health index, the method comprising the following steps: Step S1: Collect basic air quality data and personalized user characteristic data of various pollutants in real time at a dynamic frequency through a sensor network deployed in the target area. Step S2: Preprocess the raw data collected in step S1; Step S3: Calculate the general air quality health index based on the preprocessed air quality baseline data; Step S4: Initial warning level determination; Step S5: Based on various user-personalized data, correct the general air quality health index to a personalized health index; Step S6: Determine the personalized early warning level.

[0031] For ease of understanding, examples are provided for embodiments of the present invention: User Background User A, a 68-year-old retired senior citizen, has a history of chronic obstructive pulmonary disease (COPD) and belongs to a typical high-risk and sensitive group.

[0032] Detailed Explanation of Implementation Steps Step S1: Data Acquisition Environmental data: The sensor network deployed in User A's community collected the following PM2.5 data (µg / m³) at a frequency of once per hour (due to the pollution period): t-6h(09:00): 35 (Good) t-5h (10:00): 85 (lightly polluted) t-4h (11:00): 120 (Moderate pollution) t-3h (12:00): 150 (peak, moderate pollution) t-2h (13:00): 130 (Moderate pollution) t-1h (14:00): 110 (light pollution) t (current 15:00): 95 (light pollution) Personal Data: User A submitted personal information via the mobile app: Age = 68 years old, Disease = COPD. Simultaneously, the system automatically recorded his cumulative exposure time over the past 6 hours. Step S2: Data Preprocessing (S2) The PM2.5 data sequence above was examined and found to be free of missing values ​​and outliers, so it can be used directly for calculation.

[0033] Encode User A's personal information into numerical values: Age Feature Marker =1 (>65 years old), disease characteristic marker =1 (There is a respiratory disease).

[0034] Calculate the cumulative exposure time T: In the past 6 hours, starting from 10:00, the air pollution has always been above the light pollution threshold (75µg / m³), so T = 5 hours (300 minutes).

[0035] Step S3: Calculate the general health index Taking PM2.5 as an example (assuming it is the only pollutant), =1), System preset parameters: =75, =150, =0.5, =0.3, =0.2.

[0036] Calculate the peak attenuation coefficient K=3, the peak interval (3h) satisfies the interval within and Between, therefore =0.5.

[0037] Trend Item rate of change =(95-35) / 6=10µg / m³ / h Acceleration needs to be calculated. The concentration at the midpoint is taken as t-3h. =150.

[0038] Change in acceleration =95-2*150+35=-170 (a negative value, indicating that the concentration has been decreasing rapidly over the past 3 hours).

[0039] Reference value =20, =50, with weights of 0.5 and 0.5 respectively. =0.5*(10 / 20)+0.5*(-170 / 50)=0.25-1.7=-1.45 (negative value, indicating a strong improvement trend).

[0040] S(t)=0.5*1.27+0.3*0.5+0.2*(-1.45)=0.635+0.15-0.29=0.495 Interpretation: The general index is 0.495. The system mapping rule is: [0, 0.5) Good, [0.5, 1) Slightly polluted..., so the current general warning level is critically close to "Good", nearing the edge of "Slightly polluted".

[0041] Step S4: Personalized Correction System defaults: =0.2, =180 minutes (3-hour safety threshold), weights θ1=0.4, θ2=0.4, θ3=0.2.

[0042] Calculate the sensitivity of each feature: =1, =1; = =0.2*(300 / 180)=0.2*1.67=0.33 Calculate the correction factor F: F=0.4*1+0.4*1+0.2*0.33=0.4+0.4+0.066=0.866=1.73 Interpretation: Since User A is 68 years old and has COPD, his personal risk adjustment factor is as high as 1.73.

[0043] Calculate the personalization index : =1.73 * 0.495 = 0.856 Step 5: Final Warning Judgment General Warning (Initial Level): =0.495 → The mapping is good, but the trend needs to be monitored.

[0044] Personalized alerts (final level): =0.856 → Mapped to the range of light pollution risk.

[0045] The system has determined that, based on the principle of applying the highest level of alert, the personalized warning level is higher. The system has determined the final warning level to be light pollution.

[0046] It should be noted that the thresholds, intervals, and weights involved in this application are all empirical values, and the selection should be made by those skilled in the art based on the actual situation.

[0047] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A tiered personalized early warning system based on the air quality health index, characterized in that, The system includes a data acquisition module, a health index calculation module, a hierarchical early warning module, and a personalized adaptation module. The data acquisition module is used to collect basic air quality data and personalized user characteristic data; The data preprocessing module, connected to the data acquisition module, is used to clean, denoise, standardize, and fuse the collected basic air quality data and user personalized feature data. The health index calculation module is connected to the data preprocessing module and is used to calculate the air quality health index based on the preprocessed air quality baseline data. The tiered early warning module is connected to the health index calculation module. It is used to preset multi-level early warning thresholds and compare the personalized air quality health index with the preset multi-level early warning thresholds to determine the initial early warning level. The personalized adaptation module is connected to the health index calculation module and the hierarchical early warning module respectively. It performs real-time correction of the air quality health index based on personalized feature data to obtain a personalized air quality health index. The initial early warning level is adjusted according to the correction result to obtain the final early warning level.

2. The hierarchical personalized early warning system based on the air quality health index according to claim 1, characterized in that, The data acquisition module operates as follows: Basic air quality data is collected in real time through a sensor network deployed in the region, and personalized user data is collected through user terminals. The collection frequency is dynamically adjusted according to changes in air quality. The collection frequency is once per hour during periods of severe pollution and once every 6 hours during periods of good air quality.

3. The hierarchical personalized early warning system based on the air quality health index according to claim 1, characterized in that, The data preprocessing module operates as follows: Missing values ​​in the basic air quality data are supplemented using linear interpolation or mean imputation, and outliers are identified and removed using the 3σ criterion. User-personalized feature data are classified and coded, and the classified data is converted into calculable numerical data. All preprocessed data sources are normalized to ensure that the data are on the same order of magnitude.

4. The hierarchical personalized early warning system based on the air quality health index according to claim 1, characterized in that, The working process of the health index calculation module includes: Collect n important pollutants, for the i-th pollutant Indicates the current time concentration, This indicates that the i-th pollutant is present in the current set time period. The maximum value, Indicates the current set time period The rate of change of , where, ; The Air Quality Health Index is calculated by combining the current pollutant concentration, the peak pollutant concentration over the current set time period, and the rate of change of pollutant concentration over the current set time period. ; In the formula, These are the reference concentrations for the i-th pollutant set by the system. The i-th pollutant set for the system within a set time period The reference rate of change within, The indicator is the trend of change of the i-th pollutant. The weighting coefficient is the one corresponding to the i-th pollutant. , and These are the weighting coefficients corresponding to real-time values, peak values, and trends of change, respectively. It is the peak attenuation coefficient. This represents the air quality health index at the current moment.

5. The hierarchical personalized early warning system based on the air quality health index according to claim 4, characterized in that, The peak attenuation coefficient The value is assigned based on the time when the peak occurs, if the time of the peak occurrence is different from the current time. The interval is less than ,but The value is 1 if the peak occurs at a time that is 1 from the current time. The interval is greater than or equal to Less than ,but The value is 0.

5. If the peak occurs at a time that is 0.5 away from the current time... The interval is greater than or equal to ,but The value is 0.2, and the peak value is 0.2 from the current time. The farther away, the smaller the impact.

6. The hierarchical personalized early warning system based on the air quality health index according to claim 4, characterized in that, The calculation process for the indicator of the influence of the changing trend of the i-th pollutant includes: Obtain the current concentration of the i-th pollutant. Initial concentration for the current set time period Concentration at the midpoint of the current set time period The acceleration of change of the i-th pollutant in the current time period can be calculated using the following formula. : ; Let be the rate of change of the i-th pollutant in the current time period. Let be the acceleration of the change of the i-th pollutant in the current time period, so ,in, and These are the weighting coefficients corresponding to the rate of change and the acceleration, respectively. and The system is set to operate within a specified time period. The reference rate of change and acceleration within.

7. The hierarchical personalized early warning system based on the air quality health index according to claim 4, characterized in that, The working process of the hierarchical early warning module includes: The air quality health index (AQI) levels are categorized as: Good, Lightly Polluted, Moderately Polluted, and Heavily Polluted or Worst. Each pollution level has a corresponding AQI threshold range. The current AQI is then used to determine the air quality health index. Map the values ​​to the corresponding threshold ranges to determine the initial warning level.

8. The hierarchical personalized early warning system based on the air quality health index according to claim 1, characterized in that, The operation process of the personalized adaptation module includes: Personalized user data includes: age characteristics, basic respiratory or cardiovascular disease characteristics, and historical exposure accumulation characteristics; The age characteristic sensitivity index ,in, Based on user age, the criteria are: under 12 years old or over 65 years old. =1, between 12 and 65 years old =0; Sensitivity index of basic respiratory or cardiovascular disease characteristics ,in, Using categorical variables, if there is no underlying respiratory or cardiovascular disease =0, if either a basic respiratory or cardiovascular disease is present. =1, if both underlying respiratory and cardiovascular diseases exist. =2; The historical exposure cumulative characteristic sensitivity index = ,in, The cumulative effect coefficient set for the system. The cumulative exposure time within the currently set time period. The safety threshold for exposure time within the current set time period is set for the writing system; Sensitivity indicators for age characteristics and sensitivity indicators for baseline respiratory or cardiovascular disease characteristics, respectively. The correction factor is obtained by weighting and summing the sensitivity index of historical exposure cumulative characteristics. Therefore, personalized air quality health index for users ,Will Map the data to the corresponding threshold range to determine the final warning level.

9. A hierarchical personalized early warning method based on the Air Quality Health Index, said method being implemented based on the hierarchical personalized early warning system based on the Air Quality Health Index as described in any one of claims 1-8, characterized in that, The method includes the following steps: Step S1: Collect basic air quality data and personalized user characteristic data of various pollutants in real time at a dynamic frequency through a sensor network deployed in the target area. Step S2: Preprocess the raw data collected in step S1; Step S3: Calculate the general air quality health index based on the preprocessed air quality baseline data; Step S4: Initial warning level determination; Step S5: Based on various user-personalized data, correct the general air quality health index to a personalized health index; Step S6: Determine the personalized early warning level.

Images