Plateau aviation occupational health comprehensive risk assessment and monitoring system based on pollution source

Abstract

The invention discloses a pollution source-based comprehensive risk assessment and monitoring system for plateau aviation occupational health, and relates to the technical field of risk assessment. The system comprises a pollution source identification module, an individual exposure evaluation module, a physiological state analysis module, a comprehensive evaluation prediction module and a health intervention suggestion module. The technical key points are as follows: according to the scheme, multi-parameter fusion monitoring points are deployed, a three-dimensional pollution map is constructed, and a plateau height and meteorological correction function is introduced, so that the overall scheme realizes spatial distribution modeling and comprehensive strength quantification of pollution sources in complex environments such as plateau airports; the problem of inaccurate pollution identification caused by traditional static data and experience judgment is solved, a high-precision and dynamically-updated pollution data basis is provided for subsequent exposure evaluation, the real-time sensing ability of the system to environmental risks is improved, and the accuracy and intelligence of the system in an environmental sensing layer are embodied.

Description

Technical Field

[0001] This invention relates to the field of risk assessment technology, specifically to a comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources. Background Technology

[0002] Risk assessment refers to the identification and analysis of various potential hazards to evaluate the likelihood and severity of their adverse effects on individuals or groups. In the field of occupational health in high-altitude aviation, risk assessment mainly focuses on the comprehensive judgment of health risks to aviation practitioners caused by long-term exposure to air pollutants, high-intensity work, and physiological load in complex environments such as high altitude, low oxygen, and strong ultraviolet radiation. Traditional assessment methods rely on static data and experience-based judgment, usually based on pollution concentrations at fixed monitoring points, occupational exposure limits, and basic personnel information, using simple weighting or statistical methods to score and classify risk levels to guide the development of health management and intervention measures.

[0003] However, in high-altitude aviation operations, due to the high altitude, thin atmosphere, and complex weather conditions, coupled with the diverse distribution of pollution sources around some airports, traditional occupational health management systems struggle to accurately grasp individual exposure levels and changes in health risks. This can easily lead to overlooked health hazards or even sudden health incidents. For example, a ground staff member was exposed to a high PM2.5 concentration area for an extended period, but due to a lack of dynamic monitoring and individualized assessment, the persistently low blood oxygen levels were not detected in time, ultimately exacerbating altitude sickness. Traditional high-altitude aviation occupational health management systems have significant shortcomings in pollution identification, exposure assessment, and risk warning. On the one hand, pollution source identification relies heavily on fixed monitoring stations and static data analysis, which is insufficient to accurately reflect the pollution distribution characteristics under the complex terrain and meteorological conditions of high-altitude areas, such as at airports. In the surrounding area, pollutants are significantly affected by wind direction and altitude, but traditional systems cannot dynamically model this, leading to distorted exposure assessments. Furthermore, individual exposure assessments often use uniform models, neglecting personnel movement trajectories, altitude acclimatization status, and physiological differences. For example, those newly arrived at high altitudes inhale more pollutants due to increased lung ventilation, but traditional models fail to differentiate between these factors, causing assessment bias. In addition, health risk assessments often remain at the static scoring stage, lacking comprehensive consideration of physiological stress, underlying medical history, and time trends, easily leading to misjudgments or delayed warnings. Finally, traditional intervention measures lack a tiered mechanism, making it difficult to provide personalized recommendations based on risk levels, affecting the targetedness and effectiveness of health management. These problems severely restrict the scientific rigor and timeliness of health risk prevention and control for personnel working in high-altitude aviation. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: a comprehensive risk assessment and monitoring system for occupational health in high-altitude aviation based on pollution sources. This system includes: a pollution source identification module, which deploys several monitoring points in the target environment to identify the types of pollution sources and their spatial distribution characteristics; based on the calibration dataset collected from the monitoring points, an improved pollution source classification model is used to distinguish pollution sources and quantify the comprehensive pollution source intensity of each monitoring point, and to construct a three-dimensional pollution map; and an individual exposure assessment module, which, based on the three-dimensional pollution map and real-time comprehensive pollution source intensity data of each monitoring point, combined with the trajectory adaptation information set of aviation personnel, constructs an improved multi-pollutant exposure model and outputs an exposure dose report for aviation personnel, which includes at least: total exposure dose and peak exposure period. The physiological state analysis module collects physiological indicators of aviation personnel to obtain the physiological stress index. The comprehensive assessment and prediction module, based on the total exposure dose and physiological stress index of the target aviation personnel, establishes a multi-dimensional health risk assessment model within a set time period, combined with personnel registration information. The output result is the health risk index of the target aviation personnel in the Qth monitoring period. Using a fuzzy logic rule engine, the module determines the preliminary risk level of the target aviation personnel based on the output result, and performs multiple adjustments to the preliminary risk level based on the expert rule judgment library and analysis strategy to obtain the final risk level. The health intervention suggestion module, based on the obtained final risk level, triggers a suggestion adjustment mechanism and extracts the corresponding suggested measures from a pre-built strategy suggestion library according to the final risk level.

[0005] Furthermore, the types of pollution sources identified should include at least PM2.5, NO2, O3, CO, and UV; the data set identified should include at least the concentration data of various pollutants, meteorological data, and altitude data.

[0006] Furthermore, the improved pollution source classification model is applied as follows: based on the cleaned and calibrated dataset, a plateau altitude correction function and a meteorological influence correction function are introduced, and dynamic weights are set for different pollutants. The method used is: P i =(∑(w q *C i，x ))*F_al(H)*F_me(W, T, Hu); where, P i w represents the overall pollution source intensity at the i-th monitoring point. q C represents the corresponding weight of the concentration data of the qth pollutant. i，xLet F_al(H) represent the concentration data of pollutant of type x at the i-th monitoring point, where type x can represent any of the pollutant concentration data; F_al(H) represents the plateau altitude correction function, and F_me(W, T, Hu) represents the meteorological influence correction function; i = 1, 2, ..., n, where n is the number of monitoring points; where F_al(H) = 1 + k_al * ((H - H0) / Hmax); where k_al represents the empirical correction coefficient; H, H0, and Hmax represent the actual altitude, the reference altitude, and the maximum altitude of the target environment area, respectively; F_me(W, T, Hu) = e -k_w*W *(1+k_t*(T-T0))*(1+k_h*Hu); where k_w, k_t, and k_h are all corresponding empirical coefficients; W represents wind speed, T and T0 represent temperature and standard temperature, respectively, and Hu represents relative humidity.

[0007] Furthermore, the trajectory adaptation information set includes at least: the dynamic movement trajectory of aviation practitioners, stay time information, basic information data, and meteorological and altitude data in the calibration dataset; among which, the basic information data includes at least: whether it is the first time to arrive at the plateau and past health history.

[0008] Furthermore, the process of running the improved multi-pollutant exposure model is as follows: Trajectory matching: The movement paths of aviation personnel are spatiotemporally matched with monitoring points on the 3D pollution map; Time integration: Time slices are divided at 2-minute intervals, and the instantaneous exposure intensity within each time slice is calculated; Weighting: WHO standards are adopted to assign weights based on the hazard levels of different types of pollutants; Plateau and meteorological correction: The plateau altitude correction function and meteorological influence correction function are used to accumulate the instantaneous exposure intensity (exposure dose) for aviation personnel in each time slice to complete the correction action; Accumulation and report generation: The exposure doses of all time slices are accumulated to obtain the total exposure dose of aviation personnel throughout the entire monitoring period, and an individual physiological adaptation correction factor is introduced to generate an exposure dose report.

[0009] Furthermore, the improved multi-pollutant exposure model is as follows: the improved total exposure dose of the j-th aviation worker is obtained by multiplying the initial total exposure dose of the j-th aviation worker by the individual physiological adaptation correction factor; wherein, when the constraint is: if j is a newcomer to high altitude, the individual physiological adaptation correction factor is: the sum of 1 and the exposure sensitivity enhancement coefficient of newcomers to high altitude; when the constraint is: if j is not a newcomer to high altitude, the individual physiological adaptation correction factor is: 1.

[0010] Furthermore, the physiological indicators include at least heart rate, blood oxygen saturation, blood pressure, and respiratory rate. The physiological stress index of each aviation worker is obtained by weighted summation based on the physiological indicators.

[0011] Furthermore, the expert rule judgment library and analysis strategy are as follows: the preliminary risk level is adjusted once using the preset expert rule judgment library to obtain the secondary risk level; and the risk curves during the historical monitoring period are analyzed in the time series, and the results of the analysis are used to determine whether to perform a second adjustment to the secondary risk level to obtain the final risk level.

[0012] Furthermore, the process of analyzing the risk curve is as follows: Input variables: secondary risk level, time series data, and a set of evaluation indicators for the analysis dimensions; whereby, the secondary risk level includes: low, medium, high, and urgent, and the time series data includes: health risk indices R1, R2, ..., R for N historical monitoring periods. N The analysis dimension evaluation indicator set includes: risk change rate trend, maximum risk value, average risk value, and risk volatility value; trend feature extraction: comparing whether the features of each indicator in the analysis dimension evaluation indicator set exceed the limits; among them, if the risk change rate continues to rise, it indicates that the risk change rate trend exceeds the limit; when R exists... Q A value ≥0.9 indicates that the maximum risk value has exceeded the limit; an average risk value exceeding 0.6 indicates that the average risk value has exceeded the limit; a standard deviation exceeding the preset standard threshold indicates that the risk volatility value has exceeded the limit; the secondary adjustment action trigger mechanism: whether the number of indicators exceeding the limit in the analysis dimension evaluation indicator set exceeds the set value. If it exceeds the set value, a secondary adjustment action is triggered, raising the risk level by one level based on the original risk level, and using the raised risk level as the final risk level; if it does not exceed the set value, the original risk level is used as the final risk level.

[0013] This invention provides a comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources, which has the following beneficial effects: 1) By deploying monitoring points with multi-parameter fusion and constructing a three-dimensional pollution map, and by introducing plateau altitude and meteorological correction functions, this solution enables the overall solution to model the spatial distribution of pollution sources and quantify the comprehensive intensity in complex environments such as high-altitude airports. This not only overcomes the problem of inaccurate pollution identification caused by traditional static data and experience-based judgment, but also provides a high-precision, dynamically updated pollution data foundation for subsequent exposure assessment, improves the system's real-time perception capability of environmental risks, and reflects the accuracy and intelligence of the system at the environmental perception layer.

[0014] 2) This scheme performs spatiotemporal matching of individual movement trajectories with pollution maps, and combines WHO standards with individual physiological adaptation correction factors to construct a dynamic, multi-pollutant, and personalized exposure dose assessment model. On the one hand, this model solves the problem of ignoring the physiological adaptation status and individual differences at high altitudes in traditional exposure assessments. On the other hand, it provides a scientific and quantifiable input basis for subsequent health risk modeling, demonstrating the system's advanced nature in data fusion and individual modeling.

[0015] 3) This solution introduces a neural network model, a fuzzy logic rule engine, an expert rule base, and a time series analysis mechanism to construct a full-chain assessment system from exposure dose and physiological stress to risk level. This system not only achieves a preliminary judgment of risk level, but also performs two adjustments based on expert correction and historical trends, improving the accuracy of the judgment results. It also solves the problem of lack of dynamic tracking and trend prediction in traditional risk assessment, significantly improving the scientific nature and foresight of early warning, and demonstrating the system's deep integration capabilities in intelligent decision-making and risk evolution modeling.

[0016] 4) This solution combines the output of the final risk level and automatically matches the corresponding intervention recommendations from the preset strategy recommendation library. It covers multiple levels, including routine monitoring, work adjustment, oxygen inhalation recommendations, medical intervention and emergency treatment. The application of this mechanism not only solves the problem of delayed response in traditional occupational health intervention, but also realizes a complete closed-loop management from assessment to early warning to intervention, improves the pertinence and response efficiency of health intervention, and provides personalized and sustainable health management services for high-altitude aviation practitioners. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the module operation of the pollution source-based plateau aviation occupational health comprehensive risk assessment and monitoring system in this invention. Detailed Implementation

[0018] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1This embodiment provides a comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources. The system is designed for high-altitude areas where aviation professionals, such as pilots, flight attendants, and ground staff, face high occupational health risks due to high altitude, low atmospheric pressure, thin oxygen, and complex distribution of environmental pollutants. Traditional occupational health management relies heavily on static data and experience-based judgment, lacking dynamic and real-time integrated analysis of pollution exposure and individual physiological responses. The system proposed and designed in this embodiment integrates multi-source data acquisition, dynamic risk modeling, and intelligent early warning and feedback mechanisms to achieve all-weather health monitoring and personalized intervention recommendations for aviation professionals.

[0020] This system adopts a hierarchical architecture, which is divided into several sequentially running functional modules, including: 1. Pollution source identification module: Deploy several monitoring points in the target environment, identify the types of pollution sources and their spatial distribution characteristics, and based on the calibration dataset collected from the monitoring points, use an improved pollution source classification model to distinguish pollution sources and quantify the comprehensive pollution source intensity of each monitoring point, and construct a three-dimensional pollution map.

[0021] The target environment could be near an airport in a high-altitude region.

[0022] The actual number of monitoring points is set according to actual needs, ensuring even distribution. Each monitoring point is equipped with at least an environmental sensor network, a GPS module, and a weather station. The environmental sensor network uses LoRa low-power wide-area network communication to support remote deployment and is mainly used to collect real-time data on the concentration of various pollutants, including PM2.5, NO2, O3, CO, and UV. Of course, the specific types of pollutants may be much more than these; this embodiment only uses the types of pollutants given above. The GPS module is used to ensure that the location of the corresponding monitoring point is fixed, facilitating appropriate handling in case of subsequent location shifts or loss. The weather station is a small or micro weather station used to acquire meteorological data such as wind speed, wind direction, relative humidity, and temperature.

[0023] The identified pollution source type is to distinguish between the various pollutants mentioned above.

[0024] By incorporating AI image recognition algorithms into satellite remote sensing imagery, large-scale environmental information can be acquired, thereby determining the specific characteristics of spatial distribution and automatically identifying highly polluted areas and abnormally hot zones. When constructing a 3D pollution map, a GIS positioning engine is used to build a 3D spatial coordinate system, i.e., the required 3D polluted area, enabling spatial positioning of pollution sources. It also supports multi-layer map overlay, such as traffic, wind direction, and altitude, improving the accuracy of geographic analysis. The real-time comprehensive intensity of pollution sources at each monitoring point can also be reflected on this 3D pollution map.

[0025] The calibration dataset includes: various pollutant concentration data, meteorological data, and altitude data. The pollutant concentration data and meteorological data have been clearly explained and introduced above. The altitude data includes the baseline altitude, the actual altitude, and the maximum altitude of the target environment region. Specifically, traditional pollution source classification models are usually based on various pollutant concentration data, using a weighted summation method to generate the comprehensive pollution source intensity for the corresponding monitoring point. However, the improved pollution source classification model used in this embodiment is based on the cleaned calibration dataset. To more accurately reflect the health impact of pollution sources in the target environment (e.g., high-altitude areas), it considers introducing a high-altitude correction function and a meteorological impact correction function, and sets dynamic weights for different pollutants. The method used is: P i =(∑(w q *C i，x ))*F_al(H)*F_me(W, T, Hu); where, P i w represents the overall pollution source intensity at the i-th monitoring point. q This represents the weight of the concentration data for the q-th pollutant, where the maximum value of q is the number of different types of pollutant concentration data, i.e., X, C. i，x Let F_al(H) represent the concentration data of pollutant of type x at the i-th monitoring point, where type x can represent any of the pollutant concentration data, and the unit is μg / m³; F_al(H) represents the plateau altitude correction function, and F_me(W, T, Hu) represents the meteorological influence correction function; i = 1, 2, ..., n, where n is the number of monitoring points; where F_al(H) = 1 + k_al * ((H - H0) / Hmax); where k_al represents the empirical correction coefficient, reflecting the influence of the thin air of the plateau on the deposition and diffusion of pollutants, and the value range is [0, 1]; H, H0 and Hmax represent the actual altitude, the reference altitude and the maximum altitude of the target environment area, respectively.

[0026] F_me(W, T, Hu) = e -k_w*W*(1+k_t*(T-T0))*(1+k_h*Hu); where k_w, k_t, and k_h are all corresponding empirical coefficients, with values ​​ranging from [0, 1]; W represents wind speed, T and T0 represent temperature and standard temperature (e.g., 20℃) respectively, and Hu represents relative humidity. This meteorological influence correction function is used to simulate the effects of wind speed on pollutant diffusion, temperature on chemical reaction rate, and relative humidity on particulate matter deposition.

[0027] Results: By deploying multi-parameter fusion monitoring points and constructing a 3D pollution map, and by introducing plateau altitude and meteorological correction functions, the overall solution enables spatial distribution modeling and comprehensive intensity quantification of pollution sources in complex environments such as plateau airports. Utilizing an improved pollution source classification model, it not only overcomes the problem of inaccurate pollution identification caused by traditional static data and experience-based judgments, but also provides a high-precision, dynamically updated pollution data foundation for subsequent exposure assessments. It solves the problem of ambiguous pollution distribution caused by high altitude and complex meteorological conditions in plateau areas, enhances the system's real-time perception capability of environmental risks, and demonstrates the system's accuracy and intelligence at the environmental perception layer.

[0028] II. Individual Exposure Assessment Module: Based on the 3D pollution map and real-time comprehensive pollution source intensity data of each monitoring point provided in the previous module, and combined with the trajectory adaptation information set of aviation personnel (which serves as the input data source), an improved multi-pollutant exposure model is constructed to quantitatively assess the pollutant exposure level of each aviation personnel in the target environment and output an exposure dose report for each aviation personnel. The trajectory adaptation information set includes: the dynamic movement trajectory of aviation personnel, residence time information, basic information data, and meteorological and altitude data from the calibration dataset. Specifically, the dynamic movement trajectory of aviation personnel is obtained using GPS+UWB fusion positioning technology, and an exposure time integrator is used to calculate the residence time and corresponding exposure duration of aviation personnel in different pollution areas, i.e., residence time information. Basic information data includes: whether it is the first time arriving at high altitude and past health history, etc.

[0029] The process of running the improved multi-pollutant exposure model is as follows: S201, trajectory matching: the movement paths of aviation personnel are spatiotemporally matched with the monitoring points on the three-dimensional pollution map.

[0030] S202, Time Integration: Divide the time slices into segments with intervals of 2 seconds to calculate the instantaneous exposure intensity within each time slice.

[0031] S203, Weighted Approach: Using WHO standards, weights are assigned based on the hazard level of different types of pollutants.

[0032] It should be noted that the weights are set based on the relative level of harm of pollutants to human health, with reference to WHO standards or current common epidemiological research results. The level of harm is positively correlated with the weight.

[0033] S204, Plateau and Weather Correction: Continue to use the plateau altitude correction function and the weather impact correction function to complete the correction action by summing the instantaneous exposure intensity of aviation practitioners in each time slice, i.e., the exposure amount.

[0034] S205. Accumulation and Report Generation: Accumulate the exposure amounts of all time slices to obtain the total exposure dose of aviation workers during the entire monitoring period, and continue to introduce individual physiological adaptation correction factors to generate an exposure dose report. The report includes: total exposure dose, peak exposure period (i.e., the period when the exposure exceeds the benchmark threshold), and the main types of exposed pollutants.

[0035] The multi-pollutant exposure models used in S201 to S204 are as follows: In the formula, E j This represents the preliminary total exposure dose for the j-th aviation worker, in μg·hr / m³; Tr represents the total sampling time; X represents the number of different types of pollutant concentration data; and C... j，t，x w represents the concentration data of pollutant type x corresponding to the location of the j-th aviation worker at time t. x This represents the weighting factor for pollutant category x, reflecting its relative degree of harm to human health; F_al(H j ) and F_me(W t T t Hu t The same principle applies as above, so I will not elaborate further here. Δt represents the time step.

[0036] Combining S205, the improved multi-pollutant exposure model is obtained as follows: E j '=E j *A j E j ' represents the improved total exposure dose for the j-th aviation worker, A j As an individual physiological adaptation correction factor, when the constraint is: if j is a newcomer to high altitude, then A j =1+a; a represents the exposure sensitivity enhancement coefficient for those newly arrived at high altitudes, with a value range of [0.1, 0.3]. Since those newly arrived at high altitudes inhale more pollutants due to increased respiratory rate and lung ventilation, this coefficient needs adjustment. When the constraint is: if j is not a newcomer to high altitudes (i.e., a long-term resident), then A j =1.

[0037] Results: Based on GPS+UWB fusion positioning technology and time integration algorithm, the system spatiotemporally matches individual movement trajectories with pollution maps. Combining WHO standards and individual physiological adaptation correction factors, it constructs a dynamic, multi-pollutant, and personalized exposure dose assessment model. On one hand, this model solves the problem of neglecting high-altitude physiological adaptation status and individual differences in traditional exposure assessments. On the other hand, it provides scientific and quantifiable input for subsequent health risk modeling, realizing the transformation from group exposure to individual exposure, improving the accuracy and practicality of health risk assessment, and demonstrating the system's advanced nature in data fusion and individual modeling.

[0038] III. Physiological State Analysis Module: This module collects real-time physiological indicators of aviation personnel and uses a weighted summation method to obtain the physiological stress index. The collection of these indicators is accomplished through wearable devices, such as heart rate monitors, pulse oximeters, and pressure sensors. Therefore, the physiological indicators include heart rate, blood oxygen saturation, blood pressure, and respiratory rate. The physiological stress index for each aviation personnel is obtained by weighted summation of these indicators, reflecting their current stress level. For the physiological stress index over the entire monitoring period, the standard deviation can be used as the statistical unit. The calculations, for example, for heart rate, summarize the heart rate at each time point within the entire monitoring period and calculate its standard deviation as a physiological indicator. The same applies to other types of physiological indicators, so they will not be elaborated on here. In addition, a corresponding standard threshold is set for each type of data in the physiological indicators. For example, if an aviation worker is a ground staff member, and after continuous work, his heart rate rises to 110 beats / minute and his blood oxygen saturation drops to 88%, which is lower than the normal range of blood oxygen saturation (the corresponding standard threshold), an early warning will be triggered to detect potential health problems in advance.

[0039] IV. Comprehensive Assessment and Prediction Module: Based on the total exposure dose and physiological stress index of the target aviation practitioners, a multi-dimensional health risk assessment model is established within a set time period, combined with personnel registration information. The output results are: the health risk index of the target aviation practitioners in the Qth monitoring period. Using a fuzzy logic rule engine, the initial risk level of the target aviation practitioners is determined based on the output results. Combined with the expert rule judgment library, the initial risk level is adjusted once to obtain the secondary risk level. In the time series, the risk curves of historical monitoring periods are analyzed. Based on the analysis results, it is decided whether to perform a second adjustment on the secondary risk level to obtain the final risk level.

[0040] Specifically, the personnel file registration information includes age correction factor, medical history correction factor, gender coefficient, number of days of high-altitude acclimatization, and time period coefficient. This personnel file registration information serves as an input variable, and a setting submodule is built into the multi-dimensional health risk assessment model. For the age correction factor: when the actual age is <30, the age correction factor is 1.0; when the actual age is between 30 and 50, the age correction factor is 1.1; when the actual age is >50, the age correction factor is reduced to 1.2. For the medical history correction factor: when there is a pre-existing medical history, the medical history correction factor is 1.2; when there is no pre-existing medical history, the medical history correction factor is 1, with no correction. Pre-existing medical histories include asthma and hypertension, etc. For the gender coefficient: When the target aviation worker is male, the gender coefficient is set to 1; when female, it is set to 1.1. This is because women are more susceptible to certain pollutants. Regarding the number of days of acclimatization to high altitude: if the actual number of acclimatization days is less than 7, it indicates that the person is highly sensitive; if the actual number of acclimatization days is greater than 30, it indicates that the person is less sensitive. Regarding the time period coefficient: it reflects the difference in the impact of different time periods, such as non-night shifts and night shifts, on human health. Night shifts pose a greater health hazard to the human body. Therefore, when the person is on a non-night shift, the time period coefficient is set to 1; when the person is on a night shift, the time period coefficient is set to 1.1. (Night work has a higher overall risk due to factors such as fatigue and lower oxygen content. The specific value setting here can be adjusted according to actual needs and is positively correlated with the corresponding risk level.)

[0041] All of these input variables are pre-set according to requirements.

[0042] The multidimensional health risk assessment model operates based on a nonlinear evaluation function constructed using a neural network, combined with time-period coefficients. The neural network can be either an MLP or a CNN, and the model's operation is based on: R... Q =f NN (E j，Q Y j，Q Ar, Hi, Gb, Dt)*Tu Q In the formula, R Q f represents the health risk index during the Qth monitoring period. NN E represents a nonlinear evaluation function built upon a neural network. j，Q Y j，Q Ar, Hi, Gb, and Dt represent, respectively, the total exposure dose of the j-th aviation worker during the Q-th monitoring period, the physiological stress index of the j-th aviation worker during the Q-th monitoring period, the age correction factor Ar, the medical history correction factor Hi, the gender coefficient Gb, the number of days of altitude acclimatization Dt, and Tu. Q This represents the time period coefficient within the Qth monitoring period.

[0043] The process of determining the preliminary risk level of a target aviation professional based on the output results using a fuzzy logic rule engine is as follows: When R Q If R < 0.4, the initial risk level is low; if R ≤ 0.4, the risk level is low. Q If R < 0.7, the initial risk level is medium risk; if 0.7 ≤ R Q If R < 0.9, the corresponding preliminary risk level is high risk; when R Q If the value is ≥0.9, the initial risk level is emergency risk.

[0044] Based on the expert rule judgment database, the process of adjusting the initial risk level to obtain the secondary risk level is as follows: When the initial risk level is low, the process continues to determine whether the target aviation employee has a pre-existing medical history. If a pre-existing medical history exists, the low risk level is changed to medium risk; otherwise, the original low risk level is maintained. When the initial risk level is medium, the process continues to determine whether the target aviation employee has any abnormal physiological signals (e.g., SpO2 < 90%, which are considered routine abnormal physiological signals). If so, the medium risk level is changed to high risk; otherwise, the original medium risk level is maintained. When the initial risk level is high, the process continues to determine whether the target aviation employee exhibits any clinical symptoms. If so, the high risk level is changed to emergency risk; otherwise, the original high risk level is maintained. Clinical symptoms include obvious symptoms such as difficulty breathing and chest tightness and shortness of breath caused by arrhythmia. No adjustment is made for emergency risks.

[0045] Results: This system, by introducing neural network models, fuzzy logic rule engines, expert rule bases, and time series analysis mechanisms, constructs a full-chain assessment system from exposure dose and physiological stress to risk level. This system not only achieves preliminary risk level determination but also performs two adjustments based on expert correction and historical trends, avoiding misjudgments or reduced accuracy caused by single assessments. Furthermore, it addresses the lack of dynamic tracking and trend prediction in traditional risk assessments, significantly improving the scientific rigor and foresight of early warnings. This provides a reliable basis for developing health intervention strategies and demonstrates the system's deep integration capabilities in intelligent decision-making and risk evolution modeling.

[0046] In terms of time series, the risk curves within historical monitoring periods are analyzed, and the results are used to determine whether to make secondary adjustments to the secondary risk level in order to obtain the final risk level. The risk curves within historical monitoring periods represent the risk curves formed by the corresponding target aviation practitioners in historical or past monitoring periods.

[0047] The analysis process is as follows: S401, Input variables: secondary risk level, time series data, and a set of evaluation indicators for the analysis dimensions; where the secondary risk level includes: low, medium, high, and urgent, and the time series data includes: health risk indices R1, R2, ..., R for N historical monitoring periods. N The set of analytical dimension evaluation indicators includes: risk change rate trend, maximum risk value, average risk value, and risk volatility value. Among them, the set of analytical dimension evaluation indicators are all calculated based on risk curves. The risk change rate trend is the risk change rate calculated over N historical monitoring periods. The maximum risk value is the highest health risk index selected from the N historical monitoring periods. The average risk value is the mean of the health risk index corresponding to the N monitoring periods. The risk volatility value is the standard deviation of the health risk index corresponding to the N monitoring periods.

[0048] S402. Trend Feature Extraction: Comparative analysis assesses whether the features of each indicator in the indicator set exceed limits; for the trend of risk change rate: if the risk change rate continues to rise, it indicates that the limit has been exceeded, suggesting a worsening risk trend; conversely, it does not exceed the limit; for the maximum risk value: does R exist? Q ≥0.9: If present, it indicates an over-limit, suggesting an emergency risk has occurred; otherwise, it does not exceed the limit. For the average risk value: Does the average risk value exceed 0.6? If it does, it indicates an over-limit, suggesting an overall high risk; otherwise, it does not exceed the limit. For the risk volatility value: Does the standard deviation exceed the preset standard threshold? If it does, it indicates an over-limit, suggesting large fluctuations and an unstable health status; otherwise, it does not exceed the limit.

[0049] S403. Secondary Adjustment Trigger Mechanism: In the trend feature extraction results, the number of indicators exceeding the limit in the analysis dimension evaluation indicator set exceeds the set value. For example, in this embodiment, the set value is the mean + 1, rounded down. The above analysis dimension evaluation indicator set has 5 indicators, all of which are 2.5. 2.5 + 1 = 3.5, and the result after rounding down is 3. That is, if three or more of the above dimensions exceed the limit, a secondary adjustment action is triggered, that is, the risk level is increased by one level based on the original risk level, and the increased risk level is used as the final risk level. For example, if the original risk level is low, it becomes medium risk after being increased by one level. If the original risk level is emergency risk, the highest level (emergency risk) is maintained.

[0050] V. Health Intervention Recommendation Module: Based on the final risk level, a recommendation adjustment mechanism is triggered. Recommendations are extracted from a pre-built strategy recommendation library according to the final risk level. Specifically, when the final risk level is low, indicating stable health status for the target aviation personnel, the recommended measures include: routine monitoring, health education, and regular physical examinations. Specifically, this means continuing daily health data collection, without needing to adjust shifts or work intensity, strengthening self-protection awareness training in high-altitude environments, conducting basic health checks quarterly, and encouraging exercise and nutrition to improve physical adaptability. When the final risk level is medium, indicating mildly abnormal health status for the target aviation personnel, the recommended measures include: enhanced monitoring, work adjustments, and oxygen therapy recommendations. Specifically, this means increasing the frequency of daily health data collection, reducing high-intensity outdoor work time, appropriately using portable oxygen devices in high-altitude environments, recommending breathing exercises and cardiopulmonary function training, and avoiding consecutive night shifts or prolonged exposure. Regular follow-up by occupational health physicians is recommended. When the final risk level is "high risk," it indicates that the target aviation worker's health status is significantly abnormal. The recommended measures include: reducing work, oxygen therapy and rest, and medical intervention. Specifically, this means: suspending high-intensity or outdoor work, scheduling oxygen therapy and reducing continuous working hours, assessment by occupational health physicians, and, if necessary, arranging off-duty rest, avoiding night shifts and continuous work, real-time monitoring of key indicators such as SpO2, heart rate, and respiratory rate, and recording this high-risk event for subsequent analysis. When the final risk level is "emergency risk," it indicates that the target aviation worker has clinical symptoms. The recommended measures include: ceasing work, emergency medical care, and mandatory rest. Specifically, the worker should immediately leave the current work environment, contact the medical center located in the target environment or be sent to a hospital, notify relevant personnel and record the entire event. The worker should not continue working at high altitudes until recovery, and a professional physician should assess whether they are suitable to continue working at high altitudes. This assessment will be used as model feedback data to optimize future risk predictions.

[0051] The aforementioned risk level classification combined with corresponding recommendation mechanisms enables dynamic management and tiered intervention of the health status of personnel working in high-altitude aviation sectors. This not only ensures personnel health and safety but also improves operational efficiency and emergency response capabilities, providing strong support for the construction of a high-altitude occupational health monitoring system.

[0052] Results: Based on the output of the final risk level, the system automatically matches corresponding intervention suggestions from the preset strategy suggestion library, covering multiple levels such as routine monitoring, work adjustment, oxygen inhalation suggestions, medical intervention and emergency treatment. The application of this mechanism not only solves the problem of delayed response in traditional occupational health intervention, but also realizes a complete closed-loop management from assessment to early warning to intervention, improving the pertinence and response efficiency of health intervention, providing personalized and sustainable health management services for plateau aviation practitioners, and demonstrating the practicality of the system in the health management service layer.

[0053] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0054] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0055] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources, characterized in that, The system includes: a pollution source identification module, which deploys several monitoring points in the target environment to identify the types of pollution sources and their spatial distribution characteristics; based on the calibration dataset collected from the monitoring points, it adopts an improved pollution source classification model to distinguish pollution sources and quantify the comprehensive pollution source intensity of each monitoring point, and constructs a three-dimensional pollution map. The individual exposure assessment module, based on a 3D pollution map and real-time comprehensive pollution source intensity data for each monitoring point, combined with the trajectory adaptation information set of aviation personnel, constructs an improved multi-pollutant exposure model and outputs an exposure dose report for aviation personnel, which includes at least: total exposure dose and peak exposure period; The physiological state analysis module collects physiological indicators from aviation professionals and obtains the physiological stress index. The comprehensive assessment and prediction module, based on the total exposure dose and physiological stress index of the target aviation practitioners, establishes a multi-dimensional health risk assessment model within a set time period, combined with personnel registration information. The output result is the health risk index of the target aviation practitioners in the Qth monitoring period. Using a fuzzy logic rule engine, the module determines the preliminary risk level of the target aviation practitioners based on the output result, and combines the expert rule judgment library and analysis strategy to perform multiple adjustments to the preliminary risk level to obtain the final risk level.

2. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 1, characterized in that, The types of pollution sources to be calibrated should include at least PM2.5, NO2, O3, CO, and UV; the calibration dataset should include at least the concentration data of various pollutants, meteorological data, and altitude data.

3. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 1, characterized in that, The application process of the improved pollution source classification model is as follows: Based on the cleaned and calibrated dataset, a plateau altitude correction function and a meteorological impact correction function are introduced, and dynamic weights are set for different pollutants. The method used is: P i =(∑(w q *C i，x ))*F_al(H)*F_me(W, T, Hu); where, P i w represents the overall pollution source intensity at the i-th monitoring point. q C represents the corresponding weight of the concentration data of the qth pollutant. i，x This represents the concentration data of pollutant type x at the i-th monitoring point; F_al(H) represents the plateau altitude correction function, and F_me(W, T, Hu) represents the meteorological impact correction function; i = 1, 2, ..., n, where n is the number of monitoring points; where F_al(H) = 1 + k_al * ((H - H0) / Hmax); k_al represents the empirical correction coefficient; H, H0, and Hmax represent the actual altitude, the reference altitude, and the maximum altitude of the target environment area, respectively; F_me(W, T, Hu) = e -k_w*W *(1+k_t*(T-T0))*(1+k_h*Hu); where k_w, k_t, and k_h are all corresponding empirical coefficients; W represents wind speed, T and T0 represent temperature and standard temperature, respectively, and Hu represents relative humidity.

4. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 2, characterized in that, The trajectory adaptation information set includes at least: the dynamic movement trajectory of aviation personnel, stay time information, basic information data, and meteorological and altitude data in the calibration dataset; among which, the basic information data includes at least: whether it is the first time to arrive at the plateau and past health history.

5. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 3, characterized in that, The process of running the improved multi-pollutant exposure model is as follows: Trajectory matching: Spatiotemporally matching the movement paths of aviation personnel with monitoring points on a 3D pollution map; Time integration: Divide the time slices into segments with intervals of 2 seconds to calculate the instantaneous exposure intensity within each time slice; Weighted average: Using WHO standards, weights are assigned based on the hazard level of different types of pollutants; Plateau and meteorological correction: Continue to use the plateau altitude correction function and meteorological impact correction function to complete the correction action by summing the instantaneous exposure intensity of aviation practitioners in each time slice, i.e., the exposure amount; Accumulation and Report Generation: Accumulate the exposure amounts for all time slices to obtain the total exposure dose for aviation workers throughout the monitoring period, and incorporate individual physiological adaptation correction factors to generate an exposure dose report.

6. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 5, characterized in that, The improved multi-pollutant exposure model is as follows: the improved total exposure dose of the j-th aviation worker is obtained by multiplying the initial total exposure dose of the j-th aviation worker by the individual physiological adaptation correction factor; wherein, when the constraint is: if j is a newcomer to high altitude, the individual physiological adaptation correction factor is: the sum of 1 and the exposure sensitivity enhancement coefficient of newcomers to high altitude; when the constraint is: if j is not a newcomer to high altitude, the individual physiological adaptation correction factor is:

1.

7. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 1, characterized in that, Physiological indicators include at least heart rate, blood oxygen saturation, blood pressure, and respiratory rate. The physiological stress index of each aviation worker is obtained by weighted summation based on these physiological indicators.

8. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 1, characterized in that, The content of the expert rule judgment library and analysis strategy is as follows: the preliminary risk level is adjusted using the preset expert rule judgment library to obtain the secondary risk level; Furthermore, the risk curves within the historical monitoring period are analyzed in time series, and based on the analysis results, a decision is made on whether to perform a secondary adjustment to the secondary risk level in order to obtain the final risk level.

9. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 8, characterized in that, The process of analyzing the risk curve is as follows: Input variables: secondary risk level, time series data, and a set of evaluation indicators for the analysis dimensions; where secondary risk levels include: low, medium, high, and urgent; and time series data includes: health risk indices R1, R2, ..., R for N historical monitoring periods. N The set of analytical dimensions and evaluation indicators includes: risk change rate trend, maximum risk value, average risk value, and risk volatility value. Trend Feature Extraction: Comparative analysis is used to evaluate whether the features of each indicator in the indicator set exceed the limits; specifically, if the risk change rate continues to rise, it indicates that the risk change rate trend exceeds the limit; when R exists... Q A value ≥0.9 indicates that the maximum risk value has exceeded the limit; an average risk value exceeding 0.6 indicates that the average risk value has exceeded the limit; a standard deviation exceeding a preset standard threshold indicates that the risk volatility value has exceeded the limit; where R... Q This represents the health risk index during the Qth monitoring period; Secondary adjustment trigger mechanism: Analyze whether the number of features exceeding the limit in each indicator in the evaluation indicator set exceeds the set value. If it does, a secondary adjustment is triggered, raising the risk level by one level from the original risk level and using the raised risk level as the final risk level. If it does not exceed the limit, the original risk level is used as the final risk level.

10. The comprehensive occupational health risk assessment and monitoring system for high-altitude aviation based on pollution sources according to claim 1, characterized in that, The system also includes a health intervention recommendation module, which triggers a recommendation adjustment mechanism based on the obtained final risk level, and extracts corresponding recommendation measures from a pre-built strategy recommendation library according to the final risk level.

Images