A human physiological index monitoring method based on numerical analysis method

Abstract

This invention relates to the field of power grid engineering technology, and provides a method for monitoring human physiological indicators based on numerical analysis, comprising: S1, establishing a finite element simulation model of transmission line towers and working scenarios, and setting electrical parameters and boundary conditions according to actual working conditions; S2, obtaining electromagnetic field distribution data of workers at typical working positions by solving the finite element simulation model; S3, establishing a coupling relationship model between electromagnetic field strength and human physiological indicators based on the electromagnetic field distribution data; S4, constructing a multi-parameter comprehensive evaluation model, normalizing and weighting the predicted or measured values ​​of multiple physiological indicators to generate a comprehensive safety index or fatigue level; S5, classifying and assessing the human physiological safety status and providing early warning based on the comparison results of the comprehensive safety index or fatigue level with a preset threshold. This invention can better monitor human physiological indicators.

Description

Technical Field

[0001] This invention relates to the field of power grid engineering technology, and more specifically, to a method for monitoring human physiological indicators based on numerical analysis. Background Technology

[0002] With the accelerated construction of new power systems, the scale of power grid projects continues to expand, and the tasks of constructing, upgrading, and maintaining transmission lines are becoming increasingly arduous. Working at heights is a common part of power grid construction, with workers spending extended periods at altitudes of 2 meters or more above the fall reference plane, facing complex and ever-changing working conditions. Simultaneously, with the continuous increase in voltage levels, transmission lines at 110kV and above are widely used, and ±500kV, ±800kV UHVDC transmission lines and 1000kV UHVAC transmission lines have been put into operation. Workers inevitably expose themselves to strong electromagnetic fields during high-altitude operations.

[0003] According to authoritative statistics, falls from heights and being struck by objects account for over 45% of all construction safety accidents in the power industry. Prolonged work in strong electromagnetic fields can cause workers to experience headaches, dizziness, weakness in the limbs, and difficulty concentrating, potentially leading to accidents and seriously threatening personal safety. Research shows that power frequency electric and magnetic fields can induce currents that affect the central nervous system, cardiovascular system, and visual system, causing physiological changes such as increased heart rate, blood pressure fluctuations, decreased blood oxygen saturation, and abnormal body temperature regulation. Especially under the complex conditions of working at heights combined with strong electromagnetic fields, the hypoxia caused by decreased ambient air pressure coupled with the physiological effects of the electromagnetic field further exacerbates the physiological burden and safety risks for workers.

[0004] Currently, the main technical deficiencies in safety monitoring for workers operating at heights in power grid projects are as follows: (1) The monitoring methods are limited and lack the ability to make comprehensive judgments on multiple parameters. Most existing smart wearable devices can only monitor a single parameter, such as only heart rate or only electric field strength. The data between the monitoring subsystems are isolated and cannot form a fusion analysis of the physiological state of workers and environmental risks, making it difficult to achieve early identification and accurate warning of risks.

[0005] (2) Insufficient research on the mechanism of electromagnetic environment’s influence on physiological parameters. Existing technologies mostly use static threshold alarm methods, that is, when the heart rate exceeds a certain fixed value or the electric field strength exceeds a certain limit, an alarm is triggered. However, a dynamic coupling relationship model between electromagnetic field strength and human physiological indicators has not been established, and the real-time impact of electromagnetic exposure on physiological state cannot be quantified, resulting in poor accuracy and timeliness of early warning.

[0006] (3) The integration of multi-source data is low and there is a lack of comprehensive evaluation indicators. The physiological parameters (heart rate, blood pressure, blood oxygen, body temperature) and environmental parameters (electric field, magnetic field, noise, leakage current) collected by the existing monitoring system are mostly displayed independently. There is a lack of quantitative indicators that can comprehensively reflect the overall physiological load of the human body, making it difficult for on-site managers to quickly judge the actual safety status of the workers.

[0007] (4) Insufficient integration of numerical simulation and field monitoring. Existing technologies rely heavily on field measurement data, but it is difficult and costly to obtain measurement data in high-altitude strong electromagnetic environments. Furthermore, it is difficult to cover all working conditions under different voltage levels, different tower structures, and different meteorological conditions. There is a lack of research on the electromagnetic exposure characteristics of the human body based on numerical simulation methods, which cannot provide sufficient theoretical basis for setting monitoring thresholds and constructing early warning models.

[0008] To address the aforementioned technical challenges, there is an urgent need for a method for monitoring human physiological indicators that can achieve coupled analysis of electromagnetic environment and physiological parameters, multi-source data fusion assessment, and forward-looking early warning capabilities, in order to improve the safety of personnel working at heights in power grid projects. Summary of the Invention

[0009] The present invention provides a method for monitoring human physiological indicators based on numerical analysis, which can overcome some or all of the defects of the prior art.

[0010] A method for monitoring human physiological indicators based on numerical analysis according to the present invention includes the following steps: Step S1: Establish a finite element simulation model of the transmission line tower and the working scenario. The model includes a tower structure model, a conductor model and a human body model, and set electrical parameters and boundary conditions according to the actual working conditions. Step S2: By solving the finite element simulation model, electromagnetic field distribution data of the operator at a typical working position is obtained. The electromagnetic field distribution data includes the surface electric field strength, the induced current density in the body, and the magnetic induction intensity. Step S3: Based on the electromagnetic field distribution data obtained in step S2, establish a coupling relationship model between electromagnetic field strength and human physiological indicators, including blood oxygen saturation, blood pressure, heart rate and body temperature. Step S4: Construct a multi-parameter comprehensive evaluation model, normalize and weight the predicted or measured values ​​of multiple physiological indicators obtained in step S3, and generate a comprehensive safety index or fatigue level to characterize the overall physiological load of the human body. Step S5: Based on the comparison results between the comprehensive safety index or fatigue level and the preset threshold, the physiological safety status of the human body is graded, assessed, and warned.

[0011] Preferably, the human body model established in step S1 is a simplified geometric model, using a homogeneous medium approximation, with its conductivity set to 0.1 S / m and relative permittivity set to 10. 6 Furthermore, during finite element analysis, the mesh is refined for the human body surface and areas with large curvature changes; The electrical parameters and boundary conditions set in step S1 include: 1.1) Conductor voltage levels, including one or more of 10kV, 110kV, 220kV, 500kV, ±500kV, ±800kV and 1000kV; 1.2) Effective value of conductor current; 1.3) Conductor height above ground, phase spacing, and structural parameters of split conductors; 1.4) Boundary conditions include: ground potential is set to 0, conductor surface potential is set to operating voltage, and artificial boundary potential is set to nominal voltage.

[0012] Preferably, in step S2, the typical work location includes at least: 2.11) Ground location beneath the tower; 2.12) Position of the crossarm on the upper part of the tower; 2.13) Equipotential working positions on conductors; 2.14) Location of foundation construction equipment near the tower; 2.15) Work locations crossing or traversing under railway lines; Step S2 also includes considering the influence of different meteorological factors on the electromagnetic field distribution, wherein the meteorological factors include at least: 2.21) Relative humidity, ranging from 30% to 80%; 2.22) Ambient temperature, ranging from -20℃ to 40℃; 2.23) Atmospheric pressure, ranging from 50 kPa to 101.325 kPa; 2.24) Wind speed and wind direction.

[0013] Preferably, in step S3, blood oxygen saturation is established. The mathematical model specifically includes: Establish an alveolar-arterial oxygen dynamics model: ; In the formula: It is the partial pressure of oxygen in arterial blood; This refers to the partial pressure of oxygen in the alveoli. For time; This is due to the frequent occurrence of pulmonary gas exchange; For the blood volume involved in rapid gas exchange; This refers to the rate of total body oxygen consumption. alveolar oxygen partial pressure Affected by altitude: ; In the formula: The oxygen fraction inhaled. Atmospheric pressure decreases with altitude. It is the vapor pressure of water. This refers to the partial pressure of carbon dioxide in the alveoli. For respiratory quotient; Explicitly incorporate height effects and electromagnetic disturbances into the parameterization terms: ; ; In the formula: h For altitude, This is the atmospheric pressure decay constant with altitude; This refers to the basal oxygen consumption rate at rest. Operating power; It senses electrical charges within and on the body surface. Indicators of psychological stress; The contribution coefficient of workload to oxygen consumption rate; The contribution coefficient of electromagnetic field induced current to oxygen consumption rate; The contribution coefficient of psychological stress to oxygen consumption rate; As an indicator of psychological stress; Blood oxygen saturation is expressed using the Hill equation: ; in To achieve an arterial oxygen partial pressure with 50% Hb saturation, n is the Hill coefficient.

[0014] Preferably, in step S3, a mathematical model for blood pressure (BP) is established, specifically including: Mean arterial pressure Dynamic model representation: ; For cardiac output; This represents the total peripheral resistance. Cardiac output is determined by heart rate With stroke volume Decide: ; The sympathetic-parasympathetic regulation and external disturbances are modeled as state equations: ; ; In the formula: This refers to the baseline heart rate at rest. The time constant for heart rate regulation. For pressure reflection gain, The setpoint blood pressure value for pressure reflex. This is the direct driving term of the electromagnetic field on heart rate. Environmental factors are the driving factors of heart rate; Basic peripheral resistance; The coefficient representing the effect of low oxygen on peripheral resistance; This is a hypoxia indicator function; The coefficient representing the influence of electromagnetic field-induced current on peripheral resistance; This refers to the induced current density or the surface potential difference. The coefficient representing the influence of psychological stress on peripheral resistance; The linear sensitivity to short-time disturbances is approximated as follows: ; The change in blood pressure, The change in cardiac output. Based on the core output volume This represents the change in total peripheral resistance; thus, the contribution of elevation gain and EM to BP can be estimated.

[0015] Preferably, in step S3, a risk assessment model for heart rate variability (HRV) is established, including: Let the instantaneous trigger rate λ(t) be used to describe the instantaneous risk of severe cardiac arrhythmias: ; Based on the basic cardiac rhythm event trigger rate, The contribution coefficient of hypoxia to the risk of cardiac arrhythmias. This represents the contribution coefficient of electromagnetic fields to the risk of cardiac arrhythmias. g ( ) is a nonlinear mapping function for the risk of induced electrical events; The contribution coefficient of psychological stress to the risk of cardiac arrhythmic events is given within the observation window. Internal event occurrence probability : ; For integration variables; Using time-domain and frequency-domain HRV indices to Alternatively, dynamic calibration can be performed using baroreflex sensitivity.

[0016] Preferably, in step S3, a mathematical model of body temperature T is established, using the Pennes biological heat conduction equation to describe local tissue temperature changes: ; in, For tissue density, To organize specific heat capacity, Let r be the tissue temperature at time t. To improve the thermal conductivity of the tissue, For the Laplace operator, Blood density, For the specific heat capacity of blood, For blood perfusion rate, Arterial blood temperature, Metabolic heat production rate, The rate of heat generation from electromagnetic field energy absorption; Instantaneous power density This is used to quantify the thermal effects of electromagnetic field energy absorption on local tissues. Let r be the tissue conductivity at position r. Let be the electric field strength at position r at time t; Simplified full-body box model: ; In the formula: For total body heat capacity, For the core temperature, For heat dissipation, related to ambient temperature It is related to wind speed and insulating clothing. Metabolic power, This refers to electromagnetic absorption power.

[0017] Preferably, the multi-parameter comprehensive evaluation model constructed in step S4 includes the following sub-steps: Step S41: Normalize the physiological indicators: Heart rate normalization function : ; This represents the lower limit of the normal heart rate range. This represents the upper limit of the normal heart rate range. Blood pressure normalization function : ; To measure systolic blood pressure, To measure diastolic blood pressure, This represents the upper limit of the normal range for systolic blood pressure. This represents the lower limit of the normal range for diastolic blood pressure. Blood oxygen normalization function : ; Body temperature normalization function :

[0018] Step S42, weighted fusion to generate instantaneous fatigue value : ; Among them, the weighting coefficient satisfy Furthermore, adjustments are made dynamically based on the type of work and individual differences; Step S43, establish the electromagnetic field correction factor:

[0019] in, The fatigue value after electromagnetic field correction. For the normalized electric and magnetic field strengths, This is a correction factor.

[0020] Preferably, step S4 also includes a step of predicting fatigue using a deep learning model: 4.1) Construct a recurrent neural network or Transformer model. Input features include: historical physiological index time series, environmental electromagnetic parameter time series, work load information, and meteorological parameters. 4.2) Model output is the future Predicted fatigue level at any time : ; For time window The input feature sequence within; 4.3) Adopt a hybrid evaluation strategy: when or When this happens, an alert is triggered.

[0021] Preferably, in step S5, the grading assessment includes: Level I: Indicates safety; overall safety index. or fatigue The human body is in a state of physiological homeostasis; Level II: Indicates mild risk. The physiological compensation mechanism has been activated; it is recommended to strengthen monitoring. Level III: Indicates danger. When physiological regulation is out of balance, an audible and visual alarm is immediately triggered, and the warning information is uploaded to the cloud monitoring platform.

[0022] This invention combines the finite element numerical simulation method with a human physiological dynamics model, establishing a systematic mathematical model relating multiple physical parameters such as electric field strength, magnetic induction intensity, and induced current density to multiple physiological indicators such as blood oxygen saturation, blood pressure, heart rate, and body temperature. Compared with existing technologies, the advantages of this invention are: 1) The thermal effect of electromagnetic field energy absorption on local tissues is quantified using the Pennes biological heat conduction equation, solving the problem that existing technologies cannot quantitatively describe the thermal effect of electromagnetic fields. The direct driving term of the electromagnetic field on heart rate is explicitly expressed through a heart rate dynamics model, achieving the separation and quantification of electromagnetic stress and physiological regulation. Using an alveolar-arterial oxygen dynamics model and an atmospheric pressure decay formula with altitude, the coupled influence of high-altitude hypoxia and electromagnetic field exposure on blood oxygen saturation is incorporated into a unified model for the first time, solving the technical challenge of handling the superimposed effects of multiple stressors in existing technologies. Electromagnetic field distribution data at different voltage levels (10kV-1000kV) and different operating locations (under towers, crossarms, conductors at equipotential) are obtained through finite element simulation, providing a theoretical basis for the scientific setting of physiological monitoring thresholds and safety warning thresholds, avoiding false alarms and missed alarms caused by existing technologies relying on empirical thresholds.

[0023] 2) This invention constructs a multi-parameter comprehensive evaluation model that includes normalization processing, weighted fusion, and electromagnetic field correction, fusing multi-source heterogeneous data into a single quantitative index. Through heart rate normalization functions, blood pressure normalization functions, and blood oxygen normalization functions, physiological parameters with different dimensions and normal ranges are uniformly mapped to a standardized space of 0-100%, solving the problem of multi-parameter comparability. Through a weighted fusion formula, the normalization deviation of the four physiological indicators is integrated into instantaneous fatigue level, and the weighting coefficients can be dynamically adjusted according to the type of work and individual differences, solving the problem that existing technologies cannot comprehensively determine overall physiological load. Through an electromagnetic field correction factor, the influence of environmental electric field strength and magnetic induction intensity on physiological state is incorporated into the comprehensive evaluation. The correction coefficient is calibrated through finite element simulation and measured data, solving the problem that existing technologies neglect the influence of environmental factors on physiological state.

[0024] 3) This invention introduces a deep learning model to predict future fatigue levels based on historical physiological index time series, environmental electromagnetic parameter time series, workload information, and meteorological parameters. Through recurrent neural networks or Transformer models, it learns the dynamic patterns of physiological parameter changes over time, enabling early identification of risk trends during fatigue accumulation. For example, when blood oxygen saturation slowly decreases from 98% to 95%, the model can predict that it may drop below 93% in the next 10-15 minutes, triggering an early warning and buying valuable time for intervention. This early warning mechanism, combining instantaneous fatigue level with predicted fatigue level, ensures both immediate response to emergencies and early warning of progressive risks, overcoming the limitations of existing technologies that only provide post-event alerts.

[0025] This invention obtains electromagnetic field distribution data by establishing a finite element simulation model, constructs an electromagnetic-physiological coupling model and a multi-parameter comprehensive evaluation model, generates a comprehensive safety index or fatigue level, and realizes graded assessment and early warning of human physiological safety status. Attached Figure Description

[0026] Figure 1 This is a flowchart of a human physiological indicator monitoring method based on numerical analysis in Example 1; Figure 2 This is a schematic diagram of the 10kV tower model in Example 1; Figure 3 This is a schematic diagram of the human body model in Example 1; Figure 4 This is a schematic diagram of the overall electric field intensity distribution of the 10kV tower in Example 1; Figure 5 This is a schematic diagram of the surface induced electric field distribution when the worker is located under the tower in Example 1; Figure 6 This is a schematic diagram of the magnetic induction intensity distribution around the tower in Example 1; Figure 7 This is a schematic diagram of the magnetic induction intensity distribution under the tower for workers in Example 1. Detailed Implementation

[0027] To further understand the content of this invention, a detailed description of the invention will be provided in conjunction with the accompanying drawings and embodiments. It should be understood that the embodiments are merely illustrative and not limiting of the invention.

[0028] Example 1 like Figure 1 As shown, this embodiment provides a method for monitoring human physiological indicators based on numerical analysis, which includes the following steps: Step S1: Establish a finite element simulation model of the transmission line tower and the working scenario. The model includes a tower structure model, a conductor model, and a human body model. Set electrical parameters and boundary conditions according to the actual working conditions.

[0029] 10kV tower model as follows Figure 2 As shown, the tower is a self-supporting steel truss tower, constructed using a welded lattice structure of angle steel or steel pipes. The tower body is generally pyramidal in shape, composed of a truss structure that is wide at the base and gradually narrows towards the top, providing excellent wind, bending, and torsional resistance. The structure is divided into the tower base section, main trunk section, crossarm section, and top conductor support section. The model shows three layers of crossarms used for the arrangement of three-phase AC lines.

[0030] From an electromagnetic perspective, the metal lattice structure of this type of tower has good conductivity and grounding performance, and can be approximated as an ideal conductor in electromagnetic simulations to study the electric and magnetic field distribution characteristics between transmission lines and the ground. Due to the large size of the tower and its composition of multiple metal components, its influence on the distortion of the surrounding electric field is limited, and it mainly serves as a support and grounding path.

[0031] The human body model established in step S1 is a simplified geometric model, using a homogeneous medium approximation, with its conductivity set to 0.1 S / m and relative permittivity set to 10. 6 Furthermore, during finite element analysis, the mesh is refined for the human body surface and areas with significant curvature variations; such as... Figure 3 As shown.

[0032] The electrical parameters and boundary conditions set in step S1 include: 1.1) Conductor voltage levels, including one or more of 10kV, 110kV, 220kV, 500kV, ±500kV, ±800kV and 1000kV; 1.2) Effective value of conductor current; 1.3) Conductor height above ground, phase spacing, and structural parameters of split conductors; 1.4) Boundary conditions include: ground potential is set to 0, conductor surface potential is set to operating voltage, and artificial boundary potential is set to nominal voltage.

[0033] Step S2 involves solving the finite element simulation model to obtain electromagnetic field distribution data for workers at typical work locations. This data includes the surface electric field strength, induced current density within the body, and magnetic induction intensity. A schematic diagram of the overall electric field strength distribution of a 10kV tower is shown below. Figure 4 As shown in the diagram, the surface induced electric field distribution is as follows when the worker is located under the tower. Figure 5 As shown in the diagram, the magnetic field strength distribution around the tower is as follows: Figure 6 As shown in the diagram, the magnetic field strength distribution under the tower is as follows: Figure 7As shown.

[0034] In step S2, the typical work location includes at least: 2.11) Ground location beneath the tower; 2.12) Position of the crossarm on the upper part of the tower; 2.13) Equipotential working positions on conductors; 2.14) Location of foundation construction equipment near the tower; 2.15) Work locations crossing or traversing under railway lines; Step S2 also includes considering the influence of different meteorological factors on the electromagnetic field distribution, wherein the meteorological factors include at least: 2.21) Relative humidity, ranging from 30% to 80%; 2.22) Ambient temperature, ranging from -20℃ to 40℃; 2.23) Atmospheric pressure, ranging from 50 kPa to 101.325 kPa; 2.24) Wind speed and wind direction.

[0035] Step S3: Based on the electromagnetic field distribution data obtained in step S2, establish a coupling relationship model between electromagnetic field strength and human physiological indicators, including blood oxygen saturation, blood pressure, heart rate, and body temperature.

[0036] In step S3, blood oxygen saturation is established. The mathematical model specifically includes: Establish an alveolar-arterial oxygen dynamics model: ; In the formula: It is the partial pressure of oxygen in arterial blood; This refers to the partial pressure of oxygen in the alveoli. For time; This is due to the frequent occurrence of pulmonary gas exchange; For the blood volume involved in rapid gas exchange; This refers to the rate of total body oxygen consumption. alveolar oxygen partial pressure Affected by altitude: ; In the formula: The oxygen fraction inhaled. Atmospheric pressure decreases with altitude. It is the vapor pressure of water. This refers to the partial pressure of carbon dioxide in the alveoli. For respiratory quotient; Explicitly incorporate height effects and electromagnetic disturbances into the parameterization terms: ; ; In the formula: h For altitude, This is the atmospheric pressure decay constant with altitude; This refers to the basal oxygen consumption rate at rest. Operating power; It senses electrical charges within and on the body surface. Indicators of psychological stress; The contribution coefficient of workload to oxygen consumption rate; The contribution coefficient of electromagnetic field induced current to oxygen consumption rate; The contribution coefficient of psychological stress to oxygen consumption rate; As an indicator of psychological stress; Blood oxygen saturation is expressed using the Hill equation: ; in To achieve an arterial oxygen partial pressure with 50% Hb saturation, n is the Hill coefficient.

[0037] In step S3, a mathematical model for blood pressure (BP) is established, which specifically includes: Mean arterial pressure Dynamic model representation:

[0038] For cardiac output; This represents the total peripheral resistance. Cardiac output is determined by heart rate With stroke volume Decide:

[0039] The sympathetic-parasympathetic regulation and external disturbances are modeled as state equations: ; ; In the formula: This refers to the baseline heart rate at rest. The time constant for heart rate regulation. For pressure reflection gain, The setpoint blood pressure value for pressure reflex. This is the direct driving term of the electromagnetic field on heart rate. Environmental factors are the driving factors of heart rate; Basic peripheral resistance; The coefficient representing the effect of low oxygen on peripheral resistance; This is a hypoxia indicator function; The coefficient representing the influence of electromagnetic field-induced current on peripheral resistance; This refers to the induced current density or the surface potential difference. The coefficient representing the influence of psychological stress on peripheral resistance; The linear sensitivity to short-time disturbances is approximated as follows: ; The change in blood pressure, The change in cardiac output. Based on the core output volume This represents the change in total peripheral resistance; thus, the contribution of elevation gain and EM to BP can be estimated.

[0040] In step S3, a risk assessment model for heart rate variability (HRV) is established, including: Let the instantaneous trigger rate λ(t) be used to describe the instantaneous risk of severe cardiac arrhythmias: ; Based on the basic cardiac rhythm event trigger rate, The contribution coefficient of hypoxia to the risk of cardiac arrhythmias. This represents the contribution coefficient of electromagnetic fields to the risk of cardiac arrhythmias. g ( ) is a nonlinear mapping function for the risk of induced electrical events; The contribution coefficient of psychological stress to the risk of cardiac arrhythmic events is given within the observation window. Internal event occurrence probability : ; For integration variables; Using time-domain and frequency-domain HRV indices to Alternatively, dynamic calibration can be performed using baroreflex sensitivity.

[0041] In step S3, a mathematical model of body temperature T is established, and the Pennes biological heat conduction equation is used to describe the local tissue temperature change: ; in, For tissue density, To organize specific heat capacity, Let r be the tissue temperature at time t. To improve the thermal conductivity of the tissue, For the Laplace operator, Blood density, For the specific heat capacity of blood, For blood perfusion rate, Arterial blood temperature, Metabolic heat production rate, The rate of heat generation from electromagnetic field energy absorption; Instantaneous power density This is used to quantify the thermal effects of electromagnetic field energy absorption on local tissues. Let r be the tissue conductivity at position r. Let be the electric field strength at position r at time t; Simplified full-body box model: ; In the formula: For total body heat capacity, For the core temperature, For heat dissipation, related to ambient temperature It is related to wind speed and insulating clothing. Metabolic power, This refers to electromagnetic absorption power.

[0042] Step S4: Construct a multi-parameter comprehensive evaluation model. Normalize and weight the predicted or measured values ​​of multiple physiological indicators obtained in step S3 to generate a comprehensive safety index or fatigue level that characterizes the overall physiological load of the human body.

[0043] In step S4, the constructed multi-parameter comprehensive evaluation model includes the following sub-steps: Step S41: Normalize the physiological indicators: Heart rate normalization function : ; This represents the lower limit of the normal heart rate range. This represents the upper limit of the normal heart rate range. Blood pressure normalization function : ; To measure systolic blood pressure, To measure diastolic blood pressure, This represents the upper limit of the normal range for systolic blood pressure. This represents the lower limit of the normal range for diastolic blood pressure. Blood oxygen normalization function :

[0044] Body temperature normalization function : ; Step S42, weighted fusion to generate instantaneous fatigue value : ; Among them, the weighting coefficient satisfy Furthermore, adjustments are made dynamically based on the type of work and individual differences; Step S43, establish the electromagnetic field correction factor: ; in, The fatigue value after electromagnetic field correction. For the normalized electric and magnetic field strengths, This is a correction factor.

[0045] To further improve the accuracy of fatigue assessment, step S4 introduces fractional calculus theory to construct a fatigue accumulation model, replacing the traditional integer-order model. This model more accurately describes the nonlinear memory characteristics and load accumulation effects of the human physiological system under complex electromagnetic environments. Specifically, it defines the cumulative fatigue state of workers in high-altitude, strong electromagnetic environments. Satisfy the following fractional differential equations: ; in, Let be the cumulative fatigue state value at time t, ranging from 0 to 100%. It is a fractional order used to characterize the memory effect of fatigue accumulation in physiological systems. The smaller the value, the greater the impact of historical conditions on current fatigue. for A fractional-order differential operator, where t is the time variable. Let be the normalized heart rate deviation function at time t, reflecting the degree to which the heart rate deviates from the normal range. Let be the normalized blood pressure deviation function at time t, reflecting the degree to which blood pressure deviates from the normal range. Let be the normalized blood oxygen deviation function at time t, reflecting the degree to which blood oxygen saturation deviates from the normal range. Let be the normalized body temperature deviation function at time t, reflecting the degree to which body temperature deviates from the normal range. a, b, c, and d are coupling coefficients, corresponding to the weights of the contributions of heart rate, blood pressure, blood oxygen, and body temperature changes to fatigue, respectively. The recovery coefficient characterizes the rate of fatigue recovery for workers during rest.

[0046] The fractional model is discretized and solved using the Grunwald-Letnikov numerical approximation method: ; Where T is the sampling period, and n is the truncation order, representing the number of historical states considered. The coefficients are generalized binomial coefficients. Before time t The function value in units of time; Compared to integer-order models, this model can more accurately characterize the nonlinear accumulation process of fatigue and the recovery delay effect, improving prediction accuracy by about 20%. It is especially suitable for the non-stationary evolution of the physiological state of workers in high-altitude, strong electromagnetic environments.

[0047] Step S4 also includes a step of predicting fatigue using a deep learning model: 4.1) Construct a recurrent neural network or Transformer model. Input features include: historical physiological index time series, environmental electromagnetic parameter time series, work load information, and meteorological parameters. 4.2) Model output is the future Predicted fatigue level at any time :

[0048] For time window The input feature sequence within; 4.3) Adopt a hybrid evaluation strategy: when or When this happens, an alert is triggered.

[0049] To further improve the intelligence level of multi-source data fusion, the deep learning prediction model in step S4 employs a hybrid noise reduction algorithm based on adaptive Kalman filtering and variational mode decomposition for data preprocessing. This algorithm originates from advanced signal processing and adaptive control theory and is used to address the nonlinearity, non-stationarity, and strong interference characteristics of multi-source sensor signals in high-altitude, strong electromagnetic environments. Specifically, it includes the following steps: The first step is to construct a state-space model. The collected raw physiological signals (such as heart rate and blood pressure) are used as observations. The system state vector is defined as the real physiological parameters and their first derivatives. State transition equations and observation equations that include electromagnetic interference as process noise are established.

[0050] The second step is to perform variational mode decomposition. The original signal is preprocessed and decomposed into several eigenmode functions with sparse characteristics. Based on the center frequency, the effective physiological signal frequency band and the electromagnetic interference frequency band are distinguished, and high-frequency noise modes are removed.

[0051] The third step is adaptive Kalman filtering. The reconstructed signal after variational mode decomposition is used as the observation input to the Kalman filter. Mahalanobis distance is introduced as an outlier detection metric. The statistical characteristics of the innovation sequence are calculated in real time, and the process noise covariance matrix of the filter is dynamically adjusted to achieve adaptive updating of the filter gain. This algorithm can be described as follows: State estimation formula: ;

[0052] Kalman gain formula: ;

[0053] in, The state estimate at time k (the filtered physiological parameters, such as the true values ​​of heart rate and blood pressure). The state value at time k is predicted based on time k-1. The Kalman gain matrix at time k controls the fusion weights of the predicted and observed values. The value at time k is the observation (the raw physiological signal collected by the sensor). The observation matrix maps the state vector to the observation space. Let be the prediction error covariance matrix at time k. Let be the observation noise covariance matrix at time k, which characterizes the statistical properties of the sensor measurement noise; Formula for calculating Mahalanobis distance: ;

[0054] in, Mahalanobis distance is used to determine whether the current observation is an outlier. Let T be the information covariance matrix, which represents the statistical properties of the difference between the predicted and observed values. T is the matrix transpose symbol. In the above algorithm, According to Mahalanobis distance The real-time calculation results are dynamically adjusted: when When the value exceeds a preset threshold, it indicates that the observed value may be subject to strong electromagnetic interference, and the system will automatically increase the threshold. This design reduces the weight of the observation in the state estimation. It effectively suppresses periodic noise and pulse interference caused by power frequency electromagnetic fields. Compared to traditional low-pass filtering, it improves the signal-to-noise ratio by approximately 15 dB while ensuring real-time signal performance, providing high-quality input data for deep learning prediction models.

[0055] Step S5: Based on the comparison results between the comprehensive safety index or fatigue level and the preset threshold, the physiological safety status of the human body is graded, assessed, and warned.

[0056] In step S5, the grading assessment includes: Level I: Indicates safety; overall safety index. or fatigue The human body is in a state of physiological homeostasis; Level II: Indicates mild risk. The physiological compensation mechanism has been activated; it is recommended to strengthen monitoring. Level III: Indicates danger. When physiological regulation is out of balance, an audible and visual alarm is immediately triggered, and the warning information is uploaded to the cloud monitoring platform.

[0057] Following step S5, there is also an individual difference calibration step: I. The resting physiological parameters of the personnel under electromagnetic exposure-free conditions were used as the baseline; II. Dynamically adjust the reference range of each physiological indicator based on the resting baseline. and individual differences in the model ; III. Regularly update model parameters to adapt to long-term changes in personnel's physiological state.

[0058] The method also includes model validation and correction steps: i. Compare the electromagnetic field data and physiological parameters measured on-site with the model predictions; ii. Calculate the error rate. If the error exceeds the preset threshold (e.g., 5%), then correct the model parameters using the least squares method or Bayesian optimization. iii. The corrected model parameters are sent to the wearable device via the wireless communication module to achieve online model updates.

[0059] The method also includes an adaptive adjustment step for the early warning threshold: Based on the real-time fatigue level of the workers Based on historical data, the alarm thresholds for environmental parameters such as electric field, magnetic field, and leakage current are dynamically adjusted. when When the levels are high, appropriately lower the alarm threshold for environmental parameters to improve the sensitivity of the early warning. when When the alarm level is low, restore the standard alarm threshold to reduce false alarms.

[0060] To achieve holistic situational awareness of the complex coupling risks among workers, the work environment, and live equipment, a risk propagation model based on a spatiotemporal graph neural network is introduced between steps S4 and S5. This model, derived from knowledge graph and complex network theory, is used to capture the dynamic relationships and evolution paths between various risk factors (such as personnel physiological state, proximity to electrical equipment, electromagnetic field distribution, and weather conditions) in high-altitude work scenarios. The specific construction method is as follows: 1) Graph structure definition. Construct a heterogeneous graph. ,in, It is a set of nodes, including worker nodes, equipment nodes (such as towers and conductors), and environmental nodes (such as electric field regions and magnetic field regions). An edge set represents the physical association or risk propagation path between nodes. For example, the "person-conductor" edge represents near-electric risk, and the "person-environment" edge represents electromagnetic exposure.

[0061] 2) Node Feature and Edge Weight Update. At each time step, the feature vector of each node... Updated based on corresponding real-time monitoring data (such as fatigue level and electric field strength). The node state is updated by aggregating information from neighboring nodes using a graph convolutional network.

[0062] in, for Time Node The feature vector contains the real-time monitoring data corresponding to that node. for Time Node The updated feature vector, For non-linear activation functions, such as ReLU or Sigmoid, This is a self-connection weight matrix used to preserve the state information of each node. A weight matrix is ​​aggregated for neighbors, used to fuse the state information of neighbor nodes. For nodes The set of neighboring nodes, The edge weight function calculates the node weights based on factors such as physical distance and shielding efficiency. With nodes The strength of the correlation between them for Time Neighbor Nodes eigenvectors.

[0063] 3) Overall Risk Index Calculation. An attention mechanism is used to weight and pool the updated states of each node to generate an overall risk situation index for the operational scenario. : ;

[0064] in, This is a risk situation index for the overall scenario, with a value range of... A higher value indicates a higher overall risk. The query matrix is ​​obtained by linear transformation of the updated states of each node and is used to represent the risk type currently of concern. The key matrix is ​​obtained by linear transformation of the updated states of each node and is used to represent the risk information carried by each node. The value matrix is ​​obtained by linear transformation of the updated states of each node and is used to represent the contribution weight of each node to the overall risk. The dimension of the key vector is used to scale the dot product result and prevent gradient vanishing. This graph neural network model, through end-to-end learning, can uncover complex risk propagation paths that are difficult to identify using traditional methods (such as increased wind speed, conductor galloping, fluctuations in proximity to power lines, and heightened stress responses among personnel), achieving a leap from single-parameter early warning to full-scenario situational awareness. Ultimately, the calculated overall risk situation index... The fatigue level generated in step S4 or By combining these approaches, a tiered assessment and early warning system can be implemented to form a multi-dimensional and multi-layered comprehensive security system. This embodiment obtains electromagnetic field distribution data by establishing a finite element simulation model, constructs an electromagnetic-physiological coupling model and a multi-parameter comprehensive evaluation model, generates a comprehensive safety index or fatigue level, and realizes graded assessment and early warning of human physiological safety status.

[0065] Example 2 This embodiment addresses the scenario of working at heights on 110kV AC transmission line towers. It employs a numerical analysis-based method for monitoring human physiological indicators, combining finite element simulation with on-site measured data to achieve real-time monitoring and safety warnings for four core physiological indicators of workers: blood oxygen saturation, blood pressure, heart rate, and body temperature. The monitoring system is adaptable to two typical working conditions: ground work below the tower and crossarm work on the tower. It can withstand ambient temperatures ranging from -10℃ to 40℃ and relative humidity from 30% to 70%, meeting the safety requirements for operations in the strong electromagnetic environment of power grid engineering.

[0066] The implementation steps are as follows: (I) Construction of Finite Element Simulation Model Model Composition Tower structure model: A 110kV self-supporting steel structure tower is adopted, with a tower height of 28m and a crossarm length of 6m. A three-dimensional model is built according to the actual engineering drawings. Conductor model: The conductor type is LGJ-400 / 35, arranged horizontally, with a phase spacing of 4.5m, a conductor height of 18m above the ground, and a split spacing of 400mm; Human model: Simplified as a combination of a "ball head + cylinders for neck / torso / legs", height 1.72m, head radius 0.09m, torso radius 0.16m, electrical conductivity set at 0.1S / m, relative permittivity 10. 6 The mesh is locally refined, and the size of the human body surface unit is ≤5mm.

[0067] Electrical parameters and boundary conditions Conductor voltage rating: 110kV (line voltage), current RMS value 800A; Boundary conditions: Ground potential is set to 0, conductor surface potential is set to 63.5kV (phase voltage), artificial boundary is 20m horizontally from conductor, and potential is set to nominal voltage; Meteorological parameters: ambient temperature 25℃, relative humidity 60%, atmospheric pressure 101.325kPa, wind speed 1.5m / s.

[0068] (II) Acquisition of electromagnetic field distribution data The electromagnetic field distribution data for two typical working locations were obtained by solving the model using finite element simulation software: At ground level (1.5m from the tower center): maximum surface electric field strength 73.28 V / m, magnetic induction intensity 0.53 μT, maximum induced current density within the tower 8.315 × 10⁻⁶. -12 A / m²; At the crossarm position on the tower (1.2m horizontally from the conductor): the maximum surface electric field strength is 20kV / m, the magnetic induction intensity is 50μT, and the maximum induced current density within the tower is 5.165×10⁻⁶. -8 A / m².

[0069] The simulation process simultaneously considered the effect of relative humidity changes (30%~70%) on the surface field strength, and the results showed that for every 10% increase in humidity, the surface field strength decreased by an average of 3.2%.

[0070] (III) Establishment of electromagnetic field-physiological index coupling model Blood oxygen saturation model Basic parameter: Basal oxygen consumption rate at rest =0.25L / min, the blood volume involved in rapid gas exchange =5L, pulmonary gas exchange time constant =0.8min; Altitude correction: The work location is at an altitude of 100m, atmospheric pressure... =100.1 kPa, inhaled oxygen fraction =0.2095; Model equations: ; ; ;

[0071] Where n = 2.7 (Hill coefficient), =3.5kPa, =0.85 (respiratory quotient) =5.3kPa.

[0072] Blood pressure model Basic parameters: Resting heart rate =75 bpm, basic peripheral resistance =1.2×10 5 Pa・s / m³, pressure reflection gain =0.8 bpm / mmHg; Equations of state: ; ; ;

[0073] in, =2.5min, =80mL (stroke volume). =0.02, =1.2×10 -6 m² / A, This represents the induced current density.

[0074] Heart rate risk model Instantaneous trigger rate: ,in This is a nonlinear mapping function for the induced current density; Observation window Δt = 5 min, probability of cardiac rhythm events .

[0075] Body temperature model The model is simplified using the Pennes equation: ;

[0076] in, =2000J / ℃ (total heat capacity). =60W (heat dissipation power) =120W (metabolic power). =σ×|E|² (electromagnetic absorption power).

[0077] (iv) Construction of a multi-parameter comprehensive evaluation model Indicator normalization processing Heart rate normalization: Normal HR range 40~150 bpm. ;

[0078] Normalized blood pressure: Normal systolic blood pressure range is 90-140 mmHg, and normal diastolic blood pressure range is 60-90 mmHg. ;

[0079] Normalized blood oxygenation: ;

[0080] Body temperature normalization: ;

[0081] Weighted fusion computing Weighting coefficients: =0.3 (heart rate) =0.25 (blood pressure) =0.3 (blood oxygen) =0.15 (body temperature), which satisfies the requirement. ; Instantaneous fatigue level: ; Electromagnetic field correction: ,in =0.002, =0.001, , These are the normalized electric and magnetic field strengths.

[0082] (V) Tiered assessment and early warning Security level classification Level I (Safety): ≤80%, physiological indicators are in a steady state, and the system displays a green indicator light; Level II (Mild Risk): 80% If the threshold is ≤90%, an audible and visual alert will be activated, triggering an alarm every 30 seconds. Level III (Danger): >90%, continuous audible and visual alarms, and simultaneously uploads early warning information to the backend monitoring platform via a 4G module.

[0083] Early warning and response mechanism When working under the tower, a Level II warning will be triggered if the blood oxygen saturation is ≤93% or the heart rate is >120 bpm for 1 minute. When working on the tower, a Level II warning will be triggered if blood pressure fluctuates by ≥±10 mmHg or body temperature is >37.5℃. If any of the following indicators are met: blood oxygen ≤90%, systolic blood pressure >150mmHg, heart rate >140bpm, or body temperature >38℃, a Level III warning will be triggered directly.

[0084] Field application verification Wearable devices: Smart safety helmet (integrated heart rate, blood oxygen, and body temperature sensors), monitoring wristband (proximity alarm + blood pressure monitoring), and seat belt integrated module (electric field / magnetic field sensor). Data transmission: 4G CAT1 communication module is used, data acquisition frequency is 2 times / second, and warning information delay is ≤1 second; Reference equipment: Electromagnetic field strength tester (measurement error ±5%), medical-grade physiological parameter monitor (as a comparison benchmark).

[0085] Verification results Simulation and actual measurement deviations: blood oxygen saturation deviation ≤1.2%, blood pressure deviation ≤3mmHg, heart rate deviation ≤2bpm, body temperature deviation ≤0.1℃; Early warning accuracy: Level II early warning accuracy was 92.3%, Level III early warning accuracy was 100%, with no missed reports; Anti-interference performance: Under the conditions of surface electric field strength ≤20kV / m and magnetic induction intensity ≤50μT, the fluctuation range of monitoring data is ≤3%.

[0086] The present invention and its embodiments have been described above illustratively. This description is not restrictive, and the figures shown are only one embodiment of the present invention; the actual structure is not limited thereto. Therefore, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the present invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for monitoring human physiological indicators based on numerical analysis, characterized in that: Includes the following steps: Step S1: Establish a finite element simulation model of the transmission line tower and the working scenario. The model includes a tower structure model, a conductor model and a human body model, and set electrical parameters and boundary conditions according to the actual working conditions. Step S2: By solving the finite element simulation model, electromagnetic field distribution data of the operator at a typical working position is obtained. The electromagnetic field distribution data includes the surface electric field strength, the induced current density in the body, and the magnetic induction intensity. Step S3: Based on the electromagnetic field distribution data obtained in step S2, establish a coupling relationship model between electromagnetic field strength and human physiological indicators, including blood oxygen saturation, blood pressure, heart rate and body temperature. Step S4: Construct a multi-parameter comprehensive evaluation model, normalize and weight the predicted or measured values ​​of multiple physiological indicators obtained in step S3, and generate a comprehensive safety index or fatigue level to characterize the overall physiological load of the human body. Step S5: Based on the comparison results between the comprehensive safety index or fatigue level and the preset threshold, the physiological safety status of the human body is graded, assessed, and warned.

2. The method for monitoring human physiological indicators based on numerical analysis according to claim 1, characterized in that: The human body model established in step S1 is a simplified geometric model, using a homogeneous medium approximation, with its conductivity set to 0.1 S / m and relative permittivity set to 10. 6 Furthermore, during finite element analysis, the mesh is refined for the human body surface and areas with large curvature changes; The electrical parameters and boundary conditions set in step S1 include: 1.1) Conductor voltage ratings, including one or more of 10kV, 110kV, 220kV, 500kV, ±500kV, ±800kV and 1000kV; 1.2) Effective value of conductor current; 1.3) Conductor height above ground, phase spacing, and structural parameters of split conductors; 1.4) Boundary conditions include: ground potential is set to 0, conductor surface potential is set to operating voltage, and artificial boundary potential is set to nominal voltage.

3. The method for monitoring human physiological indicators based on numerical analysis according to claim 2, characterized in that: In step S2, the typical work location includes at least: 2.11) Ground location beneath the tower; 2.12) Position of the crossarm on the upper part of the tower; 2.13) Equipotential working positions on conductors; 2.14) Location of foundation construction equipment near the tower; 2.15) Work locations crossing or traversing under railway lines; Step S2 also includes considering the influence of different meteorological factors on the electromagnetic field distribution, wherein the meteorological factors include at least: 2.21) Relative humidity, ranging from 30% to 80%; 2.22) Ambient temperature, ranging from -20℃ to 40℃; 2.23) Atmospheric pressure, ranging from 50 kPa to 101.325 kPa; 2.24) Wind speed and wind direction.

4. The method for monitoring human physiological indicators based on numerical analysis according to claim 3, characterized in that: In step S3, blood oxygen saturation is established. The mathematical model specifically includes: Establish an alveolar-arterial oxygen dynamics model: ； In the formula: It is the partial pressure of oxygen in arterial blood; This refers to the partial pressure of oxygen in the alveoli. For time; This is due to the frequent occurrence of pulmonary gas exchange; For the blood volume involved in rapid gas exchange; This refers to the rate of total body oxygen consumption. alveolar oxygen partial pressure Affected by altitude: ； In the formula: The oxygen fraction inhaled. Atmospheric pressure decreases with altitude. It is the vapor pressure of water. This refers to the partial pressure of carbon dioxide in the alveoli. For respiratory quotient; Explicitly incorporate height effects and electromagnetic disturbances into the parameterization terms: ； ； In the formula: h For altitude, This is the atmospheric pressure decay constant with altitude; This refers to the basal oxygen consumption rate at rest. Operating power; It senses electrical charges within and on the body surface. Indicators of psychological stress; The contribution coefficient of workload to oxygen consumption rate; The contribution coefficient of electromagnetic field induced current to oxygen consumption rate; The contribution coefficient of psychological stress to oxygen consumption rate; As an indicator of psychological stress; Blood oxygen saturation is expressed using the Hill equation: ； in To achieve an arterial oxygen partial pressure with 50% Hb saturation, n is the Hill coefficient.

5. The method for monitoring human physiological indicators based on numerical analysis according to claim 4, characterized in that: In step S3, a mathematical model for blood pressure (BP) is established, which specifically includes: Mean arterial pressure Dynamic model representation: ； For cardiac output; This represents the total peripheral resistance. Cardiac output is determined by heart rate With stroke volume Decide: ； The sympathetic-parasympathetic regulation and external disturbances are modeled as state equations: ； ； In the formula: This refers to the baseline heart rate at rest. The time constant for heart rate regulation. For pressure reflection gain, The setpoint blood pressure value for pressure reflex. This is the direct driving term of the electromagnetic field on heart rate. Environmental factors drive heart rate; Basic peripheral resistance; The coefficient representing the effect of low oxygen on peripheral resistance; This is a hypoxia indicator function; The coefficient representing the influence of electromagnetic field-induced current on peripheral resistance; This refers to the induced current density or the surface potential difference. The coefficient representing the influence of psychological stress on peripheral resistance; The linear sensitivity to short-time disturbances is approximated as follows: ； The change in blood pressure, The change in cardiac output. Based on the core output volume This represents the change in total peripheral resistance; thus, the contribution of elevation gain and EM to BP can be estimated.

6. The method for monitoring human physiological indicators based on numerical analysis according to claim 5, characterized in that: In step S3, a risk assessment model for heart rate variability (HRV) is established, including: Let the instantaneous trigger rate λ(t) be used to describe the instantaneous risk of severe cardiac arrhythmias: ； Based on the basic cardiac rhythm event trigger rate, The contribution coefficient of hypoxia to the risk of cardiac arrhythmias. This represents the contribution coefficient of electromagnetic fields to the risk of cardiac arrhythmias. g ( ) is a nonlinear mapping function for the risk of induced electrical events; The contribution coefficient of psychological stress to the risk of cardiac arrhythmic events is given within the observation window. Internal event occurrence probability : ； For integration variables; HRV indices in the time and frequency domains Alternatively, dynamic calibration can be performed based on baroreflex sensitivity.

7. The method for monitoring human physiological indicators based on numerical analysis according to claim 6, characterized in that: In step S3, a mathematical model of body temperature T is established, and the Pennes biological heat conduction equation is used to describe the local tissue temperature change: ； in, For tissue density, To organize specific heat capacity, Let r be the tissue temperature at time t. To improve the thermal conductivity of the tissue, For the Laplace operator, Blood density, For the specific heat capacity of blood, For blood perfusion rate, Arterial blood temperature, Metabolic heat production rate, The rate of heat generation from electromagnetic field energy absorption; Instantaneous power density This is used to quantify the thermal effects of electromagnetic field energy absorption on local tissues. Let r be the tissue conductivity at position r. Let be the electric field strength at position r at time t; Simplified full-body box model: ； In the formula: For total body heat capacity, For the core temperature, For heat dissipation, related to ambient temperature It is related to wind speed and insulating clothing. Metabolic power, This refers to electromagnetic absorption power.

8. The method for monitoring human physiological indicators based on numerical analysis according to claim 7, characterized in that: In step S4, the constructed multi-parameter comprehensive evaluation model includes the following sub-steps: Step S41: Normalize the physiological indicators: Heart rate normalization function : ； This represents the lower limit of the normal heart rate range. This represents the upper limit of the normal heart rate range. Blood pressure normalization function : ； To measure systolic blood pressure, To measure diastolic blood pressure, This represents the upper limit of the normal range for systolic blood pressure. This represents the lower limit of the normal range for diastolic blood pressure. Blood oxygen normalization function : ； Body temperature normalization function : ； Step S42, weighted fusion to generate instantaneous fatigue value : ； Among them, the weighting coefficient satisfy Furthermore, adjustments are made dynamically based on the type of work and individual differences; Step S43, establish the electromagnetic field correction factor: ； in, The fatigue value after electromagnetic field correction. For the normalized electric and magnetic field strengths, This is a correction factor.

9. A method for monitoring human physiological indicators based on numerical analysis as described in claim 8, characterized in that: Step S4 also includes a step of predicting fatigue using a deep learning model: 4.1) Construct a recurrent neural network or Transformer model. Input features include: historical physiological index time series, environmental electromagnetic parameter time series, work load information, and meteorological parameters. 4.2) Model output is the future Predicted fatigue level at any time : ； For time window The input feature sequence within; 4.3) Adopt a hybrid evaluation strategy: when or At that time, an early warning is triggered.

10. A method for monitoring human physiological indicators based on numerical analysis according to claim 9, characterized in that: In step S5, the grading assessment includes: Level I: Indicates safety, overall safety index or fatigue The human body is in a state of physiological homeostasis; Level II: Indicates mild risk. The physiological compensation mechanism has been activated; it is recommended to strengthen monitoring. Level III: Indicates danger. When physiological regulation is out of balance, an audible and visual alarm is immediately triggered, and the warning information is uploaded to the cloud monitoring platform.

Images