An industrial staff real-time identity authentication and risk early warning method

Abstract

This invention discloses a method for real-time identity authentication and risk warning for industrial personnel, belonging to the field of identity authentication technology. The method includes acquiring physiological characteristic data, work environment characteristic data, and behavioral characteristic data of industrial personnel; establishing a physiological characteristic matching analysis model based on the physiological and work environment characteristic data to generate a physiological characteristic matching degree; establishing a behavioral characteristic analysis model based on the behavioral characteristic data of industrial personnel to generate a behavioral characteristic matching degree; generating a comprehensive matching degree based on the physiological and behavioral characteristic matching degrees; generating an identity authentication index based on the comprehensive matching degree; and performing real-time identity authentication and risk warning for industrial personnel based on the identity authentication index. This invention solves the problems of single authentication, susceptibility to environmental interference, and lack of continuous verification in existing technologies, achieving real-time and continuous identity authentication and risk warning for industrial personnel, reducing false alarm rates, and improving security and reliability.

Description

Technical Field

[0001] This invention belongs to the field of identity authentication technology, and in particular relates to a method for real-time identity authentication and risk warning of industrial personnel. Background Technology

[0002] In industrial safety production environments, especially in high-risk work areas such as chemical plants, mines, or construction sites, accurate and real-time identification of workers and simultaneous early warning of potential risks are core elements for ensuring operational safety, preventing unauthorized entry, and avoiding accidents caused by abnormal personnel conditions.

[0003] Traditional management methods, such as manual verification or basic access control systems, are insufficient to cope with the dynamic and changing operational needs in complex industrial scenarios, and cannot achieve continuous and accurate assessment of identity authenticity and personnel status.

[0004] Existing technical solutions mainly suffer from two limitations: First, identity recognition methods based on fixed credentials or static biometrics, such as access cards, passwords, or fingerprint scanning, can only provide single-time authentication and lack the ability to continuously verify identity during operation. Furthermore, these methods are susceptible to environmental interference; for example, insufficient lighting can cause facial recognition failure, and fluctuations in temperature and humidity can affect the accuracy of heart rate monitoring, significantly reducing authentication reliability. Second, monitoring systems based on wearable devices typically focus only on the independent analysis of a single physiological indicator (such as heart rate) or movement trajectory to determine states such as fatigue or falls. However, they fail to organically combine identity authentication with state monitoring, resulting in the system's inability to associate detected abnormal behavior (such as sudden changes in posture) with specific personnel identities or dynamically assess risk levels. A deeper problem lies in the fact that existing technologies rely excessively on single-modal data, lack real-time compensation mechanisms for environmental factors (such as temperature and noise), and fail to integrate dynamic characteristics such as behavioral acceleration and angular velocity with physiological indicators for comprehensive analysis. The analysis process often uses static threshold alarms, ignoring temporal characteristics such as behavioral continuity and attitude change rate, resulting in a high false alarm rate and severe warning delays. Ultimately, it is impossible to build a real-time intelligent perception system that coordinates "human-environment-behavior," making it difficult to meet the urgent needs of industrial safety for high-precision, low-latency risk warnings. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for real-time identity authentication and risk warning for industrial personnel, thus solving the aforementioned problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time identity authentication and risk warning for industrial personnel, comprising: Acquire physiological, work environment, and behavioral data of industrial personnel; A physiological characteristic matching analysis model is established based on the physiological characteristic data of industrial personnel and the characteristic data of their work environment, and a physiological characteristic matching degree is generated. A behavioral characteristic analysis model is established based on the behavioral characteristic data of industrial personnel to generate a behavioral characteristic matching degree; the behavioral characteristic data includes behavioral acceleration and behavioral angular velocity. A comprehensive matching score is generated based on the matching scores of physiological characteristics and behavioral characteristics. An identity authentication index is generated based on the overall matching degree; Based on the identity authentication index, real-time identity authentication and risk warning are conducted for industrial personnel.

[0007] Based on the above technical solutions, the present invention also provides the following optional technical solutions: Further technical solution: Generate a working environment correction coefficient based on working environment characteristic data; Generate physiological characteristic deviation based on physiological characteristic data; A physiological characteristic matching analysis model is established based on the work environment correction coefficient and the physiological characteristic deviation degree to generate the physiological characteristic matching degree.

[0008] Further technical solution: The method for generating the working environment correction coefficient specifically includes: Through the formula: ; Generate working environment correction factor ; In the formula, This represents the normalized value of the j-th work environment feature data. This represents the weight coefficient of the j-th work environment feature, and m represents the number of work environment features.

[0009] Further technical solution: The specific method for generating the physiological characteristic deviation degree includes: Through the formula: ; Generate physiological characteristic deviation degree ; In the formula, This represents the data of the i-th physiological characteristic at time t. This represents the baseline value of the i-th physiological characteristic data at time t. It represents the historical standard deviation of the i-th physiological characteristic data at similar times to time t in the historical data, and n represents the number of physiological characteristic data types.

[0010] Further technical solution: The specific expression of the physiological feature matching analysis model is as follows: ; In the expression, This represents the degree of physiological characteristic matching at time t. This represents the degree of deviation in physiological characteristics. The normalized value, This represents the work environment correction factor. This represents the sensitivity adjustment coefficient.

[0011] Further technical solution: The specific method for generating the behavioral feature matching degree includes: Activity intensity and attitude change rate are generated based on behavioral acceleration and behavioral angular velocity, respectively; A behavioral feature analysis model is established based on activity intensity and posture change rate to generate behavioral feature matching degree.

[0012] Further technical solution: The method for generating the activity intensity specifically includes: Through the formula: ; Generate activity intensity ; In the formula, This represents the normalized value of the acceleration in the X-axis direction at time t. This represents the normalized value of the acceleration in the y-axis direction at time t. This represents the normalized value of the acceleration in the z-axis direction at time t; The specific methods for generating the attitude change rate include: Through the formula: ; Generate attitude change rate ; In the formula, This represents the normalized value of the angular velocity of rotation about the X-axis at time t. This represents the normalized value of the angular velocity about the y-axis at time t. It represents the normalized value of the angular velocity of the rotation about the z-axis at time t.

[0013] Further technical solution: The expression of the behavioral feature analysis model is specifically as follows: ; In the expression, This represents the matching degree of behavioral features at time t. This represents the continuity index of behavior at time t. This represents the behavioral stagnation index at time t. This represents the weighting coefficient for behavioral continuity; Specifically, the generation method of the behavioral continuity index at time t includes: Through the formula: ; Generate the behavioral continuity index at time t ; In the formula, This represents the activity intensity at time t-k+1. This represents the rate of attitude change at time t-k+1. This represents the historical average activity intensity. This represents the historical average attitude change rate, and N represents the number of sampling points within the time window. , All are weighting coefficients, and ; Specifically, the generation method of the behavioral stagnation index at time t includes: Through the formula: ; Generate the behavior stagnation index at time t ; In the formula, This represents the activity intensity at time t-g+1. This represents the activity intensity threshold. M is an indicator function, representing the number of sampling points within the time window.

[0014] Further technical solutions: The method for generating the comprehensive matching degree specifically includes: Through the formula: ; Generate comprehensive matching score ; In the formula, This represents the degree of physiological characteristic matching at time t. This represents the matching degree of behavioral features at time t. This represents the weighting coefficient of the physiological characteristic matching degree at time t; The specific methods for obtaining the weight coefficients of the physiological feature matching degree at time t include: Through the formula: ; Weight coefficients for generating physiological feature matching at time t ; In the formula, This indicates the stability of the work environment. This represents the adjustment coefficient; where, the stability of the working environment refers to the normalized value of the changes in the working environment parameters.

[0015] Further technical solutions: The specific methods for generating the identity authentication index include: Through the formula: ; Generate identity authentication index ; In the formula, This represents the overall matching degree at time t. This represents the decision threshold for overall matching degree. This is the slope parameter.

[0016] This invention provides a method for real-time identity authentication and risk warning for industrial personnel, which has the following advantages compared with the prior art: This invention acquires physiological characteristic data, work environment characteristic data, and behavioral characteristic data, establishes physiological characteristic matching analysis models and behavioral characteristic analysis models to generate matching degrees, and then generates comprehensive matching degrees and identity authentication indices to achieve real-time authentication and risk warning. It solves the problems of single authentication, susceptibility to environmental interference, and lack of continuous verification in existing technologies, realizes real-time and continuous identity authentication and risk warning for industrial personnel, reduces false alarm rates, and improves security and reliability. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for real-time identity authentication and risk warning for industrial personnel provided by the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0020] Please see Figure 1 The present invention provides a method for real-time identity authentication and risk warning of industrial personnel, comprising the following steps: Step S10: Obtain physiological characteristic data, work environment characteristic data, and behavioral characteristic data of industrial personnel; Step S20: Establish a physiological characteristic matching analysis model based on the physiological characteristic data of industrial personnel and the characteristic data of the working environment, and generate the physiological characteristic matching degree; Step S30: Establish a behavioral characteristic analysis model based on the behavioral characteristic data of industrial personnel, and generate a behavioral characteristic matching degree; wherein, the behavioral characteristic data includes behavioral acceleration and behavioral angular velocity; Step S40: Generate a comprehensive matching score based on the matching scores of physiological characteristics and behavioral characteristics; Step S50: Generate an identity authentication index based on the overall matching degree; Step S60: Based on the identity authentication index, conduct real-time identity authentication and risk warning for industrial personnel; Physiological characteristic data refers to biological indicators that reflect the physical condition of industrial workers, such as heart rate, body temperature, blood pressure, and blood oxygen saturation. These data are usually collected in real time through wearable devices or non-contact sensors to assess the health status and physiological stress level of workers. Work environment characteristic data refers to environmental parameters within an industrial work area, such as ambient temperature, humidity, noise level, light intensity, and concentration of harmful gases. These data are acquired through environmental sensors and used to assess the potential impact of the environment on the physiological state and behavioral performance of personnel. Behavioral acceleration refers to the change in the speed of industrial personnel moving in three-dimensional space, reflecting the intensity of their activities; Behavioral angular velocity refers to the change in the body posture of industrial workers in three-dimensional space, reflecting the stability or frequency of change of their posture; The physiological characteristic matching analysis model is a mathematical model used to assess the degree of conformity between the current physiological state of industrial workers and their normal or baseline physiological state; the model comprehensively considers physiological characteristic data and work environment characteristic data to generate physiological characteristic matching degree; Physiological feature matching degree is the output of the physiological feature matching analysis model, which represents the degree to which the current physiological state of industrial personnel matches the preset safe or normal state; the higher the matching degree, the closer the physiological state is to normal; the lower the matching degree, the more likely there is a physiological abnormality or risk. A behavioral feature analysis model is a mathematical model used to assess the degree of conformity between the current behavioral patterns of industrial personnel and preset safe or normal behavioral patterns; the model is based on behavioral feature data to generate behavioral feature matching degrees. Behavioral feature matching degree is the output of the behavioral feature analysis model, which represents the degree to which the current behavior pattern of industrial personnel conforms to the preset safe or normal behavior pattern; the higher the matching degree, the more the behavior conforms to the norm; the lower the matching degree, the more abnormal the behavior or risk may exist. The overall matching degree is a comprehensive evaluation index obtained by fusing and calculating the matching degree of physiological characteristics and the matching degree of behavioral characteristics. This index comprehensively reflects the physiological state and behavioral performance of industrial personnel, providing a more comprehensive basis for identity authentication and risk warning. The identity authentication index is a quantitative indicator calculated based on the comprehensive matching degree. It is used to determine the authenticity of the identity of industrial personnel and the risk level of their current status. The index is usually mapped to a specific range to facilitate system decision-making and early warning.

[0021] Traditional identity authentication methods mostly rely on static biometrics or fixed credentials, such as fingerprints or access cards. These methods cannot continuously verify the authenticity of user A's identity when user A is working at heights, nor can they dynamically perceive user A's physiological and behavioral state. This application constructs a dynamic physiological feature matching analysis model and a behavioral feature analysis model by acquiring and fusing physiological feature data, work environment feature data, and behavioral feature data in real time, thereby generating a comprehensive matching degree and finally deriving an identity authentication index.

[0022] Compared to existing technologies that separate identity authentication from status monitoring, this application achieves integrated real-time intelligent perception and early warning of "human-environment-behavior". For example, in some embodiments, monitoring only heart rate may fail to distinguish between elevated heart rate caused by normal work and physiological abnormalities; monitoring only behavior may fail to determine whether strenuous exercise is normal work or an accidental fall. This application effectively compensates for the influence of environmental factors on physiological indicators by combining physiological characteristic data with work environment characteristic data, thus improving the accuracy of physiological status assessment. Simultaneously, by fusing behavioral acceleration and behavioral angular velocity, it can more precisely capture dynamic behavioral patterns of personnel. This deep fusion of multi-dimensional data enables the system to more accurately determine the true identity and potential risks of industrial personnel, avoiding the problems of high false alarm rates and untimely early warnings caused by single-modal data or lack of environmental compensation in traditional methods. Therefore, this application can provide more reliable and timely safety guarantees for industrial production.

[0023] In this application, a method for generating physiological feature matching degree is proposed to build a model based on physiological feature data and work environment feature data to generate matching degree. However, in this process, the influence of environmental factors on physiological features may not be fully considered, resulting in inaccurate matching degree calculation and inability to effectively compensate for environmental interference and physiological deviation, thereby affecting the accuracy of identity authentication.

[0024] In response, this invention further proposes a method for generating the physiological feature matching degree, specifically including: Step S21: Generate work environment correction coefficients based on work environment characteristic data; Step S22: Generate physiological characteristic deviation based on physiological characteristic data; Step S23: Establish a physiological characteristic matching analysis model based on the work environment correction coefficient and physiological characteristic deviation degree, and generate physiological characteristic matching degree; The physiological characteristic matching analysis model is a mathematical model or algorithm used to comprehensively evaluate the degree of matching between the physiological state and normal working conditions of industrial personnel. This model takes a work environment correction coefficient and a physiological characteristic deviation degree as input, and outputs a physiological characteristic matching degree. Its core function is to accurately determine whether a person's physiological state is suitable for the current work, or whether there are potential physiological risks, while considering environmental influences.

[0025] This application's solution optimizes the generation process of physiological characteristic matching degree through specific steps to address the problems of environmental interference and physiological bias. First, step S21 generates a work environment correction coefficient based on work environment characteristic data. This considers the impact of environmental factors such as temperature or noise on physiological characteristics. The correction coefficient is dynamically adjusted to reduce interference from environmental changes, ensuring the matching degree is more adapted to actual working conditions. Second, step S22 generates a physiological characteristic deviation degree based on physiological characteristic data, quantifying the difference between physiological characteristics and baseline values, helping to identify abnormal states and avoiding misjudgments due to physiological fluctuations. Finally, step S23 combines the work environment correction coefficient and the physiological characteristic deviation degree to establish a physiological characteristic matching analysis model and generate the physiological characteristic matching degree. This combination utilizes the synergistic effect of environmental correction and deviation quantification, enabling the model to more accurately integrate environmental compensation and physiological assessment, thereby improving the reliability and robustness of the matching degree. For example, when increased ambient temperature leads to a general increase in heart rate, the work environment correction coefficient can compensate for this general change, allowing the physiological characteristic deviation degree to more accurately reflect individual heart rate abnormalities caused by fatigue or illness, rather than simply environmental stimuli. This step-by-step and interconnected processing method enables the generated physiological feature matching degree to effectively distinguish between physiological fluctuations caused by the environment and physiological abnormalities caused by the individual's own state, thereby significantly improving the accuracy and robustness of physiological state assessment and providing a more reliable physiological state basis for subsequent identity authentication and risk warning.

[0026] The above technical solution fully considers the impact of work environment factors on the physiological state of industrial personnel when generating physiological characteristic matching scores, and quantifies the actual deviation of personnel's physiological characteristics. Specifically, by generating a work environment correction coefficient, the universal impact of environmental changes (such as temperature and noise) on physiological data can be dynamically compensated, avoiding misjudging normal physiological fluctuations caused by the environment as abnormalities. Simultaneously, by generating a physiological characteristic deviation score, the degree to which an individual's physiological state deviates from its baseline value can be accurately measured. Combining these two approaches to establish a physiological characteristic matching analysis model makes the calculation of physiological characteristic matching scores more accurate and robust, effectively distinguishing between environmental interference and individual physiological abnormalities, thereby significantly improving the accuracy of physiological state assessment. This provides a more reliable and refined physiological state basis for subsequent real-time identity authentication and risk warning of industrial personnel, reducing false alarm rates and improving the timeliness and effectiveness of warnings.

[0027] In the scheme proposed in this application, a work environment correction coefficient is used to adjust the matching degree of physiological characteristics. However, in its implementation, the method of generating the work environment correction coefficient is not clearly defined, which may lead to inaccurate calculation or failure to effectively compensate for the impact of environmental changes, thereby reducing the accuracy and reliability of identity authentication.

[0028] In response, this invention further proposes a method for generating the aforementioned working environment correction coefficient, specifically including: Through the formula: ; Generate working environment correction factor ; In the formula, This represents the normalized value of the j-th work environment feature data. This represents the weight coefficient of the j-th work environment feature, and m represents the number of work environment features. The work environment correction coefficient aims to quantify the impact of the current work environment on the physiological state of industrial workers, serving as a dynamic adjustment factor in the subsequent physiological characteristic matching analysis model. Its function is to compensate for the baseline or deviation of physiological characteristics based on environmental changes, thereby more accurately reflecting the true physiological state of personnel. This application comprehensively assesses the overall impact of the current work environment by weighted summation of the normalized values ​​of multiple work environment characteristics. This weighted summation method effectively integrates environmental data of different types and dimensions, assigning different weights according to their importance to the physiological state, ensuring that the calculation of the correction coefficient is both comprehensive and targeted. This represents the relative importance of the j-th work environment feature in influencing the physiological state of industrial workers. Different environmental factors have varying degrees of impact on human physiological responses; for example, extreme temperatures may have a greater impact on physiological state than slight noise. These weighting coefficients can be set through expert experience or trained and optimized through historical data analysis and machine learning algorithms (such as regression analysis) to reflect their true impact.

[0029] Specifically, this application first acquires various working environment characteristic data, such as temperature, humidity, and noise. Given that these environmental characteristics may have different dimensions and numerical ranges, in order to effectively integrate them, this scheme normalizes each environmental characteristic data to obtain... Normalized data eliminates dimensional differences, ensuring data comparability in calculations. Furthermore, considering the varying degrees of impact of different environmental characteristics on the physiological state of industrial workers, this scheme applies a different approach to each normalized environmental characteristic. Assign a weight coefficient These weighting coefficients reflect the importance of each environmental characteristic; for example, in a high-temperature working environment, temperature may have a much higher weight than humidity. Subsequently, by multiplying each normalized environmental characteristic by its corresponding weighting coefficient and summing all products, a work environment correction coefficient is obtained. This weighted summation method allows the correction coefficient to comprehensively reflect the overall impact of the current working environment and can be dynamically adjusted according to the actual importance of each environmental factor. The generated correction coefficient is then used in the physiological characteristic matching analysis model, for example, to correct for physiological characteristic deviations, thereby more accurately assessing the physiological state of industrial personnel and effectively compensating for the impact of environmental interference on the interpretation of physiological data. In this way, this solution can provide a more accurate and robust physiological characteristic matching degree, thereby improving the accuracy and reliability of real-time identity authentication and risk warning for industrial personnel.

[0030] Through the above technical solution, this application provides a clear and quantifiable method for generating work environment correction coefficients. This method, by normalizing and weighted summing multiple work environment feature data, can comprehensively and dynamically assess the impact of environmental factors on the physiological state of industrial personnel. This effectively solves the problems of unclear work environment correction coefficient generation methods and inaccurate calculations in traditional methods, thereby enabling physiological feature matching analysis models to more accurately compensate for interference caused by environmental changes. Therefore, this solution significantly improves the accuracy and reliability of real-time identity authentication for industrial personnel, reduces the misjudgment rate caused by environmental factors, and provides a more accurate risk warning basis for industrial safety production.

[0031] In the scheme proposed in this application, a physiological feature deviation degree is used to quantify the changes in physiological features to support identity authentication. However, in its implementation, due to the lack of sufficient consideration of the relative importance of different physiological features and the differences in the range of historical changes, the deviation degree calculation lacks robustness and comparability, thereby reducing the accuracy of physiological feature matching degree and the reliability of subsequent risk warning.

[0032] In response, this invention further proposes a method for generating the aforementioned physiological characteristic deviation degree, specifically including: Through the formula: ; Generate physiological characteristic deviation degree ; In the formula, This represents the data of the i-th physiological characteristic at time t. This represents the baseline value of the i-th physiological characteristic data at time t. It represents the historical standard deviation of the i-th physiological characteristic data at similar times to time t in the historical data, and n represents the number of physiological characteristic data types. Among them, the baseline value of the i-th physiological characteristic data at time t refers to the expected or standard value of the i-th physiological characteristic data of industrial personnel at time t under normal working conditions; the baseline value can be personalized, that is, pre-calibrated and stored for each industrial personnel, for example, by taking the average value of multiple measurements under the health condition of the personnel or establishing a personal physiological baseline model. The historical standard deviation of the i-th physiological characteristic data at similar times to the current time t in historical data refers to the degree of fluctuation of the i-th physiological characteristic data at times similar to the current time t in historical data (e.g., the same time period or the same work shift on the same day); this standard deviation is used to measure the natural range of variation of physiological characteristics under normal conditions. The weighting coefficient of the i-th physiological characteristic data refers to the relative importance assigned to different physiological characteristic data when calculating the deviation degree of physiological characteristics. Different physiological characteristics may have different sensitivities and indicative significance for the identification and risk warning of industrial personnel.

[0033] This application's solution generates physiological characteristic deviation using a formulaic method, solving the problem of inaccurate deviation calculation. This method comprehensively considers the weighting of multiple physiological characteristics and historical fluctuations, ensuring higher robustness and comparability of the deviation. Specifically, the method first acquires multiple physiological characteristic data of industrial personnel at time t and compares them with pre-set or dynamically adjusted benchmark values ​​to quantify the deviation of the current physiological state from the normal state. To make this deviation comparable and reflect its relative degree within the normal fluctuation range, the deviation value is further normalized by dividing it by the historical standard deviation of the i-th physiological characteristic data at similar times in historical data, thereby eliminating differences in the dimensions and fluctuation ranges of different physiological characteristics. Based on this, considering the varying importance of different physiological characteristics in indicating personnel status, weighting coefficients are introduced to weight the normalized deviation values. Finally, the squares of all weighted physiological characteristic deviation values ​​are summed and the square root is taken to generate a comprehensive physiological characteristic deviation. This calculation method not only effectively integrates deviation information from various physiological characteristics, but also prevents a single outlier from excessively affecting the overall result through square root operations, ensuring the stability and accuracy of the deviation calculation. The generated physiological characteristic deviation is then input into the physiological characteristic matching analysis model as a key parameter for generating physiological characteristic matching scores, thereby providing an accurate and reliable basis for physiological state assessment for subsequent identity authentication and risk warning.

[0034] Through the above technical solution, the calculation of physiological characteristic deviation is no longer a simple absolute difference, but rather comprehensively considers the weighting coefficients of different physiological characteristics, allowing for a more complete reflection of deviations in key physiological indicators. Simultaneously, by introducing the baseline value and historical standard deviation of the i-th physiological characteristic data at time t, dynamic and personalized assessment of the degree of physiological characteristic deviation is achieved, effectively avoiding misjudgments caused by a single static threshold. This normalization and weighted summation method makes the deviations between different physiological characteristics comparable and reasonably controls the sensitivity to outliers, significantly improving the robustness and accuracy of physiological characteristic deviation. Therefore, when this physiological characteristic deviation is used in subsequent physiological characteristic matching analysis models, it can provide more accurate and reliable input, thereby significantly improving the accuracy of real-time identity authentication for industrial personnel and the timeliness and reliability of risk warnings, effectively solving the problem of lack of robustness and comparability in deviation calculation in traditional methods.

[0035] In some of the embodiments described above in this application, a physiological feature matching analysis model is proposed to generate physiological feature matching degree. However, in its implementation process, there is a lack of a specific mathematical model to accurately calculate the matching degree, which makes it impossible to effectively compensate for the impact of changes in the working environment on physiological features, thereby affecting the accuracy and reliability of identity authentication.

[0036] In response, this invention further proposes the following expression for the physiological feature matching analysis model: ; In the expression, This represents the degree of physiological characteristic matching at time t. This represents the degree of deviation in physiological characteristics. The normalized value, This represents the work environment correction factor. This represents the sensitivity adjustment coefficient; Among them, the physiological characteristic matching analysis model aims to quantify the degree of matching between the physiological state of industrial personnel and the baseline state; its core function is to transform multi-dimensional physiological data changes into a unified matching index to reflect the health, fatigue or abnormal state of personnel. The physiological characteristic matching degree at time t represents the degree to which the physiological characteristics of industrial personnel conform to the preset normal or baseline physiological state at a specific time point t; this matching degree can serve as an important input for subsequent identity authentication and risk warning, reflecting the real-time health or abnormal level of personnel's physiological state; The sensitivity adjustment coefficient is used to adjust the sensitivity of the physiological characteristic matching degree to changes in physiological characteristic deviation. A larger sensitivity adjustment coefficient will make the matching degree respond more drastically to physiological deviations, meaning that even small physiological deviations will lead to a significant decrease in the matching degree; while a smaller sensitivity adjustment coefficient will make the matching degree respond more gradually to physiological deviations. This coefficient can be set according to the stringency requirements of risk warning in the actual application scenario. For example, in high-risk operation scenarios, a higher sensitivity adjustment coefficient can be set to improve the sensitivity of the warning.

[0037] Specifically, when the physiological characteristic data of industrial workers deviates from the baseline value, a physiological characteristic deviation degree is generated. This deviation degree is normalized to obtain a normalized value, which is used to quantify the degree of abnormality in the physiological state. Simultaneously, the system acquires work environment characteristic data in real time and generates a work environment correction coefficient accordingly. This correction coefficient reflects the potential impact of the current environment on the physiological state. For example, in harsh environments, even slight fluctuations in physiological indicators may be considered normal, while in comfortable environments they may be considered abnormal. In the expression, This method corrects for the normalized physiological deviation, achieving dynamic compensation for environmental factors. When the working environment is harsh, A larger value amplifies the negative impact of physiological bias on the matching degree; conversely, when the working environment is good, When the value of the sensitivity adjustment coefficient is small, the negative impact of physiological deviation on the matching degree is weakened; the sensitivity adjustment coefficient further controls the overall response strength of the matching degree to physiological deviation and environmental correction. The entire exponential function structure ensures that the physiological characteristic matching degree is always between 0 and 1, and that the matching degree decreases non-linearly with the increase of physiological deviation or the adjustment of the environmental correction term, thus accurately and robustly reflecting the real-time physiological state of industrial personnel. This model not only solves the problem of inaccurate calculation of physiological characteristic matching degree in traditional methods, but also effectively compensates for the impact of environmental changes on physiological characteristics by introducing a work environment correction coefficient, making the assessment of physiological state more objective and accurate, and providing a reliable physiological state basis for subsequent identity authentication and risk warning.

[0038] Through the above technical solution, this application provides a precise and robust method for calculating physiological characteristic matching degree. This method, by introducing a working environment correction coefficient, can dynamically compensate for the impact of environmental changes on physiological characteristic data, avoiding misjudgments of abnormalities due to normal fluctuations in physiological indicators under different environmental conditions, and significantly improving the accuracy of physiological state assessment. Simultaneously, through a sensitivity adjustment coefficient, the system can adjust the response intensity to physiological deviations according to actual needs, making the early warning mechanism more flexible and adaptable. This precise calculation of physiological characteristic matching degree provides a more reliable and refined physiological state basis for real-time identity authentication and risk early warning of industrial personnel, effectively reducing false alarm and false negative rates, thereby improving the intelligent level of industrial safety management.

[0039] In the scheme proposed in this application, a method for generating behavioral feature matching degree is used to generate behavioral feature matching degree as part of the comprehensive matching degree. However, in its implementation process, the behavioral feature data is not fully decomposed and quantified, resulting in a lack of fine capture of behavioral dynamics in the matching degree calculation, which affects the accuracy of identity authentication and the reliability of risk warning.

[0040] In response, this invention further proposes a method for generating the behavioral feature matching degree, specifically including: Step S31: Generate activity intensity and attitude change rate based on behavioral acceleration and behavioral angular velocity, respectively; Step S32: Establish a behavioral feature analysis model based on activity intensity and posture change rate, and generate behavioral feature matching degree; Among them, behavioral acceleration refers to the change in the speed of an industrial worker's body or worn device in space when performing activities; it can reflect the movement state of a person, such as walking, running, jumping or standing still; behavioral acceleration can be collected in real time by an inertial measurement unit (IMU) or accelerometer worn by the industrial worker. For example, a three-axis accelerometer can be used to obtain the acceleration components in the X, Y, and Z directions, or it can be obtained by an accelerometer integrated into a smart wearable device; Behavioral angular velocity refers to the speed at which an industrial worker's body or worn equipment rotates in space during activities. It can reflect changes in a person's posture, such as body rotation, bending over, or falling. Behavioral angular velocity can be collected in real time by an inertial measurement unit (IMU) or gyroscope sensor worn by the industrial worker. For example, a three-axis gyroscope can be used to obtain the angular velocity components around the X, Y, and Z axes, or it can be obtained through the gyroscope built into a smartphone. Activity intensity is a quantitative indicator of the intensity of industrial workers' behavior. It comprehensively reflects the level of activity of people within a certain period of time and can be calculated based on behavioral acceleration data. The rate of change of posture is a quantitative indicator of the dynamic changes in the body posture of industrial workers. It reflects the stability and frequency of change of the body posture of workers and can be calculated based on behavioral angular velocity data. Behavioral feature matching degree is a quantitative indicator that measures the degree to which an industrial worker's current behavioral pattern matches a preset or historical normal behavioral pattern. The higher the matching degree, the more the current behavior conforms to the normal or expected behavioral pattern; conversely, abnormal behavior may exist. The behavioral feature matching degree is usually a value between 0 and 1, where 1 represents a perfect match and 0 represents a complete mismatch.

[0041] Specifically, step S31 transforms the raw, multi-dimensional behavioral acceleration and angular velocity data into more representative and interpretable high-level behavioral characteristics, activity intensity, and posture change rate. This transformation effectively reduces data dimensionality, removes noise, and highlights the core dynamic information of the behavior, providing a more refined and effective input for subsequent behavioral analysis. Step S32 uses the activity intensity and posture change rate obtained in step S31 to construct a behavioral feature analysis model, thereby generating a behavioral feature matching degree. Through this model, these quantified behavioral features can be compared with preset normal behavioral patterns to assess the degree of matching of the current behavior.

[0042] This application further optimizes the generation process of behavioral feature matching degree based on the aforementioned real-time identity authentication and risk warning method for industrial personnel. Specifically, the scheme first refines the raw behavioral acceleration and behavioral angular velocity data obtained from industrial personnel in step S31. Behavioral acceleration data is used to generate activity intensity, making the quantification of the intensity of personnel movement more intuitive and effective, avoiding the complexity and noise interference that may result from directly using raw acceleration data. Simultaneously, behavioral angular velocity data is used to generate posture change rate, which accurately captures the dynamic changes in personnel's body posture, such as bending, turning, and falling, providing key information for identifying abnormal postures. This transformation process refines raw, high-dimensional sensor data into more physically meaningful and interpretable behavioral indicators. Subsequently, in step S32, a behavioral feature analysis model is established using the activity intensity and posture change rate generated in step S31. This model uses these refined behavioral indicators as input, and through a preset algorithm or learned patterns, comprehensively evaluates the degree of conformity between the current behavioral pattern of industrial personnel and the normal or expected behavioral pattern, and outputs the behavioral feature matching degree. This high-level behavioral feature-based analysis makes the calculation of behavioral feature matching degrees more accurate and robust, effectively distinguishing normal operational behavior from potential abnormal or dangerous behavior. Through this approach, the solution in this application addresses the problem of insufficient decomposition and quantification of behavioral feature data in traditional methods. It transforms raw behavioral data into more representative activity intensity and posture change rates, enabling the behavioral feature analysis model to capture behavioral dynamics more precisely. This improvement not only enhances the accuracy of behavioral feature matching degrees but also provides more reliable input for subsequent comprehensive matching degree calculations and identity authentication index generation, thereby improving the accuracy of real-time identity authentication for industrial personnel and the reliability of risk warnings.

[0043] Through the aforementioned technical solution, this application can transform raw behavioral acceleration and angular velocity data into more physically meaningful and interpretable activity intensity and attitude change rate, thereby enabling the behavioral feature analysis model to capture the behavioral dynamics of industrial personnel more precisely. This refined behavioral feature extraction and analysis significantly improves the accuracy and robustness of behavioral feature matching, effectively avoiding the matching degree calculation bias caused by insufficient decomposition and quantification of behavioral feature data in traditional methods. Ultimately, this provides more accurate behavioral basis for real-time identity authentication of industrial personnel, reduces false alarm rates, and makes risk warnings more timely and reliable, thereby enhancing the assurance capability of industrial safety production.

[0044] In the scheme of this application, a method for generating behavioral feature matching degree is proposed to calculate the behavioral feature matching degree. However, in its implementation process, the calculation of activity intensity and posture change rate lacks specific and quantitative methods, which may lead to inaccurate matching degree calculation and affect the reliability of identity authentication and risk warning.

[0045] In response, the present invention further proposes a method for generating the activity intensity, specifically including: Through the formula: ; Generate activity intensity ; In the formula, This represents the normalized value of the acceleration in the X-axis direction at time t. This represents the normalized value of the acceleration in the y-axis direction at time t. This represents the normalized value of the acceleration in the z-axis direction at time t; The specific methods for generating the attitude change rate include: Through the formula: ; Generate attitude change rate ; In the formula, This represents the normalized value of the angular velocity of rotation about the X-axis at time t. This represents the normalized value of the angular velocity about the y-axis at time t. It represents the normalized value of the angular velocity of rotation about the z-axis at time t; Among them, activity intensity is an indicator that measures the overall intensity of movement of industrial personnel at a certain moment t. It integrates the linear movement changes of personnel in three-dimensional space and reflects the activity level of personnel. The rate of change of posture is an indicator that measures how quickly an industrial worker's body posture or orientation changes at a given moment t. It integrates the changes in rotational motion of a person in three-dimensional space and reflects the stability or dynamism of the person's posture.

[0046] This application introduces specific mathematical formulas to accurately quantify the activity intensity and posture change rate of industrial personnel. Specifically, the activity intensity is calculated based on the square root of the sum of the squares of the normalized values ​​of acceleration values ​​in the X, Y, and Z axes at time t. This calculation method comprehensively reflects the intensity of the overall linear motion of industrial personnel in three-dimensional space, avoiding the bias that may arise from data from a single axis. Simultaneously, the posture change rate is calculated based on the square root of the sum of the squares of the normalized values ​​of the angular velocities around the X, Y, and Z axes at time t. This calculation method accurately captures the overall rotational motion changes of industrial personnel in three-dimensional space, reflecting the dynamics of their posture. By normalizing the acceleration and angular velocity values, the comparability and consistency between data from different sensors or individuals are ensured, providing standardized, high-quality input for subsequent behavioral feature analysis models. This precise quantification method makes the behavioral feature data more accurate and robust, thereby enabling more reliable generation of behavioral feature matching scores, and ultimately improving the accuracy and reliability of real-time identity authentication and risk warning for industrial personnel. Compared with traditional methods that rely solely on fuzzy or non-quantitative descriptions, this approach effectively addresses the lack of specific and quantitative methods for activity intensity and posture change rate by providing explicit calculation formulas, thus providing a solid data foundation for behavioral feature analysis models.

[0047] Through the above technical solution, this application provides a clear and quantitative calculation method for the activity intensity and posture change rate of industrial personnel. This method, based on the normalized values ​​of triaxial acceleration and triaxial angular velocity for vector magnitude calculation, can comprehensively and accurately capture the intensity of linear motion and the rate of change of rotational posture of personnel. This effectively solves the problem of the lack of specific methods for calculating activity intensity and posture change rate in existing technologies, thus providing high-quality, standardized input data for behavioral feature analysis models. Therefore, the calculation of behavioral feature matching degree will be more accurate and reliable, significantly improving the accuracy and robustness of real-time identity authentication and risk warning for industrial personnel, reducing the false alarm rate, and enabling the warning system to identify potential risks more promptly and effectively.

[0048] In this application, a behavioral feature analysis model is proposed to generate behavioral feature matching degree. However, in this process, the model lacks dynamic quantitative combination of behavioral continuity and stagnation state, resulting in incomplete calculation of behavioral feature matching degree and inability to accurately reflect the real-time behavioral state of industrial personnel, thereby affecting the reliability of identity authentication and the timeliness of risk warning.

[0049] In response, this invention further proposes the following expression for the behavioral feature analysis model: ; In the expression, This represents the matching degree of behavioral features at time t. This represents the continuity index of behavior at time t. This represents the behavioral stagnation index at time t. This represents the weighting coefficient for behavioral continuity; Specifically, the generation method of the behavioral continuity index at time t includes: Through the formula: ; Generate the behavioral continuity index at time t ; In the formula, This represents the activity intensity at time t-k+1. This represents the rate of attitude change at time t-k+1. This represents the historical average activity intensity. This represents the historical average attitude change rate, and N represents the number of sampling points within the time window. , All are weighting coefficients, and ; Specifically, the generation method of the behavioral stagnation index at time t includes: Through the formula: ; Generate the behavior stagnation index at time t ; In the formula, This represents the activity intensity at time t-g+1. This represents the activity intensity threshold. For indicator functions, M represents the number of sampling points within the time window; The expression of the behavioral feature analysis model is used to comprehensively evaluate the behavioral state of industrial personnel at time t. By weighted combination of behavioral continuity index and behavioral stagnation index, a behavioral feature matching degree is generated. Its role is to provide a quantitative indicator that reflects the normality or abnormality of industrial personnel's behavior. The behavioral characteristic matching degree at time t represents the degree to which the behavioral characteristics of industrial personnel match the preset normal behavioral pattern at a specific time t. The higher the value, the more the behavior conforms to the normal pattern; conversely, abnormal behavior may exist. The behavioral continuity index at time t is used to quantify the activity level and consistency of actions of industrial workers over a period of time; a high continuity index usually indicates that workers are engaged in purposeful and continuous work activities. The behavior stagnation index at time t is used to quantify the degree to which industrial workers are physically inactive or less active than normal over a period of time. A high stagnation index may indicate that a person is in an abnormal state such as being still, unconscious, fallen, or inactive for a long period of time. The weighting coefficient for behavioral continuity is used to adjust the relative importance of the behavioral continuity index in the calculation of behavioral feature matching degree; its value ranges from 0 to 1. When the weighting coefficient for behavioral continuity is large, it indicates that more emphasis is placed on the continuity of behavior; when the weighting coefficient for behavioral continuity is small, it indicates that more emphasis is placed on the stagnant state of behavior; this weighting coefficient can be set by expert experience, for example, by adjusting it according to the risk characteristics of different work scenarios. The behavior continuity index at time t is generated by weighting the activity intensity and posture change rate over a period of time and normalizing it with the historical average activity intensity and historical average posture change rate, thereby quantifying the continuity of behavior. Weighting coefficient , Used to balance the contributions of activity intensity and posture change rate in the calculation of the behavioral continuity index; The behavior stagnation index at time t is generated by counting the number of times the activity intensity falls below a preset threshold over a period of time, thereby quantifying the degree of behavioral stagnation. The activity intensity threshold is a preset critical value; when the activity intensity of industrial workers falls below this threshold, they are considered to be in a stagnant state. This threshold can be set according to specific industrial scenarios and operational requirements. The indicator function is a binary function that outputs 1 when the condition within its parentheses is true and 0 when the condition is false.

[0050] Through the above technical solution, this application addresses the problem in real-time identification and risk warning methods for industrial personnel that the calculation of behavioral feature matching degree lacks dynamic quantitative integration of behavioral continuity and stagnation. By introducing a behavioral continuity index and a behavioral stagnation index, and using a weighted combination to generate behavioral feature matching degree, the assessment of industrial personnel's behavioral status becomes more comprehensive and refined. Specifically, the behavioral continuity index effectively captures the activity level and movement consistency of personnel, while the behavioral stagnation index promptly identifies abnormal situations such as prolonged inactivity or lack of activity. The introduction of a weighting coefficient for behavioral continuity allows the system to flexibly adjust the importance of behavioral continuity and stagnation in the overall assessment according to different work scenarios and risk characteristics, thereby avoiding false alarms and missed alarms caused by single indicators or static threshold judgments in traditional methods. This dynamic, multi-dimensional behavioral feature analysis model significantly improves the accuracy and reliability of behavioral feature matching degree, thereby enhancing the precision of industrial personnel identification and enabling more timely and effective early warning of potential risks, ensuring industrial production safety.

[0051] In the scheme of this application, a method for generating a comprehensive matching degree is proposed to combine physiological feature matching degree and behavioral feature matching degree to generate a comprehensive matching degree. However, in its implementation process, due to the fixed weight coefficient or failure to consider the dynamic changes of the working environment, the comprehensive matching degree is not calculated accurately and cannot adapt to the authentication requirements under different environmental conditions in real time, thereby affecting the accuracy and reliability of identity authentication.

[0052] In response, this invention further proposes a method for generating the comprehensive matching degree, specifically including: Through the formula: ; Generate comprehensive matching score ; In the formula, This represents the degree of physiological characteristic matching at time t. This represents the matching degree of behavioral features at time t. This represents the weighting coefficient of the physiological characteristic matching degree at time t; The specific methods for obtaining the weight coefficients of the physiological feature matching degree at time t include: Through the formula: ; Weight coefficients for generating physiological feature matching at time t ; In the formula, This indicates the stability of the work environment. This represents the adjustment coefficient; where, the stability of the working environment refers to the normalized value of the changes in the working environment parameters. Among them, the overall matching degree is an indicator used to measure the degree to which the overall state of industrial personnel at a specific moment conforms to the expected benchmark, aiming to provide a comprehensive and quantitative basis for identity authentication and risk assessment. The weighting coefficient of the physiological characteristic matching degree at time t is a dynamically adjusted parameter used to balance the contributions of physiological characteristic matching degree and behavioral characteristic matching degree under different environmental conditions when calculating the overall matching degree. This indicates the stability of the working environment, which refers to the normalized value of the changes in working environment parameters. The stability of the working environment is an indicator that measures the degree of fluctuation of various parameters (such as temperature, humidity, noise, light, concentration of harmful gases, etc.) in the industrial working environment. It reflects the degree to which the environment may affect the physiological and behavioral state of personnel. This represents the adjustment coefficient, which is a control weighting coefficient. A parameter related to the sensitivity of the response to the stability of the working environment; this coefficient determines the steepness of the sigmoid function curve, thus affecting... Follow The speed and magnitude of change; larger The value will make right The physiological characteristics are more sensitive to changes; that is, even slight changes in environmental stability can significantly adjust the weights of physiological characteristic matching. The value will make The changes are more gradual; this coefficient can be set according to the needs and experience of actual application scenarios to achieve the best weight adjustment effect.

[0053] Through the above technical solution, this application can dynamically adjust the weight of physiological characteristic matching degree in the comprehensive evaluation according to the real-time changes in the working environment, thereby solving the problem of inaccurate comprehensive matching degree calculation caused by fixed weights or failure to consider dynamic environmental changes in traditional methods. This adaptive weight adjustment mechanism enables the comprehensive matching degree to more accurately and robustly reflect the true state of industrial personnel, especially in complex, changeable, or high-risk industrial environments, effectively reducing the impact of environmental interference on identity authentication results. Given that physiological characteristic data and behavioral characteristic data reflect the state of personnel from different dimensions, this solution uses dynamic weight fusion to enable the two to complement each other. When one dimension of data is greatly affected by the environment, the other dimension of data can play a greater role, thereby significantly improving the accuracy and reliability of real-time identity authentication of industrial personnel, providing a more solid foundation for timely detection of abnormal personnel states and risk warning.

[0054] In this application, an identity authentication index is generated based on the comprehensive matching degree for real-time authentication and risk warning. However, in this process, due to the lack of a specific quantitative model and dynamic adjustment mechanism, the authentication result may rely too much on simple threshold judgment, resulting in insufficient sensitivity to fluctuations in the matching degree, which can easily lead to false alarms or missed alarms and cannot adapt to the complex and ever-changing real-time needs in industrial scenarios.

[0055] In response, this invention further proposes a method for generating the identity authentication index, specifically including: Through the formula: ; Generate identity authentication index ; In the formula, This represents the overall matching degree at time t. This represents the decision threshold for overall matching degree. This is the slope parameter; Among them, the generation of the identity authentication index is the core output of this method, which aims to provide a quantitative and easy-to-understand indicator for real-time assessment of the authenticity of the identity of industrial personnel and their potential risk status. The identity authentication index is calculated using a sigmoid function, which maps the input value (i.e., the normalized deviation of the overall matching degree relative to the decision threshold) to a continuous output value, typically between 0 and 1. This non-linear mapping ensures that the identity authentication index transitions smoothly as the input changes, thus providing a more refined evaluation result than simple binary judgment. This represents the decision threshold for the overall matching degree. In the formula, it serves as a reference benchmark and defines the expected or acceptable level of the overall matching degree. This threshold can be set based on historical data analysis, expert experience, or determined through the system calibration process. For example, it can be set as the average value or safety lower limit of the overall matching degree under normal working conditions. The slope parameter is used to adjust the steepness of the Sigmoid curve, thereby controlling the sensitivity of the identity authentication index to changes in the overall matching degree; a larger slope parameter indicates a more stable Sigmoid curve. The value will make the curve steeper, meaning Tiny changes can lead to Significant changes, suitable for scenarios requiring highly sensitive early warning; smaller The value makes the curve smoother, allowing It has a larger fluctuation range without causing The parameter can be configured by the system administrator based on the risk level and management strategy of the actual application scenario, or it can be dynamically adjusted according to real-time operating conditions.

[0056] This method introduces the Sigmoid function to generate an identity authentication index, achieving a non-linear transformation of the overall matching degree. This continuous mapping mechanism ensures that even with small changes in the overall matching degree, the identity authentication index changes smoothly accordingly, providing a more refined and dynamic evaluation result than traditional simple threshold judgment. The decision threshold serves as a benchmark, clearly defining what kind of overall matching degree is considered normal or acceptable, giving the index generation an objective comparison standard. The slope parameter gives the system the ability to dynamically adjust its sensitivity. By adjusting the slope parameter, the response speed and magnitude of the identity authentication index to changes in the overall matching degree can be flexibly controlled according to the actual risk level and management needs of the industrial site. For example, in high-risk operating environments, a larger slope parameter can be set to enable the system to respond quickly to abnormal situations; while in general operating environments, a smaller slope parameter can be set to avoid frequent false alarms caused by oversensitivity. This mechanism, combined with the comprehensive matching degree generated by integrating physiological characteristic data, work environment characteristic data, and behavioral characteristic data in the preceding steps, enables the final identity authentication index to not only be based on multi-dimensional, real-time data, but also to conduct risk assessment in a highly configurable and dynamically adaptive manner, thereby significantly improving the accuracy and robustness of real-time identity authentication and risk warning for industrial personnel.

[0057] Through the above technical solution, this application provides a method for generating an identity authentication index based on the Sigmoid function, effectively solving the problem of the lack of a specific quantitative model and dynamic adjustment mechanism in traditional methods. This method significantly improves the system's sensitivity to fluctuations in the matching degree by mapping the comprehensive matching degree to a continuous identity authentication index, thereby avoiding false alarms and false negatives caused by simple threshold judgments. Simultaneously, the introduction of a slope parameter allows the system to flexibly adjust the strictness of authentication and the timeliness of early warnings according to the risk level and management needs of actual industrial scenarios, thus better adapting to the complex and ever-changing real-time industrial environment and improving the accuracy and reliability of real-time identity authentication and risk warning for industrial personnel.

[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for real-time identity authentication and risk warning of industrial personnel, characterized in that, The method specifically includes: Acquire physiological, work environment, and behavioral data of industrial personnel; A physiological characteristic matching analysis model is established based on the physiological characteristic data of industrial personnel and the characteristic data of their work environment, and a physiological characteristic matching degree is generated. A behavioral characteristic analysis model is established based on the behavioral characteristic data of industrial personnel to generate a behavioral characteristic matching degree; the behavioral characteristic data includes behavioral acceleration and behavioral angular velocity. A comprehensive matching score is generated based on the matching scores of physiological characteristics and behavioral characteristics. An identity authentication index is generated based on the overall matching degree; Based on the identity authentication index, real-time identity authentication and risk warning are conducted for industrial personnel.

2. The method for real-time identity authentication and risk warning of industrial personnel according to claim 1, characterized in that, The specific methods for generating the physiological feature matching degree include: Generate work environment correction coefficients based on work environment characteristic data; Generate physiological characteristic deviation based on physiological characteristic data; A physiological characteristic matching analysis model is established based on the work environment correction coefficient and the physiological characteristic deviation degree to generate the physiological characteristic matching degree.

3. The method for real-time identity authentication and risk warning of industrial personnel according to claim 2, characterized in that, The specific methods for generating the working environment correction coefficient include: Through the formula: ； Generate working environment correction factor ; In the formula, This represents the normalized value of the j-th work environment feature data. This represents the weight coefficient of the j-th work environment feature, and m represents the number of work environment features.

4. The method for real-time identity authentication and risk warning of industrial personnel according to claim 2, characterized in that, The specific methods for generating the physiological characteristic deviation include: Through the formula: ； Generate physiological characteristic deviation degree ; In the formula, This represents the data of the i-th physiological characteristic at time t. This represents the baseline value of the i-th physiological characteristic data at time t. It represents the historical standard deviation of the i-th physiological characteristic data at similar times to time t in the historical data, and n represents the number of physiological characteristic data types.

5. The method for real-time identity authentication and risk warning of industrial personnel according to claim 2, characterized in that, The specific expression of the physiological feature matching analysis model is as follows: ； In the expression, This represents the degree of physiological characteristic matching at time t. This represents the degree of deviation in physiological characteristics. The normalized value, This represents the work environment correction factor. This represents the sensitivity adjustment coefficient.

6. The method for real-time identity authentication and risk warning of industrial personnel according to claim 1, characterized in that, The specific methods for generating the behavioral feature matching degree include: Activity intensity and attitude change rate are generated based on behavioral acceleration and behavioral angular velocity, respectively; A behavioral feature analysis model is established based on activity intensity and posture change rate to generate behavioral feature matching degree.

7. The method for real-time identity authentication and risk warning of industrial personnel according to claim 6, characterized in that, The specific methods for generating the activity intensity include: Through the formula: ； Generate activity intensity ; In the formula, This represents the normalized value of the acceleration in the X-axis direction at time t. This represents the normalized value of the acceleration in the y-axis direction at time t. This represents the normalized value of the acceleration in the z-axis direction at time t; The specific methods for generating the attitude change rate include: Through the formula: ； Generate attitude change rate ; In the formula, This represents the normalized value of the angular velocity of rotation about the X-axis at time t. This represents the normalized value of the angular velocity about the y-axis at time t. It represents the normalized value of the angular velocity of the rotation about the z-axis at time t.

8. The method for real-time identity authentication and risk warning of industrial personnel according to claim 6, characterized in that, The specific expression of the behavioral feature analysis model is as follows: ； In the expression, This represents the matching degree of behavioral features at time t. This represents the continuity index of behavior at time t. This represents the behavioral stagnation index at time t. This represents the weighting coefficient for behavioral continuity; Specifically, the generation method of the behavioral continuity index at time t includes: Through the formula: ； Generate the behavioral continuity index at time t ; In the formula, This represents the activity intensity at time t-k+1. This represents the rate of attitude change at time t-k+1. This represents the historical average activity intensity. This represents the historical average attitude change rate, and N represents the number of sampling points within the time window. , All are weighting coefficients, and ; Specifically, the generation method of the behavioral stagnation index at time t includes: Through the formula: ； Generate the behavior stagnation index at time t ; In the formula, This represents the activity intensity at time t-g+1. This represents the activity intensity threshold. M is an indicator function, representing the number of sampling points within the time window.

9. The method for real-time identity authentication and risk warning of industrial personnel according to claim 1, characterized in that, The specific methods for generating the comprehensive matching degree include: Through the formula: ； Generate comprehensive matching score ; In the formula, This represents the degree of physiological characteristic matching at time t. This represents the matching degree of behavioral features at time t. This represents the weighting coefficient of the physiological characteristic matching degree at time t; The specific methods for obtaining the weight coefficients of the physiological feature matching degree at time t include: Through the formula: ； Weight coefficients for generating physiological feature matching at time t ; In the formula, This indicates the stability of the work environment. This represents the adjustment coefficient; where, the stability of the working environment refers to the normalized value of the changes in the working environment parameters.

10. The method for real-time identity authentication and risk warning of industrial personnel according to claim 1, characterized in that, The specific methods for generating the identity authentication index include: Through the formula: ； Generate identity authentication index ; In the formula, This represents the overall matching degree at time t. This represents the decision threshold for overall matching degree. This is the slope parameter.

Images