Health monitoring and early warning method and system of intelligent wearable device

Through smart wearable devices, a personalized user standard map and health warning decision-making engine are built, which solves the problem that traditional devices cannot fully reflect health conditions and inaccurate warnings, and achieves the improvement of personalized health management and early warnings.

CN119014842AInactive Publication Date: 2024-11-26SHENZHEN PERCUSSION TECH CO LTD
View PDF 0 Cites 26 Cited by

Patent Information

Application Number
CN202411362658.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-27
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional smart wearable devices have problems in health monitoring that cannot fully reflect human health status, inaccurate and timely health warnings, and it is difficult to provide personalized health management services.

Method used

By obtaining real-time and historical heart rate monitoring data, interpolation filtering optimization, personalized user feature mining, building a personalized user standard map, analyzing heart rate waveforms, calculating dynamic variation amplitudes, making multi-period situation evolution prediction, identifying abnormal stages, and based on personality Comprehensive evaluation to build a health warning decision engine.

Benefits of technology

It has achieved in-depth understanding and prediction of user heart rate changes, improved the accuracy and timeliness of health warnings, provided personalized health management services, and enhanced the application and promotion of smart wearable devices in the field of health management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119014842A_ABST
    Figure CN119014842A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of health monitoring, in particular to a health monitoring and early warning method and system for intelligent wearable equipment. The method comprises the following steps: acquiring real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on intelligent wearable equipment; performing interpolation filtering optimization on historical heart rate monitoring data, performing personalized user feature mining, and constructing a personalized user standard atlas; multi-point interval sampling is conducted on real-time user heart rate monitoring parameters, heart rate waveform fitting is conducted, and a real-time heart rate waveform curve is constructed; carrying out heart rate variability analysis on the real-time heart rate waveform curve, and carrying out dynamic variation amplitude calculation so as to obtain a user heart rate time sequence variation representation; and carrying out heart rate change trend analysis on the user heart rate time sequence change representation, and then carrying out multi-period situation evolution prediction to generate multi-period heart rate situation evolution prediction data. According to the invention, accurate and efficient health early warning is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring, and in particular to a health monitoring and early warning method and system for a smart wearable device. Background Art

[0002] As people's quality of life continues to improve, they are paying more and more attention to their health. Smart wearable devices, such as smart watches and health bracelets, have become popular health management tools due to their portability and real-time monitoring capabilities. These devices can collect users' physiological data in real time, such as heart rate, blood pressure, activity level, etc., and provide users with real-time feedback on their health status.

[0003] However, traditional smart wearable health monitoring methods have some limitations. These devices can usually only monitor a single physiological indicator and cannot fully reflect the health status of the human body. Most of the collected data are only simply analyzed and displayed, making it difficult to discover potential health risks. The existing health warning function is too simple and cannot provide personalized health tips and guidance. There are often problems with inaccurate and untimely health warnings. These defects hinder the further application and promotion of smart wearable devices in the field of health management. Therefore, there is an urgent need for a new smart wearable health monitoring and early warning method that can overcome the above problems and provide users with more intelligent, comprehensive and personalized health management services. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a health monitoring and early warning method and system for a smart wearable device to solve at least one of the above technical problems.

[0005] To achieve the above object, the present invention provides a health monitoring and early warning method for a smart wearable device, comprising the following steps: Step S1: obtaining real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on the smart wearable device; interpolating and filtering the historical heart rate monitoring data, and mining personalized user features to construct a personalized user standard map; Step S2: performing multi-point interval sampling on the real-time user heart rate monitoring parameters, and performing heart rate waveform fitting to construct a real-time heart rate waveform curve; Step S3: Analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series change; Step S4: analyzing the heart rate change trend of the user's heart rate time series change representation, and then predicting the multi-period situation evolution to generate multi-period heart rate situation evolution prediction data; Step S5: inferring the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, identifying the abnormal stage, and marking the user's abnormal heart rate situation prediction data; Step S6: Based on the personalized user standard map, the user's abnormal heart rate trend prediction data is personalized evaluated for health status, and heart rate control decisions are made to build a health warning decision engine to execute the user's health monitoring and warning operations.

[0006] The present invention constructs a personalized user standard map to better understand the user's heart rate data, providing a personalized basis for subsequent analysis, optimizing historical heart rate monitoring data through interpolation filtering, filling in data gaps, improving data integrity and accuracy, mining personalized user features to better understand the user's heart rate data change rules, providing a basis for subsequent decision-making, constructing a real-time heart rate waveform curve to monitor the user's heart rate changes in real time, and promptly discover abnormal situations, multi-point interval sampling to improve data collection efficiency while maintaining data accuracy, analyzing heart rate variability to provide a deeper understanding of heart rate changes, helping to evaluate the user's heart rate stability, calculating the dynamic variation amplitude to capture the trend and amplitude of heart rate fluctuations, providing a basis for subsequent predictions, and generating multi-period heart rate waveforms. The heart rate trend evolution prediction data helps predict the user's heart rate change trend and prepare for health intervention in advance. The multi-period trend evolution prediction data comprehensively understands the development trend of the user's heart rate and provides more comprehensive data support for health warnings. The staged activity pattern inference of the user's body posture data helps understand the user's daily activities and provides contextual support for abnormality identification. Abnormal heart rate trend prediction data is marked to promptly detect user abnormalities and improve the accuracy and timeliness of health warnings. Personalized health status assessment based on personalized user standard maps more accurately assesses the user's health status and provides a basis for health regulation. A health warning decision engine is built to automate health monitoring and warning operations to improve the accuracy and efficiency of warnings.

[0007] Preferably, step S1 comprises the following steps: Step S11: Acquire real-time user heart rate monitoring parameters and historical heart rate monitoring data based on the smart wearable device; Step S12: performing interpolation filtering and optimization on the historical heart rate monitoring data to obtain historical optimized heart rate monitoring data; Step S13: performing time granularity segmentation on the historical optimized heart rate monitoring data to obtain multiple time-series heart rate monitoring sequences; Step S14: performing user periodic behavior analysis on multiple time-series heart rate monitoring sequences to obtain user behavior characteristics of each sequence; Step S15: performing periodic change analysis on multiple time-series heart rate monitoring sequences to generate a user periodic change pattern; Step S16: Based on the user cycle change rules, personalized user feature mining is performed on the user behavior characteristics of each sequence to construct a personalized user standard map.

[0008] The present invention provides a data basis for subsequent analysis by acquiring real-time user heart rate monitoring parameters and historical heart rate monitoring data, realizes comprehensive heart rate monitoring, realizes continuous monitoring and tracking of user heart rate data by acquiring real-time and historical data, discovers potential health problems, optimizes historical heart rate monitoring data by interpolation filtering, improves data integrity and accuracy, provides a more reliable basis for subsequent analysis, divides historical optimized heart rate monitoring data into time granularity, observes the change trend of user heart rate data more carefully, and provides data support with more dimensions for subsequent analysis, reveals the heart rate behavior characteristics of users in different time periods through user periodic behavior analysis, helps understand the physiological activity law of users, analyzes the user behavior characteristics of each sequence, and conducts personalized user feature mining to better understand the meaning behind the heart rate data of each user, generates user periodic change laws to reveal the periodic change trend in user heart rate data, and helps predict future heart rate change trends, and better identifies user behavior laws through periodic change analysis, provides more in-depth data support for health warning, constructs personalized user standard maps to more comprehensively display the heart rate data characteristics of each user, provides personalized basis for health status assessment and health warning, and conducts personalized user feature mining based on user periodic change laws to realize personalized health management, and improves the accuracy and effectiveness of warning.

[0009] Preferably, the specific steps of step S12 are: Perform missing value detection on historical heart rate monitoring data to identify missing values; Perform frequency distribution statistics on missing values ​​to obtain missing value frequency distribution data; Based on the missing value frequency distribution data, the surrounding data change analysis of the missing values ​​is performed to extract the surrounding data point change data of each missing value; Calculate the mean of the change data of the surrounding data points of each missing value to obtain the mean of the surrounding data; Perform linear interpolation filling on historical heart rate monitoring data based on the mean value of surrounding data to generate interpolated and filled heart rate monitoring data; Perform outlier identification on the interpolated heart rate monitoring data and mark abnormal data points; Outlier filtering is performed based on abnormal data points to obtain historical optimized heart rate monitoring data.

[0010] The present invention ensures the integrity and accuracy of data by detecting and identifying missing values ​​in historical heart rate monitoring data, providing a reliable data basis for subsequent analysis, analyzing the patterns and trends of missing values ​​through frequency distribution statistics of missing values, helping to understand the causes and impact of data missing, analyzing the changes in data surrounding missing values ​​to understand the correlation between missing values ​​and surrounding data points, providing a basis for subsequent interpolation and filling, more accurately estimating the values ​​of missing values ​​by calculating the mean of surrounding data, improving the accuracy and reliability of filled data, filling missing values ​​through linear interpolation, restoring the integrity of historical heart rate monitoring data, making the data set more complete, and providing complete data support for subsequent analysis, marking abnormal data points to identify potential data quality problems or abnormal situations, and providing a more accurate data set for data cleaning and analysis, and improving the quality of historical heart rate monitoring data through outlier filtering, removing the interference of abnormal data points, making the data more reliable and accurate, and providing a more reliable basis for subsequent health warning methods.

[0011] Preferably, step S2 specifically comprises the following steps: Step S21: defining a time interval, performing multi-point interval sampling on the real-time user heart rate monitoring parameters based on the time interval, and extracting multiple heart rate monitoring sampling points; Step S22: Calculating the time coordinates of multiple heart rate sampling points to extract the time coordinates of each sampling point; Step S23: performing heart rate fluctuation characteristic analysis on a plurality of heart rate sampling points to generate a heart rate fluctuation characteristic; Step S24: performing heart rate waveform fitting on the heart rate fluctuation characteristics based on the time coordinate of each sampling point to construct a real-time heart rate waveform curve.

[0012] The present invention samples real-time user heart rate monitoring parameters at multi-point intervals to improve the richness and density of data, more comprehensively reflect the user's heart rate changes, define appropriate time intervals for multi-point sampling to improve time resolution, capture subtle changes in heart rate fluctuations, extract the time coordinates of each sampling point to accurately record the time point of each heart rate sampling, and provide time information support for subsequent time series analysis. By performing fluctuation feature analysis on multiple heart rate sampling points, key features of heart rate fluctuations, such as amplitude and frequency, are extracted to help understand the law of heart rate changes. Heart rate fluctuation feature analysis evaluates the user's heart rate changes, discovers abnormal fluctuations, and provides a basis for health warnings. Real-time heart rate waveform curves are constructed through heart rate waveform fitting to better simulate and understand the trends and laws of heart rate changes. The construction of real-time heart rate waveform curves intuitively displays the trend of heart rate changes, and provides users with intuitive health monitoring information.

[0013] Preferably, step S3 specifically comprises the following steps: Step S31: Calculating the global heart rate variability of the real-time heart rate waveform curve to generate a global heart rate average variability; Step S32: Divide the real-time heart rate waveform curve into stages to obtain multiple stage waveform curves; Step S33: performing heart rate variability analysis on the waveform curves of multiple stages based on the global average heart rate variation to generate heart rate variability between stages; Step S34: Calculating the dynamic variation amplitude of the heart rate variability between stages and identifying the dynamic heart rate variability amplitude; Step S35: mining the time series changes of the dynamic heart rate variability amplitude to obtain a representation of the user's heart rate time series changes.

[0014] The present invention calculates the global heart rate variability to understand the average change trend of the overall heart rate, provides a general understanding of the overall heart rate state, generates a global heart rate average variability that can be used as a benchmark for subsequent comparative analysis and health status assessment, divides the real-time heart rate waveform curve into stages to refine the analysis of heart rate changes, discovers the heart rate characteristics and change laws of different stages, and the division of multiple stage waveform curves provides a basis for realizing dynamic monitoring of heart rate changes in different time periods. The heart rate variability analysis evaluates the change amplitude and law of heart rate in different stages, helps to discover the characteristics of heart rate variability, generates heart rate variability between stages to compare the difference in heart rate changes in different stages, and provides a basis for dynamic monitoring of health status. The dynamic heart rate variability amplitude is identified to discover abnormal heart rate fluctuations, indicating potential health problems or disease risks. The calculation of the dynamic variability amplitude can be used to adjust the monitoring strategy in real time or remind the user to pay attention to abnormal heart rate fluctuations. By mining the time series changes of the dynamic heart rate variability amplitude, the time series characteristics of the user's heart rate are extracted, and the law and trend of heart rate changes are understood. The time series changes represent the changes in the heart rate state of the personalized monitoring user, and provide the user with health warnings and suggestions that are closer to individual needs.

[0015] Preferably, the specific steps of step S4 are: Step S41: performing multi-scale time-frequency decomposition on the representation of the user's heart rate time series variation to obtain a heart rate variation frequency domain of multiple frequencies; Step S42: performing frequency component distribution analysis on the heart rate variation frequency domain of multiple frequencies to generate frequency component distribution data in the frequency domain; Step S43: performing heart rate change trend analysis based on frequency domain frequency component distribution data to extract heart rate change trend features; Step S44: Predict the evolution of the heart rate trend characteristics over multiple time periods to generate prediction data for the evolution of the heart rate trend over multiple time periods.

[0016] The present invention converts the representation of the user's heart rate time series change into the frequency domain through multi-scale time-frequency decomposition, obtains the heart rate change characteristics at different frequencies, obtains the heart rate change frequency domains of multiple frequencies to deeply understand the frequency components of the heart rate change, and provides a basis for subsequent analysis. The distribution of heart rate changes at different frequencies is evaluated through frequency component distribution analysis, and the frequency characteristics of heart rate changes are revealed. Frequency domain frequency component distribution data is generated to extract the frequency domain characteristics of heart rate changes, and data support is provided for subsequent trend analysis. Based on the frequency domain frequency component distribution data, a heart rate change trend analysis is performed to extract the trend characteristics of heart rate changes, and the law of heart rate changes is revealed. The heart rate change trend is analyzed to help evaluate the heart rate change trend and discover potential abnormalities or health risks. The heart rate change trend characteristics are predicted for multiple time periods to predict the future heart rate change trend, which helps users understand the development trend of their heart rate. Multi-time period heart rate situation evolution prediction data is generated to discover the trend of heart rate changes in advance, and provide health warnings and suggestions for users.

[0017] Preferably, the specific steps of step S5 are: Step S51: performing jitter frequency statistics on the user's body posture data to obtain the user's jitter frequency parameter; Step S52: performing deep learning classification on the user's jitter frequency parameters to generate a user activity pattern; Step S53: inferring the user's activity pattern in stages based on the real-time heart rate waveform curve to obtain the user's activity pattern in each stage; Step S54: Based on the user activity pattern at each stage, the abnormal stage of the multi-period heart rate situation evolution prediction data is identified, and the abnormal heart rate situation prediction data of the user is marked.

[0018] The present invention obtains the user's jitter frequency parameters by performing jitter frequency statistics on the user's body posture data, reflecting the user's body movement and activity status. The jitter frequency parameters are a representation of the user's activity characteristics, and subsequent activity pattern classification and heart rate situation analysis. The user's jitter frequency parameters are subjected to deep learning classification to generate the user's activity pattern, which helps to understand the user's different activity states and patterns. Deep learning classification generates activity patterns according to the user's jitter frequency parameters in a personalized manner, thereby improving the accuracy of activity pattern recognition. Based on the real-time heart rate waveform curve, the user's activity pattern is inferred to identify the user's activity state at different stages, and the user behavior is more comprehensively understood in combination with the heart rate data. By inferring the user's activity pattern at each stage, the user's activity behavior in different time periods is more accurately analyzed. Based on the user's activity pattern at each stage, the multi-period heart rate situation evolution prediction data is identified for abnormal stages, and the user's abnormal heart rate situation prediction data is marked. The user data with health problems or abnormal states is identified and marked for abnormal stages, thereby improving the accuracy and timeliness of health warnings.

[0019] Preferably, the specific steps of step S6 are: Step S61: performing personalized health status assessment on the abnormal heart rate situation prediction data of the user based on the personalized user standard map to generate an abnormal heart rate health status assessment value; Step S62: comparing the abnormal heart rate health status assessment value based on a preset standard heart rate health assessment threshold, and generating a user health warning signal when the preset standard heart rate health assessment threshold is higher than the abnormal heart rate health status assessment value; Step S63: Make heart rate control decisions based on the user's health warning signal and build a health warning decision engine to execute the user's health monitoring and warning operations.

[0020] The present invention evaluates abnormal heart rate trend prediction data based on personalized user standard maps, generates personalized health status assessment values ​​according to the characteristics and historical data of each user, generates abnormal heart rate health status assessment values ​​to understand the user's current health status, and provides personalized health assessment basis, and determines whether the user's current health status is abnormal by comparing the abnormal heart rate health status assessment value with a preset standard heart rate health assessment threshold. When the preset health assessment threshold is higher than the abnormal heart rate health status assessment value, a user health warning signal is generated to remind the user of the existing health risks, and a health warning decision engine is constructed based on the user's health warning signal to realize automated health monitoring and warning operations, thereby improving response speed and accuracy, and executing the health warning decision engine to make regulation decisions on the user's heart rate according to the warning signal, so as to help the user adjust activities or seek medical treatment in time and ensure health status.

[0021] In this specification, a health monitoring and early warning system for a smart wearable device is provided, which is used to execute the health monitoring and early warning method for the smart wearable device as described above, including: Personalized feature module, used to obtain real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on smart wearable devices; interpolation filtering and optimization of historical heart rate monitoring data, and personalized user feature mining to build a personalized user standard map; The heart rate waveform module is used to perform multi-point interval sampling of real-time user heart rate monitoring parameters, perform heart rate waveform fitting, and construct a real-time heart rate waveform curve; The heart rate time series variation module is used to analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series variation; The heart rate situation prediction module is used to analyze the heart rate change trend of the user's heart rate time series change representation, and then predict the situation evolution in multiple time periods to generate multi-time period heart rate situation evolution prediction data; The abnormal heart rate recognition module is used to infer the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, and to identify the abnormal stage and mark the user's abnormal heart rate situation prediction data; The health warning module is used to conduct personalized health status assessment of the user's abnormal heart rate trend prediction data based on the personalized user standard map, make heart rate control decisions, and build a health warning decision engine to execute the user's health monitoring and warning operations.

[0022] The present invention obtains real-time user heart rate monitoring parameters and historical data, as well as user body posture data to fully understand the user's physiological condition, interpolates and filters to optimize historical heart rate monitoring data, and performs personalized user feature mining to establish a personalized user standard map, providing a basis for subsequent analysis. Through multi-point interval sampling and heart rate waveform fitting, a real-time heart rate waveform curve is constructed, providing real-time heart rate data visualization, allowing users to intuitively understand their heart rate change trends and fluctuations, analyzing heart rate variability and calculating dynamic variation amplitude to deeply understand the user's heart rate time series change characteristics. These representations can provide important references for subsequent situation prediction and health assessment. By analyzing the trend of heart rate changes and making multi-period situation evolution predictions, heart rate change patterns can be discovered in advance, so that users can understand future heart rate changes and make corresponding adjustments and preventive measures. Abnormal heart rate identification and activity pattern inference are performed based on multi-period heart rate situation evolution prediction data, and the user's abnormal heart rate is marked and identified in time. Users pay attention to health abnormalities and take necessary measures in time. Health status assessment is performed on abnormal heart rate situation prediction data based on personalized user standard maps to provide users with personalized health warnings. The constructed health warning decision engine allows users to make heart rate control decisions based on their personal circumstances, thereby improving health management and monitoring effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of the steps of a health monitoring and early warning method for a smart wearable device according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0024] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0025] The example of this application provides a method and system. The execution subjects of the method and system include but are not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of this application, and the data processing platform includes but is not limited to: at least one of audio and image management system, information management system, and cloud data management system.

[0026] See also Figures 1 to 4 The present invention provides a health monitoring and early warning method for a smart wearable device, the health monitoring and early warning method for a smart wearable device comprises the following steps: Step S1: obtaining real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on the smart wearable device; interpolating and filtering the historical heart rate monitoring data, and mining personalized user features to construct a personalized user standard map; Step S2: performing multi-point interval sampling on the real-time user heart rate monitoring parameters, and performing heart rate waveform fitting to construct a real-time heart rate waveform curve; Step S3: Analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series change; Step S4: analyzing the heart rate change trend of the user's heart rate time series change representation, and then predicting the multi-period situation evolution to generate multi-period heart rate situation evolution prediction data; Step S5: inferring the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, identifying the abnormal stage, and marking the user's abnormal heart rate situation prediction data; Step S6: Based on the personalized user standard map, the user's abnormal heart rate trend prediction data is personalized evaluated for health status, and heart rate control decisions are made to build a health warning decision engine to execute the user's health monitoring and warning operations.

[0027] The present invention constructs a personalized user standard map to better understand the user's heart rate data, providing a personalized basis for subsequent analysis, optimizing historical heart rate monitoring data through interpolation filtering, filling in data gaps, improving data integrity and accuracy, mining personalized user features to better understand the user's heart rate data change rules, providing a basis for subsequent decision-making, constructing a real-time heart rate waveform curve to monitor the user's heart rate changes in real time, and promptly discover abnormal situations, multi-point interval sampling to improve data collection efficiency while maintaining data accuracy, analyzing heart rate variability to provide a deeper understanding of heart rate changes, helping to evaluate the user's heart rate stability, calculating the dynamic variation amplitude to capture the trend and amplitude of heart rate fluctuations, providing a basis for subsequent predictions, and generating multi-period heart rate waveforms. The heart rate trend evolution prediction data helps predict the user's heart rate change trend and prepare for health intervention in advance. The multi-period trend evolution prediction data comprehensively understands the development trend of the user's heart rate and provides more comprehensive data support for health warnings. The staged activity pattern inference of the user's body posture data helps understand the user's daily activities and provides contextual support for abnormality identification. Abnormal heart rate trend prediction data is marked to promptly detect user abnormalities and improve the accuracy and timeliness of health warnings. Personalized health status assessment based on personalized user standard maps more accurately assesses the user's health status and provides a basis for health regulation. A health warning decision engine is built to automate health monitoring and warning operations to improve the accuracy and efficiency of warnings.

[0028] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a health monitoring and early warning method for a smart wearable device of the present invention. In this example, the steps of the health monitoring and early warning method for a smart wearable device include: Step S1: obtaining real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on the smart wearable device; interpolating and filtering the historical heart rate monitoring data, and mining personalized user features to construct a personalized user standard map; In this embodiment, it is ensured that the smart wearable device (such as a smart watch or a heart rate monitor) is connected to the data receiving end (such as a mobile phone application or a cloud server), usually through Bluetooth or Wi-Fi, and real-time heart rate monitoring parameters (such as current heart rate, heart rate variability, etc.) are obtained from the device at regular intervals. The data is continuously obtained by using a scheduled call method (such as setting a scheduled task or using an event-driven model). The historical heart rate monitoring data is extracted from the smart wearable device or related applications to ensure the integrity and accuracy of the data. The data includes timestamps, heart rate values, and other related parameters. The user's body posture data (such as accelerometer and gyroscope data) is obtained and directly collected through sensors in the device to record the user's motion status and posture changes. The missing values ​​in the historical heart rate monitoring data are detected using data analysis tools (such as the Pandas library), the locations of the missing values ​​are recorded, and an appropriate interpolation method (such as linear regression) is selected. Use interpolation, spline interpolation, or Lagrange interpolation to fill in the detected missing values, use the interpolate() function in Python, apply smoothing techniques (such as moving average or exponential smoothing) to the interpolated data to reduce the impact of noise and ensure the quality of the data, store the optimized historical heart rate data, and prepare for subsequent analysis and personalized feature mining. Extract key features from the optimized historical heart rate data, including average heart rate, heart rate fluctuation, maximum and minimum values, heart rate variability, etc., build a personalized user feature model based on the user's historical data and real-time monitoring parameters, use clustering algorithms (such as K-means clustering) or classification algorithms (such as decision trees) to identify the user's feature patterns, and build a personalized user standard map based on the extracted user features. This map should include the user's heart rate characteristics and body posture characteristics under different states (such as rest, exercise, stress, etc.).

[0029] Step S2: performing multi-point interval sampling on the real-time user heart rate monitoring parameters, and performing heart rate waveform fitting to construct a real-time heart rate waveform curve; In this embodiment, the user's heart rate monitoring parameters are obtained from the smart wearable device in real time, the data update frequency is ensured (for example, sampling once per second), and the time window of multi-point interval sampling is determined, for example, sampling is performed every 5 seconds or every 10 seconds to reduce the amount of data and retain key information. The heart rate data is recorded within a preset time interval. For example, if sampling is performed every 10 seconds, the data collected within 10 seconds will be integrated into one data point, and a suitable heart rate waveform model is selected. Commonly used models include sine waves, cosine waves or Gaussian functions. These models can effectively fit the periodic changes of the heart rate waveform. The sampled heart rate data is preprocessed, including removing outliers and smoothing (such as using a sliding average method) to improve the accuracy of the fitting. An optimization algorithm such as a nonlinear least squares method (such as a Levenberg-Marquardt algorithm) is used to estimate the parameters of the selected waveform model, a loss function is defined, the error between the actual data and the model data is calculated, the model parameters are iteratively adjusted, the loss function is minimized, the processed heart rate data is matched with the fitting model, and a real-time heart rate waveform curve is generated to ensure that the fitted waveform can accurately reflect the changing trend of the heart rate data.

[0030] Step S3: Analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series change; In this embodiment, the heart rate waveform curve data is organized into a time series format to ensure that each time point has a corresponding heart rate value, so as to facilitate the calculation of variability. A suitable heart rate variability calculation method is selected. Commonly used methods include standard deviation, root mean square (RMS) and heart rate variability (HRV) indicators (such as SDNN and RMSSD). The standard deviation (SD) of the heart rate data is calculated as a basic indicator of heart rate variability. The dynamic variation amplitude represents the fluctuation amplitude of the heart rate in a certain time window, which is usually achieved by calculating the standard deviation or change rate in each time window. The calculation of the dynamic variation amplitude is determined. Time windows, such as every 5 seconds or every 10 seconds, ensure that the dynamic characteristics of heart rate fluctuations can be reflected. The standard deviation of heart rate data is calculated in each time window, and the variation amplitude of each window is recorded. The data of heart rate variability and dynamic variation amplitude are integrated to form a comprehensive representation of the user's heart rate time series changes, including variability indicators and time series. The representation of the user's heart rate time series changes is stored in the database for subsequent analysis. Data visualization tools (such as Matplotlib, Plotly, etc.) are used to display the change curves of heart rate variability and dynamic variation amplitude to help users intuitively understand the dynamic changes of heart rate Step S5: inferring the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, identifying the abnormal stage, and marking the user's abnormal heart rate situation prediction data; In this embodiment, the heart rate situation evolution prediction data of multiple time periods are ensured to be complete, including timestamps, predicted heart rate values ​​and related confidence intervals. The heart rate prediction data are organized into a structured format (such as Pandas DataFrame) to facilitate subsequent analysis and processing. The user's body posture data (such as acceleration, gyroscope data) and the heart rate situation evolution prediction data are time-aligned to ensure comparison within the same time period. The body posture data are divided into multiple stages according to the time period or the changes in user activities. The sliding window method is used to ensure that each stage contains enough data points for analysis. Features are extracted from each stage, such as average acceleration, movement frequency and posture changes, and recorded in a data frame for subsequent analysis. A suitable activity pattern recognition method (such as a rule-based classification method, a machine learning model or a deep learning model) is selected to infer the user's activity pattern. If a machine learning or deep learning model is used, the model must first be trained with labeled activity data. Commonly used models include support vector machines (SVMs), Random forest or convolutional neural network (CNN) inputs the integrated body posture data into the trained model to infer the staged activity pattern, identify the activity type of each stage (such as stillness, walking, running, etc.), select appropriate anomaly detection methods (such as threshold-based detection, Z-score method or machine learning-based anomaly detection model) to identify abnormal stages, and set reasonable abnormal thresholds based on historical data and expert advice. For example, if the heart rate change exceeds a certain range or the activity pattern does not match expectations, the heart rate trend and activity pattern of each stage are compared. If it is found that the heart rate prediction value does not match the expected activity pattern (such as the heart rate is too high in the static state), it is marked as an abnormal stage, and the identified abnormal heart rate trend prediction data is marked, and the abnormal type, timestamp and heart rate value are recorded for subsequent analysis and processing.

[0031] Step S6: Based on the personalized user standard map, the user's abnormal heart rate trend prediction data is personalized evaluated for health status, and heart rate control decisions are made to build a health warning decision engine to execute the user's health monitoring and warning operations.

[0032] In this embodiment, according to the personalized user standard map, an appropriate health status assessment method (such as standardized scoring, risk assessment model, etc.) is selected to evaluate the abnormal heart rate situation, and the assessment results are stored in the database, including the assessment value, timestamp and corresponding abnormal heart rate data. The threshold of heart rate control is set according to the personalized user standard map, for example, the upper and lower limits of abnormal heart rate, and the heart rate control decision rules are formulated to clarify the control measures corresponding to different evaluation values ​​(such as increasing exercise, reducing stress, and adjusting diet). A suitable decision model (such as a rule-based model, a decision tree or a machine learning model) is selected to process the assessment results and generate specific health control suggestions, and the architecture of the health warning decision engine is designed to ensure that it can be received in real time. User health data, assessment results and abnormal marks, integrate the user's real-time heart rate monitoring data, abnormal assessment values ​​and regulatory decisions into the decision engine to ensure smooth data flow, use programming languages ​​​​(such as Python or Java) to implement the health warning decision engine, ensure that it can process data in real time and output alarms and suggestions, and generate health warning signals (such as text messages, application notifications or emails) based on assessment results and regulatory decisions to remind users to pay attention to abnormal heart rate conditions. Establish a user feedback mechanism so that users can provide feedback on the effectiveness of warning signals and the feasibility of adjustment suggestions, so as to continuously optimize the decision engine, continuously monitor the user's health status and heart rate changes, and dynamically adjust decision rules and models according to new health data and user feedback.

[0033] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Step S11: Acquire real-time user heart rate monitoring parameters and historical heart rate monitoring data based on the smart wearable device; Step S12: performing interpolation filtering and optimization on the historical heart rate monitoring data to obtain historical optimized heart rate monitoring data; Step S13: performing time granularity segmentation on the historical optimized heart rate monitoring data to obtain multiple time-series heart rate monitoring sequences; Step S14: performing user periodic behavior analysis on multiple time-series heart rate monitoring sequences to obtain user behavior characteristics of each sequence; Step S15: performing periodic change analysis on multiple time-series heart rate monitoring sequences to generate a user periodic change pattern; Step S16: Based on the user cycle change rules, personalized user feature mining is performed on the user behavior characteristics of each sequence to construct a personalized user standard map.

[0034] In this embodiment, it is ensured that the smart wearable device is connected to the data receiving end (such as a mobile phone or a cloud server), and data transmission is performed using protocols such as Bluetooth or Wi-Fi. The user's heart rate monitoring parameters (such as current heart rate, heart rate fluctuation, etc.) and historical heart rate monitoring data (such as heart rate records of the past few days or weeks) are obtained from the device in real time. A suitable interpolation method (such as linear interpolation, spline interpolation, or Lagrange interpolation) is selected to process missing values ​​in the historical heart rate monitoring data. The historical heart rate monitoring data is traversed, missing values ​​are identified, and the missing values ​​are filled using the selected interpolation method to ensure data continuity and integrity. Data smoothing techniques (such as moving average or exponential smoothing) are applied to optimize the interpolated data to reduce the impact of noise and obtain more accurate historical optimized heart rate monitoring data. The time granularity (such as every minute, every hour, or every day) is set according to the needs, and it is decided how to segment the historical optimized heart rate monitoring data. The optimized heart rate monitoring data is segmented according to the set time granularity to generate multiple time series heart rate monitoring sequences to ensure that each sequence contains heart rate data within a specific time period. A suitable period is selected Behavioral analysis models (such as Fourier transform and periodic regression analysis) are used to analyze the time-series heart rate monitoring sequence. Periodic behavior analysis is performed on each time-series heart rate monitoring sequence to extract the user's behavioral characteristics (such as activity level, heart rate change pattern, etc.). The key features of each sequence are recorded, such as the highest heart rate, the lowest heart rate, the average heart rate and the heart rate fluctuation amplitude. Time series analysis techniques (such as autocorrelation analysis and periodogram) are used to identify the periodic changes in the heart rate monitoring data. Each time-series heart rate monitoring sequence is analyzed to identify regular change patterns, such as the heart rate difference between daytime and nighttime, the recovery period after exercise, etc. The analysis results are recorded in a report, outlining the periodic change law of each sequence. According to the periodic change law of the user, a personalized feature mining model is designed to ensure that the user's unique behavioral characteristics can be extracted. Machine learning algorithms (such as cluster analysis, decision tree, support vector machine, etc.) are used to analyze the user behavior characteristics of each sequence to mine personalized user characteristics. The mined personalized user characteristics are integrated to construct a personalized user standard map, which includes the heart rate change patterns and behavioral characteristics of different users.

[0035] In this embodiment, the specific steps of step S12 are: Perform missing value detection on historical heart rate monitoring data to identify missing values; Perform frequency distribution statistics on missing values ​​to obtain missing value frequency distribution data; Based on the missing value frequency distribution data, the surrounding data change analysis of the missing values ​​is performed to extract the surrounding data point change data of each missing value; Calculate the mean of the change data of the surrounding data points of each missing value to obtain the mean of the surrounding data; Perform linear interpolation filling on historical heart rate monitoring data based on the mean value of surrounding data to generate interpolated and filled heart rate monitoring data; Perform outlier identification on the interpolated heart rate monitoring data and mark abnormal data points; Outlier filtering is performed based on abnormal data points to obtain historical optimized heart rate monitoring data.

[0036] In this embodiment, data is read from a historical heart rate monitoring data set to ensure that the data format is correct (such as a time series format), and a data analysis tool (such as a Pandas library) is used to detect missing values. The location of the missing value is identified through function detection (such as isnull()). For each missing value, data points within a certain range before and after it are extracted (such as 5 data points before and after each), and the changes between the extracted surrounding data points are calculated. For example, the change amplitude of each data point is calculated (such as the relative change percentage), and the surrounding data changes corresponding to each missing value are recorded in a list to facilitate subsequent mean calculations. For each missing value, the mean of the surrounding data points is calculated and calculated using the mean() function. The surrounding mean of each missing value is stored in a dictionary or array for subsequent interpolation. The linear interpolation method is used. Missing values ​​are filled with the mean of their surrounding data using an interpolation function (such as Pandas' interpolate()). The calculated mean replaces the original missing values ​​to generate interpolated and filled heart rate monitoring data. An appropriate outlier detection method (such as the Z-score or IQR method) is selected to identify outliers in the interpolated and filled data. The Z-score of each data point is calculated or the quartile method is used for outlier detection. Abnormal data points are marked according to the set threshold (such as a Z-score greater than 3 or less than -3). The identified abnormal data points are filtered and removed. The conditional screening method is used to retain normal data in the new data set. The historical optimized heart rate monitoring data after outlier filtering is stored as a new data set for subsequent analysis and use.

[0037] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Step S21: defining a time interval, performing multi-point interval sampling on the real-time user heart rate monitoring parameters based on the time interval, and extracting multiple heart rate monitoring sampling points; Step S22: Calculating the time coordinates of multiple heart rate sampling points to extract the time coordinates of each sampling point; Step S23: performing heart rate fluctuation characteristic analysis on a plurality of heart rate sampling points to generate a heart rate fluctuation characteristic; Step S24: performing heart rate waveform fitting on the heart rate fluctuation characteristics based on the time coordinate of each sampling point to construct a real-time heart rate waveform curve.

[0038] In this embodiment, a suitable time interval (such as every 5 seconds or every 10 seconds) is defined as the sampling frequency to ensure that the heart rate changes can be effectively captured. Within the set time interval, the heart rate data is extracted from the real-time user heart rate monitoring system. A timer or scheduler (such as time.sleep() in Python) is used to control the sampling. At each sampling, the current timestamp is recorded, and the system time function (such as datetime.now() in Python) is used to obtain the precise time. Each heart rate sampling point is paired with its corresponding timestamp to generate a time coordinate, and a dictionary or structure is created, which contains each sampling point and its time coordinate, and the heart rate data of the sampling point is analyzed. Fluctuation feature analysis: identify the change pattern of heart rate, use statistical methods (such as standard deviation, coefficient of variation) to quantify fluctuations, calculate the heart rate fluctuation characteristics of each sampling point, including maximum, minimum, average and fluctuation amplitude, etc., and store these characteristics in an array or data frame, select a suitable fitting model (such as polynomial fitting, spline fitting or other curve fitting methods) to construct a heart rate waveform curve, use time coordinates and corresponding heart rate data points to perform curve fitting, use the fitting function in the data analysis library (such as NumPy or SciPy) to generate a heart rate waveform curve, and visualize (such as using the Matplotlib library) to show the trend of heart rate changes for easy observation and analysis.

[0039] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Step S31: Calculating the global heart rate variability of the real-time heart rate waveform curve to generate a global heart rate average variability; Step S32: Divide the real-time heart rate waveform curve into stages to obtain multiple stage waveform curves; Step S33: performing heart rate variability analysis on the waveform curves of multiple stages based on the global average heart rate variation to generate heart rate variability between stages; Step S34: Calculating the dynamic variation amplitude of the heart rate variability between stages and identifying the dynamic heart rate variability amplitude; Step S35: mining the time series changes of the dynamic heart rate variability amplitude to obtain a representation of the user's heart rate time series changes.

[0040] In this embodiment, the heart rate data points are extracted from the real-time heart rate waveform curve to ensure the integrity and accuracy of the data, and the global heart rate variability is calculated. The standard deviation or coefficient of variation is usually used to reflect the volatility of the heart rate. The global variability = , σ is the standard deviation of the heart rate, μ is the average value of the heart rate, the calculated global average heart rate change is stored for subsequent analysis, the stage division standard is set according to the time period (such as every hour, every day) or the heart rate change situation, the heart rate waveform curve is divided into stages according to the set standard, and the timestamp or heart rate change threshold is used as the basis for division. The heart rate variability of the waveform curve of each stage is analyzed, and the same method as the global variability (such as standard deviation or coefficient of variation) is used for calculation. The heart rate variability of each stage is recorded in a data frame for subsequent comparison and analysis. The change of heart rate variability between stages is calculated to identify the dynamic variation amplitude. By calculating the difference of the variation of adjacent stages, the calculated dynamic heart rate variability amplitude is stored in an array, and an appropriate time series analysis method (such as time series analysis, trend analysis) is selected to mine the change pattern of the dynamic variation amplitude. The dynamic variation amplitude is mined for time series changes to identify the periodic changes and trends of the heart rate. The mining results are recorded in a data frame, and a visual chart (such as a line chart) is generated to show the time series change representation of the user's heart rate.

[0041] In this embodiment, step S4 includes the following steps: Step S41: performing multi-scale time-frequency decomposition on the representation of the user's heart rate time series variation to obtain a heart rate variation frequency domain of multiple frequencies; Step S42: performing frequency component distribution analysis on the heart rate variation frequency domain of multiple frequencies to generate frequency component distribution data in the frequency domain; Step S43: performing heart rate change trend analysis based on frequency domain frequency component distribution data to extract heart rate change trend features; Step S44: Predict the evolution of the heart rate trend characteristics over multiple time periods to generate prediction data for the evolution of the heart rate trend over multiple time periods.

[0042] In this embodiment, a suitable time-frequency decomposition method (such as wavelet transform, Fourier transform, etc.) is selected to analyze the time-series changes of the user's heart rate, and the user's heart rate time-series change data is organized into a format suitable for time-frequency analysis. The selected time-frequency decomposition algorithm is applied to decompose the heart rate time-series data into multiple frequency components to generate frequency domain data of heart rate changes. The wavelet transform provides local information of time and frequency, making the analysis more detailed. Each frequency component is extracted from the result of the time-frequency decomposition, and its distribution over time is analyzed. The amplitude and phase of each frequency component are calculated, and its changes in different time periods are analyzed. The amplitude and phase information of the frequency components are integrated to generate frequency domain frequency component distribution data, and the changes of each frequency component over time are recorded. Based on the situation, select a suitable trend analysis method (such as sliding average, linear regression or time series prediction) to extract the trend characteristics of heart rate changes, use frequency domain frequency component distribution data for processing, identify the long-term trend and short-term fluctuation of heart rate, calculate the key characteristics of heart rate changes (such as trend direction, change rate), and store these characteristics in the data frame, select a suitable prediction model (such as ARIMA, LSTM or support vector regression) for multi-period prediction according to the data characteristics, use historical heart rate data and the extracted change trend characteristics to train the prediction model, and perform cross-validation to ensure the accuracy of the model, based on the trained model, predict the heart rate changes in multiple time periods in the future, and generate multi-period heart rate situation evolution prediction data.

[0043] In this embodiment, the specific steps of step S5 are: Step S51: performing jitter frequency statistics on the user's body posture data to obtain the user's jitter frequency parameter; Step S52: performing deep learning classification on the user's jitter frequency parameters to generate a user activity pattern; Step S53: inferring the user's activity pattern in stages based on the real-time heart rate waveform curve to obtain the user's activity pattern in each stage; Step S54: Based on the user activity pattern at each stage, the abnormal stage of the multi-period heart rate situation evolution prediction data is identified, and the abnormal heart rate situation prediction data of the user is marked.

[0044] In this embodiment, the user's body posture data is collected, usually from an accelerometer or gyroscope sensor. The data should include a timestamp and acceleration values ​​of each axis. The collected posture data is analyzed to extract the characteristics of the jitter. The fast Fourier transform (FFT) is used to calculate the spectrum of the signal, identify the frequency components, determine the jitter frequency and count the frequency parameters of its occurrence, organize the jitter frequency parameters into a format suitable for deep learning model input, and usually require standardization. A suitable deep learning model (such as a convolutional neural network CNN or a recurrent neural network RNN) is selected for classification. According to the data characteristics, the model selection should be able to effectively capture the time series characteristics. The model is trained using labeled activity pattern data, and the model parameters are optimized to improve the classification accuracy. The user's jitter frequency parameters are input into the trained model to generate the user's activity pattern. The real-time heart rate waveform curve is compared with the Combined with user activity patterns to ensure time alignment, heart rate waveform data and activity pattern data are divided into multiple stages according to timestamps, which are usually related to changes in user activities. The heart rate waveform curve of each stage is analyzed, and the specific activity type of each stage (such as stillness, walking, running, etc.) is inferred based on the user's activity pattern. The user activity pattern of each stage is stored to form a list of staged activity patterns. The staged user activity pattern is combined with the multi-period heart rate trend evolution prediction data to ensure comparison and analysis, and a suitable anomaly detection method (such as threshold-based method, Z-score detection or machine learning model) is selected to identify abnormal heart rate trends. According to the staged activity pattern, the heart rate change data that does not meet the expectations is marked, the abnormal stage is identified, and the identified abnormal heart rate trend prediction data is recorded for subsequent analysis and processing to generate an abnormal situation report.

[0045] In this embodiment, the specific steps of step S6 are: Step S61: performing personalized health status assessment on the abnormal heart rate situation prediction data of the user based on the personalized user standard map to generate an abnormal heart rate health status assessment value; Step S62: comparing the abnormal heart rate health status assessment value based on a preset standard heart rate health assessment threshold, and generating a user health warning signal when the preset standard heart rate health assessment threshold is higher than the abnormal heart rate health status assessment value; Step S63: Make heart rate control decisions based on the user's health warning signal and build a health warning decision engine to execute the user's health monitoring and warning operations.

[0046] In this embodiment, a personalized user standard map is used, which includes the user's normal heart rate range, activity pattern and historical health data to ensure the accuracy and timeliness of the data. The user's abnormal heart rate situation prediction data is compared with the personalized user standard map to identify heart rate data that deviates from the normal range. The user's abnormal heart rate health status is evaluated through an algorithm (such as weighted average or standard deviation calculation) to generate an abnormal heart rate health status evaluation value, where the evaluation value = , according to clinical standards or expert advice, set one or more standard heart rate health assessment thresholds (for example, the upper and lower limits of the normal heart rate range), compare the generated abnormal heart rate health status assessment value with the preset standard threshold, and use conditional judgment statements (such as if statements) to determine whether the assessment value is lower than the threshold: if the assessment value is lower than the threshold, a health warning signal is generated, and a health warning decision engine is built to integrate user health data, activity patterns, and abnormal heart rate assessment values ​​to form a decision support system, select appropriate decision algorithms (such as rule engines, decision trees, or machine learning-based models) to implement heart rate control decisions, generate specific heart rate control measures (such as suggesting users to rest, increase exercise, or adjust their diet) based on user health warning signals and abnormal assessment values, perform health monitoring and warning operations, continuously track user heart rate changes, and optimize the decision engine based on subsequent data feedback.

[0047] In this embodiment, a health monitoring and early warning system for a smart wearable device is provided, which is used to execute the health monitoring and early warning method for the smart wearable device as described above, including: Personalized feature module, used to obtain real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on smart wearable devices; interpolation filtering and optimization of historical heart rate monitoring data, and personalized user feature mining to build a personalized user standard map; The heart rate waveform module is used to perform multi-point interval sampling of real-time user heart rate monitoring parameters, perform heart rate waveform fitting, and construct a real-time heart rate waveform curve; The heart rate time series variation module is used to analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series variation; The heart rate situation prediction module is used to analyze the heart rate change trend of the user's heart rate time series change representation, and then predict the situation evolution in multiple time periods to generate multi-time period heart rate situation evolution prediction data; The abnormal heart rate recognition module is used to infer the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, and to identify the abnormal stage and mark the user's abnormal heart rate situation prediction data; The health warning module is used to conduct personalized health status assessment of the user's abnormal heart rate trend prediction data based on the personalized user standard map, make heart rate control decisions, and build a health warning decision engine to execute the user's health monitoring and warning operations.

[0048] The present invention obtains real-time user heart rate monitoring parameters and historical data, as well as user body posture data to fully understand the user's physiological condition, interpolates and filters to optimize historical heart rate monitoring data, and performs personalized user feature mining to establish a personalized user standard map, providing a basis for subsequent analysis. Through multi-point interval sampling and heart rate waveform fitting, a real-time heart rate waveform curve is constructed, providing real-time heart rate data visualization, allowing users to intuitively understand their heart rate change trends and fluctuations, analyzing heart rate variability and calculating dynamic variation amplitude to deeply understand the user's heart rate time series change characteristics. These representations can provide important references for subsequent situation prediction and health assessment. By analyzing the trend of heart rate changes and making multi-period situation evolution predictions, heart rate change patterns can be discovered in advance, so that users can understand future heart rate changes and make corresponding adjustments and preventive measures. Abnormal heart rate identification and activity pattern inference are performed based on multi-period heart rate situation evolution prediction data, and the user's abnormal heart rate is marked and identified in time. Users pay attention to health abnormalities and take necessary measures in time. Health status assessment is performed on abnormal heart rate situation prediction data based on personalized user standard maps to provide users with personalized health warnings. The constructed health warning decision engine allows users to make heart rate control decisions based on their personal circumstances, thereby improving health management and monitoring effects.

[0049] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0050] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A health monitoring and early warning method for a smart wearable device, characterized in that: The following steps are involved: Step S1: obtaining real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on the smart wearable device; interpolating and filtering the historical heart rate monitoring data, and mining personalized user features to construct a personalized user standard map; Step S2: performing multi-point interval sampling on the real-time user heart rate monitoring parameters, and performing heart rate waveform fitting to construct a real-time heart rate waveform curve; Step S3: Analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series change; Step S4: analyzing the heart rate change trend of the user's heart rate time series change representation, and then predicting the multi-period situation evolution to generate multi-period heart rate situation evolution prediction data; Step S5: inferring the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, identifying the abnormal stage, and marking the user's abnormal heart rate situation prediction data; Step S6: Based on the personalized user standard map, the user's abnormal heart rate trend prediction data is personalized evaluated for health status, and heart rate control decisions are made to build a health warning decision engine to execute the user's health monitoring and warning operations.

2. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: Acquire real-time user heart rate monitoring parameters and historical heart rate monitoring data based on the smart wearable device; Step S12: performing interpolation filtering and optimization on the historical heart rate monitoring data to obtain historical optimized heart rate monitoring data; Step S13: performing time granularity segmentation on the historical optimized heart rate monitoring data to obtain multiple time-series heart rate monitoring sequences; Step S14: performing user periodic behavior analysis on multiple time-series heart rate monitoring sequences to obtain user behavior characteristics of each sequence; Step S15: performing periodic change analysis on multiple time-series heart rate monitoring sequences to generate a user periodic change pattern; Step S16: Based on the user cycle change rules, personalized user feature mining is performed on the user behavior characteristics of each sequence to construct a personalized user standard map.

3. The health monitoring and early warning method for a smart wearable device according to claim 2, characterized in that: The specific steps of step S12 are: Perform missing value detection on historical heart rate monitoring data to identify missing values; Perform frequency distribution statistics on missing values ​​to obtain missing value frequency distribution data; Based on the missing value frequency distribution data, the surrounding data change analysis of the missing values ​​is performed to extract the surrounding data point change data of each missing value; Calculate the mean of the change data of the surrounding data points of each missing value to obtain the mean of the surrounding data; Perform linear interpolation filling on historical heart rate monitoring data based on the mean value of surrounding data to generate interpolated and filled heart rate monitoring data; Perform outlier identification on the interpolated heart rate monitoring data and mark abnormal data points; Outlier filtering is performed based on abnormal data points to obtain historical optimized heart rate monitoring data.

4. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: defining a time interval, performing multi-point interval sampling on the real-time user heart rate monitoring parameters based on the time interval, and extracting multiple heart rate monitoring sampling points; Step S22: Calculating the time coordinates of multiple heart rate sampling points to extract the time coordinates of each sampling point; Step S23: performing heart rate fluctuation characteristic analysis on a plurality of heart rate sampling points to generate a heart rate fluctuation characteristic; Step S24: performing heart rate waveform fitting on the heart rate fluctuation characteristics based on the time coordinate of each sampling point to construct a real-time heart rate waveform curve.

5. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: Calculating the global heart rate variability of the real-time heart rate waveform curve to generate a global heart rate average variability; Step S32: Divide the real-time heart rate waveform curve into stages to obtain multiple stage waveform curves; Step S33: performing heart rate variability analysis on the waveform curves of multiple stages based on the global average heart rate variation to generate heart rate variability between stages; Step S34: Calculating the dynamic variation amplitude of the heart rate variability between stages and identifying the dynamic heart rate variability amplitude; Step S35: mining the time series changes of the dynamic heart rate variability amplitude to obtain a representation of the user's heart rate time series changes.

6. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S4 are: Step S41: performing multi-scale time-frequency decomposition on the representation of the user's heart rate time series variation to obtain a heart rate variation frequency domain of multiple frequencies; Step S42: performing frequency component distribution analysis on the heart rate variation frequency domain of multiple frequencies to generate frequency component distribution data in the frequency domain; Step S43: performing heart rate change trend analysis based on frequency domain frequency component distribution data to extract heart rate change trend features; Step S44: Predict the evolution of the heart rate trend characteristics over multiple time periods to generate prediction data for the evolution of the heart rate trend over multiple time periods.

7. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: performing jitter frequency statistics on the user's body posture data to obtain the user's jitter frequency parameter; Step S52: performing deep learning classification on the user's jitter frequency parameters to generate a user activity pattern; Step S53: inferring the user's activity pattern in stages based on the real-time heart rate waveform curve to obtain the user's activity pattern in each stage; Step S54: Based on the user activity pattern at each stage, the abnormal stage of the multi-period heart rate situation evolution prediction data is identified, and the abnormal heart rate situation prediction data of the user is marked.

8. The health monitoring and early warning method for a smart wearable device according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing personalized health status assessment on the abnormal heart rate situation prediction data of the user based on the personalized user standard map to generate an abnormal heart rate health status assessment value; Step S62: comparing the abnormal heart rate health status assessment value based on a preset standard heart rate health assessment threshold, and generating a user health warning signal when the preset standard heart rate health assessment threshold is higher than the abnormal heart rate health status assessment value; Step S63: Make heart rate control decisions based on the user's health warning signal and build a health warning decision engine to execute the user's health monitoring and warning operations.

9. A health monitoring and early warning system for smart wearable devices, characterized in that: The method for performing health monitoring and early warning of a smart wearable device according to claim 1 comprises: Personalized feature module, used to obtain real-time user heart rate monitoring parameters, historical heart rate monitoring data and user body posture data based on smart wearable devices; interpolation filtering and optimization of historical heart rate monitoring data, and personalized user feature mining to build a personalized user standard map; The heart rate waveform module is used to perform multi-point interval sampling of real-time user heart rate monitoring parameters, perform heart rate waveform fitting, and construct a real-time heart rate waveform curve; The heart rate time series variation module is used to analyze the heart rate variability of the real-time heart rate waveform curve and calculate the dynamic variation amplitude, so as to obtain the representation of the user's heart rate time series variation; The heart rate situation prediction module is used to analyze the heart rate change trend of the user's heart rate time series change representation, and then predict the situation evolution in multiple time periods to generate multi-time period heart rate situation evolution prediction data; The abnormal heart rate recognition module is used to infer the staged activity pattern of the user's body posture data based on the multi-period heart rate situation evolution prediction data, and to identify the abnormal stage and mark the user's abnormal heart rate situation prediction data; The health warning module is used to conduct personalized health status assessment of the user's abnormal heart rate trend prediction data based on the personalized user standard map, make heart rate control decisions, and build a health warning decision engine to execute the user's health monitoring and warning operations.

Citation Information

Cited By

  • Intelligent wearable device management system and method based on modular design

    CN119292708A

  • Personal health record dynamic management method and device combined with virtual technology

    CN119400338A

  • Electrocardiogram monitoring method and system based on heart rehabilitation data

    CN119454046A

  • Intelligent system for acquiring and analyzing sports data

    CN119479999A

  • Data anomaly detection method, and anomaly detection method and device for equipment monitoring indexes

    CN119577386A