Real-time monitoring method of body temperature and heart rate based on wearable devices

Abstract

This invention relates to the field of heart rate monitoring technology, specifically to a method for real-time monitoring of body temperature and heart rate based on wearable devices. The method includes: collecting a user's PPG signal data, temperature data, and exercise data through a wearable device; calculating a current cumulative exercise intensity index based on the temperature data and exercise data; identifying the corresponding heart rate waveform template and heart rate fluctuation range from a preset cluster based on the cumulative exercise intensity index; using the heart rate duration within the fluctuation range as the fluctuation interval; performing a sliding match between the current PPG signal data and the heart rate waveform template to determine the true peak value; and obtaining the user's current heart rate data based on the true peak value. Using this invention, the accuracy of heart rate monitoring by wearable devices during user exercise can be significantly improved.

Description

Technical Field

[0001] This invention relates to the field of heart rate monitoring technology, and more specifically to a method for real-time monitoring of body temperature and heart rate based on wearable devices. Background Technology

[0002] In non-contact heart rate monitoring, photoplethysmography (PPG), a common non-invasive optical technique, extracts heart rate information by emitting light of a specific wavelength (such as red, green, or infrared) onto the skin and detecting the reflected or transmitted light signals. The DC component of the signal corresponds to the average blood volume of blood and tissues, while the AC component reflects the periodic changes in blood volume caused by heartbeats. The instantaneous heart rate can be calculated by identifying the peak intervals in the AC component. However, in the actual acquisition of PPG signals from hand blood vessels using wearable devices such as smart bracelets, user movement often causes changes in contact pressure between the sensor and the skin, resulting in drastic fluctuations in blood flow velocity. This introduces significant motion artifacts, causing waveform distortion and peak positioning deviations, ultimately affecting the accuracy of heart rate calculation.

[0003] Especially during exercise, such as running, arm swinging causes changes in the relative position of the photodetector and the skin, resulting in variations in reflected light intensity independent of the pulse, exacerbating signal interference. This problem is even more complex in heat tolerance monitoring under high-temperature environments: temperature changes increase skin blood flow and sweat secretion, leading to signal baseline drift; simultaneously, the frequency and temporal characteristics of artifacts generated by the same exercise pattern differ under different heart rates and body temperatures—high heart rates are often accompanied by high-frequency artifacts, while increased temperature easily causes low-frequency drift and waveform morphology variations. Traditional artifact removal methods often fail to fully consider the coupling effect of temperature and physiological state on the signal, resulting in unstable artifact removal performance and decreased heart rate estimation accuracy in complex environments, making it difficult to meet the needs of demanding physiological monitoring scenarios such as heat tolerance. Summary of the Invention

[0004] To address the technical problem of inaccurate heart rate monitoring by wearable devices during user exercise, the present invention aims to provide a real-time body temperature and heart rate monitoring method based on wearable devices. The specific technical solution adopted is as follows:

[0005] The device collects users' PPG signal data, temperature data, and motion data through wearable devices.

[0006] The current cumulative exercise intensity index is calculated based on the temperature data and the exercise data. The cumulative exercise intensity index is used to find the cluster to which the heart rate belongs from the preset cluster group and determine the corresponding heart rate waveform template and heart rate fluctuation range.

[0007] Using the heart rate duration within the heart rate fluctuation range as the fluctuation interval, the current PPG signal data is subjected to sliding matching with the heart rate waveform template to determine the true peak value, and the user's current heart rate data is obtained based on the true peak value.

[0008] Furthermore, the process of obtaining the cumulative exercise intensity index includes:

[0009] The cumulative energy consumption index is obtained based on the exercise data within a preset time period;

[0010] The temperature change is obtained based on the temperature data within the preset time period;

[0011] The cumulative exercise intensity index within the preset time period is calculated based on the cumulative energy consumption index and the temperature change.

[0012] Furthermore, the process of obtaining the cumulative energy consumption index includes:

[0013] Based on the motion data, the user's actions are identified and the energy consumption index corresponding to the actions is calculated. The cumulative energy consumption index is calculated based on the energy consumption index corresponding to each action performed by the user within the preset time.

[0014] Furthermore, the process of obtaining the energy consumption index includes:

[0015] The maximum variation of the action is obtained based on the motion data. The energy consumption index is obtained based on the maximum variation and the mean of various data in the motion data. The maximum variation is used to quantify the peak power output of the user during the exercise. The motion data includes acceleration data and speed data.

[0016] Furthermore, the process of obtaining the temperature change includes:

[0017] The wearable device collects ambient temperature data and user skin temperature data, and obtains core temperature data based on the ambient temperature data and user skin temperature data.

[0018] The temperature change rate is obtained based on the temperature change amplitude of the core temperature data within the preset time and the duration of the preset time. The temperature change is calculated based on the temperature change rate and the average core temperature, wherein the average core temperature is the average value of the core temperature data within the preset time.

[0019] Furthermore, finding the desired cluster from a preset group of categories using the cumulative exercise intensity index includes:

[0020] From the cluster group, a target cluster that is similar to the cumulative motion intensity index is found. The target cluster is the cluster to which it belongs. PPG signal data with similar cumulative motion intensity index values ​​are classified into the same cluster. All clusters constitute the cluster group.

[0021] Furthermore, the process of obtaining the heart rate waveform template and the heart rate fluctuation range includes:

[0022] Determine the cluster center of the class to which the class belongs, and use the heart rate fluctuation corresponding to the cluster center as the heart rate waveform template;

[0023] Obtain the heart rate data corresponding to each PPG signal data contained in each type of cluster in the cluster group, and the set of heart rate data in each type of cluster constitutes the corresponding heart rate fluctuation range.

[0024] Furthermore, the process of determining the true peak value includes:

[0025] Template features of the true template peak in the heart rate waveform template and features to be tested of the peak in the PPG signal data are extracted respectively. The template features include template peak features and template peak duration features, and the features to be tested include peak features to be tested and peak duration features to be tested.

[0026] Matching is performed based on the difference between the template feature and the feature to be tested to obtain a matching value. When the matching value is greater than a preset threshold, the peak value to be tested is the true peak value; otherwise, the peak value to be tested in the PPG signal data is re-determined and matched again.

[0027] Furthermore, the process of obtaining the template features includes:

[0028] The peak characteristics of the template are obtained based on the peak values ​​of the first peak and the second peak in the heart rate waveform template.

[0029] The peak duration feature of the template is obtained based on the peak duration of the first peak and the peak duration of the second peak in the heart rate waveform template.

[0030] Furthermore, the process of obtaining the matching value includes:

[0031] The difference between the template peak feature and the peak feature to be tested is taken as the peak difference;

[0032] The difference between the peak duration feature of the template and the peak duration feature to be tested is taken as the peak duration difference;

[0033] The total feature difference is obtained based on the peak difference and the peak duration difference, and the total feature difference is subjected to negative exponential mapping to obtain the matching value.

[0034] The present invention has the following beneficial effects:

[0035] First, wearable devices collect the user's PPG signal data, temperature data, and exercise data. This forms the data foundation for subsequent analysis in this invention. Second, based on the temperature and exercise data, the current cumulative exercise intensity index is calculated. Using this index, the user's heart rate waveform template and fluctuation range are determined from a preset cluster. The heart rate waveform template is used to filter out the true peak value from the PPG signal data, and the heart rate fluctuation range is used to determine the filtering range. Since the heart rate waveform template and the PPG signal data share the same cumulative exercise intensity index, the template more closely reflects the user's heart rate during their current exercise state. Finally, using the heart rate duration within the fluctuation range as the fluctuation interval, a sliding match is performed between the current PPG signal data and the heart rate waveform template to determine the true peak value. Based on this true peak value, the user's current heart rate data is obtained. The true peak value can be determined by successfully matching the current PPG signal data with the heart rate waveform template. During the matching process, many false peak values ​​can be filtered out, and the successfully matched true peak value is more reliable. Determining heart rate data based on the true peak value of the PPG signal data is existing technology and will not be elaborated further here. Using this invention, the accuracy of obtaining the user's heart rate data during exercise is higher, and robustness under strong artifacts can be significantly improved. Attached Figure Description

[0036] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 A flowchart of the real-time body temperature and heart rate monitoring method based on wearable devices provided in the first embodiment of the present invention;

[0038] Figure 2 A flowchart illustrating the process of obtaining the cumulative exercise intensity index provided in the second embodiment of the present invention;

[0039] Figure 3 The flowchart illustrates the process for determining the true peak value provided in the third embodiment of the present invention. Detailed Implementation

[0040] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the real-time body temperature and heart rate monitoring method based on wearable devices proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0042] The specific solution of the real-time body temperature and heart rate monitoring method based on wearable devices provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0043] Please see Figure 1 The diagram illustrates a flowchart of a real-time body temperature and heart rate monitoring method based on a wearable device provided in the first embodiment of the present invention, the method comprising:

[0044] S101. Collect user's PPG signal data, temperature data, and motion data through wearable devices.

[0045] Wearable devices include portable smart bracelets, which are used in users' normal daily lives. When monitoring heart rate, users do not need to be in a static environment to perform the test. This device is a non-invasive device that can monitor the heart rate of users in motion.

[0046] The PPG signal data can be acquired by a PPG sensor. The temperature data includes ambient temperature data and user skin temperature data, acquired by an ambient temperature sensor and a skin temperature sensor, respectively. The ambient temperature sensor can be installed on the outer shell of the wearable device, and the skin temperature sensor can be installed on the back of the wearable device, directly and closely fitting the skin. The motion data includes acceleration data and angular velocity data. The acceleration data can be acquired by an accelerometer, and the angular velocity data can be acquired by a gyroscope.

[0047] The temperature data is sampled, and the sampling frequency can be set independently, such as 1Hz. The temperature data can be smoothed using a sliding window filtering method, and the window length can be set independently, such as between 5 and 30 seconds. Median filtering or mean filtering can be used.

[0048] S102. Calculate the current cumulative exercise intensity index based on the temperature data and the exercise data, and use the cumulative exercise intensity index to find the corresponding cluster from the preset cluster group and determine the corresponding heart rate waveform template and heart rate fluctuation range.

[0049] During user movement, wearable devices' PPG-based optical heart rate monitoring is primarily affected by motion artifacts, the diversity and complexity of which render traditional filtering methods ineffective. To address this, this invention employs a waveform correlation-based discrimination algorithm: utilizing the periodicity and waveform regularity of the real heart rate signal, a target signal template, i.e., a heart rate waveform template, is constructed. By comparing the real-time acquired PPG signal data with this template, matching signal data is identified as valid heart rate signals, and mismatched artifacts are eliminated, thereby achieving accurate heart rate tracking.

[0050] Furthermore, users exhibit stable heart rate responses under specific exercise intensities. First, a comprehensive quantitative index of exercise intensity is constructed by fusing data from accelerometers, gyroscopes, and temperature sensors. Then, heart rate records matching this intensity are queried from the user's historical database to determine the heart rate fluctuation range, effectively improving the robustness and continuity of heart rate monitoring under motion artifact interference.

[0051] The process of obtaining the cumulative exercise intensity index will be described in detail in the second embodiment of the present invention, and will not be repeated here.

[0052] Furthermore, finding the desired cluster from a preset group of categories using the cumulative exercise intensity index includes:

[0053] From the cluster group, a target cluster that is similar to the cumulative motion intensity index is found. The target cluster is the cluster to which it belongs. PPG signal data with similar cumulative motion intensity index values ​​are classified into the same cluster. All clusters constitute the cluster group.

[0054] It should be noted that the heart rate data or heart rate fluctuations corresponding to each PPG signal data in each cluster of the aforementioned clusters are all known and real.

[0055] In this invention, "similar" means that the cumulative motion intensity index values ​​are similar. First, based on the temperature data and motion data in the historical data acquisition scenario, the cumulative motion intensity index of the historical data is determined. Then, the DBSCAN algorithm can be used for density clustering, and the clustering results are used as different clusters to obtain cluster groups. The mean of the cumulative motion intensity index of all historical data in each cluster group is calculated as a comparison index with the current scenario. The absolute value of the difference between the current cumulative motion intensity index and the comparison index is calculated, and the cluster corresponding to the smallest absolute value of the difference is taken as the "similar" target cluster.

[0056] Furthermore, the process of obtaining the heart rate waveform template and the heart rate fluctuation range includes:

[0057] Determine the cluster center of the class to which the class belongs, and use the heart rate fluctuation corresponding to the cluster center as the heart rate waveform template;

[0058] Obtain the heart rate data corresponding to each PPG signal data contained in each type of cluster in the cluster group, and the set of heart rate data in each type of cluster constitutes the corresponding heart rate fluctuation range.

[0059] The process of obtaining the cluster center is an existing technology, such as using the K-Means algorithm to obtain the cluster center.

[0060] The heart rate fluctuation range can be understood as the range determined by the highest and lowest values ​​in the historical heart rate dataset corresponding to the current PPG signal data.

[0061] S103. Using the heart rate duration within the heart rate fluctuation range as the fluctuation interval, perform sliding matching between the current PPG signal data and the heart rate waveform template to determine the true peak value, and obtain the user's current heart rate data based on the true peak value.

[0062] First, the heart rate fluctuation range is used to predict the time window of the next heartbeat. Then, peak detection is performed only on the current PPG signal data within the prediction window, and the heart rate waveform template is used for matching analysis to distinguish between true and false peaks. The process of identifying candidate peaks during peak detection is existing technology; for example, a first-order derivative zero-crossing detection algorithm can be used to identify all candidate peaks. After identifying all candidate peaks, distinguishing between primary peaks (main peaks) and secondary peaks is also existing technology; for example, peak significance analysis can be used, i.e., setting a significance threshold. Candidate peaks with significance below this threshold are classified as secondary peaks. This threshold can be fixed or adaptive.

[0063] The principle of step S103 can be understood as follows: Define a search window, i.e., the fluctuation range. Within this search window, traverse all candidate peak points and calculate the matching value between each point and the heart rate waveform template. Find the candidate peak point with the highest matching value (i.e., the best matching point) and determine it as the true peak. Move the search window to the best peak point found this time and start searching for the next window. The step size is usually an estimated heart rate interval.

[0064] The sliding step size of the sliding match can be set independently, such as 10ms.

[0065] The process of determining the true peak value will be described in detail in the third embodiment of the present invention, and will not be repeated here.

[0066] Obtaining a user's current heart rate data based on the actual peak values ​​is a prior art technique. Simply put, it involves calculating the time difference between two consecutive actual peak values ​​and converting that time difference into an instantaneous heart rate value. Additionally, the user's historical heart rate records can be updated based on their current heart rate data.

[0067] The division of the fluctuation range involves setting a corresponding time window for the heart rate duration within the heart rate fluctuation range constructed above, based on the current PPG signal data. The time window can divide the current PPG signal data into corresponding time periods. Different windows have different time periods, but under the same division, the peak has a corresponding range of change, and there should be a real heart rate peak within this range.

[0068] It should be noted that, for the convenience of calculation, all index data involved in the calculation in this embodiment of the invention have undergone data standardization processing to eliminate the influence of dimensions and obtain dimensionless data.

[0069] Figure 2 The flowchart illustrates the process of obtaining the cumulative exercise intensity index according to the second embodiment of the present invention. The process of obtaining the cumulative exercise intensity index includes:

[0070] S201. Obtain the cumulative energy consumption index based on the motion data within a preset time period.

[0071] Specifically, the preset time can be one hour in advance, and can be adjusted according to the actual situation without any restrictions.

[0072] Furthermore, the process of obtaining the cumulative energy consumption index includes:

[0073] Based on the motion data, the user's actions are identified and the energy consumption index corresponding to the actions is calculated. The cumulative energy consumption index is calculated based on the energy consumption index corresponding to each action performed by the user within the preset time.

[0074] Identifying user actions based on the motion data is a technique that exists in the prior art, such as machine learning and deep learning.

[0075] The cumulative energy consumption index can be expressed by the formula:

[0076]

[0077] Among them, the Indicates the first The cumulative energy consumption index after completing various actions within a preset time period prior to a certain time. Indicates the first The total number of actions completed within a preset time period prior to a given moment, the Let represent the energy consumption index corresponding to the i-th action, and norm represents linear normalization.

[0078] In this embodiment of the invention, linear normalization can specifically be, for example, maximum and minimum value normalization, which normalizes the data to the range of [0,1]. Subsequent norm normalization also represents this feature, and there are no restrictions on this.

[0079] The cumulative energy consumption index can characterize the user's exercise intensity to a certain extent; the higher the cumulative energy consumption index, the greater the user's exercise intensity.

[0080] Furthermore, the process of obtaining the energy consumption index includes:

[0081] The maximum variation of the action is obtained based on the motion data. The energy consumption index is obtained based on the maximum variation and the mean of various data in the motion data. The maximum variation is used to quantify the peak power output of the user during the exercise. The motion data includes acceleration data and speed data.

[0082] The energy consumption index can be expressed by the formula:

[0083]

[0084] Among them, the The energy consumption index corresponding to the i-th action is... Representing acceleration data, the Representing speed data, the Indicates average speed, the and the All data are normalized. The normalization method can be the range normalization method. In this embodiment of the invention, maximum and minimum value normalization is applied to both acceleration and velocity data, that is, maximum and minimum value normalization is performed on the acceleration dimension to obtain acceleration data. The speed data is obtained by performing maximum and minimum value normalization on the speed dimension. This ensures that acceleration and velocity data undergo appropriate standardization. The average acceleration is represented by the following. and the All are the average values ​​of the normalized data. This represents the function for finding the maximum value. This indicates the maximum change in the action, i.e., within a preset time period of the user's movement (the first...). The peak power output of the action performed within a preset time period prior to the specified moment.

[0085] The intensity of motion is directly reflected in angular velocity (the rate of rotation) and acceleration (the rate of change of velocity). The larger these physical quantities are, the greater the kinetic energy of the body parts and the greater the impact force generated. Therefore, angular velocity and acceleration can be used as indicators to quantify energy consumption.

[0086] Furthermore, actions with higher peak power output tend to have higher total energy consumption. As a weighting factor, it has a certain degree of rationality.

[0087] S202. Obtain the temperature change based on the temperature data within the preset time period.

[0088] Furthermore, the process of obtaining the temperature change includes:

[0089] The wearable device collects ambient temperature data and user skin temperature data, and obtains core temperature data based on the ambient temperature data and user skin temperature data.

[0090] The temperature change rate is obtained based on the temperature change amplitude of the core temperature data within the preset time and the duration of the preset time. The temperature change is calculated based on the temperature change rate and the average core temperature, wherein the average core temperature is the average value of the core temperature data within the preset time.

[0091] It should be noted that obtaining core temperature data based on the ambient temperature data and the user's skin temperature data is an existing technology, such as the dual-sensor heat flux method.

[0092] The temperature change can be expressed by the following formula:

[0093]

[0094] Among them, the Indicates the first The difference between the maximum and minimum temperatures within a preset time period prior to a given moment. The duration of the preset time, the Indicates the first The average temperature over a preset time period prior to a given moment, the Indicates the first The temperature change over a preset time period prior to a given moment. `norm` represents linear normalization.

[0095] The temperature changes can also serve as an indicator of a user's exercise intensity to some extent, because a larger temperature change indicates a greater exercise intensity. The principle is that when a user exercises, their body generates heat, causing their body temperature to rise. The faster the body temperature rises and the higher the temperature it is maintained, the greater the exercise intensity.

[0096] S203. Calculate the cumulative exercise intensity index within the preset time period based on the cumulative energy consumption index and the temperature change.

[0097] The cumulative exercise intensity index can be expressed by the formula:

[0098]

[0099] Among them, the Indicates the first The cumulative exercise intensity index within a preset time period prior to a given moment, the The duration of the preset time, the Indicates the first The cumulative energy consumption index after completing various actions within a preset time period prior to a certain time. Indicates the first Temperature changes within a preset time period prior to a given moment.

[0100] Figure 3 The flowchart below shows the process for determining the true peak value provided in the third embodiment of the present invention. The process for determining the true peak value includes:

[0101] S301. Extract the template features of the true template peak in the heart rate waveform template and the test features of the peak to be tested in the PPG signal data respectively, wherein the template features include template peak features and template peak duration features, and the test features include: peak to be tested features and peak duration features.

[0102] Furthermore, the process of obtaining the template features includes:

[0103] The template peak characteristics are obtained based on the peak values ​​of the primary and secondary peaks in the heart rate waveform template. These template peak characteristics can be expressed by the following formula:

[0104]

[0105] Among them, the The template peak characteristics are represented by the This represents the peak value of the first peak in the heart rate waveform template. This represents the peak value of the second peak in the heart rate waveform template. The value is generally not 0. When it becomes 0 due to data abnormalities or other reasons, it is automatically adjusted to the preset minimum value. The preset minimum value is a safety value set according to the heart rate waveform, such as 10% of the peak value of the main wave (R wave) in the same heartbeat.

[0106] The peak duration feature of the template is obtained based on the peak duration of the primary peak and the peak duration of the secondary peak in the heart rate waveform template. The peak duration feature of the template can be expressed by the formula:

[0107]

[0108] Among them, the The template peak duration characteristic is represented by the This indicates the peak duration of the first peak in the heart rate waveform template. This indicates the peak duration of the secondary peak in the heart rate waveform template.

[0109] Furthermore, the process of obtaining the feature to be tested includes:

[0110] The peak characteristics to be tested are obtained based on the peak values ​​of the primary and secondary peaks in the PPG signal data. These peak characteristics can be expressed by the following formula:

[0111]

[0112] Among them, the The peak feature to be tested is represented by the peak feature to be tested. This represents the peak value of the first peak in the PPG signal data. This represents the peak value of the second peak in the PPG signal data, and its value is not 0. Specifically, it is related to the template peak characteristics using the formula... The processing method is similar to that of the previous one.

[0113] The peak duration characteristic to be tested is obtained based on the peak duration of the first peak and the peak duration of the second peak in the PPG signal data. The peak duration characteristic to be tested can be expressed by the formula:

[0114]

[0115] Among them, the The duration characteristic of the peak to be tested is represented by the following: This indicates the peak duration of the first peak in the PPG signal data. This indicates the peak duration of the secondary peak in the PPG signal data.

[0116] It should be noted that, in order to distinguish between the template features and the features to be tested, the subscripts in the above formula are... This indicates the meaning of the heart rate waveform template, and the subscript indicates that... This indicates the meaning of a single peak in the heart rate waveform template, indicated by the subscript. This indicates the meaning of the second peak in the heart rate waveform template, indicated by the subscript. Indicates the meaning of PPG signal data, subscript Indicates the meaning of the first peak in PPG signal data, subscript This indicates the meaning of the secondary peak in the PPG signal data.

[0117] S302. Match the template features and the features to be tested to obtain a matching value. When the matching value is greater than a preset threshold, the peak value to be tested is the true peak value. Otherwise, the peak value to be tested in the PPG signal data is re-determined and matched again.

[0118] The preset threshold can be set independently, such as 0.85.

[0119] Furthermore, the process of obtaining the matching value includes:

[0120] The difference between the template peak feature and the peak feature to be tested is taken as the peak difference, which can be expressed by the formula:

[0121]

[0122] Among them, the Indicating the peak difference, the The template peak characteristics are represented by the This indicates the peak characteristic to be tested.

[0123] The difference between the template peak duration feature and the peak duration feature to be tested is taken as the peak duration difference, which can be expressed by the formula:

[0124]

[0125] Among them, the This indicates the difference in the duration of the peak value. The template peak duration characteristic is represented by the This indicates the duration characteristic of the peak value to be tested.

[0126] The total feature difference is obtained based on the peak difference and the peak duration difference. A negative exponential mapping is then applied to the total feature difference to obtain a matching value. In this embodiment of the invention, the total feature difference represents the feature distinction between the peak to be tested and the template peak. Therefore, during the calculation process, the peak difference and the peak duration difference can be dimensionlessly processed, retaining only the numerical features. The specific dimensionless processing method is well-known to those skilled in the art and will not be elaborated upon. After dimensionless processing, the total feature difference is used to calculate the total feature difference.

[0127] The total characteristic difference can be expressed by the formula:

[0128]

[0129] Among them, the The total feature difference is represented by the following. Indicating the peak difference, the The peak duration difference is represented by norm, which represents linear normalization.

[0130] The matching value can be expressed by the formula:

[0131]

[0132] Among them, the Indicates the matching value, the This represents an exponential function with the natural constant e as its base. The total feature difference is represented by the following. This represents the absolute value function.

[0133] The negative exponential mapping indicates that there is an inverse relationship between the independent and dependent variables, where the smaller the independent variable is, the larger the dependent variable is. In this embodiment of the invention, it means that the greater the difference in total features, the smaller the matched value after mapping. Therefore, the negative exponential mapping process is achieved through the above formula.

[0134] The present invention has the following beneficial effects:

[0135] First, wearable devices collect the user's PPG signal data, temperature data, and exercise data. This forms the data foundation for subsequent analysis in this invention. Second, based on the temperature and exercise data, the current cumulative exercise intensity index is calculated. Using this index, the user's heart rate waveform template and fluctuation range are determined from a preset cluster. The heart rate waveform template is used to filter out the true peak value from the PPG signal data, and the heart rate fluctuation range is used to determine the filtering range. Since the heart rate waveform template and the PPG signal data share the same cumulative exercise intensity index, the template more closely reflects the user's heart rate during their current exercise state. Finally, using the heart rate duration within the fluctuation range as the fluctuation interval, a sliding match is performed between the current PPG signal data and the heart rate waveform template to determine the true peak value. Based on this true peak value, the user's current heart rate data is obtained. The true peak value can be determined by successfully matching the current PPG signal data with the heart rate waveform template. During the matching process, many false peak values ​​can be filtered out, and the successfully matched true peak value is more reliable. Determining heart rate data based on the true peak value of the PPG signal data is existing technology and will not be elaborated further here. Using this invention, the accuracy of obtaining the user's heart rate data during exercise is higher, and robustness under strong artifacts can be significantly improved.

[0136] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0137] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for real-time monitoring of body temperature and heart rate based on a wearable device, characterized in that, The method comprises: acquiring PPG signal data, temperature data and motion data of a user through a wearable device; calculating a current cumulative motion intensity index according to the temperature data and the motion data, finding a corresponding cluster from a preset cluster group through the cumulative motion intensity index, and determining a corresponding heart rate waveform template and a heart rate fluctuation range; taking a heart rate duration under the heart rate fluctuation range as a fluctuation interval, and performing sliding matching between the current PPG signal data and the heart rate waveform template to determine a real peak value, and acquiring current heart rate data of the user based on the real peak value; the acquisition process of the cumulative motion intensity index comprises: acquiring a cumulative energy consumption index according to the motion data within a preset time; acquiring a temperature change condition according to the temperature data within the preset time; calculating the cumulative motion intensity index within the preset time based on the cumulative energy consumption index and the temperature change condition; finding a corresponding cluster from a preset cluster group through the cumulative motion intensity index comprises: finding a target cluster similar to the cumulative motion intensity index from the cluster group, and the target cluster is the corresponding cluster, wherein the PPG signal data with similar cumulative motion intensity index values are divided into the same cluster, and all clusters form the cluster group; the acquisition process of the heart rate waveform template and the heart rate fluctuation range comprises: determining a cluster center of the corresponding cluster, and taking a heart rate fluctuation condition corresponding to the cluster center as the heart rate waveform template; acquiring heart rate data corresponding to each PPG signal data contained in each cluster in the cluster group, and the set of heart rate data in each cluster constitutes a corresponding heart rate fluctuation range. 2.The wearable device-based real-time monitoring method of body temperature and heart rate according to claim 1, wherein, the acquisition process of the cumulative energy consumption index comprises: identifying a user's action based on the motion data and calculating an energy consumption index corresponding to the action, and calculating the cumulative energy consumption index according to the energy consumption index corresponding to each action of the user within the preset time. 3.The wearable device-based real-time monitoring method of body temperature and heart rate according to claim 2, wherein, the acquisition process of the energy consumption index comprises: acquiring a maximum change condition of the action based on the motion data, and acquiring the energy consumption index according to the maximum change condition and the mean value of each type of data in the motion data, wherein the maximum change condition is used to quantify the power output peak value of the user in the motion process, and the motion data includes acceleration data and speed data. 4.The wearable device-based real-time monitoring method of body temperature and heart rate according to claim 1, wherein, the acquisition process of the temperature change condition comprises: acquiring environmental temperature data and user skin temperature data through the wearable device, and acquiring core temperature data based on the environmental temperature data and the user skin temperature data; acquiring a temperature change rate based on the temperature change amplitude of the core temperature data within the preset time and the duration of the preset time, and calculating the temperature change condition according to the temperature change rate and the average core temperature, wherein the average core temperature is the average value of the core temperature data within the preset time. 5.The wearable device-based real-time monitoring method of body temperature and heart rate according to claim 1, wherein, the determination process of the real peak value comprises: Template features of the real template peak in the heart rate waveform template and to-be-inspected features of a to-be-inspected peak in the PPG signal data are extracted respectively, wherein the template features include template peak features and template peak duration features, and the to-be-inspected features include to-be-inspected peak features and to-be-inspected peak duration features; A matching value is obtained by matching according to a difference between the template features and the to-be-inspected features, and when the matching value is greater than a preset threshold, the to-be-inspected peak is the real peak, otherwise, the to-be-inspected peak in the PPG signal data is determined again for re-matching. 6.The wearable device-based real-time monitoring method of body temperature and heart rate according to claim 5, wherein, The process of obtaining the template features includes: The template peak features are obtained according to a peak value of a primary peak and a peak value of a secondary peak in the heart rate waveform template; The template peak duration features are obtained according to a peak duration of the primary peak and a peak duration of the secondary peak in the heart rate waveform template. 7.The wearable device-based real-time temperature and heart rate monitoring method of claim 5, wherein, The process of obtaining the matching value includes: A difference between the template peak features and the to-be-inspected peak features is taken as a peak value difference; A difference between the template peak duration features and the to-be-inspected peak duration features is taken as a peak duration difference; A total feature difference is obtained according to the peak value difference and the peak duration difference, and a matching value is obtained by negative exponential mapping processing on the total feature difference.

Images