Intelligent health monitoring method and system for Internet pension

Abstract

The invention provides an intelligent health monitoring method and system for Internet pension. The method comprises the following steps: receiving sensor data of a target user transmitted by a millimeter wave radar sensor based on the Internet; the sensor data comprises pose information, track information and micro-motion information; the condition for triggering the millimeter wave radar sensor is that the infrared sensor monitors a target user; state recognition is carried out based on the pose information and the track information, and a current state mode of the target user at the current time is determined; performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain a health monitoring condition of the target user at the current time; and if the health monitoring condition is an abnormal state, monitoring information containing the health monitoring condition and the current position of the target user is sent to a mobile terminal of a preset contact person through the Internet. According to the invention, the effectiveness and timeliness of Internet old-age care health monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an intelligent health monitoring method and system for Internet-based elderly care. Background Art

[0002] With the aging of the population, internet-based elderly care has become an important way to ensure the health of the elderly, and health monitoring technology plays a key role in this. Currently, mainstream health monitoring methods rely on wearable devices, such as smart bracelets and smart watches. These devices use built-in sensors to collect real-time physiological data such as the user's heart rate and sleep, and then transmit this data to the terminal via wireless networks for analysis and processing. However, current wearable devices require regular charging and maintenance, making them inconvenient to use and making it difficult to obtain monitoring data continuously and stably. This greatly affects the effectiveness and timeliness of health monitoring, making it difficult to meet the needs of timely and effective health monitoring for the elderly in internet-based elderly care. Summary of the Invention

[0003] The present invention provides an intelligent health monitoring method and system for Internet-based elderly care, which are used to improve the effectiveness and timeliness of health monitoring for Internet-based elderly care.

[0004] In a first aspect, the present invention provides an intelligent health monitoring method for Internet-based elderly care, comprising: Receiving sensor data of a target user transmitted by a millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; Performing state recognition based on the posture information and the trajectory information to determine the current state mode of the target user at the current time; Performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain a health monitoring status of the target user at the current time; If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to a mobile terminal of a preset contact via the Internet.

[0005] In a second aspect, the present invention further provides an intelligent health monitoring system for Internet-based elderly care, which is applied to the intelligent health monitoring method for Internet-based elderly care as described in the first aspect; the intelligent health monitoring system for Internet-based elderly care includes: A receiving module is configured to receive sensor data of a target user transmitted by a millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; and the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; A state recognition module is used to perform state recognition based on the posture information and the trajectory information to determine the current state mode of the target user at the current time; a health monitoring module, configured to perform health monitoring on the target user based on the current state mode and the micro-motion information, and obtain a health monitoring status of the target user at the current time; The sending module is used to send monitoring information including the health monitoring status and the current location of the target user to a mobile terminal of a preset contact via the Internet if the health monitoring status is abnormal.

[0006] In a third aspect, the present invention also provides an electronic device comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby implementing any of the above-mentioned intelligent health monitoring methods for Internet-based elderly care.

[0007] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, it implements any of the above-mentioned intelligent health monitoring methods for Internet-oriented elderly care.

[0008] In a fifth aspect, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned intelligent health monitoring methods for Internet-oriented elderly care.

[0009] The intelligent health monitoring method for internet-based elderly care provided by an embodiment of the present invention uses millimeter-wave radar sensors for monitoring. Because millimeter-wave radar sensors can operate stably and long-term, they do not require frequent charging and maintenance, ensuring continuous acquisition of monitoring data and improving the effectiveness of internet-based elderly care health monitoring. Furthermore, the method can quickly and accurately identify the user's status pattern through posture information, trajectory information, and micro-motion information. Based on this status pattern and micro-motion information, the user's health is quickly and accurately monitored, obtaining the user's current health monitoring status. Based on this health monitoring status, monitoring information is quickly sent to the corresponding preset contacts, improving the timeliness of internet-based elderly care health monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 This is a flow chart of an intelligent health monitoring method for Internet-based elderly care provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of an intelligent health monitoring system for Internet-based elderly care provided by an embodiment of the present invention; Figure 3 An embodiment diagram of an electronic device provided by an embodiment of the present invention; Figure 4 An embodiment diagram of a computer-readable storage medium provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0011] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0012] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0013] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0014] Optional, see Figure 1 , Figure 1 This is a flow chart of the intelligent health monitoring method for Internet-based elderly care provided by the present invention. In the embodiment of the present invention, the execution subject of the intelligent health monitoring method for Internet-based elderly care is a health monitoring system. Therefore, the intelligent health monitoring method for Internet-based elderly care includes: Step 10: Receive sensor data of the target user transmitted by the millimeter-wave radar sensor over the internet. This sensor data includes position information, trajectory information, and micro-motion information. The millimeter-wave radar sensor is triggered when the infrared sensor detects the target user.

[0015] Optionally, the infrared sensor continuously monitors the surrounding environment and determines whether a target user has entered the monitoring area by detecting infrared radiation emitted by objects. When the infrared sensor detects the target user, it immediately triggers the millimeter-wave radar sensor to start working. Millimeter-wave radar sensors are installed in major activity areas such as bedrooms, living rooms, and bathrooms in nursing homes. The millimeter-wave radar sensor uses electromagnetic waves in the millimeter-wave frequency band to detect the target user and can obtain the target user's posture information, trajectory information, and micro-motion information. Posture information includes the three-dimensional coordinates and posture angles of the target user; trajectory information records the target user's speed data and acceleration data; micro-motion information refers to subtle movements of the target user's body, such as the body movement amplitude of different limb parts, the joint angles of different limb parts, and the breathing rate.

[0016] The millimeter-wave radar sensor then encodes and packages the collected sensor data and sends it to the health monitoring system's server via an Internet transmission protocol (such as TCP / IP). The server decodes the data and obtains the millimeter-wave radar sensor's sensor data.

[0017] Step 20: Perform state recognition based on the posture information and trajectory information to determine the current state mode of the target user at the current time.

[0018] Furthermore, the health monitoring system performs state recognition based on the posture information and trajectory information to determine the current state mode of the target user at the current time, wherein the current state mode includes a motion state mode, a static state mode and a sleep state mode, as specifically described in steps 201 to 205.

[0019] Step 30 : Perform health monitoring on the target user based on the current state mode and micro-motion information to obtain the health monitoring status of the target user at the current time.

[0020] Furthermore, the health monitoring system performs health monitoring on the target user according to whether the current state mode is a motion state mode, a static state mode or a sleep state mode, combined with the corresponding micro-motion information, that is, the body movement amplitude of different limb parts, the joint angles of different limb parts or the breathing frequency, to obtain the health monitoring status of the target user at the current time, wherein the health monitoring status represents whether the user has abnormalities, as specifically described in steps 301 to 313.

[0021] Step 40: If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to the mobile terminal of the preset contact via the Internet.

[0022] Furthermore, when the health monitoring status of the target user is judged to be abnormal, in order to take timely measures to ensure the safety of the target user, the health monitoring system integrates the health monitoring status and the current location of the target user (location information obtained by the millimeter wave radar sensor or determined in combination with other positioning technologies) to generate monitoring information, and sends the monitoring information to the mobile terminal of the preset contact through the Internet, using text messages, instant messaging application interfaces (such as WeChat public account message push, APP push, etc.) or emails. Among them, the preset contacts are usually the target user's family members, emergency contacts or medical staff, who are entered in advance in the health monitoring system according to actual conditions.

[0023] The embodiment of the present invention uses millimeter-wave radar sensors for monitoring. Because millimeter-wave radar sensors can operate stably over long periods of time and require no frequent charging or maintenance, they ensure continuous acquisition of monitoring data and improve the effectiveness of internet-based elderly care health monitoring. Through posture information, trajectory information, and micro-motion information, the user's status pattern can be quickly and accurately identified. Based on the status pattern combined with micro-motion information, the user's health can be quickly and accurately monitored to obtain the user's current health monitoring status. Based on the health monitoring status, monitoring information can be quickly sent to the corresponding preset contacts, improving the timeliness of internet-based elderly care health monitoring.

[0024] In one embodiment, steps 201 to 205 are described as follows: Step 201 determines a pose change vector at two adjacent time points based on the three-dimensional coordinates and attitude angles of the two adjacent time points within the current time, and determines the trajectory curvature of each time point based on the pose change vector at the two adjacent time points. The pose change vector includes a coordinate change vector and an attitude angle change vector.

[0025] Optionally, the health monitoring system obtains the three-dimensional coordinates of the target user at different time points in the current time and attitude angles (usually including pitch angles , yaw angle , roll angle ). For two adjacent time points and , by calculating the three-dimensional coordinate difference , , , get the coordinate change vector ; By calculating the attitude angle difference , , , get the attitude angle change vector , and the two together constitute the pose change vector.

[0026] Furthermore, the health monitoring system uses Frenet-Serret to combine the posture change vectors of two adjacent time points to calculate the trajectory curvature of each time point. For a three-dimensional space curve, the trajectory curvature The calculation formula is: .in, For time The velocity vector can be calculated from the posture change vector and the time interval. is the time derivative of the velocity vector, i.e. the acceleration vector.

[0027] In the home health monitoring scenario, millimeter wave radar sensors are The three-dimensional coordinates of the old man collected at all times are (1, 2, 1.5), and the posture angle is (10 , 0 , 0 );exist The three-dimensional coordinates collected at the moment are (1.2, 2.1, 1.4), and the attitude angle is (12 , 2 , 0 ). Calculate the coordinate change vector =(1.2-1, 2.1-2, 1.4-1.5)=(0.2, 0.1, -0.1), attitude angle change vector =(12 -10 , 2 -0 , 0 -0 )=(2 , 2 , 0 ). For example, the time interval =0.5s, velocity vector = =(0.4, 0.2, -0.2), by further calculating the velocity vector derivative, etc., the above formula can be used to obtain Curvature of the trajectory at a moment .

[0028] Step 202: Perform a correlation analysis based on the velocity data and trajectory curvature at each time point and the average velocity data and average trajectory curvature in the current time to determine the velocity curvature correlation, and perform a coupling analysis based on the acceleration data and attitude angle change vector at each time point and the average acceleration data and average attitude angle change vector in the current time to determine the acceleration attitude angle coupling.

[0029] Furthermore, for the velocity curvature correlation, the health monitoring system uses the canonical correlation analysis method to combine the velocity data to form a vector set and the trajectory curvature form a vector set , through CCA, find the linear combination of two sets of vectors to maximize the correlation between them. The maximum correlation coefficient obtained is the velocity curvature correlation Furthermore, for the acceleration attitude angle coupling, the health monitoring system uses the mutual information method to combine the acceleration data to form a vector set and attitude angle change vector set Calculating mutual information , that is, the acceleration attitude angle coupling is used to measure the dependency between two sets of random variables, and its calculation formula is: ,in, and They are and The probability distribution of for and The joint probability distribution of .

[0030] Continuing with the above example, we obtain the speed data and trajectory curvature of the elderly at multiple time points. We use the CCA method to process these data, find the best linear combination of speed data and trajectory curvature, and calculate the speed curvature correlation. =0.8. At the same time, the acceleration data and attitude angle change vector of the elderly are obtained, and the acceleration attitude angle coupling degree is calculated by calculating the probability distribution and using the mutual information formula. =0.7.

[0031] Step 203 : Count the number of peak values ​​of acceleration data exceeding a preset acceleration threshold, and determine the acceleration peak density based on the number of peak values ​​and the duration of the current time.

[0032] Furthermore, the health monitoring system pre-sets an acceleration threshold , traverse the acceleration data within the current time , statistics satisfy The peak number Furthermore, the health monitoring system can be used to monitor the peak value of and the duration of the current time Calculate the peak acceleration density, where the peak acceleration density The calculation formula is: .

[0033] In one embodiment, the acceleration threshold is preset =5 , during the 10s monitoring time of the elderly, the number of peak values ​​of the elderly's acceleration data exceeding the threshold is counted =8 times, calculate the peak acceleration density according to the formula times / s.

[0034] Step 204 : Count the number of times the change amplitude corresponding to the posture change vector is greater than a preset amplitude threshold, and determine the posture change frequency based on the number of changes and the duration of the current time.

[0035] Furthermore, the health monitoring system pre-sets the threshold value of the posture change amplitude , for each pose change vector , calculate its amplitude , statistics satisfy Number of changes Furthermore, the health monitoring system calculates the posture change frequency based on the number of changes and the duration of the current time, wherein the posture change frequency The specific formula is: .

[0036] In one embodiment, the positioning posture change amplitude threshold =0.5 (unit is a comprehensive unit calculated based on the actual coordinates and angles). In monitoring the elderly, the number of changes in the pose change vector amplitude greater than the threshold is counted within 10 seconds. =12 times, calculate the frequency of posture change according to the formula =1.2 times / s.

[0037] Step 205 : Perform state recognition based on the velocity curvature correlation, acceleration attitude angle coupling, acceleration peak density, and posture change frequency at the current time to determine the current state mode.

[0038] Furthermore, the health monitoring system performs state identification based on the velocity curvature correlation, acceleration attitude angle coupling, acceleration peak density and posture change frequency at the current time to determine the current state mode, as specifically described in steps 2051 to 2054.

[0039] The embodiment of the present invention comprehensively considers the interrelationships and changing characteristics of multiple factors such as speed, trajectory, acceleration, and posture during the target user's movement process. When faced with complex actual scenarios, it can more accurately distinguish different state modes, thereby quickly and accurately monitoring the user's health based on the micro-motion information under different state modes, and obtaining the user's health monitoring status at the current time, so that the monitoring information can be quickly sent to the corresponding preset contacts, thereby improving the timeliness and accuracy of Internet-based elderly health monitoring.

[0040] In one embodiment, steps 2051 to 2054 are described as follows: Step 2051 , taking each time point as a starting point, based on the speed data of each time point within the preset time window, determines the speed fluctuation degree of each time point within the preset time window.

[0041] Optional, for each time point , the health monitoring system defines a preset time window (Here is the window duration). Taking each time point as the starting point, obtain the speed data sequence of the target user within the time window. (k is the number of data points in the window). Based on the speed data at each time point in the preset time window, the standard deviation of the speed data is calculated to measure the speed fluctuation degree. The larger the standard deviation, the more severe the speed fluctuation. Therefore, the speed fluctuation degree ,in, is the average value of the velocity data in the time window, that is .

[0042] Continue to monitor the elderly in the home health monitoring scenario. For example, the preset time window is 2s. At this moment, the velocity data sequence obtained in the first 2 seconds is {1.2, 1.5, 1.0, 1.3, 1.4} (unit: m / s). First calculate the average value of this sequence =(1.2+1.5+1.0+1.3+1.4) / 5=1.28m / s, and then calculate the speed fluctuation degree according to the formula 0.17m / s.

[0043] Step 2052: Determine the attitude angle change trend vector within the time span between each time point and the first time point based on the change slope between the attitude angle at each time point and the attitude angle at the first time point.

[0044] Furthermore, the health monitoring system records the target user's posture angles at different time points, including pitch angles , yaw angle , roll angle For each time point , calculate the pitch angle, yaw angle, roll angle and the first time point respectively The slope of the corresponding attitude angle. Taking the pitch angle as an example, the slope of the attitude angle is The calculation formula is: Similarly, the slope of the yaw angle change can be obtained and the slope of the roll angle change The three change slopes are combined into the attitude angle change trend vector , the attitude angle change trend vector reflects the change trend of the attitude angle from the starting time point to the current time point.

[0045] In one embodiment, The pitch angle of the old man =10 , yaw angle =0 , roll angle =0 ; The pitch angle of the old man =20 , yaw angle =5 , roll angle =2 , total monitoring time =4s. Then the slope of pitch angle change is =(20-10) / 4=2.5 , yaw angle change slope =(5-0) / 4=1.25 , the slope of the roll angle change =(2-0) / 4=0.5 , attitude angle change trend vector =(2.5, 1.25, 0.5).

[0046] Step 2053: Determine the consistency of the trajectory direction based on the average of the cosine values ​​of the angles between the posture change vectors at all two adjacent time points.

[0047] Furthermore, the health monitoring system obtains two adjacent time points\ and The pose change vector and , calculate the cosine of the angle between them according to the vector dot product formula : Furthermore, the health monitoring system traverses the posture change vectors of all adjacent time points and calculates a series of angle cosine values (n is the total number of time points), and the consistency of the trajectory direction is obtained by taking the average value , The closer the consistency is to 1, the more consistent the trajectory direction is; the closer it is to -1, the more dramatic the change in trajectory direction is; and the closer it is to 0, the more random the change in trajectory direction is. In one embodiment, there are 10 time points in total, and the posture change vectors of adjacent time points are obtained, such as =(0.2, 0.1, -0.1), =(0.3, 0.1, -0.05), calculate the cosine of the angle between them 0.92. After calculating the cosine values ​​of the angles between all adjacent posture change vectors, the average value is used to obtain the degree of consistency of the trajectory direction. =0.85.

[0048] Step 2054, based on the current speed fluctuation degree, attitude angle change trend vector, consistency degree, speed curvature correlation, acceleration attitude angle coupling degree, acceleration peak density and posture change frequency, state identification is performed to determine the current state mode.

[0049] Furthermore, the health monitoring system performs state identification based on the current speed fluctuation degree, attitude angle change trend vector, consistency degree, speed curvature correlation, acceleration attitude angle coupling, acceleration peak density and posture change frequency to determine the current state mode, as described in steps 20541 to 25043.

[0050] The embodiment of the present invention comprehensively considers multiple factors such as the degree of speed fluctuation, attitude angle change trend vector, consistency degree, speed curvature correlation, acceleration attitude angle coupling, acceleration peak density and posture change frequency. Therefore, when faced with complex actual scenarios, it can more accurately distinguish different state modes, thereby quickly and accurately monitoring the user's health based on the micro-motion information under different state modes, and obtaining the user's health monitoring status at the current time, so that the monitoring information can be quickly sent to the corresponding preset contact, thereby improving the timeliness and accuracy of Internet elderly care health monitoring.

[0051] In one embodiment, steps 20541 to 25043 are described as follows: Step 20541: If the speed fluctuation degree at the current time is greater than the preset speed fluctuation threshold, and the acceleration peak density is greater than the preset density threshold, and the posture change frequency is greater than the preset change frequency threshold, and the speed curvature correlation is greater than the preset correlation threshold, and the acceleration attitude angle coupling is greater than the preset coupling threshold, then determine that the current state mode is the motion state mode.

[0052] Optionally, since the user's speed usually fluctuates greatly when in motion, the body generates more acceleration peaks, the posture changes frequently, and there is a strong correlation and coupling relationship between speed and trajectory curvature, and between acceleration and posture angle. Therefore, if the speed fluctuation degree at the current time is greater than the preset speed fluctuation threshold, and the acceleration peak density is greater than the preset density threshold, and the posture change frequency is greater than the preset change frequency threshold, and the speed curvature correlation is greater than the preset correlation threshold, and the acceleration posture angle coupling is greater than the preset coupling threshold, the health monitoring system determines that the current state mode is the motion state mode.

[0053] Step 20542: If the speed fluctuation degree at the current time is less than the preset speed fluctuation threshold, and the acceleration peak density is less than the preset density threshold, and the attitude angle change value corresponding to the attitude angle change trend vector is less than the preset change threshold, and the speed curvature correlation is greater than the preset correlation threshold, and the acceleration attitude angle coupling is greater than the preset coupling threshold, then determine that the current state mode is the static state mode.

[0054] Furthermore, in a stationary state, the target user's speed fluctuates slightly, acceleration peaks are low, and attitude angle changes are not significant. However, certain correlations and coupling characteristics are maintained between speed and trajectory curvature, and between acceleration and attitude angle. Therefore, if the current speed fluctuation is less than a preset speed fluctuation threshold, the acceleration peak density is less than a preset density threshold, the attitude angle change value corresponding to the attitude angle change trend vector is less than a preset change threshold, the speed curvature correlation is greater than a preset correlation threshold, and the acceleration attitude angle coupling is greater than a preset coupling threshold, the health monitoring system determines that the current state mode is a stationary state mode.

[0055] Step 20543: If the speed fluctuation degree at the current time is less than the preset speed fluctuation threshold, and the acceleration peak density is less than the preset density threshold, and the attitude angle change value corresponding to the attitude angle change trend vector is less than the preset change threshold, and the speed curvature correlation is greater than the preset correlation threshold, and the acceleration attitude angle coupling is greater than the preset coupling threshold, and the consistency degree is greater than the preset consistency degree threshold, then determine that the current state mode is the sleep state mode.

[0056] Furthermore, since the target user has minimal physical activity during sleep, speed fluctuations are minimal, acceleration peaks are almost non-existent, and attitude angle changes are minimal, while the direction of the body's motion trajectory during sleep is relatively stable, and there is still an intrinsic connection between speed and trajectory curvature, and between acceleration and attitude angle. Therefore, if the current speed fluctuation level is less than a preset speed fluctuation threshold, and the acceleration peak density is less than a preset density threshold, and the attitude angle change value corresponding to the attitude angle change trend vector is less than a preset change threshold, and the speed curvature correlation is greater than a preset correlation threshold, and the acceleration attitude angle coupling is greater than a preset coupling threshold, and the consistency level is greater than a preset consistency level threshold, the health monitoring system determines that the current state mode is a sleep state mode.

[0057] The embodiment of the present invention constructs a state recognition system based on multi-dimensional feature threshold judgment, so it can comprehensively characterize the differences in the motion characteristics of the target user in different states, so that in actual health monitoring scenarios, whether it is the motion state, static state or sleep state of daily activities, different state modes can be distinguished more accurately, so as to quickly and accurately monitor the user's health according to the micro-motion information in different state modes, and obtain the user's health monitoring status at the current time, so that the monitoring information can be quickly sent to the corresponding preset contact, thereby improving the timeliness and accuracy of Internet elderly health monitoring.

[0058] In one embodiment, steps 301 to 305 are described as follows: Step 301: If the current state mode is the motion state mode, the body movement amplitude of each limb part in the current time is compared with the body movement amplitude benchmark of each limb part to obtain the body movement amplitude deviation of each limb part.

[0059] Optionally, after determining that the target user is in motion mode, the health monitoring system obtains the body motion data of each limb of the target user (such as arms, legs, torso, etc.) at different time points in the current time. The health monitoring system pre-sets a body motion amplitude baseline value for each limb part. The baseline value is obtained based on a large amount of normal motion data statistics, reflecting the body motion amplitude range of the corresponding limb part of healthy people in normal motion state. For each limb part, the health monitoring system calculates the difference between its actual body motion amplitude and the corresponding body motion amplitude baseline value in the current time to obtain the body motion amplitude deviation, wherein the body motion amplitude deviation is ( Indicates the The calculation formula for each limb part is: .in, The first The actual body movement amplitude of each limb part is obtained by calculating the sum of the absolute values ​​of the displacement changes of the limb part at each time point; For the The body movement amplitude benchmark value of each limb part.

[0060] Continuing with the home health monitoring scenario, the elderly person is walking indoors (in motion mode). The millimeter-wave radar sensor detects that the elderly person's right arm is moving. During this time, the displacement changes at each time point are 0.1m, 0.12m, 0.08m, etc. The absolute values ​​of these displacement changes are added together to obtain the actual body movement amplitude of the right arm. =1.5m, the reference value of the right arm's body movement amplitude =1.2m, calculate the deviation of the right arm's body movement amplitude according to the formula =|(1.5-1.2) / 1.2|=0.25.

[0061] Step 302 : performing movement consistency and time synchronization analysis based on the body movement amplitudes between the various limb parts to obtain a movement coordination index between the various limb parts.

[0062] Furthermore, the health monitoring system analyzes the body movement amplitude of each limb of the target user at the same time point to evaluate the consistency of movement. At the same time, it compares the time series of the body movement amplitude changes of each limb part, analyzes the time synchronization, uses the dynamic time warping algorithm to measure the similarity between the time series of the body movement amplitude of different limb parts, and calculates the consistency of movement in combination with the cosine similarity. For example, there are m limb parts in total, and their body movement amplitude time series are First, calculate any two limb parts using the DTW algorithm. and Time series matching distance , and then calculate their cosine similarity The closer the movement coordination index is to 1, the better the movement coordination between the limbs is; the closer the movement coordination index is to 0, the worse the movement coordination is. The calculation formula is: .

[0063] In one embodiment, when the elderly person is walking, the health monitoring system obtains the time series of the body movement amplitude of the elderly person's right arm, left arm, right leg, and left leg. The time series matching distance of the right arm and left arm is calculated by the DTW algorithm. =0.3, cosine similarity =0.8; Similarly, calculate the distance and similarity between other limbs. Substitute these values ​​into the formula to calculate the movement coordination index =0.7 (other calculation items are omitted here).

[0064] Step 303 : Obtain the body movement amplitude attenuation rate of each limb part based on the attenuation rate of the body movement amplitude of each limb part at the last time point relative to the body movement amplitude at the first time point.

[0065] Furthermore, for each limb part, the health monitoring system obtains its body movement amplitude at the first time point in the current monitoring period. and the body movement amplitude at the last time point , the body motion amplitude attenuation rate is obtained by calculating the proportional relationship between the two ( Indicates the The body motion amplitude attenuation rate reflects the changing trend of the body motion amplitude of the limb during the movement. The larger the body motion amplitude attenuation rate, the more obvious the body motion amplitude attenuation. The calculation formula is: .

[0066] Continuing with the example of the old man’s right arm, During the monitoring period, the right arm Body movement amplitude at each moment =0.15m, Body movement amplitude at each moment =0.1m, calculate the body motion amplitude attenuation rate of the right arm according to the formula .

[0067] Step 304 : determining a health monitoring coefficient based on the movement coordination index between the various limb parts and the body movement amplitude deviation and body movement amplitude attenuation rate of each limb part.

[0068] Furthermore, the health monitoring system constructs a model based on support vector data description to determine the health monitoring coefficient, specifically: the movement coordination index , the deviation of the body movement amplitude of each limb and body motion amplitude attenuation Composing multidimensional feature vectors Input into the SVDD model. The SVDD model searches for a minimum volume hypersphere in high-dimensional space and contains the eigenvectors in the normal healthy state as much as possible within the sphere. , calculate its distance to the center of the hypersphere , after normalization, the health monitoring coefficient is obtained , the formula is: .in, and are the minimum and maximum distances from the feature vector to the center of the hypersphere in the training data, respectively. The closer the value is to 0, the closer the target user is to normal health status; the larger the value is, the more likely the health status is abnormal.

[0069] Continuing with the above example, the elderly's movement coordination index =0.7, right arm, left arm, right leg, left leg body movement amplitude deviation =0.25, =0.1, =0.15, =0.2, and the body motion amplitude attenuation rate =0.33, =0.2, =0.25, =0.3 to form the eigenvector , input into the trained SVDD model, and calculate the distance D=0.6. =0.1, =0.9, calculate the health monitoring coefficient according to the formula (0.6-0.1) / (0.9-0.1)=0.625.

[0070] Step 305: If the health monitoring coefficient is greater than or equal to the preset health coefficient threshold, the health monitoring status is determined to be normal, otherwise it is determined to be abnormal.

[0071] Furthermore, the health monitoring system pre-sets a preset health coefficient threshold , the calculated health monitoring coefficient is compared with the threshold. When , the health monitoring system determines that the health monitoring status of the target user is normal; when , it is determined to be in an abnormal state.

[0072] When the target user is in motion mode, the embodiment of the present invention uses a support vector data description model to combine multiple dimensions of limb movement amplitude, movement coordination, and movement amplitude change trend to perform analysis and calculate the health monitoring coefficient. Therefore, it can more accurately capture the complex relationship between the characteristics of each dimension and the health status, so that the health abnormalities of the target user can be quickly and accurately identified in complex motion scenarios, thereby improving the accuracy and reliability of health monitoring, and further quickly sending monitoring information to the corresponding preset contacts based on the health monitoring status, thereby improving the timeliness of Internet elderly health monitoring.

[0073] In one embodiment, steps 306 to 309 are described as follows: Step 306: If the current state mode is the static state mode, a correlation analysis is performed based on the joint angles between the various limb parts at the current time to obtain the correlation between the joint angles between the various limb parts.

[0074] Optionally, when the target user is determined to be in a static state, the health monitoring system obtains the joint angle data of each limb part of the target user (such as shoulder joint, elbow joint, hip joint, knee joint, etc.), and uses the canonical correlation analysis method to perform correlation analysis on the joint angle data of different limb parts. Specifically, there are m limb parts in total, and their joint angle data respectively constitute a vector set The CCA method maximizes the correlation between these linear combinations by finding the linear combinations between vector sets. The maximum correlation coefficient obtained is the correlation between the joint angles of each limb part. (Indicates the and between body parts).

[0075] Continuing with the home health monitoring scenario, the elderly man is sitting in a chair in a static state. The angle data of the elderly man’s right arm elbow is collected to form a vector , the left knee joint angle data constitutes a vector The CCA method is used to analyze these two sets of data, find their best linear combination, and calculate the joint angle correlation between the right arm elbow joint and the left leg knee joint. =0.3, the joint angle correlation reflects the correlation degree between the changes in the joint angles of the two limb parts.

[0076] Step 307 : Determine the joint angle deviation of each limb part based on the degree of deviation between the joint angle of each limb part and the angle threshold at the corresponding dwell time.

[0077] Furthermore, the health monitoring system pre-sets the corresponding angle threshold range for each limb part at different dwell times. For each limb part, its actual joint angle is obtained during the dwell time of the current static state. , and the angle threshold range under the corresponding dwell time For comparison, the joint angle deviation ( Indicates the The calculation formula for each limb part is: .

[0078] Therefore, it can be understood that when the actual joint angle is within the threshold range, the joint angle deviation is 0; otherwise, the joint angle deviation is obtained by calculating the minimum distance between the actual angle and the threshold boundary and normalizing it with the maximum value of the threshold range.

[0079] Continuing with the above example, after the elderly person sits on the chair for a while, the health monitoring system records that the actual joint angle of the elderly person's left shoulder joint is 120 , the angle threshold range of the limb part at the current residence time is [100 , 110 ] Calculate the joint angle deviation of the left shoulder joint according to the formula =|120-110| / 120≈0.083.

[0080] Step 308 : Based on the joint angle deviation of each limb part and the joint angle correlation between each limb part, determine the joint angle change value of each limb part at the corresponding dwell time.

[0081] Furthermore, the health monitoring system is constructed based on the kernel principal component analysis model, which calculates the joint angle deviation of each limb part. The correlation between the joint angles of this limb and all other limbs Composing multidimensional feature vectors The KPCA model is used to reduce the dimension and extract features of the eigenvector to obtain a comprehensive eigenvalue, which is the change value of the joint angle of each limb part under the corresponding dwell time. KPCA uses kernel functions to map original data into high-dimensional space, which can better capture the nonlinear relationship of data.

[0082] Continuing with the example of the right elbow joint of the elderly, the joint angle deviation =0.1, and the correlations with the joint angles of other limbs such as the left shoulder joint and the left knee joint are =0.2, = 0.3, etc., and these data are combined into feature vectors Input into the trained KPCA model and calculate the change value of the right elbow joint angle =0.25.

[0083] Step 309: If the joint angle change value of each limb part is within the corresponding change threshold range, the health monitoring status is determined to be normal. Otherwise, it is abnormal.

[0084] Furthermore, the health monitoring system pre-sets the corresponding joint angle change threshold range for each limb part . Change the joint angle of each limb Compare with the corresponding change threshold range. When the joint angle change values ​​of all limbs meet When , the health monitoring system determines that the health monitoring status of the target user is in normal condition; as long as the joint angle change value of any limb part exceeds the range, it is determined to be in abnormal condition.

[0085] Continuing with the above embodiment, the joint angle change threshold range of the elderly person's right elbow joint is [0, 0.2], and the calculated right elbow joint angle change value is =0.25, which exceeds the threshold range. Therefore, the health monitoring system determines that the health monitoring status of the elderly person in the current static state is abnormal.

[0086] The embodiment of the present invention evaluates the target user's physical condition in multiple dimensions, from the correlation analysis of joint angles in the static state mode, to the calculation of the degree of deviation, and then to the comprehensive determination of the joint angle change value and comparison with the threshold to determine the health status. Therefore, it can more accurately capture the complex relationship between the characteristics of each dimension and the health status, so that the health anomalies of the target user can be quickly and accurately identified in complex motion scenarios, thereby improving the accuracy and reliability of health monitoring, and further quickly sending monitoring information to the corresponding preset contacts based on the health monitoring status, thereby improving the timeliness of Internet-based elderly health monitoring.

[0087] In one embodiment, steps 310 to 313 are described as follows: Step 310: If the current state mode is the sleep state mode, the respiratory frequency fluctuation degree is determined based on the comparison between the respiratory frequency in the current time and the frequency reference value.

[0088] Optionally, when the target user is determined to be in sleep mode, the health monitoring system obtains the target user's respiratory data and converts it into respiratory rate data through a signal processing algorithm. The health monitoring system pre-sets a respiratory rate benchmark value, which is based on the statistical data of a large number of normal sleeping people and reflects the average respiratory rate of healthy people in sleep. Therefore, the health monitoring system calculates the average respiratory rate within the current time. and the respiratory rate baseline value The standard deviation is used to measure the degree of respiratory rate fluctuation. The greater the degree of respiratory rate fluctuation, the more severe the respiratory rate fluctuation. The calculation formula is: .in, For the Respiratory rate at a given time point, The number of respiratory rate data points collected during the current time.

[0089] In the home health monitoring scenario, the elderly are in sleep mode. Within minutes, respiratory rate data is collected once every minute to obtain the respiratory rate sequence {14, 15, 16, ..., 15} (unit: times / minute). First calculate the average value of the sequence =(14+15+16+...+15) / 60=15 times / minute, the baseline respiratory rate = 16 times / minute. The respiratory rate fluctuation is calculated to be 0.13 according to the formula.

[0090] Step 311: Calculate the respiratory frequency variation entropy based on the respiratory frequency variation. The respiratory frequency variation entropy represents the disorder degree of the respiratory frequency.

[0091] Furthermore, the health monitoring system analyzes the respiratory rate data within the current time and uses the information entropy calculation method to measure the disorder of the respiratory rate. First, the respiratory rate data is discretized and the probability of different respiratory rate values ​​is calculated. ( Represents different respiratory rate values), where the respiratory rate change entropy The calculation formula is: .in, is the number of different respiratory rate values ​​after discretization. The larger the entropy of respiratory rate change, the more disordered the respiratory rate change; the smaller the entropy of respiratory rate change, the more regular the respiratory rate change.

[0092] Continuing with the above embodiment, after discretizing the collected respiratory rate data, it is found that there are three different respiratory rate values ​​of 14 times / minute, 15 times / minute, and 16 times / minute, and the number of times they appear is 10 times, 35 times, and 15 times respectively. =10 / 60, =35 / 60, =15 / 60. According to the formula, the entropy of respiratory rate change is 1.48.

[0093] Step 312: Perform linear fitting based on the respiratory frequency to obtain the respiratory frequency change trend.

[0094] Furthermore, the health monitoring system considers the respiratory rate data at different time points in the current time as a time series and uses the least squares method for linear fitting. , the corresponding respiratory rate is , the equation of the fitting line is .

[0095] Furthermore, the health monitoring system solves the equations according to the least squares principle. Reaching the minimum value, thus determining the coefficient and Among them, the coefficient Reflects the changing trend of respiratory rate over time. The larger the absolute value of , the more obvious the trend of respiratory rate change; If it is close to 0, it means that the respiratory rate is relatively stable.

[0096] Continue to collect the respiratory rate data of the elderly at 60 time points during sleep, = 1 minute, = 2 minutes, ..., =60 minutes as the horizontal axis, and the corresponding respiratory rate as the vertical axis. Using the least squares method for linear fitting, the fitted straight line equation f=-0.02t+15.2 is obtained, where the coefficient a=-0.02, indicating that the elderly person's respiratory rate showed a slowly decreasing trend during this period of sleep monitoring.

[0097] In step 313, if the respiratory rate fluctuation degree is less than the frequency fluctuation threshold, the respiratory rate change entropy is less than the preset frequency change entropy threshold, and the respiratory rate change trend is less than the preset change trend threshold, then the health monitoring status is determined to be normal. Otherwise, it is abnormal.

[0098] Furthermore, the health monitoring system pre-sets a frequency fluctuation threshold, a preset frequency change entropy threshold, and a preset change trend threshold. The respiratory frequency fluctuation degree, respiratory frequency change entropy, and the absolute value of the coefficient of the respiratory frequency change trend are compared with the corresponding thresholds respectively. Only when the respiratory frequency fluctuation degree is less than the frequency fluctuation threshold, the respiratory frequency change entropy is less than the preset frequency change entropy threshold, and the absolute value of the coefficient of the respiratory frequency change trend is less than the preset change trend threshold are met at the same time, the health monitoring system determines that the health monitoring status of the target user in the sleep state is in a normal state; as long as one of the conditions is not met, the health monitoring system determines that it is in an abnormal state.

[0099] The embodiment of the present invention systematically quantifies the degree of fluctuation of respiratory frequency for the sleep state pattern, calculates the disorder of frequency change by information entropy, and then determines the change trend through linear fitting to analyze the respiratory state during sleep from multiple angles. Therefore, it can accurately capture the subtle change characteristics of respiratory frequency during sleep and accurately evaluate the health status during sleep. Whether it is abnormal fluctuation, disordered change or trend change of respiratory frequency, it can be detected in time, so that in actual health monitoring scenarios, the ability to identify potential health problems during sleep is effectively improved, thereby quickly and accurately identifying the health abnormalities of the target user, improving the accuracy and reliability of health monitoring, and further quickly sending monitoring information to the corresponding preset contact according to the health monitoring status, thereby improving the timeliness of Internet elderly health monitoring.

[0100] Furthermore, the intelligent health monitoring system for Internet-oriented elderly care provided by the present invention is described below. The intelligent health monitoring system for Internet-oriented elderly care described below and the intelligent health monitoring method for Internet-oriented elderly care described above can refer to each other.

[0101] Optional, see Figure 2 , Figure 2 This is a structural diagram of the intelligent health monitoring system for Internet-based elderly care provided by the present invention. The intelligent health monitoring system for Internet-based elderly care includes:

[0102] The receiving module 210 is configured to receive sensor data of a target user transmitted by the millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; A state recognition module 220 is used to perform state recognition based on the posture information and trajectory information to determine the current state mode of the target user at the current time; The health monitoring module 230 is used to perform health monitoring on the target user based on the current state mode and micro-motion information, and obtain the health monitoring status of the target user at the current time; The sending module 240 is configured to send monitoring information including the health monitoring status and the current location of the target user to a mobile terminal of a preset contact via the Internet if the health monitoring status is abnormal.

[0103] The embodiment of the present invention uses millimeter-wave radar sensors for monitoring. Because millimeter-wave radar sensors can operate stably over long periods of time and require no frequent charging or maintenance, they ensure continuous acquisition of monitoring data and improve the effectiveness of internet-based elderly care health monitoring. Through posture information, trajectory information, and micro-motion information, the user's status pattern can be quickly and accurately identified. Based on the status pattern combined with micro-motion information, the user's health can be quickly and accurately monitored to obtain the user's current health monitoring status. Based on the health monitoring status, monitoring information can be quickly sent to the corresponding preset contacts, improving the timeliness of internet-based elderly care health monitoring.

[0104] See also Figure 3 , Figure 3 This is a diagram of an embodiment of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented: Receive sensor data of the target user transmitted by the millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; Perform state recognition based on posture information and trajectory information to determine the current state mode of the target user at the current time; Perform health monitoring on the target user based on the current state pattern and micro-motion information to obtain the health monitoring status of the target user at the current time; If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to the mobile terminal of the preset contact via the Internet.

[0105] See also Figure 4 , Figure 4 Detailed description of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, the following steps are implemented: Receive sensor data of the target user transmitted by the millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; Perform state recognition based on posture information and trajectory information to determine the current state mode of the target user at the current time; Perform health monitoring on the target user based on the current state pattern and micro-motion information to obtain the health monitoring status of the target user at the current time; If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to the mobile terminal of the preset contact via the Internet.

[0106] On the other hand, the present invention further provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the intelligent health monitoring method for Internet-oriented elderly care provided by the above methods, which includes: Receive sensor data of the target user transmitted by the millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; Perform state recognition based on posture information and trajectory information to determine the current state mode of the target user at the current time; Perform health monitoring on the target user based on the current state pattern and micro-motion information to obtain the health monitoring status of the target user at the current time; If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to the mobile terminal of the preset contact via the Internet.

[0107] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0108] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An intelligent health monitoring method for Internet-based elderly care, characterized in that: include: Receiving sensor data of a target user transmitted by a millimeter-wave radar sensor over the Internet; The sensor data includes posture information, trajectory information and micro-motion information; The condition for triggering the millimeter wave radar sensor is that the infrared sensor detects the target user; Performing state recognition based on the posture information and the trajectory information to determine the current state mode of the target user at the current time; Performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain a health monitoring status of the target user at the current time; If the health monitoring status is abnormal, monitoring information including the health monitoring status and the current location of the target user is sent to a mobile terminal of a preset contact via the Internet.

2. The intelligent health monitoring method for Internet-based elderly care according to claim 1 is characterized in that: The posture vector includes three-dimensional coordinates and posture angles, and the trajectory information includes velocity data and acceleration data; and performing state recognition based on the posture information and the trajectory information to determine the current state mode of the target user at the current time includes: Determine, based on the three-dimensional coordinates and attitude angles of two adjacent time points within the current time, a posture change vector of the two adjacent time points, and determine, based on the posture change vectors of the two adjacent time points, a trajectory curvature of each time point; the posture change vector includes a coordinate change vector and an attitude angle change vector; Performing a correlation analysis based on the speed data and trajectory curvature at each time point and the average speed data and average trajectory curvature during the current time to determine the speed curvature correlation, and performing a coupling analysis based on the acceleration data and attitude angle change vector at each time point and the average acceleration data and average attitude angle change vector during the current time to determine the acceleration attitude angle coupling; Counting the number of peak values ​​of the acceleration data exceeding a preset acceleration threshold, and determining the acceleration peak density based on the number of peak values ​​and the duration of the current time; Counting the number of times the change amplitude corresponding to the posture change vector is greater than a preset amplitude threshold, and determining the posture change frequency based on the number of changes and the duration of the current time; State recognition is performed based on the velocity curvature correlation, acceleration attitude angle coupling, acceleration peak density and posture change frequency at the current time to determine the current state mode.

3. The intelligent health monitoring method for Internet-based elderly care according to claim 2 is characterized in that: The state identification based on the velocity curvature correlation, acceleration attitude angle coupling, acceleration peak density and posture change frequency at the current time to determine the current state mode includes: Taking each time point as a starting point, based on the speed data of each time point within a preset time window, determining the speed fluctuation degree of each time point within the preset time window; Based on the change slope between the attitude angle at each time point and the attitude angle at the first time point, determining the attitude angle change trend vector within the time span of each time point and the first time point; The consistency of the trajectory direction is determined based on the mean cosine of the angle between the pose change vectors of all two adjacent time points; State recognition is performed based on the speed fluctuation degree, attitude angle change trend vector, consistency degree, speed curvature correlation, acceleration attitude angle coupling degree, acceleration peak density and posture change frequency at the current time to determine the current state mode.

4. The intelligent health monitoring method for Internet-based elderly care according to claim 3 is characterized in that: The state identification is performed based on the speed fluctuation degree, attitude angle change trend vector, consistency degree, speed curvature correlation, acceleration attitude angle coupling degree, acceleration peak density and posture change frequency at the current time to determine the current state mode, including: If the speed fluctuation degree at the current time is greater than a preset speed fluctuation threshold, and the acceleration peak density is greater than a preset density threshold, and the posture change frequency is greater than a preset change frequency threshold, and the speed curvature correlation is greater than a preset correlation threshold, and the acceleration attitude angle coupling is greater than a preset coupling threshold, then determine that the current state mode is a motion state mode; If the speed fluctuation degree at the current time is less than a preset speed fluctuation threshold, and the acceleration peak density is less than a preset density threshold, and the attitude angle change value corresponding to the attitude angle change trend vector is less than a preset change threshold, and the speed curvature correlation is greater than a preset correlation threshold, and the acceleration attitude angle coupling is greater than a preset coupling threshold, then it is determined that the current state mode is a stationary state mode; If the speed fluctuation degree at the current time is less than the preset speed fluctuation threshold, and the acceleration peak density is less than the preset density threshold, and the attitude angle change value corresponding to the attitude angle change trend vector is less than the preset change threshold, and the speed curvature correlation is greater than the preset correlation threshold, and the acceleration attitude angle coupling is greater than the preset coupling threshold, and the consistency degree is greater than the preset consistency degree threshold, then it is determined that the current state mode is the sleep state mode.

5. The intelligent health monitoring method for Internet-based elderly care according to claim 4 is characterized in that: The micro-motion information includes the body movement amplitude of different limb parts; The performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain the health monitoring status of the target user at the current time includes: If the current state mode is a motion state mode, then comparing the body movement amplitude of each limb part during the current time with the body movement amplitude benchmark of each limb part to obtain a body movement amplitude deviation of each limb part; Based on the movement amplitude between each limb part, the movement consistency and time synchronization analysis are performed to obtain the movement coordination index between each limb part; Based on the attenuation rate of the body motion amplitude of each limb part at the last time point relative to the body motion amplitude at the first time point, the body motion amplitude attenuation rate of each limb part is obtained; Determine the health monitoring coefficient based on the movement coordination index between various limb parts and the body movement amplitude deviation and body movement amplitude attenuation rate of each limb part; If the health monitoring coefficient is greater than or equal to a preset health coefficient threshold, the health monitoring status is determined to be in a normal state; otherwise, it is determined to be in an abnormal state.

6. The intelligent health monitoring method for Internet-based elderly care according to claim 4 is characterized in that: The micro-motion information includes joint angles of different limb parts; The performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain the health monitoring status of the target user at the current time includes: If the current state mode is a static state mode, performing a correlation analysis based on the joint angles between the various limb parts within the current time to obtain the correlation between the joint angles between the various limb parts; Determining the joint angle deviation of each limb part based on the degree of deviation of the joint angle of each limb part from the angle threshold at the corresponding dwell time; Based on the joint angle deviation of each limb part and the joint angle correlation between each limb part, determining the joint angle change value of each limb part at the corresponding dwell time; If the joint angle change value of each limb part is within its corresponding change threshold range, the health monitoring status is determined to be in a normal state; otherwise, it is in an abnormal state.

7. The intelligent health monitoring method for Internet-based elderly care according to claim 4 is characterized in that: The micro-motion information includes respiratory rate; The performing health monitoring on the target user based on the current state mode and the micro-motion information to obtain the health monitoring status of the target user at the current time includes: If the current state mode is the sleep state mode, determining the degree of respiratory frequency fluctuation based on comparing the respiratory frequency in the current time with a frequency reference value; Calculating respiratory frequency variation entropy based on the frequency variation of the respiratory frequency; the respiratory frequency variation entropy represents the disorder degree of the respiratory frequency; Performing linear fitting based on the respiratory frequency to obtain a respiratory frequency change trend; If the respiratory frequency fluctuation degree is less than the frequency fluctuation threshold, and the respiratory frequency change entropy is less than the preset frequency change entropy threshold, and the respiratory frequency change trend is less than the preset change trend threshold, then the health monitoring status is determined to be in a normal state; otherwise, it is in an abnormal state.

8. An intelligent health monitoring system for Internet-based elderly care, characterized by: Applicable to the intelligent health monitoring method for Internet-oriented elderly care as described in any one of claims 1 to 7; The intelligent health monitoring system for Internet-based elderly care includes: A receiving module is configured to receive sensor data of a target user transmitted by a millimeter-wave radar sensor over the Internet; the sensor data includes position information, trajectory information, and micro-motion information; and the condition for triggering the millimeter-wave radar sensor is that the infrared sensor detects the target user; A state recognition module is used to perform state recognition based on the posture information and the trajectory information to determine the current state mode of the target user at the current time; a health monitoring module, configured to perform health monitoring on the target user based on the current state mode and the micro-motion information, and obtain a health monitoring status of the target user at the current time; The sending module is used to send monitoring information including the health monitoring status and the current location of the target user to a mobile terminal of a preset contact via the Internet if the health monitoring status is abnormal.

9. An electronic device comprising: Memory for storing computer software programs; A processor, used to read and execute the computer software program, characterized in that when the processor executes the computer software program, it implements the intelligent health monitoring method for Internet-oriented elderly care as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing a computer software program, wherein: When the computer software program is executed by a processor, the intelligent health monitoring method for Internet-oriented elderly care as claimed in any one of claims 1 to 7 is implemented.

Images