Individualized activity monitoring and whole-process management system for child health care

Abstract

The invention discloses a personalized activity monitoring and full-process management system for child health care, and relates to the technical field of internet, when the system runs, activity state information of children can be reflected in real time by acquiring a child activity state evaluation index AS, and the problems that evaluation is lagged and subjectivity is high in dependence on questionnaires, interviews and other modes are solved. The child activity state evaluation index AS is compared with the health evaluation state threshold value T through the detection and trigger module, the grade abnormity standard AB is generated, timely and personalized adjustment guidance is provided for parents, and negative effects caused by the fact that problems are not perceived in time are avoided. Besides, the improvement condition of the child activity state is evaluated in a fixed period, a new activity state evaluation index ASN is generated, and an intervention suggestion generation scheme is continuously optimized, so that accurate evaluation of the multi-dimensional child activity and psychological state can be realized, and intervention suggestions can be generated in real time and continuously optimized.

Description

Technical Field

[0001] The present invention relates to the field of Internet technology, and in particular to a personalized activity monitoring and full-process management system for children's health care. Background Art

[0002] The field of child health care is an important branch of modern medicine and public health that has received much attention. It covers comprehensive monitoring and intervention of children's physical health, psychological development and behavioral patterns. In recent years, as society has paid more and more attention to children's mental health, children's activity monitoring has gradually developed into a multi-dimensional assessment system covering physical activity, emotional behavior and psychological state. Among these systems, management platforms that integrate wearable devices, biosensors and intelligent algorithms have gradually become mainstream, and they can record children's activity data in real time.

[0003] Although some technologies and systems for monitoring children's activities and mental states have been put into use, most of them have the problems of single functions and insufficient data integration. For example, many devices can only record single indicators such as steps or heart rate, and lack multi-dimensional assessment of children's mental state. In addition, traditional mental state assessments mostly rely on questionnaires or interviews, which have a long time period, rely on the initiative of parents, and are easily affected by subjective factors, and cannot achieve real-time monitoring. For children, this method often cannot capture the dynamic characteristics of psychological changes in a timely manner, thus missing the best time to intervene, especially in complex emotional scenarios, such as emotional fluctuations caused by learning pressure, family conflicts or social problems. It is often impossible to accurately predict the child's mental state through a single activity data or emotional marker, and thus it is impossible to provide effective intervention suggestions. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a personalized activity monitoring and full-process management system for children's health care, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a personalized activity monitoring and full-process management system for children's health care, including a data monitoring module, a data processing and extraction module, an activity evaluation module, a detection and triggering module, a suggestion generation module and an iterative optimization module;

[0006] The data monitoring module collects real-time data of children through wearable devices to generate a children's activity feature vector set FV;

[0007] The data processing and extraction module performs preprocessing and feature extraction on the children's activity feature vector set FV to generate a standard feature vector set SFV;

[0008] The activity evaluation module evaluates the activity status information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and establishes an activity status evaluation algorithm formula for evaluation to obtain the child activity status evaluation index AS;

[0009] The detection and triggering module compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result, and generates the grade abnormality standard AB according to the child assessment state result;

[0010] The suggestion generation module generates a child intervention suggestion generation plan according to the abnormal grade standard AB, the activity state information and the inactivity state information, and provides prompts and notifications according to the content of the child intervention suggestion generation plan;

[0011] The iterative optimization module iteratively evaluates the child's activity status in a fixed period according to the child intervention suggestion generation plan, generates a new activity status evaluation index ASN, and compares it with the child activity status evaluation index AS to determine the effectiveness of the child intervention suggestion generation plan.

[0012] Preferably, the data monitoring module includes a data acquisition unit;

[0013] The data acquisition unit performs real-time data acquisition on the child through a wearable device, including using a three-axis acceleration sensor to record the acceleration change value per unit time to obtain the activity intensity AIt at time t; using a microphone to collect the child's voice signal SF; using a camera to obtain the child's facial video frame sequence to obtain the smile frequency EF within a fixed period; using a heart rate sensor and an acceleration sensor to collect the child's sleep information, obtain the resting heart rate HR and the number of turning over M, and generate a child activity feature vector set FV = {AIt, SF, EF, HR, M}.

[0014] Preferably, the data processing and extraction module includes a preprocessing unit and a feature extraction unit;

[0015] The preprocessing unit preprocesses the children's activity feature vector set FV, including data cleaning preprocessing, data denoising preprocessing and data normalization preprocessing, to obtain the preprocessed feature vector set FVp;

[0016] Among them, the data cleaning preprocessing includes using the linear interpolation method to process missing values ​​of the feature vector set FV, using the box plot method to process outliers of the feature vector set FV, and then checking and processing outliers and missing values ​​in the feature vector set FV;

[0017] Data denoising preprocessing includes using a sliding average filter method to perform time series data smoothing on the feature vector set FV after data cleaning preprocessing, using a low-pass filter to filter high-frequency noise on the feature vector set FV after data cleaning preprocessing, and then checking and eliminating the noise caused by equipment and environmental interference in the feature vector set FV after data cleaning preprocessing;

[0018] The data normalization preprocessing includes using the minimum-maximum normalization method to normalize the feature vector set FV after the data denoising preprocessing, and then mapping the feature vector set FV after the data denoising preprocessing to the data range of 0 to 1 to eliminate the dimension difference. The data normalization preprocessing feature vector set FV = {AItn, SFn, EFn, HRn, Mn}.

[0019] Preferably, the feature extraction unit performs feature extraction on the preprocessed feature vector set FVp to obtain activity intensity feature information, sound feature information, expression feature information and sleep feature information;

[0020] Among them, the activity intensity feature information includes the average activity intensity AIm and the activity fluctuation amplitude AIv; the sound feature information includes the sound signal energy average SFe and the sound signal peak frequency SFf; the expression feature information includes the smiling behavior proportion EFs; the sleep feature information includes the sleep score SDs, and the standard feature vector set SFV = {AIm, AIv, SFe, SFf, EFs, SDs} is generated.

[0021] Preferably, the activity evaluation module includes a state evaluation unit and an evaluation index unit;

[0022] The state evaluation unit evaluates the activity state information and inactivity state information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and obtains activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information;

[0023] The activity intensity is fitted by the average activity intensity AIm, the average sound signal energy SFe and the smiling behavior proportion EFs, and is marked as activity intensity SAI, which reflects the activity intensity of children;

[0024] The frequency fluctuation is fitted by the average activity intensity AIm, the peak frequency SFf of the sound signal and the proportion of smiling behavior EFs, and is marked as frequency fluctuation SFV, which reflects the frequency fluctuation of children's activities;

[0025] The activity duration is fitted by the sleep score SDs and the activity fluctuation amplitude AIv, and is marked as activity duration SAC.

[0026] Preferably, the evaluation index unit evaluates the activity intensity SAI, frequency fluctuation SFV and activity duration SAC according to the acquired activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information by establishing an activity state evaluation algorithm formula to reflect the comprehensive performance of activity intensity, frequency fluctuation and duration in children's activities, and obtain the children's activity state evaluation index AS;

[0027] The Children's Activity Status Assessment Index AS is The calculation formula is obtained, where e represents the exponential function.

[0028] Preferably, the detection and triggering module includes an anomaly detection unit and an intervention suggestion generation unit;

[0029] The abnormality detection unit compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result;

[0030] The child assessment status results are obtained by comparing:

[0031] When the child activity status assessment index AS is less than the health assessment status threshold T, the child assessment status result is obtained as a qualified status, marked as Status = 0, indicating that the child's activity status is normal and no special intervention is required, and the child should continue to maintain his activity habits and work and rest schedule;

[0032] When the child's activity status assessment index AS ≥ the health assessment status threshold T, the child's assessment status result is obtained as an unqualified state, marked as Status = 1, indicating that the child's activity status is unqualified and that the activity arrangement and content need to be adjusted, and the activity habits and work and rest schedule should be adjusted according to the activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information;

[0033] When the child activity status assessment index AS ≥ twice the health assessment status threshold T, the child assessment status result is obtained as an abnormal state, marked as status Status = 2, indicating that the child's activity is abnormal. Parents or managers are asked to adjust the activity content and the child's rest time. If there is an abnormal activity intensity, an abnormal report is generated based on the activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information.

[0034] Preferably, the intervention suggestion generating unit automatically generates a grade abnormality standard AB according to the child assessment status result, and generates a child intervention suggestion generating scheme according to the grade abnormality standard AB, adjusts the frequency of notification and prompt of the child assessment status result, and adjusts the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS;

[0035] The child intervention suggestion generation plan is generated in the following way;

[0036] When the child assessment status result status Status=0, the generation level abnormal standard AB=0, and the sending frequency of notifications and prompts, as well as the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS are not adjusted;

[0037] When the child assessment status result status Status=1, the abnormal level standard AB=2 is generated, the sending frequency of notifications and prompts is adjusted to 2 times of the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 2+1 times of the initial acquisition frequency;

[0038] When the child assessment status result status Status = 2, the generation level abnormality standard AB = 4, the sending frequency of notifications and prompts is adjusted to 4 times the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 4+2 times the initial acquisition frequency.

[0039] Preferably, the iterative optimization module includes an activity status updating unit and an intervention effect evaluation unit;

[0040] The activity status update unit performs fixed-period iterative evaluation of the child's activity status according to the child intervention suggestion generation plan, including fixed-period iterative evaluation of the child's activity status at a frequency of 2+1 times and fixed-period iterative evaluation of the child's activity status at a frequency of 4+2 times, generates a new activity status evaluation index ASN, and reflects the real-time activity status after the child intervention suggestion generation plan is executed.

[0041] Preferably, the intervention effect evaluation unit compares the activity state evaluation index ASN with the child activity state evaluation index AS to obtain the difference △AS, and judges the effectiveness of the child intervention suggestion generation plan through the difference △AS;

[0042] The effectiveness of the child intervention recommendation generation plan is determined by the following methods:

[0043] When the difference △AS ≥ 0, the child intervention suggestion generation plan is invalid, triggering the iterative optimization mechanism to adjust the content of the child intervention suggestion generation plan;

[0044] When the difference △AS is less than 0, the generation plan for obtaining child intervention suggestions is effective and the iterative optimization mechanism is not triggered.

[0045] The present invention provides a personalized activity monitoring and full-process management system for children's health care, which has the following beneficial effects:

[0046] (1) When the system is running, by obtaining the child activity status assessment index AS, it can reflect the child's activity status information in real time, solving the problem of delayed and highly subjective assessments based on questionnaires and interviews. The child activity status assessment index AS is compared with the health assessment status threshold T through the detection and trigger module to generate the grade abnormality standard AB, providing parents with timely and personalized adjustment guidance and avoiding adverse consequences caused by problems not being detected in time. In addition, the improvement of the child's activity status is evaluated within a fixed period and a new activity status assessment index ASN is generated, and the intervention suggestion generation plan is continuously optimized to ensure the effectiveness of the intervention measures. It can not only achieve accurate assessment of children's activities and psychological states in multiple dimensions, but also generate intervention suggestions in real time and continuously optimize them. These functions together make up for the shortcomings of the existing system in terms of real-time, comprehensiveness and personalization.

[0047] (2) By combining multi-dimensional information such as the average activity intensity AIm, the average sound signal energy SFe, and the proportion of smiling behaviors EFs, the child's activity status is dynamically fitted to reflect the overall intensity, rhythm fluctuation, and duration of the activity. By comparing the child's activity status assessment index AS with the health assessment status threshold T through the detection and trigger module, the system obtains the child's assessment status result Status in real time and generates corresponding prompts and suggestions. When the activity intensity SAI is low, the system can remind you to increase the outdoor activity time; when the activity continuity SAC is slightly insufficient, it can remind you to pay attention to the activity time arrangement; for abnormal conditions, the system not only generates a detailed abnormal report, but also can combine the performance of activity frequency fluctuation SFV to remind you to avoid fatigue accumulation. This comprehensive and dynamic activity status assessment mechanism can not only quickly discover children's potential problems in activity intensity, frequency fluctuation and continuity, but also provide accurate intervention suggestions based on real-time assessment results.

[0048] (3) According to the child's assessment status result Status, the abnormal level standard AB is automatically generated to adjust the sending frequency of notifications and prompts, as well as the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS, so as to dynamically match the actual needs of the child's health status. And use this to evaluate the effectiveness of the intervention suggestion generation plan. If the intervention is ineffective, the system will trigger the iterative optimization mechanism and automatically adjust the intervention content to ensure that the intervention measures can dynamically adapt to the actual status needs of the child. This optimization mechanism based on real-time feedback not only improves the scientificity and accuracy of the intervention plan, but also effectively reduces unnecessary intervention behaviors, while avoiding the accumulation of health risks, and provides highly adaptable dynamic optimization capabilities for children's activity health management. Compared with static intervention methods, this system has significant advantages in real-time response and continuous optimization, making children's health management more efficient and intelligent. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention is a schematic diagram of a block diagram of a personalized activity monitoring and full-process management system for children's health care. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0051] Example 1

[0052] The present invention provides a personalized activity monitoring and full-process management system for children's health care, please refer to Figure 1 , including data monitoring module, data processing and extraction module, activity evaluation module, detection and triggering module, suggestion generation module and iterative optimization module;

[0053] The data monitoring module collects real-time data of children through wearable devices to generate a children's activity feature vector set FV;

[0054] The data processing and extraction module performs preprocessing and feature extraction on the children's activity feature vector set FV to generate a standard feature vector set SFV;

[0055] The activity evaluation module evaluates the activity status information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and establishes an activity status evaluation algorithm formula for evaluation to obtain the child activity status evaluation index AS;

[0056] The detection and triggering module compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result, and generates the grade abnormality standard AB according to the child assessment state result;

[0057] The suggestion generation module generates a child intervention suggestion generation plan according to the abnormal grade standard AB, the activity state information and the inactivity state information, and provides prompts and notifications according to the content of the child intervention suggestion generation plan;

[0058] The iterative optimization module iteratively evaluates the child's activity status in a fixed period according to the child intervention suggestion generation plan, generates a new activity status evaluation index ASN, and compares it with the child activity status evaluation index AS to determine the effectiveness of the child intervention suggestion generation plan.

[0059] In this embodiment, by obtaining the child activity status assessment index AS, the child's activity status information can be reflected in real time, solving the problem of delayed and highly subjective evaluation by relying on questionnaires and interviews. The child activity status assessment index AS is compared with the health assessment status threshold T through the detection and triggering module to generate the grade abnormality standard AB, and the suggestion generation module combines the grade abnormality standard AB and other relevant information to generate a specific child intervention suggestion generation plan, providing parents with timely and personalized adjustment guidance, avoiding the adverse consequences caused by the problem not being detected in time. In addition, through the iterative optimization module, the improvement of the child's activity status is evaluated within a fixed period and a new activity status assessment index ASN is generated, and the intervention suggestion generation plan is continuously optimized to ensure the effectiveness of the intervention measures. Not only can the accurate evaluation of multi-dimensional children's activities and psychological states be achieved, but also intervention suggestions can be generated in real time and continuously optimized. These functions together make up for the shortcomings of the existing system in terms of real-time, comprehensiveness and personalization, provide a more scientific and intelligent solution for children's health management, and significantly improve the efficiency and effectiveness of children's activity status monitoring and mental health management.

[0060] Example 2

[0061] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the data monitoring module includes a data acquisition unit;

[0062] The data acquisition unit collects data of the child in real time through the wearable device, including using a three-axis acceleration sensor to record the acceleration change value per unit time, and obtaining the activity intensity AIt at time t; using a microphone to collect the child's voice signal SF; using a camera to obtain the child's facial video frame sequence, and obtaining the smile frequency EF within a fixed period; using a heart rate sensor and an acceleration sensor to collect the child's sleep information, obtain the resting heart rate HR and the number of turning over M, and generate a child activity feature vector set FV = {AIt, SF, EF, HR, M};

[0063] The activity intensity AIt is obtained by the following calculation formula:

[0064]

[0065] Where ax, ay and az represent the acceleration components in the horizontal, vertical and longitudinal directions respectively;

[0066] The speech signal SF is obtained by the following calculation formula:

[0067]

[0068] Where T represents the length of the sampling window, s(t) represents the amplitude of the sound signal at time t, |s(t)| represents the absolute value of the amplitude of the sound signal, which is used to ignore the positive and negative changes of the signal and only consider the intensity of the sound, and dt represents a small time increment, which is determined by the sampling frequency of the device;

[0069] The smile frequency EF is obtained by the following calculation formula:

[0070]

[0071] Where Nsmile represents the number of smiles, which is obtained by counting the facial expression recognition algorithm, and T represents the length of the sampling window.

[0072] The data processing and extraction module includes a preprocessing unit and a feature extraction unit;

[0073] The preprocessing unit preprocesses the children's activity feature vector set FV, including data cleaning preprocessing, data denoising preprocessing and data normalization preprocessing, to obtain the preprocessed feature vector set FVp;

[0074] Among them, the data cleaning preprocessing includes using the linear interpolation method to process missing values ​​of the feature vector set FV, using the box plot method to process outliers of the feature vector set FV, and then checking and processing outliers and missing values ​​in the feature vector set FV;

[0075] Data denoising preprocessing includes using a sliding average filter method to perform time series data smoothing on the feature vector set FV after data cleaning preprocessing, using a low-pass filter to filter high-frequency noise on the feature vector set FV after data cleaning preprocessing, and then checking and eliminating the noise caused by equipment and environmental interference in the feature vector set FV after data cleaning preprocessing;

[0076] The data normalization preprocessing includes using the minimum-maximum normalization method to normalize the feature vector set FV after the data denoising preprocessing, and then mapping the feature vector set FV after the data denoising preprocessing to the data range of 0 to 1 to eliminate the dimension difference. The data normalization preprocessing feature vector set FV = {AItn, SFn, EFn, HRn, Mn}.

[0077] The feature extraction unit performs feature extraction on the preprocessed feature vector set FVp to obtain activity intensity feature information, sound feature information, expression feature information and sleep feature information;

[0078] Among them, the activity intensity feature information includes the average activity intensity AIm and the activity fluctuation amplitude AIv; the sound feature information includes the sound signal energy average SFe and the sound signal peak frequency SFf; the expression feature information includes the smiling behavior ratio EFs; the sleep feature information includes the sleep score SDs, and the standard feature vector set SFV = {AIm, AIv, SFe, SFf, EFs, SDs} is generated;

[0079] The average activity intensity AIm is obtained by the following calculation formula:

[0080]

[0081] Where AItn,i represents the normalized activity intensity at the i-th moment, and N represents the total number of data points in the time window;

[0082] The activity fluctuation range AIv is Obtain calculation formula;

[0083] The sound signal energy average SFe is obtained by the following calculation formula:

[0084]

[0085] Where SFn,i represents the normalized speech signal at the i-th moment;

[0086] The sound signal peak frequency SFf is obtained by the following calculation formula:

[0087] SFf = argmax(FFT(SFn));

[0088] Wherein, argmax represents the maximum value function, which is specifically used to return the frequency component with the largest intensity in the frequency domain, and FFT(SFn) represents the fast Fourier transform result of the normalized speech signal, which specifically represents the intensity of each frequency component in the frequency domain;

[0089] The smiling behavior proportion EFs is obtained by the following calculation formula:

[0090]

[0091] Where EFn,i represents the smiling frequency at the i-th moment, Osmile represents the smiling behavior threshold, and || represents an exponential function, which returns 1 when the condition is true, otherwise it returns 0;

[0092] The sleep score SDs is obtained by the following calculation formula:

[0093] SDs=s1*(1-HRn)+s2*(1-Mn);

[0094] Wherein, HRn represents the normalized resting heart rate, Mn represents the normalized number of turns, s1 and s2 represent the preset weight values ​​of the normalized resting heart rate HRn and the normalized number of turns Mn, respectively, and s1+s2=1, and the specific value is set by the user.

[0095] In this embodiment, an efficient data processing link from raw data to a standard feature vector set SFV is established, making the acquisition of children's activities, emotions and sleep information more comprehensive, accurate and dynamic. With the help of multimodal devices such as three-axis accelerometers, microphones and cameras, real-time data including activity intensity AIt, voice signal SF, smile frequency EF and sleep information, resting heart rate HR and number of turning over M are collected to form a children's activity feature vector set FV. In the preprocessing stage, the system generates a normalized feature vector set FVp through data cleaning, denoising and normalization, ensuring the integrity, smoothness and consistency of the data, and effectively eliminating the interference caused by equipment noise and dimensional differences. Finally, through the feature extraction unit, a standard feature vector set SFV including average activity intensity AIm, activity fluctuation amplitude AIv, sound signal energy average SFe, sound signal peak frequency SFf, smile behavior proportion EFs and sleep score SDs is generated, providing high-quality multi-dimensional basic data for subsequent activity status evaluation. The system not only realizes the refined monitoring of children's activity status and related characteristics, but also significantly improves the reliability and availability of data, especially in dynamic environments, ensuring the real-time and accuracy of monitoring results, and providing more scientific data support for children's health assessment and behavior analysis.

[0096] Example 3

[0097] This embodiment is explained in Example 2. Please refer to Figure 1 ,Specifically: the activity evaluation module includes a state evaluation unit and an evaluation index unit;

[0098] The state evaluation unit evaluates the activity state information and inactivity state information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and obtains activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information;

[0099] The activity intensity is fitted by the average activity intensity AIm, the average sound signal energy SFe and the smiling behavior proportion EFs, and is marked as activity intensity SAI, which reflects the activity intensity of children. The greater the value, the activity in a high emotional state will further improve the score.

[0100] The frequency fluctuation is fitted by the average intensity of activity AIm, the peak frequency of sound signal SFf and the proportion of smiling behavior EFs, and is marked as frequency fluctuation SFV, which reflects the frequency fluctuation of children's activities. The more intense and the stronger the rhythm, the lower the value, which means the activity fluctuation is smaller and the rhythm is more stable;

[0101] The activity duration is fitted by the sleep score SDs and the activity fluctuation amplitude AIv, and is marked as activity duration SAC;

[0102] The activity intensity SAI is Obtain calculation formula;

[0103] The frequency fluctuation SFV is obtained by the calculation formula SFV=ln(1+AIm*SFf)*(1+EFs), where ln represents a logarithmic function;

[0104] The activity is sustained by SAC The calculation formula is obtained, where e represents an exponential function, which is specifically used to map the activity fluctuation amplitude AIv to the range of 0 to 1.

[0105] The evaluation index unit evaluates the activity intensity SAI, frequency fluctuation SFV and activity duration SAC based on the acquired activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information by establishing an activity state evaluation algorithm formula to reflect the comprehensive performance of activity intensity, frequency fluctuation and duration in children's activities, and obtain the children's activity state evaluation index AS;

[0106] The Children's Activity Status Assessment Index AS is The calculation formula is obtained, where e represents the exponential function.

[0107] The detection and triggering module includes an anomaly detection unit and an intervention suggestion generation unit;

[0108] The abnormality detection unit compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result;

[0109] The child assessment status results are obtained by comparing:

[0110] When the child activity status assessment index AS is less than the health assessment status threshold T, the child assessment status result is obtained as a qualified status, marked as Status = 0, indicating that the child's activity status is normal and no special intervention is required, and the child should continue to maintain his activity habits and work and rest schedule;

[0111] When the child's activity status assessment index AS ≥ the health assessment status threshold T, the child's assessment status result is obtained as an unqualified state, marked as Status = 1, indicating that the child's activity status is unqualified and it is necessary to adjust the activity arrangement and activity content, and adjust the activity habits and work and rest according to the activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information. When the activity intensity SAI is low, it is recommended to increase the outdoor activity time. When the frequency fluctuation SFV is normal, but the activity continuity SAC is slightly insufficient, it is necessary to pay attention to the child's activity time arrangement;

[0112] When the child's activity status assessment index AS ≥ twice the health assessment status threshold T, the child's assessment status result is abnormal and marked as Status = 2, indicating that the child's activity is abnormal. Parents or managers are asked to adjust the activity content and the child's rest time. There is an abnormal activity intensity, and an abnormal report is generated based on the activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information, including when the activity intensity SAI is low, it may indicate insufficient activity, when the frequency fluctuation SFV is high, it may lead to fatigue accumulation, and when the activity continuity SAC performance is poor, attention should be paid to the child's work and rest schedule.

[0113] In this embodiment, by combining multi-dimensional information such as the average intensity of activity AIm, the average energy of sound signal SFe, and the proportion of smiling behavior EFs, the child's activity state is dynamically fitted to reflect the overall intensity of the activity, the rhythm fluctuation and the balanced performance of the duration. Through the comparison of the child's activity state assessment index AS and the health assessment state threshold T by the detection and trigger module, the system obtains the child's assessment state result Status in real time and generates corresponding prompts and suggestions. When the activity intensity SAI is low, the system can remind you to increase the outdoor activity time; when the activity continuity SAC is slightly insufficient, you can remind you to pay attention to the activity time arrangement; for abnormal states, the system not only generates a detailed abnormal report, but also can remind you to avoid fatigue accumulation in combination with the performance of activity frequency fluctuation SFV. This comprehensive and dynamic activity state assessment mechanism can not only quickly discover children's potential problems in activity intensity, frequency fluctuation and continuity, but also provide accurate intervention suggestions based on real-time assessment results, helping parents and managers to efficiently manage children's activity habits and work and rest arrangements. Compared with the traditional single-dimensional monitoring method, the system realizes the deep linkage of activity status and intervention measures, providing an efficient, scientific and intelligent solution for children's activity health management.

[0114] Example 4

[0115] This embodiment is explained in Example 3, please refer to Figure 1Specifically: the intervention suggestion generation unit automatically generates a grade abnormality standard AB according to the child assessment status result, and generates a child intervention suggestion generation plan according to the grade abnormality standard AB, adjusts the frequency of notification and prompt of the child assessment status result, and adjusts the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS;

[0116] The child intervention suggestion generation plan is generated in the following way;

[0117] When the child assessment status result status Status=0, the generation level abnormal standard AB=0, and the sending frequency of notifications and prompts, as well as the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS are not adjusted;

[0118] When the child assessment status result status Status=1, the abnormal level standard AB=2 is generated, the sending frequency of notifications and prompts is adjusted to 2 times of the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 2+1 times of the initial acquisition frequency;

[0119] When the child assessment status result status Status = 2, the generation level abnormality standard AB = 4, the sending frequency of notifications and prompts is adjusted to 4 times the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 4+2 times the initial acquisition frequency.

[0120] The iterative optimization module includes an activity status updating unit and an intervention effect evaluation unit;

[0121] The activity status update unit performs fixed-period iterative evaluation of the child's activity status according to the child intervention suggestion generation plan, including fixed-period iterative evaluation of the child's activity status at a frequency of 2+1 times and fixed-period iterative evaluation of the child's activity status at a frequency of 4+2 times, generates a new activity status evaluation index ASN, and reflects the real-time activity status after the child intervention suggestion generation plan is executed.

[0122] The intervention effect evaluation unit compares the activity state evaluation index ASN with the child activity state evaluation index AS to obtain the difference △AS, and judges the effectiveness of the child intervention suggestion generation plan through the difference △AS;

[0123] The effectiveness of the child intervention recommendation generation plan is determined by the following methods:

[0124] When the difference △AS ≥ 0, the child intervention suggestion generation plan is invalid, triggering the iterative optimization mechanism to adjust the content of the child intervention suggestion generation plan;

[0125] When the difference △AS is less than 0, the generation plan for obtaining child intervention suggestions is effective and the iterative optimization mechanism is not triggered.

[0126] In this embodiment, the intervention suggestion generation unit automatically generates the abnormal level standard AB according to the child assessment status result Status, adjusts the sending frequency of notifications and prompts, and the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS, so as to dynamically match the actual needs of the child's health status. For unqualified and abnormal states, the system performs fixed-cycle activity status evaluation at 2+1 times the frequency and 4+2 times the frequency, respectively, and generates a new activity status evaluation index ASN to reflect the real-time activity status after the intervention plan is executed. In the iterative optimization module, the intervention effect evaluation unit calculates the difference ΔAS by comparing the new activity status evaluation index ASN with the original activity status evaluation index AS, and uses this to evaluate the effectiveness of the intervention suggestion generation plan. If the intervention is ineffective, the system will trigger the iterative optimization mechanism, automatically adjust the intervention content, and ensure that the intervention measures can dynamically adapt to the actual status needs of the child. This optimization mechanism based on real-time feedback not only improves the scientificity and accuracy of the intervention plan, but also effectively reduces unnecessary intervention behaviors, while avoiding the accumulation of health risks, and provides a highly adaptable dynamic optimization capability for children's activity health management. Compared with static intervention methods, this system has significant advantages in real-time response and continuous optimization, making children's health management more efficient and intelligent.

[0127] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A personalized activity monitoring and full-process management system for children's health care, characterized by: It includes data monitoring module, data processing and extraction module, activity evaluation module, detection and triggering module, suggestion generation module and iterative optimization module; The data monitoring module collects real-time data of children through wearable devices to generate a children's activity feature vector set FV; The data processing and extraction module performs preprocessing and feature extraction on the children's activity feature vector set FV to generate a standard feature vector set SFV; The activity evaluation module evaluates the activity status information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and establishes an activity status evaluation algorithm formula for evaluation to obtain the child activity status evaluation index AS; The detection and triggering module compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result, and generates the grade abnormality standard AB according to the child assessment state result; The suggestion generation module generates a child intervention suggestion generation plan according to the abnormal grade standard AB, the activity state information and the inactivity state information, and provides prompts and notifications according to the content of the child intervention suggestion generation plan; The iterative optimization module iteratively evaluates the child's activity status in a fixed period according to the child intervention suggestion generation plan, generates a new activity status evaluation index ASN, and compares it with the child activity status evaluation index AS to determine the effectiveness of the child intervention suggestion generation plan.

2. The personalized activity monitoring and full-process management system for children's health care according to claim 1, characterized in that: The data monitoring module includes a data acquisition unit; The data acquisition unit performs real-time data acquisition on the child through a wearable device, including using a three-axis acceleration sensor to record the acceleration change value per unit time to obtain the activity intensity AIt at time t; using a microphone to collect the child's voice signal SF; using a camera to obtain the child's facial video frame sequence to obtain the smile frequency EF within a fixed period; using a heart rate sensor and an acceleration sensor to collect the child's sleep information, obtain the resting heart rate HR and the number of turning over M, and generate a child activity feature vector set FV={AIt, SF, EF, HR, M}.

3. The personalized activity monitoring and full-process management system for children's health care according to claim 2, characterized in that: The data processing and extraction module includes a preprocessing unit and a feature extraction unit; The preprocessing unit preprocesses the children's activity feature vector set FV, including data cleaning preprocessing, data denoising preprocessing and data normalization preprocessing, to obtain the preprocessed feature vector set FVp; Among them, the data cleaning preprocessing includes using the linear interpolation method to process missing values ​​of the feature vector set FV, using the box plot method to process outliers of the feature vector set FV, and then checking and processing outliers and missing values ​​in the feature vector set FV; Data denoising preprocessing includes using a sliding average filter method to perform time series data smoothing on the feature vector set FV after data cleaning preprocessing, using a low-pass filter to filter high-frequency noise on the feature vector set FV after data cleaning preprocessing, and then checking and eliminating the noise caused by equipment and environmental interference in the feature vector set FV after data cleaning preprocessing; The data normalization preprocessing includes using the minimum-maximum normalization method to normalize the feature vector set FV after data denoising preprocessing, and then mapping the feature vector set FV after data denoising preprocessing to a data range of 0 to 1 to eliminate dimensional differences. The data normalization preprocessing feature vector set FV={AItn, SFn, EFn, HRn, Mn}.

4. The personalized activity monitoring and full-process management system for children's health care according to claim 3, characterized in that: The feature extraction unit performs feature extraction on the preprocessed feature vector set FVp to obtain activity intensity feature information, sound feature information, expression feature information and sleep feature information; Among them, the activity intensity feature information includes the average activity intensity AIm and the activity fluctuation amplitude AIv; the sound feature information includes the sound signal energy average SFe and the sound signal peak frequency SFf; the expression feature information includes the smiling behavior proportion EFs; the sleep feature information includes the sleep score SDs, generating a standard feature vector set SFV={AIm, AIv, SFe, SFf, EFs, SDs}.

5. The personalized activity monitoring and full-process management system for children's health care according to claim 3, characterized in that: The activity evaluation module includes a state evaluation unit and an evaluation index unit; The state evaluation unit evaluates the activity state information and inactivity state information of the child on the standard feature vector set SFV, including activity intensity, frequency fluctuation and activity continuity, and obtains activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information; The activity intensity is fitted by the average activity intensity AIm, the average sound signal energy SFe and the smiling behavior proportion EFs, and is marked as activity intensity SAI, which reflects the activity intensity of children; The frequency fluctuation is fitted by the average activity intensity AIm, the peak frequency SFf of the sound signal and the proportion of smiling behavior EFs, and is marked as frequency fluctuation SFV, which reflects the frequency fluctuation of children's activities; The activity duration is fitted by the sleep score SDs and the activity fluctuation amplitude AIv, and is marked as activity duration SAC.

6. The personalized activity monitoring and full-process management system for children's health care according to claim 5, characterized in that: The evaluation index unit evaluates the activity intensity SAI, frequency fluctuation SFV and activity duration SAC based on the acquired activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information by establishing an activity state evaluation algorithm formula to reflect the comprehensive performance of activity intensity, frequency fluctuation and duration in children's activities, and obtain the children's activity state evaluation index AS; The Children's Activity Status Assessment Index AS is The calculation formula is obtained, where e represents the exponential function.

7. The personalized activity monitoring and full-process management system for children's health care according to claim 1, characterized in that: The detection and triggering module includes an anomaly detection unit and an intervention suggestion generation unit; The abnormality detection unit compares the preset health assessment state threshold T with the child activity state assessment index AS to obtain the child assessment state result; The child assessment status results are obtained by comparing: When the child activity status assessment index AS is less than the health assessment status threshold T, the child assessment status result is obtained as a qualified status, marked as Status=0, indicating that the child's activity status is normal and no special intervention is required. The child should continue to maintain his activity habits and work and rest schedule. When the child's activity status assessment index AS ≥ the health assessment status threshold T, the child's assessment status result is obtained as unqualified, marked as Status = 1, indicating that the child's activity status is unqualified and the activity arrangement and content need to be adjusted, and the activity habits and work and rest schedule should be adjusted according to the activity intensity SAI, frequency fluctuation SFV and activity continuity SAC marking information; When the child activity status assessment index AS ≥ twice the health assessment status threshold T, the child assessment status result is abnormal and marked as Status=2, indicating that the child's activity is abnormal. Parents or managers are asked to adjust the activity content and the child's rest time. If the activity intensity is abnormal, an abnormal report is generated based on the activity intensity SAI, frequency fluctuation SFV and activity duration SAC marking information.

8. The personalized activity monitoring and full-process management system for children's health care according to claim 7, characterized in that: The intervention suggestion generation unit automatically generates a grade abnormality standard AB according to the child assessment status result, and generates a child intervention suggestion generation plan according to the grade abnormality standard AB, adjusts the frequency of notification and prompt of the child assessment status result, and adjusts the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS; The child intervention suggestion generation plan is generated in the following way; When the child assessment status result status Status=0, the level abnormality standard AB=0 is generated, and the sending frequency of notifications and prompts, as well as the acquisition frequency of the standard feature vector set SFV and the child activity status assessment index AS are not adjusted; When the child assessment status result status Status=1, the abnormal level standard AB=2 is generated, the sending frequency of notifications and prompts is adjusted to 2 times the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 2+1 times the initial acquisition frequency; When the child assessment status result status Status=2, the level abnormality standard AB=4 is generated, the sending frequency of notifications and prompts is adjusted to 4 times the initial frequency, and the standard feature vector set SFV and the child activity status assessment index AS are adjusted to 4+2 times the initial acquisition frequency.

9. The personalized activity monitoring and full-process management system for children's health care according to claim 1, characterized in that: The iterative optimization module includes an activity status updating unit and an intervention effect evaluation unit; The activity status update unit performs fixed-period iterative evaluation of the child's activity status according to the child intervention suggestion generation plan, including fixed-period iterative evaluation of the child's activity status at a frequency of 2+1 times and fixed-period iterative evaluation of the child's activity status at a frequency of 4+2 times, generates a new activity status evaluation index ASN, and reflects the real-time activity status after the child intervention suggestion generation plan is executed.

10. A personalized activity monitoring and full-process management system for children's health care according to claim 9, characterized in that: The intervention effect evaluation unit compares the activity state evaluation index ASN with the child activity state evaluation index AS to obtain the difference △AS, and judges the effectiveness of the child intervention suggestion generation plan through the difference △AS; The effectiveness of the child intervention recommendation generation plan is determined by the following methods: When the difference △AS ≥ 0, the child intervention suggestion generation plan is invalid, triggering the iterative optimization mechanism to adjust the content of the child intervention suggestion generation plan; When the difference △AS is less than 0, the generation plan for obtaining child intervention suggestions is effective and the iterative optimization mechanism is not triggered.

Images