Child growth multi-dimensional interaction evaluation system based on intelligent agent terminal

Abstract

The invention relates to the technical field of health-related information systems, and discloses a child growth multi-dimensional interaction evaluation system based on an intelligent agent terminal, which comprises a multi-source data acquisition terminal, a data processing module, a health suggestion generation module and a real-time feedback module, the data processing module, the health suggestion generation module and the real-time feedback module are integrated in an intelligent agent terminal, and the intelligent agent terminal is in communication connection with the multi-source data acquisition terminal. According to the application, through data acquisition of four dimensions of movement, physiology, cognition and environment, a holographic figure of child growth is constructed, body functions, psychological states and environmental influences can be reflected at the same time, a multi-dimensional data basis is realized, and a traditional Chinese medicine theory model is combined with a traditional Chinese medicine logic of symptom collection-symptom type analysis-symptomatic intervention, so that the development of children is realized. Therefore, health suggestions are more based on traditional Chinese medicine theories, comprehensiveness, accuracy and real-time performance of child growth evaluation are improved, and a reliable technology is provided for the field of intelligent education and health management.

Description

Technical Field

[0001] This application relates to the field of health-related information system technology, specifically a multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal. Background Technology

[0002] With rapid economic development, a general improvement in population quality, and the popularization of scientific parenting concepts, parents are paying close attention to their children's growth. On the other hand, children are physically vulnerable and often lack adequate health awareness during play. Therefore, utilizing smart devices for multi-dimensional interactive assessments of children's development is particularly important.

[0003] Current technologies focus on wearable health monitoring devices and educational / entertainment robots, but their data dimensions are limited and cannot comprehensively reflect children's developmental status. They lack multimodal data fusion, are not integrated with educational or entertainment scenarios, and are functionally isolated. Health assessments lack medical theoretical support and have poor real-time performance, failing to meet the immediate interactive needs of playgrounds. Summary of the Invention

[0004] The purpose of this application is to provide a multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal, so as to solve the technical problems mentioned in the background art.

[0005] To achieve the above objectives, this application discloses the following technical solution: a multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal, comprising: The multi-source data acquisition terminal includes: a motion data acquisition device configured to collect children's motion data, a physiological data acquisition device configured to collect children's physiological data, a cognitive data acquisition device configured to collect children's cognitive data, and an environmental monitoring device configured to collect environmental data. The data processing module is configured to: receive multi-source data uploaded in real time by the multi-source data acquisition terminal, and perform standardization processing and multimodal fusion analysis on the acquired multi-source data, wherein the standardization processing converts indicators of different dimensions into normalized values ​​based on preset rules; The health advice generation module is configured to generate personalized health advice based on the output of the data processing module and combined with a traditional Chinese medicine theoretical model. The health advice includes an assessment of traditional Chinese medicine constitution parameters and corresponding behavioral intervention strategies. The real-time feedback module is configured to: calculate a health status score based on a preset health status scoring rule, and provide graded early warnings for the child's health status based on the health status score; The data processing module, the health suggestion generation module, and the real-time feedback module are integrated in the intelligent agent terminal, which is communicatively connected to the multi-source data acquisition terminal.

[0006] Preferably, the standardization process includes the normalization of physiological indicators, which is based on preset medical standards or statistical models. The physiological indicators include: height or weight, dynamic heart rate, body temperature and blood oxygen.

[0007] Preferably, in the aforementioned standardization process: The normalization of height or weight was performed by combining the BMI index with the WHO standard percentile corresponding to the child's age and sex. Standardize, standardized value The calculation rule is: when percentiles hour, ;when hour, ;when hour, ; The normalization process for dynamic heart rate includes: calculating the degree to which the current heart rate deviates from the ideal heart rate based on the age-appropriate resting heart rate standard; Normalization of body temperature includes: standardization based on the normal body temperature range, and standardized values. The calculation rule is: when body temperature or hour, ;otherwise, ; Normalization of blood oxygen includes standardization based on the normal threshold for blood oxygen saturation.

[0008] Preferably, the behavioral intervention strategy includes: When the climbing behavior lasts for more than 3 minutes per instance, it is determined that the activity of the liver meridian has increased, and a suggestion to read a calming picture book is sent. When the manual operation time is less than 2 minutes, the spleen deficiency index is determined to rise, triggering tactile feedback for intervention.

[0009] Preferably, the health status score is calculated using the following formula: The calculation formula is: in, Standardized values ​​for height or weight. This is the standardized value of heart rate. This is the standardized value of body temperature. This is the standardized value of blood oxygen. , , and These are the weighting coefficients.

[0010] Preferably, the real-time feedback module is further configured to: calculate the suitability score of the puzzle item based on heart rate variability, blood oxygen, motor coordination, and overall health status. Suitable rating for brain-training projects The calculation formula is: in, Weighting for brain-training projects. As a basic weight for health, This is a standardized value for motor coordination. Preferably, the graded early warning specifically includes: When the health status score is calculated as follows At that time, a red alert is triggered; When the health status score is calculated as follows At that time, a yellow alert was triggered; When the health status score is calculated as follows When this time is reached, it indicates that the person is in a normal health condition.

[0011] Preferably, the multimodal fusion analysis specifically includes: Map real-time scene labels to scene embedding vectors Simultaneously, the standardized values ​​of motion, physiological, and cognitive data are encoded into modal feature vectors. ,in, For feature dimensions; Based on a multi-head self-attention architecture, a scene-guided modality weight generator is constructed, with the following formula: in, , Represents motion modes, Represents physiological modalities. Representing cognitive modalities; , and The weight matrix is ​​a learnable matrix; Let be the dynamic attention weights for the i-th modality at time t, and ; This is the fused multimodal feature vector; Embed vectors for real-time scene features; Let i be the feature vector of the i-th modality data; Query vectors guided by the scenario; Let be the key vector of the i-th mode; Let i be the value vector of the i-th mode; The correlation score between the scene and the i-th modality; In the amusement park scenario, motion modality weights Increase to over 60%, prioritizing the integration of gait balance and climbing frequency data; in classroom settings, cognitive modality weights Increase it to over 50%, focusing on analyzing participation and attention span in brain-training projects.

[0012] Preferably, the intelligent terminal is also configured to act as a customer service representative, interact with users in real time, and collect questions of concern to parents.

[0013] Preferably, a growth path planning module is also included, wherein the growth path planning module is configured as follows: Based on historical assessment data, a child's individual developmental trajectory model is constructed to predict growth trends over the next 3-6 months; Personalized intervention program sequences are generated using reinforcement learning algorithms; The outputs include phased goals related to movement, cognition, and environmental regulation.

[0014] Beneficial Effects: The multi-dimensional interactive assessment system for children's growth based on intelligent terminal in this application constructs a holographic profile of children's growth by collecting data from four dimensions: movement, physiology, cognition, and environment. It can simultaneously reflect physical function, psychological state, and environmental influences, achieving a multi-dimensional data foundation. Based on the TCM theoretical model and combined with the traditional TCM logic of symptom collection, syndrome analysis, and symptomatic intervention, it makes health recommendations more grounded in TCM theory, improves the comprehensiveness, accuracy, and real-time nature of children's growth assessment, and provides a reliable technology for the fields of intelligent education and health management. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a structural block diagram of a multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal, provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0018] In this document, the term "comprising" is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0019] The limited data dimensions hinder a comprehensive assessment of children's growth. Children's growth is a complex process, influenced by a combination of physical, psychological, and socio-emotional factors. A single data dimension restricts a comprehensive assessment of children's growth on multiple levels. Children's growth is a dynamic process, with different developmental priorities and characteristics at different stages. Single-dimensional data can only reflect the situation at a specific moment and cannot present the continuity and phased changes in children's growth. Secondly, poor real-time performance impacts the playground's need for instant interaction. Playgrounds are places that require rapid response to visitor needs and the provision of immediate services. Poor real-time performance negatively impacts the playground's instant interaction in several ways: cloud data platforms require highly reliable infrastructure and data backup and recovery mechanisms, and also require computation time.

[0020] Based on this, this embodiment provides a method such as Figure 1 The illustrated multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal aims to provide an innovative solution for the fields of intelligent education and health management. Specifically, the system includes: The multi-source data acquisition terminal includes: a motion data acquisition device configured to collect children's motion data, a physiological data acquisition device configured to collect children's physiological data, a cognitive data acquisition device configured to collect children's cognitive data, and an environmental monitoring device configured to collect environmental data. The data processing module is configured to: receive multi-source data uploaded in real time by the multi-source data acquisition terminal, and perform standardization processing and multimodal fusion analysis on the acquired multi-source data, wherein the standardization processing converts indicators of different dimensions into normalized values ​​based on preset rules; The health advice generation module is configured to generate personalized health advice based on the output of the data processing module and combined with the TCM theoretical model. The health advice includes TCM constitution parameter assessment (such as spleen deficiency index and liver meridian activity) and corresponding behavioral intervention strategies. The real-time feedback module is configured to: calculate a health status score based on a preset health status scoring rule, and provide graded early warnings for the child's health status based on the health status score; and trigger a three-level response mechanism (red alert, yellow alert, green normal).

[0021] The data processing module, the health suggestion generation module, and the real-time feedback module are integrated in the intelligent agent terminal, which is communicatively connected to the multi-source data acquisition terminal.

[0022] Based on the above, this system, through a traditional Chinese medicine (TCM) theoretical model, collects and summarizes symptoms, analyzes and summarizes syndrome patterns, and provides targeted treatment. Leveraging the powerful data processing and analysis capabilities of an AI agent, it comprehensively analyzes the collected multi-source data (such as exercise data, physiological data, and environmental data) to accurately analyze and summarize syndrome patterns and provide personalized professional advice. It integrates traditional TCM theory with modern artificial intelligence technology, aiming to improve the accuracy and efficiency of TCM constitution identification and provide a scientific basis for personal health management.

[0023] Feasible methods include, but are not limited to, wearable devices (such as wristbands), environmental monitoring devices (such as thermometers and decibel meters), and interactive devices for educational projects, which respectively collect motion data (gait balance, motor coordination), environmental data (comfort, safety warnings), and cognitive data (project participation, liking).

[0024] In one implementation, the standardization process includes the normalization of physiological indicators, which is based on preset medical standards or statistical models. The physiological indicators include: height or weight, dynamic heart rate, body temperature, and blood oxygen.

[0025] Furthermore, in the aforementioned standardization process: The normalization of height or weight was performed by combining the BMI index with the WHO standard percentile corresponding to the child's age and sex. Standardize, standardized value The calculation rule is: when percentiles hour, (underweight); when hour, ;when hour, (Overweight); where BMI is calculated using the following formula: The normalization of dynamic heart rate includes: based on the age-corresponding resting heart rate standard (such as the content published in the existing technology: "Mahon AD, et al. (2003). Exercise Training and Heart Rate Variability in Children. Medicine & Science in Sports & Exercise, 35(5), 818-821.", the corresponding conclusion is: the maximum heart rate (MaxHR) during exercise for children is ≈ 208 - (0.7 × age), and the heart rate during moderate-intensity exercise should be maintained at 60-80% of MaxHR), calculating the degree to which the current heart rate deviates from the ideal heart rate; the formula for calculating the standardized value SHR is: in: Current heart rate, For ideal heart rate, and These are the normal range boundary values ​​for the corresponding age group; Normalization of body temperature includes standardization based on the normal body temperature range (36.5℃~37.5℃), and the standardized values ​​are... The calculation rule is: when body temperature or hour, ;otherwise, ; Normalization of blood oxygen includes: based on blood oxygen saturation (… Normal threshold ( Standardize the values; standardized values The calculation formula is: (Results are truncated to the 0-1 interval).

[0026] In one implementation, the behavioral intervention strategy includes: When the climbing behavior lasts for more than 3 minutes per instance, it is determined that the activity of the liver meridian has increased, and a suggestion to read a calming picture book is sent. When the manual operation time is less than 2 minutes, the spleen deficiency index is determined to rise, triggering tactile feedback for intervention.

[0027] In one implementation, the health status score is calculated using the following formula: The calculation formula is: in, Standardized values ​​for height or weight. This is the standardized value of heart rate. This is the standardized value of body temperature. This is the standardized value of blood oxygen. , , and These are weighting coefficients, which can be adjusted as needed. A feasible approach is... , , , .

[0028] Furthermore, the real-time feedback module is also configured to: calculate the suitability score of the puzzle item based on heart rate variability, blood oxygen, motor coordination, and overall health status. Suitable rating for brain-training projects The calculation formula is: in, Weighting for brain-training projects. As a basic weight for health, This represents the standardized value for motion coordination. Motion coordination is collected through motion capture devices (such as inertial sensors and cameras) in a multi-source data acquisition terminal. Specific measurement indicators include: Gait balance parameters: stride uniformity, center of gravity shift, number of falls; Motion coordination parameters: climbing / jump completion time, manual operation accuracy, and specified trajectory tracking error. Standardized values ​​for motion coordination are obtained through the following steps: Collect children's gait balance data (such as the frequency of center of gravity shift) and motor coordination task data (such as climbing time); Based on developmental standards for the same age group (such as WHO percentile values ​​for children’s motor abilities), each indicator is normalized to the 0-1 range; The weighted aggregate index is calculated using the following formula: in, This is a standardized value for gait balance, a balance index calculated based on gait data (such as a normalized value of the number of times the center of gravity shifts per unit distance). The standardized value for motion coordination tasks is based on the normalized value of the completion efficiency of tasks such as climbing and manual labor. For motion coordination weights, such as , For gait balance weights, such as .

[0029] By integrating multi-dimensional motion data and normalizing it according to developmental standards, the problem of lack of quantitative basis for assessing motor ability in existing technologies has been solved, making the suitability score of intellectual development projects more accurately reflect children's actual abilities.

[0030] Furthermore, the system's three-level early warning configuration is as follows: When the health status score is calculated as follows At that time, a red alert is triggered; When the health status score is calculated as follows At that time, a yellow alert was triggered; When the health status score is calculated as follows When this time is reached, it indicates that the person is in a normal health condition.

[0031] In one implementation, the multimodal fusion analysis specifically includes: Map real-time scene labels (such as "playground mode" and "classroom mode") to scene embedding vectors. Simultaneously, the standardized values ​​of motion, physiological, and cognitive data are encoded into modal feature vectors. ,in, For feature dimensions; Based on a multi-head self-attention architecture, a scene-guided modality weight generator is constructed, with the following formula: in, , Represents motion modes, Represents physiological modalities. Representing cognitive modalities; , and The weight matrix is ​​learnable and is obtained by: initialization - random normal distribution, optimization - learning and updating on the training data through backpropagation algorithm (such as Adam optimizer); Let be the dynamic attention weights for the i-th modality at time t, and ; This is the fused multimodal feature vector; The real-time scene feature embedding vector is obtained by: scene labels (such as "playground" and "classroom") being generated by one-hot encoding and then by an embedding layer, for example by extracting visual scene features through a pre-trained scene classification model (such as a modified ResNet); The feature vector for the i-th modality is obtained as follows: the original data of each modality (such as gait balance value, heart rate, and cognitive participation) is standardized and encoded through a fully connected layer (FC) as follows. dimensional vector; Query vectors guided by the scenario; Let be the key vector of the i-th mode; Let i be the value vector of the i-th mode; The correlation score between the scene and the i-th modality; In the amusement park scenario, motion modality weights Increase to over 60%, prioritizing the integration of gait balance and climbing frequency data; in classroom settings, cognitive modality weights Increase it to over 50%, focusing on analyzing participation and attention span in brain-training projects.

[0032] Compared to traditional fuzzy logic controllers, the Transformer-based attention mechanism can automatically learn scene-modal association patterns through end-to-end training, with a weight update frequency down to the millisecond level (Δt≤100ms). It also supports expansion to more than eight modal data fusions, solving the problems of lagging weight adjustment and insufficient modal scalability in existing technologies. It achieves scene semantic-driven dynamic weight allocation, providing more accurate assessment support for multidimensional and precise evaluation of children's development.

[0033] In one implementation, the intelligent terminal is further configured to: act as a customer service representative, interact with users in real time and collect questions of concern to parents, and optimize and expand functions based on the questions of concern to parents.

[0034] In one embodiment, the system further includes a growth path planning module, which is configured to: Based on historical assessment data, a child's individual developmental trajectory model is constructed to predict growth trends over the next 3-6 months; Personalized intervention program sequences are generated using reinforcement learning algorithms; Outputs include phased goals for movement, cognition, and environmental adjustment (such as "focusing on increasing climbing training to 5 minutes / session in the next 2 weeks").

[0035] Specifically, the tasks of the growth path planning module include: (1) Multidimensional developmental trajectory modeling A developmental trajectory model for children is constructed based on Long Short-Term Memory (LSTM) networks. The expression for the developmental trajectory model is as follows: in, Predict the state vector at time t (including health score H, cognitive index E, and cumulative environmental impact value). ); Historical evaluation data for the past n periods (time resolution in weeks / months); We propose an age-sex prior mean function based on WHO child growth standards to address the balance between individual differences and group standards.

[0036] What is feasible is the cumulative environmental impact value. The acquisition of, specifically includes: Collect time-series data (time resolution ≤ 1 minute) of environmental parameters such as light intensity, air quality, and noise. Construct a model for the cumulative effects of environmental exposure, using the following formula: in, For parameter weights, The attenuation coefficient is... These are the environmental parameters at time t.

[0037] Based on the cumulative environmental impact value, joint environmental-health intervention recommendations are generated (such as adjusting the type of activity in high-noise environments).

[0038] (2) Generate personalized intervention program sequences using reinforcement learning algorithms. The expression for the intervention program sequence is: in, The optimal strategy is... This is the current state. Let the growth return function be the formula for calculating the growth return function: in, For health benefits, For cognitive development, To adapt to the environment, Score the health status at time t. This is a truncation function, restricting the range of values ​​to 1. ， For indicator functions, Rate the suitability of the puzzle activity at time t+1. Baseline level, This represents the maximum value of the cognitive ability theory. To assess project completion, This represents the cumulative environmental impact value at time t+1. The optimal environmental exposure value, This represents the maximum permissible environmental exposure value.

[0039] By synergizing the three-dimensional reward functions, a comprehensive intervention can be achieved that is based on physiological health, centered on cognitive development, and ensured by environmental adaptation, thus ensuring a balanced growth path through multi-dimensional assessment.

[0040] In summary, the multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal in this embodiment has the following technical features: (1) By using algorithms, different dimensional indicators such as height / weight, heart rate, and body temperature are normalized to standardized values ​​of 0-1 (such as SBMI and SHR) to solve the problem of inconsistent data units; (2) Based on the generation of health status scores, combined with a three-level response mechanism (red alert / yellow warning / green normal), a graded early warning system for health risks is implemented. (3) The intelligent terminal acts as a "customer service representative", collecting questions that parents are concerned about in real time (such as "how to improve children's concentration") and generating customized educational suggestions by combining cognitive data (such as participation in educational projects); (4) Long-term accumulation of multi-dimensional data can analyze the continuous changes in children's growth (such as quarterly BMI trends and cognitive ability progression curves), provide parents with phased growth reports, and assist in the formulation of long-term development plans; (5) Integrating artificial intelligence (data processing algorithms), intelligent education (adaptation of educational projects), and health monitoring (analysis of physiological indicators) to form an interdisciplinary solution, improving the comprehensiveness, accuracy, and real-time nature of children's growth assessment, and providing reliable support for the fields of intelligent education and health management.

[0041] In the embodiments provided in this application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic units designed to implement the functions described herein, or combinations thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the associated hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or code on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media may be any available medium accessible to a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code having the form of instructions or data structures and accessible to a computer.

[0042] Finally, it should be noted that the above description is only a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal, characterized in that, include: The multi-source data acquisition terminal includes: a motion data acquisition device configured to collect children's motion data, a physiological data acquisition device configured to collect children's physiological data, a cognitive data acquisition device configured to collect children's cognitive data, and an environmental monitoring device configured to collect environmental data. The data processing module is configured to: receive multi-source data uploaded in real time by the multi-source data acquisition terminal, and perform standardization processing and multimodal fusion analysis on the acquired multi-source data, wherein the standardization processing converts indicators of different dimensions into normalized values ​​based on preset rules; The health advice generation module is configured to generate personalized health advice based on the output of the data processing module and combined with a traditional Chinese medicine theoretical model. The health advice includes an assessment of traditional Chinese medicine constitution parameters and corresponding behavioral intervention strategies. The real-time feedback module is configured to: calculate a health status score based on a preset health status scoring rule, and provide graded early warnings for the child's health status based on the health status score; The data processing module, the health suggestion generation module, and the real-time feedback module are integrated in the intelligent agent terminal, which is communicatively connected to the multi-source data acquisition terminal.

2. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, The standardization process includes the normalization of physiological indicators, which is based on preset medical standards or statistical models. The physiological indicators include: height or weight, dynamic heart rate, body temperature and blood oxygen.

3. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 2, characterized in that, In the aforementioned standardization process: The normalization of height or weight was performed by combining the BMI index with the WHO standard percentile corresponding to the child's age and sex. Standardize, standardized value The calculation rule is: when percentiles hour, ;when hour, ;when hour, ; The normalization process for dynamic heart rate includes: calculating the degree to which the current heart rate deviates from the ideal heart rate based on the age-appropriate resting heart rate standard; Normalization of body temperature includes: standardization based on the normal body temperature range, and standardized values. The calculation rule is: when body temperature or hour, ;otherwise, ; Normalization of blood oxygen includes standardization based on the normal threshold for blood oxygen saturation.

4. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, The behavioral intervention strategies include: When the climbing behavior lasts for more than 3 minutes per instance, it is determined that the activity of the liver meridian has increased, and a suggestion to read a calming picture book is sent. When the manual operation time is less than 2 minutes, the spleen deficiency index is determined to rise, triggering tactile feedback for intervention.

5. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, The health status score is calculated using the following formula: The calculation formula is: in, Standardized values ​​for height or weight. This is the standardized value of heart rate. This is the standardized value of body temperature. This is the standardized value of blood oxygen. , , and These are the weighting coefficients.

6. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 5, characterized in that, The real-time feedback module is also configured to calculate a suitability score for the puzzle activity based on heart rate variability, blood oxygen, motor coordination, and overall health status. Suitable rating for brain-training projects The calculation formula is: in, Weighting for brain-training projects. As a basic weight for health, This represents the standardized value for motor coordination.

7. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 5, characterized in that, The aforementioned tiered early warning system specifically includes: When the health status score is calculated as follows At that time, a red alert is triggered; When the health status score is calculated as follows At that time, a yellow alert was triggered; When the health status score is calculated as follows When this time is reached, it indicates that the person is in a normal health condition.

8. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, The aforementioned multimodal fusion analysis specifically includes: Map real-time scene labels to scene embedding vectors Simultaneously, the standardized values ​​of motion, physiological, and cognitive data are encoded into modal feature vectors. ,in, For feature dimensions; Based on a multi-head self-attention architecture, a scene-guided modality weight generator is constructed, with the following formula: in, , Represents motion modes, Represents physiological modalities. Representing cognitive modalities; , and The weight matrix is ​​a learnable matrix; Let be the dynamic attention weights for the i-th modality at time t, and ; This is the fused multimodal feature vector; Embed vectors for real-time scene features; Let i be the feature vector of the i-th modality; Query vectors guided by the scenario; Let be the key vector of the i-th mode; Let i be the value vector of the i-th mode; The correlation score between the scene and the i-th modality; In the amusement park scenario, motion modality weights Increase to over 60%, prioritizing the integration of gait balance and climbing frequency data; in classroom settings, cognitive modality weights Increase it to over 50%, focusing on analyzing participation and attention span in brain-training projects.

9. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, The intelligent terminal is also configured to act as a customer service representative, interacting with users in real time and collecting questions that parents are concerned about.

10. The multi-dimensional interactive assessment system for children's growth based on an intelligent agent terminal according to claim 1, characterized in that, It also includes a growth path planning module, which is configured as follows: Based on historical assessment data, a child's individual developmental trajectory model is constructed to predict growth trends over the next 3-6 months; Personalized intervention program sequences are generated using reinforcement learning algorithms; The outputs include phased goals related to movement, cognition, and environmental regulation.

Images