Nutrition management system based on IoT and cloud platform

Abstract

The present application provides a nutrition management system based on the Internet of Things and a cloud platform, which relates to the field of health management. It uses artificial intelligence-based data analysis technology to embed and encode the physical index data of the monitored object at the first time point and the second time point, so as to intelligently obtain the nutritional requirement value based on the characteristics obtained by the interactive response analysis between the two time points. In this way, the physical index data is collected in real time by the Internet of Things device and uploaded to the cloud for processing, so that the nutritional recommendations can be adjusted in time according to the actual changes of the individual, and highly personalized nutritional recommendations can be provided for different individuals, solving the problem of "one size fits all" in traditional methods, ensuring that everyone can obtain the most suitable nutritional guidance for themselves, thereby realizing the intelligence of the nutrition management system.

Description

Technical Field

[0001] The present application relates to the field of health management, and more specifically, to a nutrition management system based on the Internet of Things and a cloud platform. Background Art

[0002] In modern society, health has become one of the most concerned topics. With the improvement of living standards and the popularization of health information, the public has shown great interest in how to maintain and improve health through daily diet. Reasonable nutrition can not only meet the body's basic needs for energy and various nutrients, but also help to enhance immunity, promote physical development, improve the quality of life, and to a certain extent prevent and control various chronic diseases. Therefore, more and more people are beginning to seek professional nutrition guidance, hoping to achieve the best health through personalized diet plans.

[0003] However, traditional dietary management methods often provide general nutritional advice, ignoring individual differences, such as detailed physical indicators and changing factors such as age, gender, weight, and body fat. This one-size-fits-all approach cannot meet the specific needs of different groups of people, resulting in a significant reduction in the applicability of nutritional advice. For example, the nutritional needs of the elderly and adolescents are different, and the nutritional needs of sick people and healthy people are also very different, but traditional methods often cannot provide targeted advice. In addition, due to the lack of real-time monitoring and feedback mechanisms, traditional dietary management methods are difficult to dynamically adjust according to actual changes in individuals, resulting in limited effectiveness of dietary plans.

[0004] Therefore, an optimized nutrient management system is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a nutrition management system based on the Internet of Things and a cloud platform.

[0006] According to one aspect of the present application, a nutrition management system based on the Internet of Things and a cloud platform is provided, which includes:

[0007] A monitored object body index data acquisition module, used to obtain the monitored object's body index data at a first time point and at a second time point, wherein the body index data include body weight, BMI, lean body weight, body fat percentage, muscle mass, skeletal muscle mass, bone mass and body water percentage;

[0008] A monitored object body index data embedding coding module, used to embed the body index data of the monitored object at a first time point and the body index data at a second time point using a body index embedding coding matrix to obtain a first body index embedding coding vector and a second body index embedding coding vector;

[0009] a body indicator interaction response analysis module, configured to perform a first body indicator-second body indicator interaction response analysis on the first body indicator embedded coding vector and the second body indicator embedded coding vector to obtain a body indicator interaction response feature;

[0010] A nutritional requirement generation module, used to obtain nutritional requirements based on the interactive response characteristics between the body indicators;

[0011] The body index interaction response analysis module includes:

[0012] A second body indicator embedded coding vector feature decomposition unit, used for performing feature decomposition on the second body indicator embedded coding vector to obtain a sequence of second body indicator local embedded coding vectors;

[0013] a body indicator interaction correlation unit, configured to calculate an association matrix between the first body indicator embedded coding vector and each second body indicator local embedded coding vector in a sequence of the second body indicator local embedded coding vector to obtain a sequence of body indicator interaction correlation matrices;

[0014] The local-global synergistic response analysis unit between body indicators is used to perform local-global synergistic response analysis on the sequence of the interaction correlation matrix between body indicators to obtain the local-global interaction response matrix between body indicators as the interaction response feature between body indicators.

[0015] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the body indicator interaction association unit includes: a first body indicator nonlinear transformation subunit, which is used to perform a nonlinear transformation on the first body indicator embedded coding vector to obtain the first body indicator embedded coding vector after the nonlinear transformation; a first-second body indicator interaction subunit after the nonlinear transformation, which is used to calculate the association matrix between the first body indicator embedded coding vector after the nonlinear transformation and each second body indicator local embedded coding vector in the sequence of the second body indicator local embedded coding vector to obtain a sequence of the body indicator interaction association matrices.

[0016] In the above-mentioned nutrition management system based on the Internet of Things and cloud platform, the first body indicator nonlinear transformation subunit is used to: after multiplying the first body indicator embedding coding vector with the first body indicator transformation matrix, the obtained first body indicator weight vector and the first body indicator bias vector are added by position to obtain the first body indicator embedding coding vector after the nonlinear transformation.

[0017] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the first-second body indicator interaction subunit after the nonlinear transformation is used to: vector-multiply the transpose vector of the first body indicator embedding coding vector after the nonlinear transformation and each second body indicator local embedding coding vector in the sequence of the second body indicator local embedding coding vector to obtain a sequence of interaction correlation matrices between the body indicators.

[0018] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the local-global collaborative response analysis unit between body indicators includes: a characteristic interaction energy factor calculation subunit between body indicators, which is used to calculate the characteristic interaction energy factor of each interaction association matrix between body indicators in the sequence of interaction association matrices between body indicators to obtain a sequence of characteristic interaction energy factors between body indicators; a characteristic interaction energy factor normalization subunit between body indicators, which is used to input the sequence of characteristic interaction energy factors between body indicators into a Sigmoid function to obtain a sequence of normalized characteristic interaction energy factors between body indicators; a local-global interaction response matrix calculation subunit between body indicators, which is used to use the sequence of normalized characteristic interaction energy factors between body indicators as a sequence of weights, and calculate the position-weighted sum of the sequence of interaction association matrices between body indicators to obtain the local-global interaction response matrix between body indicators.

[0019] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the body indicator characteristic interaction energy factor calculation subunit includes: a body indicator interaction correlation kurtosis value calculation secondary subunit, which is used to calculate the kurtosis value of each body indicator interaction correlation matrix in the sequence of body indicator interaction correlation matrices to obtain a sequence of body indicator interaction correlation kurtosis values; a body indicator characteristic interaction energy factor calculation secondary subunit, which is used to calculate the characteristic interaction energy factor of each body indicator interaction correlation matrix based on the each body indicator interaction correlation kurtosis value in the sequence of body indicator interaction correlation kurtosis values ​​to obtain the sequence of body indicator characteristic interaction energy factors.

[0020] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the secondary subunit for calculating the kurtosis value of interaction correlation between body indicators is used to: respectively calculate the mean and variance of the interaction correlation matrix between body indicators to obtain the mean value of interaction correlation between body indicators and the variance of interaction correlation between body indicators; calculate the positional difference between the interaction correlation matrix between body indicators and the mean value of interaction correlation between body indicators to obtain the interaction correlation difference matrix between body indicators; calculate the fourth power of each eigenvalue in the interaction correlation difference matrix between body indicators to obtain the interaction correlation modulation difference matrix between body indicators; calculate the expected value of the interaction correlation modulation difference matrix between body indicators to obtain the expected value of interaction correlation between body indicators; divide the expected value of interaction correlation between body indicators by the square of the interaction correlation variance between body indicators to obtain the interaction correlation kurtosis value between body indicators.

[0021] In the above-mentioned nutrition management system based on the Internet of Things and the cloud platform, the secondary subunit for calculating the characteristic interaction energy factor between body indicators is used to: calculate the sum of the interaction correlation variance between the body indicators and the hyperparameter to obtain the first interaction energy factor between the body indicators; calculate the square of the difference between the interaction correlation kurtosis value between the body indicators and the interaction correlation mean between the body indicators to obtain the interaction correlation difference value between the body indicators; add the value obtained by multiplying the interaction correlation variance between the body indicators by a constant of two and the value obtained by multiplying the hyperparameter by a constant of two to the interaction correlation difference value between the body indicators to obtain the second interaction energy factor between the body indicators; divide the first interaction energy factor between the body indicators by the second interaction energy factor between the body indicators to obtain the characteristic interaction energy factor between the body indicators corresponding to the interaction correlation matrix between the body indicators.

[0022] In the above-mentioned nutrition management system based on the Internet of Things and cloud platform, the nutrition requirement generation module is used to: input the local-global interaction response matrix between the body indicators into the nutrition requirement generation module based on the decoder to obtain the decoded value of the nutrition requirement.

[0023] Compared with the prior art, the nutrition management system based on the Internet of Things and the cloud platform provided by the present application adopts artificial intelligence-based data analysis technology to embed and encode the physical index data of the monitored object at the first time point and the second time point, so as to intelligently obtain the nutritional requirement value according to the characteristics obtained by the interactive response analysis between the two time points. In this way, the physical index data is collected in real time by the Internet of Things device and uploaded to the cloud for processing, so that the nutrition advice can be adjusted in time according to the actual changes of the individual, and highly personalized nutrition advice can be provided for different individuals, which solves the problem of "one size fits all" in the traditional method, and ensures that everyone can get the most suitable nutrition guidance for themselves, thereby realizing the intelligence of the nutrition management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0025] Figure 1 This is a system block diagram of a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application.

[0026] Figure 2 Schematic diagram of data flow of a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application.

[0027] Figure 3 It is a block diagram of the interactive response analysis module between body indicators in the nutrition management system based on the Internet of Things and the cloud platform according to an embodiment of the present application.

[0028] Figure 4 It is a block diagram of the interactive correlation unit between body indicators in the nutrition management system based on the Internet of Things and the cloud platform according to an embodiment of the present application.

[0029] Figure 5 It is a block diagram of a local-global collaborative response analysis unit among body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application.

[0030] Figure 6 The block diagram is a subunit for calculating characteristic interaction energy factors between body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0032] In modern society, health has received much attention, and people are increasingly aware of the importance of maintaining and improving health through a reasonable diet. However, traditional diet management methods often provide general nutritional advice, ignoring individual differences (such as age, gender, weight, body fat and other indicators and changing factors), resulting in their inability to meet the specific needs of different groups of people and poor applicability; in addition, traditional diet management methods lack real-time monitoring and feedback mechanisms, making it difficult to dynamically adjust according to actual changes in individuals, limiting the effectiveness of diet plans.

[0033] Based on this, the present application proposes a nutrition management system based on the Internet of Things and a cloud platform, which obtains the physical index data (weight, BMI, fat-free body mass, body fat percentage, muscle mass, skeletal muscle mass, bone mass and body water percentage) of the monitored object at the first time point and the second time point, and uses artificial intelligence data analysis and processing technology based on the Internet of Things and a cloud platform to embed and encode the physical index data at the first time point and the second time point, so as to intelligently obtain the nutritional requirement value according to the characteristics obtained by the interactive response analysis between the two time points. In this way, the physical index data can be collected in real time through the Internet of Things device and uploaded to the cloud for processing. At the same time, the system can adjust the nutritional recommendations in time according to the actual changes of the individual, and provide highly personalized nutritional recommendations for different individuals, solving the problem of "one size fits all" in traditional methods, ensuring that everyone can get the most suitable nutritional guidance for themselves, thereby realizing the intelligence of the nutrition management system.

[0034] Figure 1 This is a system block diagram of a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 2 FIG. 1 is a data flow diagram of a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 1 and Figure 2 As shown, in the nutrition management system 100 based on the Internet of Things and the cloud platform, it includes: a monitored object body indicator data acquisition module 110, which is used to obtain the body indicator data of the monitored object at a first time point and the body indicator data at a second time point, wherein the body indicator data include weight, BMI, fat-free body weight, body fat percentage, muscle mass, skeletal muscle mass, bone mass and body water percentage; a monitored object body indicator data embedding coding module 120, which is used to embed the body indicator data of the monitored object at the first time point and the body indicator data at the second time point using a body indicator embedding coding matrix to obtain a first body indicator embedding coding vector and a second body indicator embedding coding vector; a body indicator interaction response analysis module 130, which is used to perform a first body indicator-second body indicator interaction response analysis on the first body indicator embedding coding vector and the second body indicator embedding coding vector to obtain a body indicator interaction response feature; a nutritional requirement generation module 140, which is used to obtain the nutritional requirement based on the body indicator interaction response feature.

[0035] In the embodiment of the present application, the monitored object body index data acquisition module 110 is used to obtain the body index data of the monitored object at the first time point and the body index data at the second time point, wherein the body index data includes weight, BMI, fat-free body mass, body fat rate, muscle mass, skeletal muscle mass, bone mass and body water rate. It should be understood that body weight directly affects energy demand, and weight gain or loss usually affects the basal metabolic rate and the energy required for daily activities; BMI (body mass index) evaluates the degree of obesity of an individual through the ratio of weight to height, and helps to determine whether it is necessary to adjust the diet to maintain a healthy weight; fat-free body mass reflects the muscle and bone mass of an individual, which helps to evaluate body composition and thus affects the demand for protein and other nutrients; body fat rate directly affects energy demand and nutrient distribution; muscle mass and skeletal muscle mass are closely related to exercise capacity and basal metabolic rate, affecting the energy demand and protein demand of an individual; bone mass is related to the demand for calcium and vitamin D, affecting bone health; body water rate affects electrolyte balance and overall metabolism in the body, and water intake is closely related to the absorption and utilization of other nutrients. In general, these body index data can fully reflect the health status and nutritional needs of an individual. Considering that the physical indicators of individual subjects may change due to factors such as diet, exercise, and lifestyle, by acquiring and analyzing the physical indicator data at two time points, physical changes can be identified, and then nutritional needs can be adjusted. This dynamic monitoring and adjustment method can help better meet the health needs of individuals. In particular, in a specific implementation method of the present application, the physical indicator data related to the monitored object is collected through an Internet of Things device such as an IoT body fat scale, and the collected individual physical indicator data is transmitted to the cloud platform in real time via wireless communication technology. The collected physical indicator data is processed using a data analysis algorithm on the cloud platform to obtain personalized nutritional management recommendations.

[0036] In an embodiment of the present application, the monitored object body indicator data embedding coding module 120 is used to embed the body indicator data of the monitored object at the first time point and the body indicator data at the second time point using a body indicator embedding coding matrix to obtain a first body indicator embedding coding vector and a second body indicator embedding coding vector. Accordingly, considering that the body indicator data of the monitored object at the first time point and the body indicator data at the second time point contain a large amount of information, including important data indicator features, and there are interactions and relationships between the various indicators, based on this, in order to extract and mine the important information at each time point and reduce the complexity of the data, in the technical solution of the present application, the body indicator embedding coding matrix is ​​used to embed the body indicator data of the monitored object at the first time point and the body indicator data at the second time point to obtain a first body indicator embedding coding vector and a second body indicator embedding coding vector.

[0037] In an embodiment of the present application, the body indicator interaction response analysis module 130 is used to embed the first body indicator into the coding vector and the second body indicator into the coding vector to perform a first body indicator-second body indicator interaction response analysis to obtain a body indicator interaction response feature. It should be understood that the physiological state of the human body is complex and interrelated. That is, the body indicators between different time points are not isolated, but there are complex mutual influences and dependencies. These relationships may be both local (such as the relationship between body fat percentage and lean body weight in a certain local time period) and global (such as the overall health status reflected by all indicators). Therefore, in order to more comprehensively understand the relationship between body indicators at various time points, so as to improve the flexibility and accuracy of the model, in the technical solution of the present application, the first body indicator is embedded into the coding vector and the second body indicator is embedded into the coding vector to perform a first body indicator-second body indicator interaction response analysis to obtain a local-global interaction response matrix between body indicators as a body indicator interaction response feature.

[0038] In particular, interactive response analysis enhances the model's understanding of data by fusing local features and global features, so that the information in the feature vector can be deeply analyzed and integrated to improve the performance of the model. In detail, the second body indicator embedding coding vector is firstly subjected to feature decomposition to decompose it into a sequence of second body indicator local embedding coding vectors of multiple local body indicators, which provides a basis for subsequent fine-grained analysis and processing. Then, the first body indicator embedding coding vector is subjected to nonlinear transformation to capture these more complex relationships and patterns to obtain the first body indicator embedding coding vector after nonlinear transformation. Then, the association matrix between the first body indicator embedding coding vector after nonlinear transformation and each second body indicator local embedding coding vector is calculated to capture the correlation between the body indicator data at different time points, so as to better understand the local interactions between these indicators and obtain a sequence of interaction correlation matrices between body indicators. Furthermore, the characteristic interaction energy factor of the interaction correlation matrix between each body indicator is calculated to obtain a sequence of characteristic interaction energy factors between body indicators. In particular, the characteristic interaction energy factor is obtained by the kurtosis value, mean and variance of the interaction correlation matrix between each physical index. Specifically, the kurtosis value is a statistic that describes whether the tail of the distribution is heavier than the normal distribution. A higher kurtosis value indicates that there are more extreme values ​​or outliers. Therefore, the kurtosis value can detect abnormalities in the physical index data; the mean describes the central tendency of the data set, which can help understand the overall health status of an individual within a certain time range; the variance measures the dispersion of the data points relative to their mean, and the stability of the physical index data can be understood. In this way, by calculating the characteristic interaction energy factor, a basis can be provided for intelligent decision-making on nutritional requirements. For example, if the interaction energy factor of a feature is high, it indicates that the feature plays an important role in the assessment of the nutritional requirements of the individual. Furthermore, the characteristic interaction energy factor of each interaction matrix is ​​normalized based on the Sigmoid function to map the values ​​in any range to the [0,1] interval, and convert it into the form of weights to ensure that the model can more effectively focus on the importance of the interaction of different features, so that the model can more balancedly consider the interaction between each feature. Furthermore, by normalizing the feature interaction energy factors with the Sigmoid function, the importance of different features can be quantified and expressed on a unified scale, which helps the model to more reasonably allocate attention to the interaction of each feature, ensuring the balance and accuracy of the nutrition management model when processing multi-feature data. Finally, the sequence of normalized factors is used as the weight sequence to weight and calculate the sequence of the interaction correlation matrix to obtain the local-global interaction response matrix between body indicators. The local-global interaction response matrix between body indicators finally obtained reflects the comprehensive response of all features at the local and global levels, providing comprehensive data support for the subsequent evaluation of nutritional requirements.

[0039] Specifically, Figure 3 FIG. 1 is a block diagram of an interactive response analysis module for body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 3 As shown, the body indicator interaction response analysis module 130 includes: a second body indicator embedded coding vector feature decomposition unit 131, which is used to perform feature decomposition on the second body indicator embedded coding vector to obtain a sequence of second body indicator local embedded coding vectors; a body indicator interaction correlation unit 132, which is used to calculate the correlation matrix between each second body indicator local embedded coding vector in the sequence of the first body indicator embedded coding vector and the second body indicator local embedded coding vector to obtain a sequence of body indicator interaction correlation matrices; a body indicator local-global coordinated response analysis unit 133, which is used to perform local-global coordinated response analysis on the sequence of the body indicator interaction correlation matrices to obtain a body indicator local-global interaction response matrix as the body indicator interaction response feature.

[0040] More specifically, Figure 4 FIG. 1 is a block diagram of an interactive correlation unit between body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 4 As shown, the body indicator interaction correlation unit 132 includes: a first body indicator nonlinear transformation subunit 1321, used for performing a nonlinear transformation on the first body indicator embedded coding vector to obtain a first body indicator embedded coding vector after the nonlinear transformation; a first-second body indicator interaction subunit 1322 after the nonlinear transformation, used for calculating the correlation matrix between the first body indicator embedded coding vector after the nonlinear transformation and each second body indicator local embedded coding vector in the sequence of the second body indicator local embedded coding vector to obtain a sequence of the body indicator interaction correlation matrices.

[0041] More specifically, in an embodiment of the present application, the first body indicator nonlinear transformation subunit 1321 is used to: after multiplying the first body indicator embedding coding vector with the first body indicator transformation matrix, add the obtained first body indicator weight vector and the first body indicator bias vector by position to obtain the first body indicator embedding coding vector after the nonlinear transformation.

[0042] More specifically, in the embodiment of the present application, the first-second body indicator interaction subunit 1322 after the nonlinear transformation is used to: vector-multiply the transpose vector of the first body indicator embedding coding vector after the nonlinear transformation and each second body indicator local embedding coding vector in the sequence of the second body indicator local embedding coding vector to obtain a sequence of interaction correlation matrices between the body indicators.

[0043] More specifically, Figure 5 FIG. 1 is a block diagram of a local-global synergistic response analysis unit among body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 5 As shown, the local-global collaborative response analysis unit 133 between body indicators includes: a characteristic interaction energy factor calculation subunit 1331 between body indicators, which is used to calculate the characteristic interaction energy factor of each interaction correlation matrix between body indicators in the sequence of interaction correlation matrices between body indicators to obtain a sequence of characteristic interaction energy factors between body indicators; a characteristic interaction energy factor normalization subunit 1332 between body indicators, which is used to input the sequence of characteristic interaction energy factors between body indicators into a Sigmoid function to obtain a sequence of normalized characteristic interaction energy factors between body indicators; a local-global interaction response matrix calculation subunit 1333 between body indicators, which is used to use the sequence of normalized characteristic interaction energy factors between body indicators as a sequence of weights, and calculate the position-weighted sum of the sequence of interaction correlation matrices between body indicators to obtain the local-global interaction response matrix between body indicators.

[0044] More specifically, Figure 6 FIG. 1 is a block diagram of a subunit for calculating characteristic interaction energy factors between body indicators in a nutrition management system based on the Internet of Things and a cloud platform according to an embodiment of the present application. Figure 6 As shown, the body indicator characteristic interaction energy factor calculation subunit 1331 includes: a body indicator interaction correlation kurtosis value calculation secondary subunit 13311, which is used to calculate the kurtosis value of each body indicator interaction correlation matrix in the sequence of body indicator interaction correlation matrices to obtain a sequence of body indicator interaction correlation kurtosis values; a body indicator characteristic interaction energy factor calculation secondary subunit 13312, which is used to calculate the characteristic interaction energy factor of each body indicator interaction correlation matrix based on each body indicator interaction correlation kurtosis value in the sequence of body indicator interaction correlation kurtosis values ​​to obtain a sequence of body indicator characteristic interaction energy factors.

[0045] More specifically, in the embodiment of the present application, the secondary subunit 13311 for calculating the interaction-correlation kurtosis value between body indicators is used to: respectively calculate the mean and variance of the interaction-correlation matrix between body indicators to obtain the interaction-correlation mean and the interaction-correlation variance between body indicators; calculate the positional difference between the interaction-correlation matrix between body indicators and the interaction-correlation mean between body indicators to obtain the interaction-correlation difference matrix between body indicators; calculate the fourth power of each eigenvalue in the interaction-correlation difference matrix between body indicators to obtain the interaction-correlation modulation difference matrix between body indicators; calculate the expected value of the interaction-correlation modulation difference matrix between body indicators to obtain the interaction-correlation expected value between body indicators; divide the interaction-correlation expected value between body indicators by the square of the interaction-correlation variance between body indicators to obtain the interaction-correlation kurtosis value between body indicators.

[0046] More specifically, in the embodiment of the present application, the secondary subunit 13312 for calculating the characteristic interaction energy factor between body indicators is used to: calculate the sum of the interaction correlation variance between the body indicators and the hyperparameter to obtain the first interaction energy factor between body indicators; calculate the square of the difference between the interaction correlation kurtosis value between the body indicators and the interaction correlation mean between the body indicators to obtain the interaction correlation difference value between body indicators; add the value obtained by multiplying the interaction correlation variance between the body indicators by a constant of two and the value obtained by multiplying the hyperparameter by a constant of two to the interaction correlation difference value between the body indicators to obtain the second interaction energy factor between body indicators; divide the first interaction energy factor between body indicators by the second interaction energy factor between body indicators to obtain the characteristic interaction energy factor between body indicators corresponding to the interaction correlation matrix between body indicators.

[0047] In the embodiment of the present application, specifically, the body indicator interaction response analysis module 130 is used to: embed the first body indicator into the coding vector and the second body indicator into the coding vector to perform first body indicator-second body indicator interaction response analysis, which can be expressed by the formula: ; ; ; ; ; ; ;in, is the second body indicator embedding coding vector, is the feature decomposition operation, are the first, second, ..., first in the sequence of the local embedding coding vector of the second body indicator respectively. ,..., The second body indicator local embedding encoding vector, is the number of feature vectors in the sequence of the second body indicator local embedding coding vector, is the first body indicator embedding coding vector, and are the first body index transformation matrix and the first body index offset vector, respectively, is the first body index embedding coding vector after the nonlinear transformation, yes The transposed vector of represents vector multiplication, For the a body index interaction correlation matrix between the second body index local embedding coding vector and the first body index embedding coding vector after the nonlinear transformation, yes The eigenvalues ​​at each position in , yes The mean of To calculate the expected value of a matrix, yes The variance of yes The kurtosis value of is a hyperparameter, yes The corresponding characteristic interaction energy factor between body indicators, yes function, yes The corresponding normalized characteristic interaction energy factor between body indicators, It is the point multiplication by position. is the local-global interaction response matrix between the body indices.

[0048] In an embodiment of the present application, the nutritional requirement generation module 140 is used to obtain the nutritional requirement based on the interactive response characteristics between the body indicators. Specifically, in an embodiment of the present application, the nutritional requirement generation module 140 is used to: input the local-global interactive response matrix between the body indicators into the decoder-based nutritional requirement generation module to obtain the decoded value of the nutritional requirement. That is, the interactive response characteristics between the body indicators obtained by interactive response analysis using the first body indicator embedded coding vector and the second body indicator embedded coding vector are decoded and processed to intelligently obtain the nutritional requirement value. In this way, the body indicator data is collected in real time through the Internet of Things device and uploaded to the cloud for processing. At the same time, the system can adjust the nutritional recommendations in time according to the actual changes of the individual, provide highly personalized nutritional recommendations for different individuals, solve the problem of "one size fits all" in traditional methods, and ensure that everyone can get the most suitable nutritional guidance for themselves, thereby realizing the intelligence of the nutritional management system.

[0049] In particular, since the first body indicator embedded coding vector and the second body indicator embedded coding vector respectively represent the embedded coding representations of body indicator data at different time points, when the first body indicator-second body indicator interaction response analysis is performed, the body indicator embedded coding timing difference will also have coding feature response differences based on the local-global interaction, so that the local-global interaction response matrix between the body indicators has a diversified set expression distribution of aggregated response characteristics. Therefore, it is expected to improve the balance between the regression mapping accuracy and completeness when the local-global interaction response matrix between the body indicators is input into the decoder-based nutrient requirement generation module for decoding regression, thereby improving the accuracy of the decoded value of the obtained nutrient requirement.

[0050] In a preferred example, inputting the local-global interaction response matrix between body indicators into the decoder-based nutritional requirement generation module to obtain the decoded value of the nutritional requirement includes: determining the maximum eigenvalue of the local-global interaction between body indicators and the minimum eigenvalue of the local-global interaction between body indicators of the local-global interaction response matrix between body indicators, and calculating the local-global interaction mean between body indicators and the local-global interaction standard deviation of the feature set of the local-global interaction response matrix between body indicators; calculating the quotient of the local-global interaction mean between body indicators and the local-global interaction standard deviation between body indicators to obtain the local-global interaction statistical normalization value between body indicators; calculating the inverse of each eigenvalue of the local-global interaction response matrix between body indicators, and comparing it with the local-global interaction maximum eigenvalue and the local-global interaction standard deviation. After multiplying the difference of the minimum eigenvalues ​​of the local-global interaction between the body indicators, the difference is subtracted from the local-global interaction statistical standardized value between the body indicators to obtain the local-global interaction distribution approximation matrix between the body indicators; an exponential function with each eigenvalue of the local-global interaction distribution approximation matrix between the body indicators as the exponent is calculated with the natural constant e as the base to obtain the local-global interaction distribution class approximation matrix between the body indicators; the local-global interaction distribution class approximation matrix between the body indicators and the local-global interaction statistical standardized value between the body indicators are added, and the base-two logarithm of the absolute value of each eigenvalue of the added matrix is ​​calculated to obtain the optimized local-global interaction response matrix between the body indicators; the optimized local-global interaction response matrix between the body indicators is input into the decoder-based nutrient requirement generation module to obtain the decoded value of the nutrient requirement.

[0051] The optimization expression of the local-global interaction response matrix between the body indicators is: ; ; ; ;in, is the local-global interaction response matrix between the body indicators, and respectively represent the mean value of the local-global interaction between body indicators and the standard deviation of the local-global interaction between body indicators of the feature set of the local-global interaction response matrix between body indicators, and are respectively the maximum eigenvalue of the local-global interaction between body indicators and the minimum eigenvalue of the local-global interaction between body indicators of the local-global interaction response matrix between body indicators, is the statistically normalized value of the local-global interaction between body indices, and is the difference between the maximum eigenvalue of the local-global interaction between the body indices and the minimum eigenvalue of the local-global interaction between the body indices, It means point multiplication by position. It means to subtract by position point. represents the natural exponential function, Indicates adding by position point. represents the local-global interaction distribution approximation matrix between body indicators, To calculate the inverse of each eigenvalue of the local-global interaction response matrix between the body indicators, is the local-global interaction response matrix between the optimized body indicators.

[0052] Here, by performing random statistical standardization on the diversified feature set of the local-global interaction response matrix between the body indicators, the confidence space constructed based on the overall eigenvalue of the local-global interaction response matrix between the body indicators is approximated by the response hypothesis test of each eigenvalue of the local-global interaction response matrix between the body indicators, thereby establishing the target reachability of the diversified feature distribution of the local-global interaction response matrix between the body indicators to the unified regression target, so as to achieve the balanced executableness between the mapping accuracy and the mapping integrity in the decoding regression process based on the diversified feature distribution of the local-global interaction response matrix between the body indicators, and improve the accuracy of the decoded value of the nutritional requirement obtained by the local-global interaction response matrix between the body indicators input into the nutritional requirement generation module based on the decoder. In this way, the body indicator data can be collected in real time through the Internet of Things device and uploaded to the cloud for processing. At the same time, the system can adjust the nutritional advice in time according to the actual changes of the individual, provide highly personalized nutritional advice for different individuals, solve the problem of "one size fits all" in the traditional method, and ensure that everyone can get the most suitable nutritional guidance for themselves, thereby realizing the intelligentization of the nutritional management system.

[0053] In summary, the nutrition management system 100 based on the Internet of Things and the cloud platform according to the embodiment of the present application is explained, which uses artificial intelligence-based data analysis technology to embed and encode the body index data of the monitored object at the first time point and the second time point, so as to intelligently obtain the nutritional requirement value according to the characteristics obtained by the interactive response analysis between the two time points. In this way, by collecting body index data in real time through the Internet of Things device and uploading it to the cloud for processing, it is possible to adjust the nutrition recommendations in time according to the actual changes of the individual, and then provide highly personalized nutrition recommendations for different individuals, solving the problem of "one size fits all" in traditional methods, ensuring that everyone can obtain the most suitable nutrition guidance for themselves, thereby realizing the intelligence of the nutrition management system.

[0054] As described above, the nutrition management system 100 based on the Internet of Things and the cloud platform according to the embodiment of the present application can be implemented in various terminal devices, such as a server for nutrition management based on the Internet of Things and the cloud platform. In one example, the nutrition management system 100 based on the Internet of Things and the cloud platform according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the nutrition management system 100 based on the Internet of Things and the cloud platform can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the nutrition management system 100 based on the Internet of Things and the cloud platform can also be one of the many hardware modules of the terminal device.

[0055] Alternatively, in another example, the nutrition management system 100 based on the Internet of Things and the cloud platform and the terminal device may also be separate devices, and the nutrition management system 100 based on the Internet of Things and the cloud platform may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0056] In other embodiments of the present application, there is provided an IoT nutrition management cloud platform, which includes an APP, a PC end, and an IoT body fat scale, serving the patient end and the medical end, aiming to achieve closed-loop management of home nutrition. On the patient end, the core functions provided by the APP specifically include: a. Nutrition management: Patients can enter their own nutritional status every week, and the program will automatically evaluate according to the NRS2002 assessment scale and PG-SGA. In addition, through the connection with the IoT body fat scale, the program can automatically calculate and update the patient's nutritional needs and provide personalized diet recipes. b. Diet diary: Patients can record their daily diet, and the program will automatically calculate energy and calorie intake based on the records to ensure that nutritional goals are achieved. c. Nutritional evaluation: The program monitors key indicators such as weight and BMI in real time through the IoT body fat scale. At the same time, patients can take photos of test results and upload them to the system. The system will set a reminder function according to the patient's discharge window period to inform the patient of changes in diet plan and weight measurement. The program sets a reminder function during the window period. If weight monitoring is not performed or the weight drops by 5% in a month, the program will set a two-way reminder function to inform the patient to measure weight in time and remind the medical end. d. Nutrition consultation: Patients can consult with doctors or nutrition specialist nurses online through the APP, and the system also provides AI automatic answering services. e. Nutrition knowledge: Nutrition specialist nurses regularly update nutrition education materials to help patients better understand nutrition knowledge. On the medical and nursing side, services are provided through the corporate WeChat page or PC, including nutrition education, Q&A and patient follow-up. Medical staff can efficiently manage patient data and plan follow-up work in advance through system prompts, thereby improving follow-up efficiency.

[0057] The above is only a preferred embodiment of the present application and does not constitute any form of limitation to the present application. Although the present application has been disclosed as a preferred embodiment as above, it is not intended to limit the present application. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of the technical contents disclosed above without departing from the scope of the technical solution of the present application. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application still falls within the scope of the technical solution of the present application.

Claims

1. A nutrition management system based on the Internet of Things and cloud platform, characterized in that: include: A monitored object body index data acquisition module, used to obtain the monitored object's body index data at a first time point and at a second time point, wherein the body index data include body weight, BMI, lean body weight, body fat percentage, muscle mass, skeletal muscle mass, bone mass and body water percentage; A monitored object body index data embedding coding module, used to embed the body index data of the monitored object at a first time point and the body index data at a second time point using a body index embedding coding matrix to obtain a first body index embedding coding vector and a second body index embedding coding vector; a body indicator interaction response analysis module, configured to perform a first body indicator-second body indicator interaction response analysis on the first body indicator embedded coding vector and the second body indicator embedded coding vector to obtain a body indicator interaction response feature; A nutritional requirement generation module, used to obtain nutritional requirements based on the interactive response characteristics between the body indicators; The body index interaction response analysis module includes: A second body indicator embedded coding vector feature decomposition unit, used for performing feature decomposition on the second body indicator embedded coding vector to obtain a sequence of second body indicator local embedded coding vectors; a body indicator interaction correlation unit, configured to calculate an association matrix between the first body indicator embedded coding vector and each second body indicator local embedded coding vector in a sequence of the second body indicator local embedded coding vector to obtain a sequence of body indicator interaction correlation matrices; The local-global synergistic response analysis unit between body indicators is used to perform local-global synergistic response analysis on the sequence of the interaction correlation matrix between body indicators to obtain the local-global interaction response matrix between body indicators as the interaction response feature between body indicators.

2. The nutrition management system based on the Internet of Things and cloud platform according to claim 1, characterized in that: The body index interaction association unit comprises: A first body index nonlinear transformation subunit, configured to perform a nonlinear transformation on the first body index embedding coding vector to obtain a first body index embedding coding vector after nonlinear transformation; The first-second body indicator interaction subunit after nonlinear transformation is used to calculate the correlation matrix between each second body indicator local embedded coding vector in the sequence of the first body indicator embedded coding vector after nonlinear transformation and the second body indicator local embedded coding vector to obtain a sequence of interaction correlation matrices between the body indicators.

3. The nutrition management system based on the Internet of Things and cloud platform according to claim 2, characterized in that: The first body indicator nonlinear transformation subunit is used to: after multiplying the first body indicator embedding coding vector with the first body indicator transformation matrix, add the obtained first body indicator weight vector and the first body indicator bias vector according to the position to obtain the first body indicator embedding coding vector after the nonlinear transformation.

4. The nutrition management system based on the Internet of Things and cloud platform according to claim 3, characterized in that: The first-second body indicator interaction subunit after nonlinear transformation is used to: perform vector multiplication on the transposed vector of the first body indicator embedding coding vector after nonlinear transformation and each second body indicator local embedding coding vector in the sequence of the second body indicator local embedding coding vector to obtain a sequence of interaction correlation matrices between the body indicators.

5. The nutrition management system based on the Internet of Things and cloud platform according to claim 4, characterized in that: The local-global synergistic response analysis unit between body indicators includes: A body index characteristic interaction energy factor calculation subunit, used for calculating the characteristic interaction energy factor of each body index interaction correlation matrix in the sequence of body index interaction correlation matrices to obtain a sequence of body index characteristic interaction energy factors; A normalization subunit of characteristic interaction energy factors between body indicators, used for inputting the sequence of characteristic interaction energy factors between body indicators into a Sigmoid function to obtain a sequence of normalized characteristic interaction energy factors between body indicators; The local-global interaction response matrix calculation subunit between body indicators is used to calculate the position-weighted sum of the sequence of interaction correlation matrices between body indicators using the sequence of normalized characteristic interaction energy factors between body indicators as a sequence of weights to obtain the local-global interaction response matrix between body indicators.

6. The nutrition management system based on the Internet of Things and cloud platform according to claim 5, characterized in that: The body index characteristic interaction energy factor calculation subunit includes: A secondary subunit for calculating the kurtosis value of the interaction correlation between body indicators is used to calculate the kurtosis value of each interaction correlation matrix between body indicators in the sequence of interaction correlation matrices between body indicators to obtain a sequence of interaction correlation kurtosis values ​​between body indicators; The secondary subunit for calculating characteristic interaction energy factors between body indicators is used to calculate the characteristic interaction energy factors of the interaction correlation matrices between each body indicator based on the interaction correlation kurtosis values ​​between each body indicator in the sequence of interaction correlation kurtosis values ​​between body indicators to obtain the sequence of characteristic interaction energy factors between body indicators.

7. The nutrition management system based on the Internet of Things and cloud platform according to claim 6, characterized in that: The secondary subunit for calculating the interactive correlation kurtosis value between the physical indicators is used for: Respectively calculating the mean and variance of the interaction correlation matrix between the body indicators to obtain the mean of the interaction correlation between the body indicators and the variance of the interaction correlation between the body indicators; Calculating the positional difference between the interaction correlation matrix between the body indicators and the interaction correlation mean between the body indicators to obtain the interaction correlation difference matrix between the body indicators; Calculating the fourth power of each eigenvalue in the interaction-correlation difference matrix between the body indicators to obtain an interaction-correlation modulation difference matrix between the body indicators; Calculating the expected value of the interaction-correlation modulation difference matrix between the body indicators to obtain the expected value of the interaction-correlation between the body indicators; The expected value of the interaction correlation between the body indicators is divided by the square of the variance of the interaction correlation between the body indicators to obtain the kurtosis value of the interaction correlation between the body indicators.

8. The nutrition management system based on the Internet of Things and cloud platform according to claim 7, characterized in that: The secondary subunit for calculating the characteristic interaction energy factor between body indicators is used for: Calculating the sum of the variance of the interaction correlation between the body indicators and the hyperparameter to obtain a first energy factor of the interaction correlation between the body indicators; Calculating the square of the difference between the kurtosis value of the interaction correlation between the body indicators and the mean value of the interaction correlation between the body indicators to obtain the difference value of the interaction correlation between the body indicators; The value obtained by multiplying the variance of the mutual correlation between the body indicators by a constant of two and the value obtained by multiplying the hyperparameter by a constant of two are added to the difference value of the mutual correlation between the body indicators to obtain a second energy factor of the mutual correlation between the body indicators; The first energy factor of the interaction correlation between the body indicators is divided by the second energy factor of the interaction correlation between the body indicators to obtain the characteristic interaction energy factor between the body indicators corresponding to the interaction correlation matrix between the body indicators.

9. The nutrition management system based on the Internet of Things and cloud platform according to claim 8, characterized in that: The nutritional requirement generation module is used to: input the local-global interaction response matrix between the body indicators into the decoder-based nutritional requirement generation module to obtain the decoded value of the nutritional requirement.

Images