Physiological data-based healthy diet recommendation method, device and system and medium

Abstract

According to the embodiment of the invention, multi-dimensional health data acquired by user health data monitoring equipment is acquired and preprocessed; constructing a multi-dimensional knowledge graph, and mapping the multi-dimensional health data to corresponding nodes in the knowledge graph to obtain real-time health data of the user; based on the real-time health data of the user and the knowledge graph analysis result, utilizing an artificial intelligence algorithm to dynamically generate a personalized target diet scheme; the method deeply depends on the individual biological data of the user, the current physical state of the user is comprehensively analyzed through precise real-time physical sign data collection in combination with an advanced knowledge graph technology, meanwhile, potential health risks are accurately predicted, and comprehensiveness and accuracy of analysis are ensured; the diet suggestions are flexibly, accurately and dynamically adjusted according to the real-time health data of the user by using the AI large model, so that the short-term health improvement target of the user is supported, and the user is assisted to realize long-term health planning for a long time.

Description

Technical Field

[0001] The present invention relates to the field of health management, and in particular to a method, device, system and medium for recommending healthy diet based on physiological data. Background Art

[0002] As people's quality of life and health awareness improve, the demand for personalized health management services is growing at an unprecedented rate. However, most of the healthy diet recommendation systems currently on the market are limited to the analysis of static information, and it is difficult to flexibly capture and respond to dynamic changes in users' health conditions. What is more worrying is that these systems generally lack real-time monitoring and early warning mechanisms for health risks, and the generation of their dietary recommendations often relies on fixed rules, making it difficult to achieve efficient, accurate and personalized dietary guidance. In view of this, how to rely on real-time updated health data to dynamically and scientifically formulate a reasonable and personalized diet plan has become a key issue that needs to be overcome. Summary of the invention

[0003] In view of this, the purpose of the embodiments of the present invention is to provide a healthy diet recommendation method, device, system and medium based on physiological data, so as to solve the real-time updated health data, dynamically and scientifically formulate a reasonable and personalized diet plan, and realize efficient, accurate and personalized diet guidance.

[0004] In one aspect, an embodiment of the present invention provides a method for recommending a healthy diet based on physiological data, comprising: Obtain multi-dimensional health data collected by the user's health data monitoring device and perform pre-processing; Construct a multi-dimensional knowledge graph, map the multi-dimensional health data to corresponding nodes in the knowledge graph, and obtain the user's real-time health data; wherein the real-time health data represents the user's health status and potential health risks; Based on the user's real-time health data and knowledge graph analysis results, artificial intelligence algorithms are used to dynamically generate personalized target diet plans.

[0005] Optionally, the acquiring multi-dimensional health data collected by the user health data monitoring device and preprocessing the data includes: The time series data is smoothed based on the sliding window filtering method, and the formula is: ; in, is the original data at time t, and N is the window size.

[0006] Optionally, the smoothing of the time series data based on the sliding window filtering method includes: Normalize the smoothed data, where the normalization is to the interval [0, 1], and the formula is as follows: ; Wherein, and are the first historical target value and the second historical target value of this index respectively.

[0007] Optionally, after normalizing the smoothed data, it includes; Detect outliers based on the three - standard - deviation method, and the formula is as follows: If , then is an outlier; Wherein, and are the mean and standard deviation of this index respectively; Receive time - series data from multiple data sources, and use the interpolation method to align missing and / or inconsistent time points, and the formula is as follows: ; Wherein, and are the timestamps of two adjacent data points respectively.

[0008] Optionally, constructing the multi - dimensional knowledge graph includes: Construct a multi - dimensional knowledge graph in the form of triples; Obtain health and nutrition - related data based on multiple data sources, store the processed data in a graph database, and represent nodes and their relationships in the form of an adjacency matrix. The formula is as follows: ; Where w ij is the relationship weight between node i and node j; judge whether there is a relationship between node i and node j. If there is a relationship between node i and node j, the element A[i, j] at the corresponding position in the adjacency matrix A is assigned a non - zero value w ij ; if there is no relationship between node i and node j, the element A[i, j] at the corresponding position in the adjacency matrix A is assigned a zero value w ij ; Use natural language processing technology to extract text data, and use the BERT or BioBERT model to perform entity recognition and relationship classification on the processed text data; the relationship weight formula is as follows: ; Where: represents the relationship score based on the document frequency; represents the calibration score based on domain expert scoring or experimental data; is a weight factor, satisfying ; Store the extracted entities and their relationships into a graph database using Neo4j, and build a visualization interface based on the graph database to graphically display the entity and its relationship network so as to realize the functions of searching, filtering, and expanding the knowledge of the knowledge graph.

[0009] Optionally, mapping the multi-dimensional health data to the corresponding nodes in the knowledge graph to obtain the user's real-time health data, including: Map the user's health indicators to the knowledge graph nodes and calculate the matching score of each node: ; where is the comprehensive score of the entity ; is the weight of the health indicator; is the influence function of the health indicator value.

[0010] Optionally, the dynamically generating a personalized target diet plan based on the user's real-time health data and the knowledge graph analysis results includes: Define an optimization objective function according to the user's health data and goals: ; where: is the health score for achieving the short-term goal; is the trend score for achieving the long-term goal; is the satisfaction score of the user's preference; is a weight factor that is dynamically adjusted according to the user's priority.

[0011] On the other hand, an embodiment of the present invention provides a health diet recommendation device based on physiological data, including: The first module is used to obtain the multi-dimensional health data collected by the user health data monitoring device and perform preprocessing; The second module is used to build a multi-dimensional knowledge graph, map the multi-dimensional health data to the corresponding nodes in the knowledge graph, and obtain the user's real-time health data; wherein, the real-time health data represents the user's health status and potential health risks; The third module is used to dynamically generate a personalized target diet plan based on the user's real-time health data and the knowledge graph analysis results.

[0012] On the other hand, an embodiment of the present invention provides a health diet recommendation system based on physiological data, and the system includes: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method for recommending healthy diet based on physiological data as described in any one of the above.

[0013] On the other hand, an embodiment of the present invention provides a computer-readable storage medium, in which a program executable by a processor is stored. The program executable by the processor is used to execute a method for recommending healthy diet based on physiological data as described in any one of the above when executed by the processor.

[0014] In the embodiment of the present invention, multi-dimensional health data collected by a user's health data monitoring device is obtained and preprocessed; a multi-dimensional knowledge graph is constructed, and the multi-dimensional health data is mapped to corresponding nodes in the knowledge graph to obtain the user's real-time health data; based on the user's real-time health data and the analysis result of the knowledge graph, an artificial intelligence algorithm is used to dynamically generate a personalized target diet plan. The present invention deeply relies on the user's individual biological data, through precise real-time physical sign data collection (covering key indicators such as blood glucose level, blood pressure status, body fat ratio, etc.), and combines advanced knowledge graph technology to comprehensively analyze the user's current physical state, and at the same time accurately predict potential health risks. This algorithm integrates three core modules: an efficient data collection module, a precise knowledge graph matching module, and an intelligent optimization generation module. By widely compatible with various health data collection devices, the system can build a complex association model between diseases, foods, and nutritional components. Based on the profound foundation of the knowledge graph, this model ensures the comprehensiveness and accuracy of the analysis. The system uses the powerful computing power of the AI large model and the flexibility of the adaptive algorithm to flexibly and accurately dynamically adjust diet recommendations according to the user's real-time health data, not only supporting the user's short-term health improvement goals, but also helping them achieve long-term health plans in the long run.

[0015] The application scope of the present invention is wide, which is not only applicable to the refined requirements of personalized health management, but also provides strong technical support for the effective prevention and control of chronic diseases and targeted nutritional interventions, demonstrating its great potential in promoting human health and well-being. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of a healthy diet recommendation method based on physiological data provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of the overall process provided by an embodiment of the present invention; Figure 3 A schematic diagram of knowledge graph construction provided by an embodiment of the present invention; Figure 4 It is another schematic diagram of knowledge graph construction provided by an embodiment of the present invention; Figure 5 It is a schematic diagram of the visualization interface of the knowledge graph in neo4j provided by an embodiment of the present invention; Figure 6 It is a structural block diagram of a system provided by an embodiment of the present invention; Figure 7 It is a structural block diagram of a device provided by an embodiment of the present invention. Detailed implementation manners

[0018] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0019] It should be noted that although functional module division is performed in the device schematic diagram and the logical sequence is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order from the module division in the device or the sequence in the flowchart. Terms such as "first", "second", etc. in the description, claims and the above-mentioned drawings are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0021] In addition, the described features, structures or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, many specific details are provided to give a full understanding of the embodiments of the present application. However, those skilled in the art will realize that the technical solutions of the present application can be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. can be adopted. In other cases, well-known methods, devices, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present application.

[0022] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software form, or implemented in one or more hardware charging modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0023] The flowcharts shown in the accompanying drawings are merely exemplary descriptions, and do not necessarily include all contents and operations / steps, nor are they necessarily executed in the described order. For example, some operations / steps can be decomposed, while some operations / steps can be combined or partially combined, so the actual execution order may change according to the actual situation.

[0024] Refer to Figures 1 to 5 As shown, a method for recommending healthy diet based on physiological data provided by an embodiment of the present invention includes: S1. Obtain multi-dimensional health data collected by a user health data monitoring device, and perform preprocessing; Exemplarily, multi-dimensional health data can be obtained from the user side in real time and preprocessed to provide accurate input for subsequent knowledge graph matching and optimization calculation.

[0025] It can be understood that the user health data monitoring device includes a variety of health monitoring devices, such as: Smart watch / wristband: used to collect heart rate, blood oxygen saturation, exercise amount, sleep quality, etc.; Blood glucose monitor: collect dynamic blood glucose data, including fasting blood glucose value and postprandial blood glucose change curve; Blood pressure monitor: monitor the systolic blood pressure and diastolic blood pressure of the user, and calculate the blood pressure fluctuation range; Body fat scale: collect indicators such as user weight, body fat percentage, muscle content, etc.; Optional devices: such as thermometer, uric acid detector, etc., used to supplement other health data.

[0026] All devices are connected to the system host through Bluetooth, Wi-Fi, NFC or USB interface to ensure real-time transmission of data to the cloud or local processing unit.

[0027] It should be noted that in order to achieve seamless docking of hardware data and the system, a unified data acquisition API has been developed, which supports the following functions: Device compatibility support: the developed API supports the calling of driver programs of mainstream brand devices (such as Apple HealthKit, Google Fit, Huawei Sports Health, Xiaomi Sports Health, oppo, vivo, etc.).

[0028] Data Parsing and Validation: After receiving the raw data, the API parses it into a standardized format and eliminates abnormal data through validation rules.

[0029] Data Transmission and Encryption: The data is transmitted using the AES-256 encryption algorithm to ensure user privacy and security.

[0030] S2. Construct a multi-dimensional knowledge graph, map the multi-dimensional health data to the corresponding nodes in the knowledge graph to obtain the user's real-time health data; wherein, the real-time health data represents the user's health status and potential health risks. Exemplarily, receive the user's multi-dimensional health data, including physiological indicators, behavior data, and eating habits, and accurately map the user's health data to the corresponding nodes in the knowledge graph; analyze the user's health status based on the knowledge graph and identify potential health risks.

[0031] S3. Based on the user's real-time health data and the analysis results of the knowledge graph, use artificial intelligence algorithms to dynamically generate personalized target diet plans.

[0032] Exemplarily, use natural language processing (NLP) to understand user preferences and feedback. Analyze the eating patterns of similar users based on the collaborative filtering algorithm. Use reinforcement learning to optimize diet recommendations to adapt to dynamic adjustments based on user feedback. Generate personalized health reports according to the analysis results and recommend targeted intervention measures or diet plans.

[0033] Specifically, the acquisition of multi-dimensional health data collected by the user health data monitoring device and the preprocessing thereof include: Perform smoothing processing on time series data based on the sliding window filtering method, and its formula is: ; Wherein, is the raw data at time t, and N is the window size.

[0034] Exemplarily, perform denoising on multi-dimensional health data, etc.

[0035] Optionally, after performing smoothing processing on time series data based on the sliding window filtering method, it includes: Normalize the smoothed data, wherein it is normalized to the interval [0, 1], and its formula is as follows: ; Wherein, and are the first historical target value and the second historical target value of this indicator, respectively.

[0036] Exemplarily, the first historical target value is the historical minimum value, and the second historical target value is the historical maximum value.

[0037] Optionally, after normalizing the smoothed data, it includes: Detect outliers based on the three - standard - deviation method, and its formula is as follows: If , then is an outlier; where and are the mean and standard deviation of this indicator respectively; Exemplarily, receive multi - dimensional time - series health data, and calculate the mean and of the data; detect outliers based on the three - standard - deviation method, remove the outliers that meet the above conditions, and perform complementation or smoothing processing on the remaining data.

[0038] Receive time - series data from multiple data sources, and use the interpolation method to align the missing and / or inconsistent time points, and its formula is as follows: ; where and are the timestamps of two adjacent data points respectively.

[0039] Exemplarily, a unified reference time axis can be set; for the missing and / or inconsistent time points, use the interpolation method to align the data to generate time - series data with consistent timestamps. For non - linearly changing data sequences, use spline interpolation or polynomial interpolation for alignment. For large data - missing regions, use the moving window average method for filling processing to maintain data smoothness and continuity. The time alignment process supports dynamic window adjustment to adaptively select the interpolation method according to the change of the timestamp interval.

[0040] For example, a blood glucose monitor: the blood glucose value is 8.7 mmol / L; a sphygmomanometer: the blood pressure is 140 / 90 mmHg.

[0041] Pretreatment result: After filtering the blood glucose data, the smoothed value is 8.6 mmol / L; The normalized value of the blood pressure data is {0.7, 0.6}; After the above processing, the data can be input into the knowledge graph matching module for subsequent analysis and optimization calculation.

[0042] Optionally, the construction of the multi - dimensional knowledge graph includes: Construct a multi - dimensional knowledge graph in the form of triples; Exemplarily, the knowledge graph is constructed in the form of triples (Entity, Relation, Entity), and the main elements can be: Entity Related to the user's health status: such as "high blood sugar", "hypertension", "vitamin D deficiency", etc.; Food and ingredients: such as "brown rice", "vitamin C", "calcium ions", etc.; Diseases and risks: such as "diabetes", "cardiovascular diseases", etc.; Relation Health status and diseases: such as "associated with", "trigger"; Food and ingredients: such as "contains", "is rich in"; Ingredients and health needs: such as "improve", "regulate"; For example: (High blood sugar --> trigger --> diabetes) (Brown rice --> is rich in --> complex carbohydrates) (Vitamin C --> improve --> immunity) Attributes Each entity and relation has quantified attribute values, such as weights, dosage ranges, etc. For example: the carbohydrate content attribute of the food "brown rice": 75g / 100g; the association weight between "high blood sugar" and "excessive carbohydrate intake": 0.8; Obtain health and nutrition-related data from multiple data sources, store the processed data in a graph database, and represent the nodes and their relationships in the form of an adjacency matrix. The formula is as follows: ; where w ij is the relationship weight between node i and node j; to determine whether there is a relationship between node i and node j, if there is a relationship between node i and node j, the element A[i,j] at the corresponding position in the adjacency matrix A is assigned a non-zero value w ij ; if there is no relationship between node i and node j, the element A[i,j] at the corresponding position in the adjacency matrix A is assigned a zero value w ij ; It should be noted that this formula aims to describe the associations between nodes in the graph database. In the form of an adjacency matrix, it clearly and intuitively shows the relationship connections between nodes.

[0043] Exemplarily, health-related data is collected from multiple data sources, where the data sources include medical literature, public health databases, and nutritional composition tables; the collected data is reviewed and verified in combination with domain expert annotations to improve data accuracy and reliability; the processed data is stored in a knowledge graph and is represented and managed in a structured manner.

[0044] It can be understood that the data sources of medical literature include the PubMed database, the Cochrane Library, and other peer-reviewed literature resources. The public health databases include the WHO Nutrition Database, the FAO Food Composition Database, and other international authoritative health data resources. The nutritional composition tables include the USDA Nutrition Database, the CNF (Canadian Nutrient File), and other national food nutrition composition standard databases. The domain expert annotations adopt a combination of manual review and machine learning-assisted annotation to improve efficiency and accuracy.

[0045] Natural language processing techniques are used to extract text data, and the BERT or BioBERT model is used to perform entity recognition and relationship classification on the processed text data; the relationship weight formula is as follows: ; Where: represents the relationship score based on the literature frequency; represents the calibration score based on domain expert scoring or experimental data; is the weight factor, satisfying ; The extracted entities and their relationships are stored in a graph database using Neo4j, and a visualization interface is constructed based on the graph database to graphically display the entity and its relationship network so as to realize the functions of searching, filtering, and expanding the knowledge of the graph.

[0046] Exemplarily, text data is obtained and preprocessed, including word segmentation, stop word removal, and syntactic parsing. The BERT or BioBERT model is used to perform entity recognition and relationship classification on the processed text data; the relationship weight between entities is calculated. Then the extracted entities and their relationships are stored in a graph database using Neo4j; a visualization interface is constructed based on the graph database to graphically display the entity and its relationship network, thereby realizing the functions of searching, filtering, and expanding the knowledge of the graph.

[0047] It can be understood that entity recognition includes the extraction of disease names, nutritional components, food names, and their associated information; relationship classification includes the recognition and annotation of causal relationships, correlation relationships, and hierarchical dependency relationships; the entities and relationships stored in the Neo4j graph database are represented in the form of nodes and edges and support dynamic expansion and real-time update.

[0048] Exemplarily, the visualization interface also supports graphical display of entities and relationship networks, locating specified entities and relationships in the search box, filtering specific information according to filtering conditions; dragging to adjust the node layout and relationship display mode.

[0049] Optionally, mapping the multi-dimensional health data to the corresponding nodes in the knowledge graph to obtain the user's real-time health data includes: Mapping user health indicators to the knowledge graph nodes and calculating the matching score for each node: ; where is the comprehensive score of entity ; is the weight of the health indicator; is the influence function of the health indicator value.

[0050] Exemplarily, assume the user's blood glucose is 8.5 mmol / L (higher than the normal value), and the system maps it to the "hyperglycemia" node with a matching score S(hyperglycemia) = 0.9. Using the relationships in the knowledge graph for derivation, generate health suggestions: starting from the "hyperglycemia" node, derive the associated disease "diabetes"; through the "rich in" relationship, find suitable foods (such as "brown rice"). According to the user's historical data and current status, adjust the node weights in the graph. After the user has long-term intake of vitamin C and the immunity is improved, the weight of "vitamin C - improve - immunity" is reduced, indicating a decrease in the recommended dosage.

[0051] The specific operation steps are as follows; 1. Input data: User's blood glucose value: 8.5 mmol / L, weight: 80 kg, exercise volume: less than 5000 steps per day; 2. Health status: Map the blood glucose value to the "hyperglycemia" node with a score of 0.9; insufficient exercise is mapped to the "insufficient exercise" node with a score of 0.8; 3. Knowledge derivation: From the "hyperglycemia" node, derive the risk disease "diabetes"; combined with the "insufficient exercise" node, recommend an exercise plan and diet optimization: recommend a low-carbohydrate diet; 4. Recommended foods: Foods associated with "low-carbohydrate diet" in the graph: such as "brown rice", "spinach", "chicken breast"; dynamically generate a specific food combination plan; Through the processing of the knowledge graph matching module, the user's health data is transformed into structured knowledge, providing a basis for generating personalized diet plans. This matching method improves the accuracy and scientificity of health suggestions, and at the same time supports dynamic updates to adapt to changes in the user's status.

[0052] Optionally, the dynamic generation of a personalized target diet plan based on the user's real-time health data and the knowledge graph analysis results includes: Define an optimization objective function according to the user's health data and goals: ; Where: is the health score for achieving short-term goals; is the trend score for achieving long-term goals; is the satisfaction score for the user's preferences; is a weight factor, dynamically adjusted according to the user's priorities.

[0053] Exemplarily, analyze the user's health status based on the knowledge graph to identify short-term and long-term health needs; through the goal setting module, the user can set short-term and long-term goals, including blood sugar control, fat loss, or immune enhancement; use the optimization algorithm engine, adopt a multi-objective optimization algorithm, and comprehensively consider the user's health needs, dietary preferences, and food availability to dynamically generate a recommended personalized diet plan; based on the user feedback learning module, update the diet plan in real time to adapt to changes in user needs and optimize the recommendation effect.

[0054] It can be understood that short-term goals include blood sugar control, blood pressure reduction, or acute nutritional supplementation, and long-term goals include fat loss, immune enhancement, and chronic disease prevention; the system automatically recommends goal priorities through health assessment.

[0055] The optimization algorithm engine adopts a multi-objective optimization algorithm, comprehensively considers health needs, dietary preferences, and food availability, and dynamically generates a recommended plan; perform multi-objective optimization calculations on the diet plan based on genetic algorithms, particle swarm optimization algorithms, or reinforcement learning algorithms.

[0056] The user feedback learning module adjusts the recommended plan according to the user's actual feedback to make it more in line with the user's needs; dynamically adjust the recommendation results based on collaborative filtering algorithms or gradient boosting decision trees (GBDT).

[0057] Specifically, the constraint conditions can be set as Total daily energy intake: meet the user's metabolic needs (BMR): ; Where, is the intake of food i, is its calorie value; The nutritional balance constraint can be set as: Ensure that the proportions of macronutrients (carbohydrates, proteins, fats) meet the standards: ; Generate an initial solution using a genetic algorithm; evaluate the fitness of each solution according to the objective function and constraints: ; where is the health score of the index, is the index weight; Use the Particle Swarm Optimization (PSO) algorithm or Reinforcement Learning (RL) to further adjust the solution: ; where, is the particle velocity, is the particle position, and are the individual best solution and the global best solution respectively; The optimized diet plan is presented in the form of a list and a timeline, including the following: Specific food recommendations; Breakfast: 50g of oats, 200ml of milk, 30g of blueberries; Lunch: 100g of brown rice, 150g of steamed chicken breast, 100g of broccoli; Dinner: 80g of sweet potatoes, 2 hard-boiled eggs, 150g of spinach; Drug and nutrient recommendations (if needed): Vitamin D3 supplement 1000IU / day; Health goal matching degree: The target adaptation score of each meal plan (such as glycemic load, calorie requirement); Dynamic adjustment suggestions; If the user feedback is "too salty", adjust the plan to reduce salt intake; If the blood sugar fluctuation exceeds the target value, recommend reducing the carbohydrate ratio; The optimization generation module introduces a user feedback mechanism to improve the plan effect through iteration: The user submits feedback information through the APP, including food preferences (such as "likes sweet taste") and health feedback (such as "lack of satiety"); The system collects health data, analyzes the deviation between the actual effect and the expectation; Update the user profile according to the feedback data and adjust the weight: ; where, is the learning rate; Iteratively train based on the feedback data through a reinforcement learning model (such as DQN) to improve the next recommendation plan: ; where, is the value function of choosing action a in state s, and r is the immediate reward.

[0058] For example: Personalized diet optimization for diabetic users Input health data: Blood glucose value: 8.5 mmol / L; Weight: 80 kg; Goal: Control blood sugar and lose weight; Initial recommendation: Breakfast: 40g of oats, 200ml of milk, 10g of walnuts; Lunch: 80g of brown rice, 100g of steamed fish, 150g of green leafy vegetables; Dinner: 60g of purple sweet potatoes, 1 hard-boiled egg, 100g of chicken breast salad; User feedback: "Lack of satiety in breakfast"; Blood sugar fluctuations slightly exceed the expected value; Scheme optimization: Adjust breakfast to: 50g of oats, 200ml of milk, 15g of walnuts; Reduce the carbohydrate ratio for dinner: Purple sweet potatoes reduced to 50g, increase chicken breast to 120g.

[0059] Through the optimization generation module, the system realizes the closed-loop optimization from data collection, knowledge graph analysis to personalized recommendation scheme, improves the scientificity and adaptability of healthy diet suggestions, and at the same time supports multiple iterative optimizations to meet the dynamic needs of users.

[0060] In the embodiment of the present invention, multi-dimensional health data collected by a user's health data monitoring device is obtained and preprocessed; a multi-dimensional knowledge graph is constructed, and the multi-dimensional health data is mapped to corresponding nodes in the knowledge graph to obtain the user's real-time health data; based on the user's real-time health data and the analysis results of the knowledge graph, an artificial intelligence algorithm is used to dynamically generate a personalized target diet plan. The present invention deeply relies on the user's individual biological data, through precise real-time physical sign data collection (covering key indicators such as blood sugar level, blood pressure condition, body fat ratio, etc.), and combines advanced knowledge graph technology to comprehensively analyze the user's current physical state, and at the same time accurately predict potential health risks. This algorithm integrates three core modules: an efficient data collection module, an accurate knowledge graph matching module, and an intelligent optimization generation module. By widely compatible with various health data collection devices, the system can build a complex association model between diseases, foods and nutritional components, which, based on the profound foundation of the knowledge graph, ensures the comprehensiveness and accuracy of the analysis. The system uses the powerful computing power of the AI large model and the flexibility of the adaptive algorithm to flexibly and accurately dynamically adjust diet suggestions according to the user's real-time health data, not only supporting the user's short-term health improvement goals, but also helping them achieve long-term health plans in the long run.

[0061] The application scope of the present invention is wide, not only applicable to the refined needs of personalized health management, but also providing strong technical support for the effective prevention and control of chronic diseases and targeted nutritional interventions, demonstrating its great potential in promoting human health and well-being.

[0062] Refer to Figure 6 , the embodiment of the present invention provides a healthy diet recommendation device based on physiological data, including: The first module is used to obtain multi-dimensional health data collected by a user health data monitoring device and perform preprocessing; Exemplarily, the first module can be a data collection and processing module: This module integrates a variety of user health data monitoring devices (such as smart watches, blood glucose monitors, etc.), realizes real-time capture and preliminary processing of user physiological indicators, and ensures the accuracy and timeliness of the data.

[0063] The second module is used to construct a multi-dimensional knowledge graph, map the multi-dimensional health data to corresponding nodes in the knowledge graph, and obtain the real-time health data of the user; wherein, the real-time health data represents the user's health status and potential health risks; Exemplarily, the second module can be a knowledge graph intelligent matching module: We have constructed a professional knowledge graph covering extensive disease information, nutritional requirements, and food ingredients. By accurately mapping the user's health data to the corresponding nodes in the graph, the system can deeply analyze the user's health status and potential health risks, providing a scientific basis for personalized diet recommendations.

[0064] The third module is used to dynamically generate a personalized target diet plan based on the user's real-time health data and the analysis results of the knowledge graph using artificial intelligence algorithms.

[0065] Exemplarily, the third module can be: an AI optimized plan generation module: Relying on a powerful artificial intelligence large model and an adaptive optimization algorithm, this module can flexibly generate and adjust a personalized dynamic diet plan according to the user's short-term health goals and long-term health management goals, ensuring that the diet recommendations not only meet the user's needs but also fit the actual health status. Through interacting with the user, the system continuously iterates and optimizes the recommendation results, featuring high efficiency, flexibility, and accuracy, and is widely applicable to the fields of personalized health management, chronic disease prevention, and nutritional intervention.

[0066] See Figure 7 , an embodiment of the present invention provides a health diet recommendation system based on physiological data, including: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the health diet recommendation method based on physiological data as described in any one of the above.

[0067] It can be seen that the content in the above method embodiments is applicable to the system embodiments of the present invention. The functions specifically implemented by the system embodiments are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those in the above method embodiments.

[0068] On the other hand, an embodiment of the present invention provides a computer-readable storage medium storing a program executable by a processor. The program executable by the processor, when executed by the processor, is used to execute a method for recommending a healthy diet based on physiological data as described in any one of the above.

[0069] In addition, an embodiment of the present application also discloses a computer program product or a computer program. The computer program product or the computer program is stored in a computer-readable storage medium. The processor of the computer device can read the computer program from the computer-readable storage medium, and the processor executes the computer program, so that the computer device executes the above method. Similarly, the content in the above method embodiments is applicable to this storage medium embodiment. The functions specifically implemented by this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those in the above method embodiments.

[0070] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed to multiple network units. Some or all of the charging modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0071] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional charging modules / units in the systems and devices, can be implemented as software, firmware, hardware, and appropriate combinations thereof.

[0072] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above drawings are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0073] It should be understood that in this application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the association relationship of associated objects and indicates that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously. Here, A and B can be singular or plural. The character " / " generally indicates an "or" relationship between the associated objects before and after. "At least one (item) of the following" or a similar expression refers to any combination of these items, including any combination of single items or plural items. For example, at least one (item) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0074] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be in electrical, mechanical, or other forms.

[0075] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0076] In addition, the functional units in each embodiment of this application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0077] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media that can store programs, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0078] The preferred embodiments of the embodiments of this application have been described above with reference to the accompanying drawings, which does not limit the scope of rights of the embodiments of this application. Any modifications, equivalent replacements, and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of this application shall fall within the scope of rights of the embodiments of this application.

Claims

1. A healthy diet recommendation method based on physiological data, characterized in that: include: Obtain multi-dimensional health data collected by the user's health data monitoring device and perform pre-processing; Construct a multi-dimensional knowledge graph, map the multi-dimensional health data to corresponding nodes in the knowledge graph, and obtain the user's real-time health data; wherein the real-time health data represents the user's health status and potential health risks; Based on the user's real-time health data and knowledge graph analysis results, artificial intelligence algorithms are used to dynamically generate personalized target diet plans.

2. The method according to claim 1, characterized in that The obtaining of multi-dimensional health data collected by the user health data monitoring device and preprocessing includes: The time series data is smoothed based on the sliding window filtering method, and the formula is: ； in, is the original data at time t, and N is the window size.

3. The method according to claim 2, characterized in that The smoothing process of the time series data based on the sliding window filtering method includes: The smoothed data is normalized to the interval [0, 1], and the formula is as follows: ； in, and They are the first historical target value and the second historical target value of the indicator respectively.

4. The method according to claim 3, characterized in that The step of normalizing the smoothed data comprises: The outlier detection method is based on the triple standard deviation method, and the formula is as follows: like ,but is an outlier; in, and are the mean and standard deviation of the indicator respectively; Receive time series data from multiple data sources, where missing and / or inconsistent time points are aligned using interpolation, using the following formula: ； in, and are the timestamps of two adjacent data points.

5. The method according to claim 1, characterized in that: The construction of a multi-dimensional knowledge graph includes: Construct a multi-dimensional knowledge graph in the form of triples; Health and nutrition-related data are obtained from multiple data sources, and the processed data is stored in a graph database. The nodes and their relationships are represented in the form of an adjacency matrix. The formula is as follows: ； Among them, w ij is the relationship weight between node i and node j; it is determined whether there is a relationship between node i and node j. If there is a relationship between node i and node j, the element A[i,j] at the corresponding position in the adjacency matrix A is assigned a non-zero value w ij ; If there is no relationship between node i and node j, the element A[i,j] at the corresponding position in the adjacency matrix A is assigned a zero value w ij ; Natural language processing technology is used to extract text data, and the BERT or BioBERT model is used to perform entity recognition and relationship classification on the processed text data; the relationship weight formula is as follows: ； in: represents the relationship score based on the document frequency; Indicates the corrected score based on domain expert scoring or experimental data; is the weight factor, satisfying ; The extracted entities and their relationships are stored in a graph database using Neo4j, and a visualization interface is built based on the graph database to graphically display the entities and their relationship network so that it can realize the functions of searching, filtering and expanding the graph knowledge.

6. The method according to claim 5, characterized in that Mapping the multi-dimensional health data to corresponding nodes in the knowledge graph to obtain the user's real-time health data includes: User health indicators Map to knowledge graph nodes and calculate the matching score of each node: ； in, For Entity The overall score of For health indicators The weight of is the influence function of the health index value.

7. The method according to claim 6, characterized in that The method of dynamically generating a personalized target diet plan based on the user's real-time health data and knowledge graph analysis results using an artificial intelligence algorithm includes: Based on user health data and goals, define the optimization objective function: ； in: health scores for short-term goal achievement; Score trends toward long-term goal achievement; Rate the satisfaction of user preferences; is a weight factor, which is adjusted dynamically according to user priority.

8. A healthy diet recommendation device based on physiological data, characterized in that: include: The first module is used to obtain multi-dimensional health data collected by the user's health data monitoring device and perform pre-processing; The second module is used to construct a multi-dimensional knowledge graph, map the multi-dimensional health data to the corresponding nodes in the knowledge graph, and obtain the user's real-time health data; wherein the real-time health data represents the user's health status and potential health risks; The third module is used to dynamically generate personalized target diet plans based on the user's real-time health data and knowledge graph analysis results using artificial intelligence algorithms.

9. A healthy diet recommendation system based on physiological data, characterized in that: The system comprises: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the healthy diet recommendation method based on physiological data as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to execute a healthy diet recommendation method based on physiological data as described in any one of claims 1-7 when executed by the processor.

Images