Foot health management method based on decision tree model and intelligent socks

Abstract

The invention relates to the technical field of artificial intelligence, intelligent wearing systems and the like, and provides a foot health management method based on a decision tree model and intelligent socks. The multi-category foot health associated data comprises intelligent sock internal temperature data and foot motion data, inputting the intelligent sock internal temperature data and the foot motion data into a trained decision tree model to perform foot health state classification so as to obtain a foot health state classification result, and according to the foot health state classification result, determining the foot health state of the intelligent sock according to the foot health state classification result. A foot health prediction report is generated and provided for a wearing user of the intelligent sock, correlation analysis of the heating temperature is carried out on a foot health state classification result to obtain a foot health temperature adjusting signal, the heating temperature of a heating assembly in the intelligent sock is adjusted, and therefore foot health management is provided for the wearing user of the intelligent sock; and the use experience of the intelligent sock product is improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of artificial intelligence, intelligent wearable systems, and in particular to a foot health management method and intelligent socks based on a decision tree model. Background Art

[0002] With the development of smart health technology, traditional smart wearable products, such as smart watches and smart bracelets, have integrated functions such as temperature, motion monitoring, and user physiological characteristics monitoring, providing users with more and more intelligent services. However, traditional smart wearable products are not commonly used items, and the user usage rate is not high. For this reason, new smart wearable products such as smart clothing, smart gloves, and smart socks have appeared on the market. Taking smart socks as an example, the current smart socks can monitor the temperature of the user's feet, and then control the heating of the heating components through the preset temperature adjustment gear, giving the user a comfortable wearing experience. However, for the intelligent management of the user's foot health, the current smart socks still do not provide too much functional support, which is not conducive to the intelligent management of the user's foot health, and the user experience of smart socks products needs to be further improved.

[0003] To sum up, existing smart socks have technical problems such as the lack of foot health management function and the user experience of smart socks products needs to be improved. Summary of the invention

[0004] In view of the deficiencies in the above-mentioned prior art, the present invention provides a foot health management method and smart socks based on a decision tree model, so as to provide foot health management for users who wear the smart socks and enhance the user experience of the smart socks products.

[0005] In a first aspect, the present invention provides a foot health management method based on a decision tree model, comprising: Acquire multi-category foot health-related data monitored by the smart socks sensor group when the user uses the smart socks, wherein the multi-category foot health-related data includes temperature data inside the smart socks and foot movement data; Inputting the internal temperature data of the smart socks and the foot movement data into the trained decision tree model to classify the foot health status, so as to obtain the foot health status classification result; Based on the foot health status classification results, a foot health prediction report is generated and provided to the wearer of the smart socks, and the foot health status classification results are subjected to a correlation analysis of heating temperature to obtain a foot health temperature adjustment signal to adjust the heating temperature of the heating component in the smart socks.

[0006] In a second aspect, the present invention provides a smart sock, comprising a main controller, wherein the main controller runs the above-mentioned foot health management method based on the decision tree model.

[0007] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a foot health management method and smart socks based on a decision tree model. The method obtains multi-category foot health-related data monitored by a smart socks sensor group when a user uses the smart socks, and the multi-category foot health-related data includes smart socks internal temperature data and foot movement data. The smart socks internal temperature data and foot movement data are input into a trained decision tree model to classify the foot health status to obtain a foot health status classification result. According to the foot health status classification result, a foot health prediction report is generated and provided to the wearer of the smart socks. The foot health status classification result is subjected to a correlation analysis of the heating temperature to obtain a foot health temperature adjustment signal, and the heating temperature of the heating component in the smart socks is adjusted, thereby providing foot health management for the wearer of the smart socks and improving the wearer's experience of the smart socks product. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention and do not constitute an improper limitation of the present invention. Some specific embodiments of the present invention will be described in detail in an exemplary and non-restrictive manner with reference to the drawings. The same reference numerals in the drawings indicate the same or similar components or parts. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 It is a schematic diagram of the architecture of a foot health management method based on a decision tree model according to an embodiment of the present invention; Figure 2 It is a schematic diagram of an architecture for connecting and communicating smart socks with a smart user terminal according to an embodiment of the present invention. DETAILED DESCRIPTION

[0009] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only an embodiment of a part of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0010] Embodiment 1 See also Figure 1-Figure 2 , this embodiment provides a foot health management method based on a decision tree model, comprising the following steps: S101, obtaining multi-category foot health-related data monitored by the smart socks sensor group when the user uses the smart socks, wherein the multi-category foot health-related data includes temperature data inside the smart socks and foot motion data; wherein the temperature data inside the smart socks can be provided by monitoring the temperature sensor in the smart socks sensor group, and the foot motion data can be provided by monitoring the motion state monitoring sensor in the smart socks sensor group. The temperature sensor can be installed in the smart socks near the middle of the user's instep, and the motion state monitoring sensor can be installed at the sock opening of the smart socks.

[0011] S102, inputting the internal temperature data of the smart socks and the foot movement data into the trained decision tree model to classify the foot health status, so as to obtain the foot health status classification result; S103. Generate a foot health prediction report based on the foot health status classification result and provide it to the wearer of the smart socks, and perform a correlation analysis of the heating temperature on the foot health status classification result to obtain a foot health temperature adjustment signal to adjust the heating temperature of the heating component in the smart socks.

[0012] It should be noted that temperature data can reflect whether the foot is at risk of overheating or overcooling, and motion data (such as acceleration, number of steps or activity intensity) can identify the user's current activity status and intensity. It is difficult to comprehensively evaluate foot health based on a single data. For example, only temperature data may not be able to distinguish whether the temperature rise is caused by activity or the environment is overheated; only motion data cannot determine whether the foot is overheated. In step S101, through the comprehensive monitoring of temperature and motion status, deeper associations can be found (such as long-term overheating and large amounts of exercise may cause foot fatigue, increased risk of fungal infection, etc.). In order to use the decision tree model to classify the health status of the feet in the future, accurate and multi-dimensional foot health-related data must be available as input. Step S101 lays a complete data foundation for subsequent processing. In step S102, the internal temperature data of the smart socks and the foot motion data are input into the trained decision tree model to classify the health status of the feet to obtain the classification result of the health status of the feet. The decision tree performs well on structured data (such as temperature, exercise intensity, step frequency, etc.) and has good interpretability, which is suitable for wearable devices (smart socks) to interpret and verify real-time prediction results. Smart socks usually need to make quick decisions in a limited resource environment. The decision tree can quickly produce classification results through simple feature comparison, meeting the requirements of wearable devices for computing efficiency and latency. In step S103, based on the classification results of the foot health status, a foot health prediction report is generated and provided to the wearer of the smart socks, and the foot health status classification results are analyzed for association with the heating temperature to obtain a foot health temperature adjustment signal to adjust the heating temperature of the heating component in the smart socks. On the one hand, the smart socks can automatically adjust the temperature through the foot health status classification results, and on the other hand, the user can get a health prediction report to understand the condition of their own feet, thereby realizing user foot health management through smart socks and improving the user's experience of using smart socks products.

[0013] In some preferred embodiments, when training the decision tree model, it includes: collecting the internal temperature data of the smart socks and the foot movement data and then performing data cleaning to obtain the cleaned sensor data; binding the foot health status label to the cleaned sensor data to obtain the foot health status classification sample for training the decision tree model; training the decision tree model through the foot health status classification sample to obtain the trained decision tree model. It should be noted that the original sensor data often contains noise, outliers or missing values. If it is not cleaned, it will cause the decision tree model to be biased or overfitted if it is directly used for training. The data cleaning process can effectively eliminate extreme data caused by sensor failures, etc., and provide an accurate and stable data basis for subsequent model training. In addition, the decision tree is a supervised learning model that requires training samples with "correct answers" (i.e., with health status labels) to learn how to classify new data. In this embodiment, the foot health status label is bound to the cleaned sensor data to obtain foot health status classification samples for training the decision tree model. By training the decision tree model with the foot health status classification samples, the model can learn the foot health status corresponding to each sensor data, thereby making accurate judgments in actual use.

[0014] In some further embodiments, binding a foot health status label to the cleaned sensor data comprises: analyzing the smart sock internal temperature data and the foot motion data in the cleaned sensor data, and binding a foot health risk label of excessive exercise to the cleaned sensor data when the internal temperature of the smart sock does not exceed the temperature risk threshold and the motion intensity of the user's foot exceeds the motion intensity risk threshold; binding a foot health risk label of overheated feet to the cleaned sensor data when the internal temperature of the smart sock exceeds the temperature risk threshold and the motion intensity of the user's foot does not exceed the motion intensity risk threshold; binding a foot health risk label of overheated feet to the cleaned sensor data when the internal temperature of the smart sock exceeds the temperature risk threshold and the motion intensity of the user's foot exceeds the motion intensity risk threshold; binding a foot health risk label of overheated feet and excessive exercise to the cleaned sensor data; and binding a foot health risk label of low temperature and lack of activity to the cleaned sensor data when the internal temperature of the smart sock is lower than the temperature safety lower threshold and the motion intensity of the user's foot is lower than the motion intensity safety lower threshold. It should be noted that even if the temperature is normal, if the exercise intensity is too high, the feet are prone to fatigue, blisters or other overuse injuries, so they need to be marked separately as excessive exercise. In addition, the user may be in a hot environment or the heating level of the socks is too high, causing the feet to overheat; at this time, even if the exercise intensity is not high, there will be a risk of stuffy feet, fungal growth or skin damage. If only the intensity of exercise is judged, the high temperature caused by the environment or heating may be ignored; therefore, setting a separate foot overheating label can more accurately distinguish the factors that cause overheating. In addition, when the user is in a state of high temperature and high-intensity exercise at the same time, the feet are very prone to blisters, inflammation, and even abrasions; it is a "dangerous superposition" and requires a separate and clear label. Through combined labels, more abundant label samples can be provided for decision tree training, improving the model's ability to identify high-risk scenarios. In addition, too low foot temperature may cause poor blood circulation, numbness and other problems; if the amount of exercise is insufficient at this time, it is more likely to cause foot stiffness or other chronic injury risks.

[0015] In some preferred embodiments, according to the classification result of the foot health status, a foot health prediction report is generated and provided to the wearer of the smart socks, including: obtaining the classification result of the foot health status and parsing it to obtain the classification of foot health temperature; generating a foot health prediction report according to the classification of foot health temperature and providing it to the wearer of the smart socks. In addition, according to the classification result of the foot health status, a foot health prediction report is generated and provided to the wearer of the smart socks, including: obtaining the classification result of the foot health status and parsing it to obtain the classification of foot health movement status; generating a foot health prediction report according to the classification of foot health movement status and providing it to the wearer of the smart socks. It should be noted that for foot health, too high or too low temperature will directly affect foot comfort and health risks (such as fungal infection, frostbite, etc.). Parsing the temperature classification separately and generating a report can allow users to understand their current foot health status more intuitively. In addition, high-intensity exercise may cause foot fatigue, wear, blisters, etc.; lack of exercise may cause poor blood circulation, stiff feet, etc. Analyzing the classification of exercise status helps users understand the impact of the current activity on foot health.

[0016] In some preferred embodiments, when providing the foot health prediction report to the user wearing the smart socks, it includes: connecting and communicating with the smart user terminal of the user wearing the smart socks, responding to the report viewing request initiated by the smart user terminal, and transmitting the foot health prediction report to the smart user terminal. It should be noted that smart user terminals (such as smart phones, tablets or other mobile devices) are the most commonly used personal devices by users. Transmitting the foot health prediction report to the user terminal allows the user to easily view it through the interface or application anytime and anywhere, and make corresponding adjustments based on the report (such as changing the heating level of the socks, etc.).

[0017] In some preferred embodiments, when generating a foot health prediction report based on the foot health status classification result, it includes: obtaining the foot health status classification result and inputting the obtained foot health status classification result into a foot health prediction model, and the foot health prediction model analyzes the input foot health status classification result to generate the foot health prediction report. It should be noted that after the foot health status classification has been obtained, the current foot condition can be qualitatively characterized, such as "high temperature", "excessive exercise", "normal", etc. However, the user also needs a comprehensive foot health conclusion for reference. In this embodiment, the foot health prediction model analyzes the input foot health status classification result to generate the foot health prediction report, thereby realizing the hierarchical processing of the model and improving the processing accuracy of the model.

[0018] In some preferred embodiments, the foot health status classification results include foot health temperature classification and foot health movement status classification, and the foot health prediction model uses a trained gradient boosting decision tree model, and the trained gradient boosting decision tree model analyzes the input foot health temperature classification and the foot health movement status classification to generate the foot health prediction report. It should be noted that gradient boosting decision trees (such as XGBoost, LightGBM, CatBoost, etc.) usually perform well in processing structured features, and have fast training and prediction speeds. When the input is discrete or numerical foot health temperature classification and the foot health movement status classification features, the gradient boosting decision tree can perform efficient predictions and output accurate prediction results.

[0019] In some preferred embodiments, when training the gradient boosting decision tree model, it includes: binding the foot health temperature classification and the foot health movement state classification with a foot health prediction label to obtain a foot health prediction training sample for training the gradient boosting decision tree model; training the gradient boosting decision tree model with the foot health prediction training sample to obtain the trained gradient boosting decision tree model. It should be noted that the gradient boosting decision tree, as a supervised learning model, requires "features + annotation labels" for training. In this embodiment, the foot health temperature classification and the foot health movement state classification are input features, and the foot health prediction label is the target to be learned by the model (such as "risk score" or "health outcome", etc.).

[0020] In some further embodiments, when binding the foot health temperature classification and the foot health exercise state classification with foot health prediction labels, it includes: when the user's foot exercises excessively, binding the foot exercise excessively with a prediction label of easy foot strain; when the user's foot is overheated, binding the foot overheating with a prediction label of easy breeding of fungi or bacteria; when the user's foot is overheated and exercises excessively, binding the foot overheating and excessive exercise with a prediction label of easy foot strain, easy breeding of fungi or bacteria; when the user's foot is cold and lacks activity, binding the foot cold and lack of activity with a prediction label of poor foot blood circulation. It should be noted that after the decision tree model classifies the foot health state, the classification results of the foot health state are obtained, such as overheated feet, overheated feet and excessive exercise, cold feet and lack of activity, etc. The gradient boosting decision tree model then performs a higher-level foot health prediction based on the foot health state classification results, such as easy breeding of fungi or bacteria, easy strain, poor blood circulation, etc. When the feet are overheated, the hot and stuffy environment makes it easy for bacteria or fungi to multiply, causing athlete's foot or infectious problems. When the feet are overheated and exercised excessively, there is a risk of microbial reproduction caused by high temperature, and there is also a risk of muscle and joint strain caused by high exercise load. When the feet are cold and inactive, the cold environment plus lack of exercise can easily lead to insufficient local blood flow. In this embodiment, when the user's feet exercise excessively, the prediction label of easy foot strain is bound to the excessive foot exercise; when the user's feet are overheated, the prediction label of easy breeding of fungi or bacteria is bound to the overheated feet; when the user's feet are overheated and exercise excessively, the prediction labels of easy foot strain, easy breeding of fungi or bacteria are bound to the overheated feet and excessive exercise; when the user's feet are cold and lack activity, the prediction label of poor foot blood circulation is bound to the cold feet and lack of activity, thereby obtaining foot health prediction training samples for training the gradient boosting decision tree model, training the gradient boosting decision tree model with the foot health prediction training samples to obtain the trained gradient boosting decision tree model, so that the gradient boosting decision tree model can achieve comprehensive and detailed foot health prediction.

[0021] In some preferred embodiments, when the foot health status classification result is subjected to the correlation analysis of the heating temperature, it includes: analyzing the correlation relationship between the foot health movement status classification and the foot health temperature classification in the foot health status classification result, obtaining the foot health temperature adjustment signal according to the correlation relationship between the foot health movement status classification and the foot health temperature classification, and adjusting the heating temperature of the heating component in the smart socks. It should be noted that if the sock temperature is adjusted only according to one of the high temperature or excessive exercise, the inherent correlation between the two may be ignored, which is not conducive to the accurate and comprehensive health management of the feet. In this embodiment, by analyzing the correlation relationship between the foot health movement status classification and the foot health temperature classification in the foot health status classification result, according to the correlation relationship between the foot health movement status classification and the foot health temperature classification, a foot health temperature adjustment signal is obtained, and the heating temperature of the heating component in the smart socks is adjusted, thereby dynamically adjusting the temperature in multiple dimensions, better protecting the health and safety of the user's feet, and avoiding invalid or excessive temperature control operations, bringing a more comfortable wearing experience to the user.

[0022] In some further embodiments, when adjusting the heating temperature of the heating component in the smart socks, if the user's feet are overheated and the amount of exercise is normal, it is determined that the overheating of the user's feet is not related to the amount of exercise of the user's feet, and the heating temperature of the heating component in the smart socks is reduced to a preset comfortable heating temperature range; if the user's feet are overheated and the amount of exercise is excessive, it is determined that the overheating of the user's feet is related to the amount of exercise of the user's feet, and the heating component in the smart socks is controlled to stop heating; if the user's feet are cold and lack activity, it is determined that the low temperature of the user's feet is related to the amount of exercise of the user's feet, and the heating temperature of the heating component in the smart socks is increased to a preset rapid heating temperature range; the temperature value of the rapid heating temperature range is higher than the temperature value of the comfortable heating temperature range. It should be noted that when the foot temperature is too high and the amount of exercise is normal, it means that the high temperature mainly comes from the heat of the smart socks themselves or the external environment temperature, rather than the extra heat generated by the large amount of exercise. Then it can be moderately cooled to the comfortable heating temperature range. When the foot temperature is too high and the amount of exercise is excessive, it means that the user's feet are under the dual load of high temperature and large amount of exercise, which may cause serious health risks. In this case, completely stopping the heating can quickly relieve the pressure on the feet and avoid aggravation of problems such as blisters and fungal infections. When the foot temperature is low and there is a lack of activity, it is determined that the low temperature is related to insufficient exercise. In this type of scenario, a simple moderate warming or comfortable temperature cannot quickly solve problems such as stiff feet and poor blood circulation. Therefore, the heating temperature needs to be increased to the rapid heating temperature range (higher than the comfort range) to increase the foot temperature as soon as possible to help improve circulation. In this embodiment, whether the overheating of the feet is related to the amount of exercise and whether the low temperature of the feet is related to the amount of exercise are included in the temperature control logic, so as to more accurately determine the cause of overheating or low temperature, and adopt a more targeted temperature control strategy (lowering to the comfort zone, stopping heating, or increasing to the rapid heating temperature) to achieve timely stop-loss in high-risk scenarios or rapid improvement in low-temperature scenarios, which not only ensures the health of the feet, but also improves user comfort and smart experience.

[0023] Embodiment 2 See also Figure 1-Figure 2The present embodiment provides a smart sock, comprising a main controller, wherein the main controller runs the foot health management method based on the decision tree model described in any of the above embodiments, by acquiring multi-category foot health-related data monitored by the smart sock sensor group when the user uses the smart socks, wherein the multi-category foot health-related data includes internal temperature data of the smart socks and foot movement data, and inputting the internal temperature data of the smart socks and the foot movement data into a trained decision tree model to classify the foot health status, so as to obtain a foot health status classification result, and generating a foot health prediction report according to the foot health status classification result and providing it to the wearer of the smart socks, and performing a correlation analysis of the heating temperature on the foot health status classification result to obtain a foot health temperature adjustment signal, and adjusting the heating temperature of the heating component in the smart socks, thereby providing foot health management for the wearer of the smart socks and improving the wearer's experience of the smart socks product.

[0024] It should be pointed out that the above embodiments are only preferred specific implementation modes of the present invention, and the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. The protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A foot health management method based on a decision tree model, characterized in that: include: Acquire multi-category foot health-related data monitored by the smart socks sensor group when the user uses the smart socks, wherein the multi-category foot health-related data includes temperature data inside the smart socks and foot movement data; Inputting the internal temperature data of the smart socks and the foot movement data into the trained decision tree model to classify the foot health status, so as to obtain the foot health status classification result; Based on the foot health status classification results, a foot health prediction report is generated and provided to the wearer of the smart socks, and the foot health status classification results are subjected to a correlation analysis of heating temperature to obtain a foot health temperature adjustment signal to adjust the heating temperature of the heating component in the smart socks.

2. The foot health management method based on the decision tree model according to claim 1, characterized in that: When training the decision tree model, it includes: collecting the internal temperature data of the smart socks and the foot movement data, performing data cleaning to obtain the cleaned sensor data; binding the foot health status label to the cleaned sensor data to obtain the foot health status classification sample for training the decision tree model; training the decision tree model through the foot health status classification sample to obtain the trained decision tree model.

3. The foot health management method based on the decision tree model according to claim 1, characterized in that: Based on the foot health status classification result, a foot health prediction report is generated and provided to the user wearing the smart socks, including: obtaining the foot health status classification result and analyzing it to obtain a foot health temperature classification; based on the foot health temperature classification, a foot health prediction report is generated and provided to the user wearing the smart socks.

4. The foot health management method based on the decision tree model according to claim 1, characterized in that: According to the foot health status classification result, a foot health prediction report is generated and provided to the wearer of the smart socks, including: obtaining the foot health status classification result and analyzing it to obtain the foot health movement status classification; according to the foot health movement status classification, a foot health prediction report is generated and provided to the wearer of the smart socks.

5. The foot health management method based on the decision tree model according to claim 1, characterized in that: When providing the foot health prediction report to the wearer of the smart socks, it includes: connecting and communicating with the smart user terminal of the wearer of the smart socks, responding to a report viewing request initiated by the smart user terminal, and transmitting the foot health prediction report to the smart user terminal.

6. The foot health management method based on the decision tree model according to claim 1, characterized in that: When generating a foot health prediction report according to the foot health status classification result, it includes: obtaining the foot health status classification result and inputting the obtained foot health status classification result into a foot health prediction model, the foot health prediction model analyzes the input foot health status classification result to generate the foot health prediction report.

7. The foot health management method based on the decision tree model according to claim 6, characterized in that: The foot health status classification results include foot health temperature classification and foot health movement status classification. The foot health prediction model adopts a trained gradient boosting decision tree model. The trained gradient boosting decision tree model analyzes the input foot health temperature classification and the foot health movement status classification to generate the foot health prediction report.

8. The foot health management method based on the decision tree model according to claim 7, characterized in that: When training the gradient boosting decision tree model, it includes: binding the foot health prediction label to the foot health temperature classification and the foot health movement state classification to obtain the foot health prediction training sample for training the gradient boosting decision tree model; training the gradient boosting decision tree model through the foot health prediction training sample to obtain the trained gradient boosting decision tree model.

9. The foot health management method based on the decision tree model according to claim 1, characterized in that: When performing a correlation analysis of the heating temperature on the foot health status classification results, it includes: analyzing the correlation between the foot health movement status classification and the foot health temperature classification in the foot health status classification results, and obtaining a foot health temperature adjustment signal based on the correlation between the foot health movement status classification and the foot health temperature classification, and adjusting the heating temperature of the heating component in the smart socks.

10. A smart sock, comprising a main controller, characterized in that: The main controller runs the foot health management method based on the decision tree model as described in any one of claims 1-9.

Images