Intelligent management system and method for dynamically adjusting food-grade capsule taking time

Abstract

The present invention discloses an intelligent management system and method for dynamically adjusting food-grade capsule taking time, which specifically relates to the field of personalized nutrition management and is used to solve the problem of dynamically optimizing food-grade capsule intake time based on individual physiological data. The method realizes personalized food-grade capsule taking time regulation by establishing capsule characteristic data analysis, physiological characteristic extraction and matching, thereby improving the bioavailability of capsules and the effectiveness of user health intervention; utilizes multidimensional physiological data collected by intelligent wearable devices to extract physiological characteristic control indicators with actual correlation, and accurately matches them with capsule taking key parameter sets; adopts mutual information measurement and regression model to perform quantitative analysis, and through a dynamic tuning strategy, updates the taking time in real time according to the individual physiological state, so that the taking time can flexibly adapt to the changes in the user's physiological rhythm, ensure the optimal state of ingredient intake and absorption, thereby realizing fast, flexible and highly reliable personalized capsule taking management.

Description

Technical Field

[0001] The present invention relates to the field of personalized nutrition management, and more specifically, to an intelligent management system and method for dynamically adjusting the taking time of food-grade capsules. Background Art

[0002] In the application scenarios of modern health management and personalized nutritional supplementation, food-grade capsules are an important form of nutritional supplements and are widely used to improve individual health and prevent diseases. With the popularization of smart wearable devices (such as smart watches), users can obtain multi-dimensional physiological data in real time, including heart rate variability, skin galvanic response, exercise volume, and work and rest patterns. The continuous monitoring and analysis of these data provide a technical basis for optimizing the intake time of capsules, aiming to achieve the best absorption and utilization of nutrients by accurately matching individual physiological states with capsule characteristics, thereby improving the effectiveness and personalization of health interventions.

[0003] However, existing technologies still face significant challenges in accurately matching and dynamically adjusting multi-source physiological data with key parameters for capsule administration. First, the physiological data collected by smart devices are multi-dimensional and complex, and there is a lack of efficient algorithms to screen out core indicators that are highly correlated with the effect of capsule intake from massive data. Secondly, it is difficult for existing association analysis methods to effectively integrate key parameters for administration (such as ingredient release curves, optimal absorption windows, and contraindications) provided by capsule manufacturers with real-time physiological data, resulting in the recommendation of administration time still relying on experience or static rules, and unable to achieve dynamic and personalized adjustments. In addition, the existing models are insufficient in processing real-time data streams and making quick decisions, which limits the response speed and accuracy of intelligent management systems in practical applications. In response to the above problems, it is urgent to develop a set of systematic technical processes that can efficiently analyze key parameters for capsule administration, accurately screen and match individual physiological characteristics, and use advanced statistical models for quantitative analysis, so as to achieve dynamic and intelligent management of food-grade capsule administration time. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide an intelligent management system and method for dynamically adjusting the taking time of food-grade capsules. By establishing a systematic technical process of capsule characteristic data analysis, physiological characteristic extraction and matching, and dynamic optimization of taking time, accurate regulation of personalized food-grade capsule taking time is achieved, and the bioavailability of capsules and the effectiveness of user health intervention are significantly improved; multi-dimensional physiological data collected by smart wearable devices are used to extract physiological characteristic control indicators with practical relevance, and accurately match them with the key parameter set for capsule taking; mutual information measurement and regression model are used for quantitative analysis, and through a dynamic tuning strategy, the taking time is updated in real time according to the individual physiological state, so that the taking time can flexibly adapt to the changes in the user's physiological rhythm, ensuring the optimal state of ingredient intake and absorption, thereby achieving fast, flexible and highly reliable personalized capsule taking management to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for dynamically adjusting the time of taking food-grade capsules and intelligently managing the time of taking the food-grade capsules comprises the following steps:

[0007] S1: parse and structure the capsule characteristic data provided by the manufacturer, and obtain and store the capsule taking key parameter set for capsule taking;

[0008] S2: Using the multi-dimensional physiological signals collected by the smartwatch, after data cleaning, time series alignment and feature dimension reduction, a set of physiological feature candidate indicators that can potentially be used to infer the timing of capsule taking is generated;

[0009] S3: Compare the candidate physiological characteristic indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and select the physiological characteristic control indicators that meet the capsule taking requirements from the candidate physiological characteristic indicators through correlation analysis and mutual information measurement;

[0010] S4: Input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the judgment quantitative value, and quantitatively evaluate the suitability of the current taking time point;

[0011] S5: Based on the judgment result of the determination quantization value, if the current time period does not meet the ideal conditions, the corresponding time period is marked and a dynamic tuning program is entered to seek a better taking time recommendation;

[0012] S6: For the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set for capsule administration are used to re-optimize the time, and the recommended administration time point with higher bioavailability potential is selected from the adjustable interval.

[0013] In a preferred embodiment, step S1 includes the following contents:

[0014] First, the source format of the capsule characteristic data provided by the manufacturer is parsed to construct a parameter subset;

[0015] In the data analysis phase, the initial data is accurately differentiated according to the following parameter dimensions:

[0016] Absorption period parameters: including the starting time and ending time of the capsule, described in hours as the basic unit, recorded as the starting point of the absorption period End of absorption period ;

[0017] Taboo interval parameter: contains certain physiological state intervals that must be avoided during the use of the drug, expressed as a specific upper and lower limit range, recorded as the taboo lower limit Taboo Cap ;

[0018] Ingredient release parameters: describe the release rate and absorption peak time of a specific ingredient in the capsule under ideal conditions, and release the peak time as a time scalar Carry out calibration;

[0019] Absorption gain parameter: used to indicate the increase in the absorption efficiency of the capsule under specific physiological conditions, recorded as the absorption gain factor ;

[0020] All the parameters obtained by analysis are stored in order to form the following vector-based key parameter set for capsule administration: ; Then, the capsule administration key parameter set is checked for consistency and format converted.

[0021] In a preferred embodiment, step S2 includes the following contents:

[0022] Firstly, data cleaning and limiting processing are performed on each type of data in the multidimensional physiological signals of the smart watch, and the records that are out of the physiological reasonable interval are eliminated; the signals with inconsistent time series are unified to the same time base through linear interpolation method to obtain multiple physiological data sequences that can be compared under the same time axis; the physiological sequences obtained after alignment are assumed to be:

[0023] : At the time The heart rate parameter sequence below, is the discrete time index;

[0024] : At the time The skin conductance parameter sequence under

[0025] : At the time The sequence of movement rhythm parameters under

[0026] Then, a variety of feature functions that can be used to evaluate the fluctuation characteristics of individual physiological states are extracted from the aligned sequences; multidimensional discrete jump degree measurement and interval consistency evaluation are used in the feature extraction process; For example, define the jump degree function It is used to measure the cumulative average of absolute differences between adjacent samples, thereby evaluating the magnitude of signal changes over time. and definition and , so as to obtain the changes in skin conductance and movement rhythm;

[0027] After obtaining these variation indexes, constraints are set for each feature sequence: the variation must fall within a specific interval to ensure that the corresponding feature is neither too stable nor extremely drastic. Only features whose variation meets the requirements of the corresponding interval are included in the set of candidate physiological feature indicators;

[0028] Among the features that meet the degree of variation condition, interval consistency evaluation is also performed for each signal;

[0029] Finally, through this multiple conditional constraints, a set of information parameters that meet the dynamic range restrictions, have an effective change pattern and do not violate physiological rationality are extracted from the original multidimensional signal and recorded as a set of candidate physiological feature indicators.

[0030] In a preferred embodiment, among the features that meet the variation condition, interval consistency evaluation is also performed on each signal, and the specific steps are as follows:

[0031] In the right When three sequences are screened for physiological laws, a fixed length is first determined is the analysis window, and Each sequence is divided into multiple continuous sub-segments for the step size; for each sub-segment, its data point sequence is checked in turn to see whether it shows a monotonically increasing or monotonically decreasing trend. If there is no rated number of repeated changes in the direction of the data value in the entire sub-segment, the corresponding sub-segment is counted as a local monotonic segment; then the proportion of the local monotonic segment in the current sequence relative to all sub-segments is counted If the ratio If the value is lower than the preset lower limit, it indicates that the corresponding sequence lacks stable physiological regulation characteristics within the preset window and has multiple fluctuations and repetitions, which is regarded as a characteristic sequence with too high a disorder level. The corresponding sequence is thus eliminated and not included in the set of candidate physiological characteristic indicators.

[0032] In a preferred embodiment, step S3 includes the following contents:

[0033] The time period on the time axis that meets the absorption conditions and does not trigger the taboo interval is defined as the ideal condition subset ; The ideal condition subset represents an ideal data set within the appropriate time interval and the physiological state falls outside the taboo interval; for each feature sequence in the candidate physiological feature index Do the following:

[0034] a1. Mutual information measurement analysis: Each feature sequence is regarded as a continuous numerical signal, and the feature sequence is discretized into value intervals to obtain the probability distribution ; At the same time, construct a binary indicator sequence corresponding to the ideal condition subset :when The conditions that meet the definition of the ideal condition subset are ,otherwise , thus obtaining The joint distribution of ; under this condition, calculate and The mutual information value of for: ;in for and The joint probability of for The probability of a single for The probability of a single Indicates Candidate indicators of physiological characteristics, Is a binary indicator sequence used to mark the time moment Whether the ideal absorption conditions defined by the key parameter set for capsule administration are met; when the mutual information value is higher than the set threshold, it means that the corresponding physiological characteristics have a significant information correlation with the ideal condition subset when it appears, and are preliminarily selected as candidates for physiological characteristic control indicators;

[0035] a2. Correlation test of release peak time: In the subset of candidate physiological characteristics that have passed the mutual information test, in order to screen out the characteristics that are closely related to the release peak time, The local extreme point information of is compared with the position of the release peak moment; when When a high proportion of local maximum values ​​appear within the range, the corresponding proportion is defined as ;in, Indicates Physiological characteristic candidate indicators In the moment The specific value of A time window range representing the critical release peak moment of the capsule, Indicates the deviation tolerance of the peak moment, which is a positive number used to define a time interval to allow time fluctuations within a certain range; Indicates Physiological characteristic candidate indicators In the time window The number of local maxima in ; Indicates Physiological characteristic candidate indicators The total number of local maxima in the entire time series; if the corresponding ratio is higher than the peak threshold, it means that the corresponding feature presents a stable and significant peak behavior of the physiological signal in the period close to the peak release moment, confirming its consistency with the optimal absorption rhythm of the capsule;

[0036] Through the mathematical measurement and screening strategies of step a1 and step a2, the features that meet the standards of mutual information and local peak location are recorded as member elements of the physiological feature control index.

[0037] In a preferred embodiment, step S4 includes the following contents:

[0038] First, obtain the characteristic sequence set based on the physiological characteristic control index At the same time, parameters that can quantitatively describe the characteristics of the taking window are extracted from the key parameters of capsule taking, including the starting point of the absorption period, the end point of the absorption period, the upper limit of the taboo, the lower limit of the taboo, the peak release time and the absorption gain factor; all time parameters are normalized based on a unified time reference;

[0039] Defined in time The linear combination expression of for: ;in, is the model intercept term; For the corresponding The model coefficients of Through The normalized deviation from the peak release moment within a suitable time interval is measured; is the influence coefficient of absorption gain factor on the judgment; The taboo interval is then measured symmetrically to measure the risk ratio of the current state inside and outside the taboo interval, thereby adjusting the sensitivity of the judgment result to potential taboo conditions; After that, it is mapped to the (0,1) interval through the Logistic function to obtain the judgment quantization value .

[0040] In a preferred embodiment, step S5 includes the following contents:

[0041] First, set two threshold parameters:

[0042] Used to identify periods of high suitability; Used to identify undesirable conditions; in actual operation, when receiving the output judgment quantization value After the sequence, for each time moment Perform the following decision test:

[0043] like , it indicates that the physiological characteristic control index and the capsule taking key parameter set have a high matching degree in the corresponding period, and no adjustment is required. The corresponding period is recorded as the passing interval;

[0044] like , it means that although this period has not completely reached the ideal standard, it still maintains a certain degree of suitability. It will not be marked for the time being, and only the judgment quantification value of the corresponding period will be recorded for reference in the later fine-tuning;

[0045] like , it means that there is a risk of significant deviation from the ideal conditions during this period, and it is necessary to enter the subsequent optimization process. The corresponding time period is marked as the re-optimization period;

[0046] All time moments that are judged to need further optimization The corresponding physiological characteristics and control index characteristic values ​​are classified and stored, while retaining the mapping relationship with the key information of the capsule taking key parameter set.

[0047] In a preferred embodiment, step S6 includes the following contents:

[0048] First, based on the marked re-optimization period, the adjustable interval is extracted from the key parameter set of capsule administration, that is, the candidate time set within the appropriate time interval and does not trigger the taboo interval ; For each candidate time point in the candidate time set Call the physiological characteristic control index again to recalculate the candidate time point The quantitative value of , The calculation is still based on the previously trained Logistic regression model, but the input time is changed to ; In this way, all candidate moments within the adjustable interval are obtained distributed.

[0049] In a preferred embodiment, step S6 further includes the following contents:

[0050] The time point with the greatest potential for bioavailability is selected from these candidate moments, taking into account the peak release moment of the capsule and the position of the contraindication interval; the intermediate value is the center position of the taboo interval; calculate the synergy measure , used in Based on the above, the capsule release peak and the taboo interval position relationship are coordinated; the synergy measurement value is defined as: Through the synergistic measurement value, it is possible to quickly determine within the adjustable range which new time point can maintain a higher judgment quantification value while taking into account the favorable position relationship between the release peak moment and the intermediate value; finally, the candidate time point that makes the synergistic measurement value reach the maximum value is selected as the new recommended time point for taking.

[0051] An intelligent management system for dynamically adjusting the time of taking food-grade capsules, comprising: a data analysis module, a physiological preprocessing module, a feature screening module, a model evaluation module, a conditional marking module and a time optimization module;

[0052] Data parsing module: parses and structures the capsule characteristic data provided by the manufacturer, obtains and stores the key parameter set of capsule taking; outputs the key parameter set of capsule taking as one of the inputs of the feature screening module;

[0053] Physiological preprocessing module: uses the multi-dimensional physiological signals collected by the smart watch to generate a set of physiological characteristic candidate indicators that can be used to infer the timing of capsule taking after data cleaning, time series alignment and feature dimension reduction; outputs the physiological characteristic candidate indicators as another input of the feature screening module;

[0054] Feature screening module: compare the candidate physiological feature indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and screen out the physiological feature control indicators that meet the capsule taking requirements from the candidate physiological feature indicators through correlation analysis and mutual information measurement; output the physiological feature control indicators and input them together with the capsule taking key parameter set into the model evaluation module;

[0055] Model evaluation module: input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the judgment quantitative value, and quantitatively evaluate the suitability of the current taking time point; output the judgment quantitative value as the input of the conditional marking module;

[0056] Conditional marking module: Based on the judgment result of the quantitative value, if the current time period does not meet the ideal conditions, the corresponding time period will be marked and enter the dynamic tuning program to seek a better time recommendation; the time period to be optimized is marked according to the result of the quantitative value, which serves as the input of the time optimization module;

[0057] Time optimization module: for the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set of capsule taking are used to re-optimize the time, and the recommended taking time point with higher bioavailability potential is selected from the adjustable interval.

[0058] Technical effects and advantages of the intelligent management system and method for dynamically adjusting the time of taking food-grade capsules of the present invention:

[0059] The present invention establishes a systematic technical process from capsule characteristic data analysis to physiological characteristic extraction and matching, and then to dynamically optimize the timing of taking, so that the taking time of personalized food-grade capsules can be accurately controlled, thereby significantly improving the bioavailability of capsules and the effectiveness of user health intervention. Compared with the existing traditional methods that rely on experience or static rules, the present invention makes full use of the multi-dimensional physiological data collected by smart wearable devices, obtains physiological characteristic control indicators with practical relevance through rigorous data cleaning, time series alignment and feature dimension reduction technology, and accurately matches them with the key parameter set of capsule taking (including ingredient release characteristics, optimal absorption window and contraindications). On this basis, mutual information measurement, regression quantitative analysis and result-oriented dynamic tuning strategies are adopted to achieve real-time decision updates for individual physiological states, so that the taking time point can flexibly adapt to the micro-dynamic changes of the user's physiological rhythm, thereby ensuring that the intake and absorption of ingredients reach a better state, so as to achieve fast, flexible and highly reliable personalized capsule taking management. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A schematic diagram of a process of an intelligent management method for dynamically adjusting the taking time of food-grade capsules according to the present invention;

[0061] Figure 2 The present invention is a structural schematic diagram of an intelligent management system for dynamically adjusting the time of taking food-grade capsules. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Embodiment 1: Figure 1 The present invention provides an intelligent management method for dynamically adjusting the taking time of food-grade capsules, comprising:

[0064] S1: parse and structure the capsule characteristic data provided by the manufacturer, and obtain and store the capsule taking key parameter set for capsule taking;

[0065] S2: Using the multi-dimensional physiological signals collected by the smartwatch, after data cleaning, time series alignment and feature dimension reduction, a set of physiological feature candidate indicators that can potentially be used to infer the timing of capsule taking is generated;

[0066] S3: Compare the candidate physiological characteristic indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and select the physiological characteristic control indicators that match the capsule taking requirements from the candidate physiological characteristic indicators through correlation analysis and mutual information measurement;

[0067] S4: Input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the judgment quantitative value, and quantitatively evaluate the suitability of the current taking time point;

[0068] S5: Based on the judgment result of the determination quantization value, if the current time period does not meet the ideal conditions, the corresponding time period is marked and a dynamic tuning program is entered to seek a better taking time recommendation;

[0069] S6: For the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set for capsule administration are used to re-optimize the time, and the recommended administration time point with higher bioavailability potential is selected from the adjustable interval.

[0070] In order to realize personalized dynamic management of food-grade capsule taking time, it is necessary to first parse and structure the capsule characteristic data provided by the manufacturer to obtain the key parameter set of capsule taking. The In subsequent steps, the candidate indicators of physiological characteristics will be associated, matched and screened to provide a solid data foundation for the optimization of the administration time.

[0071] Step S1 includes the following contents:

[0072] Firstly, the source format of the capsule characteristic data provided by the manufacturer is rigorously parsed, and a set of parameter subsets for subsequent quantitative calculations and logical judgments is constructed based on the information in text, table or metadata format such as capsule composition, release characteristics, optimal absorption range and contraindications.

[0073] In the data analysis phase, the initial data is accurately differentiated according to the following parameter dimensions:

[0074] Absorption period parameters: including the starting time and ending time of the capsule, described in hours as the basic unit, recorded as the starting point of the absorption period End of absorption period .

[0075] Taboo interval parameters: Contains certain physiological state intervals that must be avoided during use (such as critical values ​​under special metabolic states), expressed as specific upper and lower limits, recorded as the taboo lower limit Taboo Cap .

[0076] Ingredient release parameters: describe the release rate and absorption peak time of a specific ingredient in the capsule under ideal conditions, and release the peak time as a time scalar Perform calibration.

[0077] Absorption gain parameter: used to indicate the increase in the absorption efficiency of the capsule under specific physiological conditions (such as heart rate variability reaching a certain range), recorded as the absorption gain factor .

[0078] After collecting and separating the above parameters, As the target data structure, all the parameters obtained by parsing are stored in order to form the following vector-based capsule key parameter set: : ;

[0079] In this process, in order to ensure the applicability and rigor of subsequent links, the capsule taking key parameter set is checked for consistency and format converted:

[0080] Consistency check: Check the logical relationship between the absorption period start and the absorption period end to ensure that the absorption period end is after the absorption period start; verify the validity of the taboo lower limit and the taboo upper limit to ensure that the taboo upper limit is greater than the taboo lower limit; check whether the release peak moment falls within the appropriate time interval Confirm that the absorption gain factor is greater than zero. If any abnormal values ​​that do not conform to the logical relationship are found between the parameters, the elimination or correction operation is performed.

[0081] Format conversion: convert all time parameters Convert to the same time base (such as the zero point of the user's daily physiological rhythm as the base); Absorption gain factor Standardize to dimensionless value to avoid dimension confusion in subsequent analysis and calculation.

[0082] After completing the above operations, the capsule taking key parameter set is integrated into a quantitative, structured and interactive parameter set, and stored in a dedicated data storage unit. The existence of the capsule taking key parameter set provides a clear input basis for the subsequent steps. The subsequent steps will directly retrieve the parameters in the capsule taking key parameter set to achieve the screening and matching of physiological characteristics, and finally derive the suitability judgment on the timing of capsule taking through logistic regression analysis.

[0083] Through this step, the capsule characteristic data originally provided by the manufacturer in an unstructured or semi-structured form is successfully converted into a structured capsule taking key parameter set with multi-dimensional parameters such as time period, contraindications, release characteristics and gain factors. The capsule taking key parameter set plays a decisive role in the subsequent physiological characteristic matching, determination of quantitative value calculation and time dynamic optimization process, ensuring that the present invention can quickly and accurately call relevant parameters in subsequent steps, thereby realizing dynamic optimization management of individualized nutritional supplement intake time.

[0084] In step S1, the present invention has parsed and structured the capsule characteristic data provided by the manufacturer to obtain the key parameter set for capsule taking. Next, this step will use the multi-dimensional physiological signal data (such as heart rate variability, skin conductance value and movement rhythm parameters) obtained by the smart watch for processing to produce a set of candidate physiological characteristic indicators that can be used as inferences for subsequent taking time. These candidate physiological characteristic indicators will be associated with the capsule taking key parameter set in subsequent steps for analysis and screening to achieve accurate determination and optimization of individualized capsule taking time.

[0085] Step S2 includes the following contents:

[0086] First, data cleaning and limiting are performed on each type of data in the multidimensional physiological signals of the smart watch, and records that are significantly out of the physiological reasonable range are eliminated to ensure that there is no risk of error propagation in subsequent calculations. Signals with inconsistent time series are unified to the same time base through linear interpolation methods to obtain multiple physiological data sequences that can be compared on the same time axis. Assume that the physiological sequences obtained after alignment are:

[0087] : At the time The heart rate parameter sequence below, is the discrete time index;

[0088] : At the time The skin conductance parameter sequence under

[0089] : At the time The following movement rhythm parameter sequence.

[0090] Next, a variety of feature functions that can be used to evaluate the fluctuation characteristics of individual physiological states are extracted from the aligned sequences for subsequent screening and association. Multidimensional discrete jump degree measurement and interval consistency evaluation are used in the feature extraction process. For example, define the jump degree function It is used to measure the cumulative average of absolute differences between adjacent samples to evaluate the magnitude of signal changes over time: ; here for The total number of valid sampling points, The higher the value, the more significant the change in the corresponding physiological parameter. and definition and , in order to obtain the changes in skin conduction and movement rhythm.

[0091] After obtaining these change indexes, in order to further screen out the features with usability, constraints are set for each feature sequence: the change degree must fall within a specific range to ensure that the corresponding features are neither too stable (lack of information value) nor too drastic (difficult to be effectively used in subsequent modeling). For example, setting a lower limit With upper limit As The allowed range is: ; Only features whose degree of variation meets the requirements of the corresponding interval are included in the set of candidate physiological feature indicators.

[0092] Among the features that meet the above-mentioned variation conditions, in order to improve the effectiveness of the features, each signal is also evaluated for interval consistency. The local monotonic fragment count within the preset window length is used to screen out feature sequences with no obvious physiological regularity or with too high a disorder. The specific steps are as follows:

[0093] In the right When three sequences are screened for physiological laws, a fixed length is first determined is the analysis window, and Each sequence is divided into multiple continuous sub-segments for the step size; for each sub-segment, its data point sequence is checked in turn to see whether it shows a strictly monotonically increasing or strictly monotonically decreasing trend. If there is no rated number of repeated changes in the direction of the data value in the entire sub-segment, the corresponding sub-segment is counted as a local monotonic segment; then the proportion of the local monotonic segment in the current sequence relative to all sub-segments is counted ( is the dimensionless ratio of the number of monotonic segments divided by the total number of subsegments); if the ratio If the value is lower than the preset lower limit, it indicates that the corresponding sequence lacks stable physiological regulation characteristics within the preset window and has multiple fluctuations and repetitions, which is regarded as a characteristic sequence with too high a disorder level. The corresponding sequence is thus eliminated and not included in the set of candidate physiological characteristic indicators for subsequent analysis.

[0094] Finally, through this multiple conditional constraints, a set of information parameters that meet the dynamic range restrictions, have an effective change pattern and do not violate physiological rationality are extracted from the original multidimensional signal and recorded as a set of candidate physiological feature indicators.

[0095] After extracting the candidate physiological characteristic indicators, the set of candidate physiological characteristic indicators is stored in a structured manner for subsequent steps. The next step is to retrieve the physiological characteristic control indicators that match the absorption conditions and contraindication thresholds in the key parameter set of capsule taking from the set of candidate physiological characteristic indicators. Therefore, the data consistency, time alignment accuracy, and feature selection strictness defined and maintained in this step provide a reliable premise for subsequent matching and quantitative analysis.

[0096] Through this step, the multi-source physiological signals of the smart watch have undergone strict data cleaning, time alignment, and feature extraction based on the degree of change and interval consistency, successfully producing a set of candidate physiological characteristic indicators. The features in the candidate physiological characteristic indicator set have a clear time reference and information value, which can support the subsequent steps to associate them with the key parameter set for capsule taking and screen them, thereby achieving accurate dynamic management of individualized capsule taking time.

[0097] In step S2, the capsule characteristic data provided by the manufacturer has been rigorously analyzed and structured to successfully obtain the key parameter set for capsule taking. In the earlier step, the user's physiological data has been collected and processed using a smart watch to generate a set of physiological characteristic candidate indicators that can potentially be used to infer the timing of capsule taking. This step will quantitatively match and screen the physiological characteristic candidate indicators based on the absorption conditions and contraindication thresholds defined by the capsule taking key parameter set to form physiological characteristic control indicators, providing high-precision feature input for subsequent quantitative judgment and dynamic time optimization.

[0098] Step S3 includes the following contents:

[0099] First, the key constraint factors for association analysis are selected from the key parameter set of capsule administration, including the appropriate time interval defined by the absorption conditions. , Taboo Zone And release peak moments and absorption gain parameters On this basis, the time period on the time axis that meets the absorption conditions and does not trigger the taboo interval is defined as the ideal condition subset The ideal condition subset represents an ideal data set within the appropriate time interval and the physiological state falls outside the taboo interval.

[0100] In order to achieve accurate correlation analysis between candidate physiological characteristics and key parameters of capsule taking, each characteristic sequence in the candidate physiological characteristics is Do the following:

[0101] a1. Mutual information measurement analysis: Each feature sequence is regarded as a continuous numerical signal, and the feature sequence is discretized into value intervals to obtain the probability distribution ; At the same time, construct a binary indicator sequence corresponding to the ideal condition subset :when The conditions that meet the definition of the ideal condition subset are ,otherwise , thus obtaining Under this condition, calculate The mutual information value of for: ;in for The joint probability of The probability of a single for The probability of a single Indicates Candidate indicators of physiological characteristics, Is a binary indicator sequence used to mark the time moment Whether the ideal absorption conditions defined by the key parameter set for capsule administration are met. When the mutual information value is higher than the set threshold, it means that the corresponding physiological characteristics have a significant information correlation with the ideal condition subset when it appears, and can be preliminarily selected as a candidate for physiological characteristic control indicator.

[0102] a2. Correlation test of release peak time: In the subset of candidate physiological characteristics that have passed the mutual information test, in order to further screen out the physiological characteristics that are related to the release peak time, (reflecting the critical release peak moment of the capsule) has a closely related feature. The local extreme point information of is compared with the release peak moment position. When a high proportion of local maximum values ​​appear within the range, the corresponding proportion is defined as ;in, Indicates Physiological characteristic candidate indicators In the moment The specific value of A time window range representing the critical release peak moment of the capsule, Indicates the deviation tolerance of the peak moment. It is a positive number used to define a time interval to allow time fluctuations within a certain range.

[0103] Indicates Physiological characteristic candidate indicators In the time window The number of local maxima in the . It is used to measure the activity and volatility of the candidate physiological feature indicators near the peak moment of capsule release.

[0104] Indicates Physiological characteristic candidate indicators The total number of local maxima in the entire time series. It reflects the overall fluctuation characteristics of the candidate indicators of physiological characteristics and shows the fluctuation frequency of the characteristic during the entire monitoring time.

[0105] If the corresponding ratio is higher than the peak threshold, it means that the corresponding feature presents a stable and significant physiological signal peak behavior in the period close to the peak release moment, further confirming its consistency with the optimal absorption rhythm of the capsule.

[0106] Through the mathematical measurement and screening strategies in the above two steps, the features that meet the standards for mutual information and local peak positioning are recorded as member elements of the physiological characteristic control index. The set of physiological characteristic control indexes will be used as input features in subsequent steps to achieve quantitative evaluation of the suitability of the current taking time.

[0107] This step achieves a double screening of physiological feature vectors by measuring the mutual information of candidate physiological feature indicators and the key parameter set for capsule administration and testing the local peak correlation of the time window. After processing in this step, the initially generated set of physiological feature control indicators strictly selects feature indicators that are highly consistent with ideal absorption conditions and key release windows, laying a high-precision and strongly correlated feature foundation for the subsequent Logistic regression-based calculation of quantitative values ​​and dynamic optimization of administration time points. This not only improves the model's matching degree between individual physiological rhythms and capsule characteristics, but also provides solid technical support for the realization of personalized and precise nutritional supplementation.

[0108] In step S3, physiological characteristic control indicators that are highly consistent with the capsule taking key parameter set have been screened out from the physiological characteristic candidate indicators through correlation analysis and mutual information measurement. This step aims to use the physiological characteristic control indicators and the quantitative time parameters and contraindication interval elements in the capsule taking key parameter set to construct and input a pre-trained Logistic regression model to calculate the judgment quantitative value, thereby quantitatively evaluating the suitability of the current candidate taking period. The processing results of this step will provide a clear numerical basis for the dynamic tuning of the next step.

[0109] Step S4 includes the following contents:

[0110] First, obtain the characteristic sequence set based on the physiological characteristic control index , where each In the moment All of the above have been proved to be of practical significance in the release and absorption conditions of capsule ingredients through mutual information and local peak tests. At the same time, parameters that can quantitatively describe the characteristics of the taking window are extracted from the key parameters of capsule taking, including the starting point of the absorption period, the end point of the absorption period, the upper limit of the taboo, the lower limit of the taboo, the peak release time and the absorption gain factor. To facilitate the subsequent model input, all time parameters are normalized based on a unified time reference (such as the user's daily physiological rhythm zero point as a reference) to ensure that each feature quantity is comparable at the same scale.

[0111] Defined in time The linear combination expression of ;

[0112] in:

[0113] is the model intercept term, determined by historical training data, and is used to provide a baseline estimate in the absence of other inputs.

[0114] For the corresponding The model coefficient comes from the pre-training process and determines the direction and degree of influence of the feature on the final judgment.

[0115] Through The normalized deviation from the release peak moment within a suitable time interval is measured to quantify the relative relationship between the current time and the release peak moment.

[0116] It is the influence coefficient of the absorption gain factor on the judgment, reflecting the adjustment of the judgment result by the absorption enhancement potential under a specific physiological state.

[0117] The taboo interval is then measured symmetrically to measure the risk ratio of the current state inside and outside the taboo interval, thereby adjusting the sensitivity of the judgment result to potential taboo conditions.

[0118] In determining After that, it is mapped to the (0,1) interval through the Logistic function to obtain the judgment quantization value ;

[0119] When the judgment quantization value is close to 1, it means that the physiological and capsule release conditions of the current period are highly matched; when the judgment quantization value is close to 0, it means that the current period does not meet the ideal absorption conditions or there is a potential risk.

[0120] Through the above strict and clear processing logic, this step inputs the physiological characteristic control index and the key parameters of capsule taking into the trained Logistic regression model, and successfully obtains the quantitative judgment value of the judgment quantification value. In the next step, if the judgment quantification value shows that the current time period is not ideal, the time period is marked according to the output of this step and the dynamic tuning process is started to find a time to take the capsule that is more conducive to the absorption and utilization of the capsule ingredients from other adjustable intervals. This step ensures a seamless connection from feature screening to quantitative judgment, and provides an accurate and solid data decision-making basis for realizing personalized and dynamic optimization of capsule taking time.

[0121] In step S4, the present invention has quantified the time period parameters and contraindication intervals defined by the physiological characteristic control index and the capsule taking key parameter set, and calculated the judgment quantization value using the pre-trained Logistic regression model. The judgment quantization value describes the suitability of the current taking period in the interval (0,1). If the judgment quantization value is close to 1, it means that the current period meets the capsule taking requirements; if the judgment quantization value is close to 0, it indicates that there is a high risk of deviation from the ideal conditions in the current period.

[0122] In this step, based on step S4, a decision interpretation is performed on the determined quantization value. If it is determined that the ideal condition is not met, the time period is marked as a target object that needs to be optimized again to support the dynamic tuning process of the next step.

[0123] Step S5 includes the following contents:

[0124] First, set two threshold parameters:

[0125] Used to identify high suitability periods. , then the current time is considered The expected requirements are met and no subsequent marking actions are performed;

[0126] Used to identify undesirable conditions. , then for this time Make special marks.

[0127] In actual operation, when receiving the output judgment quantization value After the sequence (the set of decision quantization values ​​calculated one by one at the relevant time moments), for each time moment Perform the following decision test:

[0128] like , it indicates that the physiological characteristic control index and the capsule taking key parameter set have a high matching degree in the corresponding period, and no adjustment is required. The corresponding period is recorded as the passing interval;

[0129] like , it means that although this period has not completely reached the ideal standard, it still maintains a certain degree of suitability. It can be temporarily not marked, and only the judgment quantization value of this period is recorded for reference in the later fine-tuning;

[0130] like , it means that there is a risk of significant deviation from the ideal conditions during this period, and it is necessary to enter the subsequent optimization process. The corresponding time period is marked as the re-optimization period.

[0131] All time moments that are judged to need further optimization The corresponding physiological characteristics and control index characteristic values ​​are classified and stored, while retaining the mapping relationship with the key information of the capsule taking key parameter set. This marking action forms a re-optimized time list so that the recommended taking time point can be reselected from the adjustable interval in the next step to improve the bioavailability of the capsule.

[0132] Through the decision-making inspection and marking process of this step, the present invention accurately extracts the time period when the judgment quantization value is not ideal, providing a clear target object for the subsequent dynamic tuning program. This ensures that the processing chain from the judgment quantization value calculation to the time fine-tuning is closely connected, so that the system can perform targeted optimization and adjustment for the time period that does not meet the requirements, and finally realizes high-precision, dynamic and intelligent management of individualized capsule taking time.

[0133] In step S4, the present invention has calculated the judgment quantification value through the Logistic regression model, and marked the time period that does not meet the ideal conditions according to the judgment quantification value. The marked time period represents the weak link under the current administration strategy, and a recommended administration time point with higher bioavailability potential must be reselected within the adjustable interval. This step will use the previously obtained physiological characteristic control index and the absorption window specified in the capsule administration key parameter set to select a better administration time point from the adjustable interval, so that the final time recommendation can not only fit the capsule release peak characteristics, but also avoid the contraindication interval, thereby improving the bioavailability of the capsule.

[0134] Step S6 includes the following contents:

[0135] First, based on the re-optimization period marked in the previous step, the adjustable interval is extracted from the key parameter set of capsule administration, that is, the candidate time set within the appropriate time interval and does not trigger the taboo interval For each candidate time point in the candidate time set Re-call the physiological characteristic control index to recalculate the candidate time point The quantitative value of The calculation is still based on the previously trained Logistic regression model, but the input time is changed to In this way, all candidate moments within the adjustable interval are obtained. distributed.

[0136] In order to select the time point with the greatest potential bioavailability from these candidate moments, it is necessary to comprehensively consider the capsule release peak moment and the position of the contraindication interval. is the center position of the taboo interval. To improve the accuracy of the selection, calculate the synergy measure , used in On this basis, the relationship between the capsule release peak and the taboo interval position is coordinated.

[0137] Define the synergy measure as: ;

[0138] In this formula:

[0139] For the moment in time The closer the calculated judgment quantization value is to 1, the higher the suitability is;

[0140] The absolute deviation from the capsule release peak moment is smaller, which means The closer to the ideal release peak, the more efficient the absorption of ingredients;

[0141] The absolute distance relative to the center point of the taboo interval is larger when this value is larger. The further away from potential risk areas, the better it is for avoiding adverse conditions;

[0142] The purpose of adding the constant 1 is to avoid meaningless ratios in the denominator when the distance is zero and to ensure the stability of the calculation process.

[0143] Through the collaborative measurement value, it is possible to quickly determine which new time point can maintain a higher judgment quantization value and take into account the favorable position relationship between the release peak moment and the intermediate value within the adjustable range. Finally, the candidate time point that makes the collaborative measurement value reach the maximum value is selected as the new recommended time point for taking the drug, and the analysis results of the new recommended time point and its collaborative measurement process are recorded and archived to provide users with a reasonable and well-founded dynamic optimization strategy.

[0144] Through this step, the time period marked as undesirable conditions can be recalibrated within the adjustable range. The optimization process not only refers to the original judgment quantitative value indicators, but also incorporates the precise consideration of the relative position of the release peak moment and the contraindication interval, thereby achieving a higher-dimensional optimization of the taking time point. This ensures that the final output time recommendation point can achieve a dynamic balance between bioavailability, risk avoidance and individual physiological state matching, laying a comprehensive and accurate technical foundation for personalized food-grade capsule taking time management.

[0145] Embodiment 2: Figure 2 The present invention provides an intelligent management system for dynamically adjusting the time of taking food-grade capsules, comprising: a data analysis module, a physiological preprocessing module, a feature screening module, a model evaluation module, a condition marking module and a time optimization module;

[0146] Data parsing module: parses and structures the capsule characteristic data provided by the manufacturer, obtains and stores the key parameter set of capsule taking; outputs the key parameter set of capsule taking as one of the inputs of the feature screening module;

[0147] Physiological preprocessing module: uses the multi-dimensional physiological signals collected by the smart watch to generate a set of physiological characteristic candidate indicators that can be used to infer the timing of capsule taking after data cleaning, time series alignment and feature dimension reduction; outputs the physiological characteristic candidate indicators as another input of the feature screening module;

[0148] Feature screening module: compare the candidate physiological feature indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and screen out the physiological feature control indicators that meet the capsule taking requirements from the candidate physiological feature indicators through correlation analysis and mutual information measurement; output the physiological feature control indicators and input them together with the capsule taking key parameter set into the model evaluation module;

[0149] Model evaluation module: input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the judgment quantitative value, and quantitatively evaluate the suitability of the current taking time point; output the judgment quantitative value as the input of the conditional marking module;

[0150] Conditional marking module: Based on the judgment result of the quantitative value, if the current time period does not meet the ideal conditions, the corresponding time period will be marked and enter the dynamic tuning program to seek a better time recommendation; the time period to be optimized is marked according to the result of the quantitative value, which serves as the input of the time optimization module;

[0151] Time optimization module: for the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set of capsule taking are used to re-optimize the time, and the recommended taking time point with higher bioavailability potential is selected from the adjustable interval.

[0152] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0153] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0154] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0155] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for dynamically adjusting the time of taking food-grade capsules and intelligent management, characterized in that: Includes steps: S1: parse and structure the capsule characteristic data provided by the manufacturer, and obtain and store the capsule taking key parameter set for capsule taking; S2: Using the multi-dimensional physiological signals collected by the smartwatch, after data cleaning, time series alignment and feature dimension reduction, a set of physiological feature candidate indicators that can potentially be used to infer the timing of capsule taking is generated; S3: Compare the candidate physiological characteristic indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and select the physiological characteristic control indicators that match the capsule taking requirements from the candidate physiological characteristic indicators through correlation analysis and mutual information measurement; S4: Input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the judgment quantitative value, and quantitatively evaluate the suitability of the current taking time point; S5: Based on the judgment result of the determination quantization value, if the current time period does not meet the ideal conditions, the corresponding time period is marked and a dynamic tuning program is entered to seek a better taking time recommendation; S6: For the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set for capsule administration are used to re-optimize the time, and the recommended administration time point with higher bioavailability potential is selected from the adjustable interval.

2. According to claim 1, a method for dynamically adjusting the intelligent management of food-grade capsule taking time is characterized in that: Step S1 includes the following contents: First, the source format of the capsule characteristic data provided by the manufacturer is parsed to construct a parameter subset; In the data analysis phase, the initial data is accurately differentiated according to the following parameter dimensions: Absorption period parameters: including the starting time and ending time of the capsule, described in hours as the basic unit, recorded as the starting point of the absorption period End of absorption period ; Taboo interval parameter: contains certain physiological state intervals that must be avoided during the use of the drug, expressed as a specific upper and lower limit range, recorded as the taboo lower limit Taboo Cap ; Ingredient release parameters: describe the release rate and absorption peak time of a specific ingredient in the capsule under ideal conditions, and release the peak time as a time scalar Carry out calibration; Absorption gain parameter: used to indicate the increase in the absorption efficiency of the capsule under specific physiological conditions, recorded as the absorption gain factor ; All the parameters obtained by analysis are stored in order to form the following vector-based key parameter set for capsule administration: : ; Then, the capsule administration key parameter set is checked for consistency and format converted.

3. The method for dynamically adjusting the time of taking food-grade capsules according to claim 2 is characterized in that: Step S2 includes the following contents: Firstly, data cleaning and limiting processing are performed on each type of data in the multidimensional physiological signals of the smart watch, and the records that are out of the physiological reasonable interval are eliminated; the signals with inconsistent time series are unified to the same time base through linear interpolation method to obtain multiple physiological data sequences that can be compared under the same time axis; the physiological sequences obtained after alignment are assumed to be: : At the time The heart rate parameter sequence is as follows: is the discrete time index; : At the time The skin conductance parameter sequence under : At the time The sequence of movement rhythm parameters under Next, a variety of feature functions that can be used to evaluate the fluctuation characteristics of individual physiological states are extracted from the aligned sequences; In the feature extraction process, multi-dimensional discrete jump degree measurement and interval consistency evaluation are used; based on , define the jump degree function It is used to measure the cumulative average of absolute differences between adjacent samples, thereby evaluating the magnitude of signal changes over time. and definition and , so as to obtain the changes in skin conductance and movement rhythm; After obtaining these variation indexes, constraints are set for each feature sequence: the variation must fall within a specific interval to ensure that the corresponding feature is neither too stable nor extremely drastic. Only features whose variation meets the requirements of the corresponding interval are included in the set of candidate physiological feature indicators; Among the features that meet the degree of variation condition, interval consistency evaluation is also performed for each signal; Finally, through this multiple conditional constraints, a set of information parameters that meet the dynamic range restrictions, have an effective change pattern and do not violate physiological rationality are extracted from the original multidimensional signal and recorded as a set of candidate physiological feature indicators.

4. The method for dynamically adjusting the time of taking food-grade capsules according to claim 3 is characterized in that: Among the features that meet the variation condition, each signal is also evaluated for interval consistency. The specific steps are as follows: In the right , , When three sequences are screened for physiological laws, a fixed length is first determined is the analysis window, and Each sequence is divided into multiple continuous sub-segments for the step size; for each sub-segment, its data point sequence is checked in turn to see whether it shows a monotonically increasing or monotonically decreasing trend. If there is no rated number of repeated changes in the direction of the data value in the entire sub-segment, the corresponding sub-segment is counted as a local monotonic segment; then the proportion of the local monotonic segment in the current sequence relative to all sub-segments is counted ; If the ratio If the value is lower than the preset lower limit, it means that the corresponding sequence lacks stable physiological regulation characteristics within the preset window and has multiple fluctuations and repetitions, which is regarded as a characteristic sequence with too high disorder. The corresponding sequence is thus eliminated and not included in the set of candidate physiological characteristic indicators.

5. The method for dynamically adjusting the time of taking food-grade capsules according to claim 3 is characterized in that: Step S3 includes the following contents: The time period on the time axis that satisfies the absorption conditions and does not trigger the taboo interval is defined as the ideal condition subset ; The ideal condition subset represents an ideal data set within the appropriate time interval and the physiological state falls outside the taboo interval; for each feature sequence in the candidate physiological feature index Do the following: a1. Mutual information measurement analysis: Each feature sequence is regarded as a continuous numerical signal, and the feature sequence is discretized into value intervals to obtain the probability distribution ; At the same time, construct a binary indicator sequence corresponding to the ideal condition subset :when The conditions that meet the definition of the ideal condition subset are ,otherwise , thus obtaining and and The joint distribution of ; under this condition, calculate and The mutual information value of for: ;in for and The joint probability of for The probability of a single for The probability of a single Indicates Candidate indicators of physiological characteristics, Is a binary indicator sequence used to mark the time moment Whether the ideal absorption conditions defined by the key parameter set for capsule administration are met; when the mutual information value is higher than the set threshold, it means that the corresponding physiological characteristics have a significant information correlation with the ideal condition subset when it appears, and are preliminarily selected as candidates for physiological characteristic control indicators; a2. Correlation test of release peak time: In the subset of candidate physiological characteristics that have passed the mutual information test, in order to screen out the characteristics that are closely related to the release peak time, The local extreme point information of is compared with the position of the release peak moment; when exist When a high proportion of local maximum values ​​appear within the range, the corresponding proportion is defined as : ;in, Indicates Physiological characteristic candidate indicators In the moment The specific value of A time window range representing the critical release peak moment of the capsule, Indicates the deviation tolerance of the peak moment, which is a positive number used to define a time interval to allow time fluctuations within a certain range; Indicates Physiological characteristic candidate indicators In the time window The number of local maxima in ; Indicates Physiological characteristic candidate indicators The total number of local maxima in the entire time series; if the corresponding ratio is higher than the peak threshold, it means that the corresponding feature presents a stable and significant peak behavior of the physiological signal in the period close to the peak release moment, confirming its consistency with the optimal absorption rhythm of the capsule; Through the mathematical measurement and screening strategies of step a1 and step a2, the features that meet the standards of mutual information and local peak location are recorded as member elements of the physiological feature control index.

6. The method for dynamically adjusting the time of taking food-grade capsules according to claim 5, characterized in that: Step S4 includes the following contents: First, obtain the characteristic sequence set based on the physiological characteristic control index At the same time, parameters that can quantitatively describe the characteristics of the taking window are extracted from the key parameters of capsule taking, including the starting point of the absorption period, the end point of the absorption period, the upper limit of the taboo, the lower limit of the taboo, the peak release time and the absorption gain factor; all time parameters are normalized based on a unified time reference; Defined in time The linear combination expression of for: ;in, is the model intercept term; For the corresponding The model coefficients of Through It is measured by the normalized deviation from the peak release moment within a suitable time interval; is the influence coefficient of absorption gain factor on the judgment; The taboo interval is then measured symmetrically to measure the risk ratio of the current state inside and outside the taboo interval, thereby adjusting the sensitivity of the judgment result to potential taboo conditions; After that, it is mapped to the (0,1) interval through the Logistic function to obtain the judgment quantization value .

7. The method for dynamically adjusting the time of taking food-grade capsules according to claim 6, characterized in that: Step S5 includes the following contents: First, set two threshold parameters: Used to identify periods of high suitability; Used to identify undesirable conditions; in actual operation, when receiving the output judgment quantization value After the sequence, for each time moment Perform the following decision test: like , it indicates that the physiological characteristic control index and the capsule taking key parameter set have a high matching degree in the corresponding period, and no adjustment is required. The corresponding period is recorded as the passing interval; like , it means that although this period has not completely reached the ideal standard, it still maintains a certain degree of suitability. It will not be marked for the time being, and only the judgment quantification value of the corresponding period will be recorded for reference in the later fine-tuning; like , it means that there is a risk of significant deviation from the ideal conditions during this period, and it is necessary to enter the subsequent optimization process. The corresponding time period is marked as the re-optimization period; All time moments that are judged to need further optimization The corresponding physiological characteristics and control index characteristic values ​​are classified and stored, while retaining the mapping relationship with the key information of the capsule taking key parameter set.

8. The method for dynamically adjusting the time of taking food-grade capsules according to claim 7 is characterized in that: Step S6 includes the following contents: First, based on the marked re-optimization period, the adjustable interval is extracted from the key parameter set of capsule administration, that is, the candidate time set within the appropriate time interval and does not trigger the taboo interval ; For each candidate time point in the candidate time set Call the physiological characteristic control index again to recalculate the candidate time point The quantitative value of , The calculation is still based on the previously trained Logistic regression model, but the input time is changed to ; In this way, all candidate moments within the adjustable interval are obtained distributed.

9. The method for dynamically adjusting the time of taking food-grade capsules according to claim 8, characterized in that: Step S6 also includes the following contents: The time point with the greatest potential for bioavailability is selected from these candidate moments, taking into account the peak release moment of the capsule and the position of the contraindication interval; the intermediate value is the center position of the taboo interval; Calculating Synergy Measures , used in On this basis, the relationship between the capsule release peak and the taboo interval position is coordinated; Define the synergy measure as: Through the synergistic measurement value, it is possible to quickly determine within the adjustable range which new time point can maintain a higher judgment quantification value while taking into account the favorable position relationship between the release peak moment and the intermediate value; finally, the candidate time point that makes the synergistic measurement value reach the maximum value is selected as the new recommended time point for taking.

10. An intelligent management system for dynamically adjusting the time of taking food-grade capsules, used to implement the intelligent management method for dynamically adjusting the time of taking food-grade capsules according to any one of claims 1 to 9, characterized in that: include: Data parsing module, physiological preprocessing module, feature screening module, model evaluation module, conditional labeling module and time optimization module; Data parsing module: parses and structures the capsule characteristic data provided by the manufacturer, obtains and stores the key parameter set of capsule taking; outputs the key parameter set of capsule taking as one of the inputs of the feature screening module; Physiological preprocessing module: The multi-dimensional physiological signals collected by the smartwatch are cleaned, time series aligned, and feature dimension reduced to generate a set of physiological characteristic candidate indicators that can potentially be used to infer the timing of capsule taking. Output candidate physiological feature indicators as another input to the feature screening module; Feature screening module: compares the candidate physiological feature indicators with the absorption conditions and contraindication thresholds defined in the capsule taking key parameter set one by one, and selects the physiological feature control indicators that meet the capsule taking requirements from the candidate physiological feature indicators through correlation analysis and mutual information measurement; Output physiological characteristic control indexes and input them into the model evaluation module together with the key parameter set of capsule administration; Model evaluation module: input the physiological characteristic control index and the time period parameters and contraindication interval indicators of the key parameters of capsule taking into the pre-trained Logistic regression model, calculate the quantitative value of the judgment, and quantitatively evaluate the suitability of the current taking time point; Output the decision quantization value as the input of the conditional marking module; Conditional marking module: Based on the judgment result of the quantitative value, if the current time period does not meet the ideal conditions, the corresponding time period will be marked and enter the dynamic tuning program to seek a better taking time recommendation; Mark the time period to be optimized according to the result of the quantitative value determination as the input of the time optimization module; Time optimization module: for the marked time period, the physiological state characteristics reflected by the physiological characteristic control index and the absorption window specified by the key parameter set of capsule taking are used to re-optimize the time, and the recommended taking time point with higher bioavailability potential is selected from the adjustable interval.

Images