Wearable physical ability monitoring system based on graphene

Abstract

The invention provides a wearable physical ability monitoring system based on graphene, and relates to the technical field of physical ability monitoring. The system comprises a motion direction identification module, a motion amplitude monitoring module, a dynamic sensitive area adjustment module, a physical fitness state evaluation module and a physical fitness index monitoring module. According to the invention, through combined analysis of the skin shear strain and the subcutaneous vertical strain, the deformation difference of different layers of tissues in the movement process can be caught more finely, so that the physical ability monitoring is more accurate. By means of a strain gradient calculation mode, the skin deformation direction and trend can be recognized, and high-precision description of the motion state is achieved. And the change trend of the strain amplitude is dynamically monitored, so that data processing is more continuous, and the movement intensity fluctuation condition can be accurately judged. The weight distribution of the monitoring area is adjusted according to the change of the exercise load, so that the sensor resources are optimally configured, and the accuracy of physical fitness state evaluation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of physical fitness monitoring, and in particular to a wearable physical fitness monitoring system based on graphene. Background Art

[0002] The field of physical fitness monitoring technology includes systems and methods for measuring and analyzing human physiological parameters and motion status. The core content of this technical field involves biological signal acquisition, motion data analysis, and health status assessment. Biosignal acquisition includes the acquisition of information such as heart rate, electrophysiological signals, and skin conductance. Motion data analysis involves the processing of motion parameters such as gait, acceleration, and angular velocity. Health status assessment determines the individual's physical fitness based on the changing trends of physiological and motion data. Existing physical fitness monitoring technology mainly relies on wearable devices to complete data collection and analysis. Common systems include heart rate monitors based on optical sensors, pedometers based on acceleration sensors, and electrocardiogram monitoring devices based on electrophysiological signals. With the development of wearable devices, the application of flexible electronic materials, miniaturized sensors, and wireless communication technologies has improved the convenience and real-time performance of physical fitness monitoring.

[0003] The graphene-based wearable fitness monitoring system uses graphene materials to construct sensing units, combined with signal acquisition and data transmission technologies, to monitor the human body's physiological and motor status. The core of this system includes a graphene-based flexible sensor that detects bioelectrical signals, skin electrical impedance, or mechanical deformation through skin contact and converts the acquired signals into electrical output signals. Signal acquisition utilizes a highly sensitive conductive path structure to detect minute changes in skin potential or stress. Data transmission utilizes a wireless communication interface to transmit the acquired data to a terminal device. The system can be powered by an integrated micropower supply and combined with a low-power data processing circuit to perform signal preprocessing and preliminary feature extraction.

[0004] Current fitness monitoring primarily relies on single or limited biosignal sensing technologies, which are limited in their ability to analyze complex motion states. Most monitoring systems analyze only heart rate, step count, or single-dimensional motion parameters, making it difficult to effectively capture the true deformation of tissues at different levels during exercise, limiting monitoring accuracy. Motion state assessments are often based on a fixed monitoring area and fail to optimize for individual movement patterns. This makes monitoring data susceptible to individual differences, reducing assessment accuracy. Data fusion methods primarily rely on independent parameter analysis, failing to deeply explore the inherent connections between different monitoring indicators, resulting in a rather one-sided assessment of exercise load. In practical applications, uneven movement amplitude and load distribution make monitoring equipment unstable in complex exercise environments, making it difficult to adapt to dynamically changing exercise demands. Monitoring systems lack intelligent optimization methods for regional sensitivity adjustment, making it difficult to adapt to different exercise scenarios. This results in poorly balanced data collection and compromises the reliability of the final assessment results. Summary of the Invention

[0005] In order to solve the technical problems existing in the prior art, an embodiment of the present invention provides a wearable physical fitness monitoring system based on graphene. The technical solution is as follows:

[0006] A graphene-based wearable fitness monitoring system comprises: a motion direction recognition module that obtains skin surface shear strain and subcutaneous vertical strain data collected by a graphene strain sensor, calculates skin strain gradients based on the strain change rates of the skin surface and subcutaneous deep layers in adjacent time periods, and obtains a strain direction change state; a motion amplitude monitoring module that, based on the strain direction change state, retrieves the skin strain gradient change value for each time period, calculates the skin strain amplitude for adjacent time periods, and obtains a strain amplitude change trend; a dynamic sensitive area adjustment module that, referring to the distribution deviation between current strain data and the motion load monitoring area in the strain amplitude change trend, identifies monitoring areas exceeding a preset deviation threshold as sensitive areas, adjusts the strain weights of the sensitive areas, and obtains strain sensitive area optimization results; a physical fitness assessment module that determines the motion load change trend of each sensitive area in the strain sensitive area optimization results, classifies physical fitness strain response types for the same exercise intensity category, and obtains a physical fitness classification result; and a physical fitness index monitoring module that, based on the physical fitness classification results, analyzes the mutual influence of each type of physical fitness index, classifies physical fitness levels according to the influence, and obtains a physical fitness monitoring assessment result.

[0007] Optionally, the strain direction change state includes shear strain change, vertical strain change, and strain gradient distribution; the strain amplitude change trend includes the strain gradient change amplitude in adjacent time periods, the skin surface strain change rate, and the subcutaneous vertical strain change rate; the skin strain sensitive area optimization result includes the strain data deviation value, the sensitive area adjustment threshold, and the corrected monitoring area weight; the physical fitness status classification result includes the exercise load change trend, physical fitness strain response type, and exercise intensity classification; the physical fitness monitoring evaluation result includes monitoring indicator correlation analysis, physical fitness status evaluation, and exercise load impact analysis.

[0008] Optionally, the movement direction identification module includes: a strain data acquisition submodule that acquires data collected by a graphene strain sensor, calculates the change in shear strain of the skin surface and the change in vertical strain of the deep subcutaneous layer, matches the change in shear strain and vertical strain according to a time series, establishes a data correspondence, and obtains a matched strain data set; a strain gradient calculation submodule extracts the difference between the change in shear strain of the surface layer and the change in the deep vertical strain based on the matched strain data set, calculates the skin strain gradient value, filters out abnormal points that exceed a set strain gradient abnormality threshold based on a time series change range, calculates the average of the gradient changes of multiple consecutive time series before and after the abnormal point as a reference value, corrects the abnormal point, and obtains a corrected skin strain gradient value; a strain direction change identification submodule calls the corrected skin strain gradient value, calculates the direction angle change under multiple time series, judges the strain direction change trend, filters the time period where the change rate exceeds the angle change rate threshold, classifies the change trend, and obtains the strain direction change state.

[0009] Optionally, for calculating the direction angle change θ under multiple time series t , using the formula:

[0010]

[0011] Among them, Δε i,t represents the strain gradient change value of the i-th monitoring point at time t, Δε i,t-1 represents the strain gradient change value of the i-th monitoring point at the previous moment t-1, n represents the total number of monitoring points involved in the calculation in the current time series, d j,t represents the distance measurement data of the jth monitoring point at time t, d j,t-1 represents the ranging data of the jth monitoring point at time t-1, and m represents the number of monitoring points in the current time series whose ranging data participate in the calculation.

[0012] Optionally, the motion amplitude monitoring module includes: a strain amplitude extraction submodule calls the skin strain gradient change value of each time period in the strain direction change state, sorts it according to the time series, records the strain amplitude information of the complete time period, and obtains the time period strain amplitude data; an adjacent time period amplitude calculation submodule calculates the skin strain amplitude change value of the adjacent time period based on the time period strain amplitude data, extracts the change rate average of the time period whose change rate exceeds the amplitude fluctuation threshold, sets the mutation point according to the average and marks the time series, calculates the mean deviation of the two time periods before and after the mutation point, adjusts the mutation point data whose deviation exceeds the set deviation range, recalculates the adjusted amplitude change rate, and obtains the amplitude change rate of the adjacent time period; an amplitude change trend analysis submodule calls the amplitude change rate of the adjacent time period, calculates the amplitude change trend values of multiple time series, filters the time period where the change rate continuously increases or decreases, classifies the trend pattern, and obtains the strain amplitude change trend.

[0013] Optionally, for calculating the amplitude change trend value ΔV of different time series trend , using the formula:

[0014]

[0015] Among them, V t1 Represents the amplitude value at time point t1, V t1-1 Represents the amplitude value at the previous time point t1-1, Δt1 represents the time interval, usually the measurement period, V avg represents the average value of the amplitude value in all measurement periods, T represents the length of the total time series, ΔV trend Represents the calculated amplitude change trend value.

[0016] Optionally, the dynamic sensitive area adjustment module includes: a deviation calculation submodule calls the strain amplitude change trend, calculates the deviation value of the current strain data and the motion load monitoring area distribution, calculates the deviation mean of the area beyond the set deviation range, and obtains the regional deviation distribution result; a sensitive area adjustment threshold setting submodule calculates the deviation fluctuation rate of the area beyond the set deviation range based on the regional deviation distribution result, calculates the sensitive area adjustment threshold according to the deviation mean and the fluctuation rate, judges the change trend of the sensitive area adjustment threshold, adjusts the adjustment range of the abnormal area, and obtains the updated sensitive area adjustment threshold; the monitoring area weight optimization submodule calls the updated sensitive area adjustment threshold, adjusts the sensitive area weight, calculates the adjusted weight balance, filters the area where the weight change exceeds the adjustment range, and obtains the skin strain sensitive area optimization result.

[0017] Optionally, the physical fitness status assessment module includes: a motion load trend analysis submodule calling the motion load change data of each sensitive area in the skin strain sensitive area optimization result, calculating the load change rate under the time series, screening the areas with change rates exceeding the set change rate threshold, and obtaining the motion load change trend; a strain response classification submodule extracting sensitive areas with similar load change rates based on the motion load change trend, calculating the load change similarity between areas, screening areas with similarities exceeding the set similarity threshold, classifying similar load change areas, and obtaining physical fitness strain response types; a physical fitness status division submodule calling the physical fitness strain response type, classifying physical fitness strain areas of the same category, calculating the load balance between the classified areas, screening the categories with load balance deviation exceeding the load balance deviation threshold, dividing multiple physical fitness status areas, and obtaining physical fitness status classification results.

[0018] Optionally, for calculating the load change similarity S between regions i1,j1 , using the formula:

[0019]

[0020] Among them, R i1,t2 represents the load change rate of region i1 at time t2, R j1,t2 represents the load change rate of region j1 at time t2, T represents the total time series length, S i1,j1 Represents the load change similarity between region i1 and region j1.

[0021] Optionally, the physical fitness indicator monitoring module includes: a physical fitness monitoring data integration submodule extracts the corresponding monitoring data from the physical fitness state classification results, refers to the exercise load, physiological feedback and external environmental parameters, and matches the time series, eliminates the data with synchronization errors exceeding the error range, and obtains time-aligned monitoring data; a monitoring indicator impact analysis submodule analyzes the correlation of each monitoring indicator based on the time-aligned monitoring data, calculates the contribution of each indicator to the physical fitness state classification, screens indicators whose contributions exceed the set contribution standards, summarizes the influence relationship of the contribution indicators, and obtains the influence degree of the monitoring indicators; a physical fitness monitoring evaluation generation submodule calls the monitoring indicator influence degree, calculates the indicator change amplitude under each physical fitness state, analyzes the differences in indicators under each state category, screens indicators whose differences exceed the physical fitness state difference threshold as feature indicators, extracts the change trend of the feature indicators and divides the physical fitness levels, and obtains the physical fitness monitoring evaluation results.

[0022] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0023] By combining analysis of skin shear strain and subcutaneous vertical strain, the system can more precisely capture the deformation differences between tissue layers during exercise, making physical fitness monitoring more accurate. Strain gradient calculation can identify the direction and trend of skin deformation, achieving highly accurate characterization of exercise status. Dynamic monitoring of strain amplitude trends ensures more continuous data processing, helping to accurately assess fluctuations in exercise intensity. Adjusting the weight distribution of monitoring areas based on changes in exercise load optimizes sensor resource allocation and improves the accuracy of physical fitness assessment. Based on the changing trends of physical fitness load, dynamic classification of exercise status can be achieved. By integrating corresponding monitoring data and analyzing the interactions between different indicators, the overall assessment of physical fitness can be optimized, ensuring that the monitoring results are more consistent with actual exercise needs. By accurately identifying exercise patterns and physical fitness load, the distribution of physical fitness monitoring areas is optimized, effectively reducing data errors and improving the monitoring system's adaptability to complex exercise scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0025] Figure 1 is a system flow chart of the present invention;

[0026] Figure 2 It is a submodule flow chart of the present invention;

[0027] Figure 3 This is a flow chart of the motion direction identification module of the present invention;

[0028] Figure 4 This is a flow chart of the motion amplitude monitoring module of the present invention;

[0029] Figure 5 This is a flow chart of the dynamic sensitive area adjustment module of the present invention;

[0030] Figure 6 This is a flow chart of the physical state assessment module of the present invention;

[0031] Figure 7 This is a flow chart of the physical fitness indicator monitoring module of the present invention. DETAILED DESCRIPTION

[0032] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0033] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0034] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0035] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0036] like Figure 1 As shown, the present invention provides a graphene-based wearable fitness monitoring system, comprising: a motion direction recognition module that obtains skin surface shear strain and subcutaneous vertical strain data collected by a graphene strain sensor, calculates the skin strain gradient based on the strain change rate of the skin surface and subcutaneous deep layers in adjacent time periods, and obtains a strain direction change state; a motion amplitude monitoring module that, based on the strain direction change state, calls the skin strain gradient change value for each time period, calculates the skin strain amplitude for adjacent time periods, and obtains a strain amplitude change trend; a dynamic sensitive area adjustment module that, referring to the distribution deviation between the current strain data and the motion load monitoring area in the strain amplitude change trend, identifies the monitoring area exceeding a preset deviation threshold as a sensitive area, adjusts the strain weight of the sensitive area, and obtains a strain sensitive area optimization result; a physical state assessment module that determines the motion load change trend of each sensitive area in the strain sensitive area optimization result, classifies the physical strain response types of the same motion intensity category, and obtains a physical state classification result; and a physical fitness index monitoring module that, based on the physical state classification result, analyzes the mutual influence of each type of physical state index, classifies physical fitness levels according to the influence degree, and obtains a physical fitness monitoring assessment result. The strain direction change states include shear strain changes, vertical strain changes, and strain gradient distribution. The strain amplitude change trends include the strain gradient change amplitudes in adjacent time periods, the skin surface strain change rate, and the subcutaneous vertical strain change rate. The skin strain sensitive area optimization results include strain data deviation values, sensitive area adjustment thresholds, and corrected monitoring area weights. The physical fitness status classification results include exercise load change trends, physical strain response types, and exercise intensity classifications. The physical fitness monitoring and evaluation results include monitoring indicator correlation analysis, physical fitness status evaluation, and exercise load impact analysis.

[0037] See also Figure 2 and Figure 3 , the motion direction recognition module includes:

[0038] The strain data acquisition submodule obtains data collected by the graphene strain sensor, calculates the change in shear strain of the skin surface and the change in vertical strain of the deep subcutaneous layer, matches the changes in shear strain and vertical strain according to the time series, establishes a data correspondence, and obtains a matching strain data set;

[0039] The strain data acquisition submodule is used to obtain the strain data collected by the graphene strain sensor, including but not limited to timestamp, shear strain change and vertical strain change. First, the sensor obtains real-time strain data. The data format includes sensor ID, timestamp, shear strain change and vertical strain change, and is stored in the data buffer in chronological order. For example, the shear strain change collected at a certain time t1 is 0.0023, and the vertical strain change is 0.0011. The data is marked and stored. Then, the collected data is preprocessed to filter out abnormal data, such as whether the data exceeds the measurement range of the sensor (assuming the shear strain change is 0.0023 and the vertical strain change is 0.0011). The range of shear strain change is -0.005-0.005, and the range of vertical strain change is -0.003-0.003). If the data exceeds this range, the data point is eliminated. Subsequently, the shear strain change and the vertical strain change are matched in time series. The data with the same timestamp are compared by index and arranged in ascending order. For example, the data point at time t1 is (0.0023, 0.0011), and the data point at time t2 is (0.0025, 0.0013). Finally, a matching strain dataset is formed, which stores the corresponding relationship between the shear strain change of the skin surface and the vertical strain change of the deep subcutaneous layer at the same timestamp.

[0040] The strain gradient calculation submodule extracts the difference between the surface shear strain change and the deep vertical strain change based on the matching strain dataset, calculates the skin strain gradient value, and screens outliers that exceed the set strain gradient anomaly threshold based on the time series change range. The average of the gradient changes of multiple consecutive time series before and after the outlier is calculated as the baseline value, and the outlier is corrected to obtain the corrected skin strain gradient value.

[0041] The strain gradient calculation submodule is based on the matching strain data set. First, the difference between the surface shear strain change and the deep vertical strain change, that is, the strain gradient value, is calculated as Δε=ε s -ε d , where ε s represents the shear strain change, ε d Indicates the vertical strain change, for example, at time t1, ε s =0.0023,ε d=0.0011, then Δε=0.0023-0.0011=0.0012. Then, the strain gradient value is screened in time series to determine the abnormal point. Assuming that the strain gradient abnormal threshold is set to 0.002, if |Δε| exceeds the threshold, it is marked as an abnormal point. For example, if Δε at a certain time t5 is 0.0025, then the point is marked as an abnormal point. Subsequently, the abnormal point is corrected, and the mean of the gradient changes of multiple consecutive time series before and after the abnormal point is calculated as the reference value. Assuming that the three points before and after are selected for mean calculation, for example, Δε at time t2, t3, and t4 are 0.0011, 0.0013, and 0.0012 respectively, then the reference value is The strain gradient value of the abnormal point is adjusted to the reference value, and finally the corrected skin strain gradient value is obtained.

[0042] The strain direction change identification submodule calls the corrected skin strain gradient value, calculates the direction angle change under multiple time series, determines the strain direction change trend, screens the time period where the change rate exceeds the angle change rate threshold, classifies the change trend, and obtains the strain direction change status;

[0043] Calculate the direction angle change θ under multiple time series t , using the formula:

[0044]

[0045] Calculate the direction angle change, determine the strain direction change trend, filter the time period where the change rate exceeds the angle change rate threshold, classify the change trend, and obtain the strain direction change state; where θ t Represents the direction angle change at time t of the time series, Δε i,t represents the strain gradient change value of the i-th monitoring point at time t, Δε i,t-1 represents the strain gradient change value of the same monitoring point at the previous time t-1, n represents the total number of monitoring points involved in the calculation in the current time series, d j,t represents the distance measurement data of the jth monitoring point at time t, d j,t-1 represents the ranging data of the same monitoring point at time t-1, and m represents the number of monitoring points whose ranging data are involved in the calculation in the current time series;

[0046] Detailed explanation of the formula and the process of formula calculation: The parameter n represents the number of strain sensors used for calculation in the current time series. The number of sensors obtained from the field test is 5, among which the strain gradient values of sensor No. 1 at time t and t-1 are Δε respectively. 1,t =0.0020 and Δε 1,t-1 =0.0018, the calculated gradient change is:

[0047] |Δε 1,t-Δε 1,t-1 |=|0.0020-0.0018|=0.0002;

[0048] The data of sensor No. 2 is Δε 2,t =0.0023, Δε 2,t-1 =0.0021, we get:

[0049] |0.0023-0.0021|=0.0002;

[0050] The data of sensor No. 3 is Δε 3,t =0.0019, Δε 3,t-1 =0.0017,

[0051] |0.0019-0.0017|=0.0002;

[0052] The data of sensor No. 4 is Δε 4,t =0.0021, Δε 4,t-1 =0.0020,

[0053] |0.0021-0.0020|=0.0001;

[0054] The data of sensor No. 5 is Δε 5,t =0.0024, Δε 5,t-1 =0.0022,

[0055] |0.0024-0.0022|=0.0002;

[0056] The sum of the absolute values of all gradient changes is:

[0057] 0.0002+0.0002+0.0002+0.0001+0.0002=0.0009;

[0058] Find the average:

[0059] The parameter m represents the number of ranging sensors. The number of ranging sensors confirmed by on-site detection is 3. The ranging data of sensor No. 1 at time t and t-1 are d 1,t =2.00mm and d 1,t-1 =2.01mm, calculate the difference: d 1,t -d 1,t-1 =2.00-2.01=-0.01;

[0060] After squaring: (-0.01) 2 =0.0001;

[0061] No. 2 distance measurement data d 2,t=2.05mm,d 2,t-1 =2.04mm, calculated as:

[0062] (2.05-2.04) 2 =0.0001;

[0063] No. 3 distance measurement data d 3,t =1.98mm,d 3,t-1 =1.99mm,

[0064] (1.98-1.99) 2 =0.0001;

[0065] The sum of the squares of all distance measurement changes is:

[0066] 0.0001+0.0001+0.0001=0.0003;

[0067] Calculate the square root:

[0068] Add 1 and take the reciprocal:

[0069] Calculate the product of the average strain gradient difference and the inverse of the distance measurement: 0.00018×0.98299=0.000177;

[0070] The final calculated angle change: θ t =arctan(0.000177);

[0071] Since the angle change is very small, the small angle approximation method can be used: θ t ≈0.000177;

[0072] The results show that the direction angle change at time t in the time series is 0.000177, which means that the direction of the skin strain gradient has a small angle change at this time point. The direction change is small, which means that the skin deformation trend at this moment is relatively stable. This value is further used to judge the direction change trend. If the θ t If the value shows an increasing or decreasing trend, it can be determined that the current skin strain direction has changed significantly. Combined with the angle change rate threshold, the time period of drastic change is screened out, and the change trend is classified to finally obtain the strain direction change status.

[0073] like Figure 2 and Figure 4 As shown, the motion amplitude monitoring module includes:

[0074] The strain amplitude extraction submodule calls the skin strain gradient change value of each period in the strain direction change state, sorts it according to the time series, records the strain amplitude information of the complete period, and obtains the strain amplitude data of the period;

[0075] The strain amplitude extraction submodule calls the skin strain gradient change value of each period in the strain direction change state. First, the strain gradient change values of each period are sorted according to the time series to ensure that the data are arranged in chronological order. For example, in the time series t1, t2, t3, and t4, the skin strain gradient change values are 0.0012, 0.0015, 0.0018, and 0.0013 respectively. The data are stored in the order of t1→t2→t3→t4. Then, the strain amplitude of each period is calculated. The amplitude calculation method is the maximum gradient change value in the period and The difference in the minimum gradient change value, for example, in the period t1→t3, the maximum value is 0.0018 and the minimum value is 0.0012, then the strain amplitude of this period is 0.0018-0.0012=0.0006. Subsequently, all time periods are traversed and the strain amplitude information of each complete time period is recorded to form a strain amplitude data set. This data set stores the strain amplitude values in all time periods and is arranged in time series order. For example, the amplitude of t1→t3 is 0.0006, and the amplitude of t2→t4 is 0.0005. Finally, the strain amplitude data of the time period is obtained.

[0076] The adjacent time period amplitude calculation submodule calculates the skin strain amplitude change in adjacent time periods based on the time period strain amplitude data, extracts the average change rate of the time period where the change rate exceeds the amplitude fluctuation threshold, sets the mutation point based on the average value and marks the time sequence, calculates the mean deviation of the two time periods before and after the mutation point, adjusts the mutation point data that exceeds the set deviation range, recalculates the adjusted amplitude change rate, and obtains the amplitude change rate of the adjacent time periods;

[0077] The adjacent time period amplitude calculation submodule is based on the time period strain amplitude data. First, the skin strain amplitude change in adjacent time periods is calculated. The amplitude change is calculated as the difference between the amplitudes of adjacent time periods. For example, the amplitude from t1 to t3 is 0.0006, and the amplitude from t2 to t4 is 0.0005. The change is 0.0006-0.0005=0.0001. Then, the time period with a change rate exceeding the amplitude fluctuation threshold is screened. Assuming that the amplitude fluctuation threshold is set to 0.00015, if the change rate is greater than the threshold, the time period is marked. For example, in the time period from t3 to t5, the amplitude change rate is calculated as follows: If the calculated result is 0.0002, which is greater than 0.00015, the period is marked. Subsequently, the mean deviation of the two periods before and after the mutation point is calculated. The mean deviation is calculated as the difference between the means of the two periods before and after t3. For example, the mean of the two periods before t3 is 0.00055, and the mean of the two periods after t3 is 0.00065. The mean deviation is 0.00065-0.00055=0.0001. Next, it is determined whether the deviation exceeds the set deviation range. Assume that the deviation range is set to 0.00012. If the calculated value is less than the range, no adjustment is required. Otherwise, adjustment is performed. For example, if the calculated value is 0.00014, which is out of range, the mutation point amplitude is adjusted to the mean of the adjacent period. For example, the mutation point data is adjusted to Finally, based on the adjusted amplitude data, the adjusted amplitude change rate is recalculated to obtain the amplitude change rate of adjacent time periods.

[0078] The amplitude change trend analysis submodule calls the amplitude change rate of adjacent time periods, calculates the amplitude change trend values of multiple time series, filters the time periods with continuous growth or decline in the change rate, classifies the trend pattern, and obtains the strain amplitude change trend;

[0079] Calculate the amplitude change trend value of different time series using the formula:

[0080]

[0081] Calculate the amplitude change trend value, filter the time period with continuous growth or decline in the change rate, classify the trend pattern, and obtain the strain amplitude change trend;

[0082] Among them, V t1 Represents the amplitude value at time point t1, V t1-1 Represents the amplitude value at the previous time point t1-1, Δt1 represents the time interval, usually the measurement period, V avg represents the average value of the amplitude value in all measurement periods, T represents the length of the total time series, ΔV trend Represents the calculated amplitude change trend value;

[0083] Detailed explanation of the formula and the process of formula calculation and derivation:

[0084] Consider a practical scenario where the strain amplitude is measured over a time series (T=4 moments). Assume that the time interval Δt1 is 1 second and the amplitude value V t1 At four consecutive time points t1 = 1, 2, 3, and 4, the records are 2.0, 2.5, 3.0, and 2.8 units respectively. First, calculate the amplitude change rate at each time point, then calculate the weight of each rate, and finally get the trend value.

[0085] Calculate the rate of change of amplitude at each time point: from t1=1 to t1=2:

[0086] From t1=2 to t1=3: From t1=3 to t1=4:

[0087] Calculate the average amplitude value V avg :

[0088] Calculate the weighted amplitude change rate: |V2-V avg |=|2.5-2.575|=0.075;

[0089] |V3-V avg |=|3.0-2.575|=0.425;

[0090] |V4-V avg =|2.8-2.575|=0.225;

[0091] Apply the formula to calculate the trend value ΔV trend :

[0092]

[0093] The results indicate an average amplitude trend value of 0.0683, reflecting the weighted value of the average rate of change across the entire time series, taking into account the degree to which the amplitude of each rate deviates from the average. This numerical trend indicates that the strain amplitude exhibits an overall increasing trend over the considered time period, with the strain amplitude at t = 3 being particularly high, exceeding the average and contributing most to the overall trend. In this way, the model can effectively classify and identify periods of sustained growth or decline, thereby assisting in analyzing the overall dynamics of strain amplitude changes.

[0094] like Figure 2 and Figure 5 As shown, the dynamic sensitive area adjustment module includes:

[0095] The deviation calculation submodule calls the strain amplitude change trend, calculates the deviation value between the current strain data and the motion load monitoring area distribution, calculates the deviation mean of the area beyond the set deviation range, and obtains the regional deviation distribution result;

[0096] The deviation calculation submodule calls the strain amplitude change trend to calculate the deviation value between the current strain data and the distribution of the motion load monitoring area. First, the current strain data is determined. The data format contains the timestamp, the strain amplitude change trend value and the corresponding monitoring area number. For example, the strain amplitude change trend value at a certain time t1 is 0.0008, and the monitoring area number is A1. Secondly, the historical distribution data of the motion load monitoring area is extracted. The data stores the mean and variance of the strain amplitude in multiple time series in the past of the area. For example, the historical mean of area A1 is 0.0006 and the variance is 0.0001. Then, the deviation value between the current strain data and the historical distribution data of the area is calculated. The deviation value calculation method is |ε 当前 -ε 历史均值 |, where ε 当前 Indicates the current strain amplitude change trend value, ε 历史均值 Indicates the historical mean of the area, for example, |0.0008-0.0006|=0.0002. If the deviation value exceeds the set deviation range, the area is marked as an out-of-range area. Subsequently, the deviation mean is calculated for all areas that exceed the set deviation range. The deviation mean is calculated as the average of the deviation values of the out-of-range areas. For example, the deviation values of areas A1, A2, and A3 are 0.0002, 0.0003, and 0.00025 respectively, then the deviation mean is Finally, the regional deviation distribution results are obtained.

[0097] The sensitive area adjustment threshold setting submodule calculates the deviation fluctuation rate of the area beyond the set deviation range based on the regional deviation distribution results, calculates the sensitive area adjustment threshold according to the deviation mean and the fluctuation rate, determines the change trend of the sensitive area adjustment threshold, adjusts the adjustment range of the abnormal area, and obtains the updated sensitive area adjustment threshold;

[0098] The sensitive area adjustment threshold setting submodule calculates the deviation fluctuation rate of the area beyond the set deviation range based on the regional deviation distribution results. First, all areas beyond the deviation range are determined and the deviation values of the area in multiple time series before and after are extracted. For example, the deviation values of area A1 at t1, t2, and t3 are 0.0002, 0.00015, and 0.0003 respectively. Then, the deviation fluctuation rate is calculated as the ratio of the change in the deviation value at adjacent moments to the time interval, for example Assuming Δt = 0.2s, the fluctuation rate is Next, the sensitive area adjustment threshold is calculated based on the deviation mean and fluctuation rate. The adjustment threshold is calculated as K·ε 偏差均值 +B·ε 波动速率, where K and B are adjustment coefficients. Assuming K = 2 and B = 1, the adjustment threshold is 2×0.00025+1×0.00075=0.00125. Subsequently, the changing trend of the adjustment threshold of the sensitive area is determined. The trend determination method is the change direction of the adjustment threshold at multiple adjacent time points. For example, if the adjustment thresholds of t1, t2, and t3 are 0.001, 0.0012, and 0.00125 respectively, it is determined to be an upward trend. If the adjustment threshold of a certain area fluctuates abnormally, the adjustment range of the area is adjusted. The adjustment method is to smooth the adjustment threshold of the abnormal area to the average value of the adjacent areas. For example, if the average value of the threshold of the adjacent areas is 0.0011, the abnormal area is adjusted to 0.0011, and finally the updated adjustment threshold of the sensitive area is obtained.

[0099] The monitoring area weight optimization submodule calls the updated sensitive area adjustment threshold, adjusts the sensitive area weight, calculates the adjusted weight balance, filters the areas where the weight change exceeds the adjustment range, and obtains the optimization results of the skin strain sensitive area;

[0100] The monitoring area weight optimization submodule calls the updated sensitive area adjustment threshold to adjust the sensitive area weight. First, the weight of each monitoring area is calculated based on the updated adjustment threshold. The weight calculation method is: For example, if the adjustment threshold of a region is 0.00125, its weight calculation is Next, calculate the adjusted weight balance. The balance is calculated as the standard deviation of all region weights. For example, the weights of regions A1, A2, and A3 are 800, 850, and 780 respectively. The standard deviation is calculated as Then, the areas whose weight changes exceed the adjustment range are screened. The screening method is to determine whether the weight change of a certain area exceeds the set weight fluctuation threshold. Assuming that the weight fluctuation threshold is set to 30, if the weight change of a certain area is greater than 30, the area is marked. For example, the weight of a certain area changes from 780 to 850, and the change is 70, which is greater than 30, then the area is marked. Finally, the weight of the marked area is adjusted to make it tend to the mean of the adjacent areas. For example, if the mean of the adjacent areas is 815, the weight of the area is adjusted to 815, and finally the optimization result of the skin strain sensitive area is obtained.

[0101] like Figure 2 and Figure 6 As shown, the physical state assessment module includes:

[0102] The motion load trend analysis submodule calls the motion load change data of each sensitive area in the skin strain sensitive area optimization results, calculates the load change rate under the time series, filters the areas where the change rate exceeds the set change rate threshold, and obtains the motion load change trend;

[0103] The motion load trend analysis submodule calls the motion load change data of each sensitive area in the skin strain sensitive area optimization results. First, the motion load data of each sensitive area is extracted. The data format includes the timestamp, area number and the load value of the area at that moment. For example, at time point t1, the motion load value of area A1 is 5.2N, and the load value of area A2 is 4.8N. Then, the load change rate under the time series is calculated. The calculation method is the ratio of the load value change at adjacent time points to the time interval, that is, Among them F t1 ,F t2 Represent the load values at time t1 and t2 respectively. Assuming Δt = 0.5s, if F t1 =5.2N, F t2 =5.5N, then the load change rate is Next, filter out areas where the rate of change exceeds the set change rate threshold, and set the threshold to 0.5N / s. If the calculated result is greater than the threshold, mark the area. For example, if the rate of change in area A1 is 0.6N / s, which exceeds the threshold, area A1 is marked. Traverse all areas and record all time periods that exceed the threshold to finally obtain the trend of exercise load changes.

[0104] The strain response classification submodule extracts sensitive areas with similar load change rates based on the trend of exercise load changes, calculates the load change similarity between areas, filters areas with similarity exceeding the set similarity threshold, classifies similar load change areas, and obtains the physical strain response type;

[0105] Calculate the load change similarity S between regions ij , using the formula:

[0106] Calculate the similarity of load changes, filter out areas where the similarity exceeds the set similarity threshold, classify similar load change areas, and obtain the physical strain response type;

[0107] Among them, R i1,t2 represents the load change rate of region i1 at time t2, R j1,t2 represents the load change rate of region j1 at time t2, T represents the total time series length, S i1,j1 Represents the load change similarity between regions i1 and j1. The calculation method is based on the improved Jaccard similarity. At each time point t2, the minimum and maximum load rates of the two regions are calculated and the ratio is accumulated in the entire time series. When the change trends of the two regions are close, the similarity score is higher. If the load change value of one region is large for a long time, the similarity is reduced. This ensures that the local matching of the change trend is more sensitive and avoids the dependence of the absolute value size of the methods such as Euclidean distance.

[0108] Formula explanation and calculation process: Load change rate R of area A and area B in time series T = 5 moments i,t The data is collected by the sensor in Newtons per second. The data is as follows: R A,t2 =[2.5,3.2,2.9,3.5,3.1]; R B,t2 =[2.7,3.0,2.8,3.6,3.3];

[0109] Calculate the minimum load change rate at each time point in the time series: min(2.5,2.7)=2.5; min(3.2,3.0)=3.0; min(2.9,2.8)=2.8;

[0110] min(3.5,3.6)=3.5; min(3.1,3.3)=3.1;

[0111] Total = 2.5 + 3.0 + 2.8 + 3.5 + 3.1 = 14.9;

[0112] Calculate the maximum load change rate at each time point in the time series: max(2.5,2.7)=2.7; max(3.2,3.0)=3.2; max(2.9,2.8)=2.9; max(3.5,3.6)=3.6; max(3.1,3.3)=3.3; total=2.7+3.2+2.9+3.6+3.3=15.7;

[0113] Calculate the load change similarity S ij :

[0114] The results show that the load change trends of region A and region B at the five time points are highly similar, with a similarity of 0.948, close to 1, which means that the load change rates of the two regions are highly consistent in the time series. This value can be used to screen regions with similar load changes. If the value exceeds the set similarity threshold, it can be determined that the two regions belong to the same load change type and are then classified into the corresponding physical strain response type.

[0115] The physical state classification submodule calls the physical strain response type, classifies the physical strain areas of the same category, calculates the load balance between the classified areas, filters out the categories whose load balance deviation exceeds the load balance deviation threshold, divides the physical state into multiple areas, and obtains the physical state classification results;

[0116] The physical state classification submodule calls the physical strain response type to classify the physical strain areas of the same category. First, the load balance between the classified areas is calculated. The calculation method is the standard deviation of the load values of all classified areas. For example, the load values of areas A1, A2, and A3 are 5.2N, 5.4N, and 5.1N respectively. The standard deviation is calculated as follows: Then, the categories whose load balance deviation exceeds the load balance deviation threshold are screened. Assume that the threshold is set to 0.15N. If the standard deviation is less than the threshold, the category is determined to be load balanced. Otherwise, it is divided. For example, if the standard deviation of a certain category is 0.18N, which is greater than the threshold, the category needs to be further split. Then, multiple physical status areas are divided according to the load balance situation. If the load deviation of the internal area of a category is large, it is split into two independent categories. For example, the original category B1 contains areas A1, A2, and A3, but because the load balance of area A3 exceeds the range, it is split into B1 and B2, and finally the physical status classification result is obtained.

[0117] like Figure 2 and Figure 7 As shown, the physical fitness indicator monitoring module includes:

[0118] The physical fitness monitoring data integration submodule extracts the corresponding monitoring data from the physical fitness status classification results, refers to the exercise load, physiological feedback and external environmental parameters, and matches the time series, eliminating the data with synchronization errors outside the error range to obtain time-series aligned monitoring data;

[0119] The physical fitness monitoring data integration submodule extracts the corresponding monitoring data from the physical fitness state classification results. First, the monitoring data under each physical fitness state category is extracted. The data format includes timestamp, exercise load value, physiological feedback parameters (such as heart rate, blood oxygen saturation, etc.) and external environmental parameters (such as temperature, humidity, etc.). For example, at a certain time t1, the monitoring data includes a load value of 6.2N, a heart rate of 80bpm, a blood oxygen saturation of 97, and an ambient temperature of 25°C. Secondly, all monitoring data are matched according to the time series to ensure that each data source corresponds to the same timestamp. For example, at time point t1, the values of each parameter are 6.2N, 8 0bpm,97, then, calculate the synchronization error between the data. The calculation method is the time deviation between different data sources. For example, if the recording time of the exercise load data is t1=10:00:00, and the recording time of the physiological feedback data is t1'=10:00:02, then the synchronization error is t1'-t1=2s. Then, filter out the data that exceeds the set error range. Assume that the error range is set to 1s. If the data error exceeds the range, the data point is eliminated. For example, if the error of a data point is 3s, the data is excluded. Traverse all time points and eliminate the data with errors exceeding the range to finally obtain the time-series aligned monitoring data.

[0120] The monitoring indicator impact analysis submodule analyzes the correlation of each monitoring indicator based on the time-series alignment monitoring data, calculates the contribution of each indicator to the physical status classification, selects indicators whose contribution exceeds the set contribution standard, summarizes the impact relationship of the contribution indicators, and obtains the impact degree of the monitoring indicators;

[0121] The monitoring indicator impact analysis submodule analyzes the correlation of each monitoring indicator based on the time-series alignment monitoring data. First, the correlation coefficient between each monitoring indicator is calculated using the Pearson correlation coefficient. Where Cov(X,Y) is the covariance of indicators X and Y, σ X ,σ Y are the standard deviations of indicators X and Y, respectively. For example, the correlation coefficient between heart rate and load is 0.85, indicating that there is a strong positive correlation between the two. Next, the contribution of each indicator to the physical state classification is calculated. The contribution calculation method is the discrimination of a certain indicator in different physical state categories. The calculation method is Among them, μ1 and μ2 are the means of the indicators in the two categories respectively, and σ is the standard deviation. For example, in physical state 1 and state 2, the mean heart rate is 80bpm and 90bpm respectively, and the standard deviation is 5, then the contribution is Then, screen the indicators whose contribution exceeds the set contribution standard. Assume that the set standard is 1.5. If the contribution of an indicator is greater than the standard, the indicator is screened. For example, the contribution of heart rate is 2.0, which is greater than 1.5, then the indicator is selected. Traverse all indicators and summarize the influence relationship of contribution indicators to finally obtain the influence degree of monitoring indicators.

[0122] The physical fitness monitoring and evaluation generation submodule calls the impact degree of monitoring indicators, calculates the change range of indicators under each physical fitness state, analyzes the differences of indicators under each state category, selects indicators with differences exceeding the physical fitness state difference threshold as characteristic indicators, extracts the change trend of characteristic indicators and divides them into physical fitness levels to obtain physical fitness monitoring and evaluation results;

[0123] The physical fitness monitoring evaluation generation submodule calls the degree of influence of the monitoring indicators and calculates the range of changes in the indicators under each physical fitness state. First, the range of changes of each indicator in different physical fitness state categories is calculated as the difference between the maximum and minimum values. For example, in physical fitness state 1, the heart rate range is 75bpm-85bpm, and the range of change is 85-75=10bpm. Then, the difference of the indicators in each state category is analyzed. The calculation method is the difference value of the mean value of the indicators in different categories. For example, the mean heart rate of state 1 is 80bpm, and the mean heart rate of state 2 is 90bpm, then the difference value is 90-80=10bpm. Then, the indicators whose differences exceed the physical fitness state difference threshold are selected as feature indicators. Assuming that the threshold is set to 8bpm, if the difference value of an indicator is greater than the threshold, it is selected as a feature indicator. For example, the difference value of the heart rate is 10bpm, and it is greater than 8bpm, it is selected. Subsequently, the change trend of the feature indicator is extracted, and the trend value is calculated as the rate of change of adjacent time points, for example If Δt=2s, the rate of change is Finally, the physical fitness level is divided according to the changing trend of the characteristic indicators. For example, if the heart rate change trend exceeds 1.2bpm / s, it is classified as a high-intensity state, otherwise it is classified as a moderate-intensity state, and finally the physical fitness monitoring evaluation results are obtained.

[0124] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A wearable fitness monitoring system based on graphene, characterized by: The system comprises: The movement direction recognition module obtains the skin surface shear strain and subcutaneous vertical strain data collected by the graphene strain sensor, calculates the skin strain gradient based on the strain change rate of the skin surface and subcutaneous deep layer in adjacent time periods, and obtains the strain direction change state; The movement amplitude monitoring module calls the skin strain gradient change value of each time period based on the strain direction change state, calculates the skin strain amplitude of adjacent time periods, and obtains the strain amplitude change trend; The dynamic sensitive area adjustment module refers to the distribution deviation between the current strain data and the motion load monitoring area in the strain amplitude change trend, takes the monitoring area exceeding the preset deviation threshold as the sensitive area, adjusts the strain weight of the sensitive area, and obtains the strain sensitive area optimization result; The physical state evaluation module determines the exercise load change trend of each sensitive area in the strain sensitive area optimization result, classifies the physical strain response types of the same exercise intensity category, and obtains a physical state classification result; The physical fitness indicator monitoring module analyzes the mutual influence degree of each type of physical fitness indicator based on the physical fitness status classification result, divides the physical fitness level according to the influence degree, and obtains the physical fitness monitoring evaluation result.

2. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: The strain direction change state includes shear strain change, vertical strain change, and strain gradient distribution. The strain amplitude change trend includes the strain gradient change amplitude in adjacent time periods, the skin surface strain change rate, and the subcutaneous vertical strain change rate. The skin strain sensitive area optimization result includes the strain data deviation value, the sensitive area adjustment threshold, and the corrected monitoring area weight. The physical status classification result includes the exercise load change trend, physical strain response type, and exercise intensity classification. The physical monitoring evaluation result includes monitoring indicator correlation analysis, physical status evaluation, and exercise load impact analysis.

3. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: The motion direction recognition module includes: The strain data acquisition submodule obtains data collected by the graphene strain sensor, calculates the change in shear strain of the skin surface and the change in vertical strain of the deep subcutaneous layer, matches the changes in shear strain and vertical strain according to the time series, establishes a data correspondence, and obtains a matching strain data set; The strain gradient calculation submodule extracts the difference between the surface shear strain change and the deep vertical strain change based on the matching strain data set, calculates the skin strain gradient value, screens outliers that exceed the set strain gradient anomaly threshold based on the time series change range, calculates the average of the gradient changes of multiple consecutive time series before and after the outlier as the baseline value, corrects the outlier, and obtains the corrected skin strain gradient value; The strain direction change identification submodule calls the corrected skin strain gradient value, calculates the direction angle change under multiple time series, determines the strain direction change trend, screens the time period where the change rate exceeds the angle change rate threshold, classifies the change trend, and obtains the strain direction change status.

4. The graphene-based wearable fitness monitoring system according to claim 1, wherein: For calculating the direction angle change θ under multiple time series t , using the formula: Among them, Δε i,t represents the strain gradient change value of the i-th monitoring point at time t, Δε i,t-1 represents the strain gradient change value of the i-th monitoring point at the previous moment t-1, n represents the number of monitoring points involved in the calculation in the current time series, d j,t represents the distance measurement data of the jth monitoring point at time t, d j,t-1 represents the ranging data of the jth monitoring point at time t-1, and m represents the number of monitoring points whose ranging data are involved in the calculation in the current time series.

5. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: The motion amplitude monitoring module includes: The strain amplitude extraction submodule calls the skin strain gradient change value of each period in the strain direction change state, sorts it according to the time series, records the strain amplitude information of the complete period, and obtains the strain amplitude data of the period; The adjacent time period amplitude calculation submodule calculates the skin strain amplitude change in adjacent time periods based on the strain amplitude data of the time periods, extracts the average change rate of the time periods where the change rate exceeds the amplitude fluctuation threshold, sets the mutation point based on the average value and marks the time sequence, calculates the average deviation of the two time periods before and after the mutation point, adjusts the mutation point data where the deviation exceeds the set deviation range, recalculates the adjusted amplitude change rate, and obtains the amplitude change rate of the adjacent time periods; The amplitude change trend analysis submodule calls the amplitude change rate of the adjacent time periods, calculates the amplitude change trend values of multiple time series, filters the time periods with continuous growth or decline in the change rate, classifies the trend pattern, and obtains the strain amplitude change trend.

6. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: For calculating the amplitude change trend value ΔV of different time series trend , using the formula: Among them, V t1 Represents the amplitude value at time point t1, V t1-1 represents the amplitude value at the previous time point t1-1, Δt represents the time interval, usually the measurement period, V avg represents the average value of the amplitude value in all measurement periods, T represents the length of the total time series, ΔV trend Represents the calculated amplitude change trend value.

7. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: The dynamic sensitive area adjustment module includes: The deviation calculation submodule calls the strain amplitude change trend, calculates the deviation value between the current strain data and the motion load monitoring area distribution, calculates the deviation mean of the area beyond the set deviation range, and obtains the regional deviation distribution result; The sensitive area adjustment threshold setting submodule calculates the deviation fluctuation rate of the area exceeding the set deviation range based on the regional deviation distribution result, calculates the sensitive area adjustment threshold according to the deviation mean and the fluctuation rate, determines the change trend of the sensitive area adjustment threshold, adjusts the adjustment range of the abnormal area, and obtains the updated sensitive area adjustment threshold; The monitoring area weight optimization submodule calls the updated sensitive area adjustment threshold, adjusts the sensitive area weight, calculates the adjusted weight balance, filters the area where the weight change exceeds the adjustment range, and obtains the skin strain sensitive area optimization result.

8. The graphene-based wearable fitness monitoring system according to claim 1, wherein: The physical condition assessment module includes: The motion load trend analysis submodule calls the motion load change data of each sensitive area in the skin strain sensitive area optimization result, calculates the load change rate in the time series, filters the areas where the change rate exceeds the set change rate threshold, and obtains the motion load change trend; The strain response classification submodule extracts sensitive areas with similar load change rates based on the exercise load change trend, calculates the load change similarity between the areas, filters areas with similarity exceeding a set similarity threshold, classifies similar load change areas, and obtains the physical strain response type; The physical state classification submodule calls the physical strain response type, classifies physical strain areas of the same category, calculates the load balance between the classified areas, filters out categories whose load balance deviation exceeds the load balance deviation threshold, divides the area into multiple physical state areas, and obtains physical state classification results.

9. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: For the calculation of the load change similarity S between regions i1,j1 , using the formula: Among them, R i1,t2 represents the load change rate of region i1 at time t2, R j1,t2 represents the load change rate of region j1 at time t2, T represents the total time series length, S i1,j1 Represents the load change similarity between region i1 and region j1.

10. The graphene-based wearable fitness monitoring system according to claim 1, characterized in that: The physical fitness indicator monitoring module includes: The physical fitness monitoring data integration submodule extracts the corresponding monitoring data from the physical fitness status classification results, refers to the exercise load, physiological feedback and external environmental parameters, matches the time series, eliminates the data with synchronization errors exceeding the error range, and obtains time-series aligned monitoring data; The monitoring indicator impact analysis submodule analyzes the correlation of each monitoring indicator based on the time-series aligned monitoring data, calculates the contribution of each indicator to the physical state classification, selects indicators whose contribution exceeds the set contribution standard, summarizes the impact relationship of the contribution indicators, and obtains the impact degree of the monitoring indicators; The physical fitness monitoring and evaluation generation submodule calls the influence degree of the monitoring indicators, calculates the indicator change range under each physical fitness state, analyzes the differences of the indicators under each state category, selects indicators whose differences exceed the physical fitness state difference threshold as feature indicators, extracts the change trend of the feature indicators and divides the physical fitness levels to obtain the physical fitness monitoring and evaluation results.

Images