An information-based sports education platform management system

Abstract

The present invention discloses an information-based sports education platform management system, which specifically relates to the field of sports training data management, and is used to solve the problem of accurate classification and feedback of multi-device training data. Firstly, based on the fluctuation characteristics of physiological data, the switching period data is captured to ensure the continuity and timeliness of the training data; then, the transition response consistency index and the target device feature matching degree are used to accurately define the effective transition data, clearly distinguish the physiological feature data of each device, and improve the classification accuracy; the defined effective transition data is finely classified and labeled through the target device dynamic feature model to ensure the independence and consistency of the data; finally, an adaptive rejection mechanism is used to screen out abnormal data, retain high-quality data to generate personalized feedback, enhance the pertinence of the feedback, and build a closed-loop system from data collection to feedback generation, providing scientific support for sports training data management.

Description

Technical Field

[0001] The present invention relates to the field of sports training data management, and more specifically, to an information-based sports education platform management system. Background Art

[0002] With the in-depth development of information technology, modern sports training has gradually integrated intelligent equipment. Users can obtain detailed physiological data feedback by wearing physiological data monitoring devices and combining them with a variety of training equipment. These devices can dynamically observe the user's training intensity and physical response through real-time monitoring of physiological data such as heart rate, respiratory rate, and blood oxygen saturation, thereby providing data support for scientific and personalized training plans. In multi-equipment training, physiological data monitoring can accurately reflect the user's physiological changes under different equipment, gradually forming a multidimensional data set of training characteristics of different equipment, and providing precise guidance for subsequent training optimization.

[0003] However, in the continuous switching of multi-machine training, the boundary problem of physiological data gradually emerges, especially when users frequently switch between different types of machines (such as switching from aerobic machines to strength training machines), the physiological data will have a certain lag effect. For example, indicators such as heart rate and respiratory rate may not have returned to a stable state after strength training, that is, the user has switched to another machine to continue training, which causes the data to have a "transition segment" phenomenon in time. The physiological data of this transition segment is difficult to accurately classify to a specific machine, affecting the clarity and accuracy of the evaluation of the training effect. In the prior art, the system lacks a data recognition and elimination mechanism for the transition period of machine switching, and cannot accurately divide the physiological data characteristics of each machine, resulting in deviations in the training feedback data and affecting the accuracy of personalized training recommendations.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an information-based sports education platform management system. First, when the equipment is switched, the transition state is identified through the fluctuation characteristics of physiological data, and the physiological data during the switching period is fully recorded to ensure the continuity and timeliness of data acquisition, and to construct a complete training data stream, laying the foundation for subsequent analysis. Secondly, based on the transition response consistency index and the target equipment feature matching degree, the physiological data transition segment in the multi-equipment switching process is clearly defined, and the feature data of each equipment is accurately divided to ensure the accuracy and independence of classification, avoid the analysis deviation caused by data crossover, and form a high-quality training data management model. For effective transition data segments, the scheme uses the dynamic feature model of the target equipment to finely classify and generate labels for the data, so that the data classification strictly corresponds to the characteristics of the target equipment, and improves the independence and scientificity of data management. Finally, the classified data segments are screened through an adaptive elimination mechanism, and data with large volatility or abnormality are eliminated, and personalized feedback is generated by retaining data with high stability and good quality, so that the feedback content is more in line with the user's actual training performance. The overall solution uses multi-module deep data management to form a closed-loop system from physiological data collection, transition period identification to dynamic classification and elimination, which solves the challenges of multi-device training data in terms of continuity, classification accuracy and feedback adaptability, significantly improves the scientific nature of sports training data management and the accuracy of training analysis, and provides users with accurate and personalized training guidance to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An information-based sports education platform management system, comprising: a fluctuation detection module, a transition determination module, a dynamic classification module and an adaptive elimination module;

[0008] Fluctuation detection module: collects physiological data monitored by the device worn by the user, identifies the data fluctuation pattern when the user switches the device, generates a transition signal, and transmits the identified transition data to the transition determination module;

[0009] Transition determination module: Based on the transition response consistency index and the matching degree of the target device characteristics, it determines whether the transition data segment meets the valid transition conditions, generates valid transition data labels, and passes the labeled data to the dynamic classification module;

[0010] Dynamic classification module: According to the target device feature model, the defined valid transition data is dynamically classified, the corresponding device data label is generated, and the classified data is passed to the adaptive rejection module;

[0011] Adaptive elimination module: Adaptively eliminates classified data based on elimination rules and outputs personalized training feedback results.

[0012] In a preferred embodiment, the fluctuation detection module includes the following contents:

[0013] S1.1, physiological data is collected in real time under the monitoring of the device worn by the user; based on the segmentation strategy of the time window, the physiological data is divided into several time series segments according to each time point of the device switching, and the time of device switching is marked as the starting point of the time period to capture a period of data fluctuation before and after the switching;

[0014] S1.2, for each segment of physiological data, calculate its instantaneous rate of change in time, that is, the amplitude of change; by comparing the degree of change of physiological data at adjacent time points, form a gradient vector of the data, so as to analyze the change trend of physiological data when switching; then perform statistics on the gradient values ​​of each data and calculate its degree of deviation from the standard distribution;

[0015] S1.3, calculate its fluctuation stability index, and judge the stable trend of fluctuation by analyzing the difference between the change amplitudes at adjacent time points;

[0016] S1.4. Finally, by performing comprehensive matching calculations on the fluctuation stability index and change deviation of each physiological data parameter, the matching function value is obtained to determine whether the corresponding data segment is a valid transition signal; if the matching function value exceeds the set threshold, it is considered that the corresponding moment is in a transition state, and the corresponding data segment is marked as transition data.

[0017] In a preferred embodiment, the transition determination module includes the following contents:

[0018] S2.1, the logic for obtaining the transition response consistency index is:

[0019] The calculation of the instantaneous change rate vector is used to obtain the change rate of the transition data in time to capture the fluctuation degree of each physiological parameter during the switching process; specifically, for the data vector at each moment, the transition data at the current moment is subtracted from the corresponding data at the previous moment to obtain the instantaneous change amplitude of each parameter;

[0020] The multidimensional smoothness vector is used to evaluate whether the transition data gradually tends to be stable during the switching process; specifically, the instantaneous change rate vector at the current moment is calculated first, and then it is subtracted from the instantaneous change rate vector at the previous moment to obtain the fluctuation change amplitude of each physiological parameter;

[0021] The consistent cumulative offset is used to evaluate the fluctuation consistency and trend change of the transition data during the entire transition period. First, traverse each moment of the entire transition period and calculate the smoothness vector and instantaneous change rate vector of the current moment. Then, multiply the absolute value of the smoothness vector by the instantaneous change rate, and then divide it by the sequence number of the current moment plus 1, and use this value as the offset of the corresponding moment.

[0022] The transition response consistency index is used to comprehensively evaluate the smoothing trend and consistency of the data during the entire transition period. First, the reciprocal of the consistency cumulative offset is taken to ensure that the smoothly transitioned data has a higher weight in the index calculation. Then, the smoothness vector and the instantaneous change rate at each moment during the transition period are traversed, and the absolute value of the smoothness vector is divided by the instantaneous change rate plus 1 to obtain the smooth fluctuation ratio at each moment. Finally, the smooth fluctuation ratios of all moments are accumulated and multiplied by the reciprocal of the consistency cumulative offset to generate the transition response consistency index.

[0023] In a preferred embodiment, S2.2, the logic for obtaining the target device feature matching degree is:

[0024] At each moment during the transition period, the user's transition data is collected to form an actual transition data feature vector, which is used to represent the user's physiological state at the corresponding moment;

[0025] According to the preset characteristics of the target device, a target device feature vector is generated, which represents the physiological state of the user when using the target device under ideal conditions; it is used as a comparison benchmark and matched one by one with the actual transition data during the transition period;

[0026] At each moment during the transition period, the deviation between the actual transition data feature vector and the target device feature vector is calculated to generate a feature deviation vector; the difference value of each physiological parameter is calculated parameter by parameter to obtain the degree of difference between the actual data and the target feature;

[0027] The feature deviation vectors at each moment are accumulated in sequence to generate the feature matching offset accumulation. In order to give higher weight to the later data in the accumulation process and make it have a greater impact on the matching offset, the deviation vector uses the square of the time factor as the weight when accumulating, so that the later deviation increases the impact on the accumulation.

[0028] The target device feature matching degree is calculated based on the inverse of the cumulative feature matching offset. First, the inverse of the deviation is calculated. Then, the deviations at all moments during the entire transition period are traversed and introduced one by one into the calculation of the final matching degree. Finally, the target device feature matching degree is obtained by accumulation.

[0029] In a preferred embodiment, S2.3, first, the transition response consistency index and the target device characteristic matching degree are dimensionlessly processed, and the two are converted to the same scale; after the dimensionless conversion is completed, the cross-state fitting index is constructed; if the cross-state fitting index is less than or equal to the transition threshold, it means that the physiological data of the transition data segment during the transition period is stable and highly matched with the target device characteristics, and meets the effective transition conditions, then the corresponding transition data segment is marked with a valid transition data label; if the cross-state fitting index is greater than the transition threshold, it indicates that the physiological data fluctuates greatly and does not match the target device characteristics, and the effective transition conditions are not met, so it is not marked.

[0030] In a preferred embodiment, the dynamic classification module includes the following contents:

[0031] S3.1, first, establish the characteristic model of the target device based on the typical physiological data characteristics of the corresponding device; generate the characteristic vector of the target device by analyzing the physiological response curve of the corresponding device under the ideal training state, denoted as M i ;

[0032] S3.2, within the effective transition data segment, the physiological data at each moment is integrated into an instantaneous feature vector, denoted as D i ;

[0033] S3.3, for each instantaneous characteristic vector and target device characteristic vector, calculate the dynamic characteristic deviation between the two to quantify the degree of difference between the actual data and the target device characteristics; the formula for calculating the dynamic characteristic deviation is as follows: Where: Δ i : The dynamic characteristic deviation at the current moment i; and Respectively represent the physiological parameter values ​​of the instantaneous characteristic vector and the target device characteristic vector in the kth dimension at the current moment; n: the number of dimensions of physiological data; ln(t+1): time adjustment factor;

[0034] S3.4, within the valid transition data segment, determine whether the data meets the matching criteria of the target device based on the dynamic feature deviation, and generate the corresponding device data label; if the dynamic feature deviation is less than or equal to the matching threshold, the current valid transition data segment is marked as the target device data; if the dynamic feature deviation is greater than the matching threshold, the corresponding data segment does not meet the target device requirements and is not marked.

[0035] In a preferred embodiment, the adaptive culling module includes the following contents:

[0036] S4.1. For each target device data segment that has been classified, calculate the abnormal deviation index. The calculation of the abnormal deviation index evaluates the stability and deviation of the data segment by combining the deviation accumulation and overall volatility of the multidimensional physiological parameters in the effective transition data segment. First, for each effective transition data segment, calculate the numerical difference between it and the target device characteristics dimension by dimension, and accumulate the total deviation. In the calculation of the deviation, use the time increment factor to make the later data deviation more weighted, so as to pay more attention to the stability performance of the later data segment. In addition, the sum of the fluctuation changes of the physiological parameters at each moment in the effective transition data segment is added to the calculation process to evaluate the overall volatility of the corresponding data segment.

[0037] S4.2, after obtaining the abnormal deviation index, compare it with the adaptive rejection benchmark to screen the target device data segment that meets the rejection conditions; the rejection judgment rule is: if the abnormal deviation index exceeds the adaptive rejection benchmark, the corresponding target device data segment is regarded as an abnormal data segment with large volatility and is rejected; if the abnormal deviation index is lower than or equal to the rejection benchmark, the corresponding target device data segment is retained and used to generate personalized feedback results;

[0038] S4.3, after eliminating invalid data segments, the retained target device data segments are used to generate personalized training feedback.

[0039] The technical effects and advantages of an information-based sports education platform management system of the present invention are as follows:

[0040] Through rigorous multi-module processing, the complete collection, accurate classification and personalized feedback of multi-device training data are achieved, and the accuracy of data management and training analysis are improved. First, when switching between devices, the complete switching period data is captured by identifying the fluctuation characteristics of physiological data to ensure the continuity and timeliness of the data. Subsequently, the transition data segment is accurately defined by combining the transition response consistency index with the matching degree of the target device characteristics, and the physiological response characteristics of each device are identified, which improves the accuracy of data classification. The transition data segments defined as valid are classified and labeled by the dynamic feature model of the target device, so that each data segment is strictly matched with the characteristics of the corresponding device, ensuring the independence and clarity of the data. Finally, the adaptive elimination mechanism is used to screen the classified data segments, eliminate data with large fluctuations or abnormalities, and retain high-quality data for personalized feedback generation, so that the feedback content is more in line with the user's actual training performance. The modules of the solution are closely connected, from the continuous collection of physiological data, the precise definition of the transition period, to the dynamic adjustment of classification and elimination, to build a complete closed loop for multi-device training data processing. It has greatly improved the data processing depth and feedback accuracy of the information-based sports education platform, providing users with scientific and customized training guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a structural schematic diagram of an information-based sports education platform management system of the present invention. DETAILED DESCRIPTION

[0042] 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.

[0043] Embodiment 1: Figure 1 The invention provides an information-based sports education platform management system, comprising: a fluctuation detection module, a transition determination module, a dynamic classification module and an adaptive elimination module;

[0044] Fluctuation detection module: collects physiological data monitored by the device worn by the user, identifies the data fluctuation pattern when the user switches the device, generates a transition signal, and transmits the identified transition data to the transition determination module;

[0045] Transition determination module: Based on the transition response consistency index and the matching degree of the target device characteristics, it determines whether the transition data segment meets the valid transition conditions, generates valid transition data labels, and passes the labeled data to the dynamic classification module;

[0046] Dynamic classification module: According to the target device feature model, the defined valid transition data is dynamically classified, the corresponding device data label is generated, and the classified data is passed to the adaptive rejection module;

[0047] Adaptive elimination module: Adaptively eliminates classified data based on elimination rules and outputs personalized training feedback results.

[0048] In the information-based sports education platform, in order to achieve accurate classification and analysis of multi-device training data, users wear devices to monitor physiological data (such as heart rate, respiratory rate, blood oxygen saturation, etc.) in real time. Since users frequently switch devices during multi-device training, physiological data will experience short-term transition fluctuations when switching, which may lead to data misclassification. Therefore, identifying these fluctuation characteristics and generating transition signals becomes a key link to ensure that the system can identify the data characteristics of different devices, thereby achieving accurate data classification and analysis.

[0049] The fluctuation detection module includes the following:

[0050] S1.1, real-time collection of physiological data (denoted as {X i}, where X irepresents the physiological data collected at the i-th second), and the data includes parameters such as heart rate (HR), respiratory rate (RR), blood oxygen saturation (SpO2), etc. Based on the time window segmentation strategy, the physiological data is divided into several time series segments at each time point of device switching, and the moment of device switching (when the device detects the user switching action) is marked as the starting point of the time period to capture a period of data fluctuation before and after the switching.

[0051] S1.2, for each segment of physiological data, calculate its instantaneous rate of change over time, that is, the amplitude of change. By comparing the degree of change of physiological data (such as heart rate or respiratory rate) at adjacent time points, a gradient vector of the data is formed to analyze the changing trend of physiological data when switching. Then, statistics are taken on the gradient values ​​of each data, and the degree of deviation from the standard distribution is calculated. If the amplitude of change of physiological data is large within a specific time period, it can be preliminarily judged as a transition state of the data. The formula for the gradient vector is as follows: The gradient vector reflects the rate of change of physiological data before and after the switch. For a larger rate of change, it can be considered that the user has entered a transition state. Combined with the Z-score method in statistics, the deviation degree of each parameter change is calculated to generate a set of fluctuation anomaly values ​​{Z i}, used to indicate the fluctuation range.

[0052] S1.3, based on the data change rate in the previous step, further calculate its fluctuation stability index. This index judges the stable trend of fluctuations by analyzing the difference between the change amplitudes of adjacent time points. If the change trend of physiological data tends to be gentle and the fluctuation amplitude gradually narrows, it indicates that the data stability is high, which means that it gradually returns to a stable state. This stability index will be used for subsequent comprehensive judgment to ensure the accuracy of transition state identification. The fluctuation stability index is calculated as follows: Where N represents the time length of the transition period data, |G i -G i-1 | represents the fluctuation amplitude of the data at adjacent moments. When the fluctuation stability index is lower than the set stability threshold, the data is considered to be gradually becoming stable. Through this process, the system can identify the stable and unstable stages of data fluctuations.

[0053] S1.4, finally, by performing comprehensive matching calculation on the fluctuation stability index and change deviation of each physiological data parameter, the matching function value is obtained to determine whether the data segment is a valid transition signal. match Make a comprehensive judgment, the formula is as follows: Where m is the total number of physiological data parameters (e.g., heart rate, respiratory rate, etc.), Z k For each parameter, the deviation value, SI kis the stability index of each parameter. The value generated by this matching function is used to determine whether the transition signal is established. If the matching function value exceeds the set threshold, it is considered that the corresponding moment is in a transition state, and the corresponding data segment is marked as transition data, otherwise it is marked as a stable segment.

[0054] The fluctuation detection module successfully achieves accurate identification of physiological data in the transition period by segmenting the physiological data collected in real time by the user's worn device. First, the physiological data is segmented by setting a time window to ensure that the complete data flow before and after the device switch is captured. Then, based on the data gradient change and fluctuation pattern, the instantaneous fluctuation characteristics of the physiological data are carefully analyzed to determine whether the data has entered the transition state. Subsequently, the stability trend of the data fluctuation is determined by further calculation of the fluctuation stability index to ensure that the stable and non-stationary stages of the transition data can be effectively distinguished. Finally, the transition signal is generated by matching the comprehensive volatility and the change trend, so that the system can identify reasonable and stable transition data segments, laying a solid foundation for the subsequent module's transition data definition, classification and effective label generation. This module strictly guarantees the continuity and stability of data processing. Through innovative multi-dimensional fluctuation analysis and stability judgment, it ensures the accuracy and effectiveness of the data in the transition stage to the greatest extent, providing highly accurate support for the classification of multi-device training data.

[0055] During multi-device training, due to the volatility of transition data when users switch devices, the system needs to define which data are valid transition data, so as to achieve accurate classification of data. To this end, two indicators, transition response consistency index and target device feature matching, are introduced to ensure the continuity and rationality of transition data during device switching. These two indicators determine whether the transition data segment meets the training response characteristics of the target device through precise calculation and matching analysis, and generate valid transition data labels for qualified transition data to avoid affecting the accuracy of subsequent training feedback due to unclear data.

[0056] The transition determination module includes the following:

[0057] S2.1, first, through time series analysis, feature extraction is performed on the physiological data segment after switching, and the response trend of the data after the switch is analyzed. The transition response consistency index determines the validity of the data during the transition period by detecting whether the physiological data shows a reasonable continuous response after switching. Specifically, the transition response consistency index evaluates whether the fluctuation trend of the physiological data is stable, that is, whether it conforms to the pattern of gradually transitioning to the target device characteristics within the time period.

[0058] The logic for obtaining the transition response consistency index is:

[0059] The calculation of the instantaneous change rate vector is used to obtain the change rate of the transition data over time to capture the degree of fluctuation of each physiological parameter during the switching process. The vector is obtained by comparing the difference between the transition data at the current moment and the previous moment. Specifically, for the data vector at each moment, the transition data at the current moment (including parameters such as heart rate, respiratory rate, blood oxygen saturation, etc.) is subtracted from the corresponding data at the previous moment to obtain the instantaneous change amplitude of each parameter. This vector represents the instantaneous fluctuation degree of each physiological parameter at the current moment, providing a basis for subsequent judgment of whether the data is in a transition state.

[0060] The multidimensional smoothness vector is used to evaluate whether the transition data gradually stabilizes during the switching process. The vector is obtained by comparing the difference between the instantaneous change rate at the current moment and the instantaneous change rate at the previous moment. Specifically, the instantaneous change rate vector at the current moment is calculated first, and then it is subtracted from the instantaneous change rate vector at the previous moment to obtain the fluctuation amplitude of each physiological parameter. The smaller the absolute value of the smoothness vector, the higher the degree of gradual smoothing of the data fluctuation, thus reflecting whether the data is gradually stabilizing.

[0061] The consistent cumulative offset is used to evaluate the fluctuation consistency and trend change of the transition data throughout the transition period. First, traverse each moment of the entire transition period and calculate the smoothness vector and instantaneous change rate vector of the current moment. Then, multiply the absolute value of the smoothness vector by the instantaneous change rate, and then divide it by the sequence number of the current moment plus 1, and use this value as the offset of the moment. During the entire transition period, add up the offsets of all moments one by one to obtain the consistent cumulative offset. The smaller the value of this offset, the more consistent and stable the fluctuation of the transition data is.

[0062] The transition response consistency index is used to comprehensively evaluate the smoothing trend and consistency of the data during the entire transition period. First, take the inverse of the consistency cumulative offset to ensure that the smoothly transitioned data has a higher weight in the index calculation. Then, traverse the smoothness vector and instantaneous change rate at each moment during the transition period, divide the absolute value of the smoothness vector by the instantaneous change rate plus 1, and obtain the smooth fluctuation ratio at each moment. Finally, add up the smooth fluctuation ratios of all moments and multiply them by the inverse of the consistency cumulative offset to generate the transition response consistency index. The larger the value of this index, the more stable and consistent the transition data is during the transition period, and the judgment conditions for effective transition are met.

[0063] S2.2, on the basis of obtaining the transition response consistency index, further calculate the matching degree of the switched physiological data with the characteristic model of the target device. The target device characteristic matching degree is used to measure the similarity between the characteristic changes of the transition data and the characteristic pattern of the target device. Specifically, the matching degree is compared one by one with the characteristic vector of the target device through multi-dimensional features (such as heart rate change pattern, respiratory rate adjustment trend, etc.). This matching process does not rely on simple point-by-point similarity, but through multi-dimensional trend consistency, comprehensively analyzes whether the gradual changes of multiple physiological parameters are close to the model of the target device.

[0064] The logic for obtaining the target device feature matching degree is:

[0065] At each moment during the transition period, the user's transition data is collected to form a feature vector to represent the user's physiological state at that moment. The feature vector at each moment contains multiple physiological parameters, such as heart rate, respiratory rate, and blood oxygen saturation. These physiological data are integrated into a vector for subsequent comparison with the target device features. The construction of the feature vector is repeated at each time point to ensure that the physiological state of the entire transition period is fully covered.

[0066] Based on the preset characteristics of the target device, an ideal physiological feature vector of the target device is generated. This vector defines the standard physiological response of the user when using the device, including the expected heart rate, breathing rate, and blood oxygen level. This target feature vector does not change over time and represents the physiological state of the user when using the target device under ideal conditions. This target feature vector will be used as a comparison benchmark and matched one by one with the actual transition data during the transition period.

[0067] At each moment during the transition period, the deviation between the actual transition data feature vector and the target device feature vector is calculated to generate a feature deviation vector. This vector calculates the difference value of each physiological parameter parameter by parameter to obtain the degree of difference between the actual data and the target feature. Each dimension of the deviation vector reflects the deviation of the corresponding physiological parameter, ensuring the independent judgment of each physiological feature, thereby providing a basis for the subsequent overall matching degree.

[0068] The feature deviation vectors at each moment are accumulated in sequence to generate the feature matching offset accumulation. In order to give higher weight to the later data in the accumulation process and make it have a greater impact on the matching offset, the deviation vector uses the square of the time factor as the weight when accumulating, so that the later deviation increases the impact on the accumulation. This offset accumulation is used to evaluate the overall deviation of the physiological data relative to the target feature during the entire transition period. The smaller the value, the more the data is gradually moving toward the target feature.

[0069] The target device feature matching degree is calculated based on the inverse of the cumulative feature matching offset. First, the inverse of the deviation is calculated to ensure that the data with large deviations are suppressed in the matching degree evaluation. Then, the deviations at all moments during the entire transition period are traversed and introduced one by one into the calculation of the final matching degree. Finally, the value of the target device feature matching degree is obtained by accumulation. The higher the value of the target device feature matching degree, the more the transition data conforms to the characteristics of the target device, and the physiological data that meets the requirements will be marked as valid data.

[0070] S2.3, first, the transition response consistency index and the target device feature matching degree are dimensionless and converted to the same scale for comprehensive comparison. In order to avoid the use of traditional statistical methods such as mean and standard deviation, logarithmic mapping and nonlinear normalization methods are used.

[0071] After completing the dimensionless transformation, construct the cross-state fitting explicitness, for example, using the following calculation formula: Among them, f1 and f2 are the preliminary scale coefficients. The formula is based on the concept of the unit circle. The two normalized indicators are mapped to the two-dimensional space, and the comprehensive matching degree is evaluated by calculating their distance to the origin.

[0072] Cross-state fit explicitness is used to evaluate the degree of match between physiological data and target device characteristics during the transition period. The cross-state fit explicitness value indicates the stability of physiological data and its consistency with the target device characteristics during the multi-device training switching process. The smaller the value, the closer the physiological data is to the target device characteristics during the transition period, the smaller the fluctuation, the better the matching effect, and thus more in line with the effective transition conditions; on the contrary, the larger the cross-state fit explicitness value, the greater the data fluctuation, the lower the matching degree with the target device characteristics, and does not meet the requirements of effective transition.

[0073] If the cross-state fitting significance is less than or equal to the transition threshold, it means that the physiological data of the transition data segment during the transition period is stable and highly matched with the characteristics of the target device, and meets the effective transition conditions, then the corresponding transition data segment is marked with a valid transition data label; if the cross-state fitting significance is greater than the transition threshold, it means that the physiological data fluctuates greatly and does not match the characteristics of the target device, and does not meet the effective transition conditions, so it is not suitable to be marked as valid transition data.

[0074] The transition judgment module systematically realizes the accurate judgment of the effectiveness of the user's physiological data in the multi-device training switching process by introducing the transition response consistency index and the target device feature matching degree. First, the transition response consistency index ensures the continuity and stability of the data during the transition period through multi-level judgments such as time series analysis, physiological data fluctuation trend, and smoothness evaluation, thereby confirming whether the data presents a reasonable transition response. Secondly, the target device feature matching degree carefully evaluates whether the change trend of the transition data in multiple physiological dimensions gradually approaches the target device feature model through multi-dimensional comparison of the data feature vector and the target feature at each moment, ensuring the compliance and fit of the physiological response. The two work together, relying on the comprehensive judgment of dimensionless transformation and cross-state fitting explicitness, to transform the indicators of different units into a unified scale, and realize the accurate calculation of the matching degree through the distance evaluation in two-dimensional space. In the comprehensive judgment process, the conditions for effective transition are determined by the numerical value of cross-state fitting explicitness. If the value is lower than the threshold, the data is considered to be stable and highly matched with the target device feature, thereby generating an effective transition label for the data. Combining dynamic physiological responses with target device feature matching solves the data classification problem in multi-device training switching, provides a reliable basis for subsequent personalized training feedback and accurate data labeling, and ensures the integrity and rigor of multi-device training data processing.

[0075] During the multi-device training process, the identified valid transition data needs to be further dynamically matched and classified with the target device characteristics to ensure the accurate attribution of each data segment. This classification aims to accurately mark the defined valid transition data as target device data based on the characteristic models of different devices to ensure the independence and clarity of data classification during multi-device training. To this end, based on the target device characteristic model, an innovative dynamic classification method is used to finely match the valid transition data to generate the corresponding device data label.

[0076] The dynamic classification module includes the following:

[0077] S3.1, first, establish a characteristic model of the target device. The model is constructed based on the typical physiological data characteristics of the device, including the ideal values ​​or ranges of multi-dimensional physiological parameters such as heart rate, respiratory rate, and blood oxygen saturation. By analyzing the physiological response curve of the corresponding device under the ideal training state, the characteristic vector of the target device is generated, denoted as M i This feature vector is used to define the dynamic physiological characteristics of the device and serves as a benchmark for valid transition data matching. i : The dynamic feature vector of the target device, representing the physiological parameters of the device under ideal training conditions.

[0078] S3.2, in the effective transition data segment, the physiological data (heart rate, respiratory rate, blood oxygen saturation, etc.) at each moment are integrated into an instantaneous feature vector, denoted as D i , to represent the actual physiological state of the user at that moment. The instantaneous feature vector contains multi-dimensional physiological parameter values, reflecting the specific physiological characteristics of the effective transition data segment. By performing feature vectorization processing on the data at each moment in the effective transition data segment, it is ensured that the complete physiological response is captured. i : The instantaneous feature vector at the current moment in the valid transition data segment, containing multi-dimensional physiological data.

[0079] S3.3, for each instantaneous feature vector and target device feature vector, calculate the dynamic feature deviation between the two to quantify the degree of difference between the actual data and the target device feature. By introducing the time adjustment factor t, the later valid transition data is given a higher matching weight, focusing on evaluating the convergence of the later data segment. The formula for calculating the dynamic feature deviation is as follows:

[0080] in:

[0081] Δ i : The dynamic feature deviation at the current moment i.

[0082] and They represent the physiological parameter values ​​of the instantaneous feature vector and the target device feature vector in the kth dimension at the current moment respectively.

[0083] n: The number of dimensions of physiological data (such as heart rate, respiratory rate, blood oxygen saturation, etc.).

[0084] ln(t+1): Time adjustment factor, which makes the deviation of the later data segments more weighted and gradually moves closer to the target feature model.

[0085] The formula reflects the instant deviation of the effective transition data segment by combining the deviation accumulation of multi-dimensional parameters with the time adjustment factor.

[0086] S3.4, within the effective transition data segment, determine whether the data meets the matching standard of the target device based on the dynamic feature deviation, and generate the corresponding device data label. The matching threshold is used to determine whether the feature deviation meets the requirements of the target device:

[0087] If the dynamic feature deviation is less than or equal to the matching threshold, the current valid transition data segment is marked as target device data;

[0088] If the dynamic feature deviation is greater than the matching threshold, the data segment does not meet the requirements of the target device and is not marked or temporarily stored as unclassified data.

[0089] This determination process is performed moment by moment in each valid transition data segment to ensure the accuracy of the data label. When the label is generated, key information such as the timestamp of the data segment and the target device number are recorded for subsequent analysis and call. Ultimately, all valid transition data segments that meet the conditions are marked as corresponding device data, laying the foundation for independent data classification for multi-device training.

[0090] The dynamic classification module accurately classifies valid transition data segments through the dynamic feature model of the target device, achieving deep matching of physiological data with target device features. First, by constructing the feature vector of the target device and the instantaneous feature vector of the transition data, it ensures that the multi-dimensional physiological parameters can correspond one by one. Then, the dynamic feature deviation is introduced to quantify the difference between the actual data and the target features, and the influence of the later data on the matching results is increased by the time adjustment factor, making the data classification more dynamically responsive. After setting the matching threshold, the corresponding device data label is generated for the valid transition data segment that meets the conditions, and the detailed information is recorded to ensure the accuracy and independence of the data classification. This module combines the dynamic deviation of data with the time series weight, laying a stable foundation for the clear identification and subsequent analysis of multi-device training data, making the multi-device data management more rigorous and effective.

[0091] In the processing of multi-device training data, in order to ensure the accuracy of training feedback, it is necessary to further screen the classified target device data segments to eliminate data segments that affect the analysis due to noise, short-term fluctuations or abnormal deviations. This module uses an adaptive elimination mechanism to dynamically identify and filter abnormal data based on elimination rules of data features, retaining only stable data segments that represent training characteristics, and generating personalized training feedback results for users. This process introduces a unique calculation method to adaptively process the target device data to ensure the accuracy and personalization of the feedback results.

[0092] The adaptive culling module includes the following:

[0093] S4.1, for each target device data segment that has been classified, calculate the abnormal deviation index. The calculation of the abnormal deviation index evaluates the stability and deviation of the data segment by combining the deviation accumulation and overall volatility of the multidimensional physiological parameters in the effective transition data segment; first, for each effective transition data segment, calculate the numerical difference between it and the target device characteristics dimension by dimension, and accumulate the total deviation; in the calculation of the deviation, use the time increment factor to make the later data deviation more weighted, so as to pay more attention to the stability performance of the data segment in the later period; in addition, the sum of the fluctuation changes of the physiological parameters at each moment in the effective transition data segment is added to the calculation process to evaluate the overall volatility of the data segment. This cumulative fluctuation amplitude is used as the denominator to suppress the overall volatility of the data segment, so that the abnormal deviation index can distinguish between stable data and abnormal data in the target device data. Finally, the larger the value of the abnormal deviation index, the greater the deviation of the data segment in multiple dimensions and the higher the overall volatility. Conversely, it indicates that the data segment is more stable and close to the target device characteristics, so it is suitable as a data source for personalized training feedback. The calculation formula of the abnormal deviation index is:

[0094] in:

[0095] EDI i : Abnormal deviation index, used to evaluate the compliance of the deviation of the target device data segment with the rejection conditions.

[0096] and They respectively represent the numerical difference between the data segment and the target feature in the kth dimension.

[0097] n: the number of dimensions of the physiological parameter.

[0098] t+1) 2 : Time increment factor, giving higher weight to later target device data segments.

[0099] Cumulative volatility to quantify the overall volatility of the data segment during the transition period.

[0100] The formula generates an abnormal deviation index by combining the multi-dimensional deviation accumulation with the overall volatility, thus reflecting the overall stability of each target device data segment.

[0101] S4.2, after obtaining the abnormal deviation index, compare it with the adaptive rejection benchmark to screen the target device data segments that meet the rejection criteria. The rejection judgment rules are:

[0102] If the abnormal deviation index exceeds the adaptive rejection benchmark, the corresponding target device data segment is regarded as an abnormal data segment with large volatility and is rejected.

[0103] If the abnormal deviation index is lower than or equal to the rejection benchmark, the corresponding target device data segment is retained to generate personalized feedback results.

[0104] S4.3, after eliminating invalid data segments, the retained target device data segments are used to generate personalized training feedback. The physiological parameters of the remaining data segments are aggregated and analyzed to extract the average response and fluctuation characteristics of each indicator to further evaluate the user's training effect. The personalized feedback results include the physiological response trend of each device, training adaptability analysis and dynamic adjustment suggestions. Through personalized interpretation of the trend and deviation characteristics of the user's current training data, targeted training optimization suggestions are generated to improve the user's training effect and experience.

[0105] The adaptive rejection module realizes the detection of volatility and validity judgment of the target device data segment through adaptive adjustment of the rejection rules. It uses the matching mechanism of abnormal deviation index and adaptive rejection benchmark to accurately eliminate invalid target device data segments, ensuring the high stability and representativeness of the final feedback data. The generated personalized training feedback not only provides users with detailed analysis and optimization suggestions of physiological data, but also provides solid support for the accurate management and application of multi-device training data.

[0106] 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.

[0107] 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.

[0108] 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.

[0109] 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. An information-based sports education platform management system, characterized in that: include: Fluctuation detection module, transition determination module, dynamic classification module and adaptive elimination module; Fluctuation detection module: collects physiological data monitored by the device worn by the user, identifies the data fluctuation pattern when the user switches the device, generates a transition signal, and transmits the identified transition data to the transition determination module; Transition determination module: Based on the transition response consistency index and the matching degree of the target device characteristics, it determines whether the transition data segment meets the valid transition conditions, generates valid transition data labels, and passes the labeled data to the dynamic classification module; Dynamic classification module: According to the target device feature model, the defined valid transition data is dynamically classified, the corresponding device data label is generated, and the classified data is passed to the adaptive rejection module; Adaptive elimination module: Adaptively eliminates classified data based on elimination rules and outputs personalized training feedback results.

2. The information-based sports education platform management system according to claim 1 is characterized by: The fluctuation detection module includes the following: S1.1, physiological data is collected in real time under the monitoring of the device worn by the user; based on the segmentation strategy of the time window, the physiological data is divided into several time series segments according to each time point of the device switching, and the time of device switching is marked as the starting point of the time period to capture a period of data fluctuation before and after the switching; S1.2, for each segment of physiological data, calculate its instantaneous rate of change in time, that is, the amplitude of change; by comparing the degree of change of physiological data at adjacent time points, form a gradient vector of the data, so as to analyze the change trend of physiological data when switching; then perform statistics on the gradient values ​​of each data and calculate its degree of deviation from the standard distribution; S1.3, calculate its fluctuation stability index, and judge the stable trend of fluctuation by analyzing the difference between the change amplitudes at adjacent time points; S1.

4. Finally, by performing comprehensive matching calculations on the fluctuation stability index and change deviation of each physiological data parameter, the matching function value is obtained to determine whether the corresponding data segment is a valid transition signal; if the matching function value exceeds the set threshold, it is considered that the corresponding moment is in a transition state, and the corresponding data segment is marked as transition data.

3. The information-based sports education platform management system according to claim 2 is characterized by: The transition determination module includes the following: S2.1, the logic for obtaining the transition response consistency index is: The calculation of the instantaneous change rate vector is used to obtain the change rate of the transition data in time to capture the fluctuation degree of each physiological parameter during the switching process; specifically, for the data vector at each moment, the transition data at the current moment is subtracted from the corresponding data at the previous moment to obtain the instantaneous change amplitude of each parameter; The multidimensional smoothness vector is used to evaluate whether the transition data gradually tends to be stable during the switching process; specifically, the instantaneous change rate vector at the current moment is calculated first, and then it is subtracted from the instantaneous change rate vector at the previous moment to obtain the fluctuation change amplitude of each physiological parameter; The consistent cumulative offset is used to evaluate the fluctuation consistency and trend change of the transition data during the entire transition period. First, traverse each moment of the entire transition period and calculate the smoothness vector and instantaneous change rate vector of the current moment. Then, multiply the absolute value of the smoothness vector by the instantaneous change rate, and then divide it by the sequence number of the current moment plus 1, and use this value as the offset of the corresponding moment. The transition response consistency index is used to comprehensively evaluate the smoothing trend and consistency of the data during the entire transition period. First, the reciprocal of the consistency cumulative offset is taken to ensure that the smoothly transitioned data has a higher weight in the index calculation. Then, the smoothness vector and the instantaneous change rate at each moment during the transition period are traversed, and the absolute value of the smoothness vector is divided by the instantaneous change rate plus 1 to obtain the smooth fluctuation ratio at each moment. Finally, the smooth fluctuation ratios of all moments are accumulated and multiplied by the reciprocal of the consistency cumulative offset to generate the transition response consistency index.

4. The information-based sports education platform management system according to claim 3 is characterized by: S2.2, the logic for obtaining the target device feature matching degree is: At each moment during the transition period, the user's transition data is collected to form an actual transition data feature vector, which is used to represent the user's physiological state at the corresponding moment; According to the preset characteristics of the target device, a target device feature vector is generated, which represents the physiological state of the user when using the target device under ideal conditions; it is used as a comparison benchmark and matched one by one with the actual transition data during the transition period; At each moment during the transition period, the deviation between the actual transition data feature vector and the target device feature vector is calculated to generate a feature deviation vector; the difference value of each physiological parameter is calculated parameter by parameter to obtain the degree of difference between the actual data and the target feature; The feature deviation vectors at each moment are accumulated in sequence to generate the feature matching offset accumulation. In order to give higher weight to the later data in the accumulation process and make it have a greater impact on the matching offset, the deviation vector uses the square of the time factor as the weight when accumulating, so that the later deviation increases the impact on the accumulation. Calculate the target device feature matching degree according to the inverse of the feature matching offset accumulation; First, the inverse of the deviation is calculated; then, the deviations at all moments during the entire transition period are traversed and introduced one by one into the calculation of the final matching degree; finally, the matching degree of the target device feature is obtained by accumulation.

5. The information-based sports education platform management system according to claim 4 is characterized by: S2.3, first, the transition response consistency index and the target device characteristic matching degree are dimensionless, and the two are converted to the same scale; after the dimensionless conversion is completed, the cross-state fitting index is constructed; if the cross-state fitting index is less than or equal to the transition threshold, it means that the physiological data of the transition data segment during the transition period is stable and highly matched with the target device characteristics, and meets the effective transition conditions, then the corresponding transition data segment is marked with a valid transition data label; if the cross-state fitting index is greater than the transition threshold, it indicates that the physiological data fluctuates greatly and does not match the target device characteristics, and the effective transition conditions are not met, so it is not marked.

6. The information-based sports education platform management system according to claim 5 is characterized by: The dynamic classification module includes the following: S3.1, first, establish a characteristic model of the target device based on the typical physiological data characteristics of the corresponding device; By analyzing the physiological response curve of the corresponding device under the ideal training state, the target device feature vector is generated, denoted as M i ; S3.2, within the effective transition data segment, the physiological data at each moment is integrated into an instantaneous feature vector, denoted as D i ; S3.3, for each instantaneous characteristic vector and target device characteristic vector, calculate the dynamic characteristic deviation between the two to quantify the degree of difference between the actual data and the target device characteristics; the formula for calculating the dynamic characteristic deviation is as follows: Where: Δ i : The dynamic characteristic deviation at the current moment i; and Respectively represent the physiological parameter values ​​of the instantaneous characteristic vector and the target device characteristic vector in the kth dimension at the current moment; n: the number of dimensions of physiological data; ln ( t+1 ) : time adjustment factor; S3.4, within the valid transition data segment, determine whether the data meets the matching criteria of the target device based on the dynamic feature deviation, and generate the corresponding device data label; if the dynamic feature deviation is less than or equal to the matching threshold, the current valid transition data segment is marked as the target device data; if the dynamic feature deviation is greater than the matching threshold, the corresponding data segment does not meet the target device requirements and is not marked.

7. The information-based sports education platform management system according to claim 6 is characterized by: The adaptive culling module includes the following: S4.

1. For each target device data segment that has been classified, calculate the abnormal deviation index. The calculation of the abnormal deviation index evaluates the stability and deviation of the data segment by combining the deviation accumulation and overall volatility of the multidimensional physiological parameters in the effective transition data segment. First, for each effective transition data segment, calculate the numerical difference between it and the target device characteristics dimension by dimension, and accumulate the total deviation. In the calculation of the deviation, use the time increment factor to make the later data deviation more weighted, so as to pay more attention to the stability performance of the later data segment. In addition, the sum of the fluctuation changes of the physiological parameters at each moment in the effective transition data segment is added to the calculation process to evaluate the overall volatility of the corresponding data segment. S4.2, after obtaining the abnormal deviation index, compare it with the adaptive rejection benchmark to screen the target device data segment that meets the rejection condition; the rejection judgment rule is: if the abnormal deviation index exceeds the adaptive rejection benchmark, the corresponding target device data segment is regarded as an abnormal data segment with large volatility and is rejected; If the abnormal deviation index is lower than or equal to the rejection benchmark, the corresponding target device data segment is retained to generate personalized feedback results; S4.3, after eliminating invalid data segments, the retained target device data segments are used to generate personalized training feedback.

Images