Personalized prescription generation method based on motion big data

Abstract

The invention relates to the technical field of exercise health big data processing and personalized exercise prescription generation, and discloses a personalized prescription generation method based on exercise big data. The method covers the whole process of motion data acquisition, time sequence standardization, feature extraction, grading judgment and prescription output. By collecting continuous heartbeat, displacement and time consumption information, an equal-interval heartbeat sequence is reconstructed and normalized, frequency domain transformation and integral analysis are combined, and physiological and behavioral characteristics of a sporter are accurately obtained. And on the basis of low-frequency and high-frequency energy integrals, a training intensity index is constructed in combination with an average speed, and a grading threshold value is set, so that automatic personalized exercise prescription output is realized. The whole process supports historical data archiving and trend analysis. According to the method, the problem that fine evaluation is difficult due to the fact that traditional methods depend on single indexes and subjective experience is solved, and the scientificity, pertinence and intelligent level of exercise prescriptions are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sports health big data processing and personalized sports prescription generation, in particular to a personalized prescription generation method based on sports big data. BACKGROUND

[0002] With the popularization of health management and scientific fitness concepts, sports physiological parameter collection and data analysis technology based on wearable devices have been widely used. Existing sports health big data analysis methods mainly rely on basic data such as exercise heart rate, step frequency, and calorie consumption, and through device or cloud data aggregation and simple statistics, general health suggestions and exercise interventions are provided for users. However, traditional methods mostly use single or static indicators such as average heart rate, maximum heart rate, and energy consumption as the main criteria for individual exercise load and health level, lacking in-depth analysis of dynamic characteristics such as time-varying heart rate sequence, frequency domain response, and exercise intensity and cardiopulmonary regulation coupling. Therefore, the existing technology has great limitations in truly reflecting the cardiopulmonary function state, training load changes, and individual differences in the exercise process.

[0003] Currently, the mainstream sports prescription generation technology mostly relies on preset models, expert experience, or templates trained based on historical data, lacking dynamic and personalized modeling capabilities in terms of heart rate response patterns, exercise intensity regulation, and phase trends for different users and training cycles. On the one hand, general templates cannot adapt to individual differences in physiological age, health basis, training goals, and other aspects of different exercisers, resulting in insufficient precision and applicability of exercise recommendations. On the other hand, existing analysis methods generally lack multi-scale and multi-domain dynamic big data processing algorithms, and cannot fully utilize time domain, frequency domain, and other composite features to systematically quantify exercise health status, which is prone to misjudgment of abnormal data, extreme working conditions, or data missing situations. In addition, traditional methods mostly use qualitative description or only focus on extreme points in frequency domain analysis, failing to effectively integrate the frequency spectrum structure of continuous heart rate variation with actual training load, resulting in a lack of scientific and quantifiable decision basis for exercise prescription intensity classification and training plan parameter recommendation. With the improvement of wearable device collection accuracy and the rapid growth of health data scale, existing sports health data analysis and personalized prescription generation methods need to break through the limitations of relying only on static mean or empirical rules, introduce more rigorous, continuous, and fine-grained time series modeling and frequency domain quantization mechanisms, and achieve comprehensive extraction and structured output of key physiological information such as individual training response process, heart rate variation dynamics, and training intensity and cardiopulmonary load coupling. At the same time, for complex scenarios such as multi-source heterogeneous big data, extreme or abnormal samples, and training history trends, existing technology also needs to improve the robustness and adaptive ability of the algorithm to avoid analysis failure or unscientific recommendations due to special data conditions.

[0004] To this end, the case aims to propose a personalized prescription generation method based on motion big data, which is based on the multi-source original data of heart rate change, motion displacement, time and other data in the process of motion, and automatically outputs the personalized training prescription with fine, scientific and dynamic adjustment through data structuring, standardization processing, frequency domain feature extraction, hierarchical determination and parameterization generation. SUMMARY

[0005] The application provides a personalized prescription generation method based on motion big data, which promotes the solution to the problems mentioned in the background art.

[0006] The application provides the following technical scheme: a personalized prescription generation method based on motion big data, comprising: Collecting continuous heartbeat time information and displacement and time information in the process of motion, calculating the time difference of adjacent heartbeats to form an RR interval sequence, and calculating the average speed to generate a structured data package with a unified field; Standardizing the RR interval sequence, setting a fixed sampling step to establish an equidistant time axis, reconstructing the equidistant RR estimation sequence according to the boundary and internal rules using linear interpolation, and normalizing according to the maximum and minimum values; According to the sampling step, a frequency axis is established, and the normalized RR estimation sequence is executed to perform real cosine frequency domain transformation to obtain a discrete power spectral density; Setting two frequency bands of low frequency and high frequency, performing discrete integration of the power spectrum in the corresponding frequency band to obtain low frequency integral value and high frequency integral value respectively; According to the low frequency integral value and the high frequency integral value, the frequency band ratio is calculated, and the average speed is combined to form a training intensity index; Setting two threshold values of the training intensity index, and determining the training intensity index as one of the three prescription levels according to the threshold interval; According to the prescription level, the training time, the recommended running speed, the interval distance, the interval group number and the interval rest time are determined from the preset parameter table to form a structured prescription output; Archiving the daily training intensity index, calculating the average intensity according to the stage and generating a stage trend output.

[0007] Optionally, the collecting of continuous heartbeat time information and displacement and time information in the process of motion, the calculation of the time difference of adjacent heartbeats to form an RR interval sequence, and the calculation of the average speed to generate a structured data package with a unified field, specifically comprises: A chest strap heart rate device is used to record the occurrence time of continuous heartbeats during running training; The time difference between the time of adjacent two heartbeats is obtained to form an RR interval sequence; When the number of RR interval samples is less than two, the process is terminated, and a structured prompt containing the reason and operation field is output; The positioning and timing device is used to record the total distance and total time, and the average speed is calculated, wherein the distance and time are the same training process data; The RR interval sequence, average speed, total distance and total time are uniformly named and packaged to generate a structured data packet corresponding to one training.

[0008] Optionally, the RR interval sequence is standardized, a fixed sampling step is set to establish an equidistant time axis, linear interpolation is used to reconstruct the equidistant RR estimation sequence according to the boundary and internal rules, and normalization is performed according to the maximum and minimum value, which specifically includes: The occurrence time of each heartbeat is converted into a relative time relative to the first heartbeat, and the total duration of the recording period is determined; The equidistant sampling step is set to 200 milliseconds, the equidistant time axis from the beginning to the end of the recording period is established, and the total number of sampling points is determined; When the total number of sampling points is zero, it is adjusted to one sampling point and the first sampling time is set to the last relative heartbeat time; In the case of only two heartbeats, all sampling times are directly assigned the same RR estimation value as the second heartbeat; When the number of heartbeats is not less than three and any sampling time is located in the interval defined by the adjacent two heartbeats, the RR estimation value of the sampling time is calculated according to the linear interpolation rule of the interval; When the sampling time is earlier than the relative time of the second heartbeat, the RR estimation value is set to the same as the second heartbeat; When the sampling time is located in the interval of the last two heartbeats, the RR estimation value is calculated according to the linear interpolation rule of the interval; When the sampling time is aligned with the end of the recording, the last RR interval is directly taken as the RR estimation value; The maximum and minimum values of the RR estimation sequence are obtained, and when the maximum value is greater than the minimum value, the interval linear normalization is performed, and when the two are equal, the normalization result of each sampling point is set to zero.

[0009] Optionally, the frequency axis is established according to the sampling step, and the real cosine type frequency domain transformation is performed on the normalized RR estimation sequence to obtain the discrete power spectral density, which specifically includes: The sampling step is converted to seconds; The frequency resolution is set according to the sample number and the sampling step, and a discrete frequency point sequence is generated which increases by a fixed step from zero; The real cosine type transformation is performed on the normalized RR estimation sequence to obtain the frequency domain coefficients corresponding to each discrete frequency point; The power spectral density of each discrete frequency point is calculated according to the frequency domain coefficient.

[0010] Optionally, the low frequency and high frequency bands are set, and the power spectrum in the corresponding frequency band is discretely integrated to obtain a low frequency integral value and a high frequency integral value, specifically including: The low frequency band cutoff frequency is set to 0.15 Hz, the high frequency band cutoff frequency is set to 0.40 Hz, and the Nyquist condition related to the sampling frequency is satisfied; The upper index of the two frequency bands on the discrete frequency axis is determined according to the frequency resolution, and the direct current component is excluded; When the upper index of the low frequency band satisfies the integral condition, the low frequency integral value is obtained by discretely integrating the low frequency band power spectrum, otherwise the low frequency integral value is set to zero; When the index relationship between the high frequency band and the low frequency band satisfies the integral condition, the high frequency integral value is obtained by discretely integrating the high frequency band power spectrum, otherwise the high frequency integral value is set to zero.

[0011] Optionally, the low frequency integral value and the high frequency integral value are used to calculate the frequency band ratio, and the average speed is combined to form a training intensity index, specifically including: When the high frequency integral value is greater than zero, the ratio of the low frequency integral value to the high frequency integral value is calculated; When the high frequency integral value is equal to zero and the low frequency integral value is greater than zero, the low frequency integral value is scaled according to the frequency resolution and used as the ratio; When the low frequency integral value and the high frequency integral value are both equal to zero, the ratio is recorded as one; The ratio is multiplied by the average speed to obtain the training intensity index, with a unit of meters per second.

[0012] Optionally, two threshold values of the training intensity index are set, and the training intensity index is determined as one of three prescription levels according to the threshold interval, specifically including: The two segment thresholds of the training intensity index are set to 0.8 meters per second and 1.8 meters per second; When the training intensity index is less than 0.8 meters per second, it is determined as prescription level one; When the training intensity index is between 0.8 meters per second and 1.8 meters per second, it is determined as prescription level two; When the training intensity index is not less than 1.8 meters per second, it is determined as prescription level three.

[0013] Optionally, the training duration, recommended running speed, interval distance, interval group number and group rest time are determined from the preset parameter table according to the prescription level to form a structured prescription output, specifically including: Establish a prescription parameter table: Prescription level one: training duration 30 minutes, recommended running speed 70% of average speed, no interval distance, 1 group of interval groups, and 0 seconds of group rest time; Prescription level two: training duration is 45 minutes, recommended running speed is 90% of average speed, no interval distance is set, interval group number is 1, and group interval rest time is 0 seconds; Prescription level three: training duration is 40 minutes, recommended running speed is equal to average speed, single interval distance is 400 meters, interval group number is 6, and group interval rest time is 90 seconds; The prescription level, historical index, user identifier, recommended running speed, training duration, interval distance, interval group number and group interval rest time are packaged into a structured prescription output package.

[0014] Optionally, the daily training intensity index is historically archived, the average intensity is calculated according to stages, and a stage trend output is generated, specifically including: The training intensity index is recorded according to days, and a historical intensity sequence arranged in chronological order is generated; When the historical days are not less than one day, the stage average training intensity is calculated in an equal-weight average manner; When the historical days are zero, the stage average training intensity is set as the current training intensity index; The stage trend structured result containing the stage average training intensity and the historical intensity sequence is output.

[0015] The present application has the following beneficial effects:

[0016] The heartbeat interval and the multi-dimensional original data such as movement displacement and time are synchronously collected, and are uniformly packaged into a structured data package, so that the consistency and space-time coupling of the movement physiology and behavior information are ensured. The standardized collection mechanism of the whole process not only ensures the data integrity of subsequent analysis, but also facilitates cross-device, cross-scene data comparison and processing. Compared with the existing method which only focuses on single heart rate or displacement and ignores the collection specification and time sequence correspondence, the present application improves the scientificity and usability of movement data. Especially in the actual movement scene, data anomalies and loss are easy to cause analysis distortion, and the specially designed anomaly judgment and prompt (such as automatically terminating the process and giving a prompt when the number of heartbeat samples is insufficient) in the scheme enhances the fault tolerance and user experience of the system, and effectively avoids the negative influence of non-standard data on the generation of personalized prescription.

[0017] The standardization processing around continuous heartbeat interval data adopts fixed sampling step to establish equidistant time axis, and realizes linear interpolation reconstruction and normalization of heartbeat interval based on boundary and interval rules. The innovation lies in fully considering the actual problems of irregular motion center rate sampling point time, data sparsity and deletion, realizing high-quality continuity and normalization of heart rate data through boundary condition and piecewise interpolation algorithm, and ensuring the accuracy of subsequent feature extraction. Compared with the existing method of directly using original heartbeat data or only simple average, the application can more truly reflect the dynamic change of heart rate in the exercise process, and provide a solid data basis for frequency domain analysis and intensity discrimination. In addition, through the normalization processing, the comparability and universality of data of different users, different devices and different exercise states are enhanced, and the wide application ability of the method in the big data environment is expanded.

[0018] The scheme adopts real cosine frequency domain transformation to map the normalized RR estimation sequence to the frequency domain space, calculates the discrete power spectral density, and deeply mines the frequency domain features of heart rate variation. This technical route breaks through the traditional idea of only paying attention to the time domain mean or peak value, converts the time-varying information of physiological signals into the energy distribution in the frequency domain, and can better reveal the complex mechanism of autonomic nervous regulation and heart response under exercise state. Compared with the existing traditional method of only analyzing the average heart rate or the maximum heart rate, the method has higher sensitivity and scientificity in reflecting the heart function and fatigue state, capturing the training load and recovery state, and laying a solid foundation for subsequent intensity quantitative discrimination based on frequency band energy, and makes the intelligent generation of exercise prescription closer to the individual physiological nature.

[0019] The application clearly divides the low frequency and high frequency two classic frequency bands of heart rate variation analysis, adopts discrete integral to extract the energy of each frequency band, avoids the limitation that a single index is easy to lose information. By setting the integral boundary in detail, excluding the direct current component, and dynamically adjusting the integral range according to the sampling resolution, the scientificity and comparability of the analysis results are guaranteed. Compared with the previous method of manually dividing or being easily affected by sampling anomalies during integration, the method has improved in robustness and applicability. It can accurately capture the changes of autonomic nervous activity induced by exercise, and provides more objective data support for exercise load, training recovery and health risk assessment applications.

[0020] The present application innovatively combines the ratio of low frequency and high frequency integration with the average speed, comprehensively combines the heart rate variability characteristics and the movement behavior parameters, and constructs the training intensity index. This comprehensive index can more comprehensively and dynamically reflect the training load and individual response, and avoids the one-sidedness of relying on only a single physiological signal or behavior index. Compared with the existing method of relying only on subjective scoring and single threshold determination, the training intensity index organically combines the complexity of heart rate variability and the objectivity of movement performance, enhances the scientific basis of personalized evaluation, and also improves the quantification and refinement level of individual prescription. Especially under the big data platform, this comprehensive index is convenient for batch analysis, group intervention and longitudinal trend tracking, which expands the depth and breadth of sports health management.

[0021] The present application sets double thresholds according to the training intensity index, realizes scientific grading of training load, automatically determines three movement prescription grades, and corresponds to different training needs such as recovery, aerobic and interval. This grading determination not only dynamically generates according to individual data, but also quantitatively maps combined with the scientific principles of sports physiology, overcoming the disadvantages of previous experience-based and subjective classification. By setting reasonable segmentation intervals and classification labels, the scheme makes the intervention measures for different training goals have clear data support, effectively improving the pertinence and operability of the movement prescription. At the same time, the grading system also provides a standardized basis for subsequent training plan parameterization, movement risk control and continuous tracking, promoting the intelligent, personalized and standardized development of sports intervention.

[0022] Based on the grading determination result, the present application intelligently matches and recommends specific parameters such as training duration, running speed, interval distance, group number and rest time from the preset parameter table, and automatically outputs a structured personalized movement prescription. The innovation lies in that the parameter configuration not only comes from sports physiology theory, but also corresponds to the actual training intensity of the individual, realizing the dynamic, data-based and automatic customization of the prescription. Compared with the traditional way of adjusting manually or using a fixed template for "one-size-fits-all", the present method improves the individual adaptability and scientificity of prescription recommendation, helps to avoid the risks caused by excessive or insufficient training, and comprehensively improves the actual effect of sports intervention and user satisfaction.

[0023] The present application archives the daily training intensity index, establishes a historical intensity sequence, dynamically calculates the stage average intensity and trend, and forms a visual evaluation of the stage training performance. This mechanism not only helps users and coaches to scientifically track the effect of sports intervention, but also provides data support for subsequent personalized adjustment and risk warning. Compared with the traditional method of only looking at immediate indicators and lacking historical longitudinal analysis, the present application can more systematically and dynamically reflect the changing trend of individual movement performance, providing a solid data foundation and decision support for scientific management of sports health and optimization of training path. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The flowchart of the present application is shown. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0026] Embodiments, with reference to Figure 1 A personalized prescription generation method based on sports big data, comprising: Collecting continuous heartbeat time information and displacement, time information during exercise, calculating the time difference of adjacent heartbeats to form an RR interval sequence, and calculating the average speed to generate a structured data packet with a unified field; Standardizing the RR interval sequence, setting a fixed sampling step to establish an equidistant time axis, using linear interpolation to reconstruct the equidistant RR estimation sequence according to boundary and internal rules, and normalizing according to the maximum and minimum values; According to the sampling step, a frequency axis is established, and real cosine frequency domain transformation is performed on the normalized RR estimation sequence to obtain discrete power spectral density; Setting two frequency bands of low frequency and high frequency, performing discrete integration of the power spectrum in the corresponding frequency band to obtain low frequency integral value and high frequency integral value respectively; According to the low frequency integral value and the high frequency integral value, the frequency band ratio is calculated, and the average speed is combined to form a training intensity index; Setting two threshold values of the training intensity index, and determining the training intensity index as one of three prescription grades according to the threshold interval; According to the prescription grade, the training duration, recommended running speed, interval distance, interval group number and group rest time are determined from the preset parameter table to form a structured prescription output; Archiving the daily training intensity index, calculating the average intensity according to the stage and generating a stage trend output.

[0027] It covers the whole process of data collection, time series processing, feature extraction, hierarchical judgment and structured output. By collecting the continuous heartbeat interval, displacement and time information during the movement process, the system can obtain the basic data reflecting the physiological and movement behavior of the mover. Subsequently, through standardization and interpolation steps, the heartbeat interval sequence is reconstructed into equidistant data, avoiding the influence of uneven data distribution on the accuracy of feature analysis. Using frequency domain transformation to extract the energy distribution in time series data, and then integrating the low and high frequency intervals, can better reveal the law of movement heart rate change. By setting a threshold, the training intensity is quantitatively classified, and further combined with the preset parameter table to automatically output personalized prescriptions such as training time and recommended running speed, realizing the whole-link automatic processing from data to movement intervention suggestions. This method overcomes the problem of relying on single index or subjective experience to evaluate individual differences, improves the scientificity, pertinence and operability of the movement prescription, and can realize historical data archiving and trend tracking, facilitating subsequent personalized optimization and health management.

[0028] The continuous heartbeat time information and displacement, time information during the movement process are collected, the time difference between adjacent heartbeats is calculated to form an RR interval sequence, and the average speed is calculated to generate a structured data packet with a unified field, specifically including: A chest strap heart rate device is used to record the occurrence time of continuous heartbeats during running training; The time difference between adjacent two heartbeats is obtained to get the RR interval sequence; When the number of RR interval samples is less than two, terminate this process and output a structured prompt containing the reason and operation field; The total distance and total time are recorded synchronously by positioning and timing devices, and the average speed is calculated, wherein the distance and time are both data of the same training process; The RR interval sequence, average speed, total distance and total time are uniformly named and packaged to generate a structured data packet corresponding to one training.

[0029] Further specific implementation operations include: During running training, continuous heartbeat interval data is collected through a chest strap heart rate belt; the number of continuous heartbeats recorded during running is denoted as K, and the continuous heartbeat time is denoted as ; wherein, is the time of the kth heartbeat, unit: millisecond; k is the heartbeat order index; The RR interval between adjacent heartbeats is set as: ; wherein, is the RR interval between the kth and k-1th heartbeats, unit: millisecond; The complete RR interval sequence is obtained as: ; wherein, is the RR interval sequence set; If K≥2, continue; if K<2, terminate the current prescription generation process and output a structured prompt: ; wherein, is an abnormal output set; record the total distance of the training through GPS and timing device synchronization and the total time consumption , calculate the average speed index ; wherein, in meters, in seconds; unify the naming and structured storage of all data as: ; wherein, is the structured data package of the current training.

[0030] The detailed process of data collection and structured processing is specifically defined. By using chest straps and other equipment to collect continuous heart beat intervals during exercise, and synchronously collecting the displacement and time of exercise, the consistency and integrity of all basic data are ensured. The system automatically calculates the time difference between adjacent heartbeats to obtain a sequence of continuous heart beat intervals, while the average speed of exercise is also calculated. If the number of heart beat samples is insufficient, the process is directly terminated and a structured prompt is output to avoid interference of non-standard data on the analysis results. Finally, all raw data is standardized named and uniformly packaged to form a structured data package, which is convenient for subsequent processing. This not only effectively improves the scientificity and comparability of the data, but also reduces the influence of abnormal data or collection errors on the generation of individualized prescription. Compared with the existing method which often ignores the problems of multi-source data synchronization and automatic processing of abnormal data, the method guarantees the systematicness and reliability of data processing, and lays a foundation for subsequent high-quality analysis.

[0031] The RR interval sequence is standardized, a fixed sampling step is set to establish an equidistant time axis, a linear interpolation is used to reconstruct an equidistant RR estimation sequence according to boundary and internal rules, and normalization is performed according to the maximum and minimum value, which specifically includes: Convert each heartbeat occurrence time to a relative time relative to the first jump, and determine the total duration of the recording period; Set the equidistant sampling step to 200 milliseconds, establish an equidistant time axis from the beginning to the end of the recording period, and determine the total number of sampling points; When the total number of sampling points is zero, adjust it to one sampling point and set the first sampling time to the last relative heartbeat time; In the case of only two heartbeats, all sampling times are directly assigned the same RR estimation value as the second heartbeat; When the number of heartbeats is not less than three and any sampling time is located in the interval defined by the adjacent two heartbeats, the RR estimation value of the sampling time is calculated according to the linear interpolation rule of the interval; When the sampling time is earlier than the relative time of the second heartbeat, the RR estimation value is set to be the same as the second heartbeat; When the sampling time is located in the interval of the last two heartbeats, the RR estimation value is calculated according to the linear interpolation rule of the interval; When the sampling time is aligned with the end of the record, the last RR interval is directly taken as the RR estimation value; The maximum and minimum values of the RR estimation sequence are obtained, and when the maximum value is greater than the minimum value, normalization is performed in the interval linear manner, and when the two values are equal, the normalization results of each sampling point are set to zero.

[0032] Further specific implementation operations include: Setting the relative heartbeat time: ; wherein, is the time of the kth heartbeat relative to the first heartbeat, in milliseconds; Let the total duration of the recording period be: ; wherein, is the total duration of the RR recording interval, in milliseconds; is the value of the relative time at k=K; Set the equal-interval sampling step size , and construct an equal-interval relative time axis: ; wherein, is the nth equal-interval sampling time; n is the sampling point number; is the total number of sampling points; If N=0 occurs, let N=1, ; For any sampling time , steps S201 to S205 are executed: S201, only two heartbeats: if K=2, then for all n, directly take ; wherein, is the RR estimation value corresponding to the sampling time ; S202, if K≥3 and there is a unique such that , then the linear interpolation of adjacent RR is used to calculate ; wherein, j is the left end RR index of the interpolation interval; S203, starting boundary: if , then ; wherein, is the time relative to the second heartbeat; S204, end interval interpolation: if , let j=K-1, the interpolation calculation is equivalent to ; wherein, , is the relative time of the last two heartbeats; S205, end boundary: if the integer alignment condition occurs, the interpolation directly takes ; wherein, is the last RR interval; Let ; wherein, is the maximum value of the RR estimation sequence ; is the minimum value of the RR estimation sequence ; If , normalize , ; wherein, is the normalized RR value of the nth sampling point; If , let , .

[0033] The standardization and reconstruction method of the heartbeat interval sequence is specified in detail. By converting the original heartbeat occurrence time into relative time and setting a fixed equidistant sampling step, the system can uniformly reconstruct the heart rate data sequence. In view of the insufficient data samples or special boundary conditions, the boundary conditions and interpolation rules are designed in the claim, so that all sampling points can obtain appropriate RR interval estimation values. Finally, through normalization processing, data of different ranges are converted to a unified interval, improving the comparability between different samples and the stability of subsequent analysis. This method eliminates the problems of uneven sampling and inconsistent scales of the original data through interpolation and normalization, improving the scientificity and accuracy of subsequent frequency domain analysis. Compared with the existing simple mean or ignoring abnormal point processing method, this method can more truly reflect the continuous change of exercise heart rate, making the subsequent feature extraction and personalized judgment more accurate.

[0034] The frequency axis is established according to the sampling step, and the real cosine type frequency domain transformation is performed on the normalized RR estimation sequence to obtain the discrete power spectral density, which specifically includes: Convert the sampling step to seconds; Set the frequency resolution according to the number of samples and the sampling step to generate a discrete frequency point sequence that increases by a fixed step from zero; Perform real cosine type transformation on the normalized RR estimation sequence to obtain the frequency domain coefficients corresponding to each discrete frequency point; Calculate the power spectral density of each discrete frequency point according to the frequency domain coefficients.

[0035] Further specific implementation steps include: Convert the sampling step size to seconds: ;in, The sampling step size is expressed in seconds. Set the frequency resolution to The unit is Hertz; Set the m-th frequency point as Where m is the frequency index; right Perform a real cosine transformation: , m=1,2,...,N; where, Let m be the m-th real frequency domain coefficient; Calculate the m-th discrete power spectral density value , m=1,2,...,N.

[0036] This paper proposes specific steps to obtain the discrete power spectral density from the reconstructed normalized heart rate interval sequence through a real-cosine frequency domain transform. By mapping the equally spaced sampling sequence to the frequency domain and calculating the energy distribution at each frequency point, the system can uncover the time-varying patterns and frequency domain characteristics behind the heart rate signal. This processing step can identify physiological phenomena such as heart rate fluctuations during exercise, fatigue states, and autonomic nervous system regulation, providing a scientific basis for the quantitative assessment of exercise intensity and recovery status. Unlike traditional approaches that only focus on time-domain characteristics or simple maximum / minimum value statistics, this method can comprehensively reflect the dynamic trends and complexity of heart rate changes during exercise, improving the depth and accuracy of sports and health data analysis.

[0037] The process involves setting up two frequency bands, low frequency and high frequency, and then performing discrete integration on the power spectrum within the corresponding frequency bands to obtain the low-frequency integral value and the high-frequency integral value, specifically including: Set the low-frequency cutoff frequency to 0.15 Hz and the high-frequency cutoff frequency to 0.40 Hz, while satisfying the Nyquist condition related to the sampling frequency; Determine the upper bound index of the two frequency bands on the discrete frequency axis based on the frequency resolution, and exclude the DC component; When the upper bound index of the low-frequency band satisfies the integration condition, the low-frequency band power spectrum is discretely integrated to obtain the low-frequency integral value; otherwise, the low-frequency integral value is set to zero. When the index relationship between the high-frequency band and the low-frequency band satisfies the integration condition, the high-frequency power spectrum is discretely integrated to obtain the high-frequency integral value; otherwise, the high-frequency integral value is set to zero.

[0038] Further specific implementation steps include: Set the low-frequency band cutoff frequency to Set the high-frequency band cutoff frequency to and satisfy ; The upper bound of the corresponding index is obtained as: , ; wherein, is the upper bound of the frequency point index of the low frequency band; is the upper bound of the frequency point index of the high frequency band; and the direct current component m = 1 is excluded; If , set , otherwise let ; wherein, is the low frequency band power spectrum integral value, unit: hertz; If , set , otherwise ; wherein, is the high frequency band power spectrum integral value.

[0039] The specific integral operation of extracting the energy of the low frequency and high frequency two typical physiological frequency bands based on the frequency domain feature is specified. By scientifically setting the frequency band boundary, excluding the direct current component, and adapting the integral range according to the sampling resolution, the system can accurately calculate the low frequency and high frequency energy reflecting the autonomic nervous activity. This method solves the one-sidedness of the previous direct total energy or only analyzing a single interval, and can more carefully reflect the dynamic response of the exercise center rate under different physiological regulation mechanisms, providing more powerful quantitative support for training load, recovery level and health risk analysis. Unlike the existing analysis method of subjective division or rough statistics based on experience alone, this method enhances the scientificity, robustness and applicability of frequency domain feature extraction, and improves the objective basis for subsequent personalized prescription generation.

[0040] The frequency band ratio is calculated according to the low frequency integral value and the high frequency integral value, and combined with the average speed to form the training intensity index, specifically including: When the high frequency integral value is greater than zero, the ratio of the low frequency integral value to the high frequency integral value is calculated; When the high frequency integral value is equal to zero and the low frequency integral value is greater than zero, the low frequency integral value is scaled according to the frequency resolution and used as the ratio; When the low frequency integral value and the high frequency integral value are both equal to zero, the ratio is recorded as one; The ratio is multiplied by the average speed to obtain the training intensity index, unit: meter per second.

[0041] Further specific implementation operations include: Constructing the frequency domain ratio index ; Synthesizing the training intensity index , unit: m / s.

[0042] The steps of constructing a ratio index of low frequency and high frequency integral values and combining average speed to form a training intensity index are proposed. By fusing the frequency domain ratio reflecting heart rate variation and the average speed reflecting actual sports performance, the training load of the athlete can be more comprehensively and dynamically evaluated. The method solves the problem that only using a single physiological signal or behavior index is easy to be distorted or not comprehensive, and improves the scientificity and fineness of training load evaluation. By systematically integrating physiological and behavior data, not only the individualized evaluation ability is enhanced, but also batch, trend and subgroup analysis are facilitated. Compared with the existing experience scoring or item judgment method, the present application realizes the automatic quantization of sports load, reduces subjectivity, and provides a data basis for precise sports prescription.

[0043] The two thresholds of the training intensity index are set, and the training intensity index is determined as one of the three prescription levels according to the threshold interval, specifically including: The two segmented thresholds of the training intensity index are set as 0.8 meters per second and 1.8 meters per second; When the training intensity index is less than 0.8 meters per second, it is determined as prescription level one; When the training intensity index is between 0.8 meters per second and 1.8 meters per second, it is determined as prescription level two; When the training intensity index is not less than 1.8 meters per second, it is determined as prescription level three.

[0044] Further specific implementation operations include: The segmented boundaries of the training intensity index are set as: , ; wherein, , are the prescription segmented thresholds; A classification mapping rule is established: ; wherein, A, B, and C are category labels.

[0045] The segmented thresholds and classification rules of the training intensity index are clearly specified, and the training load is quantitatively divided into three prescription levels. By formulating scientific and reasonable thresholds and classification labels, the system can automatically map individual training intensity to different intervention suggestions such as recovery, aerobic and interval. This step solves the problem of inconsistent grading standards and difficulty in implementation caused by subjective grading or reliance on expert experience in the past. Quantitative grading not only provides a clear scientific basis for sports intervention, but also facilitates the parameterization and standardized management of subsequent training plans. Compared with traditional experience-based or manual adjustment methods, the present method improves the scientificity, consistency and operability of grading, and helps to automatically generate sports prescriptions and promote batch promotion.

[0046] The training duration, recommended running speed, interval distance, interval group number and inter-group rest time are determined from the preset parameter table according to the prescription level to form a structured prescription output, which specifically includes: A prescription parameter table is established: Prescription level one: training duration 30 minutes, recommended running speed 70% of average speed, no interval distance, 1 group of interval groups, and 0 seconds of inter-group rest time; Prescription level two: training duration 45 minutes, recommended running speed 90% of average speed, no interval distance, 1 group of interval groups, and 0 seconds of inter-group rest time; Prescription level three: training duration 40 minutes, recommended running speed equal to average speed, single interval distance 400 meters, 6 groups of interval groups, and 90 seconds of inter-group rest time; The prescription level, historical index, user identifier, recommended running speed, training duration, interval distance, interval group number and inter-group rest time are packaged as a structured prescription output package.

[0047] Further specific implementation operations include: A prescription generation table is constructed: ; wherein, is the recommended training duration; is the recommended running speed; is the recommended distance of a single interval section; G is the recommended number of interval groups; is the rest time between two groups; The training plan is packaged as a structured output: ; wherein, is the prescription output package.

[0048] Based on the hierarchical results, the complete exercise prescription parameters such as training duration, recommended running speed, interval distance, group number and rest time are automatically matched and output from the preset parameter table in a structured data form. Through this automatic parameter configuration, the most suitable exercise prescription can be automatically recommended for different training levels, realizing intelligent customization and dynamic adjustment of exercise intervention. This method solves the problems of traditional methods, such as relying on manual experience, lack of individual differences and dynamic adaptability in prescription content, etc. Compared with the fixed template of one-size-fits-all, the present invention effectively improves the personalization and scientificity, reduces the risk of human intervention and operation errors, and makes the execution of exercise programs more targeted and efficient.

[0049] The daily training intensity index is historically archived, and the average intensity is calculated by stage and a stage trend output is generated, which specifically includes: The training intensity index is recorded by day and a historical intensity sequence arranged in chronological order is generated; When the history days are not less than one day, the phase average training intensity is calculated by equal weight average method; When the history days are zero, the phase average training intensity is set as the current training intensity index; The phase trend structured result containing the phase average training intensity and the history intensity sequence is output.

[0050] The further specific implementation operation includes: The training intensity index of the user on the dth day is recorded as The history days are set as ; wherein, d is the day index; is a non-negative integer; The history intensity sequence is ; If , the phase average training intensity is set as ; wherein, is the phase average training intensity; If , the phase average training intensity is set as ; The structured output is formed as ; wherein, is the phase trend output package.

[0051] The history archiving of the training intensity index and the process of the phase trend analysis are clarified, the phase average value and the history change sequence are dynamically calculated by recording the daily training intensity, and the phase trend result is output. The method not only facilitates the scientific tracking of the training effect of the user and the professional personnel, but also provides data support for the continuous optimization of the exercise scheme and the health risk monitoring. Through the longitudinal trend analysis, the change of the exercise capacity and the intervention effect can be dynamically reflected, the individualized adjustment and the intelligent risk warning are realized. Compared with the previous method of only paying attention to the instant index and ignoring the history change, the present application realizes the systematicness, the scientificness and the forward-looking of the exercise health management, and effectively improves the precision and the depth of the exercise prescription management.

[0052] It should be noted that, in the present text, the relationship terms such as first and second 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 the entities or operations. Moreover, the term "comprises", "includes" or any other variant thereof is intended to cover the non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment.

[0053] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, several improvements and refinements can be made without departing from the technical principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method for generating personalized prescriptions based on sports big data, characterized in that, include: Collect continuous heartbeat time information, displacement, and time information during the movement, calculate the time difference between adjacent heartbeats to form an RR interval sequence, and calculate the average velocity to generate a structured data packet with unified fields. The RR interval sequence is standardized, a fixed sampling step size is set to establish an equally spaced time axis, and the equally spaced RR estimation sequence is reconstructed by linear interpolation according to boundary and internal rules, and normalized according to the maximum and minimum values. A frequency axis is established based on the sampling step size. A real cosine frequency domain transform is performed on the normalized RR estimation sequence to obtain the discrete power spectral density. Two frequency bands, low frequency and high frequency, are set, and the power spectrum is discretely integrated within the corresponding frequency bands to obtain the low frequency integral value and the high frequency integral value respectively; The bandwidth ratio is calculated based on the low-frequency integral value and the high-frequency integral value, and then combined with the average speed to form the training intensity index. Set two thresholds for the training intensity index, and determine the training intensity index as one of three prescription levels according to the threshold range; Based on the prescription level, the training duration, recommended running speed, interval distance, number of interval sets, and rest time between sets are determined from the preset parameter table to form a structured prescription output; The daily training intensity index is archived historically, and the average intensity is calculated by stage, generating stage trend output.

2. The personalized prescription generation method based on sports big data according to claim 1, characterized in that, The process involves collecting continuous heartbeat time information, displacement, and time information during the movement; calculating the time difference between adjacent heartbeats to form an RR interval sequence; simultaneously calculating the average velocity; and generating a structured data packet with unified fields, specifically including: During running training, a chest strap heart rate monitor was used to record the moments of continuous heartbeats. The difference between two consecutive heartbeat times is used to obtain the RR interval sequence; The process terminates when the number of RR interval samples is less than two, and outputs a structured prompt containing the reason and operation fields. The total distance and total time were recorded synchronously using positioning and timing devices, and the average speed was calculated. The distance and time were data from the same training process. The RR interval sequence, average speed, total distance, and total time are uniformly named and encapsulated to generate a structured data packet corresponding to one training session.

3. The personalized prescription generation method based on sports big data according to claim 2, characterized in that, The standardization process for the RR interval sequence involves setting a fixed sampling step size to establish an equally spaced time axis, reconstructing the equally spaced RR estimation sequence using linear interpolation according to boundary and internal rules, and normalizing it based on the maximum and minimum values. Specifically, this includes: Convert the time of each heartbeat to the relative time of the first heartbeat, and determine the total duration of the recording period; Set the equal-interval sampling step size to 200 milliseconds, establish an equal-interval time axis from the start to the end of the recording period, and determine the total number of sampling points; When the total number of sampling points is zero, it is adjusted to one sampling point and the first sampling time is set as the last relative heartbeat time; In the case of only two heartbeats, the same RR estimate as the second heartbeat is directly assigned to all sampling times; When the heart rate is no less than three times and any sampling time is within the interval defined by two adjacent heartbeats, the estimated RR value for that sampling time is calculated according to the linear interpolation rule of that interval; When the sampling time is earlier than the relative time of the second heartbeat, the RR estimate is set to be the same as the second heartbeat. When the sampling time is within the interval between the last two heartbeats, the estimated RR value is calculated according to the linear interpolation rule of that interval; When the sampling time is aligned with the end of the record, the last RR interval is directly taken as the RR estimate. Obtain the maximum and minimum values ​​of the RR estimation sequence. When the maximum value is greater than the minimum value, normalize it in an interval linear manner. When the two values ​​are equal, set the normalization result of each sampling point to zero.

4. The personalized prescription generation method based on sports big data according to claim 3, characterized in that, The process of establishing a frequency axis based on the sampling step size and performing a real cosine frequency domain transform on the normalized RR estimation sequence to obtain the discrete power spectral density specifically includes: Convert the sampling step size to seconds; The frequency resolution is set according to the number of samples and the sampling step size to generate a discrete frequency point sequence that starts from zero and increases with a fixed step size. Perform a real cosine transform on the normalized RR estimation sequence to obtain the frequency domain coefficients corresponding to each discrete frequency point; Calculate the power spectral density at each discrete frequency point based on the frequency domain coefficients.

5. The personalized prescription generation method based on sports big data according to claim 4, characterized in that, The process involves setting up two frequency bands, low frequency and high frequency, and then performing discrete integration on the power spectrum within the corresponding frequency bands to obtain the low-frequency integral value and the high-frequency integral value, specifically including: Set the low-frequency cutoff frequency to 0.15 Hz and the high-frequency cutoff frequency to 0.40 Hz, while satisfying the Nyquist condition related to the sampling frequency; Determine the upper bound index of the two frequency bands on the discrete frequency axis based on the frequency resolution, and exclude the DC component; When the upper bound index of the low-frequency band satisfies the integration condition, the low-frequency band power spectrum is discretely integrated to obtain the low-frequency integral value; otherwise, the low-frequency integral value is set to zero. When the index relationship between the high-frequency band and the low-frequency band satisfies the integration condition, the high-frequency power spectrum is discretely integrated to obtain the high-frequency integral value; otherwise, the high-frequency integral value is set to zero.

6. The personalized prescription generation method based on sports big data according to claim 5, characterized in that, The calculation of the bandwidth ratio based on the low-frequency integral value and the high-frequency integral value, and the combination with the average velocity to form the training intensity index, specifically includes: When the high-frequency integral value is greater than zero, calculate the ratio of the low-frequency integral value to the high-frequency integral value; When the high-frequency integral value is equal to zero and the low-frequency integral value is greater than zero, the low-frequency integral value is scaled according to the frequency resolution and used as the ratio. When both the low-frequency integral value and the high-frequency integral value are equal to zero, the ratio is recorded as one. Multiplying the ratio by the average speed yields the training intensity index, measured in meters per second.

7. The personalized prescription generation method based on sports big data according to claim 6, characterized in that, The setting of two thresholds for the training intensity index, and the determination of the training intensity index into one of three prescription levels according to the threshold range, specifically includes: Set the two segment thresholds for the training intensity index to 0.8 m / s and 1.8 m / s; When the training intensity index is less than 0.8 meters per second, it is determined to be prescription level one; When the training intensity index is between 0.8 m / s and 1.8 m / s, it is determined to be prescription level two; When the training intensity index is not less than 1.8 meters per second, it is determined to be prescription level three.

8. The method for generating personalized prescriptions based on sports big data according to claim 7, characterized in that, The process of determining training duration, recommended running speed, interval distance, number of interval sets, and rest time between sets from a preset parameter table according to prescription level to form a structured prescription output specifically includes: Establish a prescription parameter table: Prescription Level 1: Training duration 30 minutes, recommended running speed is 70% of average speed, no interval distance set, 1 set of intervals, and 0 seconds of rest time between sets; Prescription Level 2: Training duration 45 minutes, recommended running speed 90% of average speed, no interval distance set, 1 set of intervals, 0 seconds of rest time between sets; Prescription Level 3: Training duration 40 minutes, recommended running speed equal to average speed, single interval distance 400 meters, number of interval sets 6, rest time between sets 90 seconds; The prescription level, historical index, user ID, recommended running speed, training duration, interval distance, number of interval sets, and rest time between sets are encapsulated into a structured prescription output package.

9. The method for generating personalized prescriptions based on sports big data according to claim 8, characterized in that, The process of archiving daily training intensity indices historically, calculating average intensity by stage, and generating stage trend output specifically includes: Record training intensity indices daily and generate historical intensity sequences arranged chronologically by date; When the number of historical days is not less than one day, the average training intensity of the stage shall be calculated using an equal-weighted average method. When the number of historical days is zero, the average training intensity of the phase is set as the current training intensity index; The output includes a stage trend structured result containing the stage average training intensity and the historical intensity sequence.

Images