Sports event full-process simulation training method and system based on digital twinning

Abstract

The invention relates to the technical field of sports training, and particularly discloses a digital twinning-based sports event full-process simulation training method and a digital twinning-based sports event full-process simulation training system, which are characterized in that physiological parameters of oxyhemoglobin saturation and respiratory frequency of athletes are monitored in real time, and fluctuation characteristic values are calculated by using Morlet wavelet transform and fast Fourier transform advanced signal processing technologies; according to the method, the physical state of the athlete and the adaptability of the athlete to the current training intensity are evaluated, and the comprehensive feature vector is constructed by the physiological indexes and is input into the pre-trained gradient boosting tree model, so that the optimal training scheme can be customized for each athlete, the training efficiency is improved, and the training cost is reduced. And the problem of overtraining or insufficient training can be effectively prevented, the training intensity, the intermittent time and the recovery strategy are dynamically adjusted by adopting an iterative optimization strategy until an optimal combination most suitable for individual requirements is found, and it is ensured that athletes reach the optimal competitive level while keeping a good health state.

Description

Technical Field

[0001] The present invention relates to the technical field of sports training, and particularly to a full-process simulation training method and system for sports events based on digital twin. Background Art

[0002] With the continuous development of sports science and the increasing improvement of competitive levels, the methods and techniques of athlete training are also constantly innovating. In order to gain an advantage in the fierce competition, athletes not only need to have excellent physical fitness and skills, but also need scientific and systematic training programs to optimize their personal performance and reduce the risk of injury. Traditional training methods usually rely on the experience of coaches and limited physiological data, such as heart rate monitoring, etc. Although these methods can guide training to a certain extent, they often lack a comprehensive understanding of the overall state of athletes and personalized adjustment. In recent years, with the development of digital twin technology and big data analysis, it has become possible to simulate training by constructing a virtual model of an athlete and combining real-time physiological data. This digital twin-based method can more accurately reflect the real state of athletes and provide a new way for formulating more scientific and reasonable training plans.

[0003] The existing technology has the following deficiencies: The body's response of each athlete to training and competition is unique. If the digital twin system fails to accurately capture some athlete-specific physiological responses (for example, rare metabolic disorders or heart problems), it may give training suggestions that are not suitable for the athlete, thereby increasing the risk of injury and even endangering life safety. Therefore, there is an urgent need for a more intelligent, flexible and accurate training method and system to meet the needs of modern sports training. Summary of the Invention

[0004] The purpose of the present invention is to provide a full-process simulation training method and system for sports events based on digital twin to solve the problems in the above background.

[0005] The purpose of the present invention can be achieved by the following technical solutions: A full-process simulation training method for sports events based on digital twin includes the following steps: S1: During the process of athlete training and competition simulation, real-time monitor the blood oxygen saturation parameter and respiratory rate parameter of the athlete; S2: According to the change amplitude of the monitored blood oxygen saturation data, calculate the blood oxygen saturation fluctuation characteristic value to evaluate the oxygen supply ability of the athlete under the current training intensity; S3: Based on the fluctuation amplitude of the measured respiratory rate, calculate the respiratory rate abnormality characteristic value to judge the body adaptability of the athlete; S4: Comprehensively analyze the blood oxygen saturation fluctuation eigenvalue and the respiration frequency anomaly eigenvalue, and based on the analysis result, determine whether the current training plan is suitable for the corresponding athlete; S5: If the analysis result shows that the current training plan is not suitable for the athlete, dynamically adjust the training intensity, rest interval, and recovery strategy, and optimize the blood oxygen saturation parameter and the respiration frequency parameter until the most suitable training plan for the corresponding athlete is found.

[0006] As a further solution of the present invention: The evaluation of the oxygen supply capacity of the athlete under the current training intensity specifically includes: During the process of simulating the athlete's training and competition, collect the athlete's blood oxygen saturation data in real time. According to the change range of the blood oxygen saturation data, calculate the blood oxygen saturation fluctuation eigenvalue, and determine whether the blood oxygen saturation fluctuation eigenvalue is greater than or equal to the preset threshold. If so, the oxygen supply capacity of the athlete under the current training intensity is abnormal; if not, the oxygen supply capacity of the athlete under the current training intensity is normal.

[0007] As a further solution of the present invention: The process of obtaining the blood oxygen saturation fluctuation eigenvalue is as follows: During the process of simulating the athlete's training and competition, collect the athlete's blood oxygen saturation data in real time according to the time series. Use the Morlet wavelet function to process the athlete's blood oxygen saturation data, apply continuous wavelet transform to the blood oxygen saturation time series to obtain the wavelet coefficients at different scales; according to the wavelet coefficients at different scales, calculate the energy spectral density at each scale; Calculate the average value of the energy spectral density at all scales. Denote all scales greater than the average value of the energy spectral density as high-frequency scales, and calculate the ratio of the sum of the energy spectral density of all high-frequency scales to the sum of the energy spectral density of all scales to obtain the blood oxygen saturation fluctuation eigenvalue.

[0008] As a further solution of the present invention: The judgment of the athlete's physical adaptability specifically includes: During the process of simulating the athlete's training and competition, collect the athlete's respiration frequency data in real time. According to the fluctuation range of the respiration frequency, calculate the respiration frequency anomaly eigenvalue, and determine whether the respiration frequency anomaly eigenvalue is greater than or equal to the preset threshold. If so, the athlete's physical adaptability is abnormal; if not, the athlete's physical adaptability is normal.

[0009] As a further solution of the present invention: The process of obtaining the respiration frequency anomaly eigenvalue is as follows: During the process of simulating the athlete's training and competition, collect the athlete's respiration frequency data in real time according to the time series. Apply the fast Fourier transform to the collected respiration frequency data to obtain its frequency domain representation, and based on the fast Fourier transform result, calculate the power spectral density on each frequency component; Power spectral density range of the preset normal breathing pattern , where represents the minimum value of the power spectral density of the preset normal breathing pattern, represents the maximum value of the power spectral density of the preset normal breathing pattern. According to the minimum value of the power spectral density of the preset normal breathing pattern, the abnormal breathing frequency eigenvalue is calculated.

[0010] As a further aspect of the present invention: The comprehensive analysis of the blood oxygen saturation fluctuation eigenvalue and the abnormal breathing frequency eigenvalue specifically includes: During the athlete training and competition simulation process, the blood oxygen saturation fluctuation eigenvalue and the abnormal breathing frequency eigenvalue of the athlete are obtained in real time. The blood oxygen saturation fluctuation eigenvalue and the abnormal breathing frequency eigenvalue are constructed into a comprehensive feature vector and used as the input of the machine learning model. Taking minimizing the error between the predicted training effect score and the actual training effect score as the training objective, the machine learning model is trained. According to the trained machine learning model, the training effect score is output. The machine learning model is a gradient boosting tree model.

[0011] As a further aspect of the present invention: The training process of the machine learning model is as follows: The obtained comprehensive feature vector is input into the gradient boosting tree model. Taking minimizing the error between the predicted training effect score and the actual training effect score as the objective function, the gradient boosting tree model is trained. During the training process, the model gradually adjusts the decision tree structure and weights through an iterative optimization method to reduce the difference between the predicted value and the true value. The mean square error is used to quantify the prediction error, and the optimal parameter set is found through the gradient descent method. The gradient boosting tree model after sufficient training can accurately output the training effect score according to the physiological data of the athlete.

[0012] As a further aspect of the present invention: The judgment of whether the current training plan is suitable for the corresponding athlete specifically includes: According to the training effect score output by the machine learning model, it is judged whether the training effect score during the athlete training and competition simulation process is greater than or equal to the preset threshold. If so, the current training plan is suitable for the corresponding athlete; if not, the current training plan is not suitable for the corresponding athlete.

[0013] As a further aspect of the present invention: The optimization of the blood oxygen saturation parameter and the breathing frequency parameter until the most suitable training plan for the corresponding athlete is found specifically includes: Based on the real-time collected blood oxygen saturation data and respiratory rate data of athletes, the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value are calculated respectively. By constructing the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value into a comprehensive feature vector and inputting it into a pre-trained gradient boosting tree model, the training effect score under the current training plan is predicted. With the goal of improving the training effect score, the training intensity, duration, and rest interval in the training plan are adjusted, and the corresponding blood oxygen saturation fluctuation characteristic value and respiratory rate abnormality characteristic value are recalculated to form a new comprehensive feature vector. In an iterative optimization manner, after each adjustment of the training plan, the gradient boosting tree model is used to re-predict the training effect score and compare it with the historical training effect score. If the predicted score of the new plan is higher, the plan is retained as the current optimal plan; otherwise, the training parameters are continuously adjusted until the best combination is found.

[0014] A full-process simulation training system for sports events based on digital twin, comprising: A data acquisition module, which, during the training and competition simulation of athletes, monitors the blood oxygen saturation parameter and the respiratory rate parameter of athletes in real time; An oxygen supply capacity evaluation module, which calculates the blood oxygen saturation fluctuation characteristic value according to the change range of the monitored blood oxygen saturation data to evaluate the oxygen supply capacity of athletes under the current training intensity; An adaptability evaluation module, which calculates the respiratory rate abnormality characteristic value according to the fluctuation range of the measured respiratory rate to judge the physical adaptability of athletes; A training plan judgment module, which comprehensively analyzes the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value, and judges whether the current training plan is suitable for the corresponding athlete according to the analysis result; A training adjustment module, if the analysis result shows that the current training plan is not suitable for the athlete, dynamically adjusts the training intensity, rest interval, and recovery strategy, and optimizes the blood oxygen saturation parameter and the respiratory rate parameter until the most suitable training plan for the corresponding athlete is found.

[0015] The beneficial effects of the present invention: (1) By integrating real-time monitoring technology with advanced signal processing algorithms such as Morlet wavelet transform and fast Fourier transform, the present invention realizes in-depth analysis of the blood oxygen saturation and respiratory frequency parameters of athletes, thereby accurately calculating the fluctuation characteristic values that reflect the physical state and training adaptability. This process can not only meticulously capture the physiological response changes of athletes under different training intensities, but also provide solid data support for the design of personalized training programs. The comprehensive feature vector constructed based on these key physiological indicators is input into a machine learning model (gradient boosting tree model). Through learning a large amount of historical data, the model can accurately predict and recommend the training strategies most suitable for each athlete's current situation, ensuring that the training plan can maximize sports performance while effectively avoiding the risks brought by overtraining or under-training. By continuously monitoring and dynamically adjusting training parameters, a closed-loop feedback system is formed, enabling coaches to optimize training arrangements in a timely manner according to the latest physiological data, and promoting athletes to reach their personal best competitive levels while maintaining good health.

[0016] (2) The present invention adopts an advanced iterative optimization strategy. Based on physiological data such as the real-time feedback of athletes' blood oxygen saturation and respiratory frequency, it dynamically adjusts training intensity, interval time, and recovery strategies until the ideal training combination that best meets individual needs is determined. By applying advanced signal processing techniques such as Morlet wavelet transform and fast Fourier transform to calculate the fluctuation characteristic values, and combining with a machine learning model (gradient boosting tree) to analyze the comprehensive feature vector, this method realizes the accurate assessment of each athlete's state and the continuous optimization of personalized training programs. This approach not only significantly improves the scientificity and flexibility of the training plan, enabling coaches to make precise adjustments in a timely manner according to the specific situation of athletes and getting rid of the limitations of traditional fixed templates; at the same time, by setting a reasonable threshold range to monitor the changes of key physiological parameters, the present invention can identify potential health problems at an early stage, thereby taking preventive measures to effectively prevent physical injuries caused by overtraining or improper training. This data-driven decision support system not only provides more scientific and personalized training guidance for athletes, but also greatly improves the quality and effect of training management, ensuring that athletes steadily improve their competitive levels while maintaining the best physical state, reflecting the deep integration of modern sports technology and health management concepts. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below with reference to the drawings.

[0018] Figure 1 is a flowchart of a full-process simulation training method for sports events based on digital twin according to the present invention; Figure 2 is a flowchart of a full-process simulation training system for sports events based on digital twin according to the present invention. Detailed implementation manners

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Please refer to Figure 1 As shown, the present invention is a full-process simulation training method for sports events based on digital twins, including the following steps: S1: During the simulation of athletes' training and competitions, real-time monitor the blood oxygen saturation parameters and respiratory rate parameters of the athletes; S2: Calculate the blood oxygen saturation fluctuation characteristic value according to the change range of the monitored blood oxygen saturation data to evaluate the oxygen supply capacity of the athlete under the current training intensity; S3: Calculate the respiratory rate abnormality characteristic value based on the fluctuation range of the measured respiratory rate to judge the physical adaptability of the athlete; S4: Conduct a comprehensive analysis of the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value, and judge whether the current training plan is suitable for the corresponding athlete according to the analysis result; S5: If the analysis result shows that the current training plan is not suitable for the athlete, dynamically adjust the training intensity, rest interval and recovery strategy, and optimize the blood oxygen saturation parameters and respiratory rate parameters until the most suitable training plan for the corresponding athlete is found.

[0021] In S1, during the simulation of athletes' training and competitions, real-time monitor the blood oxygen saturation parameters and respiratory rate parameters of the athletes, specifically including: During the athlete training and competition simulation process, the athlete's blood oxygen saturation parameters are first obtained in real time through a wearable physiological monitoring device. The device includes a plurality of optical sensors that emit light of a specific wavelength through the skin and measure the change in light intensity after being absorbed by the blood to calculate the blood oxygen saturation. The process is based on the Lambert-Beer law and determines the oxygen content in the blood by comparing the absorption difference of light of different wavelengths. Specifically, the device continuously records the blood oxygen saturation values ​​at a series of discrete time points to represent the sampling time. The collected data not only reflects the athlete's current blood oxygen level, but can also be used for subsequent analysis of the blood oxygen saturation fluctuation characteristic value to evaluate the athlete's physical load and recovery status. At the same time, in order to monitor the athlete's respiratory frequency parameters in real time, the present invention uses a chest strap non-invasive device to indirectly measure the respiratory frequency by detecting chest movement. This type of device is equipped with a flexible strain sensor that can generate corresponding electrical signal changes as the chest expands and contracts when the athlete breathes. The accurate acquisition of the above two parameters lays a solid foundation for constructing a comprehensive feature vector and optimizing the training program.

[0022] In S2, according to the variation range of the monitored blood oxygen saturation data, the blood oxygen saturation fluctuation characteristic value is calculated to evaluate the oxygen supply capacity of the athlete under the current training intensity, including: During athlete training and competition simulation, the athlete's blood oxygen saturation data is collected in real time. According to the change amplitude of the blood oxygen saturation data, the blood oxygen saturation fluctuation characteristic value is calculated to determine whether the blood oxygen saturation fluctuation characteristic value is greater than or equal to the preset threshold. If so, the athlete's oxygen supply capacity is abnormal under the current training intensity. If not, the athlete's oxygen supply capacity is normal under the current training intensity.

[0023] During the athlete training and competition simulation process, the athlete's blood oxygen saturation data is collected in real time according to the time series, and the athlete's blood oxygen saturation data is processed using the Morlet wavelet function. The continuous wavelet transform is applied to the blood oxygen saturation time series to obtain its wavelet coefficients at different scales; according to the wavelet coefficients at different scales, the energy spectrum density at each scale is calculated, and the calculation expression is: ;in, represents the number of scales, represents the position of the wavelet coefficient along the time axis, Indicates The energy spectral density of the scale, Indicates Scale position The wavelet coefficients at Represents the integral variable Small changes in Calculate the average value of the energy spectral density at all scales. Denote the scales where the energy spectral density is greater than the average value of the energy spectral density as high-frequency scales. Calculate the ratio of the sum of the energy spectral densities of all high-frequency scales to the sum of the energy spectral densities at all scales to obtain the blood oxygen saturation fluctuation eigenvalue.

[0024] It should be noted that: By performing continuous wavelet transform on the blood oxygen saturation data of athletes using the Morlet wavelet function, and evaluating the oxygen supply ability of athletes under the current training intensity by calculating the energy spectral density at different scales and analyzing its distribution characteristics, this method can not only more accurately capture the dynamic characteristics of blood oxygen saturation changing with time, but also identify the high-frequency components hidden in the signal. These high-frequency components may indicate abnormal conditions such as transient hypoxia or hyperventilation that occur when athletes are training at high intensity. By defining all scales higher than the average energy spectral density as high-frequency scales and calculating the proportion of the sum of their energies as the blood oxygen saturation fluctuation eigenvalue, the present invention provides a new way to quantitatively evaluate the oxygen supply ability and training adaptability of athletes, which helps coaches adjust the training plan in time to optimize the performance of athletes. At the same time, it has higher sensitivity and specificity and is applicable to personalized training monitoring and health management.

[0025] In S3, according to the fluctuation amplitude of the measured respiratory rate, calculate the respiratory rate abnormality eigenvalue for judging the physical adaptability of athletes, specifically including: During the process of simulating athletes' training and competitions, collect the respiratory rate data of athletes in real time. According to the fluctuation amplitude of the respiratory rate, calculate the respiratory rate abnormality eigenvalue, and judge whether the respiratory rate abnormality eigenvalue is greater than or equal to the preset threshold. If so, the physical adaptability of the athlete is abnormal; if not, the physical adaptability of the athlete is normal.

[0026] The process of obtaining the respiratory rate abnormality eigenvalue is as follows: During the process of simulating athletes' training and competitions, collect the respiratory rate data of athletes in real time according to the time series. Apply the fast Fourier transform to the collected respiratory rate data to obtain its frequency domain representation. Based on the fast Fourier transform result, calculate the power spectral density on each frequency component. The calculation expression is: ; where represents the number of frequency components, represents the th power spectral density of the frequency component, represents the th frequency domain representation of the frequency component, represents the total number of frequency components; The power spectral density range of the preset normal breathing pattern , where Represents the minimum value of the power spectral density of the preset normal breathing pattern. Represents the maximum value of the power spectral density of the preset normal breathing pattern. According to the minimum value of the power spectral density of the preset normal breathing pattern, calculate the respiratory frequency anomaly eigenvalue. The calculation expression is: ; In the formula, Represents the respiratory frequency anomaly eigenvalue. Represents the frequency component corresponding to the maximum power spectral density of the preset normal breathing pattern. Represents the frequency component corresponding to the minimum power spectral density of the preset normal breathing pattern. Represents the indicator function, which returns 1 when the condition is satisfied and 0 otherwise.

[0027] It should be noted that: By applying the fast Fourier transform to process the respiratory frequency data collected in real time by the athlete, and calculating the power spectral density on each frequency component based on the frequency domain analysis method, the abnormal fluctuation characteristics in the breathing pattern can be effectively identified. The present invention can accurately quantify the dynamic changes of the respiratory frequency, and judge the physical adaptability state of the athlete by setting the power spectral density range of the normal breathing pattern. By calculating the proportion of the frequency components exceeding the preset normal range as the respiratory frequency anomaly eigenvalue, not only can potential physical adaptability problems such as irregular breathing or dyspnea be sensitively detected, providing a scientific basis for coaches to adjust the training intensity. The method of the present invention improves the ability to capture subtle changes in the breathing pattern, has higher accuracy and reliability, and is applicable to personalized sports training monitoring and health management.

[0028] In S4, comprehensively analyze the blood oxygen saturation fluctuation eigenvalue and the respiratory frequency anomaly eigenvalue, and according to the analysis result, judge whether the current training plan is suitable for the corresponding athlete, specifically including: During the athlete's training and competition simulation, obtain the blood oxygen saturation fluctuation eigenvalue and the respiratory frequency anomaly eigenvalue of the athlete in real time, construct the blood oxygen saturation fluctuation eigenvalue and the respiratory frequency anomaly eigenvalue into a comprehensive feature vector, and use it as the input of the machine learning model. Taking minimizing the error between the predicted training effect score and the actual training effect score as the training objective, train the machine learning model, and according to the trained machine learning model, output the training effect score. The machine learning model is a gradient boosting tree model.

[0029] The training process of the machine learning model is: The obtained comprehensive feature vector is input into the gradient boosting tree model. With the objective of minimizing the error between the predicted training effect score and the actual training effect score, the gradient boosting tree model is trained. During the training process, the model gradually adjusts the decision tree structure and weights through iterative optimization to reduce the difference between the predicted value and the true value. The mean squared error is used to quantify the prediction error, and the optimal parameter set is found through the gradient descent method. After sufficient training, the gradient boosting tree model can accurately output the training effect score based on the physiological data of the athlete.

[0030] Judging whether the current training plan is suitable for the corresponding athlete specifically includes: According to the training effect score output by the machine learning model, judge whether the training effect score during the athlete's training and competition simulation process is greater than or equal to the preset threshold. If so, the current training plan is suitable for the corresponding athlete; if not, the current training plan is not suitable for the corresponding athlete.

[0031] In S5, if the analysis result shows that the current training plan is not suitable for the athlete, the training intensity, rest interval, and recovery strategy are dynamically adjusted, and the blood oxygen saturation parameter and respiratory frequency parameter are optimized until the most suitable training plan for the corresponding athlete is found. Specifically, it includes: Based on the real-time collected blood oxygen saturation data and respiratory frequency data of the athlete, the blood oxygen saturation fluctuation eigenvalue and the respiratory frequency anomaly eigenvalue are calculated respectively. These two eigenvalues reflect the physiological state changes of the athlete during the training process and are key indicators for evaluating the training load and physical adaptability. By inputting these eigenvalues into the pre-trained gradient boosting tree model, the training effect score under the current training plan is predicted. With the goal of improving the training effect score, the training intensity, duration, and rest interval in the training plan are adjusted, and the corresponding blood oxygen saturation fluctuation eigenvalue and respiratory frequency anomaly eigenvalue are recalculated to form a new comprehensive feature vector.

[0032] Adopting the iterative optimization method, after each adjustment of the training plan, the gradient boosting tree model is used to re-predict the training effect score and compare it with the historical training effect score. If the predicted score of the new plan is higher, the plan is retained as the current optimal plan; otherwise, the training parameters are continuously adjusted until the best combination is found. During this process, the safety and feasibility of the training plan can be ensured by setting the upper and lower limits of the blood oxygen saturation fluctuation eigenvalue and the threshold of the respiratory frequency anomaly eigenvalue.

[0033] Finally, a personalized training plan can be generated for each athlete, making the blood oxygen saturation fluctuation eigenvalue and the respiratory frequency anomaly eigenvalue within the optimal range, thereby maximizing the training effect score.

[0034] Please refer to Figure 2As shown in the figure, a full-process simulation training system for sports events based on digital twins includes: A data acquisition module that, during the training and competition simulation of athletes, monitors the blood oxygen saturation parameters and respiratory rate parameters of athletes in real time; An oxygen supply capacity evaluation module that calculates the blood oxygen saturation fluctuation characteristic value according to the change range of the monitored blood oxygen saturation data to evaluate the oxygen supply capacity of athletes under the current training intensity; An adaptability evaluation module that calculates the respiratory rate abnormality characteristic value according to the fluctuation range of the measured respiratory rate to judge the physical adaptability of athletes; A training plan judgment module that comprehensively analyzes the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value, and judges whether the current training plan is suitable for the corresponding athlete according to the analysis result; A training adjustment module that, if the analysis result shows that the current training plan is not suitable for the athlete, dynamically adjusts the training intensity, rest interval, and recovery strategy, and optimizes the blood oxygen saturation parameters and respiratory rate parameters until the most suitable training plan for the corresponding athlete is found.

[0035] Working principle of the present invention: Optimize the training plan by real-time monitoring and analyzing the physiological parameters of athletes to improve the training effect and personal performance. During the training and competition simulation of athletes, wearable devices and chest straps are used to obtain the blood oxygen saturation parameter and respiratory rate parameter in real time respectively. These data not only reflect the current physical state of the athletes, but also lay the foundation for subsequent analysis. The continuous wavelet transform is performed on the blood oxygen saturation data using the Morlet wavelet function, the energy spectral density at different scales is calculated, and the blood oxygen saturation fluctuation eigenvalue is determined according to the proportion higher than the average energy spectral density, so as to evaluate whether the oxygen supply capacity of the athlete is normal. For the respiratory rate, the fast Fourier transform is applied to convert the respiratory rate data into the frequency domain, the power spectral density on each frequency component is calculated, and it is judged whether the abnormal eigenvalue of the respiratory rate exceeds the preset range, so as to evaluate the physical adaptability of the athlete. The above two eigenvalues are constructed into a comprehensive feature vector and input into a pre-trained gradient boosting tree model. With the goal of minimizing the error between the prediction and the actual training effect score, the training effect score is output, and then it is judged whether the existing training plan is suitable for a specific athlete. If the existing plan is not suitable, the training intensity, interval time and recovery strategy are dynamically adjusted, and the iteration is repeated until the optimal training plan is found. This method realizes the optimization of the personalized training plan by accurately capturing the subtle changes of the athletes' physiological signals and combining machine learning technology. It not only improves the training efficiency and effect, but also effectively prevents the health risks brought by overtraining, and has important practical significance and broad application prospects. Through this systematic process, coaches can scientifically monitor and guide the training process of athletes, ensure that each athlete can participate in the competition in their best state, and maximize their potential.

[0036] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0037] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0038] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. Additionally, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context before and after.

[0039] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not indicate the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0040] The above has described in detail one embodiment of the present invention, but the content described is only the preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. A full-process simulation training method for sports events based on digital twins, characterized in that, It includes the following steps: S1: During the process of simulating athletes' training and competitions, real-time monitor the blood oxygen saturation parameters and respiratory rate parameters of athletes; S2: According to the change range of the monitored blood oxygen saturation data, calculate the blood oxygen saturation fluctuation characteristic value to evaluate the oxygen supply capacity of athletes under the current training intensity; S3: Based on the fluctuation range of the measured respiratory rate, calculate the respiratory rate abnormality characteristic value to judge the physical adaptability of athletes; S4: Conduct a comprehensive analysis of the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value, and according to the analysis results, judge whether the current training plan is suitable for the corresponding athlete; S5: If the analysis results show that the current training plan is not suitable for the athlete, dynamically adjust the training intensity, interval time and recovery strategy, and optimize the blood oxygen saturation parameters and respiratory rate parameters until the most suitable training plan for the corresponding athlete is found.

2. The method for full-process simulation training of a sports event based on digital twin according to claim 1, wherein The evaluation of the oxygen supply capacity of athletes under the current training intensity specifically includes: During the process of simulating athletes' training and competitions, real-time collect the blood oxygen saturation data of athletes. According to the change range of the blood oxygen saturation data, calculate the blood oxygen saturation fluctuation characteristic value, and judge whether the blood oxygen saturation fluctuation characteristic value is greater than or equal to the preset threshold. If so, the oxygen supply capacity of athletes under the current training intensity is abnormal; if not, the oxygen supply capacity of athletes under the current training intensity is normal.

3. A full-process simulation training method for sports events based on digital twin according to claim 2, characterized in that, The process of obtaining the blood oxygen saturation fluctuation characteristic value is as follows: During the process of simulating athletes' training and competitions, real-time collect the blood oxygen saturation data of athletes according to the time series. Use the Morlet wavelet function to process the blood oxygen saturation data of athletes, apply continuous wavelet transform to the blood oxygen saturation time series, and obtain the wavelet coefficients at different scales; according to the wavelet coefficients at different scales, calculate the energy spectral density at each scale; Calculate the average value of the energy spectral density at all scales, record the scales where all energy spectral densities are greater than the average value of the energy spectral density as high-frequency scales, and calculate the ratio of the sum of the energy spectral densities of all high-frequency scales to the sum of the energy spectral densities of all scales to obtain the blood oxygen saturation fluctuation characteristic value.

4. A full-process simulation training method for sports events based on digital twins according to claim 1, characterized in that The judgment of the physical adaptability of athletes specifically includes: During the process of simulating athletes' training and competitions, real-time collect the respiratory rate data of athletes. According to the fluctuation range of the respiratory rate, calculate the respiratory rate abnormality characteristic value, and judge whether the respiratory rate abnormality characteristic value is greater than or equal to the preset threshold. If so, the physical adaptability of the athlete is abnormal; if not, the physical adaptability of the athlete is normal.

5. A full-process simulation training method for sports events based on digital twins according to claim 2, characterized in that, The process of obtaining the respiratory rate abnormality characteristic value is as follows: During the process of simulating athletes' training and competitions, real-time collect the respiratory rate data of athletes according to the time series. Apply the fast Fourier transform to the collected respiratory rate data to obtain its frequency domain representation, and based on the fast Fourier transform result, calculate the power spectral density on each frequency component; Power spectral density range of the preset normal breathing pattern , where represents the minimum value of the power spectral density of the preset normal breathing pattern, represents the maximum value of the power spectral density of the preset normal breathing pattern, and according to the minimum value of the power spectral density of the preset normal breathing pattern, the abnormal breathing frequency eigenvalue is calculated.

6. A full-process simulation training method for sports events based on digital twins according to claim 1, characterized in that The comprehensive analysis of the blood oxygen saturation fluctuation characteristic value and the respiratory rate abnormality characteristic value specifically includes: During the process of simulating athletes' training and competitions, the characteristic values of the fluctuations in blood oxygen saturation and the abnormal characteristic values of the breathing frequency of the athletes are obtained in real time. The characteristic values of the fluctuations in blood oxygen saturation and the abnormal characteristic values of the breathing frequency are constructed into a comprehensive feature vector, which is used as the input of a machine learning model. Taking minimizing the error between the predicted training effect score and the actual training effect score as the training objective, the machine learning model is trained. According to the trained machine learning model, the training effect score is output, and the machine learning model is a gradient boosting tree model.

7. A full-process simulation training method for sports events based on digital twin according to claim 6, characterized in that, The training process of the machine learning model is as follows: The obtained comprehensive feature vector is input into the gradient boosting tree model. Taking minimizing the error between the predicted training effect score and the actual training effect score as the objective function, the gradient boosting tree model is trained. During the training process, the model gradually adjusts the decision tree structure and weights through an iterative optimization method to reduce the difference between the predicted value and the true value. The mean square error is used to quantify the prediction error, and the optimal parameter set is found through the gradient descent method. After sufficient training, the gradient boosting tree model can accurately output the training effect score according to the physiological data of the athletes.

8. A full-process simulation training method for sports events based on digital twins according to claim 1, characterized in that, The judgment of whether the current training plan is suitable for the corresponding athlete specifically includes: According to the training effect score output by the machine learning model, it is judged whether the training effect score during the process of simulating the athletes' training and competitions is greater than or equal to the preset threshold. If so, the current training plan is suitable for the corresponding athlete; if not, the current training plan is not suitable for the corresponding athlete.

9. A full-process simulation training method for sports events based on digital twins according to claim 1, characterized in that The optimization of the blood oxygen saturation parameter and the breathing frequency parameter until the most suitable training plan for the corresponding athlete is found specifically includes: According to the blood oxygen saturation data and breathing frequency data collected in real time of the athletes, the characteristic values of the fluctuations in blood oxygen saturation and the abnormal characteristic values of the breathing frequency are calculated respectively. By constructing the characteristic values of the fluctuations in blood oxygen saturation and the abnormal characteristic values of the breathing frequency into a comprehensive feature vector and inputting it into the pre-trained gradient boosting tree model, the training effect score under the current training plan is predicted. Taking improving the training effect score as the goal, the training intensity, duration, and rest interval in the training plan are adjusted, and the corresponding characteristic values of the fluctuations in blood oxygen saturation and the abnormal characteristic values of the breathing frequency are recalculated to form a new comprehensive feature vector. In an iterative optimization manner, after each adjustment of the training plan, the gradient boosting tree model is used to re-predict the training effect score and compare it with the historical training effect score. If the predicted score of the new plan is higher, the plan is retained as the current optimal plan; otherwise, the training parameters are continuously adjusted until the best combination is found.

10. A full-process simulation training system for sports events based on digital twin, which is used for a full-process simulation training method for sports events based on digital twin according to any one of claims 1-9, characterized in that, It includes: A data acquisition module that, during the process of simulating the athletes' training and competitions, monitors the blood oxygen saturation parameter and the breathing frequency parameter of the athletes in real time; An oxygen supply capacity evaluation module that calculates the characteristic values of the fluctuations in blood oxygen saturation according to the change range of the monitored blood oxygen saturation data to evaluate the oxygen supply capacity of the athletes under the current training intensity; An adaptability evaluation module, which calculates an abnormal respiratory rate eigenvalue based on the fluctuation amplitude of the measured respiratory rate to determine the physical adaptability of the athlete; A training plan judgment module, which comprehensively analyzes the blood oxygen saturation fluctuation eigenvalue and the abnormal respiratory rate eigenvalue, and determines whether the current training plan is suitable for the corresponding athlete according to the analysis results; A training adjustment module, if the analysis results show that the current training plan is not suitable for the athlete, dynamically adjusts the training intensity, rest interval, and recovery strategy, and optimizes the blood oxygen saturation parameter and the respiratory rate parameter until the most suitable training plan for the corresponding athlete is found.

Images