Intelligent management system and method for physical training

Abstract

The invention relates to the technical field of sports management, and discloses an intelligent management system and method for sports training, and the method comprises the steps: collecting the physiological, kinematic and subjective feedback data of an athlete at a training moment in real time, and carrying out the preprocessing, and generating multi-modal data; performing fatigue analysis at the training moment of the athlete, performing cumulative load analysis in combination with heart rate and acceleration data of the athlete, and calculating a comprehensive load index of the athlete in a training state; calculating a mean value and a standard deviation of historical comprehensive load indexes by combining historical training data of the athletes to set a judgment standard, judging whether the athletes maintain or enhance or reduce the current training intensity, and outputting a control instruction for training equipment to realize dynamic adjustment of the training intensity of the athletes; the system can comprehensively sense the training states of the athletes and accurately quantify the fatigue and load states of the athletes, so that the scientificity of training state evaluation of the athletes and the dynamic adjustment training intensity are improved.

Description

Technical Field

[0001] This invention relates to the field of sports management technology, specifically to an intelligent management system and method for sports training. Background Technology

[0002] Most existing sports training management systems focus on the collection of training data (such as heart rate, speed, and trajectory) and post-training analysis. Coaches and athletes need to rely on experience to judge the athlete's condition and adjust subsequent training plans by reviewing data reports or charts. This approach has the following drawbacks: 1. Lag; adjustment decisions are based on historical data, making it impossible to make real-time, adaptive interventions for the current training session; 2. One-sidedness; it mostly relies on single-dimensional physiological or kinematic data and lacks a comprehensive assessment of the athlete's overall condition, such as fatigue and mental state; 3. Experience dependence; the adjustment of training plans relies heavily on the coach's personal experience, lacks objective and quantitative decision support, and is difficult to standardize and scale. 4. Ignoring individual differences; existing systems typically use generic thresholds, such as percentage of maximum heart rate, and fail to personalize them based on the athlete’s real-time condition and individual recovery ability. Summary of the Invention

[0003] To address the aforementioned shortcomings in the existing technology, this invention provides an intelligent management system and method for sports training, which solves the problem that existing sports training management systems rely on human experience, resulting in the inability to assess and adjust athletes' training status in real time.

[0004] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A smart management system for sports training includes: The multimodal data acquisition module is used to collect athletes' physiological data, kinematic data, and subjective feedback data in real time during training. After data preprocessing, multimodal data is generated. The data fusion and state assessment module is used to receive multimodal data, analyze the fatigue of athletes during training, and combine the athletes' current heart rate and acceleration data to perform cumulative load analysis and calculate the comprehensive load index of athletes in training state. The dynamic adjustment module receives the comprehensive load index, combines it with the athlete's historical training data to obtain the statistical value of the historical comprehensive load index, compares the statistical value with the athlete's comprehensive load index at the current training moment, and outputs control commands to the training equipment in real time based on the comparison results, so as to realize the dynamic adjustment of the athlete's training intensity.

[0005] A method for applying a smart management system for sports training includes the following steps: S1. Real-time collection of athletes' physiological data, kinematic data, and subjective feedback data during training, data preprocessing, and generation of multimodal data; S2. Based on multimodal data, analyze the fatigue of athletes during training, and combine the athletes' current heart rate and acceleration data to analyze the cumulative load and calculate the comprehensive load index of athletes in training. S3. Based on the comprehensive load index and combined with the athlete's historical training data, calculate the historical comprehensive load index mean and historical comprehensive load index standard deviation to set judgment criteria, determine whether the athlete should maintain, increase or decrease the current training intensity, and output control commands to the training equipment to realize dynamic adjustment of the athlete's training intensity.

[0006] The present invention has the following beneficial effects: 1. The intelligent management system and method for sports training proposed in this invention achieves comprehensive perception of athletes' training status through multimodal data fusion; that is, by collecting physiological, kinematic, and subjective feedback data and performing normalization, variance and other preprocessing, a multi-dimensional and standardized training data system is constructed, which breaks through the limitations of single data dimension in traditional sports training methods or systems, and provides a comprehensive and comparable multimodal data foundation for athletes' training status. 2. This invention also precisely quantifies athletes' fatigue and load status, improving the scientific rigor of athlete training status assessment. Specifically, by integrating weights based on cardiovascular, muscular, and subjective feedback dimensions, a real-time fatigue index is constructed, achieving multi-dimensional and personalized quantification of athletes' fatigue status and avoiding the one-sidedness of single-indicator assessment. Simultaneously, by combining heart rate and acceleration data, a comprehensive load index of physiological and biomechanical load is calculated, accurately quantifying cumulative training load and providing an objective and dynamic quantitative basis for judging training intensity. 3. This invention can also dynamically adjust the training intensity of athletes for personalized and scientific training. Specifically, it utilizes statistical data from historical training, including the mean and standard deviation, to set personalized load intervals. By comparing the current comprehensive load index in real time, it automatically decides whether to maintain, reduce, or increase the athlete's training intensity, achieving dynamic optimization of training intensity. Simultaneously, it effectively balances training stimulation and recovery, avoiding the risk of injury from overtraining while preventing undertraining from hindering performance improvement. Ultimately, it helps improve athletes' training efficiency and athletic performance, promoting the development of sports training towards intelligence and personalization. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the structure of an intelligent management system for sports training proposed in this invention; Figure 2 This is a schematic diagram of the structure of the multimodal data acquisition module in the embodiment; Figure 3 This is a flowchart illustrating a smart management method for sports training proposed in this invention. Detailed Implementation

[0008] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0009] like Figure 1 As shown, a smart management system for sports training includes: The multimodal data acquisition module is used to collect athletes' physiological data, kinematic data, and subjective feedback data in real time during training. After data preprocessing, multimodal data is generated.

[0010] In this embodiment, the multimodal data acquisition module integrates various data such as the athlete's physiological state, athletic performance, and subjective feelings during training through real-time acquisition of multi-dimensional data. This constructs a comprehensive multimodal training data system, providing a rich data foundation for subsequent training status analysis, fatigue monitoring, and athletic performance evaluation. Simultaneously, through preprocessing operations such as normalization, the module eliminates dimensional differences between different types of data, making multi-source data comparable and analyzable, thus improving the accuracy and efficiency of subsequent data modeling and analysis. Furthermore, the introduced real-time acquisition and processing mechanism can provide timely feedback on the athlete's training status, offering real-time decision-making support for the coaching team or intelligent training system, and facilitating the dynamic adjustment and optimization of training programs.

[0011] The structure and connection relationship of the multimodal data acquisition module are as follows: Figure 2 As shown, it includes a physiological data acquisition module, a kinematic data acquisition module, a subjective feedback data acquisition module, and a data preprocessing module.

[0012] The physiological data acquisition module is used to collect physiological data of athletes during training, including heart rate, heart rate variability, and surface electromyography signals.

[0013] In this embodiment, a smart wearable device, such as a smart bracelet, is used to collect the athlete's heart rate in real time during training, so as to intuitively reflect the athlete's exercise intensity and cardiovascular system load. Based on continuous heart rate data, the degree of variability of adjacent heartbeat intervals (RR interval) is calculated to obtain the heart rate variability (HRV index), so as to effectively assess the athlete's autonomic nervous system balance, fatigue level and recovery status. At the same time, surface electromyography (SEMG) electrodes are attached to the athlete's muscle skin surface to detect muscle contraction electrical signals. After amplification and filtering, SEMG signal data is obtained, thereby quantifying the degree of muscle activation, force pattern and fatigue status, and helping to analyze the rationality of muscle participation in the movement. Therefore, detecting the athlete's physiological data during training can accurately monitor the athlete's physiological state, muscle activity and fatigue level, and provide data support for scientifically adjusting training plans, preventing injuries and developing personalized solutions.

[0014] The kinematic data acquisition module is used to collect kinematic data of athletes during training, including speed and acceleration.

[0015] In this embodiment, the module directly acquires high-frequency triaxial acceleration data through an inertial measurement unit integrated into a wearable device (such as a smart bracelet or a professional motion sensor). Movement speed is acquired in different ways depending on the training scenario: in outdoor environments, it is primarily measured through the device's built-in global navigation satellite system receiver; in indoor environments, it is preferentially acquired through connected training equipment (such as a treadmill or stationary bike), or by processing acceleration data and combining it with information such as stride frequency for estimation. Therefore, through kinematic data acquisition, the athlete's exercise intensity, movement rhythm, and explosive power are quantified, providing objective data for evaluating the athlete's performance during training, optimizing technical movements, and adjusting training load, thus helping to improve training efficiency and athletic performance.

[0016] The subjective feedback data acquisition module is used to collect athletes' subjective feedback data after training, including fatigue level scores.

[0017] In this embodiment, the Borg Rating of Perceived Exertion is used, and the rating scale uses a 0-10 scale, where 0 represents no perceived fatigue and 1 represents the highest perceived fatigue. This is to investigate the fatigue level of athletes after training and to collect fatigue level scores at different training times. This is to supplement objective data that cannot cover individual feelings, accurately capture the physical and mental fatigue state of athletes, and provide subjective basis for adjusting training intensity, ensuring recovery, and customizing personalized plans, thereby avoiding overtraining.

[0018] The data preprocessing module receives physiological data, kinematic data, and subjective feedback data, performs data preprocessing, and generates multimodal data, specifically: The heart rate, heart rate variability, and fatigue score were normalized to generate normalized heart rate, normalized heart rate variability, and normalized fatigue score, i.e.:

[0019]

[0020]

[0021] in, Indicates that the athlete is during training. Heart rate below This represents an athlete's individual resting heart rate. Indicates that the athlete is during training. Normalized heart rate Indicates that the athlete is during training. Heart rate variability under the following conditions This represents the individual baseline heart rate variability of an athlete, and is the mean of the heart rate variability. Indicates that the athlete is during training. Normalized heart rate variability under normalization, Indicates that the athlete is during training. The fatigue level rating below, This represents the maximum value on the athlete's individual fatigue level score scale. Indicates that the athlete is during training. Normalized fatigue level.

[0022] Obtain the time window for athletes during training sessions. The surface electromyography (EMG) signals of the target muscle groups were analyzed, and their variance was calculated to obtain the surface EMG signals of the athlete's target muscle groups within a time window. variance within .

[0023] In this embodiment, the heart rate, heart rate variability and fatigue score data are normalized to eliminate the dimensional differences between these three different types of data, making them comparable and analyzable, thereby improving the accuracy and efficiency of subsequent data modeling and analysis. The absolute value of surface electromyography (EMG) signals is affected by factors such as electrode placement, muscle fiber orientation, and device sensitivity, limiting the direct comparative significance between individuals. Therefore, in practical analysis, the focus is often on the trend of EMG changes during the individual's own movements (such as peak EMG values ​​and integral EMG within the movement cycle), or by extracting time-domain / frequency-domain features (such as root mean square and spectral entropy) to transform them into comparable features, rather than directly normalizing the original signal. Therefore, this invention considers the time window when subsequently introducing surface EMG signals for real-time fatigue index calculation. Within a continuous training time interval, the variance of surface electromyography (EMG) signals is calculated to achieve dynamic, real-time monitoring of muscle state. This ensures that the variance index reflects the characteristics of muscle electrical activity at a specific stage of training, rather than isolated instantaneous values. Therefore, this variance accurately reflects the degree of fluctuation in muscle electrical activity. When muscles are fatigued, the coordination of muscle fiber activation decreases, signal fluctuations increase, and the variance rises accordingly. Through this variance index, the fatigue state and stability of muscle activation of the target muscle group can be objectively quantified, compensating for the limitations of subjective feedback and single physiological indicators. Furthermore, in the fatigue index formula, this index serves as a core parameter of the muscle dimension, in conjunction with the cardiovascular dimension. Subjective feedback dimension This combination enables a comprehensive assessment of athletes' fatigue status from multiple dimensions, including physiological, muscular, and subjective feelings, making the fatigue index more comprehensive and accurate.

[0024] The data fusion and state assessment module receives multimodal data, analyzes the athlete's fatigue level during training, and combines this with the athlete's current heart rate and acceleration data to perform cumulative load analysis and calculate the athlete's comprehensive load index during training. Specifically: Based on the athlete's normalized heart rate, normalized heart rate variability, normalized fatigue level, and surface electromyography signals, a comprehensive fatigue analysis was performed considering cardiovascular, muscular, and subjective feedback dimensions. Through weighted allocation, a real-time fatigue index was calculated, namely:

[0025] in, Indicates that the athlete is during training. Real-time fatigue index under the current conditions For cardiovascular dimensions indicators, This indicates the electromyographic signals on the surface of the athlete's target muscle group within a time window. The variance within the range is an indicator of muscle dimension. For subjective feedback dimension indicators, , , These represent the weighting coefficients for the cardiovascular dimension, muscle dimension, and subjective feedback dimension, respectively.

[0026] In this embodiment, by integrating the cardiovascular dimension (the ratio of normalized heart rate to normalized heart rate variability, reflecting the load and regulatory capacity of the cardiovascular system), the muscle dimension (surface electromyography signal variance, quantifying the stability of muscle activation and fatigue level), and the subjective feedback dimension (normalized fatigue level, reflecting the athlete's subjective fatigue experience), fatigue status is comprehensively assessed from three levels: objective physiological indicators, objective muscle indicators, and subjective feelings, avoiding the one-sidedness of single-dimensional assessment; simultaneously, during training... Real-time fatigue index calculation allows coaches or training systems to instantly monitor athletes' fatigue progress based on real-time data, providing timely evidence for dynamic adjustments to training intensity and optimization of training rhythm, effectively avoiding overtraining or undertraining. Secondly, the indicators for the cardiovascular and subjective feedback dimensions are normalized based on the athlete's individual baseline values ​​(resting heart rate, HRV baseline, and maximum value of the fatigue rating scale), reflecting personalization and making the assessment results more consistent with the athlete's own physiological and subjective characteristics. Furthermore, the surface electromyography signal variance of the muscle dimension is an objective result of physiological signal analysis, combining objectivity to ensure that fatigue assessment is supported by solid physiological data. In addition, the weighting coefficients for the cardiovascular dimension, muscle dimension, and subjective feedback dimension, i.e. , , This is achieved through personalized calibration using machine learning models for different athletes and different sports, specifically: First, collect the athletes' historical training data, including indicators across various dimensions, namely... , , , The data includes the corresponding actual fatigue label, namely the post-match recovery time; among which, the post-match recovery time is an objective and valuable continuous variable label, indicating that the longer the recovery time, the deeper the fatigue level. Secondly, historical training data (i.e., multi-dimensional indicators) is used as input, and real fatigue labels are used as output to train the machine learning model. The model fits the data to find the strength of influence (i.e., contribution) of each indicator on the fatigue outcome. The choice of model is not limited; models capable of handling non-linear relationships between features, such as random forest regression models, can be used as training models to learn the contribution weights of each indicator to the degree of fatigue. After model training, the contribution weights of the three dimensions of indicators to the fatigue outcome are extracted through the model's feature importance mechanism, such as the feature importance score of random forest. These weights are used as weight coefficients for the cardiovascular, muscle, and subjective feedback dimensions, and must be normalized to ensure accuracy. Ultimately, we obtain the trained model and the weights of the athlete's cardiovascular, muscle, and subjective feedback dimensions in that sport. The purpose of the above-mentioned personalized weight adaptation is: (1) the weight differences of different sports; endurance sports (such as marathons) are more sensitive to cardiovascular fatigue, therefore The weighting is too high; strength sports (such as weightlifting) are more sensitive to muscle fatigue, therefore The weighting is too high; projects that require subjective perception (such as ball games) therefore... The weight can be appropriately increased, and (2) Weight differences among different athletes; Combine the athletes' physiological characteristics (such as resting heart rate, muscle fiber type) and training history (such as injury history, fatigue recovery cycle) to fine-tune the weights of the model output to ensure that they are adapted to individual differences. Finally, as athlete training data continues to accumulate, the model needs to be retrained regularly (e.g., every two weeks) with new data to achieve dynamic optimization of weight coefficients and ensure the timeliness and accuracy of the evaluation.

[0027] Based on the real-time fatigue index and the athlete's heart rate during training, the accumulated physiological load of the athlete during training is calculated. Based on the athlete's acceleration during training, the accumulated biomechanical load of the athlete during training is calculated. Weights are allocated to the physiological load and biomechanical load to calculate the athlete's comprehensive load index during training, i.e.:

[0028]

[0029]

[0030] in, Indicates that the athlete is during training. The comprehensive load index under, , These represent the weighting coefficients for balancing physiological load and biomechanical load, respectively. Indicates that the athlete is during training. The physiological load under the condition, This indicates the start time of the athlete's training. Indicates that the athlete is during training. Real-time fatigue index under the current conditions Indicates that the athlete is during training. Heart rate below Indicates that the athlete is during training. Under biomechanical load, Indicates that the athlete is during training. The acceleration under the current.

[0031] In this embodiment, physiological load To measure the real-time fatigue index With heart rate Integrating the product over time precisely quantifies the cumulative load on physiological systems, such as cardiovascular and muscle fatigue, reflecting the superimposed effect of fatigue state and heart rate activation; biomechanical load. , in order to obtain the square of the acceleration By integrating data, the cumulative biomechanical load on the body from exercise impact and movement intensity is scientifically simulated, such as the ground impact during running and the acceleration changes during jumping; therefore, the comprehensive load index... Through weighting coefficients , It integrates physiological and biomechanical loads, taking into account the differences in characteristics of different sports, such as endurance sports focusing on physiology and weight sports focusing on biomechanics, and also achieves unified quantification of multi-dimensional loads. In addition, the weighting coefficients for balancing physiological load and biomechanical load, i.e. , Based on the varying degrees of dependence on physiological and biomechanical loads in different sports, the following settings were implemented: For endurance events, such as marathons and long-distance swimming: the physiological load is higher, at which point... , ; For strength and explosive power sports, such as weightlifting and sprinting, biomechanical load (impact of movement, changes in acceleration) is more critical. , ; For comprehensive sports such as soccer and basketball, both physiological and biomechanical loads are important. , .

[0032] Therefore, calculating the comprehensive load index of athletes during training not only captures the cumulative physiological fatigue (such as the superposition of prolonged high heart rate combined with high fatigue) and biomechanical impact (such as the repeated impact of high-intensity movements) during training, but also avoids the one-sidedness of single-dimensional assessment; at the same time, it also characterizes the process accumulation of load, i.e., integral operation, from the start time of training. up to the current moment This transforms instantaneous load into cumulative process load, accurately reflecting the impact of training duration on load; and the comprehensive load index It can be directly used to dynamically adjust training intensity, such as reducing intensity when the load is too high, and to predict recovery status, such as high... The recovery time needs to be extended and the risk of injury needs to be assessed, such as the risk of joint damage caused by excessive biomechanical load over a long period of time. This provides the core quantitative basis for subsequent intelligent management functions. In summary, this step, through multi-dimensional integration, process accumulation and personalized adaptation, has built a scientific and comprehensive training load assessment system, which provides key support for subsequent functions such as dynamic adjustment, recovery prediction and plan generation.

[0033] The dynamic adjustment module receives the comprehensive load index, combines it with the athlete's historical training data to obtain a statistical measure of the historical comprehensive load index, compares this statistical measure with the athlete's comprehensive load index at the current training moment, and outputs control commands to the training equipment in real time based on the comparison results. This enables dynamic adjustment of the athlete's training intensity. Specifically: Based on athletes' historical training data, the statistics of the historical comprehensive load index are calculated, including the historical comprehensive load index mean and the historical comprehensive load index standard deviation. The calculation formula is as follows:

[0034]

[0035] in, This represents the historical average comprehensive load index. Indicates the number of training sessions. Indicates the athlete in the The overall load index under this training session This represents the standard deviation of the historical composite load index.

[0036] In this embodiment, the mean and standard deviation of historical comprehensive load are calculated to construct a normal load fluctuation range based on the athlete's individual historical data. This fully utilizes the personalized characteristics of individual data, that is, the statistics are calculated based on the athlete's own historical data, rather than a general threshold, which fully adapts to the differences in individual load tolerance and recovery ability. For example, athletes with strong physical fitness have high historical load mean and large standard deviation, and their normal fluctuation range will naturally adapt to their ability; athletes with weaker physical fitness have a more conservative range, avoiding the one-sidedness of general thresholds.

[0037] Perform the first judgment to determine when the athlete is training. The comprehensive load index below Is it within the standard range? If yes, then the athlete's current training load is within the normal fluctuation range of their individual ability, and the current training intensity should be maintained; otherwise, the second judgment should be executed.

[0038] Perform the second judgment to determine when the athlete is training. The comprehensive load index below Is it greater than If so, the athlete's current training load is significantly higher than the upper limit of its normal fluctuation range, posing a risk of overtraining and injury. An overload command is sent to the training equipment to reduce the athlete's training intensity. Otherwise, the third judgment is executed.

[0039] Perform the third judgment to determine when the athlete is training. The comprehensive load index below Is it less than If so, the athlete's current training load is below the lower limit of its normal fluctuation range, indicating that the athlete is in a recovery period or the training intensity is insufficient. A low-load instruction is sent to the training equipment to increase the athlete's training intensity.

[0040] In this embodiment, a three-stage judgment is adopted: from normal range to overload to low load, clearly covering all possible load states and ensuring no omissions in decision-making. Simultaneously, instructions to maintain, reduce, or increase intensity are output according to different load states, perfectly meeting the needs of dynamically adjusting training intensity. When the intensity is reduced, timely triggering of the reduction instruction can effectively avoid the risk of injury caused by overtraining, such as muscle strain and cardiovascular overload; when At the same time, it triggers an increased intensity command to avoid slow improvement in physical fitness due to insufficient training stimulation, ensuring training efficiency; secondly, historical data is dynamically updated, meaning the number of training sessions... As training progresses and the mean and standard deviation increase, they iteratively optimize as the athlete's physical fitness improves and the training stage changes, ensuring that the normal fluctuation range always matches the athlete's current true ability and achieving long-term dynamic adaptation. In summary, this step, through personalized statistical benchmarks combined with hierarchical decision-making logic, not only ensures the accuracy and safety of training adjustments but also possesses dynamic adaptability and interpretability, making it an efficient and practical dynamic adjustment scheme.

[0041] like Figure 3 As shown, a method for applying a smart management system for sports training includes the following steps: S1. Real-time collection of athletes' physiological data, kinematic data, and subjective feedback data during training, followed by data preprocessing to generate multimodal data.

[0042] S2. Based on multimodal data, analyze the fatigue of athletes during training, and combine the athletes' current heart rate and acceleration data to analyze the cumulative load and calculate the comprehensive load index of athletes in training.

[0043] S3. Based on the comprehensive load index and combined with the athlete's historical training data, calculate the historical comprehensive load index mean and historical comprehensive load index standard deviation to set judgment criteria, determine whether the athlete should maintain, increase or decrease the current training intensity, and output control commands to the training equipment to realize dynamic adjustment of the athlete's training intensity.

[0044] In summary, the intelligent management system and method for sports training proposed in this invention firstly achieves a comprehensive perception of athletes' training status through multimodal data fusion. Specifically, by collecting physiological, kinematic, and subjective feedback data and performing preprocessing such as normalization and variance analysis, a multi-dimensional and standardized training data system is constructed, overcoming the limitations of single data dimensions in traditional sports training methods or systems. This provides a comprehensive and comparable multimodal data foundation for assessing athletes' training status. Secondly, it accurately quantifies athletes' fatigue and load status, improving the scientific rigor of athlete training status assessment. Specifically, by constructing a real-time fatigue index based on weighted fusion of cardiovascular, muscular, and subjective feedback dimensions, a multi-dimensional and personalized quantification of athletes' fatigue status is achieved, avoiding the one-sidedness of single-indicator assessments. Simultaneously, by combining heart rate and acceleration data, a comprehensive load index of physiological and biomechanical load is calculated, accurately quantifying the cumulative training load and providing an objective and dynamic quantitative basis for judging training intensity. Finally, it can also dynamically adjust the athlete's training intensity for personalized and scientific training. That is, by using statistical measures of historical training data, including mean and standard deviation, personalized load intervals are set, and by comparing the current comprehensive load index in real time, the system automatically decides whether to maintain, reduce, or increase the athlete's training intensity, thus achieving dynamic optimization of training intensity. At the same time, it effectively balances training stimulation and recovery, avoiding the risk of injury caused by overtraining and preventing undertraining from affecting performance improvement, ultimately helping to improve the athlete's training efficiency and athletic performance, and promoting the development of sports training towards intelligence and personalization.

[0045] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0046] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A smart management system for sports training, characterized in that, include: The multimodal data acquisition module is used to collect athletes' physiological data, kinematic data, and subjective feedback data in real time during training. After data preprocessing, multimodal data is generated. The data fusion and state assessment module is used to receive multimodal data, analyze the fatigue of athletes during training, and combine the athletes' current heart rate and acceleration data to perform cumulative load analysis and calculate the comprehensive load index of athletes in training state. The dynamic adjustment module receives the comprehensive load index, combines it with the athlete's historical training data to obtain the statistical value of the historical comprehensive load index, compares the statistical value with the athlete's comprehensive load index at the current training moment, and outputs control commands to the training equipment in real time based on the comparison results, so as to realize the dynamic adjustment of the athlete's training intensity.

2. The intelligent management system for sports training according to claim 1, characterized in that, The multimodal data acquisition module includes a physiological data acquisition module, a kinematic data acquisition module, a subjective feedback data acquisition module, and a data preprocessing module; The physiological data acquisition module is used to collect physiological data of athletes during training, including heart rate, heart rate variability, and surface electromyography signals. The kinematic data acquisition module is used to collect kinematic data of athletes during training, including movement speed and acceleration. The subjective feedback data acquisition module is used to collect athletes' subjective feedback data after training, including fatigue level scores. The data preprocessing module receives physiological data, kinematic data, and subjective feedback data, performs data preprocessing, and generates multimodal data, specifically: The heart rate, heart rate variability, and fatigue score were normalized to generate normalized heart rate, normalized heart rate variability, and normalized fatigue score, i.e.: in, Indicates that the athlete is during training. Heart rate below This represents an athlete's individual resting heart rate. Indicates that the athlete is during training. Normalized heart rate Indicates that the athlete is during training. Heart rate variability under the following conditions This represents the individual baseline heart rate variability of an athlete, and is the mean of the heart rate variability. Indicates that the athlete is during training. Normalized heart rate variability under normalization, Indicates that the athlete is during training. The fatigue level rating below, This represents the maximum value on the athlete's individual fatigue level score scale. Indicates that the athlete is during training. Normalized fatigue rate; Obtain the time window for athletes during training sessions. The surface electromyography (EMG) signals of the target muscle groups were analyzed, and their variance was calculated to obtain the surface EMG signals of the athlete's target muscle groups within a time window. variance within .

3. The intelligent management system for sports training according to claim 2, characterized in that, The data fusion and state assessment module receives multimodal data, analyzes the athlete's fatigue level during training, and combines this with the athlete's current heart rate and acceleration data to perform cumulative load analysis and calculate the athlete's comprehensive load index during training. Specifically: Based on the athlete's normalized heart rate, normalized heart rate variability, normalized fatigue level, and surface electromyography signals, a comprehensive fatigue analysis was performed considering cardiovascular, muscular, and subjective feedback dimensions. Through weighted allocation, a real-time fatigue index was calculated, namely: in, Indicates that the athlete is during training. Real-time fatigue index under the current conditions For cardiovascular dimensions indicators, This indicates the electromyographic signals on the surface of the athlete's target muscle group within a time window. The variance within the range is an indicator of muscle dimension. For subjective feedback dimension indicators, , , These represent the weighting coefficients for the cardiovascular dimension, muscle dimension, and subjective feedback dimension, respectively. Based on the real-time fatigue index and the athlete's heart rate during training, the accumulated physiological load of the athlete during training is calculated. Based on the athlete's acceleration during training, the accumulated biomechanical load of the athlete during training is calculated. Weights are allocated to the physiological load and biomechanical load to calculate the athlete's comprehensive load index during training, i.e.: in, Indicates that the athlete is during training. The comprehensive load index under, , These represent the weighting coefficients for balancing physiological load and biomechanical load, respectively. Indicates that the athlete is during training. The physiological load under the condition, This indicates the start time of the athlete's training. Indicates that the athlete is during training. Real-time fatigue index under the current conditions Indicates that the athlete is during training. Heart rate below Indicates that the athlete is during training. Under biomechanical load, Indicates that the athlete is during training. The acceleration under the current.

4. The intelligent management system for sports training according to claim 1, characterized in that, The dynamic adjustment module receives the comprehensive load index, combines it with the athlete's historical training data to obtain a statistical measure of the historical comprehensive load index, compares this statistical measure with the athlete's comprehensive load index at the current training moment, and outputs control commands to the training equipment in real time based on the comparison results. This enables dynamic adjustment of the athlete's training intensity. Specifically: Based on athletes' historical training data, the statistics of the historical comprehensive load index are calculated, including the historical comprehensive load index mean and the historical comprehensive load index standard deviation. Perform the first judgment to determine when the athlete is training. The comprehensive load index below Is it within the standard range? If yes, then the athlete's current training load is within the normal fluctuation range of their individual ability, and the current training intensity should be maintained; otherwise, the second judgment should be executed. Perform the second judgment to determine when the athlete is training. The comprehensive load index below Is it greater than If so, the athlete's current training load is significantly higher than the upper limit of its normal fluctuation range, posing a risk of overtraining and injury. An overload command is sent to the training equipment to reduce the athlete's training intensity. Otherwise, the third judgment is executed. Perform the third judgment to determine when the athlete is training. The comprehensive load index below Is it less than If so, the athlete's current training load is below the lower limit of its normal fluctuation range, indicating that the athlete is in a recovery period or the training intensity is insufficient. A low-load instruction is sent to the training equipment to increase the athlete's training intensity.

5. The intelligent management system for sports training according to claim 4, characterized in that, The formulas for calculating the historical composite load index mean and the historical composite load index standard deviation are as follows: in, This represents the historical average comprehensive load index. Indicates the number of training sessions. Indicates the athlete in the The overall load index under this training session This represents the standard deviation of the historical composite load index.

6. A smart management method for sports training, characterized in that, Applied to the system as described in any one of claims 1-5, comprising the following steps: S1. Real-time collection of athletes' physiological data, kinematic data, and subjective feedback data during training, data preprocessing, and generation of multimodal data; S2. Based on multimodal data, analyze the fatigue of athletes during training, and combine the athletes' current heart rate and acceleration data to analyze the cumulative load and calculate the comprehensive load index of athletes in training. S3. Based on the comprehensive load index and combined with the athlete's historical training data, calculate the historical comprehensive load index mean and historical comprehensive load index standard deviation to set judgment criteria, determine whether the athlete should maintain, increase or decrease the current training intensity, and output control commands to the training equipment to realize dynamic adjustment of the athlete's training intensity.

Images